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": 33.28109618647526,
+      "learning_rate": 2.1321961620469085e-08,
+      "loss": 1.6853,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 43.91919769806288,
+      "learning_rate": 4.264392324093817e-08,
+      "loss": 1.7153,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 26.493951361893362,
+      "learning_rate": 6.396588486140725e-08,
+      "loss": 1.7954,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 43.92807300827429,
+      "learning_rate": 8.528784648187634e-08,
+      "loss": 1.8073,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 58.180183674726806,
+      "learning_rate": 1.0660980810234542e-07,
+      "loss": 1.7118,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 26.44801442680443,
+      "learning_rate": 1.279317697228145e-07,
+      "loss": 1.6998,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 38.797917765531864,
+      "learning_rate": 1.4925373134328358e-07,
+      "loss": 1.7311,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 30.32320938499203,
+      "learning_rate": 1.7057569296375268e-07,
+      "loss": 1.6486,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 28.49349847788429,
+      "learning_rate": 1.918976545842218e-07,
+      "loss": 1.6517,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 32.79107536538845,
+      "learning_rate": 2.1321961620469084e-07,
+      "loss": 1.8598,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 34.36028757063917,
+      "learning_rate": 2.3454157782515995e-07,
+      "loss": 1.7304,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 46.55815652480895,
+      "learning_rate": 2.55863539445629e-07,
+      "loss": 1.8716,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 44.215660342011304,
+      "learning_rate": 2.771855010660981e-07,
+      "loss": 1.674,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 58.171568200547135,
+      "learning_rate": 2.9850746268656716e-07,
+      "loss": 1.6648,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 35.84419817305837,
+      "learning_rate": 3.1982942430703626e-07,
+      "loss": 1.7075,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 24.095802889577822,
+      "learning_rate": 3.4115138592750537e-07,
+      "loss": 1.6886,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 53.69964021365882,
+      "learning_rate": 3.624733475479744e-07,
+      "loss": 1.6703,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 41.320239751317054,
+      "learning_rate": 3.837953091684436e-07,
+      "loss": 1.6843,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 24.64051043671449,
+      "learning_rate": 4.0511727078891263e-07,
+      "loss": 1.6917,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.67229676736006,
+      "learning_rate": 4.264392324093817e-07,
+      "loss": 0.9704,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 26.505636626413466,
+      "learning_rate": 4.4776119402985074e-07,
+      "loss": 1.5296,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 25.41847323132133,
+      "learning_rate": 4.690831556503199e-07,
+      "loss": 1.7048,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 40.05396139173693,
+      "learning_rate": 4.904051172707889e-07,
+      "loss": 1.6831,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 36.112945174577916,
+      "learning_rate": 5.11727078891258e-07,
+      "loss": 1.5345,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 78.26556048628788,
+      "learning_rate": 5.33049040511727e-07,
+      "loss": 1.7003,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 30.974730496776676,
+      "learning_rate": 5.543710021321962e-07,
+      "loss": 1.6566,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 32.341242650010926,
+      "learning_rate": 5.756929637526653e-07,
+      "loss": 1.6316,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 29.52790206866178,
+      "learning_rate": 5.970149253731343e-07,
+      "loss": 1.6454,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 121.52544518757914,
+      "learning_rate": 6.183368869936035e-07,
+      "loss": 1.5992,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 58.87909324421914,
+      "learning_rate": 6.396588486140725e-07,
+      "loss": 1.4659,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 22.390343102212537,
+      "learning_rate": 6.609808102345417e-07,
+      "loss": 1.5027,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 16.821348196406937,
+      "learning_rate": 6.823027718550107e-07,
+      "loss": 1.5023,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 30.912026285215315,
+      "learning_rate": 7.036247334754798e-07,
+      "loss": 1.514,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.710835670239849,
+      "learning_rate": 7.249466950959488e-07,
+      "loss": 0.8645,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 47.661061843744804,
+      "learning_rate": 7.462686567164179e-07,
+      "loss": 1.3893,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 43.77943528547016,
+      "learning_rate": 7.675906183368872e-07,
+      "loss": 1.617,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.073080820719719,
+      "learning_rate": 7.889125799573562e-07,
+      "loss": 0.9159,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 17.914303117647346,
+      "learning_rate": 8.102345415778253e-07,
+      "loss": 1.4791,
+      "step": 38
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 22.04582731348449,
+      "learning_rate": 8.315565031982943e-07,
+      "loss": 1.5458,
+      "step": 39
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 20.569336222007667,
+      "learning_rate": 8.528784648187634e-07,
+      "loss": 1.6503,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.335303173500554,
+      "learning_rate": 8.742004264392324e-07,
+      "loss": 1.3518,
+      "step": 41
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 12.285589073639093,
+      "learning_rate": 8.955223880597015e-07,
+      "loss": 1.3644,
+      "step": 42
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.204318710480241,
+      "learning_rate": 9.168443496801707e-07,
+      "loss": 0.7834,
+      "step": 43
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.514401861155571,
+      "learning_rate": 9.381663113006398e-07,
+      "loss": 1.3165,
+      "step": 44
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 29.35042626971408,
+      "learning_rate": 9.594882729211088e-07,
+      "loss": 1.1144,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 19.51813694801758,
+      "learning_rate": 9.808102345415779e-07,
+      "loss": 1.3104,
+      "step": 46
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.044837482042877,
+      "learning_rate": 1.002132196162047e-06,
+      "loss": 0.7352,
+      "step": 47
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.507555239994229,
+      "learning_rate": 1.023454157782516e-06,
+      "loss": 1.3496,
+      "step": 48
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 17.641291019045845,
+      "learning_rate": 1.044776119402985e-06,
+      "loss": 1.2677,
+      "step": 49
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.452194188386725,
+      "learning_rate": 1.066098081023454e-06,
+      "loss": 1.1843,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 18.745519779386633,
+      "learning_rate": 1.0874200426439234e-06,
+      "loss": 1.1364,
+      "step": 51
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.325018102946773,
+      "learning_rate": 1.1087420042643924e-06,
+      "loss": 1.2334,
+      "step": 52
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.060514126076129,
+      "learning_rate": 1.1300639658848615e-06,
+      "loss": 1.2729,
+      "step": 53
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.313532195845921,
+      "learning_rate": 1.1513859275053305e-06,
+      "loss": 1.2096,
+      "step": 54
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 20.73004170915379,
+      "learning_rate": 1.1727078891257996e-06,
+      "loss": 1.1782,
+      "step": 55
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 41.91159941228401,
+      "learning_rate": 1.1940298507462686e-06,
+      "loss": 1.156,
+      "step": 56
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.50204818593784,
+      "learning_rate": 1.2153518123667379e-06,
+      "loss": 1.1062,
+      "step": 57
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.9269541823559155,
+      "learning_rate": 1.236673773987207e-06,
+      "loss": 1.3527,
+      "step": 58
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.035524359349103,
+      "learning_rate": 1.257995735607676e-06,
+      "loss": 1.3356,
+      "step": 59
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 18.089053033456228,
+      "learning_rate": 1.279317697228145e-06,
+      "loss": 1.016,
+      "step": 60
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.977577583387989,
+      "learning_rate": 1.300639658848614e-06,
+      "loss": 1.1849,
+      "step": 61
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.718300044001928,
+      "learning_rate": 1.3219616204690834e-06,
+      "loss": 1.0936,
+      "step": 62
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.71932622730894,
+      "learning_rate": 1.3432835820895524e-06,
+      "loss": 0.9974,
+      "step": 63
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.810181640431733,
+      "learning_rate": 1.3646055437100215e-06,
+      "loss": 1.1842,
+      "step": 64
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.887985228748798,
+      "learning_rate": 1.3859275053304905e-06,
+      "loss": 1.0962,
+      "step": 65
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.503072022752859,
+      "learning_rate": 1.4072494669509596e-06,
+      "loss": 0.9473,
+      "step": 66
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.774473669301928,
+      "learning_rate": 1.4285714285714286e-06,
+      "loss": 0.9622,
+      "step": 67
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.085713476021658,
+      "learning_rate": 1.4498933901918977e-06,
+      "loss": 1.0176,
+      "step": 68
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.300411457207263,
+      "learning_rate": 1.4712153518123667e-06,
+      "loss": 0.9344,
+      "step": 69
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.061396778960698,
+      "learning_rate": 1.4925373134328358e-06,
+      "loss": 1.2402,
+      "step": 70
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 22.894902565860114,
+      "learning_rate": 1.5138592750533053e-06,
+      "loss": 1.0362,
+      "step": 71
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.762727633227999,
+      "learning_rate": 1.5351812366737743e-06,
+      "loss": 1.1305,
+      "step": 72
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 105.63944556072738,
+      "learning_rate": 1.5565031982942434e-06,
+      "loss": 1.0224,
+      "step": 73
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.126900664639504,
+      "learning_rate": 1.5778251599147124e-06,
+      "loss": 1.0977,
+      "step": 74
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.0642716588778063,
+      "learning_rate": 1.5991471215351815e-06,
+      "loss": 0.7756,
+      "step": 75
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.076125720054462,
+      "learning_rate": 1.6204690831556505e-06,
+      "loss": 1.2334,
+      "step": 76
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.049680777198999,
+      "learning_rate": 1.6417910447761196e-06,
+      "loss": 1.1486,
+      "step": 77
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.5008090522365065,
+      "learning_rate": 1.6631130063965886e-06,
+      "loss": 0.9857,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 21.851817338942503,
+      "learning_rate": 1.6844349680170577e-06,
+      "loss": 1.0045,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.831141047496596,
+      "learning_rate": 1.7057569296375267e-06,
+      "loss": 1.017,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.659410042207726,
+      "learning_rate": 1.7270788912579958e-06,
+      "loss": 0.7508,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.46281665477027,
+      "learning_rate": 1.7484008528784648e-06,
+      "loss": 0.9435,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3756606768285917,
+      "learning_rate": 1.7697228144989339e-06,
+      "loss": 0.6065,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 32.83071500967826,
+      "learning_rate": 1.791044776119403e-06,
+      "loss": 1.1042,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 17.06280239391578,
+      "learning_rate": 1.812366737739872e-06,
+      "loss": 0.9341,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.9236229662710995,
+      "learning_rate": 1.8336886993603415e-06,
+      "loss": 1.0786,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.387614428225236,
+      "learning_rate": 1.8550106609808105e-06,
+      "loss": 1.0608,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.487695640957417,
+      "learning_rate": 1.8763326226012796e-06,
+      "loss": 1.1262,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.160797226384473,
+      "learning_rate": 1.8976545842217486e-06,
+      "loss": 1.0698,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.582951632818221,
+      "learning_rate": 1.9189765458422177e-06,
+      "loss": 1.205,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.022060565866631,
+      "learning_rate": 1.9402985074626867e-06,
+      "loss": 1.1028,
+      "step": 91
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.5902223098010975,
+      "learning_rate": 1.9616204690831558e-06,
+      "loss": 0.9946,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 14.26513626880042,
+      "learning_rate": 1.982942430703625e-06,
+      "loss": 0.9767,
+      "step": 93
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.109841157023887,
+      "learning_rate": 2.004264392324094e-06,
+      "loss": 0.9839,
+      "step": 94
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.992306241315927,
+      "learning_rate": 2.025586353944563e-06,
+      "loss": 1.0251,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.425795802744814,
+      "learning_rate": 2.046908315565032e-06,
+      "loss": 0.91,
+      "step": 96
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.830748774046691,
+      "learning_rate": 2.068230277185501e-06,
+      "loss": 1.0416,
+      "step": 97
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 13.842578707980532,
+      "learning_rate": 2.08955223880597e-06,
+      "loss": 0.9737,
+      "step": 98
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.170106890328179,
+      "learning_rate": 2.110874200426439e-06,
+      "loss": 1.0923,
+      "step": 99
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1654057857482294,
+      "learning_rate": 2.132196162046908e-06,
+      "loss": 1.0553,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.845749324713476,
+      "learning_rate": 2.1535181236673773e-06,
+      "loss": 0.9814,
+      "step": 101
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.118400092779808,
+      "learning_rate": 2.1748400852878467e-06,
+      "loss": 1.1737,
+      "step": 102
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.6573065317514395,
+      "learning_rate": 2.1961620469083158e-06,
+      "loss": 0.9337,
+      "step": 103
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5190986153661197,
+      "learning_rate": 2.217484008528785e-06,
+      "loss": 0.71,
+      "step": 104
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7184842474902253,
+      "learning_rate": 2.238805970149254e-06,
+      "loss": 0.9609,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 16.455965680055677,
+      "learning_rate": 2.260127931769723e-06,
+      "loss": 1.1558,
+      "step": 106
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.0291060221570767,
+      "learning_rate": 2.281449893390192e-06,
+      "loss": 1.02,
+      "step": 107
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.461925330336962,
+      "learning_rate": 2.302771855010661e-06,
+      "loss": 0.9642,
+      "step": 108
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.847184657523936,
+      "learning_rate": 2.32409381663113e-06,
+      "loss": 1.0422,
+      "step": 109
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 16.334579194092356,
+      "learning_rate": 2.345415778251599e-06,
+      "loss": 0.8588,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2991492931604323,
+      "learning_rate": 2.366737739872068e-06,
+      "loss": 0.7227,
+      "step": 111
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.243999455831442,
+      "learning_rate": 2.3880597014925373e-06,
+      "loss": 1.0679,
+      "step": 112
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.524780125498815,
+      "learning_rate": 2.4093816631130067e-06,
+      "loss": 0.9137,
+      "step": 113
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.121840166788039,
+      "learning_rate": 2.4307036247334758e-06,
+      "loss": 1.1445,
+      "step": 114
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5944945010771112,
+      "learning_rate": 2.452025586353945e-06,
+      "loss": 0.7606,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.864451524636321,
+      "learning_rate": 2.473347547974414e-06,
+      "loss": 0.8845,
+      "step": 116
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.15555543937216,
+      "learning_rate": 2.494669509594883e-06,
+      "loss": 0.9833,
+      "step": 117
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8983264453809827,
+      "learning_rate": 2.515991471215352e-06,
+      "loss": 0.9713,
+      "step": 118
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.238615953982402,
+      "learning_rate": 2.537313432835821e-06,
+      "loss": 0.9636,
+      "step": 119
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.300137305865406,
+      "learning_rate": 2.55863539445629e-06,
+      "loss": 0.9997,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.475837603827229,
+      "learning_rate": 2.579957356076759e-06,
+      "loss": 1.0213,
+      "step": 121
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.698070038052698,
+      "learning_rate": 2.601279317697228e-06,
+      "loss": 1.0057,
+      "step": 122
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7661361158857336,
+      "learning_rate": 2.6226012793176977e-06,
+      "loss": 0.9957,
+      "step": 123
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.544926309588696,
+      "learning_rate": 2.6439232409381667e-06,
+      "loss": 0.9575,
+      "step": 124
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2771517554345198,
+      "learning_rate": 2.6652452025586358e-06,
+      "loss": 0.9118,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9337710806652064,
+      "learning_rate": 2.686567164179105e-06,
+      "loss": 0.6142,
+      "step": 126
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.03716310658932,
+      "learning_rate": 2.707889125799574e-06,
+      "loss": 0.6895,
+      "step": 127
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.806268591173616,
+      "learning_rate": 2.729211087420043e-06,
+      "loss": 0.9687,
+      "step": 128
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.418631706454691,
+      "learning_rate": 2.750533049040512e-06,
+      "loss": 0.9688,
+      "step": 129
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8226156888365255,
+      "learning_rate": 2.771855010660981e-06,
+      "loss": 0.9298,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.221359307740677,
+      "learning_rate": 2.79317697228145e-06,
+      "loss": 0.9129,
+      "step": 131
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.305448060065418,
+      "learning_rate": 2.814498933901919e-06,
+      "loss": 0.9641,
+      "step": 132
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.812669317723893,
+      "learning_rate": 2.835820895522388e-06,
+      "loss": 0.9453,
+      "step": 133
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0157888658069583,
+      "learning_rate": 2.8571428571428573e-06,
+      "loss": 1.0847,
+      "step": 134
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 14.619911311813524,
+      "learning_rate": 2.8784648187633263e-06,
+      "loss": 0.9711,
+      "step": 135
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.0013286684831595,
+      "learning_rate": 2.8997867803837954e-06,
+      "loss": 1.0344,
+      "step": 136
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.676266412748936,
+      "learning_rate": 2.9211087420042644e-06,
+      "loss": 0.9177,
+      "step": 137
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3282768586075457,
+      "learning_rate": 2.9424307036247335e-06,
+      "loss": 0.8637,
+      "step": 138
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.694430710853514,
+      "learning_rate": 2.9637526652452025e-06,
+      "loss": 0.9844,
+      "step": 139
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 13.199916441783287,
+      "learning_rate": 2.9850746268656716e-06,
+      "loss": 0.8897,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.36364980848261,
+      "learning_rate": 3.006396588486141e-06,
+      "loss": 1.0678,
+      "step": 141
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.325047118752239,
+      "learning_rate": 3.0277185501066105e-06,
+      "loss": 1.0294,
+      "step": 142
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4907221622981632,
+      "learning_rate": 3.0490405117270796e-06,
+      "loss": 0.6422,
+      "step": 143
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 18.75523783245584,
+      "learning_rate": 3.0703624733475486e-06,
+      "loss": 0.994,
+      "step": 144
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.489988387072182,
+      "learning_rate": 3.0916844349680177e-06,
+      "loss": 0.8042,
+      "step": 145
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.037320022217296,
+      "learning_rate": 3.1130063965884867e-06,
+      "loss": 1.0591,
+      "step": 146
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4841989188819533,
+      "learning_rate": 3.1343283582089558e-06,
+      "loss": 0.95,
+      "step": 147
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 13.21398264863175,
+      "learning_rate": 3.155650319829425e-06,
+      "loss": 0.9383,
+      "step": 148
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.3097331411851245,
+      "learning_rate": 3.176972281449894e-06,
+      "loss": 1.0406,
+      "step": 149
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4881261907101067,
+      "learning_rate": 3.198294243070363e-06,
+      "loss": 0.8697,
+      "step": 150
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3565115674249983,
+      "learning_rate": 3.219616204690832e-06,
+      "loss": 0.5717,
+      "step": 151
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.554089660026784,
+      "learning_rate": 3.240938166311301e-06,
+      "loss": 0.993,
+      "step": 152
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.6217722558904852,
+      "learning_rate": 3.26226012793177e-06,
+      "loss": 0.7775,
+      "step": 153
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.311960969437155,
+      "learning_rate": 3.283582089552239e-06,
+      "loss": 0.9188,
+      "step": 154
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5152769664699304,
+      "learning_rate": 3.304904051172708e-06,
+      "loss": 0.9481,
+      "step": 155
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 20.48795902812986,
+      "learning_rate": 3.3262260127931773e-06,
+      "loss": 0.8538,
+      "step": 156
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7258139132709713,
+      "learning_rate": 3.3475479744136463e-06,
+      "loss": 1.2716,
+      "step": 157
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.785478785678323,
+      "learning_rate": 3.3688699360341154e-06,
+      "loss": 0.9657,
+      "step": 158
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1549569855655704,
+      "learning_rate": 3.3901918976545844e-06,
+      "loss": 0.9599,
+      "step": 159
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 33.806673557254236,
+      "learning_rate": 3.4115138592750535e-06,
+      "loss": 0.9413,
+      "step": 160
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8500678518202247,
+      "learning_rate": 3.4328358208955225e-06,
+      "loss": 0.8829,
+      "step": 161
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.123757889779569,
+      "learning_rate": 3.4541577825159916e-06,
+      "loss": 0.9855,
+      "step": 162
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.063840972092507,
+      "learning_rate": 3.4754797441364606e-06,
+      "loss": 0.9585,
+      "step": 163
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.430035930192975,
+      "learning_rate": 3.4968017057569297e-06,
+      "loss": 0.814,
+      "step": 164
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.431186991683928,
+      "learning_rate": 3.5181236673773987e-06,
+      "loss": 0.9032,
+      "step": 165
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.617681076559489,
+      "learning_rate": 3.5394456289978678e-06,
+      "loss": 0.9051,
+      "step": 166
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.059125628629912,
+      "learning_rate": 3.560767590618337e-06,
+      "loss": 1.0229,
+      "step": 167
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.352799322728352,
+      "learning_rate": 3.582089552238806e-06,
+      "loss": 0.9071,
+      "step": 168
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.747547246062081,
+      "learning_rate": 3.603411513859275e-06,
+      "loss": 0.9204,
+      "step": 169
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3553005834996004,
+      "learning_rate": 3.624733475479744e-06,
+      "loss": 0.9097,
+      "step": 170
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.564693041544327,
+      "learning_rate": 3.6460554371002135e-06,
+      "loss": 1.1482,
+      "step": 171
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.9407359792550327,
+      "learning_rate": 3.667377398720683e-06,
+      "loss": 0.9342,
+      "step": 172
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.916357278283614,
+      "learning_rate": 3.688699360341152e-06,
+      "loss": 0.8994,
+      "step": 173
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.656146297275596,
+      "learning_rate": 3.710021321961621e-06,
+      "loss": 1.1006,
+      "step": 174
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.4440874159806425,
+      "learning_rate": 3.73134328358209e-06,
+      "loss": 0.918,
+      "step": 175
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.64765945372912,
+      "learning_rate": 3.752665245202559e-06,
+      "loss": 0.835,
+      "step": 176
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.730042712108293,
+      "learning_rate": 3.773987206823028e-06,
+      "loss": 0.889,
+      "step": 177
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.995650978232185,
+      "learning_rate": 3.7953091684434973e-06,
+      "loss": 0.8872,
+      "step": 178
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.275097731261329,
+      "learning_rate": 3.816631130063966e-06,
+      "loss": 0.8636,
+      "step": 179
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.612993660016237,
+      "learning_rate": 3.837953091684435e-06,
+      "loss": 0.8616,
+      "step": 180
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4184141817815052,
+      "learning_rate": 3.859275053304904e-06,
+      "loss": 0.8963,
+      "step": 181
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0452142228166075,
+      "learning_rate": 3.8805970149253735e-06,
+      "loss": 0.972,
+      "step": 182
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.045445477246414,
+      "learning_rate": 3.9019189765458425e-06,
+      "loss": 0.923,
+      "step": 183
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.052934472857606,
+      "learning_rate": 3.9232409381663116e-06,
+      "loss": 0.909,
+      "step": 184
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.267536892597484,
+      "learning_rate": 3.944562899786781e-06,
+      "loss": 0.8609,
+      "step": 185
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.515742061413083,
+      "learning_rate": 3.96588486140725e-06,
+      "loss": 0.936,
+      "step": 186
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4517820484549033,
+      "learning_rate": 3.987206823027719e-06,
+      "loss": 0.8401,
+      "step": 187
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7693547501328712,
+      "learning_rate": 4.008528784648188e-06,
+      "loss": 0.9571,
+      "step": 188
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.129768289954375,
+      "learning_rate": 4.029850746268657e-06,
+      "loss": 0.8356,
+      "step": 189
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.066862609655172,
+      "learning_rate": 4.051172707889126e-06,
+      "loss": 0.8844,
+      "step": 190
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2513715213072805,
+      "learning_rate": 4.072494669509595e-06,
+      "loss": 0.6412,
+      "step": 191
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.290706660788013,
+      "learning_rate": 4.093816631130064e-06,
+      "loss": 0.8773,
+      "step": 192
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4085165248876925,
+      "learning_rate": 4.115138592750533e-06,
+      "loss": 0.8908,
+      "step": 193
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.545288321922942,
+      "learning_rate": 4.136460554371002e-06,
+      "loss": 0.999,
+      "step": 194
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1171864721084317,
+      "learning_rate": 4.157782515991471e-06,
+      "loss": 0.7962,
+      "step": 195
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4961924288934565,
+      "learning_rate": 4.17910447761194e-06,
+      "loss": 0.9436,
+      "step": 196
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.529223882794142,
+      "learning_rate": 4.200426439232409e-06,
+      "loss": 0.8447,
+      "step": 197
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.6651537980277205,
+      "learning_rate": 4.221748400852878e-06,
+      "loss": 0.8326,
+      "step": 198
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.04110352524989,
+      "learning_rate": 4.243070362473347e-06,
+      "loss": 0.8798,
+      "step": 199
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.952009550602607,
+      "learning_rate": 4.264392324093816e-06,
+      "loss": 1.001,
+      "step": 200
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.782712581691702,
+      "learning_rate": 4.2857142857142855e-06,
+      "loss": 0.9274,
+      "step": 201
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.664061904823161,
+      "learning_rate": 4.3070362473347545e-06,
+      "loss": 0.8381,
+      "step": 202
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.046437738062339,
+      "learning_rate": 4.3283582089552236e-06,
+      "loss": 0.8648,
+      "step": 203
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.855328609925418,
+      "learning_rate": 4.3496801705756935e-06,
+      "loss": 0.8942,
+      "step": 204
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.374922264570846,
+      "learning_rate": 4.3710021321961625e-06,
+      "loss": 0.9647,
+      "step": 205
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.944863829560073,
+      "learning_rate": 4.3923240938166316e-06,
+      "loss": 1.0317,
+      "step": 206
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4842297780995455,
+      "learning_rate": 4.413646055437101e-06,
+      "loss": 0.6611,
+      "step": 207
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4313353207420145,
+      "learning_rate": 4.43496801705757e-06,
+      "loss": 1.0392,
+      "step": 208
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.042011683751464,
+      "learning_rate": 4.456289978678039e-06,
+      "loss": 0.7003,
+      "step": 209
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2660982495812614,
+      "learning_rate": 4.477611940298508e-06,
+      "loss": 0.9729,
+      "step": 210
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.706153593860751,
+      "learning_rate": 4.498933901918977e-06,
+      "loss": 0.8499,
+      "step": 211
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9286962422594953,
+      "learning_rate": 4.520255863539446e-06,
+      "loss": 0.8982,
+      "step": 212
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.88638770889486,
+      "learning_rate": 4.541577825159915e-06,
+      "loss": 0.9703,
+      "step": 213
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.956724929118262,
+      "learning_rate": 4.562899786780384e-06,
+      "loss": 0.8855,
+      "step": 214
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2363793916309191,
+      "learning_rate": 4.584221748400853e-06,
+      "loss": 0.6638,
+      "step": 215
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.641341971585338,
+      "learning_rate": 4.605543710021322e-06,
+      "loss": 0.9343,
+      "step": 216
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8516073755516718,
+      "learning_rate": 4.626865671641791e-06,
+      "loss": 0.9428,
+      "step": 217
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.242545242421389,
+      "learning_rate": 4.64818763326226e-06,
+      "loss": 0.7724,
+      "step": 218
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7928480258442065,
+      "learning_rate": 4.669509594882729e-06,
+      "loss": 0.8646,
+      "step": 219
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5133702324918845,
+      "learning_rate": 4.690831556503198e-06,
+      "loss": 0.9245,
+      "step": 220
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.016135805850737,
+      "learning_rate": 4.712153518123667e-06,
+      "loss": 0.9937,
+      "step": 221
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7898541058915605,
+      "learning_rate": 4.733475479744136e-06,
+      "loss": 0.9412,
+      "step": 222
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.461020197017766,
+      "learning_rate": 4.7547974413646055e-06,
+      "loss": 0.8462,
+      "step": 223
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.109725783623522,
+      "learning_rate": 4.7761194029850745e-06,
+      "loss": 0.8916,
+      "step": 224
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.720639790315238,
+      "learning_rate": 4.797441364605544e-06,
+      "loss": 0.8646,
+      "step": 225
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.036879250535004,
+      "learning_rate": 4.8187633262260135e-06,
+      "loss": 0.8672,
+      "step": 226
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.227794899121631,
+      "learning_rate": 4.8400852878464825e-06,
+      "loss": 0.9027,
+      "step": 227
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.3916936916827884,
+      "learning_rate": 4.8614072494669516e-06,
+      "loss": 0.9507,
+      "step": 228
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4372675790453313,
+      "learning_rate": 4.882729211087421e-06,
+      "loss": 0.726,
+      "step": 229
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6047383727212114,
+      "learning_rate": 4.90405117270789e-06,
+      "loss": 0.8921,
+      "step": 230
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6869487756566772,
+      "learning_rate": 4.925373134328359e-06,
+      "loss": 0.7894,
+      "step": 231
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.9682289097752417,
+      "learning_rate": 4.946695095948828e-06,
+      "loss": 0.9254,
+      "step": 232
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3018081117851836,
+      "learning_rate": 4.968017057569297e-06,
+      "loss": 0.6537,
+      "step": 233
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.547336618748181,
+      "learning_rate": 4.989339019189766e-06,
+      "loss": 1.2988,
+      "step": 234
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.698320849234,
+      "learning_rate": 5.010660980810235e-06,
+      "loss": 0.8734,
+      "step": 235
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.97302445547019,
+      "learning_rate": 5.031982942430704e-06,
+      "loss": 0.9705,
+      "step": 236
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.936263811692734,
+      "learning_rate": 5.053304904051173e-06,
+      "loss": 0.8931,
+      "step": 237
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.774599842877439,
+      "learning_rate": 5.074626865671642e-06,
+      "loss": 1.0795,
+      "step": 238
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.889146339326743,
+      "learning_rate": 5.095948827292111e-06,
+      "loss": 0.8782,
+      "step": 239
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.424275969077103,
+      "learning_rate": 5.11727078891258e-06,
+      "loss": 1.0022,
+      "step": 240
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.215445191304509,
+      "learning_rate": 5.138592750533049e-06,
+      "loss": 0.7835,
+      "step": 241
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.642347299484213,
+      "learning_rate": 5.159914712153518e-06,
+      "loss": 0.9534,
+      "step": 242
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.6302375488920724,
+      "learning_rate": 5.181236673773987e-06,
+      "loss": 0.9852,
+      "step": 243
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.135461449290248,
+      "learning_rate": 5.202558635394456e-06,
+      "loss": 0.8947,
+      "step": 244
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4047069838306774,
+      "learning_rate": 5.2238805970149255e-06,
+      "loss": 0.8689,
+      "step": 245
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.030754089294014,
+      "learning_rate": 5.245202558635395e-06,
+      "loss": 0.8237,
+      "step": 246
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.149151726154507,
+      "learning_rate": 5.2665245202558636e-06,
+      "loss": 0.8962,
+      "step": 247
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1119460674647184,
+      "learning_rate": 5.2878464818763335e-06,
+      "loss": 0.895,
+      "step": 248
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4494396724703513,
+      "learning_rate": 5.309168443496802e-06,
+      "loss": 0.9663,
+      "step": 249
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.549365836796633,
+      "learning_rate": 5.3304904051172716e-06,
+      "loss": 0.9582,
+      "step": 250
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1331634763010077,
+      "learning_rate": 5.351812366737741e-06,
+      "loss": 0.7914,
+      "step": 251
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.9042477429028826,
+      "learning_rate": 5.37313432835821e-06,
+      "loss": 0.8767,
+      "step": 252
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9404179654650973,
+      "learning_rate": 5.394456289978679e-06,
+      "loss": 0.8914,
+      "step": 253
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.427084017925242,
+      "learning_rate": 5.415778251599148e-06,
+      "loss": 0.9263,
+      "step": 254
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5950757853291981,
+      "learning_rate": 5.437100213219617e-06,
+      "loss": 0.8237,
+      "step": 255
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6302790123084876,
+      "learning_rate": 5.458422174840086e-06,
+      "loss": 0.8676,
+      "step": 256
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.261900447288023,
+      "learning_rate": 5.479744136460555e-06,
+      "loss": 0.9144,
+      "step": 257
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.82746148703367,
+      "learning_rate": 5.501066098081024e-06,
+      "loss": 0.9367,
+      "step": 258
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.698071680146352,
+      "learning_rate": 5.522388059701493e-06,
+      "loss": 0.932,
+      "step": 259
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6292152653169052,
+      "learning_rate": 5.543710021321962e-06,
+      "loss": 0.851,
+      "step": 260
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5810393153605595,
+      "learning_rate": 5.565031982942431e-06,
+      "loss": 0.924,
+      "step": 261
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9040652851347637,
+      "learning_rate": 5.5863539445629e-06,
+      "loss": 0.7535,
+      "step": 262
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7467264946821004,
+      "learning_rate": 5.607675906183369e-06,
+      "loss": 0.8923,
+      "step": 263
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.45603540126598,
+      "learning_rate": 5.628997867803838e-06,
+      "loss": 0.877,
+      "step": 264
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.114425973218077,
+      "learning_rate": 5.650319829424308e-06,
+      "loss": 0.7943,
+      "step": 265
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9256413284309144,
+      "learning_rate": 5.671641791044776e-06,
+      "loss": 0.8731,
+      "step": 266
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6755989158754974,
+      "learning_rate": 5.692963752665246e-06,
+      "loss": 0.7802,
+      "step": 267
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7251486637653968,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": 0.9381,
+      "step": 268
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7087512230905726,
+      "learning_rate": 5.735607675906184e-06,
+      "loss": 0.8462,
+      "step": 269
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5058439306218099,
+      "learning_rate": 5.756929637526653e-06,
+      "loss": 0.7829,
+      "step": 270
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0538826821665683,
+      "learning_rate": 5.7782515991471225e-06,
+      "loss": 0.907,
+      "step": 271
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9104991670731144,
+      "learning_rate": 5.799573560767591e-06,
+      "loss": 0.853,
+      "step": 272
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8466670585698823,
+      "learning_rate": 5.820895522388061e-06,
+      "loss": 0.8258,
+      "step": 273
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.127780229336552,
+      "learning_rate": 5.842217484008529e-06,
+      "loss": 0.8855,
+      "step": 274
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3554831744291005,
+      "learning_rate": 5.863539445628999e-06,
+      "loss": 0.9649,
+      "step": 275
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5639229742213794,
+      "learning_rate": 5.884861407249467e-06,
+      "loss": 0.9024,
+      "step": 276
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3644502136290888,
+      "learning_rate": 5.906183368869937e-06,
+      "loss": 0.8515,
+      "step": 277
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.604855429260883,
+      "learning_rate": 5.927505330490405e-06,
+      "loss": 0.8132,
+      "step": 278
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0207704283253114,
+      "learning_rate": 5.948827292110875e-06,
+      "loss": 0.8851,
+      "step": 279
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.837925012830386,
+      "learning_rate": 5.970149253731343e-06,
+      "loss": 0.8836,
+      "step": 280
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5823605827341223,
+      "learning_rate": 5.991471215351813e-06,
+      "loss": 0.8029,
+      "step": 281
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.650169324013565,
+      "learning_rate": 6.012793176972282e-06,
+      "loss": 0.7742,
+      "step": 282
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8601975100826227,
+      "learning_rate": 6.034115138592751e-06,
+      "loss": 0.8027,
+      "step": 283
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5640175447867397,
+      "learning_rate": 6.055437100213221e-06,
+      "loss": 0.8612,
+      "step": 284
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7362421737323575,
+      "learning_rate": 6.076759061833689e-06,
+      "loss": 0.9553,
+      "step": 285
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8524799432838135,
+      "learning_rate": 6.098081023454159e-06,
+      "loss": 0.9207,
+      "step": 286
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6446778767565309,
+      "learning_rate": 6.119402985074627e-06,
+      "loss": 0.8557,
+      "step": 287
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.9879020939411802,
+      "learning_rate": 6.140724946695097e-06,
+      "loss": 0.863,
+      "step": 288
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5819026194189083,
+      "learning_rate": 6.1620469083155655e-06,
+      "loss": 0.7106,
+      "step": 289
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7513932200445823,
+      "learning_rate": 6.183368869936035e-06,
+      "loss": 0.872,
+      "step": 290
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4335608982003227,
+      "learning_rate": 6.2046908315565036e-06,
+      "loss": 0.7937,
+      "step": 291
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.404518267181221,
+      "learning_rate": 6.2260127931769735e-06,
+      "loss": 0.8005,
+      "step": 292
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3843994446959893,
+      "learning_rate": 6.247334754797442e-06,
+      "loss": 0.8029,
+      "step": 293
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.593855391605365,
+      "learning_rate": 6.2686567164179116e-06,
+      "loss": 0.9137,
+      "step": 294
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5405891875137774,
+      "learning_rate": 6.28997867803838e-06,
+      "loss": 0.8349,
+      "step": 295
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4119971101729367,
+      "learning_rate": 6.31130063965885e-06,
+      "loss": 0.8999,
+      "step": 296
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6980223951537599,
+      "learning_rate": 6.332622601279318e-06,
+      "loss": 0.7455,
+      "step": 297
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5246567146308294,
+      "learning_rate": 6.353944562899788e-06,
+      "loss": 0.8397,
+      "step": 298
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4768564970886024,
+      "learning_rate": 6.375266524520256e-06,
+      "loss": 0.6593,
+      "step": 299
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7312472391038634,
+      "learning_rate": 6.396588486140726e-06,
+      "loss": 0.8332,
+      "step": 300
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2098484531938167,
+      "learning_rate": 6.417910447761194e-06,
+      "loss": 0.8891,
+      "step": 301
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7504990502521154,
+      "learning_rate": 6.439232409381664e-06,
+      "loss": 0.8441,
+      "step": 302
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1670659434279793,
+      "learning_rate": 6.460554371002132e-06,
+      "loss": 1.0605,
+      "step": 303
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.125840741462889,
+      "learning_rate": 6.481876332622602e-06,
+      "loss": 0.8816,
+      "step": 304
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5401127563465664,
+      "learning_rate": 6.50319829424307e-06,
+      "loss": 0.8159,
+      "step": 305
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5643804773780103,
+      "learning_rate": 6.52452025586354e-06,
+      "loss": 0.9299,
+      "step": 306
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8634757257702286,
+      "learning_rate": 6.545842217484008e-06,
+      "loss": 0.842,
+      "step": 307
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8339566718158282,
+      "learning_rate": 6.567164179104478e-06,
+      "loss": 0.8246,
+      "step": 308
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6894496703095183,
+      "learning_rate": 6.5884861407249465e-06,
+      "loss": 0.7945,
+      "step": 309
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.664407105258195,
+      "learning_rate": 6.609808102345416e-06,
+      "loss": 0.9428,
+      "step": 310
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4852768908131249,
+      "learning_rate": 6.631130063965885e-06,
+      "loss": 0.8121,
+      "step": 311
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5553362284054018,
+      "learning_rate": 6.6524520255863545e-06,
+      "loss": 0.8599,
+      "step": 312
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6780833512400715,
+      "learning_rate": 6.673773987206824e-06,
+      "loss": 0.923,
+      "step": 313
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7124248725710796,
+      "learning_rate": 6.695095948827293e-06,
+      "loss": 0.8637,
+      "step": 314
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.671375679672624,
+      "learning_rate": 6.7164179104477625e-06,
+      "loss": 0.8923,
+      "step": 315
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5626097792646259,
+      "learning_rate": 6.737739872068231e-06,
+      "loss": 0.9052,
+      "step": 316
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6960109957030158,
+      "learning_rate": 6.759061833688701e-06,
+      "loss": 1.0029,
+      "step": 317
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.825024869765912,
+      "learning_rate": 6.780383795309169e-06,
+      "loss": 1.0025,
+      "step": 318
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6849359197203946,
+      "learning_rate": 6.801705756929639e-06,
+      "loss": 0.7686,
+      "step": 319
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.772919900704352,
+      "learning_rate": 6.823027718550107e-06,
+      "loss": 0.8337,
+      "step": 320
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.522374295365717,
+      "learning_rate": 6.844349680170577e-06,
+      "loss": 0.633,
+      "step": 321
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8658387561986682,
+      "learning_rate": 6.865671641791045e-06,
+      "loss": 0.6401,
+      "step": 322
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.625438736831122,
+      "learning_rate": 6.886993603411515e-06,
+      "loss": 0.7494,
+      "step": 323
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8433211002273284,
+      "learning_rate": 6.908315565031983e-06,
+      "loss": 0.8869,
+      "step": 324
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4853103278004465,
+      "learning_rate": 6.929637526652453e-06,
+      "loss": 1.027,
+      "step": 325
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8855719557342234,
+      "learning_rate": 6.950959488272921e-06,
+      "loss": 0.8404,
+      "step": 326
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4002449733501796,
+      "learning_rate": 6.972281449893391e-06,
+      "loss": 0.7027,
+      "step": 327
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5629855510118467,
+      "learning_rate": 6.993603411513859e-06,
+      "loss": 0.8903,
+      "step": 328
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8994182828693875,
+      "learning_rate": 7.014925373134329e-06,
+      "loss": 0.872,
+      "step": 329
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6226308101736715,
+      "learning_rate": 7.0362473347547975e-06,
+      "loss": 0.7942,
+      "step": 330
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3562889833043585,
+      "learning_rate": 7.057569296375267e-06,
+      "loss": 0.7553,
+      "step": 331
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.057313437337635,
+      "learning_rate": 7.0788912579957356e-06,
+      "loss": 0.7777,
+      "step": 332
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.63144288682161,
+      "learning_rate": 7.1002132196162055e-06,
+      "loss": 0.8235,
+      "step": 333
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.663765771133885,
+      "learning_rate": 7.121535181236674e-06,
+      "loss": 0.9124,
+      "step": 334
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.298957713961099,
+      "learning_rate": 7.1428571428571436e-06,
+      "loss": 0.6316,
+      "step": 335
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7812930618774812,
+      "learning_rate": 7.164179104477612e-06,
+      "loss": 0.7674,
+      "step": 336
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4625102343077834,
+      "learning_rate": 7.185501066098082e-06,
+      "loss": 0.8279,
+      "step": 337
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6887718984965672,
+      "learning_rate": 7.20682302771855e-06,
+      "loss": 0.9254,
+      "step": 338
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.536777030526388,
+      "learning_rate": 7.22814498933902e-06,
+      "loss": 1.0118,
+      "step": 339
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9578352561456323,
+      "learning_rate": 7.249466950959488e-06,
+      "loss": 0.8746,
+      "step": 340
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4601306317009732,
+      "learning_rate": 7.270788912579958e-06,
+      "loss": 0.8054,
+      "step": 341
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4177355797532556,
+      "learning_rate": 7.292110874200427e-06,
+      "loss": 0.8424,
+      "step": 342
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.590306654833927,
+      "learning_rate": 7.313432835820896e-06,
+      "loss": 0.7771,
+      "step": 343
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5411226459138594,
+      "learning_rate": 7.334754797441366e-06,
+      "loss": 0.6267,
+      "step": 344
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.406503484209324,
+      "learning_rate": 7.356076759061834e-06,
+      "loss": 0.9298,
+      "step": 345
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.710821769397449,
+      "learning_rate": 7.377398720682304e-06,
+      "loss": 0.8857,
+      "step": 346
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.764722890838849,
+      "learning_rate": 7.398720682302772e-06,
+      "loss": 0.8578,
+      "step": 347
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5982530819401932,
+      "learning_rate": 7.420042643923242e-06,
+      "loss": 0.8263,
+      "step": 348
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8356025362477826,
+      "learning_rate": 7.44136460554371e-06,
+      "loss": 0.8962,
+      "step": 349
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7127210367473786,
+      "learning_rate": 7.46268656716418e-06,
+      "loss": 0.8702,
+      "step": 350
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7540603955296479,
+      "learning_rate": 7.484008528784648e-06,
+      "loss": 1.0047,
+      "step": 351
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7277412182061842,
+      "learning_rate": 7.505330490405118e-06,
+      "loss": 0.9081,
+      "step": 352
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5128315512544093,
+      "learning_rate": 7.5266524520255865e-06,
+      "loss": 0.8291,
+      "step": 353
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.504747871065002,
+      "learning_rate": 7.547974413646056e-06,
+      "loss": 0.8049,
+      "step": 354
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.083763528274316,
+      "learning_rate": 7.569296375266525e-06,
+      "loss": 0.8554,
+      "step": 355
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8044329512919008,
+      "learning_rate": 7.5906183368869945e-06,
+      "loss": 0.954,
+      "step": 356
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7780348159419814,
+      "learning_rate": 7.611940298507463e-06,
+      "loss": 0.9087,
+      "step": 357
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3873914864248862,
+      "learning_rate": 7.633262260127933e-06,
+      "loss": 0.639,
+      "step": 358
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.811043380894575,
+      "learning_rate": 7.654584221748402e-06,
+      "loss": 0.903,
+      "step": 359
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6044826107232015,
+      "learning_rate": 7.67590618336887e-06,
+      "loss": 0.755,
+      "step": 360
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8934635586848603,
+      "learning_rate": 7.69722814498934e-06,
+      "loss": 0.8661,
+      "step": 361
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.397059192421763,
+      "learning_rate": 7.718550106609809e-06,
+      "loss": 0.7023,
+      "step": 362
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6032267335125872,
+      "learning_rate": 7.739872068230278e-06,
+      "loss": 0.7723,
+      "step": 363
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2884008360275507,
+      "learning_rate": 7.761194029850747e-06,
+      "loss": 0.7748,
+      "step": 364
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7120206587705817,
+      "learning_rate": 7.782515991471216e-06,
+      "loss": 0.8828,
+      "step": 365
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3748113570340001,
+      "learning_rate": 7.803837953091685e-06,
+      "loss": 0.7452,
+      "step": 366
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2299149761955874,
+      "learning_rate": 7.825159914712154e-06,
+      "loss": 1.0347,
+      "step": 367
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3670041140618245,
+      "learning_rate": 7.846481876332623e-06,
+      "loss": 0.7206,
+      "step": 368
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1428006925913308,
+      "learning_rate": 7.867803837953092e-06,
+      "loss": 0.8633,
+      "step": 369
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4408799799871943,
+      "learning_rate": 7.889125799573561e-06,
+      "loss": 0.7668,
+      "step": 370
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6140821733622184,
+      "learning_rate": 7.91044776119403e-06,
+      "loss": 0.8712,
+      "step": 371
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6782602490877185,
+      "learning_rate": 7.9317697228145e-06,
+      "loss": 0.9207,
+      "step": 372
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7452583046718573,
+      "learning_rate": 7.953091684434968e-06,
+      "loss": 0.8897,
+      "step": 373
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3827079929835304,
+      "learning_rate": 7.974413646055437e-06,
+      "loss": 0.7436,
+      "step": 374
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6309046532312728,
+      "learning_rate": 7.995735607675907e-06,
+      "loss": 0.8235,
+      "step": 375
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6581558275849468,
+      "learning_rate": 8.017057569296376e-06,
+      "loss": 0.8417,
+      "step": 376
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4360052102114114,
+      "learning_rate": 8.038379530916846e-06,
+      "loss": 0.8148,
+      "step": 377
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6990835964021551,
+      "learning_rate": 8.059701492537314e-06,
+      "loss": 0.8446,
+      "step": 378
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5388168526829336,
+      "learning_rate": 8.081023454157784e-06,
+      "loss": 0.8557,
+      "step": 379
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4678982828357883,
+      "learning_rate": 8.102345415778252e-06,
+      "loss": 0.7846,
+      "step": 380
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7110537143201716,
+      "learning_rate": 8.123667377398723e-06,
+      "loss": 0.8489,
+      "step": 381
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.611102256441078,
+      "learning_rate": 8.14498933901919e-06,
+      "loss": 0.7288,
+      "step": 382
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6355786576059468,
+      "learning_rate": 8.16631130063966e-06,
+      "loss": 0.8285,
+      "step": 383
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.64171364041531,
+      "learning_rate": 8.187633262260128e-06,
+      "loss": 0.8827,
+      "step": 384
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.595933294372878,
+      "learning_rate": 8.208955223880599e-06,
+      "loss": 0.9234,
+      "step": 385
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.654265208932199,
+      "learning_rate": 8.230277185501066e-06,
+      "loss": 0.8058,
+      "step": 386
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5054557088974927,
+      "learning_rate": 8.251599147121537e-06,
+      "loss": 0.6222,
+      "step": 387
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3558906109267668,
+      "learning_rate": 8.272921108742004e-06,
+      "loss": 0.6057,
+      "step": 388
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9399222263124087,
+      "learning_rate": 8.294243070362475e-06,
+      "loss": 0.8107,
+      "step": 389
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6762587530738358,
+      "learning_rate": 8.315565031982942e-06,
+      "loss": 0.856,
+      "step": 390
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6122736190165747,
+      "learning_rate": 8.336886993603413e-06,
+      "loss": 0.8957,
+      "step": 391
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.401073498293609,
+      "learning_rate": 8.35820895522388e-06,
+      "loss": 0.6931,
+      "step": 392
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1117984661709674,
+      "learning_rate": 8.379530916844351e-06,
+      "loss": 1.1456,
+      "step": 393
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7442856947459806,
+      "learning_rate": 8.400852878464819e-06,
+      "loss": 0.8895,
+      "step": 394
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7017537698898382,
+      "learning_rate": 8.42217484008529e-06,
+      "loss": 1.0621,
+      "step": 395
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.761216721982463,
+      "learning_rate": 8.443496801705757e-06,
+      "loss": 0.8559,
+      "step": 396
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.741370281948164,
+      "learning_rate": 8.464818763326227e-06,
+      "loss": 0.8657,
+      "step": 397
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.671995222592096,
+      "learning_rate": 8.486140724946695e-06,
+      "loss": 0.8323,
+      "step": 398
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.757925705306604,
+      "learning_rate": 8.507462686567165e-06,
+      "loss": 0.7502,
+      "step": 399
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5085307541389716,
+      "learning_rate": 8.528784648187633e-06,
+      "loss": 0.7941,
+      "step": 400
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3082285454192426,
+      "learning_rate": 8.550106609808104e-06,
+      "loss": 0.8082,
+      "step": 401
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8825476148829077,
+      "learning_rate": 8.571428571428571e-06,
+      "loss": 0.9117,
+      "step": 402
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4104937405963727,
+      "learning_rate": 8.592750533049042e-06,
+      "loss": 0.7041,
+      "step": 403
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.497902074620827,
+      "learning_rate": 8.614072494669509e-06,
+      "loss": 0.7545,
+      "step": 404
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5766426618300466,
+      "learning_rate": 8.63539445628998e-06,
+      "loss": 0.9195,
+      "step": 405
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6844229510985422,
+      "learning_rate": 8.656716417910447e-06,
+      "loss": 0.8435,
+      "step": 406
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.757836241994965,
+      "learning_rate": 8.678038379530918e-06,
+      "loss": 0.8316,
+      "step": 407
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7911942864704007,
+      "learning_rate": 8.699360341151387e-06,
+      "loss": 0.838,
+      "step": 408
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.600360708745728,
+      "learning_rate": 8.720682302771856e-06,
+      "loss": 0.7094,
+      "step": 409
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8202991489103166,
+      "learning_rate": 8.742004264392325e-06,
+      "loss": 0.9076,
+      "step": 410
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7024055741787603,
+      "learning_rate": 8.763326226012794e-06,
+      "loss": 0.8779,
+      "step": 411
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.546979200395418,
+      "learning_rate": 8.784648187633263e-06,
+      "loss": 0.8957,
+      "step": 412
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2971536878515493,
+      "learning_rate": 8.805970149253732e-06,
+      "loss": 0.6483,
+      "step": 413
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4996686489692168,
+      "learning_rate": 8.827292110874201e-06,
+      "loss": 0.82,
+      "step": 414
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7707747190546252,
+      "learning_rate": 8.84861407249467e-06,
+      "loss": 0.8427,
+      "step": 415
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4063848952175155,
+      "learning_rate": 8.86993603411514e-06,
+      "loss": 0.655,
+      "step": 416
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6359768899208984,
+      "learning_rate": 8.891257995735608e-06,
+      "loss": 0.8342,
+      "step": 417
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5713059156124183,
+      "learning_rate": 8.912579957356077e-06,
+      "loss": 0.8448,
+      "step": 418
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.58649240039141,
+      "learning_rate": 8.933901918976547e-06,
+      "loss": 0.8732,
+      "step": 419
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5817705503814337,
+      "learning_rate": 8.955223880597016e-06,
+      "loss": 0.9685,
+      "step": 420
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5597636442459164,
+      "learning_rate": 8.976545842217485e-06,
+      "loss": 0.7766,
+      "step": 421
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5597855774773655,
+      "learning_rate": 8.997867803837954e-06,
+      "loss": 0.8189,
+      "step": 422
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.628980067591861,
+      "learning_rate": 9.019189765458423e-06,
+      "loss": 0.9341,
+      "step": 423
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3486050077297413,
+      "learning_rate": 9.040511727078892e-06,
+      "loss": 0.7473,
+      "step": 424
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3823273116866808,
+      "learning_rate": 9.06183368869936e-06,
+      "loss": 0.6634,
+      "step": 425
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5293351544994187,
+      "learning_rate": 9.08315565031983e-06,
+      "loss": 0.858,
+      "step": 426
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5821583286967735,
+      "learning_rate": 9.104477611940299e-06,
+      "loss": 0.7563,
+      "step": 427
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.452500136780569,
+      "learning_rate": 9.125799573560768e-06,
+      "loss": 0.6816,
+      "step": 428
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5816116773261724,
+      "learning_rate": 9.147121535181237e-06,
+      "loss": 0.7875,
+      "step": 429
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5902726307553974,
+      "learning_rate": 9.168443496801706e-06,
+      "loss": 0.7738,
+      "step": 430
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4362909964776196,
+      "learning_rate": 9.189765458422175e-06,
+      "loss": 0.6812,
+      "step": 431
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8254528283039806,
+      "learning_rate": 9.211087420042644e-06,
+      "loss": 0.7495,
+      "step": 432
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5308661023191446,
+      "learning_rate": 9.232409381663113e-06,
+      "loss": 0.8184,
+      "step": 433
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5981364368494806,
+      "learning_rate": 9.253731343283582e-06,
+      "loss": 0.8132,
+      "step": 434
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5136420017391101,
+      "learning_rate": 9.275053304904051e-06,
+      "loss": 0.7181,
+      "step": 435
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6095781076439561,
+      "learning_rate": 9.29637526652452e-06,
+      "loss": 1.0604,
+      "step": 436
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5784651177750026,
+      "learning_rate": 9.31769722814499e-06,
+      "loss": 0.8455,
+      "step": 437
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3654592436205628,
+      "learning_rate": 9.339019189765458e-06,
+      "loss": 0.7787,
+      "step": 438
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.728331635417254,
+      "learning_rate": 9.36034115138593e-06,
+      "loss": 0.7895,
+      "step": 439
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3736342342368286,
+      "learning_rate": 9.381663113006397e-06,
+      "loss": 0.731,
+      "step": 440
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5911108090890478,
+      "learning_rate": 9.402985074626867e-06,
+      "loss": 0.8574,
+      "step": 441
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9555822641105496,
+      "learning_rate": 9.424307036247335e-06,
+      "loss": 0.9233,
+      "step": 442
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5079799030526357,
+      "learning_rate": 9.445628997867805e-06,
+      "loss": 0.7672,
+      "step": 443
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.202537238101738,
+      "learning_rate": 9.466950959488273e-06,
+      "loss": 1.0329,
+      "step": 444
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3742096901615435,
+      "learning_rate": 9.488272921108744e-06,
+      "loss": 0.7299,
+      "step": 445
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4504698826141889,
+      "learning_rate": 9.509594882729211e-06,
+      "loss": 0.7206,
+      "step": 446
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4942988052022523,
+      "learning_rate": 9.530916844349682e-06,
+      "loss": 0.7063,
+      "step": 447
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0745183580494495,
+      "learning_rate": 9.552238805970149e-06,
+      "loss": 0.8962,
+      "step": 448
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5430400635959165,
+      "learning_rate": 9.57356076759062e-06,
+      "loss": 0.7913,
+      "step": 449
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7171294431056987,
+      "learning_rate": 9.594882729211089e-06,
+      "loss": 0.8454,
+      "step": 450
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8617385432953328,
+      "learning_rate": 9.616204690831558e-06,
+      "loss": 0.8348,
+      "step": 451
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6266707761519836,
+      "learning_rate": 9.637526652452027e-06,
+      "loss": 0.8061,
+      "step": 452
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5735539104885543,
+      "learning_rate": 9.658848614072496e-06,
+      "loss": 0.6099,
+      "step": 453
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6626248629912752,
+      "learning_rate": 9.680170575692965e-06,
+      "loss": 0.7931,
+      "step": 454
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5135679938217836,
+      "learning_rate": 9.701492537313434e-06,
+      "loss": 0.6666,
+      "step": 455
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5127774030978827,
+      "learning_rate": 9.722814498933903e-06,
+      "loss": 0.8212,
+      "step": 456
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5279528886984677,
+      "learning_rate": 9.744136460554372e-06,
+      "loss": 0.833,
+      "step": 457
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5248562721011865,
+      "learning_rate": 9.765458422174841e-06,
+      "loss": 0.7547,
+      "step": 458
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.846153037404608,
+      "learning_rate": 9.78678038379531e-06,
+      "loss": 0.8493,
+      "step": 459
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4271502544262937,
+      "learning_rate": 9.80810234541578e-06,
+      "loss": 0.669,
+      "step": 460
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5009122372264254,
+      "learning_rate": 9.829424307036248e-06,
+      "loss": 0.7968,
+      "step": 461
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.651894396818958,
+      "learning_rate": 9.850746268656717e-06,
+      "loss": 0.8823,
+      "step": 462
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5495607242796068,
+      "learning_rate": 9.872068230277187e-06,
+      "loss": 0.8184,
+      "step": 463
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.66168959141969,
+      "learning_rate": 9.893390191897656e-06,
+      "loss": 0.8841,
+      "step": 464
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6162602497435663,
+      "learning_rate": 9.914712153518125e-06,
+      "loss": 0.8583,
+      "step": 465
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3270889684695664,
+      "learning_rate": 9.936034115138594e-06,
+      "loss": 0.842,
+      "step": 466
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4480730579411794,
+      "learning_rate": 9.957356076759063e-06,
+      "loss": 0.7873,
+      "step": 467
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4656897831038973,
+      "learning_rate": 9.978678038379532e-06,
+      "loss": 0.7509,
+      "step": 468
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2450958261447511,
+      "learning_rate": 1e-05,
+      "loss": 0.717,
+      "step": 469
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.878153859076313,
+      "learning_rate": 9.999999892569433e-06,
+      "loss": 0.9428,
+      "step": 470
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9016925946403571,
+      "learning_rate": 9.999999570277734e-06,
+      "loss": 0.8474,
+      "step": 471
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1490215601033071,
+      "learning_rate": 9.999999033124919e-06,
+      "loss": 0.7013,
+      "step": 472
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5060622506403898,
+      "learning_rate": 9.999998281111008e-06,
+      "loss": 0.7401,
+      "step": 473
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5527876358323738,
+      "learning_rate": 9.999997314236036e-06,
+      "loss": 0.7846,
+      "step": 474
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5713282762492382,
+      "learning_rate": 9.999996132500046e-06,
+      "loss": 0.8824,
+      "step": 475
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.4720631158265998,
+      "learning_rate": 9.999994735903083e-06,
+      "loss": 0.7962,
+      "step": 476
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3364552989446468,
+      "learning_rate": 9.999993124445213e-06,
+      "loss": 0.7237,
+      "step": 477
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6758630943434483,
+      "learning_rate": 9.999991298126501e-06,
+      "loss": 0.8803,
+      "step": 478
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.715701482286779,
+      "learning_rate": 9.999989256947029e-06,
+      "loss": 0.7846,
+      "step": 479
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7960390639024864,
+      "learning_rate": 9.999987000906883e-06,
+      "loss": 0.9051,
+      "step": 480
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7084511353415914,
+      "learning_rate": 9.99998453000616e-06,
+      "loss": 0.9287,
+      "step": 481
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9787045504875047,
+      "learning_rate": 9.999981844244966e-06,
+      "loss": 0.8802,
+      "step": 482
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.630363814803144,
+      "learning_rate": 9.999978943623417e-06,
+      "loss": 0.9169,
+      "step": 483
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3770468766932624,
+      "learning_rate": 9.999975828141635e-06,
+      "loss": 0.7192,
+      "step": 484
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.567907500182171,
+      "learning_rate": 9.999972497799758e-06,
+      "loss": 0.8179,
+      "step": 485
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8313714377601422,
+      "learning_rate": 9.999968952597926e-06,
+      "loss": 0.7439,
+      "step": 486
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.584107464914844,
+      "learning_rate": 9.999965192536295e-06,
+      "loss": 0.8949,
+      "step": 487
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.189438972331427,
+      "learning_rate": 9.999961217615022e-06,
+      "loss": 0.8021,
+      "step": 488
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3473216101352339,
+      "learning_rate": 9.999957027834282e-06,
+      "loss": 0.7439,
+      "step": 489
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6009455366610275,
+      "learning_rate": 9.999952623194252e-06,
+      "loss": 0.8319,
+      "step": 490
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3555233977478438,
+      "learning_rate": 9.99994800369512e-06,
+      "loss": 0.7935,
+      "step": 491
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0525703112500313,
+      "learning_rate": 9.99994316933709e-06,
+      "loss": 0.95,
+      "step": 492
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6842835093285387,
+      "learning_rate": 9.999938120120364e-06,
+      "loss": 0.7696,
+      "step": 493
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5315256301148386,
+      "learning_rate": 9.999932856045164e-06,
+      "loss": 0.8192,
+      "step": 494
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5617673747825906,
+      "learning_rate": 9.999927377111712e-06,
+      "loss": 0.8507,
+      "step": 495
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6436892307959907,
+      "learning_rate": 9.999921683320247e-06,
+      "loss": 0.7346,
+      "step": 496
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6365686386403502,
+      "learning_rate": 9.999915774671009e-06,
+      "loss": 0.8315,
+      "step": 497
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5002172071600035,
+      "learning_rate": 9.999909651164258e-06,
+      "loss": 0.7193,
+      "step": 498
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.426367485606446,
+      "learning_rate": 9.99990331280025e-06,
+      "loss": 0.7921,
+      "step": 499
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.270200023340624,
+      "learning_rate": 9.999896759579263e-06,
+      "loss": 0.7538,
+      "step": 500
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8099502584867322,
+      "learning_rate": 9.99988999150158e-06,
+      "loss": 0.9522,
+      "step": 501
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7442925804480571,
+      "learning_rate": 9.999883008567487e-06,
+      "loss": 0.7989,
+      "step": 502
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5792907387343391,
+      "learning_rate": 9.999875810777283e-06,
+      "loss": 0.8579,
+      "step": 503
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3386600683850423,
+      "learning_rate": 9.999868398131282e-06,
+      "loss": 0.7705,
+      "step": 504
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9439838339974918,
+      "learning_rate": 9.9998607706298e-06,
+      "loss": 0.7556,
+      "step": 505
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6916047318272367,
+      "learning_rate": 9.999852928273166e-06,
+      "loss": 0.829,
+      "step": 506
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6038243890001214,
+      "learning_rate": 9.999844871061718e-06,
+      "loss": 0.8858,
+      "step": 507
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4697367490605777,
+      "learning_rate": 9.999836598995799e-06,
+      "loss": 0.7791,
+      "step": 508
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5492055574502421,
+      "learning_rate": 9.999828112075764e-06,
+      "loss": 0.903,
+      "step": 509
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9481799931056822,
+      "learning_rate": 9.999819410301983e-06,
+      "loss": 0.6471,
+      "step": 510
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.532566731967807,
+      "learning_rate": 9.999810493674826e-06,
+      "loss": 0.8129,
+      "step": 511
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5785357054900324,
+      "learning_rate": 9.999801362194676e-06,
+      "loss": 0.7903,
+      "step": 512
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.479979576700831,
+      "learning_rate": 9.999792015861927e-06,
+      "loss": 0.7692,
+      "step": 513
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2564462111459878,
+      "learning_rate": 9.99978245467698e-06,
+      "loss": 0.6234,
+      "step": 514
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.541119582892493,
+      "learning_rate": 9.999772678640245e-06,
+      "loss": 0.9434,
+      "step": 515
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4181116260936175,
+      "learning_rate": 9.999762687752143e-06,
+      "loss": 0.7009,
+      "step": 516
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.228774210094001,
+      "learning_rate": 9.999752482013105e-06,
+      "loss": 0.7675,
+      "step": 517
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4474004757771448,
+      "learning_rate": 9.999742061423567e-06,
+      "loss": 0.8344,
+      "step": 518
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4691922131158641,
+      "learning_rate": 9.999731425983975e-06,
+      "loss": 0.7852,
+      "step": 519
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2430091965287802,
+      "learning_rate": 9.999720575694792e-06,
+      "loss": 0.766,
+      "step": 520
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5193634872246968,
+      "learning_rate": 9.99970951055648e-06,
+      "loss": 0.8109,
+      "step": 521
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5780778431675593,
+      "learning_rate": 9.999698230569515e-06,
+      "loss": 0.8743,
+      "step": 522
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.497172005680606,
+      "learning_rate": 9.999686735734382e-06,
+      "loss": 0.9453,
+      "step": 523
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5336442329851234,
+      "learning_rate": 9.999675026051576e-06,
+      "loss": 0.7784,
+      "step": 524
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3429826184537308,
+      "learning_rate": 9.999663101521599e-06,
+      "loss": 0.7953,
+      "step": 525
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7580816831329364,
+      "learning_rate": 9.999650962144962e-06,
+      "loss": 0.7385,
+      "step": 526
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8377516854540006,
+      "learning_rate": 9.999638607922191e-06,
+      "loss": 0.8444,
+      "step": 527
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.793435441170837,
+      "learning_rate": 9.99962603885381e-06,
+      "loss": 0.8806,
+      "step": 528
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3337020172842504,
+      "learning_rate": 9.999613254940368e-06,
+      "loss": 0.753,
+      "step": 529
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2084384307892415,
+      "learning_rate": 9.999600256182408e-06,
+      "loss": 0.7545,
+      "step": 530
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8779947881259353,
+      "learning_rate": 9.99958704258049e-06,
+      "loss": 0.8723,
+      "step": 531
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.135364192761596,
+      "learning_rate": 9.999573614135183e-06,
+      "loss": 0.6852,
+      "step": 532
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.475853096753333,
+      "learning_rate": 9.999559970847061e-06,
+      "loss": 0.8381,
+      "step": 533
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5142185973951299,
+      "learning_rate": 9.999546112716715e-06,
+      "loss": 0.7625,
+      "step": 534
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6034850181361069,
+      "learning_rate": 9.999532039744736e-06,
+      "loss": 0.8708,
+      "step": 535
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5840219052497961,
+      "learning_rate": 9.999517751931731e-06,
+      "loss": 0.7644,
+      "step": 536
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6429983699055624,
+      "learning_rate": 9.999503249278315e-06,
+      "loss": 0.8962,
+      "step": 537
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5772691332703774,
+      "learning_rate": 9.99948853178511e-06,
+      "loss": 0.8687,
+      "step": 538
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2976495566484472,
+      "learning_rate": 9.999473599452746e-06,
+      "loss": 0.6525,
+      "step": 539
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.625643301238781,
+      "learning_rate": 9.999458452281868e-06,
+      "loss": 0.7876,
+      "step": 540
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5003746626103744,
+      "learning_rate": 9.999443090273126e-06,
+      "loss": 0.6793,
+      "step": 541
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4162682380126543,
+      "learning_rate": 9.99942751342718e-06,
+      "loss": 0.7362,
+      "step": 542
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5431561223526225,
+      "learning_rate": 9.9994117217447e-06,
+      "loss": 0.7614,
+      "step": 543
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4413985642081044,
+      "learning_rate": 9.999395715226365e-06,
+      "loss": 0.702,
+      "step": 544
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5028066475554953,
+      "learning_rate": 9.99937949387286e-06,
+      "loss": 0.9294,
+      "step": 545
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2905220707414953,
+      "learning_rate": 9.999363057684885e-06,
+      "loss": 0.6749,
+      "step": 546
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7077843806386575,
+      "learning_rate": 9.999346406663144e-06,
+      "loss": 0.7211,
+      "step": 547
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8645146251292017,
+      "learning_rate": 9.999329540808355e-06,
+      "loss": 0.7761,
+      "step": 548
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6966831053547342,
+      "learning_rate": 9.999312460121242e-06,
+      "loss": 0.6828,
+      "step": 549
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.354764582858276,
+      "learning_rate": 9.999295164602536e-06,
+      "loss": 0.7188,
+      "step": 550
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.02211808061413,
+      "learning_rate": 9.999277654252985e-06,
+      "loss": 0.9135,
+      "step": 551
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.466523141935564,
+      "learning_rate": 9.999259929073339e-06,
+      "loss": 0.8751,
+      "step": 552
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.291628426096786,
+      "learning_rate": 9.999241989064358e-06,
+      "loss": 0.777,
+      "step": 553
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.633671121869679,
+      "learning_rate": 9.999223834226817e-06,
+      "loss": 0.7622,
+      "step": 554
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6189079438193514,
+      "learning_rate": 9.999205464561494e-06,
+      "loss": 0.651,
+      "step": 555
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7015971901980596,
+      "learning_rate": 9.999186880069178e-06,
+      "loss": 0.8285,
+      "step": 556
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6676466899704157,
+      "learning_rate": 9.999168080750666e-06,
+      "loss": 0.7894,
+      "step": 557
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.169030050305818,
+      "learning_rate": 9.99914906660677e-06,
+      "loss": 0.8555,
+      "step": 558
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6939788942255085,
+      "learning_rate": 9.999129837638303e-06,
+      "loss": 0.9121,
+      "step": 559
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8721923886112048,
+      "learning_rate": 9.999110393846097e-06,
+      "loss": 0.8119,
+      "step": 560
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.718674725683557,
+      "learning_rate": 9.999090735230981e-06,
+      "loss": 0.7745,
+      "step": 561
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8796841311652908,
+      "learning_rate": 9.999070861793803e-06,
+      "loss": 0.9364,
+      "step": 562
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.120298976940141,
+      "learning_rate": 9.999050773535414e-06,
+      "loss": 0.6535,
+      "step": 563
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3859911084528376,
+      "learning_rate": 9.999030470456684e-06,
+      "loss": 0.672,
+      "step": 564
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6201036257884096,
+      "learning_rate": 9.999009952558478e-06,
+      "loss": 0.7695,
+      "step": 565
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5392196696072675,
+      "learning_rate": 9.998989219841684e-06,
+      "loss": 0.7421,
+      "step": 566
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8077612494150117,
+      "learning_rate": 9.998968272307187e-06,
+      "loss": 0.779,
+      "step": 567
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4896025264181894,
+      "learning_rate": 9.998947109955892e-06,
+      "loss": 0.8397,
+      "step": 568
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.50050543276988,
+      "learning_rate": 9.998925732788706e-06,
+      "loss": 0.7243,
+      "step": 569
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.265751562273945,
+      "learning_rate": 9.998904140806547e-06,
+      "loss": 0.6385,
+      "step": 570
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4974591580153214,
+      "learning_rate": 9.998882334010345e-06,
+      "loss": 0.7733,
+      "step": 571
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5921818154561553,
+      "learning_rate": 9.998860312401035e-06,
+      "loss": 0.8704,
+      "step": 572
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5357889616088047,
+      "learning_rate": 9.998838075979566e-06,
+      "loss": 0.8352,
+      "step": 573
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1558763348497503,
+      "learning_rate": 9.99881562474689e-06,
+      "loss": 0.5842,
+      "step": 574
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4191118442846555,
+      "learning_rate": 9.998792958703975e-06,
+      "loss": 0.8107,
+      "step": 575
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5430368076487433,
+      "learning_rate": 9.998770077851794e-06,
+      "loss": 0.8073,
+      "step": 576
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.765954808843168,
+      "learning_rate": 9.998746982191327e-06,
+      "loss": 0.7993,
+      "step": 577
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6818372493770566,
+      "learning_rate": 9.998723671723572e-06,
+      "loss": 0.8789,
+      "step": 578
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4584136064772384,
+      "learning_rate": 9.998700146449528e-06,
+      "loss": 0.78,
+      "step": 579
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2151978471934914,
+      "learning_rate": 9.998676406370205e-06,
+      "loss": 0.9839,
+      "step": 580
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.417255787055671,
+      "learning_rate": 9.998652451486626e-06,
+      "loss": 0.7722,
+      "step": 581
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5767554632299468,
+      "learning_rate": 9.998628281799816e-06,
+      "loss": 0.8028,
+      "step": 582
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.570386876990803,
+      "learning_rate": 9.998603897310818e-06,
+      "loss": 0.798,
+      "step": 583
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6709126386405366,
+      "learning_rate": 9.998579298020676e-06,
+      "loss": 0.7998,
+      "step": 584
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4650928887191086,
+      "learning_rate": 9.998554483930452e-06,
+      "loss": 0.8184,
+      "step": 585
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1919091798378005,
+      "learning_rate": 9.998529455041209e-06,
+      "loss": 0.6358,
+      "step": 586
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8159410727410465,
+      "learning_rate": 9.998504211354021e-06,
+      "loss": 0.8036,
+      "step": 587
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.098924733930881,
+      "learning_rate": 9.998478752869976e-06,
+      "loss": 0.7978,
+      "step": 588
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4447979588218574,
+      "learning_rate": 9.998453079590167e-06,
+      "loss": 0.821,
+      "step": 589
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8721936741091754,
+      "learning_rate": 9.998427191515696e-06,
+      "loss": 0.8832,
+      "step": 590
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6165523417443415,
+      "learning_rate": 9.998401088647677e-06,
+      "loss": 0.7558,
+      "step": 591
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.419199439922824,
+      "learning_rate": 9.998374770987232e-06,
+      "loss": 0.8327,
+      "step": 592
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.899524768301563,
+      "learning_rate": 9.99834823853549e-06,
+      "loss": 0.8531,
+      "step": 593
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7578705837811506,
+      "learning_rate": 9.998321491293592e-06,
+      "loss": 0.5982,
+      "step": 594
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5606144108508713,
+      "learning_rate": 9.998294529262688e-06,
+      "loss": 0.739,
+      "step": 595
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.594192983880369,
+      "learning_rate": 9.998267352443938e-06,
+      "loss": 0.9028,
+      "step": 596
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4353398311000283,
+      "learning_rate": 9.998239960838505e-06,
+      "loss": 0.6723,
+      "step": 597
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6947419902060226,
+      "learning_rate": 9.998212354447572e-06,
+      "loss": 0.7175,
+      "step": 598
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6434854972413513,
+      "learning_rate": 9.998184533272321e-06,
+      "loss": 0.7827,
+      "step": 599
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8983108489094564,
+      "learning_rate": 9.99815649731395e-06,
+      "loss": 0.9743,
+      "step": 600
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.486330480408818,
+      "learning_rate": 9.998128246573663e-06,
+      "loss": 0.7312,
+      "step": 601
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4641545717696802,
+      "learning_rate": 9.998099781052673e-06,
+      "loss": 0.7932,
+      "step": 602
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5955912791316091,
+      "learning_rate": 9.998071100752203e-06,
+      "loss": 0.6652,
+      "step": 603
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6901262846925518,
+      "learning_rate": 9.998042205673489e-06,
+      "loss": 0.7382,
+      "step": 604
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5774658837389164,
+      "learning_rate": 9.998013095817768e-06,
+      "loss": 0.7281,
+      "step": 605
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7542567977936638,
+      "learning_rate": 9.997983771186293e-06,
+      "loss": 0.8649,
+      "step": 606
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4636519227704559,
+      "learning_rate": 9.997954231780326e-06,
+      "loss": 0.8167,
+      "step": 607
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8454284658264304,
+      "learning_rate": 9.997924477601134e-06,
+      "loss": 0.6861,
+      "step": 608
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4094483875553796,
+      "learning_rate": 9.997894508649995e-06,
+      "loss": 0.7427,
+      "step": 609
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7247405615294136,
+      "learning_rate": 9.9978643249282e-06,
+      "loss": 0.8366,
+      "step": 610
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5630259632203023,
+      "learning_rate": 9.997833926437043e-06,
+      "loss": 0.768,
+      "step": 611
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2602972271714756,
+      "learning_rate": 9.997803313177831e-06,
+      "loss": 0.5678,
+      "step": 612
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3262721353370739,
+      "learning_rate": 9.997772485151881e-06,
+      "loss": 0.7127,
+      "step": 613
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8140468857027885,
+      "learning_rate": 9.997741442360515e-06,
+      "loss": 0.8126,
+      "step": 614
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.806286979887251,
+      "learning_rate": 9.997710184805072e-06,
+      "loss": 0.9103,
+      "step": 615
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7776945836577387,
+      "learning_rate": 9.997678712486889e-06,
+      "loss": 0.7663,
+      "step": 616
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6118492725468505,
+      "learning_rate": 9.997647025407322e-06,
+      "loss": 0.6903,
+      "step": 617
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5937834754505305,
+      "learning_rate": 9.997615123567733e-06,
+      "loss": 0.9052,
+      "step": 618
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3949802428818217,
+      "learning_rate": 9.99758300696949e-06,
+      "loss": 0.7665,
+      "step": 619
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3816031391261951,
+      "learning_rate": 9.997550675613976e-06,
+      "loss": 0.7242,
+      "step": 620
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6825532915814525,
+      "learning_rate": 9.997518129502579e-06,
+      "loss": 0.8389,
+      "step": 621
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8352224272236262,
+      "learning_rate": 9.997485368636698e-06,
+      "loss": 0.8714,
+      "step": 622
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.558489019990855,
+      "learning_rate": 9.99745239301774e-06,
+      "loss": 0.8361,
+      "step": 623
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.205840534571298,
+      "learning_rate": 9.997419202647124e-06,
+      "loss": 0.6272,
+      "step": 624
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1826038522530846,
+      "learning_rate": 9.997385797526275e-06,
+      "loss": 0.6266,
+      "step": 625
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8073298449323252,
+      "learning_rate": 9.997352177656627e-06,
+      "loss": 0.8375,
+      "step": 626
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7203924107938608,
+      "learning_rate": 9.997318343039629e-06,
+      "loss": 0.8378,
+      "step": 627
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6029589317026098,
+      "learning_rate": 9.99728429367673e-06,
+      "loss": 0.804,
+      "step": 628
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7833056780259575,
+      "learning_rate": 9.997250029569395e-06,
+      "loss": 0.7631,
+      "step": 629
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7882802219826175,
+      "learning_rate": 9.997215550719097e-06,
+      "loss": 0.7756,
+      "step": 630
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4215676842517058,
+      "learning_rate": 9.997180857127318e-06,
+      "loss": 0.8698,
+      "step": 631
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.011851386944492,
+      "learning_rate": 9.997145948795548e-06,
+      "loss": 0.9596,
+      "step": 632
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5771492091319215,
+      "learning_rate": 9.997110825725288e-06,
+      "loss": 0.8693,
+      "step": 633
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4871410074707887,
+      "learning_rate": 9.997075487918047e-06,
+      "loss": 0.7571,
+      "step": 634
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.830874071614606,
+      "learning_rate": 9.99703993537534e-06,
+      "loss": 1.1158,
+      "step": 635
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6172975825543174,
+      "learning_rate": 9.997004168098703e-06,
+      "loss": 0.7389,
+      "step": 636
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.332175196514336,
+      "learning_rate": 9.996968186089664e-06,
+      "loss": 0.6217,
+      "step": 637
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6607378507528774,
+      "learning_rate": 9.996931989349775e-06,
+      "loss": 0.8262,
+      "step": 638
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.936509870574924,
+      "learning_rate": 9.99689557788059e-06,
+      "loss": 0.8812,
+      "step": 639
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.786608052551075,
+      "learning_rate": 9.996858951683672e-06,
+      "loss": 0.8193,
+      "step": 640
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7263513212428598,
+      "learning_rate": 9.996822110760595e-06,
+      "loss": 0.9127,
+      "step": 641
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3872039372754963,
+      "learning_rate": 9.996785055112947e-06,
+      "loss": 0.6825,
+      "step": 642
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4667322468753154,
+      "learning_rate": 9.996747784742314e-06,
+      "loss": 0.7867,
+      "step": 643
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4212674301200703,
+      "learning_rate": 9.996710299650302e-06,
+      "loss": 0.7104,
+      "step": 644
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9163607518496073,
+      "learning_rate": 9.99667259983852e-06,
+      "loss": 0.882,
+      "step": 645
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6058840925306783,
+      "learning_rate": 9.996634685308586e-06,
+      "loss": 0.8026,
+      "step": 646
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6288386044734222,
+      "learning_rate": 9.996596556062133e-06,
+      "loss": 0.8388,
+      "step": 647
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6570123799187884,
+      "learning_rate": 9.9965582121008e-06,
+      "loss": 0.8757,
+      "step": 648
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5267847501888014,
+      "learning_rate": 9.996519653426229e-06,
+      "loss": 0.7734,
+      "step": 649
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.576437562353175,
+      "learning_rate": 9.996480880040083e-06,
+      "loss": 0.7492,
+      "step": 650
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5051918409882359,
+      "learning_rate": 9.996441891944023e-06,
+      "loss": 0.7906,
+      "step": 651
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5109491989325714,
+      "learning_rate": 9.99640268913973e-06,
+      "loss": 0.8823,
+      "step": 652
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.682476387531278,
+      "learning_rate": 9.996363271628886e-06,
+      "loss": 0.8827,
+      "step": 653
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5243867057288945,
+      "learning_rate": 9.996323639413185e-06,
+      "loss": 0.7378,
+      "step": 654
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5114140407458632,
+      "learning_rate": 9.996283792494328e-06,
+      "loss": 0.8315,
+      "step": 655
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5667705496342657,
+      "learning_rate": 9.99624373087403e-06,
+      "loss": 0.7782,
+      "step": 656
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8254264123871016,
+      "learning_rate": 9.996203454554011e-06,
+      "loss": 0.8301,
+      "step": 657
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.307035373021903,
+      "learning_rate": 9.996162963536004e-06,
+      "loss": 0.739,
+      "step": 658
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1772878161716185,
+      "learning_rate": 9.996122257821746e-06,
+      "loss": 0.6603,
+      "step": 659
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9367530726944149,
+      "learning_rate": 9.996081337412988e-06,
+      "loss": 0.9538,
+      "step": 660
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6426923257760477,
+      "learning_rate": 9.996040202311488e-06,
+      "loss": 0.7262,
+      "step": 661
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0757299788265535,
+      "learning_rate": 9.995998852519014e-06,
+      "loss": 0.844,
+      "step": 662
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4577252990046023,
+      "learning_rate": 9.995957288037343e-06,
+      "loss": 0.6609,
+      "step": 663
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0990968971823214,
+      "learning_rate": 9.99591550886826e-06,
+      "loss": 0.776,
+      "step": 664
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.576628400109749,
+      "learning_rate": 9.995873515013562e-06,
+      "loss": 0.7811,
+      "step": 665
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7387887780069196,
+      "learning_rate": 9.99583130647505e-06,
+      "loss": 0.824,
+      "step": 666
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4681399060498563,
+      "learning_rate": 9.995788883254543e-06,
+      "loss": 0.8129,
+      "step": 667
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.412267691155963,
+      "learning_rate": 9.995746245353861e-06,
+      "loss": 0.6962,
+      "step": 668
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5352054379482114,
+      "learning_rate": 9.995703392774836e-06,
+      "loss": 0.7657,
+      "step": 669
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8353661452140775,
+      "learning_rate": 9.99566032551931e-06,
+      "loss": 0.856,
+      "step": 670
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.774765604512492,
+      "learning_rate": 9.995617043589135e-06,
+      "loss": 0.856,
+      "step": 671
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.67435423607112,
+      "learning_rate": 9.99557354698617e-06,
+      "loss": 0.852,
+      "step": 672
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1940727552660373,
+      "learning_rate": 9.995529835712283e-06,
+      "loss": 0.6694,
+      "step": 673
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2460212704511493,
+      "learning_rate": 9.995485909769354e-06,
+      "loss": 0.6281,
+      "step": 674
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5094744939907019,
+      "learning_rate": 9.995441769159269e-06,
+      "loss": 0.8,
+      "step": 675
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.659940303240236,
+      "learning_rate": 9.995397413883928e-06,
+      "loss": 0.8572,
+      "step": 676
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6540406366221783,
+      "learning_rate": 9.995352843945232e-06,
+      "loss": 0.7957,
+      "step": 677
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1554529153903637,
+      "learning_rate": 9.9953080593451e-06,
+      "loss": 0.7171,
+      "step": 678
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5076892035124057,
+      "learning_rate": 9.995263060085456e-06,
+      "loss": 0.8208,
+      "step": 679
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5241604951535201,
+      "learning_rate": 9.995217846168232e-06,
+      "loss": 0.6904,
+      "step": 680
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5547734645324214,
+      "learning_rate": 9.995172417595373e-06,
+      "loss": 0.7302,
+      "step": 681
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7403901452804906,
+      "learning_rate": 9.995126774368831e-06,
+      "loss": 0.7959,
+      "step": 682
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3361193308930064,
+      "learning_rate": 9.995080916490566e-06,
+      "loss": 0.7494,
+      "step": 683
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.094592950209753,
+      "learning_rate": 9.99503484396255e-06,
+      "loss": 0.7282,
+      "step": 684
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7206738311990621,
+      "learning_rate": 9.994988556786763e-06,
+      "loss": 0.8929,
+      "step": 685
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7295973259876762,
+      "learning_rate": 9.99494205496519e-06,
+      "loss": 0.8781,
+      "step": 686
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5140171529179172,
+      "learning_rate": 9.994895338499835e-06,
+      "loss": 0.8003,
+      "step": 687
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2514944700748996,
+      "learning_rate": 9.994848407392703e-06,
+      "loss": 0.6955,
+      "step": 688
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6312451439084852,
+      "learning_rate": 9.99480126164581e-06,
+      "loss": 0.78,
+      "step": 689
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4269210303663333,
+      "learning_rate": 9.994753901261183e-06,
+      "loss": 0.7705,
+      "step": 690
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6345059064988963,
+      "learning_rate": 9.994706326240858e-06,
+      "loss": 0.7398,
+      "step": 691
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7254283609407082,
+      "learning_rate": 9.994658536586878e-06,
+      "loss": 0.7593,
+      "step": 692
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.78123373340574,
+      "learning_rate": 9.994610532301296e-06,
+      "loss": 0.7423,
+      "step": 693
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5574500334947974,
+      "learning_rate": 9.994562313386177e-06,
+      "loss": 0.8099,
+      "step": 694
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4478823180962226,
+      "learning_rate": 9.99451387984359e-06,
+      "loss": 0.721,
+      "step": 695
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7343600905145107,
+      "learning_rate": 9.99446523167562e-06,
+      "loss": 0.8405,
+      "step": 696
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.277869348811036,
+      "learning_rate": 9.994416368884357e-06,
+      "loss": 0.7618,
+      "step": 697
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.51641359510978,
+      "learning_rate": 9.994367291471896e-06,
+      "loss": 0.899,
+      "step": 698
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6003214811452804,
+      "learning_rate": 9.994317999440351e-06,
+      "loss": 0.8283,
+      "step": 699
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6907591293862059,
+      "learning_rate": 9.99426849279184e-06,
+      "loss": 0.879,
+      "step": 700
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3744964818437047,
+      "learning_rate": 9.994218771528487e-06,
+      "loss": 0.7881,
+      "step": 701
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6644791527987162,
+      "learning_rate": 9.994168835652432e-06,
+      "loss": 0.7813,
+      "step": 702
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4920224537504456,
+      "learning_rate": 9.994118685165819e-06,
+      "loss": 0.8213,
+      "step": 703
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4920660572150584,
+      "learning_rate": 9.994068320070805e-06,
+      "loss": 0.7693,
+      "step": 704
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.551103258889897,
+      "learning_rate": 9.99401774036955e-06,
+      "loss": 0.8714,
+      "step": 705
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4666304633960856,
+      "learning_rate": 9.993966946064233e-06,
+      "loss": 0.766,
+      "step": 706
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0687464072716213,
+      "learning_rate": 9.993915937157033e-06,
+      "loss": 0.8798,
+      "step": 707
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4299739298224974,
+      "learning_rate": 9.993864713650143e-06,
+      "loss": 0.7312,
+      "step": 708
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.52701283720996,
+      "learning_rate": 9.993813275545764e-06,
+      "loss": 0.6479,
+      "step": 709
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6083409958604795,
+      "learning_rate": 9.993761622846107e-06,
+      "loss": 0.7232,
+      "step": 710
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.63022099263237,
+      "learning_rate": 9.99370975555339e-06,
+      "loss": 0.7641,
+      "step": 711
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2001075238309435,
+      "learning_rate": 9.993657673669848e-06,
+      "loss": 0.644,
+      "step": 712
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4639716233520481,
+      "learning_rate": 9.993605377197711e-06,
+      "loss": 0.7113,
+      "step": 713
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7406947105221242,
+      "learning_rate": 9.99355286613923e-06,
+      "loss": 0.8243,
+      "step": 714
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.665629327096355,
+      "learning_rate": 9.993500140496659e-06,
+      "loss": 0.9279,
+      "step": 715
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5445325179303495,
+      "learning_rate": 9.993447200272269e-06,
+      "loss": 0.7614,
+      "step": 716
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2086699930659421,
+      "learning_rate": 9.993394045468329e-06,
+      "loss": 0.6907,
+      "step": 717
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.415467118169142,
+      "learning_rate": 9.993340676087129e-06,
+      "loss": 0.8052,
+      "step": 718
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5716561071007749,
+      "learning_rate": 9.993287092130956e-06,
+      "loss": 0.8043,
+      "step": 719
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6761006432833203,
+      "learning_rate": 9.993233293602119e-06,
+      "loss": 0.9314,
+      "step": 720
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5694130890070195,
+      "learning_rate": 9.993179280502926e-06,
+      "loss": 0.8933,
+      "step": 721
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6059396851546626,
+      "learning_rate": 9.993125052835699e-06,
+      "loss": 0.7841,
+      "step": 722
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7935783540347578,
+      "learning_rate": 9.993070610602766e-06,
+      "loss": 0.9074,
+      "step": 723
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2591328325992546,
+      "learning_rate": 9.993015953806472e-06,
+      "loss": 0.7316,
+      "step": 724
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.380257156855885,
+      "learning_rate": 9.992961082449161e-06,
+      "loss": 0.7417,
+      "step": 725
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5597324816988445,
+      "learning_rate": 9.992905996533193e-06,
+      "loss": 0.8766,
+      "step": 726
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4561974876748667,
+      "learning_rate": 9.992850696060933e-06,
+      "loss": 0.7449,
+      "step": 727
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.106977861335658,
+      "learning_rate": 9.99279518103476e-06,
+      "loss": 0.7833,
+      "step": 728
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.8983321297430464,
+      "learning_rate": 9.992739451457058e-06,
+      "loss": 0.9396,
+      "step": 729
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.71067169796286,
+      "learning_rate": 9.992683507330223e-06,
+      "loss": 0.8165,
+      "step": 730
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5313311558298586,
+      "learning_rate": 9.992627348656658e-06,
+      "loss": 0.6732,
+      "step": 731
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.554056034930593,
+      "learning_rate": 9.992570975438776e-06,
+      "loss": 0.7817,
+      "step": 732
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.523886631225325,
+      "learning_rate": 9.992514387679e-06,
+      "loss": 1.0578,
+      "step": 733
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.051964609866181,
+      "learning_rate": 9.992457585379764e-06,
+      "loss": 0.9321,
+      "step": 734
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.515196479870815,
+      "learning_rate": 9.992400568543506e-06,
+      "loss": 0.9258,
+      "step": 735
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5939299270502183,
+      "learning_rate": 9.992343337172674e-06,
+      "loss": 0.7249,
+      "step": 736
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1397833707569083,
+      "learning_rate": 9.992285891269732e-06,
+      "loss": 0.7532,
+      "step": 737
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.043068893389243,
+      "learning_rate": 9.992228230837148e-06,
+      "loss": 0.8538,
+      "step": 738
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9478029018971497,
+      "learning_rate": 9.992170355877398e-06,
+      "loss": 0.8092,
+      "step": 739
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5296696311009412,
+      "learning_rate": 9.99211226639297e-06,
+      "loss": 0.7692,
+      "step": 740
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6592072270660787,
+      "learning_rate": 9.99205396238636e-06,
+      "loss": 0.6896,
+      "step": 741
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4497205700598956,
+      "learning_rate": 9.991995443860074e-06,
+      "loss": 0.7478,
+      "step": 742
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6310665216402622,
+      "learning_rate": 9.991936710816624e-06,
+      "loss": 0.702,
+      "step": 743
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3491054020404742,
+      "learning_rate": 9.991877763258538e-06,
+      "loss": 0.7368,
+      "step": 744
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4713446591123842,
+      "learning_rate": 9.991818601188345e-06,
+      "loss": 0.7064,
+      "step": 745
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6374076128994866,
+      "learning_rate": 9.99175922460859e-06,
+      "loss": 0.8609,
+      "step": 746
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7942063262509733,
+      "learning_rate": 9.991699633521823e-06,
+      "loss": 0.7517,
+      "step": 747
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.334875118392853,
+      "learning_rate": 9.991639827930607e-06,
+      "loss": 0.7067,
+      "step": 748
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4233647248532624,
+      "learning_rate": 9.991579807837511e-06,
+      "loss": 0.7232,
+      "step": 749
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4350257169197969,
+      "learning_rate": 9.991519573245112e-06,
+      "loss": 0.7131,
+      "step": 750
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5841033206828634,
+      "learning_rate": 9.991459124156003e-06,
+      "loss": 0.7589,
+      "step": 751
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3955686145530095,
+      "learning_rate": 9.991398460572776e-06,
+      "loss": 0.5223,
+      "step": 752
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.607012287585254,
+      "learning_rate": 9.991337582498042e-06,
+      "loss": 0.6962,
+      "step": 753
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7370936067726643,
+      "learning_rate": 9.991276489934416e-06,
+      "loss": 0.779,
+      "step": 754
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7146269719932536,
+      "learning_rate": 9.991215182884521e-06,
+      "loss": 0.739,
+      "step": 755
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1005234908097354,
+      "learning_rate": 9.991153661350996e-06,
+      "loss": 0.6635,
+      "step": 756
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.590575434506917,
+      "learning_rate": 9.991091925336481e-06,
+      "loss": 0.8011,
+      "step": 757
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.435210385568162,
+      "learning_rate": 9.991029974843632e-06,
+      "loss": 0.8395,
+      "step": 758
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3008508587823944,
+      "learning_rate": 9.990967809875107e-06,
+      "loss": 0.7517,
+      "step": 759
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.339979237340746,
+      "learning_rate": 9.990905430433582e-06,
+      "loss": 0.7339,
+      "step": 760
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.524662406856987,
+      "learning_rate": 9.990842836521732e-06,
+      "loss": 0.8228,
+      "step": 761
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5081144657319643,
+      "learning_rate": 9.990780028142253e-06,
+      "loss": 0.7656,
+      "step": 762
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6297379556449385,
+      "learning_rate": 9.990717005297841e-06,
+      "loss": 0.793,
+      "step": 763
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6116074333070705,
+      "learning_rate": 9.990653767991203e-06,
+      "loss": 0.7763,
+      "step": 764
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9898165165271076,
+      "learning_rate": 9.99059031622506e-06,
+      "loss": 0.8302,
+      "step": 765
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6125367609675323,
+      "learning_rate": 9.990526650002133e-06,
+      "loss": 0.6633,
+      "step": 766
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.203256634978752,
+      "learning_rate": 9.990462769325163e-06,
+      "loss": 0.5886,
+      "step": 767
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5923288914580793,
+      "learning_rate": 9.990398674196896e-06,
+      "loss": 0.6333,
+      "step": 768
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.377716946045291,
+      "learning_rate": 9.99033436462008e-06,
+      "loss": 0.7678,
+      "step": 769
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9379049321348518,
+      "learning_rate": 9.990269840597484e-06,
+      "loss": 0.7638,
+      "step": 770
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0367670434211376,
+      "learning_rate": 9.990205102131879e-06,
+      "loss": 0.7593,
+      "step": 771
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7114712001086383,
+      "learning_rate": 9.990140149226047e-06,
+      "loss": 0.9211,
+      "step": 772
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0084905858388926,
+      "learning_rate": 9.99007498188278e-06,
+      "loss": 0.9186,
+      "step": 773
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5289701917812106,
+      "learning_rate": 9.990009600104875e-06,
+      "loss": 0.8137,
+      "step": 774
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6295325112570145,
+      "learning_rate": 9.989944003895146e-06,
+      "loss": 0.8573,
+      "step": 775
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7844677650268335,
+      "learning_rate": 9.98987819325641e-06,
+      "loss": 0.9399,
+      "step": 776
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6109613623110102,
+      "learning_rate": 9.989812168191495e-06,
+      "loss": 0.7476,
+      "step": 777
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.220162448451518,
+      "learning_rate": 9.989745928703237e-06,
+      "loss": 0.8039,
+      "step": 778
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3928814528473954,
+      "learning_rate": 9.989679474794484e-06,
+      "loss": 0.7423,
+      "step": 779
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7772449950311437,
+      "learning_rate": 9.989612806468092e-06,
+      "loss": 0.6911,
+      "step": 780
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.673261178948494,
+      "learning_rate": 9.989545923726925e-06,
+      "loss": 0.7538,
+      "step": 781
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4862195622540189,
+      "learning_rate": 9.989478826573857e-06,
+      "loss": 0.7448,
+      "step": 782
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7096936451613214,
+      "learning_rate": 9.989411515011773e-06,
+      "loss": 0.7385,
+      "step": 783
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.323055410230329,
+      "learning_rate": 9.989343989043563e-06,
+      "loss": 0.6382,
+      "step": 784
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6358318767269098,
+      "learning_rate": 9.989276248672131e-06,
+      "loss": 0.7371,
+      "step": 785
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1492799447584254,
+      "learning_rate": 9.989208293900386e-06,
+      "loss": 0.5387,
+      "step": 786
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7840822516489887,
+      "learning_rate": 9.98914012473125e-06,
+      "loss": 0.8833,
+      "step": 787
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9095437239016235,
+      "learning_rate": 9.989071741167649e-06,
+      "loss": 0.8076,
+      "step": 788
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0866691016427183,
+      "learning_rate": 9.989003143212526e-06,
+      "loss": 0.9125,
+      "step": 789
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1564159693869627,
+      "learning_rate": 9.988934330868826e-06,
+      "loss": 0.6883,
+      "step": 790
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.409940115818162,
+      "learning_rate": 9.988865304139509e-06,
+      "loss": 0.7212,
+      "step": 791
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.559365707326197,
+      "learning_rate": 9.988796063027534e-06,
+      "loss": 0.8168,
+      "step": 792
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2802357127172104,
+      "learning_rate": 9.988726607535885e-06,
+      "loss": 0.69,
+      "step": 793
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5862594186195915,
+      "learning_rate": 9.988656937667544e-06,
+      "loss": 0.7974,
+      "step": 794
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7357515737405722,
+      "learning_rate": 9.988587053425504e-06,
+      "loss": 0.7265,
+      "step": 795
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8112849706495042,
+      "learning_rate": 9.988516954812767e-06,
+      "loss": 0.7959,
+      "step": 796
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8360955384670659,
+      "learning_rate": 9.988446641832346e-06,
+      "loss": 0.7453,
+      "step": 797
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6108993465133228,
+      "learning_rate": 9.988376114487264e-06,
+      "loss": 0.7272,
+      "step": 798
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.416756999404159,
+      "learning_rate": 9.98830537278055e-06,
+      "loss": 0.7798,
+      "step": 799
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4276806485913762,
+      "learning_rate": 9.988234416715246e-06,
+      "loss": 0.6968,
+      "step": 800
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4559551033828702,
+      "learning_rate": 9.988163246294398e-06,
+      "loss": 0.9253,
+      "step": 801
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4475217449803595,
+      "learning_rate": 9.988091861521067e-06,
+      "loss": 0.8093,
+      "step": 802
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5199572681580293,
+      "learning_rate": 9.98802026239832e-06,
+      "loss": 0.7029,
+      "step": 803
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4470682440133003,
+      "learning_rate": 9.987948448929232e-06,
+      "loss": 0.7726,
+      "step": 804
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4408837834322978,
+      "learning_rate": 9.98787642111689e-06,
+      "loss": 0.7823,
+      "step": 805
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6192825163341769,
+      "learning_rate": 9.987804178964391e-06,
+      "loss": 0.7519,
+      "step": 806
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.255764223472271,
+      "learning_rate": 9.987731722474839e-06,
+      "loss": 0.7868,
+      "step": 807
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5247949911388754,
+      "learning_rate": 9.987659051651345e-06,
+      "loss": 0.8115,
+      "step": 808
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5369432894426167,
+      "learning_rate": 9.987586166497032e-06,
+      "loss": 0.7575,
+      "step": 809
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6503639500092357,
+      "learning_rate": 9.987513067015035e-06,
+      "loss": 0.7252,
+      "step": 810
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3702536161050305,
+      "learning_rate": 9.987439753208494e-06,
+      "loss": 0.6852,
+      "step": 811
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2337481805534065,
+      "learning_rate": 9.98736622508056e-06,
+      "loss": 0.8343,
+      "step": 812
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3739949078010092,
+      "learning_rate": 9.98729248263439e-06,
+      "loss": 0.7893,
+      "step": 813
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5818165488896252,
+      "learning_rate": 9.987218525873155e-06,
+      "loss": 0.8002,
+      "step": 814
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7852045779687962,
+      "learning_rate": 9.987144354800035e-06,
+      "loss": 0.7671,
+      "step": 815
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5968947084205105,
+      "learning_rate": 9.987069969418213e-06,
+      "loss": 0.7247,
+      "step": 816
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.560298757719467,
+      "learning_rate": 9.986995369730887e-06,
+      "loss": 0.6499,
+      "step": 817
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2770427124559096,
+      "learning_rate": 9.986920555741267e-06,
+      "loss": 0.6556,
+      "step": 818
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6807255851607232,
+      "learning_rate": 9.98684552745256e-06,
+      "loss": 0.8014,
+      "step": 819
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2082912048605636,
+      "learning_rate": 9.986770284867997e-06,
+      "loss": 0.6001,
+      "step": 820
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2435639956363929,
+      "learning_rate": 9.986694827990807e-06,
+      "loss": 0.6616,
+      "step": 821
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4980625614830672,
+      "learning_rate": 9.986619156824236e-06,
+      "loss": 0.8192,
+      "step": 822
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.084197794241925,
+      "learning_rate": 9.986543271371532e-06,
+      "loss": 0.7742,
+      "step": 823
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5788059512562183,
+      "learning_rate": 9.98646717163596e-06,
+      "loss": 0.7961,
+      "step": 824
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8341409285305872,
+      "learning_rate": 9.986390857620787e-06,
+      "loss": 0.6895,
+      "step": 825
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3663460363450484,
+      "learning_rate": 9.986314329329294e-06,
+      "loss": 0.6638,
+      "step": 826
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.086700222154728,
+      "learning_rate": 9.986237586764768e-06,
+      "loss": 0.7938,
+      "step": 827
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2688056400741798,
+      "learning_rate": 9.98616062993051e-06,
+      "loss": 0.6934,
+      "step": 828
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4264849692413857,
+      "learning_rate": 9.986083458829824e-06,
+      "loss": 0.7621,
+      "step": 829
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4603038045864911,
+      "learning_rate": 9.986006073466027e-06,
+      "loss": 0.8242,
+      "step": 830
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4394162466129767,
+      "learning_rate": 9.985928473842445e-06,
+      "loss": 0.7042,
+      "step": 831
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.440842049890511,
+      "learning_rate": 9.985850659962411e-06,
+      "loss": 0.9094,
+      "step": 832
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6419488598886696,
+      "learning_rate": 9.985772631829272e-06,
+      "loss": 0.948,
+      "step": 833
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.476025153470772,
+      "learning_rate": 9.985694389446378e-06,
+      "loss": 0.6633,
+      "step": 834
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4479239602240426,
+      "learning_rate": 9.985615932817092e-06,
+      "loss": 0.8071,
+      "step": 835
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6830362644231764,
+      "learning_rate": 9.985537261944787e-06,
+      "loss": 0.8798,
+      "step": 836
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5655057819136522,
+      "learning_rate": 9.985458376832842e-06,
+      "loss": 0.7484,
+      "step": 837
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5831508123206581,
+      "learning_rate": 9.985379277484646e-06,
+      "loss": 0.882,
+      "step": 838
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.606311993027777,
+      "learning_rate": 9.9852999639036e-06,
+      "loss": 0.7666,
+      "step": 839
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5059559504383295,
+      "learning_rate": 9.985220436093112e-06,
+      "loss": 0.6986,
+      "step": 840
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.436043796468328,
+      "learning_rate": 9.985140694056601e-06,
+      "loss": 0.7709,
+      "step": 841
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1313512006570012,
+      "learning_rate": 9.985060737797489e-06,
+      "loss": 0.699,
+      "step": 842
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.437950207602335,
+      "learning_rate": 9.984980567319216e-06,
+      "loss": 0.771,
+      "step": 843
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5303123583819955,
+      "learning_rate": 9.984900182625226e-06,
+      "loss": 0.834,
+      "step": 844
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6681756727573642,
+      "learning_rate": 9.984819583718973e-06,
+      "loss": 0.9195,
+      "step": 845
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4175089494252087,
+      "learning_rate": 9.984738770603921e-06,
+      "loss": 0.8247,
+      "step": 846
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5051551886702215,
+      "learning_rate": 9.984657743283543e-06,
+      "loss": 0.7598,
+      "step": 847
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9180944466059937,
+      "learning_rate": 9.984576501761319e-06,
+      "loss": 0.8588,
+      "step": 848
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5307633866924202,
+      "learning_rate": 9.98449504604074e-06,
+      "loss": 0.7389,
+      "step": 849
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5066027364595385,
+      "learning_rate": 9.984413376125312e-06,
+      "loss": 0.7307,
+      "step": 850
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6042232406718793,
+      "learning_rate": 9.984331492018537e-06,
+      "loss": 0.7546,
+      "step": 851
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.527145830755461,
+      "learning_rate": 9.984249393723939e-06,
+      "loss": 0.7487,
+      "step": 852
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4600225262092703,
+      "learning_rate": 9.984167081245044e-06,
+      "loss": 0.7214,
+      "step": 853
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.422102451802642,
+      "learning_rate": 9.984084554585387e-06,
+      "loss": 0.7883,
+      "step": 854
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.531600126827573,
+      "learning_rate": 9.984001813748519e-06,
+      "loss": 0.7611,
+      "step": 855
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.990055486015273,
+      "learning_rate": 9.98391885873799e-06,
+      "loss": 0.7727,
+      "step": 856
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3672617587814788,
+      "learning_rate": 9.98383568955737e-06,
+      "loss": 0.6614,
+      "step": 857
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7368032918325147,
+      "learning_rate": 9.983752306210233e-06,
+      "loss": 0.8304,
+      "step": 858
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8806788899725522,
+      "learning_rate": 9.983668708700156e-06,
+      "loss": 0.8252,
+      "step": 859
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3572170728743553,
+      "learning_rate": 9.983584897030737e-06,
+      "loss": 0.6484,
+      "step": 860
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.41570165589477,
+      "learning_rate": 9.983500871205577e-06,
+      "loss": 0.7518,
+      "step": 861
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.713516725934832,
+      "learning_rate": 9.983416631228284e-06,
+      "loss": 0.7218,
+      "step": 862
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.707835499805939,
+      "learning_rate": 9.98333217710248e-06,
+      "loss": 0.9031,
+      "step": 863
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4453355188703811,
+      "learning_rate": 9.983247508831795e-06,
+      "loss": 0.758,
+      "step": 864
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6724583445432488,
+      "learning_rate": 9.983162626419866e-06,
+      "loss": 0.8765,
+      "step": 865
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.273533466866713,
+      "learning_rate": 9.98307752987034e-06,
+      "loss": 0.705,
+      "step": 866
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.470419712845699,
+      "learning_rate": 9.982992219186876e-06,
+      "loss": 0.6854,
+      "step": 867
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4118965857659616,
+      "learning_rate": 9.982906694373136e-06,
+      "loss": 0.7857,
+      "step": 868
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2588748690079887,
+      "learning_rate": 9.9828209554328e-06,
+      "loss": 0.7014,
+      "step": 869
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4222880244864218,
+      "learning_rate": 9.98273500236955e-06,
+      "loss": 0.7625,
+      "step": 870
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4969173909178917,
+      "learning_rate": 9.98264883518708e-06,
+      "loss": 0.8275,
+      "step": 871
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4245415957553595,
+      "learning_rate": 9.98256245388909e-06,
+      "loss": 0.7442,
+      "step": 872
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5472555592956156,
+      "learning_rate": 9.982475858479298e-06,
+      "loss": 0.6869,
+      "step": 873
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3747385233658231,
+      "learning_rate": 9.982389048961421e-06,
+      "loss": 0.7638,
+      "step": 874
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.665993496135593,
+      "learning_rate": 9.98230202533919e-06,
+      "loss": 0.8272,
+      "step": 875
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4347795671673889,
+      "learning_rate": 9.982214787616345e-06,
+      "loss": 0.802,
+      "step": 876
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3886616312396816,
+      "learning_rate": 9.982127335796631e-06,
+      "loss": 0.7306,
+      "step": 877
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0228818758583589,
+      "learning_rate": 9.982039669883815e-06,
+      "loss": 0.6055,
+      "step": 878
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0093942691806634,
+      "learning_rate": 9.981951789881657e-06,
+      "loss": 0.7667,
+      "step": 879
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7158639583158275,
+      "learning_rate": 9.981863695793934e-06,
+      "loss": 0.7419,
+      "step": 880
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5372538476679733,
+      "learning_rate": 9.981775387624433e-06,
+      "loss": 0.732,
+      "step": 881
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.668774429246153,
+      "learning_rate": 9.98168686537695e-06,
+      "loss": 0.9609,
+      "step": 882
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1454317205876723,
+      "learning_rate": 9.981598129055286e-06,
+      "loss": 0.6216,
+      "step": 883
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5038348664029046,
+      "learning_rate": 9.981509178663256e-06,
+      "loss": 0.7959,
+      "step": 884
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4243800241234545,
+      "learning_rate": 9.981420014204682e-06,
+      "loss": 0.7763,
+      "step": 885
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3502935267003755,
+      "learning_rate": 9.981330635683396e-06,
+      "loss": 0.7304,
+      "step": 886
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.421944841429418,
+      "learning_rate": 9.981241043103239e-06,
+      "loss": 0.7715,
+      "step": 887
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.669720812137144,
+      "learning_rate": 9.981151236468061e-06,
+      "loss": 0.7994,
+      "step": 888
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8725203903020344,
+      "learning_rate": 9.98106121578172e-06,
+      "loss": 0.9168,
+      "step": 889
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6040375720536968,
+      "learning_rate": 9.980970981048086e-06,
+      "loss": 0.8126,
+      "step": 890
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5989003783269793,
+      "learning_rate": 9.980880532271033e-06,
+      "loss": 0.7351,
+      "step": 891
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8673370092028685,
+      "learning_rate": 9.980789869454454e-06,
+      "loss": 0.7332,
+      "step": 892
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4452713205803338,
+      "learning_rate": 9.980698992602241e-06,
+      "loss": 0.7645,
+      "step": 893
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1487240134662833,
+      "learning_rate": 9.980607901718297e-06,
+      "loss": 0.6425,
+      "step": 894
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5288340260799809,
+      "learning_rate": 9.98051659680654e-06,
+      "loss": 0.745,
+      "step": 895
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4085066262932264,
+      "learning_rate": 9.980425077870895e-06,
+      "loss": 0.7852,
+      "step": 896
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.549462476732744,
+      "learning_rate": 9.98033334491529e-06,
+      "loss": 0.7079,
+      "step": 897
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6545204938090148,
+      "learning_rate": 9.98024139794367e-06,
+      "loss": 0.8482,
+      "step": 898
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6917841038958217,
+      "learning_rate": 9.980149236959986e-06,
+      "loss": 0.8136,
+      "step": 899
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.670354681193323,
+      "learning_rate": 9.980056861968197e-06,
+      "loss": 0.91,
+      "step": 900
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3649241160215586,
+      "learning_rate": 9.979964272972275e-06,
+      "loss": 0.8138,
+      "step": 901
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4489443872692795,
+      "learning_rate": 9.979871469976197e-06,
+      "loss": 0.8266,
+      "step": 902
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6128665912193416,
+      "learning_rate": 9.979778452983949e-06,
+      "loss": 0.7788,
+      "step": 903
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6344427007093028,
+      "learning_rate": 9.979685221999532e-06,
+      "loss": 0.6287,
+      "step": 904
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4423209846499008,
+      "learning_rate": 9.979591777026952e-06,
+      "loss": 0.7434,
+      "step": 905
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2933420260697805,
+      "learning_rate": 9.97949811807022e-06,
+      "loss": 0.7062,
+      "step": 906
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5913735412386918,
+      "learning_rate": 9.979404245133367e-06,
+      "loss": 0.7579,
+      "step": 907
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4994182902866042,
+      "learning_rate": 9.979310158220422e-06,
+      "loss": 0.7817,
+      "step": 908
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.692203668483005,
+      "learning_rate": 9.97921585733543e-06,
+      "loss": 0.7709,
+      "step": 909
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2509906319087505,
+      "learning_rate": 9.979121342482442e-06,
+      "loss": 0.6486,
+      "step": 910
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4007142522182847,
+      "learning_rate": 9.979026613665523e-06,
+      "loss": 0.7588,
+      "step": 911
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7146573838959782,
+      "learning_rate": 9.978931670888741e-06,
+      "loss": 0.8581,
+      "step": 912
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5046549375254576,
+      "learning_rate": 9.978836514156177e-06,
+      "loss": 0.8049,
+      "step": 913
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3750841553482822,
+      "learning_rate": 9.97874114347192e-06,
+      "loss": 0.7924,
+      "step": 914
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4310361727957712,
+      "learning_rate": 9.978645558840066e-06,
+      "loss": 0.8045,
+      "step": 915
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5049039692444999,
+      "learning_rate": 9.978549760264724e-06,
+      "loss": 0.8005,
+      "step": 916
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.338773002737649,
+      "learning_rate": 9.978453747750012e-06,
+      "loss": 0.8265,
+      "step": 917
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2981322903053525,
+      "learning_rate": 9.978357521300054e-06,
+      "loss": 0.6986,
+      "step": 918
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.623212463571559,
+      "learning_rate": 9.978261080918988e-06,
+      "loss": 0.8052,
+      "step": 919
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.0017534455630144,
+      "learning_rate": 9.978164426610956e-06,
+      "loss": 0.7295,
+      "step": 920
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3882566761689419,
+      "learning_rate": 9.97806755838011e-06,
+      "loss": 0.641,
+      "step": 921
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.394941937979967,
+      "learning_rate": 9.977970476230615e-06,
+      "loss": 0.6499,
+      "step": 922
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6978325547635944,
+      "learning_rate": 9.977873180166641e-06,
+      "loss": 0.7099,
+      "step": 923
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.169602944613949,
+      "learning_rate": 9.977775670192373e-06,
+      "loss": 0.8611,
+      "step": 924
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.438587596730339,
+      "learning_rate": 9.977677946311995e-06,
+      "loss": 0.7406,
+      "step": 925
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0238625341860008,
+      "learning_rate": 9.977580008529713e-06,
+      "loss": 0.6044,
+      "step": 926
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0943177493923784,
+      "learning_rate": 9.977481856849729e-06,
+      "loss": 0.6236,
+      "step": 927
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7347402153349691,
+      "learning_rate": 9.977383491276267e-06,
+      "loss": 0.7795,
+      "step": 928
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.397393173468354,
+      "learning_rate": 9.977284911813549e-06,
+      "loss": 0.7679,
+      "step": 929
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6640700219258673,
+      "learning_rate": 9.977186118465813e-06,
+      "loss": 0.8255,
+      "step": 930
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1659772479767097,
+      "learning_rate": 9.977087111237307e-06,
+      "loss": 0.5998,
+      "step": 931
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.344975235211494,
+      "learning_rate": 9.976987890132283e-06,
+      "loss": 0.825,
+      "step": 932
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2608727056889502,
+      "learning_rate": 9.976888455155002e-06,
+      "loss": 0.7473,
+      "step": 933
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6447757116383552,
+      "learning_rate": 9.976788806309742e-06,
+      "loss": 0.8042,
+      "step": 934
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.614159287316912,
+      "learning_rate": 9.976688943600786e-06,
+      "loss": 0.7979,
+      "step": 935
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.377420901229514,
+      "learning_rate": 9.97658886703242e-06,
+      "loss": 0.6644,
+      "step": 936
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5069631669175447,
+      "learning_rate": 9.976488576608945e-06,
+      "loss": 0.771,
+      "step": 937
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4070591064568188,
+      "learning_rate": 9.976388072334674e-06,
+      "loss": 0.6778,
+      "step": 938
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4770312148516982,
+      "learning_rate": 9.976287354213924e-06,
+      "loss": 0.7282,
+      "step": 939
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5282788493265222,
+      "learning_rate": 9.976186422251026e-06,
+      "loss": 0.8803,
+      "step": 940
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.332540461199421,
+      "learning_rate": 9.976085276450312e-06,
+      "loss": 0.7322,
+      "step": 941
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6206942960111619,
+      "learning_rate": 9.975983916816134e-06,
+      "loss": 0.8609,
+      "step": 942
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5770395463520734,
+      "learning_rate": 9.975882343352843e-06,
+      "loss": 0.8688,
+      "step": 943
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.312454207326133,
+      "learning_rate": 9.975780556064806e-06,
+      "loss": 0.6902,
+      "step": 944
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4662466782932841,
+      "learning_rate": 9.975678554956397e-06,
+      "loss": 0.7324,
+      "step": 945
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.128694577445927,
+      "learning_rate": 9.975576340032e-06,
+      "loss": 0.7387,
+      "step": 946
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.757637841903817,
+      "learning_rate": 9.975473911296005e-06,
+      "loss": 0.8967,
+      "step": 947
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.491865287749678,
+      "learning_rate": 9.975371268752816e-06,
+      "loss": 0.635,
+      "step": 948
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.757728291770356,
+      "learning_rate": 9.975268412406842e-06,
+      "loss": 0.8377,
+      "step": 949
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.464525482983179,
+      "learning_rate": 9.975165342262505e-06,
+      "loss": 0.8317,
+      "step": 950
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4780282520908854,
+      "learning_rate": 9.975062058324232e-06,
+      "loss": 0.7859,
+      "step": 951
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3304594188342351,
+      "learning_rate": 9.974958560596464e-06,
+      "loss": 0.6919,
+      "step": 952
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3653446050451592,
+      "learning_rate": 9.974854849083645e-06,
+      "loss": 0.6577,
+      "step": 953
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1005251324546168,
+      "learning_rate": 9.974750923790234e-06,
+      "loss": 0.6367,
+      "step": 954
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.797794216443401,
+      "learning_rate": 9.974646784720696e-06,
+      "loss": 0.7141,
+      "step": 955
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5834169015323658,
+      "learning_rate": 9.974542431879507e-06,
+      "loss": 0.7143,
+      "step": 956
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4092542486516018,
+      "learning_rate": 9.974437865271152e-06,
+      "loss": 0.8469,
+      "step": 957
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5765229132191503,
+      "learning_rate": 9.97433308490012e-06,
+      "loss": 0.798,
+      "step": 958
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4076559089669929,
+      "learning_rate": 9.97422809077092e-06,
+      "loss": 0.8923,
+      "step": 959
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3161971137992907,
+      "learning_rate": 9.97412288288806e-06,
+      "loss": 0.7004,
+      "step": 960
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6848627604320172,
+      "learning_rate": 9.974017461256062e-06,
+      "loss": 0.7453,
+      "step": 961
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5178924480567844,
+      "learning_rate": 9.973911825879454e-06,
+      "loss": 0.6455,
+      "step": 962
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5844487649610157,
+      "learning_rate": 9.97380597676278e-06,
+      "loss": 0.756,
+      "step": 963
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5121264907372594,
+      "learning_rate": 9.973699913910584e-06,
+      "loss": 1.0965,
+      "step": 964
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7288905619193375,
+      "learning_rate": 9.973593637327427e-06,
+      "loss": 0.7374,
+      "step": 965
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5584896153281236,
+      "learning_rate": 9.973487147017874e-06,
+      "loss": 0.8119,
+      "step": 966
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5068536341132224,
+      "learning_rate": 9.9733804429865e-06,
+      "loss": 0.6998,
+      "step": 967
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3040698676189462,
+      "learning_rate": 9.973273525237895e-06,
+      "loss": 0.6812,
+      "step": 968
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.462002796278723,
+      "learning_rate": 9.97316639377665e-06,
+      "loss": 0.7449,
+      "step": 969
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6578050384544005,
+      "learning_rate": 9.973059048607367e-06,
+      "loss": 0.8086,
+      "step": 970
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.537855842975262,
+      "learning_rate": 9.972951489734664e-06,
+      "loss": 0.7228,
+      "step": 971
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6176696464181268,
+      "learning_rate": 9.972843717163159e-06,
+      "loss": 0.8255,
+      "step": 972
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5809793269127372,
+      "learning_rate": 9.972735730897484e-06,
+      "loss": 0.7891,
+      "step": 973
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2057708923011243,
+      "learning_rate": 9.97262753094228e-06,
+      "loss": 0.7147,
+      "step": 974
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5640582386193078,
+      "learning_rate": 9.972519117302195e-06,
+      "loss": 0.8228,
+      "step": 975
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.370547818201438,
+      "learning_rate": 9.972410489981892e-06,
+      "loss": 0.701,
+      "step": 976
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4524894969587567,
+      "learning_rate": 9.972301648986033e-06,
+      "loss": 0.7005,
+      "step": 977
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7812795876958123,
+      "learning_rate": 9.972192594319301e-06,
+      "loss": 0.8208,
+      "step": 978
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5421665025103681,
+      "learning_rate": 9.972083325986377e-06,
+      "loss": 0.7787,
+      "step": 979
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5227855039897964,
+      "learning_rate": 9.97197384399196e-06,
+      "loss": 0.8645,
+      "step": 980
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.40487966161507,
+      "learning_rate": 9.971864148340755e-06,
+      "loss": 0.7292,
+      "step": 981
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5455482141123547,
+      "learning_rate": 9.971754239037474e-06,
+      "loss": 0.8145,
+      "step": 982
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.333093762909377,
+      "learning_rate": 9.971644116086841e-06,
+      "loss": 0.7706,
+      "step": 983
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.28082572706841,
+      "learning_rate": 9.971533779493586e-06,
+      "loss": 0.713,
+      "step": 984
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4344620584506502,
+      "learning_rate": 9.971423229262454e-06,
+      "loss": 0.7226,
+      "step": 985
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2096311344218116,
+      "learning_rate": 9.971312465398192e-06,
+      "loss": 0.6484,
+      "step": 986
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.623140076610551,
+      "learning_rate": 9.971201487905563e-06,
+      "loss": 0.776,
+      "step": 987
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.517856341968722,
+      "learning_rate": 9.971090296789336e-06,
+      "loss": 0.7603,
+      "step": 988
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6159152159598302,
+      "learning_rate": 9.970978892054286e-06,
+      "loss": 0.7388,
+      "step": 989
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6901285248521503,
+      "learning_rate": 9.970867273705203e-06,
+      "loss": 0.8709,
+      "step": 990
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0705962695385953,
+      "learning_rate": 9.970755441746882e-06,
+      "loss": 0.6545,
+      "step": 991
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4197848808580078,
+      "learning_rate": 9.970643396184128e-06,
+      "loss": 0.7805,
+      "step": 992
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6103598546638338,
+      "learning_rate": 9.970531137021759e-06,
+      "loss": 0.7984,
+      "step": 993
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4999511276720219,
+      "learning_rate": 9.970418664264596e-06,
+      "loss": 0.8221,
+      "step": 994
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0109366998341818,
+      "learning_rate": 9.970305977917473e-06,
+      "loss": 0.6647,
+      "step": 995
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.431338846814968,
+      "learning_rate": 9.970193077985235e-06,
+      "loss": 0.6949,
+      "step": 996
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3795705857852487,
+      "learning_rate": 9.97007996447273e-06,
+      "loss": 0.7785,
+      "step": 997
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3870873424428178,
+      "learning_rate": 9.969966637384818e-06,
+      "loss": 0.8496,
+      "step": 998
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4073751391001152,
+      "learning_rate": 9.969853096726372e-06,
+      "loss": 0.73,
+      "step": 999
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5245403001290596,
+      "learning_rate": 9.969739342502272e-06,
+      "loss": 0.7799,
+      "step": 1000
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5112876954661647,
+      "learning_rate": 9.969625374717401e-06,
+      "loss": 0.7343,
+      "step": 1001
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1234183677280627,
+      "learning_rate": 9.969511193376661e-06,
+      "loss": 0.6894,
+      "step": 1002
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1647709246945053,
+      "learning_rate": 9.969396798484958e-06,
+      "loss": 0.6899,
+      "step": 1003
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.469171609268134,
+      "learning_rate": 9.969282190047207e-06,
+      "loss": 0.8521,
+      "step": 1004
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.328008296745504,
+      "learning_rate": 9.969167368068334e-06,
+      "loss": 0.8008,
+      "step": 1005
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3759221864359548,
+      "learning_rate": 9.96905233255327e-06,
+      "loss": 0.7882,
+      "step": 1006
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.36529192051721,
+      "learning_rate": 9.968937083506962e-06,
+      "loss": 0.6963,
+      "step": 1007
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4201639282408849,
+      "learning_rate": 9.96882162093436e-06,
+      "loss": 0.7875,
+      "step": 1008
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1664215579810546,
+      "learning_rate": 9.968705944840428e-06,
+      "loss": 0.6588,
+      "step": 1009
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4310599367003038,
+      "learning_rate": 9.968590055230135e-06,
+      "loss": 0.7613,
+      "step": 1010
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2261948222740777,
+      "learning_rate": 9.968473952108462e-06,
+      "loss": 0.616,
+      "step": 1011
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9320418698965276,
+      "learning_rate": 9.968357635480397e-06,
+      "loss": 0.7979,
+      "step": 1012
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3489028525149145,
+      "learning_rate": 9.96824110535094e-06,
+      "loss": 0.7413,
+      "step": 1013
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.282180583700217,
+      "learning_rate": 9.968124361725098e-06,
+      "loss": 0.6515,
+      "step": 1014
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3745697747560224,
+      "learning_rate": 9.968007404607887e-06,
+      "loss": 0.7257,
+      "step": 1015
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.685605530691689,
+      "learning_rate": 9.967890234004335e-06,
+      "loss": 0.8592,
+      "step": 1016
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3043684490321477,
+      "learning_rate": 9.967772849919474e-06,
+      "loss": 0.791,
+      "step": 1017
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5690873090157484,
+      "learning_rate": 9.96765525235835e-06,
+      "loss": 0.9271,
+      "step": 1018
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2797990539351942,
+      "learning_rate": 9.967537441326018e-06,
+      "loss": 0.7267,
+      "step": 1019
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0663210985582467,
+      "learning_rate": 9.967419416827537e-06,
+      "loss": 0.6412,
+      "step": 1020
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.304664752108742,
+      "learning_rate": 9.967301178867981e-06,
+      "loss": 0.619,
+      "step": 1021
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.473794240095715,
+      "learning_rate": 9.96718272745243e-06,
+      "loss": 0.7399,
+      "step": 1022
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.440976969261973,
+      "learning_rate": 9.967064062585977e-06,
+      "loss": 0.7645,
+      "step": 1023
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3193311928623717,
+      "learning_rate": 9.966945184273716e-06,
+      "loss": 0.7498,
+      "step": 1024
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4554127927364628,
+      "learning_rate": 9.96682609252076e-06,
+      "loss": 0.7633,
+      "step": 1025
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4178773508664824,
+      "learning_rate": 9.966706787332224e-06,
+      "loss": 0.7373,
+      "step": 1026
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.633016856213013,
+      "learning_rate": 9.966587268713237e-06,
+      "loss": 0.99,
+      "step": 1027
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6774208921159635,
+      "learning_rate": 9.966467536668933e-06,
+      "loss": 0.8695,
+      "step": 1028
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4130520415062775,
+      "learning_rate": 9.966347591204459e-06,
+      "loss": 0.6985,
+      "step": 1029
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1171342619552873,
+      "learning_rate": 9.96622743232497e-06,
+      "loss": 0.6334,
+      "step": 1030
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2926654579472736,
+      "learning_rate": 9.966107060035624e-06,
+      "loss": 0.7145,
+      "step": 1031
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5301859051381543,
+      "learning_rate": 9.9659864743416e-06,
+      "loss": 0.7674,
+      "step": 1032
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4232575056074788,
+      "learning_rate": 9.965865675248077e-06,
+      "loss": 0.7626,
+      "step": 1033
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7367717102600604,
+      "learning_rate": 9.965744662760246e-06,
+      "loss": 0.7425,
+      "step": 1034
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4401724311998467,
+      "learning_rate": 9.965623436883307e-06,
+      "loss": 0.7392,
+      "step": 1035
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7379899924091982,
+      "learning_rate": 9.96550199762247e-06,
+      "loss": 0.8594,
+      "step": 1036
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1279479738552551,
+      "learning_rate": 9.965380344982955e-06,
+      "loss": 0.7062,
+      "step": 1037
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3902868181846249,
+      "learning_rate": 9.965258478969985e-06,
+      "loss": 0.7421,
+      "step": 1038
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5979537010218583,
+      "learning_rate": 9.965136399588803e-06,
+      "loss": 0.8032,
+      "step": 1039
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5579175037224913,
+      "learning_rate": 9.96501410684465e-06,
+      "loss": 0.6912,
+      "step": 1040
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.62920122554738,
+      "learning_rate": 9.964891600742785e-06,
+      "loss": 0.8981,
+      "step": 1041
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2644728768862872,
+      "learning_rate": 9.96476888128847e-06,
+      "loss": 0.623,
+      "step": 1042
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.006709092916133,
+      "learning_rate": 9.964645948486978e-06,
+      "loss": 0.818,
+      "step": 1043
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5636030492765145,
+      "learning_rate": 9.964522802343593e-06,
+      "loss": 0.8655,
+      "step": 1044
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5145524779655672,
+      "learning_rate": 9.964399442863609e-06,
+      "loss": 0.7489,
+      "step": 1045
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.425063755151717,
+      "learning_rate": 9.96427587005232e-06,
+      "loss": 0.7846,
+      "step": 1046
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1716614466486919,
+      "learning_rate": 9.964152083915045e-06,
+      "loss": 0.7275,
+      "step": 1047
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4351425321766238,
+      "learning_rate": 9.964028084457098e-06,
+      "loss": 0.7181,
+      "step": 1048
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.678895887566619,
+      "learning_rate": 9.963903871683806e-06,
+      "loss": 0.7184,
+      "step": 1049
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7579156130649625,
+      "learning_rate": 9.963779445600512e-06,
+      "loss": 0.788,
+      "step": 1050
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.37665656494089,
+      "learning_rate": 9.963654806212561e-06,
+      "loss": 0.7602,
+      "step": 1051
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.092339200300182,
+      "learning_rate": 9.963529953525307e-06,
+      "loss": 0.6876,
+      "step": 1052
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5854751130327003,
+      "learning_rate": 9.963404887544117e-06,
+      "loss": 0.7326,
+      "step": 1053
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.545842172675303,
+      "learning_rate": 9.963279608274364e-06,
+      "loss": 0.8108,
+      "step": 1054
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5764641908985573,
+      "learning_rate": 9.963154115721433e-06,
+      "loss": 0.7991,
+      "step": 1055
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.32816417621882,
+      "learning_rate": 9.963028409890716e-06,
+      "loss": 0.6682,
+      "step": 1056
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3804199532239938,
+      "learning_rate": 9.962902490787616e-06,
+      "loss": 0.6923,
+      "step": 1057
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.488327493016544,
+      "learning_rate": 9.962776358417543e-06,
+      "loss": 0.7973,
+      "step": 1058
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.415938158153086,
+      "learning_rate": 9.962650012785917e-06,
+      "loss": 0.7966,
+      "step": 1059
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5550933818239903,
+      "learning_rate": 9.962523453898166e-06,
+      "loss": 0.796,
+      "step": 1060
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7345691691302696,
+      "learning_rate": 9.96239668175973e-06,
+      "loss": 0.8071,
+      "step": 1061
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2700530855703283,
+      "learning_rate": 9.962269696376059e-06,
+      "loss": 0.6529,
+      "step": 1062
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6269259548969104,
+      "learning_rate": 9.962142497752605e-06,
+      "loss": 0.8671,
+      "step": 1063
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7645633234847893,
+      "learning_rate": 9.962015085894838e-06,
+      "loss": 0.8162,
+      "step": 1064
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3891695860575586,
+      "learning_rate": 9.961887460808232e-06,
+      "loss": 0.675,
+      "step": 1065
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3283754808653299,
+      "learning_rate": 9.96175962249827e-06,
+      "loss": 0.7186,
+      "step": 1066
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3700273076508402,
+      "learning_rate": 9.961631570970449e-06,
+      "loss": 0.7955,
+      "step": 1067
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4899604545790617,
+      "learning_rate": 9.961503306230267e-06,
+      "loss": 0.7321,
+      "step": 1068
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5065192731637094,
+      "learning_rate": 9.961374828283239e-06,
+      "loss": 0.8206,
+      "step": 1069
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.578397065698884,
+      "learning_rate": 9.961246137134883e-06,
+      "loss": 0.6845,
+      "step": 1070
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7105374991279036,
+      "learning_rate": 9.961117232790734e-06,
+      "loss": 0.8154,
+      "step": 1071
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.419205634188292,
+      "learning_rate": 9.960988115256328e-06,
+      "loss": 0.7218,
+      "step": 1072
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5542302138893205,
+      "learning_rate": 9.960858784537213e-06,
+      "loss": 0.7048,
+      "step": 1073
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.477886412986774,
+      "learning_rate": 9.960729240638947e-06,
+      "loss": 0.7585,
+      "step": 1074
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4384093807304728,
+      "learning_rate": 9.9605994835671e-06,
+      "loss": 0.7269,
+      "step": 1075
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4924457779262743,
+      "learning_rate": 9.960469513327242e-06,
+      "loss": 0.7764,
+      "step": 1076
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5045799656795746,
+      "learning_rate": 9.960339329924964e-06,
+      "loss": 0.7792,
+      "step": 1077
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.458909529329674,
+      "learning_rate": 9.960208933365857e-06,
+      "loss": 0.7025,
+      "step": 1078
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3966914234648968,
+      "learning_rate": 9.960078323655524e-06,
+      "loss": 0.7903,
+      "step": 1079
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6111122236304771,
+      "learning_rate": 9.95994750079958e-06,
+      "loss": 0.8498,
+      "step": 1080
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3088995980101938,
+      "learning_rate": 9.959816464803647e-06,
+      "loss": 0.6804,
+      "step": 1081
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5485260619438657,
+      "learning_rate": 9.959685215673351e-06,
+      "loss": 0.8818,
+      "step": 1082
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4930459173390669,
+      "learning_rate": 9.959553753414338e-06,
+      "loss": 0.8314,
+      "step": 1083
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.707614690506651,
+      "learning_rate": 9.959422078032253e-06,
+      "loss": 0.628,
+      "step": 1084
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3903937552681864,
+      "learning_rate": 9.959290189532757e-06,
+      "loss": 0.6621,
+      "step": 1085
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.855445294270785,
+      "learning_rate": 9.959158087921516e-06,
+      "loss": 0.7725,
+      "step": 1086
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7326595321155287,
+      "learning_rate": 9.95902577320421e-06,
+      "loss": 0.8664,
+      "step": 1087
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3291417453160799,
+      "learning_rate": 9.95889324538652e-06,
+      "loss": 0.7555,
+      "step": 1088
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4294065913334262,
+      "learning_rate": 9.958760504474144e-06,
+      "loss": 0.734,
+      "step": 1089
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3947381188780459,
+      "learning_rate": 9.958627550472784e-06,
+      "loss": 0.6902,
+      "step": 1090
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2876560777795787,
+      "learning_rate": 9.958494383388155e-06,
+      "loss": 0.6635,
+      "step": 1091
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5109206942746114,
+      "learning_rate": 9.958361003225979e-06,
+      "loss": 0.8023,
+      "step": 1092
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.902434123335468,
+      "learning_rate": 9.958227409991988e-06,
+      "loss": 0.7539,
+      "step": 1093
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.474934971298451,
+      "learning_rate": 9.958093603691923e-06,
+      "loss": 0.8117,
+      "step": 1094
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4352357992211562,
+      "learning_rate": 9.957959584331533e-06,
+      "loss": 0.8105,
+      "step": 1095
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5532402911513472,
+      "learning_rate": 9.957825351916579e-06,
+      "loss": 0.842,
+      "step": 1096
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2931418465431734,
+      "learning_rate": 9.957690906452827e-06,
+      "loss": 0.6864,
+      "step": 1097
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3487154197414588,
+      "learning_rate": 9.957556247946057e-06,
+      "loss": 0.6893,
+      "step": 1098
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.554784531707466,
+      "learning_rate": 9.957421376402053e-06,
+      "loss": 0.8138,
+      "step": 1099
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4122561019435116,
+      "learning_rate": 9.95728629182661e-06,
+      "loss": 0.7703,
+      "step": 1100
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5480733606681705,
+      "learning_rate": 9.957150994225535e-06,
+      "loss": 0.7545,
+      "step": 1101
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3056409906908368,
+      "learning_rate": 9.957015483604643e-06,
+      "loss": 0.7525,
+      "step": 1102
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5494519073725135,
+      "learning_rate": 9.956879759969755e-06,
+      "loss": 0.75,
+      "step": 1103
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5246476003219893,
+      "learning_rate": 9.956743823326704e-06,
+      "loss": 0.7207,
+      "step": 1104
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.375655688695232,
+      "learning_rate": 9.956607673681334e-06,
+      "loss": 0.7893,
+      "step": 1105
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7187655090364773,
+      "learning_rate": 9.956471311039491e-06,
+      "loss": 0.8689,
+      "step": 1106
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3750326188659447,
+      "learning_rate": 9.956334735407037e-06,
+      "loss": 0.8704,
+      "step": 1107
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4542550094995488,
+      "learning_rate": 9.95619794678984e-06,
+      "loss": 0.8567,
+      "step": 1108
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.306482188105742,
+      "learning_rate": 9.956060945193781e-06,
+      "loss": 0.7612,
+      "step": 1109
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5164092513808332,
+      "learning_rate": 9.955923730624745e-06,
+      "loss": 0.8214,
+      "step": 1110
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3817366112337555,
+      "learning_rate": 9.95578630308863e-06,
+      "loss": 0.6656,
+      "step": 1111
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6840459161432975,
+      "learning_rate": 9.95564866259134e-06,
+      "loss": 0.7798,
+      "step": 1112
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3848969875955028,
+      "learning_rate": 9.95551080913879e-06,
+      "loss": 0.76,
+      "step": 1113
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5158614908476191,
+      "learning_rate": 9.955372742736903e-06,
+      "loss": 0.7754,
+      "step": 1114
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5242943228569963,
+      "learning_rate": 9.955234463391612e-06,
+      "loss": 0.805,
+      "step": 1115
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4822494577380447,
+      "learning_rate": 9.955095971108863e-06,
+      "loss": 0.6617,
+      "step": 1116
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3607063701239954,
+      "learning_rate": 9.954957265894602e-06,
+      "loss": 0.6808,
+      "step": 1117
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7163106483919226,
+      "learning_rate": 9.954818347754795e-06,
+      "loss": 0.751,
+      "step": 1118
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2905145476933149,
+      "learning_rate": 9.954679216695406e-06,
+      "loss": 0.6896,
+      "step": 1119
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2837371398528215,
+      "learning_rate": 9.954539872722417e-06,
+      "loss": 0.7231,
+      "step": 1120
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.604039322076373,
+      "learning_rate": 9.954400315841817e-06,
+      "loss": 0.7189,
+      "step": 1121
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7018880151978377,
+      "learning_rate": 9.9542605460596e-06,
+      "loss": 0.9418,
+      "step": 1122
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3355750066259942,
+      "learning_rate": 9.954120563381772e-06,
+      "loss": 0.6279,
+      "step": 1123
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7661026982187222,
+      "learning_rate": 9.953980367814354e-06,
+      "loss": 0.6259,
+      "step": 1124
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3809831915711486,
+      "learning_rate": 9.953839959363363e-06,
+      "loss": 0.9388,
+      "step": 1125
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4697802260221948,
+      "learning_rate": 9.953699338034838e-06,
+      "loss": 0.6926,
+      "step": 1126
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5474379094469868,
+      "learning_rate": 9.953558503834819e-06,
+      "loss": 0.7441,
+      "step": 1127
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5208335698563589,
+      "learning_rate": 9.95341745676936e-06,
+      "loss": 0.7894,
+      "step": 1128
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5634526279495362,
+      "learning_rate": 9.953276196844519e-06,
+      "loss": 0.8179,
+      "step": 1129
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3966556803901218,
+      "learning_rate": 9.95313472406637e-06,
+      "loss": 0.6963,
+      "step": 1130
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.196545766894864,
+      "learning_rate": 9.95299303844099e-06,
+      "loss": 0.7751,
+      "step": 1131
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.404980448427074,
+      "learning_rate": 9.952851139974468e-06,
+      "loss": 0.7317,
+      "step": 1132
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4186318696692488,
+      "learning_rate": 9.952709028672902e-06,
+      "loss": 0.7058,
+      "step": 1133
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4811396056825972,
+      "learning_rate": 9.9525667045424e-06,
+      "loss": 0.8992,
+      "step": 1134
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4437094830750548,
+      "learning_rate": 9.952424167589076e-06,
+      "loss": 0.8493,
+      "step": 1135
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.349682230446703,
+      "learning_rate": 9.952281417819057e-06,
+      "loss": 0.7314,
+      "step": 1136
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5846939881699553,
+      "learning_rate": 9.952138455238474e-06,
+      "loss": 0.7602,
+      "step": 1137
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3742182444286741,
+      "learning_rate": 9.951995279853472e-06,
+      "loss": 0.6965,
+      "step": 1138
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3991425100802024,
+      "learning_rate": 9.951851891670206e-06,
+      "loss": 0.7535,
+      "step": 1139
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1897615371892722,
+      "learning_rate": 9.951708290694837e-06,
+      "loss": 0.6831,
+      "step": 1140
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4589345360061752,
+      "learning_rate": 9.951564476933534e-06,
+      "loss": 0.8002,
+      "step": 1141
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4520008085268288,
+      "learning_rate": 9.951420450392476e-06,
+      "loss": 0.8263,
+      "step": 1142
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4080082299606038,
+      "learning_rate": 9.951276211077854e-06,
+      "loss": 0.7173,
+      "step": 1143
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4635131845894216,
+      "learning_rate": 9.951131758995866e-06,
+      "loss": 0.7494,
+      "step": 1144
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5237822686869567,
+      "learning_rate": 9.950987094152721e-06,
+      "loss": 0.6344,
+      "step": 1145
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5063410611942785,
+      "learning_rate": 9.950842216554633e-06,
+      "loss": 0.8463,
+      "step": 1146
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4644020446892363,
+      "learning_rate": 9.95069712620783e-06,
+      "loss": 0.7253,
+      "step": 1147
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0744151093151246,
+      "learning_rate": 9.950551823118544e-06,
+      "loss": 0.9038,
+      "step": 1148
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5890623507396244,
+      "learning_rate": 9.950406307293023e-06,
+      "loss": 0.7377,
+      "step": 1149
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5013689452603372,
+      "learning_rate": 9.950260578737517e-06,
+      "loss": 0.8752,
+      "step": 1150
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.416939068893771,
+      "learning_rate": 9.950114637458288e-06,
+      "loss": 0.7304,
+      "step": 1151
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6634206013589024,
+      "learning_rate": 9.949968483461609e-06,
+      "loss": 0.7885,
+      "step": 1152
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5095279764309433,
+      "learning_rate": 9.94982211675376e-06,
+      "loss": 0.8309,
+      "step": 1153
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1296342611040513,
+      "learning_rate": 9.949675537341031e-06,
+      "loss": 0.657,
+      "step": 1154
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5100447273139974,
+      "learning_rate": 9.949528745229721e-06,
+      "loss": 0.678,
+      "step": 1155
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4297065964648237,
+      "learning_rate": 9.949381740426139e-06,
+      "loss": 0.7516,
+      "step": 1156
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3025261544188496,
+      "learning_rate": 9.949234522936597e-06,
+      "loss": 0.7641,
+      "step": 1157
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4523821078564112,
+      "learning_rate": 9.949087092767428e-06,
+      "loss": 0.6678,
+      "step": 1158
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3696761096146903,
+      "learning_rate": 9.948939449924964e-06,
+      "loss": 0.7515,
+      "step": 1159
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4946282634770574,
+      "learning_rate": 9.94879159441555e-06,
+      "loss": 0.7707,
+      "step": 1160
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.375676817827109,
+      "learning_rate": 9.948643526245538e-06,
+      "loss": 0.7629,
+      "step": 1161
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5009291175174952,
+      "learning_rate": 9.948495245421294e-06,
+      "loss": 0.7127,
+      "step": 1162
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6144360444750356,
+      "learning_rate": 9.948346751949188e-06,
+      "loss": 0.718,
+      "step": 1163
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2581506205689483,
+      "learning_rate": 9.948198045835601e-06,
+      "loss": 0.6135,
+      "step": 1164
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5727628771314688,
+      "learning_rate": 9.948049127086925e-06,
+      "loss": 0.8349,
+      "step": 1165
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0071865763849315,
+      "learning_rate": 9.947899995709556e-06,
+      "loss": 0.7034,
+      "step": 1166
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6206688551042063,
+      "learning_rate": 9.947750651709907e-06,
+      "loss": 0.8559,
+      "step": 1167
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3135114331091284,
+      "learning_rate": 9.947601095094392e-06,
+      "loss": 0.6534,
+      "step": 1168
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2786743025136194,
+      "learning_rate": 9.94745132586944e-06,
+      "loss": 0.6605,
+      "step": 1169
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5541663139649207,
+      "learning_rate": 9.947301344041485e-06,
+      "loss": 0.8613,
+      "step": 1170
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5605883374504632,
+      "learning_rate": 9.947151149616972e-06,
+      "loss": 0.8673,
+      "step": 1171
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6038296646222074,
+      "learning_rate": 9.947000742602358e-06,
+      "loss": 0.7745,
+      "step": 1172
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.603381615235763,
+      "learning_rate": 9.946850123004103e-06,
+      "loss": 0.727,
+      "step": 1173
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3395687815578958,
+      "learning_rate": 9.946699290828683e-06,
+      "loss": 0.7114,
+      "step": 1174
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4912752725858651,
+      "learning_rate": 9.946548246082577e-06,
+      "loss": 0.7201,
+      "step": 1175
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2951998172760644,
+      "learning_rate": 9.946396988772275e-06,
+      "loss": 0.7381,
+      "step": 1176
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.466891782837542,
+      "learning_rate": 9.94624551890428e-06,
+      "loss": 0.8777,
+      "step": 1177
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4502186569023026,
+      "learning_rate": 9.946093836485097e-06,
+      "loss": 0.7543,
+      "step": 1178
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4072127482251433,
+      "learning_rate": 9.94594194152125e-06,
+      "loss": 0.654,
+      "step": 1179
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2457677814558674,
+      "learning_rate": 9.945789834019259e-06,
+      "loss": 0.7502,
+      "step": 1180
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6496618628364104,
+      "learning_rate": 9.945637513985664e-06,
+      "loss": 0.9341,
+      "step": 1181
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5441502736444397,
+      "learning_rate": 9.945484981427011e-06,
+      "loss": 0.7946,
+      "step": 1182
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6214910299789467,
+      "learning_rate": 9.945332236349857e-06,
+      "loss": 0.8424,
+      "step": 1183
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2441035188412262,
+      "learning_rate": 9.945179278760759e-06,
+      "loss": 0.6399,
+      "step": 1184
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1605032763823027,
+      "learning_rate": 9.945026108666296e-06,
+      "loss": 0.7218,
+      "step": 1185
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4267648781996694,
+      "learning_rate": 9.944872726073048e-06,
+      "loss": 0.6613,
+      "step": 1186
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4772323067124995,
+      "learning_rate": 9.944719130987606e-06,
+      "loss": 0.8182,
+      "step": 1187
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4511279238043988,
+      "learning_rate": 9.94456532341657e-06,
+      "loss": 0.8135,
+      "step": 1188
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.398449345376766,
+      "learning_rate": 9.94441130336655e-06,
+      "loss": 0.7591,
+      "step": 1189
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5970628868260213,
+      "learning_rate": 9.944257070844165e-06,
+      "loss": 0.7989,
+      "step": 1190
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4343256275384757,
+      "learning_rate": 9.944102625856042e-06,
+      "loss": 0.7372,
+      "step": 1191
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.415193500178158,
+      "learning_rate": 9.943947968408816e-06,
+      "loss": 0.6513,
+      "step": 1192
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4184196287155193,
+      "learning_rate": 9.943793098509138e-06,
+      "loss": 0.7506,
+      "step": 1193
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7350800440305925,
+      "learning_rate": 9.943638016163658e-06,
+      "loss": 0.7047,
+      "step": 1194
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3849197507605642,
+      "learning_rate": 9.943482721379045e-06,
+      "loss": 0.7113,
+      "step": 1195
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5350209154193313,
+      "learning_rate": 9.943327214161968e-06,
+      "loss": 0.6994,
+      "step": 1196
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5494641904853668,
+      "learning_rate": 9.943171494519111e-06,
+      "loss": 0.8377,
+      "step": 1197
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4615381565970362,
+      "learning_rate": 9.943015562457166e-06,
+      "loss": 0.6978,
+      "step": 1198
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2112985276183914,
+      "learning_rate": 9.942859417982833e-06,
+      "loss": 0.6072,
+      "step": 1199
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0904782566566409,
+      "learning_rate": 9.942703061102823e-06,
+      "loss": 0.7209,
+      "step": 1200
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5605859757391725,
+      "learning_rate": 9.942546491823856e-06,
+      "loss": 0.7599,
+      "step": 1201
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5383649448940293,
+      "learning_rate": 9.942389710152657e-06,
+      "loss": 0.8444,
+      "step": 1202
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4943593073895745,
+      "learning_rate": 9.942232716095965e-06,
+      "loss": 0.8618,
+      "step": 1203
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8506653655187,
+      "learning_rate": 9.942075509660527e-06,
+      "loss": 0.7547,
+      "step": 1204
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3986247457396719,
+      "learning_rate": 9.941918090853096e-06,
+      "loss": 0.8071,
+      "step": 1205
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1555964934950818,
+      "learning_rate": 9.941760459680441e-06,
+      "loss": 0.7143,
+      "step": 1206
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.169446113731559,
+      "learning_rate": 9.94160261614933e-06,
+      "loss": 0.6989,
+      "step": 1207
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3645866382059917,
+      "learning_rate": 9.941444560266551e-06,
+      "loss": 0.7745,
+      "step": 1208
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3612001564552119,
+      "learning_rate": 9.941286292038894e-06,
+      "loss": 0.6421,
+      "step": 1209
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3546493627864729,
+      "learning_rate": 9.941127811473162e-06,
+      "loss": 0.8224,
+      "step": 1210
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2346326351865067,
+      "learning_rate": 9.94096911857616e-06,
+      "loss": 0.7064,
+      "step": 1211
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0944971055327692,
+      "learning_rate": 9.940810213354713e-06,
+      "loss": 0.75,
+      "step": 1212
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3277713093994177,
+      "learning_rate": 9.940651095815646e-06,
+      "loss": 0.7047,
+      "step": 1213
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5115439900555994,
+      "learning_rate": 9.940491765965798e-06,
+      "loss": 0.8251,
+      "step": 1214
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5032567946718058,
+      "learning_rate": 9.940332223812017e-06,
+      "loss": 0.8021,
+      "step": 1215
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3670371503867538,
+      "learning_rate": 9.940172469361158e-06,
+      "loss": 0.7207,
+      "step": 1216
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.226334740182117,
+      "learning_rate": 9.940012502620085e-06,
+      "loss": 0.6881,
+      "step": 1217
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4116254417965337,
+      "learning_rate": 9.939852323595671e-06,
+      "loss": 0.7273,
+      "step": 1218
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.033009054771052,
+      "learning_rate": 9.939691932294804e-06,
+      "loss": 0.6233,
+      "step": 1219
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4666353179555804,
+      "learning_rate": 9.939531328724371e-06,
+      "loss": 0.8017,
+      "step": 1220
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1559997248193212,
+      "learning_rate": 9.939370512891277e-06,
+      "loss": 0.6837,
+      "step": 1221
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.321561061643898,
+      "learning_rate": 9.93920948480243e-06,
+      "loss": 0.7055,
+      "step": 1222
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5291548803246706,
+      "learning_rate": 9.939048244464753e-06,
+      "loss": 0.6898,
+      "step": 1223
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6060845097102512,
+      "learning_rate": 9.938886791885172e-06,
+      "loss": 0.6989,
+      "step": 1224
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3475677852731536,
+      "learning_rate": 9.938725127070628e-06,
+      "loss": 0.6223,
+      "step": 1225
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.576626677556507,
+      "learning_rate": 9.938563250028063e-06,
+      "loss": 0.7954,
+      "step": 1226
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4284271827505064,
+      "learning_rate": 9.93840116076444e-06,
+      "loss": 0.7735,
+      "step": 1227
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5029368650641246,
+      "learning_rate": 9.938238859286719e-06,
+      "loss": 0.7733,
+      "step": 1228
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6240607455490152,
+      "learning_rate": 9.938076345601875e-06,
+      "loss": 0.6273,
+      "step": 1229
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.456989513524123,
+      "learning_rate": 9.937913619716895e-06,
+      "loss": 0.7179,
+      "step": 1230
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.250362262195705,
+      "learning_rate": 9.937750681638766e-06,
+      "loss": 0.7491,
+      "step": 1231
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3778123778913947,
+      "learning_rate": 9.937587531374497e-06,
+      "loss": 0.7154,
+      "step": 1232
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.518737814535401,
+      "learning_rate": 9.937424168931094e-06,
+      "loss": 0.7231,
+      "step": 1233
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7115670183614968,
+      "learning_rate": 9.937260594315578e-06,
+      "loss": 0.6859,
+      "step": 1234
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.391614814428079,
+      "learning_rate": 9.937096807534978e-06,
+      "loss": 0.6886,
+      "step": 1235
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4456584456376518,
+      "learning_rate": 9.936932808596332e-06,
+      "loss": 0.7167,
+      "step": 1236
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2070024709554565,
+      "learning_rate": 9.936768597506689e-06,
+      "loss": 0.733,
+      "step": 1237
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5503176872578672,
+      "learning_rate": 9.936604174273105e-06,
+      "loss": 0.718,
+      "step": 1238
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5118981745501232,
+      "learning_rate": 9.936439538902644e-06,
+      "loss": 0.8791,
+      "step": 1239
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5362310801551962,
+      "learning_rate": 9.936274691402384e-06,
+      "loss": 0.7463,
+      "step": 1240
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7190425112533456,
+      "learning_rate": 9.936109631779404e-06,
+      "loss": 0.7879,
+      "step": 1241
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6884145414477008,
+      "learning_rate": 9.935944360040803e-06,
+      "loss": 0.9981,
+      "step": 1242
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.419959298430799,
+      "learning_rate": 9.935778876193678e-06,
+      "loss": 0.8183,
+      "step": 1243
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.74946350653532,
+      "learning_rate": 9.935613180245143e-06,
+      "loss": 0.7981,
+      "step": 1244
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5863845361952928,
+      "learning_rate": 9.935447272202317e-06,
+      "loss": 0.7303,
+      "step": 1245
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.405945079657888,
+      "learning_rate": 9.935281152072329e-06,
+      "loss": 0.78,
+      "step": 1246
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2594486569291925,
+      "learning_rate": 9.93511481986232e-06,
+      "loss": 0.7553,
+      "step": 1247
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5634695851624107,
+      "learning_rate": 9.934948275579435e-06,
+      "loss": 0.7623,
+      "step": 1248
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2008909079783674,
+      "learning_rate": 9.934781519230832e-06,
+      "loss": 0.6294,
+      "step": 1249
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3415242269458851,
+      "learning_rate": 9.934614550823679e-06,
+      "loss": 0.7451,
+      "step": 1250
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1708770147731706,
+      "learning_rate": 9.934447370365146e-06,
+      "loss": 0.8428,
+      "step": 1251
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.68321456363142,
+      "learning_rate": 9.93427997786242e-06,
+      "loss": 0.7023,
+      "step": 1252
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3128688044676453,
+      "learning_rate": 9.934112373322695e-06,
+      "loss": 0.6547,
+      "step": 1253
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4719319342805528,
+      "learning_rate": 9.933944556753173e-06,
+      "loss": 0.7597,
+      "step": 1254
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4431448083514675,
+      "learning_rate": 9.933776528161063e-06,
+      "loss": 0.768,
+      "step": 1255
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1886299272286354,
+      "learning_rate": 9.933608287553589e-06,
+      "loss": 0.6614,
+      "step": 1256
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.438852841321342,
+      "learning_rate": 9.93343983493798e-06,
+      "loss": 0.6863,
+      "step": 1257
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2908367115027735,
+      "learning_rate": 9.93327117032147e-06,
+      "loss": 0.725,
+      "step": 1258
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.039891054585343,
+      "learning_rate": 9.933102293711314e-06,
+      "loss": 0.7613,
+      "step": 1259
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3317836664226064,
+      "learning_rate": 9.932933205114766e-06,
+      "loss": 0.6118,
+      "step": 1260
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5046340197575347,
+      "learning_rate": 9.932763904539091e-06,
+      "loss": 0.731,
+      "step": 1261
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3280505939147804,
+      "learning_rate": 9.932594391991566e-06,
+      "loss": 0.6486,
+      "step": 1262
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6643405284578874,
+      "learning_rate": 9.932424667479475e-06,
+      "loss": 0.775,
+      "step": 1263
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.727070335988034,
+      "learning_rate": 9.932254731010108e-06,
+      "loss": 0.852,
+      "step": 1264
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9319346180057095,
+      "learning_rate": 9.932084582590773e-06,
+      "loss": 0.7722,
+      "step": 1265
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2880004470782216,
+      "learning_rate": 9.93191422222878e-06,
+      "loss": 0.7538,
+      "step": 1266
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2182675301880632,
+      "learning_rate": 9.931743649931446e-06,
+      "loss": 0.7365,
+      "step": 1267
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3539494729784372,
+      "learning_rate": 9.931572865706105e-06,
+      "loss": 0.7778,
+      "step": 1268
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4588321512405353,
+      "learning_rate": 9.931401869560096e-06,
+      "loss": 0.7676,
+      "step": 1269
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5251663609857837,
+      "learning_rate": 9.931230661500764e-06,
+      "loss": 0.8814,
+      "step": 1270
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3879280871710948,
+      "learning_rate": 9.93105924153547e-06,
+      "loss": 0.7782,
+      "step": 1271
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2728427925847872,
+      "learning_rate": 9.930887609671577e-06,
+      "loss": 0.6956,
+      "step": 1272
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3876397639273834,
+      "learning_rate": 9.930715765916462e-06,
+      "loss": 0.7327,
+      "step": 1273
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3389202144495542,
+      "learning_rate": 9.93054371027751e-06,
+      "loss": 0.7474,
+      "step": 1274
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.575708961283556,
+      "learning_rate": 9.930371442762112e-06,
+      "loss": 0.8151,
+      "step": 1275
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5464237512456327,
+      "learning_rate": 9.930198963377672e-06,
+      "loss": 0.7888,
+      "step": 1276
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.568100399915259,
+      "learning_rate": 9.930026272131605e-06,
+      "loss": 0.7236,
+      "step": 1277
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4923540379295375,
+      "learning_rate": 9.929853369031328e-06,
+      "loss": 0.6689,
+      "step": 1278
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4104731187402653,
+      "learning_rate": 9.929680254084273e-06,
+      "loss": 0.7871,
+      "step": 1279
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5900208444168236,
+      "learning_rate": 9.929506927297878e-06,
+      "loss": 0.7999,
+      "step": 1280
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.12231226239516,
+      "learning_rate": 9.929333388679593e-06,
+      "loss": 0.7827,
+      "step": 1281
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3130411282701575,
+      "learning_rate": 9.929159638236872e-06,
+      "loss": 0.7106,
+      "step": 1282
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.182954336327062,
+      "learning_rate": 9.928985675977185e-06,
+      "loss": 0.6137,
+      "step": 1283
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.332324360261561,
+      "learning_rate": 9.928811501908006e-06,
+      "loss": 0.7461,
+      "step": 1284
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2754785037283554,
+      "learning_rate": 9.928637116036818e-06,
+      "loss": 0.7175,
+      "step": 1285
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4326593819861055,
+      "learning_rate": 9.928462518371119e-06,
+      "loss": 0.7627,
+      "step": 1286
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4929423241623043,
+      "learning_rate": 9.928287708918408e-06,
+      "loss": 0.7755,
+      "step": 1287
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1378918279543264,
+      "learning_rate": 9.928112687686197e-06,
+      "loss": 0.6887,
+      "step": 1288
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3326866893637195,
+      "learning_rate": 9.92793745468201e-06,
+      "loss": 0.6988,
+      "step": 1289
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.470407384519956,
+      "learning_rate": 9.927762009913377e-06,
+      "loss": 0.7253,
+      "step": 1290
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.679735323869398,
+      "learning_rate": 9.927586353387832e-06,
+      "loss": 0.8294,
+      "step": 1291
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4065774178713741,
+      "learning_rate": 9.92741048511293e-06,
+      "loss": 0.8278,
+      "step": 1292
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4518861131290117,
+      "learning_rate": 9.927234405096224e-06,
+      "loss": 0.7343,
+      "step": 1293
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3648515647793065,
+      "learning_rate": 9.927058113345282e-06,
+      "loss": 0.7507,
+      "step": 1294
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4833628078857088,
+      "learning_rate": 9.92688160986768e-06,
+      "loss": 0.5854,
+      "step": 1295
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5457349860912066,
+      "learning_rate": 9.926704894671004e-06,
+      "loss": 0.8442,
+      "step": 1296
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3364137303301498,
+      "learning_rate": 9.926527967762845e-06,
+      "loss": 0.6616,
+      "step": 1297
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.132273777833393,
+      "learning_rate": 9.926350829150806e-06,
+      "loss": 0.7396,
+      "step": 1298
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.476489135341046,
+      "learning_rate": 9.926173478842502e-06,
+      "loss": 0.7484,
+      "step": 1299
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3348330857147594,
+      "learning_rate": 9.925995916845553e-06,
+      "loss": 0.6218,
+      "step": 1300
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3521164691904546,
+      "learning_rate": 9.925818143167587e-06,
+      "loss": 0.6468,
+      "step": 1301
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.589307254402672,
+      "learning_rate": 9.925640157816246e-06,
+      "loss": 0.9114,
+      "step": 1302
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5296485739150254,
+      "learning_rate": 9.925461960799177e-06,
+      "loss": 0.8161,
+      "step": 1303
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7616934572499887,
+      "learning_rate": 9.925283552124039e-06,
+      "loss": 0.6515,
+      "step": 1304
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.389055456377223,
+      "learning_rate": 9.925104931798496e-06,
+      "loss": 0.7179,
+      "step": 1305
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3048832743046872,
+      "learning_rate": 9.924926099830226e-06,
+      "loss": 0.62,
+      "step": 1306
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4715951596055823,
+      "learning_rate": 9.924747056226913e-06,
+      "loss": 0.8156,
+      "step": 1307
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9739088315436646,
+      "learning_rate": 9.92456780099625e-06,
+      "loss": 0.8777,
+      "step": 1308
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2966900971817468,
+      "learning_rate": 9.924388334145943e-06,
+      "loss": 0.8029,
+      "step": 1309
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1184989145999689,
+      "learning_rate": 9.9242086556837e-06,
+      "loss": 0.6399,
+      "step": 1310
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4521449473035042,
+      "learning_rate": 9.924028765617247e-06,
+      "loss": 0.7215,
+      "step": 1311
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6069579176802717,
+      "learning_rate": 9.92384866395431e-06,
+      "loss": 0.728,
+      "step": 1312
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2919521174350943,
+      "learning_rate": 9.923668350702628e-06,
+      "loss": 0.6779,
+      "step": 1313
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2621966546734387,
+      "learning_rate": 9.923487825869955e-06,
+      "loss": 0.7114,
+      "step": 1314
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3863658409840676,
+      "learning_rate": 9.923307089464043e-06,
+      "loss": 0.7663,
+      "step": 1315
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1427702298249325,
+      "learning_rate": 9.92312614149266e-06,
+      "loss": 0.695,
+      "step": 1316
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.29097115068362,
+      "learning_rate": 9.922944981963583e-06,
+      "loss": 0.5983,
+      "step": 1317
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4261076064319356,
+      "learning_rate": 9.922763610884596e-06,
+      "loss": 0.7712,
+      "step": 1318
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2581345209876282,
+      "learning_rate": 9.922582028263495e-06,
+      "loss": 0.6937,
+      "step": 1319
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3589482081524868,
+      "learning_rate": 9.922400234108079e-06,
+      "loss": 0.75,
+      "step": 1320
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3255554814151895,
+      "learning_rate": 9.922218228426162e-06,
+      "loss": 0.7262,
+      "step": 1321
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2407432579567053,
+      "learning_rate": 9.922036011225564e-06,
+      "loss": 0.5898,
+      "step": 1322
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3709663795035236,
+      "learning_rate": 9.92185358251412e-06,
+      "loss": 0.7101,
+      "step": 1323
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.407523963785978,
+      "learning_rate": 9.921670942299664e-06,
+      "loss": 0.8192,
+      "step": 1324
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4841437438213598,
+      "learning_rate": 9.921488090590047e-06,
+      "loss": 0.7726,
+      "step": 1325
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2955109880000522,
+      "learning_rate": 9.921305027393125e-06,
+      "loss": 0.5896,
+      "step": 1326
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3872149074563789,
+      "learning_rate": 9.921121752716767e-06,
+      "loss": 0.7334,
+      "step": 1327
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2511829070493181,
+      "learning_rate": 9.920938266568847e-06,
+      "loss": 0.6929,
+      "step": 1328
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1972858522910506,
+      "learning_rate": 9.92075456895725e-06,
+      "loss": 0.6682,
+      "step": 1329
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3366448369960078,
+      "learning_rate": 9.92057065988987e-06,
+      "loss": 0.6696,
+      "step": 1330
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2426443090781887,
+      "learning_rate": 9.92038653937461e-06,
+      "loss": 0.6772,
+      "step": 1331
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2620667940302575,
+      "learning_rate": 9.920202207419382e-06,
+      "loss": 0.5725,
+      "step": 1332
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3279056684228425,
+      "learning_rate": 9.920017664032109e-06,
+      "loss": 0.7439,
+      "step": 1333
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.513852315889344,
+      "learning_rate": 9.919832909220717e-06,
+      "loss": 0.8287,
+      "step": 1334
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.297171639094171,
+      "learning_rate": 9.91964794299315e-06,
+      "loss": 0.7809,
+      "step": 1335
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4530110484134477,
+      "learning_rate": 9.919462765357353e-06,
+      "loss": 0.6698,
+      "step": 1336
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2562950849895538,
+      "learning_rate": 9.919277376321284e-06,
+      "loss": 0.8173,
+      "step": 1337
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0920980563414584,
+      "learning_rate": 9.91909177589291e-06,
+      "loss": 0.6574,
+      "step": 1338
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8957431349666582,
+      "learning_rate": 9.91890596408021e-06,
+      "loss": 0.6834,
+      "step": 1339
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5622446218441264,
+      "learning_rate": 9.918719940891165e-06,
+      "loss": 0.8228,
+      "step": 1340
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.459086809766253,
+      "learning_rate": 9.918533706333767e-06,
+      "loss": 0.698,
+      "step": 1341
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3301302873460614,
+      "learning_rate": 9.918347260416024e-06,
+      "loss": 0.6943,
+      "step": 1342
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4642754059646756,
+      "learning_rate": 9.918160603145943e-06,
+      "loss": 0.8006,
+      "step": 1343
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3991732791252074,
+      "learning_rate": 9.917973734531549e-06,
+      "loss": 0.7892,
+      "step": 1344
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6433933189805496,
+      "learning_rate": 9.917786654580872e-06,
+      "loss": 0.8342,
+      "step": 1345
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.651507214511156,
+      "learning_rate": 9.917599363301947e-06,
+      "loss": 0.708,
+      "step": 1346
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2422621274328043,
+      "learning_rate": 9.917411860702828e-06,
+      "loss": 0.6073,
+      "step": 1347
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.15270767942106,
+      "learning_rate": 9.917224146791568e-06,
+      "loss": 0.6762,
+      "step": 1348
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4409216805336815,
+      "learning_rate": 9.917036221576235e-06,
+      "loss": 0.7639,
+      "step": 1349
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2426596734990003,
+      "learning_rate": 9.916848085064906e-06,
+      "loss": 0.7087,
+      "step": 1350
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4429490395566662,
+      "learning_rate": 9.916659737265664e-06,
+      "loss": 0.7619,
+      "step": 1351
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4960939607651498,
+      "learning_rate": 9.916471178186602e-06,
+      "loss": 0.8682,
+      "step": 1352
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.021951037895212,
+      "learning_rate": 9.916282407835826e-06,
+      "loss": 0.5758,
+      "step": 1353
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2892712500208534,
+      "learning_rate": 9.916093426221445e-06,
+      "loss": 0.7318,
+      "step": 1354
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5781359519489961,
+      "learning_rate": 9.91590423335158e-06,
+      "loss": 0.6762,
+      "step": 1355
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3995866587520818,
+      "learning_rate": 9.915714829234361e-06,
+      "loss": 0.8286,
+      "step": 1356
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5779717854263557,
+      "learning_rate": 9.91552521387793e-06,
+      "loss": 0.6942,
+      "step": 1357
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2685737076278494,
+      "learning_rate": 9.915335387290432e-06,
+      "loss": 0.6394,
+      "step": 1358
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4234674460288028,
+      "learning_rate": 9.915145349480027e-06,
+      "loss": 0.9025,
+      "step": 1359
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3807263320399543,
+      "learning_rate": 9.914955100454878e-06,
+      "loss": 0.7692,
+      "step": 1360
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4544625819923187,
+      "learning_rate": 9.914764640223162e-06,
+      "loss": 0.7897,
+      "step": 1361
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6801527027077272,
+      "learning_rate": 9.914573968793064e-06,
+      "loss": 0.8941,
+      "step": 1362
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4170686819554466,
+      "learning_rate": 9.914383086172778e-06,
+      "loss": 0.7731,
+      "step": 1363
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.335096195682461,
+      "learning_rate": 9.914191992370504e-06,
+      "loss": 0.6679,
+      "step": 1364
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.454596396590096,
+      "learning_rate": 9.914000687394457e-06,
+      "loss": 0.8572,
+      "step": 1365
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3266762289659528,
+      "learning_rate": 9.913809171252856e-06,
+      "loss": 0.7604,
+      "step": 1366
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3785977354362573,
+      "learning_rate": 9.913617443953931e-06,
+      "loss": 0.7368,
+      "step": 1367
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.416507845506652,
+      "learning_rate": 9.913425505505923e-06,
+      "loss": 0.6983,
+      "step": 1368
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5658508400934787,
+      "learning_rate": 9.913233355917075e-06,
+      "loss": 0.7333,
+      "step": 1369
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4264228844371916,
+      "learning_rate": 9.913040995195651e-06,
+      "loss": 0.7663,
+      "step": 1370
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.633675858110399,
+      "learning_rate": 9.91284842334991e-06,
+      "loss": 0.8298,
+      "step": 1371
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4797036738389568,
+      "learning_rate": 9.912655640388134e-06,
+      "loss": 0.7835,
+      "step": 1372
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2693454040477172,
+      "learning_rate": 9.912462646318602e-06,
+      "loss": 0.7829,
+      "step": 1373
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4289645378345803,
+      "learning_rate": 9.91226944114961e-06,
+      "loss": 0.7234,
+      "step": 1374
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4455607919981495,
+      "learning_rate": 9.91207602488946e-06,
+      "loss": 0.7573,
+      "step": 1375
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6223291795135275,
+      "learning_rate": 9.911882397546464e-06,
+      "loss": 0.7729,
+      "step": 1376
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3554578327730495,
+      "learning_rate": 9.91168855912894e-06,
+      "loss": 0.7752,
+      "step": 1377
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4065602317676347,
+      "learning_rate": 9.911494509645222e-06,
+      "loss": 0.7287,
+      "step": 1378
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4405620402100734,
+      "learning_rate": 9.911300249103646e-06,
+      "loss": 0.6787,
+      "step": 1379
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3146856379347127,
+      "learning_rate": 9.91110577751256e-06,
+      "loss": 0.7081,
+      "step": 1380
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4569106578833069,
+      "learning_rate": 9.91091109488032e-06,
+      "loss": 0.7223,
+      "step": 1381
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5972392728755185,
+      "learning_rate": 9.910716201215294e-06,
+      "loss": 0.8073,
+      "step": 1382
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7597417440408898,
+      "learning_rate": 9.910521096525856e-06,
+      "loss": 0.7417,
+      "step": 1383
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2767616680302658,
+      "learning_rate": 9.910325780820391e-06,
+      "loss": 0.6548,
+      "step": 1384
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3944606476370005,
+      "learning_rate": 9.910130254107288e-06,
+      "loss": 0.9207,
+      "step": 1385
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.611540670308349,
+      "learning_rate": 9.909934516394957e-06,
+      "loss": 0.7478,
+      "step": 1386
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2563370205193056,
+      "learning_rate": 9.909738567691803e-06,
+      "loss": 0.7411,
+      "step": 1387
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3657620221497828,
+      "learning_rate": 9.909542408006247e-06,
+      "loss": 0.7752,
+      "step": 1388
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4022460614843948,
+      "learning_rate": 9.90934603734672e-06,
+      "loss": 0.6923,
+      "step": 1389
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3800192725152578,
+      "learning_rate": 9.909149455721661e-06,
+      "loss": 0.7262,
+      "step": 1390
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.483169663204944,
+      "learning_rate": 9.908952663139518e-06,
+      "loss": 0.7043,
+      "step": 1391
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8218021551081895,
+      "learning_rate": 9.908755659608743e-06,
+      "loss": 0.8071,
+      "step": 1392
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3490637335701112,
+      "learning_rate": 9.908558445137807e-06,
+      "loss": 0.7095,
+      "step": 1393
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3821142953236845,
+      "learning_rate": 9.908361019735181e-06,
+      "loss": 0.7986,
+      "step": 1394
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.10487817169928,
+      "learning_rate": 9.908163383409353e-06,
+      "loss": 0.8147,
+      "step": 1395
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4403411805050077,
+      "learning_rate": 9.907965536168811e-06,
+      "loss": 0.7158,
+      "step": 1396
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.119391179714961,
+      "learning_rate": 9.907767478022061e-06,
+      "loss": 0.6277,
+      "step": 1397
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6391739224404258,
+      "learning_rate": 9.90756920897761e-06,
+      "loss": 0.7567,
+      "step": 1398
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.313789712166129,
+      "learning_rate": 9.907370729043984e-06,
+      "loss": 0.5925,
+      "step": 1399
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4194301119603794,
+      "learning_rate": 9.907172038229706e-06,
+      "loss": 0.8006,
+      "step": 1400
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3325456231952648,
+      "learning_rate": 9.906973136543317e-06,
+      "loss": 0.7285,
+      "step": 1401
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.76474799364662,
+      "learning_rate": 9.906774023993362e-06,
+      "loss": 0.9363,
+      "step": 1402
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3545847314867883,
+      "learning_rate": 9.906574700588403e-06,
+      "loss": 0.85,
+      "step": 1403
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4625350555347931,
+      "learning_rate": 9.906375166336998e-06,
+      "loss": 0.7758,
+      "step": 1404
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4621611812549165,
+      "learning_rate": 9.906175421247728e-06,
+      "loss": 0.8119,
+      "step": 1405
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2220689595930014,
+      "learning_rate": 9.905975465329172e-06,
+      "loss": 0.733,
+      "step": 1406
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3272172340064163,
+      "learning_rate": 9.905775298589923e-06,
+      "loss": 0.7093,
+      "step": 1407
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5091579394091876,
+      "learning_rate": 9.905574921038586e-06,
+      "loss": 0.758,
+      "step": 1408
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4505689596546405,
+      "learning_rate": 9.905374332683768e-06,
+      "loss": 0.7439,
+      "step": 1409
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2136601346811704,
+      "learning_rate": 9.90517353353409e-06,
+      "loss": 0.6008,
+      "step": 1410
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5490958147488663,
+      "learning_rate": 9.904972523598183e-06,
+      "loss": 0.7339,
+      "step": 1411
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2947191248615995,
+      "learning_rate": 9.90477130288468e-06,
+      "loss": 0.7012,
+      "step": 1412
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3063342131113673,
+      "learning_rate": 9.904569871402233e-06,
+      "loss": 0.6706,
+      "step": 1413
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3096896216227345,
+      "learning_rate": 9.904368229159494e-06,
+      "loss": 0.6435,
+      "step": 1414
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6467255528455575,
+      "learning_rate": 9.904166376165129e-06,
+      "loss": 0.6607,
+      "step": 1415
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7700450467982736,
+      "learning_rate": 9.903964312427813e-06,
+      "loss": 0.8918,
+      "step": 1416
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5811057279306977,
+      "learning_rate": 9.903762037956228e-06,
+      "loss": 0.8014,
+      "step": 1417
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4149681985781246,
+      "learning_rate": 9.903559552759069e-06,
+      "loss": 0.7202,
+      "step": 1418
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3289718297187525,
+      "learning_rate": 9.903356856845035e-06,
+      "loss": 0.6887,
+      "step": 1419
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5317156830274028,
+      "learning_rate": 9.903153950222834e-06,
+      "loss": 0.8591,
+      "step": 1420
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2910803937260569,
+      "learning_rate": 9.90295083290119e-06,
+      "loss": 0.6871,
+      "step": 1421
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4159547205549252,
+      "learning_rate": 9.90274750488883e-06,
+      "loss": 0.8687,
+      "step": 1422
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.319302823989454,
+      "learning_rate": 9.902543966194489e-06,
+      "loss": 0.7665,
+      "step": 1423
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1806195595871098,
+      "learning_rate": 9.902340216826915e-06,
+      "loss": 0.6087,
+      "step": 1424
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4634787041664405,
+      "learning_rate": 9.902136256794865e-06,
+      "loss": 0.7561,
+      "step": 1425
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.725094377436612,
+      "learning_rate": 9.901932086107103e-06,
+      "loss": 0.833,
+      "step": 1426
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3198731669687667,
+      "learning_rate": 9.901727704772401e-06,
+      "loss": 0.7633,
+      "step": 1427
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.469278475896954,
+      "learning_rate": 9.901523112799543e-06,
+      "loss": 0.7565,
+      "step": 1428
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4785247224797673,
+      "learning_rate": 9.90131831019732e-06,
+      "loss": 0.7342,
+      "step": 1429
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3525252624569986,
+      "learning_rate": 9.901113296974533e-06,
+      "loss": 0.6946,
+      "step": 1430
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3952572630859739,
+      "learning_rate": 9.900908073139995e-06,
+      "loss": 0.6967,
+      "step": 1431
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1068000266548033,
+      "learning_rate": 9.90070263870252e-06,
+      "loss": 0.697,
+      "step": 1432
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.531577640465834,
+      "learning_rate": 9.900496993670937e-06,
+      "loss": 0.7806,
+      "step": 1433
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2589887209274164,
+      "learning_rate": 9.900291138054086e-06,
+      "loss": 0.6309,
+      "step": 1434
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4732422150586115,
+      "learning_rate": 9.90008507186081e-06,
+      "loss": 0.7302,
+      "step": 1435
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.64145257345001,
+      "learning_rate": 9.899878795099966e-06,
+      "loss": 0.6814,
+      "step": 1436
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.295957481955746,
+      "learning_rate": 9.899672307780419e-06,
+      "loss": 0.6598,
+      "step": 1437
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2529718049074012,
+      "learning_rate": 9.89946560991104e-06,
+      "loss": 0.6662,
+      "step": 1438
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5636297106605896,
+      "learning_rate": 9.899258701500712e-06,
+      "loss": 0.8581,
+      "step": 1439
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3048122674249705,
+      "learning_rate": 9.899051582558325e-06,
+      "loss": 0.6386,
+      "step": 1440
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5011985720782475,
+      "learning_rate": 9.898844253092782e-06,
+      "loss": 0.7792,
+      "step": 1441
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3102350928707627,
+      "learning_rate": 9.898636713112992e-06,
+      "loss": 0.7042,
+      "step": 1442
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.430047495925579,
+      "learning_rate": 9.89842896262787e-06,
+      "loss": 0.7588,
+      "step": 1443
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2871797144752044,
+      "learning_rate": 9.89822100164635e-06,
+      "loss": 0.794,
+      "step": 1444
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1715386066752975,
+      "learning_rate": 9.898012830177361e-06,
+      "loss": 0.6459,
+      "step": 1445
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5458492197431508,
+      "learning_rate": 9.897804448229854e-06,
+      "loss": 0.8206,
+      "step": 1446
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.470285594297649,
+      "learning_rate": 9.89759585581278e-06,
+      "loss": 0.8133,
+      "step": 1447
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4669566596391603,
+      "learning_rate": 9.897387052935106e-06,
+      "loss": 0.6693,
+      "step": 1448
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6035498289966104,
+      "learning_rate": 9.897178039605803e-06,
+      "loss": 0.9135,
+      "step": 1449
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.34592215197732,
+      "learning_rate": 9.896968815833853e-06,
+      "loss": 0.759,
+      "step": 1450
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.607125061342193,
+      "learning_rate": 9.896759381628247e-06,
+      "loss": 0.7792,
+      "step": 1451
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9842395453276144,
+      "learning_rate": 9.896549736997985e-06,
+      "loss": 0.5992,
+      "step": 1452
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3536294334952517,
+      "learning_rate": 9.896339881952076e-06,
+      "loss": 0.7126,
+      "step": 1453
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6141711543954398,
+      "learning_rate": 9.896129816499535e-06,
+      "loss": 0.7689,
+      "step": 1454
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3331200741882718,
+      "learning_rate": 9.895919540649395e-06,
+      "loss": 0.6145,
+      "step": 1455
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.512390905653172,
+      "learning_rate": 9.895709054410686e-06,
+      "loss": 0.8069,
+      "step": 1456
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9086203290624029,
+      "learning_rate": 9.895498357792456e-06,
+      "loss": 0.7666,
+      "step": 1457
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6025547684908183,
+      "learning_rate": 9.89528745080376e-06,
+      "loss": 0.714,
+      "step": 1458
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4755013236512209,
+      "learning_rate": 9.89507633345366e-06,
+      "loss": 0.7904,
+      "step": 1459
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.286472761022284,
+      "learning_rate": 9.894865005751227e-06,
+      "loss": 0.7451,
+      "step": 1460
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.569028256932131,
+      "learning_rate": 9.894653467705542e-06,
+      "loss": 0.7489,
+      "step": 1461
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.290591056680593,
+      "learning_rate": 9.894441719325698e-06,
+      "loss": 0.6986,
+      "step": 1462
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5543036078311596,
+      "learning_rate": 9.894229760620793e-06,
+      "loss": 0.7154,
+      "step": 1463
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2885785660784685,
+      "learning_rate": 9.894017591599934e-06,
+      "loss": 0.667,
+      "step": 1464
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.53165778851626,
+      "learning_rate": 9.89380521227224e-06,
+      "loss": 0.6029,
+      "step": 1465
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2775122382162996,
+      "learning_rate": 9.893592622646838e-06,
+      "loss": 0.6841,
+      "step": 1466
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4776944410254291,
+      "learning_rate": 9.893379822732863e-06,
+      "loss": 0.7072,
+      "step": 1467
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.506544733452296,
+      "learning_rate": 9.893166812539456e-06,
+      "loss": 0.6749,
+      "step": 1468
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9267190065492337,
+      "learning_rate": 9.892953592075776e-06,
+      "loss": 0.6199,
+      "step": 1469
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4689164269075332,
+      "learning_rate": 9.892740161350981e-06,
+      "loss": 0.7903,
+      "step": 1470
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.629015051606841,
+      "learning_rate": 9.892526520374246e-06,
+      "loss": 0.8299,
+      "step": 1471
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2949568221248184,
+      "learning_rate": 9.892312669154749e-06,
+      "loss": 0.7699,
+      "step": 1472
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3253810215568416,
+      "learning_rate": 9.892098607701682e-06,
+      "loss": 0.7152,
+      "step": 1473
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4940534618720749,
+      "learning_rate": 9.891884336024242e-06,
+      "loss": 0.6853,
+      "step": 1474
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5380214134370123,
+      "learning_rate": 9.891669854131636e-06,
+      "loss": 0.7164,
+      "step": 1475
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4687610447429182,
+      "learning_rate": 9.891455162033085e-06,
+      "loss": 0.645,
+      "step": 1476
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2826199044073539,
+      "learning_rate": 9.891240259737809e-06,
+      "loss": 0.7855,
+      "step": 1477
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4940076766380508,
+      "learning_rate": 9.891025147255047e-06,
+      "loss": 0.7549,
+      "step": 1478
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6101920523885056,
+      "learning_rate": 9.890809824594041e-06,
+      "loss": 0.7343,
+      "step": 1479
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3744158366123562,
+      "learning_rate": 9.890594291764044e-06,
+      "loss": 0.7327,
+      "step": 1480
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4914374963094867,
+      "learning_rate": 9.89037854877432e-06,
+      "loss": 0.8033,
+      "step": 1481
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.218055417547801,
+      "learning_rate": 9.890162595634138e-06,
+      "loss": 0.7627,
+      "step": 1482
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3153103159779034,
+      "learning_rate": 9.889946432352776e-06,
+      "loss": 0.6689,
+      "step": 1483
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2195020029901324,
+      "learning_rate": 9.889730058939529e-06,
+      "loss": 0.6661,
+      "step": 1484
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.542706768551151,
+      "learning_rate": 9.889513475403689e-06,
+      "loss": 0.7778,
+      "step": 1485
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.539622110520928,
+      "learning_rate": 9.889296681754565e-06,
+      "loss": 0.8004,
+      "step": 1486
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.479333705643111,
+      "learning_rate": 9.889079678001474e-06,
+      "loss": 0.7729,
+      "step": 1487
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.379099757208603,
+      "learning_rate": 9.88886246415374e-06,
+      "loss": 0.7857,
+      "step": 1488
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3043699804426698,
+      "learning_rate": 9.8886450402207e-06,
+      "loss": 0.7238,
+      "step": 1489
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3420320567631099,
+      "learning_rate": 9.888427406211694e-06,
+      "loss": 0.7318,
+      "step": 1490
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.464920259853647,
+      "learning_rate": 9.888209562136074e-06,
+      "loss": 0.833,
+      "step": 1491
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5551600755057395,
+      "learning_rate": 9.887991508003205e-06,
+      "loss": 0.7456,
+      "step": 1492
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5139594129399618,
+      "learning_rate": 9.887773243822454e-06,
+      "loss": 0.7721,
+      "step": 1493
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3393508939681005,
+      "learning_rate": 9.8875547696032e-06,
+      "loss": 0.7093,
+      "step": 1494
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3528359390946814,
+      "learning_rate": 9.887336085354834e-06,
+      "loss": 0.8131,
+      "step": 1495
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2530784554116228,
+      "learning_rate": 9.887117191086749e-06,
+      "loss": 0.7814,
+      "step": 1496
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.303704980703188,
+      "learning_rate": 9.886898086808356e-06,
+      "loss": 0.6393,
+      "step": 1497
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4095189600814837,
+      "learning_rate": 9.886678772529069e-06,
+      "loss": 0.7366,
+      "step": 1498
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3316906495911232,
+      "learning_rate": 9.88645924825831e-06,
+      "loss": 0.6801,
+      "step": 1499
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3180097155269928,
+      "learning_rate": 9.886239514005515e-06,
+      "loss": 0.7199,
+      "step": 1500
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8862834175948822,
+      "learning_rate": 9.886019569780128e-06,
+      "loss": 0.8624,
+      "step": 1501
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4030111315697285,
+      "learning_rate": 9.885799415591594e-06,
+      "loss": 0.7314,
+      "step": 1502
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3702231847470827,
+      "learning_rate": 9.88557905144938e-06,
+      "loss": 0.8323,
+      "step": 1503
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.325840833884715,
+      "learning_rate": 9.885358477362956e-06,
+      "loss": 0.7012,
+      "step": 1504
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3604915398075448,
+      "learning_rate": 9.885137693341795e-06,
+      "loss": 0.7328,
+      "step": 1505
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2738481235865677,
+      "learning_rate": 9.884916699395389e-06,
+      "loss": 0.8738,
+      "step": 1506
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1544901642857908,
+      "learning_rate": 9.884695495533232e-06,
+      "loss": 0.7042,
+      "step": 1507
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7584737207427608,
+      "learning_rate": 9.884474081764833e-06,
+      "loss": 0.7679,
+      "step": 1508
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.449579573761931,
+      "learning_rate": 9.8842524580997e-06,
+      "loss": 0.6156,
+      "step": 1509
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5067030384406428,
+      "learning_rate": 9.884030624547364e-06,
+      "loss": 0.8081,
+      "step": 1510
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4330803147401245,
+      "learning_rate": 9.883808581117355e-06,
+      "loss": 0.7718,
+      "step": 1511
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0798929314862558,
+      "learning_rate": 9.883586327819214e-06,
+      "loss": 0.7569,
+      "step": 1512
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4105288222930303,
+      "learning_rate": 9.883363864662491e-06,
+      "loss": 0.6585,
+      "step": 1513
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1693316089918548,
+      "learning_rate": 9.883141191656748e-06,
+      "loss": 0.7336,
+      "step": 1514
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4258679459342618,
+      "learning_rate": 9.882918308811554e-06,
+      "loss": 0.7016,
+      "step": 1515
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6083265200493457,
+      "learning_rate": 9.882695216136485e-06,
+      "loss": 0.8348,
+      "step": 1516
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2766900530397212,
+      "learning_rate": 9.882471913641126e-06,
+      "loss": 0.735,
+      "step": 1517
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2985304107406863,
+      "learning_rate": 9.882248401335076e-06,
+      "loss": 0.7094,
+      "step": 1518
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4087034999592394,
+      "learning_rate": 9.88202467922794e-06,
+      "loss": 0.8332,
+      "step": 1519
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3278761315079906,
+      "learning_rate": 9.881800747329328e-06,
+      "loss": 0.7097,
+      "step": 1520
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4891284255095731,
+      "learning_rate": 9.881576605648869e-06,
+      "loss": 0.8016,
+      "step": 1521
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3494938102092233,
+      "learning_rate": 9.881352254196189e-06,
+      "loss": 0.8119,
+      "step": 1522
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2626772950527076,
+      "learning_rate": 9.881127692980932e-06,
+      "loss": 0.6651,
+      "step": 1523
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.350927418737862,
+      "learning_rate": 9.880902922012747e-06,
+      "loss": 0.797,
+      "step": 1524
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7856336667593022,
+      "learning_rate": 9.880677941301293e-06,
+      "loss": 0.6935,
+      "step": 1525
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3776905831117299,
+      "learning_rate": 9.880452750856239e-06,
+      "loss": 0.7057,
+      "step": 1526
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4892846063586183,
+      "learning_rate": 9.880227350687259e-06,
+      "loss": 0.7179,
+      "step": 1527
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5056038500831765,
+      "learning_rate": 9.880001740804043e-06,
+      "loss": 0.7085,
+      "step": 1528
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5788632772287885,
+      "learning_rate": 9.879775921216284e-06,
+      "loss": 0.8797,
+      "step": 1529
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4558190773516335,
+      "learning_rate": 9.879549891933684e-06,
+      "loss": 0.8795,
+      "step": 1530
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.323258781351218,
+      "learning_rate": 9.879323652965957e-06,
+      "loss": 0.6094,
+      "step": 1531
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2566108012761574,
+      "learning_rate": 9.879097204322828e-06,
+      "loss": 0.7536,
+      "step": 1532
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1255124950064712,
+      "learning_rate": 9.878870546014025e-06,
+      "loss": 0.7077,
+      "step": 1533
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2921394015962648,
+      "learning_rate": 9.87864367804929e-06,
+      "loss": 0.6025,
+      "step": 1534
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2165872975008372,
+      "learning_rate": 9.87841660043837e-06,
+      "loss": 0.6398,
+      "step": 1535
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2271175847407227,
+      "learning_rate": 9.878189313191023e-06,
+      "loss": 0.6447,
+      "step": 1536
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3642076959229852,
+      "learning_rate": 9.877961816317018e-06,
+      "loss": 0.7501,
+      "step": 1537
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2979987390409422,
+      "learning_rate": 9.87773410982613e-06,
+      "loss": 0.7319,
+      "step": 1538
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2294184089512752,
+      "learning_rate": 9.877506193728144e-06,
+      "loss": 0.6253,
+      "step": 1539
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1733961624410847,
+      "learning_rate": 9.877278068032852e-06,
+      "loss": 0.6458,
+      "step": 1540
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2389153440269476,
+      "learning_rate": 9.877049732750062e-06,
+      "loss": 0.7185,
+      "step": 1541
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.238413328346642,
+      "learning_rate": 9.876821187889581e-06,
+      "loss": 0.6948,
+      "step": 1542
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.57112777988007,
+      "learning_rate": 9.876592433461234e-06,
+      "loss": 0.7148,
+      "step": 1543
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1108082168362943,
+      "learning_rate": 9.876363469474848e-06,
+      "loss": 0.8541,
+      "step": 1544
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.38498513022297,
+      "learning_rate": 9.876134295940264e-06,
+      "loss": 0.7208,
+      "step": 1545
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.384945268764757,
+      "learning_rate": 9.875904912867328e-06,
+      "loss": 0.6877,
+      "step": 1546
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1369131782749635,
+      "learning_rate": 9.875675320265903e-06,
+      "loss": 0.5932,
+      "step": 1547
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6285879385423705,
+      "learning_rate": 9.875445518145848e-06,
+      "loss": 0.7073,
+      "step": 1548
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4145072407867187,
+      "learning_rate": 9.87521550651704e-06,
+      "loss": 0.6656,
+      "step": 1549
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3594058740595536,
+      "learning_rate": 9.874985285389364e-06,
+      "loss": 0.7496,
+      "step": 1550
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5217185980879224,
+      "learning_rate": 9.874754854772714e-06,
+      "loss": 0.7855,
+      "step": 1551
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6120401232334802,
+      "learning_rate": 9.874524214676993e-06,
+      "loss": 0.7536,
+      "step": 1552
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6090737779260242,
+      "learning_rate": 9.874293365112109e-06,
+      "loss": 0.9206,
+      "step": 1553
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3291424803011553,
+      "learning_rate": 9.874062306087983e-06,
+      "loss": 0.7051,
+      "step": 1554
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5704939667327242,
+      "learning_rate": 9.873831037614544e-06,
+      "loss": 0.7867,
+      "step": 1555
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5648370383972716,
+      "learning_rate": 9.873599559701732e-06,
+      "loss": 0.7143,
+      "step": 1556
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4437343217957446,
+      "learning_rate": 9.873367872359492e-06,
+      "loss": 0.6964,
+      "step": 1557
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7044052284953242,
+      "learning_rate": 9.873135975597781e-06,
+      "loss": 0.8681,
+      "step": 1558
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3037308806154975,
+      "learning_rate": 9.872903869426564e-06,
+      "loss": 0.691,
+      "step": 1559
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3769851120235457,
+      "learning_rate": 9.872671553855817e-06,
+      "loss": 0.7962,
+      "step": 1560
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.527576993795424,
+      "learning_rate": 9.872439028895518e-06,
+      "loss": 0.6371,
+      "step": 1561
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6547589127529696,
+      "learning_rate": 9.872206294555665e-06,
+      "loss": 0.8058,
+      "step": 1562
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.08525887625499,
+      "learning_rate": 9.871973350846255e-06,
+      "loss": 0.6803,
+      "step": 1563
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4519489568619008,
+      "learning_rate": 9.8717401977773e-06,
+      "loss": 0.8786,
+      "step": 1564
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6211566857888462,
+      "learning_rate": 9.87150683535882e-06,
+      "loss": 0.7141,
+      "step": 1565
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3650728281587128,
+      "learning_rate": 9.871273263600842e-06,
+      "loss": 0.6612,
+      "step": 1566
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1966430752452422,
+      "learning_rate": 9.871039482513402e-06,
+      "loss": 0.7041,
+      "step": 1567
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4509562179470408,
+      "learning_rate": 9.870805492106546e-06,
+      "loss": 0.79,
+      "step": 1568
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.810819462620897,
+      "learning_rate": 9.870571292390331e-06,
+      "loss": 0.9465,
+      "step": 1569
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0995368679985913,
+      "learning_rate": 9.87033688337482e-06,
+      "loss": 0.68,
+      "step": 1570
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8320194108048027,
+      "learning_rate": 9.870102265070087e-06,
+      "loss": 0.7695,
+      "step": 1571
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4328083703049317,
+      "learning_rate": 9.869867437486214e-06,
+      "loss": 0.8028,
+      "step": 1572
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5561432239060031,
+      "learning_rate": 9.86963240063329e-06,
+      "loss": 0.7614,
+      "step": 1573
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.451176802082999,
+      "learning_rate": 9.869397154521418e-06,
+      "loss": 0.8298,
+      "step": 1574
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6099267685447174,
+      "learning_rate": 9.869161699160704e-06,
+      "loss": 0.7778,
+      "step": 1575
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.178366036635672,
+      "learning_rate": 9.868926034561268e-06,
+      "loss": 0.6785,
+      "step": 1576
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3524600993772682,
+      "learning_rate": 9.868690160733237e-06,
+      "loss": 0.7731,
+      "step": 1577
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2560544101819364,
+      "learning_rate": 9.868454077686746e-06,
+      "loss": 0.6821,
+      "step": 1578
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6327050547536683,
+      "learning_rate": 9.868217785431942e-06,
+      "loss": 0.8272,
+      "step": 1579
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6418500203654134,
+      "learning_rate": 9.867981283978975e-06,
+      "loss": 0.8292,
+      "step": 1580
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4196611610922554,
+      "learning_rate": 9.867744573338012e-06,
+      "loss": 0.6506,
+      "step": 1581
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.408761783296569,
+      "learning_rate": 9.867507653519225e-06,
+      "loss": 0.808,
+      "step": 1582
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.499616179608551,
+      "learning_rate": 9.867270524532792e-06,
+      "loss": 0.7159,
+      "step": 1583
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.243497541846184,
+      "learning_rate": 9.867033186388906e-06,
+      "loss": 0.6916,
+      "step": 1584
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2792941427911688,
+      "learning_rate": 9.866795639097763e-06,
+      "loss": 0.7326,
+      "step": 1585
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.194174164956123,
+      "learning_rate": 9.866557882669574e-06,
+      "loss": 0.687,
+      "step": 1586
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.412985614880353,
+      "learning_rate": 9.866319917114554e-06,
+      "loss": 0.6952,
+      "step": 1587
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2699052985091652,
+      "learning_rate": 9.866081742442928e-06,
+      "loss": 0.6897,
+      "step": 1588
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5232279195803435,
+      "learning_rate": 9.865843358664933e-06,
+      "loss": 0.6241,
+      "step": 1589
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1350995016374887,
+      "learning_rate": 9.865604765790812e-06,
+      "loss": 0.6979,
+      "step": 1590
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4602990938457547,
+      "learning_rate": 9.865365963830818e-06,
+      "loss": 0.6511,
+      "step": 1591
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4551484494420848,
+      "learning_rate": 9.865126952795212e-06,
+      "loss": 0.8495,
+      "step": 1592
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5320683402860527,
+      "learning_rate": 9.864887732694266e-06,
+      "loss": 0.7338,
+      "step": 1593
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3871034979389751,
+      "learning_rate": 9.86464830353826e-06,
+      "loss": 0.7619,
+      "step": 1594
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2965412686585696,
+      "learning_rate": 9.864408665337482e-06,
+      "loss": 0.7814,
+      "step": 1595
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5701456636484532,
+      "learning_rate": 9.86416881810223e-06,
+      "loss": 0.7539,
+      "step": 1596
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3881790007445158,
+      "learning_rate": 9.86392876184281e-06,
+      "loss": 0.6857,
+      "step": 1597
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5117878175598554,
+      "learning_rate": 9.86368849656954e-06,
+      "loss": 0.725,
+      "step": 1598
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5941347490671047,
+      "learning_rate": 9.863448022292742e-06,
+      "loss": 0.8321,
+      "step": 1599
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4779058092452557,
+      "learning_rate": 9.863207339022753e-06,
+      "loss": 0.7988,
+      "step": 1600
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.438324775165498,
+      "learning_rate": 9.862966446769912e-06,
+      "loss": 0.6841,
+      "step": 1601
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2988970165251548,
+      "learning_rate": 9.862725345544573e-06,
+      "loss": 0.6988,
+      "step": 1602
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3573413096269609,
+      "learning_rate": 9.862484035357095e-06,
+      "loss": 0.6599,
+      "step": 1603
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8298909966181272,
+      "learning_rate": 9.86224251621785e-06,
+      "loss": 0.662,
+      "step": 1604
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3859610657793595,
+      "learning_rate": 9.862000788137215e-06,
+      "loss": 0.7355,
+      "step": 1605
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6142528524960238,
+      "learning_rate": 9.861758851125578e-06,
+      "loss": 0.8345,
+      "step": 1606
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.353046873004782,
+      "learning_rate": 9.861516705193334e-06,
+      "loss": 0.757,
+      "step": 1607
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.436861448403439,
+      "learning_rate": 9.861274350350891e-06,
+      "loss": 0.7612,
+      "step": 1608
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5221464047072308,
+      "learning_rate": 9.861031786608663e-06,
+      "loss": 0.7211,
+      "step": 1609
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2854077649237974,
+      "learning_rate": 9.860789013977074e-06,
+      "loss": 0.694,
+      "step": 1610
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.210183001804378,
+      "learning_rate": 9.860546032466553e-06,
+      "loss": 0.8932,
+      "step": 1611
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3958677776375534,
+      "learning_rate": 9.860302842087545e-06,
+      "loss": 0.7517,
+      "step": 1612
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3250188070812383,
+      "learning_rate": 9.8600594428505e-06,
+      "loss": 0.7859,
+      "step": 1613
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8668615987248773,
+      "learning_rate": 9.859815834765875e-06,
+      "loss": 0.9125,
+      "step": 1614
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6610093413514426,
+      "learning_rate": 9.859572017844143e-06,
+      "loss": 0.7288,
+      "step": 1615
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7663457017489361,
+      "learning_rate": 9.859327992095777e-06,
+      "loss": 0.6602,
+      "step": 1616
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4192531260474062,
+      "learning_rate": 9.859083757531265e-06,
+      "loss": 0.7095,
+      "step": 1617
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.282100326343385,
+      "learning_rate": 9.858839314161101e-06,
+      "loss": 0.9241,
+      "step": 1618
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7279866780698196,
+      "learning_rate": 9.858594661995792e-06,
+      "loss": 0.8639,
+      "step": 1619
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2952442402649338,
+      "learning_rate": 9.85834980104585e-06,
+      "loss": 0.6831,
+      "step": 1620
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7020770808929733,
+      "learning_rate": 9.858104731321796e-06,
+      "loss": 0.9396,
+      "step": 1621
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9857761138490233,
+      "learning_rate": 9.857859452834161e-06,
+      "loss": 0.6159,
+      "step": 1622
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4867042934050658,
+      "learning_rate": 9.857613965593488e-06,
+      "loss": 0.7427,
+      "step": 1623
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4145159088066266,
+      "learning_rate": 9.857368269610325e-06,
+      "loss": 0.745,
+      "step": 1624
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3594782970204837,
+      "learning_rate": 9.857122364895229e-06,
+      "loss": 0.7547,
+      "step": 1625
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4518653752097646,
+      "learning_rate": 9.856876251458767e-06,
+      "loss": 0.7957,
+      "step": 1626
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5498435254010199,
+      "learning_rate": 9.856629929311516e-06,
+      "loss": 0.7571,
+      "step": 1627
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2367346477765861,
+      "learning_rate": 9.85638339846406e-06,
+      "loss": 0.6333,
+      "step": 1628
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3251630583474996,
+      "learning_rate": 9.856136658926993e-06,
+      "loss": 0.6125,
+      "step": 1629
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.607134948967391,
+      "learning_rate": 9.855889710710919e-06,
+      "loss": 0.6666,
+      "step": 1630
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4404956548259233,
+      "learning_rate": 9.85564255382645e-06,
+      "loss": 0.7114,
+      "step": 1631
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3140515319051362,
+      "learning_rate": 9.855395188284206e-06,
+      "loss": 0.7337,
+      "step": 1632
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3468655633631128,
+      "learning_rate": 9.85514761409482e-06,
+      "loss": 0.6473,
+      "step": 1633
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.209056025474934,
+      "learning_rate": 9.854899831268926e-06,
+      "loss": 0.6675,
+      "step": 1634
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4620028076975495,
+      "learning_rate": 9.854651839817174e-06,
+      "loss": 0.7302,
+      "step": 1635
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3566544520157948,
+      "learning_rate": 9.854403639750219e-06,
+      "loss": 0.8002,
+      "step": 1636
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.116489478128584,
+      "learning_rate": 9.854155231078728e-06,
+      "loss": 0.6614,
+      "step": 1637
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5849833337793127,
+      "learning_rate": 9.853906613813378e-06,
+      "loss": 0.827,
+      "step": 1638
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4666532619998818,
+      "learning_rate": 9.85365778796485e-06,
+      "loss": 0.7449,
+      "step": 1639
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5046606853282956,
+      "learning_rate": 9.853408753543835e-06,
+      "loss": 0.7138,
+      "step": 1640
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.694490121623458,
+      "learning_rate": 9.85315951056104e-06,
+      "loss": 0.6995,
+      "step": 1641
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.330337006712443,
+      "learning_rate": 9.852910059027172e-06,
+      "loss": 0.711,
+      "step": 1642
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5377438936952281,
+      "learning_rate": 9.852660398952949e-06,
+      "loss": 0.6884,
+      "step": 1643
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.461524690873491,
+      "learning_rate": 9.852410530349102e-06,
+      "loss": 0.7794,
+      "step": 1644
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7372386763671266,
+      "learning_rate": 9.852160453226367e-06,
+      "loss": 0.8266,
+      "step": 1645
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4750703057719667,
+      "learning_rate": 9.851910167595493e-06,
+      "loss": 0.9716,
+      "step": 1646
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.528567953098995,
+      "learning_rate": 9.851659673467232e-06,
+      "loss": 0.7319,
+      "step": 1647
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4857779039532129,
+      "learning_rate": 9.85140897085235e-06,
+      "loss": 0.8226,
+      "step": 1648
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4847457552394414,
+      "learning_rate": 9.851158059761617e-06,
+      "loss": 0.704,
+      "step": 1649
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.311002809438783,
+      "learning_rate": 9.850906940205822e-06,
+      "loss": 0.7459,
+      "step": 1650
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2991368863559665,
+      "learning_rate": 9.85065561219575e-06,
+      "loss": 0.6674,
+      "step": 1651
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2086709369621544,
+      "learning_rate": 9.850404075742204e-06,
+      "loss": 0.6455,
+      "step": 1652
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4167654592897831,
+      "learning_rate": 9.850152330855992e-06,
+      "loss": 0.793,
+      "step": 1653
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3582572858903028,
+      "learning_rate": 9.849900377547933e-06,
+      "loss": 0.7332,
+      "step": 1654
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3879063301991819,
+      "learning_rate": 9.849648215828853e-06,
+      "loss": 0.7589,
+      "step": 1655
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7520992872423269,
+      "learning_rate": 9.849395845709588e-06,
+      "loss": 0.7433,
+      "step": 1656
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3906529514359678,
+      "learning_rate": 9.849143267200985e-06,
+      "loss": 0.5895,
+      "step": 1657
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4553140271897538,
+      "learning_rate": 9.848890480313894e-06,
+      "loss": 0.6301,
+      "step": 1658
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6335540471885384,
+      "learning_rate": 9.848637485059183e-06,
+      "loss": 0.8015,
+      "step": 1659
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2271862118017247,
+      "learning_rate": 9.848384281447718e-06,
+      "loss": 0.7804,
+      "step": 1660
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6069080131009856,
+      "learning_rate": 9.848130869490384e-06,
+      "loss": 0.8154,
+      "step": 1661
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1364066902186678,
+      "learning_rate": 9.847877249198069e-06,
+      "loss": 0.6709,
+      "step": 1662
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.615394664194317,
+      "learning_rate": 9.84762342058167e-06,
+      "loss": 0.7193,
+      "step": 1663
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6350732377498904,
+      "learning_rate": 9.8473693836521e-06,
+      "loss": 0.8349,
+      "step": 1664
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3387156642918492,
+      "learning_rate": 9.84711513842027e-06,
+      "loss": 0.7652,
+      "step": 1665
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4312306959485888,
+      "learning_rate": 9.846860684897107e-06,
+      "loss": 0.7845,
+      "step": 1666
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5340594790348903,
+      "learning_rate": 9.846606023093545e-06,
+      "loss": 0.7241,
+      "step": 1667
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.265107719094014,
+      "learning_rate": 9.84635115302053e-06,
+      "loss": 0.6474,
+      "step": 1668
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5721555145890485,
+      "learning_rate": 9.846096074689012e-06,
+      "loss": 0.7291,
+      "step": 1669
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.383116349166173,
+      "learning_rate": 9.845840788109953e-06,
+      "loss": 0.7125,
+      "step": 1670
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3211618795633047,
+      "learning_rate": 9.845585293294321e-06,
+      "loss": 0.7831,
+      "step": 1671
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7473646557724374,
+      "learning_rate": 9.845329590253098e-06,
+      "loss": 0.7216,
+      "step": 1672
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3549696665855155,
+      "learning_rate": 9.845073678997275e-06,
+      "loss": 0.7202,
+      "step": 1673
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2723324839250885,
+      "learning_rate": 9.844817559537841e-06,
+      "loss": 0.63,
+      "step": 1674
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1607032074110721,
+      "learning_rate": 9.84456123188581e-06,
+      "loss": 0.6874,
+      "step": 1675
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2503253390748106,
+      "learning_rate": 9.84430469605219e-06,
+      "loss": 0.7408,
+      "step": 1676
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.622158747903021,
+      "learning_rate": 9.84404795204801e-06,
+      "loss": 0.7065,
+      "step": 1677
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7701619637466308,
+      "learning_rate": 9.843790999884301e-06,
+      "loss": 0.7114,
+      "step": 1678
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8028988122762502,
+      "learning_rate": 9.843533839572105e-06,
+      "loss": 0.7698,
+      "step": 1679
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5536670513053916,
+      "learning_rate": 9.843276471122473e-06,
+      "loss": 0.6899,
+      "step": 1680
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3069467843037987,
+      "learning_rate": 9.843018894546466e-06,
+      "loss": 0.8164,
+      "step": 1681
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.116459765382908,
+      "learning_rate": 9.84276110985515e-06,
+      "loss": 0.7718,
+      "step": 1682
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.135526434530252,
+      "learning_rate": 9.842503117059604e-06,
+      "loss": 0.6623,
+      "step": 1683
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3427076517476155,
+      "learning_rate": 9.842244916170913e-06,
+      "loss": 0.7178,
+      "step": 1684
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.36544970004464,
+      "learning_rate": 9.841986507200174e-06,
+      "loss": 0.7907,
+      "step": 1685
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.43793146821756,
+      "learning_rate": 9.841727890158493e-06,
+      "loss": 0.8295,
+      "step": 1686
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3343583723411188,
+      "learning_rate": 9.84146906505698e-06,
+      "loss": 0.6753,
+      "step": 1687
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3031595196761563,
+      "learning_rate": 9.841210031906758e-06,
+      "loss": 0.7045,
+      "step": 1688
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5032081294005413,
+      "learning_rate": 9.840950790718959e-06,
+      "loss": 0.9593,
+      "step": 1689
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.419458868072523,
+      "learning_rate": 9.840691341504725e-06,
+      "loss": 0.7572,
+      "step": 1690
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3404651765476772,
+      "learning_rate": 9.8404316842752e-06,
+      "loss": 0.6517,
+      "step": 1691
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4694383935781925,
+      "learning_rate": 9.84017181904155e-06,
+      "loss": 0.7817,
+      "step": 1692
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3758378724895284,
+      "learning_rate": 9.839911745814933e-06,
+      "loss": 0.7616,
+      "step": 1693
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3610041448806038,
+      "learning_rate": 9.83965146460653e-06,
+      "loss": 0.7893,
+      "step": 1694
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.491847949799077,
+      "learning_rate": 9.839390975427526e-06,
+      "loss": 0.7399,
+      "step": 1695
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3739075142276111,
+      "learning_rate": 9.839130278289115e-06,
+      "loss": 0.686,
+      "step": 1696
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.535436557320033,
+      "learning_rate": 9.838869373202498e-06,
+      "loss": 0.7938,
+      "step": 1697
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5378663708127513,
+      "learning_rate": 9.838608260178885e-06,
+      "loss": 0.7738,
+      "step": 1698
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.265901065907532,
+      "learning_rate": 9.838346939229501e-06,
+      "loss": 0.5671,
+      "step": 1699
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2507328997388043,
+      "learning_rate": 9.838085410365573e-06,
+      "loss": 0.6471,
+      "step": 1700
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7356531244442819,
+      "learning_rate": 9.83782367359834e-06,
+      "loss": 0.7345,
+      "step": 1701
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1078386336202442,
+      "learning_rate": 9.837561728939048e-06,
+      "loss": 0.6973,
+      "step": 1702
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4573785829595223,
+      "learning_rate": 9.837299576398958e-06,
+      "loss": 0.8051,
+      "step": 1703
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2906799401079243,
+      "learning_rate": 9.83703721598933e-06,
+      "loss": 0.7435,
+      "step": 1704
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3537842710225345,
+      "learning_rate": 9.83677464772144e-06,
+      "loss": 0.7151,
+      "step": 1705
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8517073906179378,
+      "learning_rate": 9.83651187160657e-06,
+      "loss": 0.6363,
+      "step": 1706
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5755557718344682,
+      "learning_rate": 9.836248887656016e-06,
+      "loss": 0.7525,
+      "step": 1707
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4090629188550237,
+      "learning_rate": 9.835985695881076e-06,
+      "loss": 0.7234,
+      "step": 1708
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3592872048129627,
+      "learning_rate": 9.835722296293058e-06,
+      "loss": 0.8357,
+      "step": 1709
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2750280102012967,
+      "learning_rate": 9.835458688903286e-06,
+      "loss": 0.705,
+      "step": 1710
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8866936647918342,
+      "learning_rate": 9.835194873723085e-06,
+      "loss": 0.8702,
+      "step": 1711
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3769292672830264,
+      "learning_rate": 9.834930850763789e-06,
+      "loss": 0.7081,
+      "step": 1712
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5067054327500258,
+      "learning_rate": 9.834666620036748e-06,
+      "loss": 0.8303,
+      "step": 1713
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3739746262966812,
+      "learning_rate": 9.834402181553314e-06,
+      "loss": 0.748,
+      "step": 1714
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.587878689167956,
+      "learning_rate": 9.834137535324852e-06,
+      "loss": 0.7585,
+      "step": 1715
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2926273270169992,
+      "learning_rate": 9.833872681362735e-06,
+      "loss": 0.6982,
+      "step": 1716
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4702071037407274,
+      "learning_rate": 9.833607619678343e-06,
+      "loss": 0.7161,
+      "step": 1717
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7337271774277208,
+      "learning_rate": 9.833342350283067e-06,
+      "loss": 0.7859,
+      "step": 1718
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1438476194184786,
+      "learning_rate": 9.833076873188303e-06,
+      "loss": 0.5884,
+      "step": 1719
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4728119099510868,
+      "learning_rate": 9.832811188405464e-06,
+      "loss": 0.8206,
+      "step": 1720
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2297705130999632,
+      "learning_rate": 9.832545295945964e-06,
+      "loss": 0.6754,
+      "step": 1721
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3653191543880245,
+      "learning_rate": 9.83227919582123e-06,
+      "loss": 0.6019,
+      "step": 1722
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.537371164040746,
+      "learning_rate": 9.832012888042699e-06,
+      "loss": 0.7765,
+      "step": 1723
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4686498430775767,
+      "learning_rate": 9.831746372621811e-06,
+      "loss": 0.7803,
+      "step": 1724
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3752676962517882,
+      "learning_rate": 9.831479649570021e-06,
+      "loss": 0.5603,
+      "step": 1725
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3956244521908507,
+      "learning_rate": 9.83121271889879e-06,
+      "loss": 0.6883,
+      "step": 1726
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1286797710134062,
+      "learning_rate": 9.830945580619588e-06,
+      "loss": 0.7112,
+      "step": 1727
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4709410725822472,
+      "learning_rate": 9.830678234743896e-06,
+      "loss": 0.7619,
+      "step": 1728
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4563676822666491,
+      "learning_rate": 9.830410681283203e-06,
+      "loss": 0.7577,
+      "step": 1729
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6918650761402927,
+      "learning_rate": 9.830142920249004e-06,
+      "loss": 0.7656,
+      "step": 1730
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.250711274228375,
+      "learning_rate": 9.829874951652807e-06,
+      "loss": 0.6177,
+      "step": 1731
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.202672686086619,
+      "learning_rate": 9.829606775506124e-06,
+      "loss": 0.6812,
+      "step": 1732
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5549207239186587,
+      "learning_rate": 9.829338391820484e-06,
+      "loss": 0.8038,
+      "step": 1733
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7426020292123292,
+      "learning_rate": 9.829069800607418e-06,
+      "loss": 0.7497,
+      "step": 1734
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6595410148742198,
+      "learning_rate": 9.828801001878468e-06,
+      "loss": 0.6322,
+      "step": 1735
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5583638894133287,
+      "learning_rate": 9.828531995645183e-06,
+      "loss": 0.7118,
+      "step": 1736
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.345748230390202,
+      "learning_rate": 9.828262781919124e-06,
+      "loss": 0.6854,
+      "step": 1737
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.261866995877306,
+      "learning_rate": 9.827993360711863e-06,
+      "loss": 0.6679,
+      "step": 1738
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4420622093919346,
+      "learning_rate": 9.827723732034972e-06,
+      "loss": 0.7296,
+      "step": 1739
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7830671039982011,
+      "learning_rate": 9.827453895900043e-06,
+      "loss": 0.8901,
+      "step": 1740
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3789582223052963,
+      "learning_rate": 9.827183852318666e-06,
+      "loss": 0.8288,
+      "step": 1741
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2732650040416924,
+      "learning_rate": 9.82691360130245e-06,
+      "loss": 0.6553,
+      "step": 1742
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.220943915753849,
+      "learning_rate": 9.826643142863006e-06,
+      "loss": 0.6739,
+      "step": 1743
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4587345726654013,
+      "learning_rate": 9.826372477011956e-06,
+      "loss": 0.7382,
+      "step": 1744
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3248084681010313,
+      "learning_rate": 9.826101603760934e-06,
+      "loss": 0.7184,
+      "step": 1745
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2158364231108407,
+      "learning_rate": 9.825830523121575e-06,
+      "loss": 0.707,
+      "step": 1746
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2981764212178903,
+      "learning_rate": 9.825559235105532e-06,
+      "loss": 0.6314,
+      "step": 1747
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4043953912061669,
+      "learning_rate": 9.82528773972446e-06,
+      "loss": 0.6348,
+      "step": 1748
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.387902042262686,
+      "learning_rate": 9.825016036990029e-06,
+      "loss": 0.7268,
+      "step": 1749
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5113712510502137,
+      "learning_rate": 9.824744126913914e-06,
+      "loss": 0.7286,
+      "step": 1750
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.564342886485657,
+      "learning_rate": 9.824472009507796e-06,
+      "loss": 0.8318,
+      "step": 1751
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3806166631488086,
+      "learning_rate": 9.824199684783373e-06,
+      "loss": 0.6825,
+      "step": 1752
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.758620509847144,
+      "learning_rate": 9.823927152752344e-06,
+      "loss": 0.8506,
+      "step": 1753
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.8301189905511914,
+      "learning_rate": 9.823654413426424e-06,
+      "loss": 0.7048,
+      "step": 1754
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5367389436820789,
+      "learning_rate": 9.82338146681733e-06,
+      "loss": 0.6383,
+      "step": 1755
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3806893081551903,
+      "learning_rate": 9.823108312936791e-06,
+      "loss": 0.6622,
+      "step": 1756
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2545267341769222,
+      "learning_rate": 9.822834951796547e-06,
+      "loss": 0.6134,
+      "step": 1757
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3880612801138235,
+      "learning_rate": 9.822561383408345e-06,
+      "loss": 0.731,
+      "step": 1758
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5521628444244833,
+      "learning_rate": 9.822287607783938e-06,
+      "loss": 0.6913,
+      "step": 1759
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3751043290482317,
+      "learning_rate": 9.822013624935093e-06,
+      "loss": 0.6778,
+      "step": 1760
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6299029242149035,
+      "learning_rate": 9.821739434873586e-06,
+      "loss": 0.9049,
+      "step": 1761
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.552228435553787,
+      "learning_rate": 9.821465037611194e-06,
+      "loss": 0.7503,
+      "step": 1762
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3970540750115874,
+      "learning_rate": 9.821190433159711e-06,
+      "loss": 0.7217,
+      "step": 1763
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.542349401204882,
+      "learning_rate": 9.820915621530939e-06,
+      "loss": 0.6815,
+      "step": 1764
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3110400197409964,
+      "learning_rate": 9.820640602736687e-06,
+      "loss": 0.7029,
+      "step": 1765
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2945772934583513,
+      "learning_rate": 9.82036537678877e-06,
+      "loss": 0.6661,
+      "step": 1766
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4901071225268259,
+      "learning_rate": 9.820089943699018e-06,
+      "loss": 0.7846,
+      "step": 1767
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3406578849427973,
+      "learning_rate": 9.819814303479268e-06,
+      "loss": 0.6263,
+      "step": 1768
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5144997689045208,
+      "learning_rate": 9.81953845614136e-06,
+      "loss": 0.7977,
+      "step": 1769
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5637678018940189,
+      "learning_rate": 9.819262401697154e-06,
+      "loss": 0.6826,
+      "step": 1770
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2866564541205974,
+      "learning_rate": 9.818986140158507e-06,
+      "loss": 0.6619,
+      "step": 1771
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4141948992953997,
+      "learning_rate": 9.818709671537295e-06,
+      "loss": 0.6439,
+      "step": 1772
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6625026199116923,
+      "learning_rate": 9.818432995845394e-06,
+      "loss": 0.7712,
+      "step": 1773
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6389572386092144,
+      "learning_rate": 9.818156113094699e-06,
+      "loss": 0.7155,
+      "step": 1774
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4698482951578549,
+      "learning_rate": 9.817879023297103e-06,
+      "loss": 0.7738,
+      "step": 1775
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.455467281741214,
+      "learning_rate": 9.817601726464516e-06,
+      "loss": 0.7064,
+      "step": 1776
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6296466591052103,
+      "learning_rate": 9.817324222608853e-06,
+      "loss": 0.6706,
+      "step": 1777
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.559503591765765,
+      "learning_rate": 9.817046511742042e-06,
+      "loss": 0.784,
+      "step": 1778
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.550507259027689,
+      "learning_rate": 9.816768593876012e-06,
+      "loss": 0.7725,
+      "step": 1779
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.542642153157851,
+      "learning_rate": 9.816490469022708e-06,
+      "loss": 0.7376,
+      "step": 1780
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6737891511273262,
+      "learning_rate": 9.816212137194081e-06,
+      "loss": 0.8268,
+      "step": 1781
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4722263441091783,
+      "learning_rate": 9.815933598402095e-06,
+      "loss": 0.7383,
+      "step": 1782
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4928440190036387,
+      "learning_rate": 9.815654852658714e-06,
+      "loss": 0.7112,
+      "step": 1783
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5278369410387802,
+      "learning_rate": 9.81537589997592e-06,
+      "loss": 0.7237,
+      "step": 1784
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3370114084718554,
+      "learning_rate": 9.815096740365698e-06,
+      "loss": 0.7005,
+      "step": 1785
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3264378479840717,
+      "learning_rate": 9.814817373840046e-06,
+      "loss": 0.7105,
+      "step": 1786
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.599939982713163,
+      "learning_rate": 9.814537800410968e-06,
+      "loss": 0.7813,
+      "step": 1787
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.445299511106832,
+      "learning_rate": 9.814258020090478e-06,
+      "loss": 0.7,
+      "step": 1788
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3624488023806534,
+      "learning_rate": 9.8139780328906e-06,
+      "loss": 0.6946,
+      "step": 1789
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1658951773091597,
+      "learning_rate": 9.813697838823364e-06,
+      "loss": 0.6845,
+      "step": 1790
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.331648987320329,
+      "learning_rate": 9.813417437900812e-06,
+      "loss": 0.7189,
+      "step": 1791
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4803887667607003,
+      "learning_rate": 9.81313683013499e-06,
+      "loss": 0.7267,
+      "step": 1792
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2429305735902998,
+      "learning_rate": 9.81285601553796e-06,
+      "loss": 0.6319,
+      "step": 1793
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2254001360713562,
+      "learning_rate": 9.812574994121791e-06,
+      "loss": 0.8553,
+      "step": 1794
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0314019108480963,
+      "learning_rate": 9.812293765898555e-06,
+      "loss": 0.6017,
+      "step": 1795
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3869068240844185,
+      "learning_rate": 9.812012330880337e-06,
+      "loss": 0.7167,
+      "step": 1796
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4117271149367578,
+      "learning_rate": 9.811730689079233e-06,
+      "loss": 0.6859,
+      "step": 1797
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2787256912739773,
+      "learning_rate": 9.811448840507345e-06,
+      "loss": 0.7129,
+      "step": 1798
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3475157688582047,
+      "learning_rate": 9.811166785176785e-06,
+      "loss": 0.6894,
+      "step": 1799
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3667053674562193,
+      "learning_rate": 9.810884523099673e-06,
+      "loss": 0.7027,
+      "step": 1800
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3046082815487476,
+      "learning_rate": 9.810602054288138e-06,
+      "loss": 0.6884,
+      "step": 1801
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.355474031352781,
+      "learning_rate": 9.81031937875432e-06,
+      "loss": 0.7623,
+      "step": 1802
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7681728982312386,
+      "learning_rate": 9.810036496510364e-06,
+      "loss": 0.7517,
+      "step": 1803
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1730831736688296,
+      "learning_rate": 9.809753407568427e-06,
+      "loss": 0.7101,
+      "step": 1804
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7993012145324516,
+      "learning_rate": 9.809470111940675e-06,
+      "loss": 0.849,
+      "step": 1805
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2874001974782545,
+      "learning_rate": 9.809186609639281e-06,
+      "loss": 0.7795,
+      "step": 1806
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3166963804737515,
+      "learning_rate": 9.808902900676429e-06,
+      "loss": 0.7075,
+      "step": 1807
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4336062833903584,
+      "learning_rate": 9.808618985064307e-06,
+      "loss": 0.7585,
+      "step": 1808
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1590070298390938,
+      "learning_rate": 9.80833486281512e-06,
+      "loss": 0.6547,
+      "step": 1809
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.74659812471405,
+      "learning_rate": 9.808050533941074e-06,
+      "loss": 0.848,
+      "step": 1810
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5481975052889123,
+      "learning_rate": 9.807765998454388e-06,
+      "loss": 0.7333,
+      "step": 1811
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3375824806070162,
+      "learning_rate": 9.80748125636729e-06,
+      "loss": 0.6755,
+      "step": 1812
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3045230564750643,
+      "learning_rate": 9.807196307692015e-06,
+      "loss": 0.7069,
+      "step": 1813
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7804371347859382,
+      "learning_rate": 9.80691115244081e-06,
+      "loss": 0.7449,
+      "step": 1814
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.150159504095301,
+      "learning_rate": 9.806625790625925e-06,
+      "loss": 0.6681,
+      "step": 1815
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3574260211211482,
+      "learning_rate": 9.806340222259627e-06,
+      "loss": 0.7306,
+      "step": 1816
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3084353542757445,
+      "learning_rate": 9.806054447354185e-06,
+      "loss": 0.7754,
+      "step": 1817
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3935276255952873,
+      "learning_rate": 9.805768465921878e-06,
+      "loss": 0.7696,
+      "step": 1818
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3051558623446968,
+      "learning_rate": 9.805482277974999e-06,
+      "loss": 0.5973,
+      "step": 1819
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.001799988787791,
+      "learning_rate": 9.805195883525844e-06,
+      "loss": 0.61,
+      "step": 1820
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1159989744991037,
+      "learning_rate": 9.804909282586718e-06,
+      "loss": 0.6835,
+      "step": 1821
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4657724286567022,
+      "learning_rate": 9.804622475169943e-06,
+      "loss": 0.7182,
+      "step": 1822
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.492885764931635,
+      "learning_rate": 9.804335461287835e-06,
+      "loss": 0.7664,
+      "step": 1823
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2907162785304405,
+      "learning_rate": 9.804048240952736e-06,
+      "loss": 0.6451,
+      "step": 1824
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2495963894289202,
+      "learning_rate": 9.803760814176983e-06,
+      "loss": 0.7564,
+      "step": 1825
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9233913459114904,
+      "learning_rate": 9.80347318097293e-06,
+      "loss": 0.7941,
+      "step": 1826
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3723085833585238,
+      "learning_rate": 9.803185341352936e-06,
+      "loss": 0.8869,
+      "step": 1827
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3786947805087704,
+      "learning_rate": 9.802897295329373e-06,
+      "loss": 0.7762,
+      "step": 1828
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5121999208423682,
+      "learning_rate": 9.802609042914614e-06,
+      "loss": 0.6373,
+      "step": 1829
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0428933011852999,
+      "learning_rate": 9.80232058412105e-06,
+      "loss": 0.6566,
+      "step": 1830
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5644154018916112,
+      "learning_rate": 9.802031918961073e-06,
+      "loss": 0.8387,
+      "step": 1831
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.288046572460731,
+      "learning_rate": 9.801743047447092e-06,
+      "loss": 0.7822,
+      "step": 1832
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4267830455753807,
+      "learning_rate": 9.801453969591517e-06,
+      "loss": 0.6422,
+      "step": 1833
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3981678408029128,
+      "learning_rate": 9.80116468540677e-06,
+      "loss": 0.7323,
+      "step": 1834
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3969987080940083,
+      "learning_rate": 9.800875194905286e-06,
+      "loss": 0.7126,
+      "step": 1835
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.466886389718225,
+      "learning_rate": 9.800585498099501e-06,
+      "loss": 0.7857,
+      "step": 1836
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4076220644816415,
+      "learning_rate": 9.800295595001868e-06,
+      "loss": 0.8056,
+      "step": 1837
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3845511666823935,
+      "learning_rate": 9.800005485624842e-06,
+      "loss": 0.7632,
+      "step": 1838
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.612296910280989,
+      "learning_rate": 9.79971516998089e-06,
+      "loss": 0.8143,
+      "step": 1839
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4479811751950002,
+      "learning_rate": 9.799424648082486e-06,
+      "loss": 0.7164,
+      "step": 1840
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7172135303107015,
+      "learning_rate": 9.799133919942117e-06,
+      "loss": 0.8818,
+      "step": 1841
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3021643792527202,
+      "learning_rate": 9.798842985572274e-06,
+      "loss": 0.7448,
+      "step": 1842
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5348119959788178,
+      "learning_rate": 9.798551844985463e-06,
+      "loss": 0.9422,
+      "step": 1843
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3679578160414916,
+      "learning_rate": 9.79826049819419e-06,
+      "loss": 0.6996,
+      "step": 1844
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4254986083479926,
+      "learning_rate": 9.797968945210978e-06,
+      "loss": 0.7269,
+      "step": 1845
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4532083367342143,
+      "learning_rate": 9.797677186048356e-06,
+      "loss": 0.7032,
+      "step": 1846
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5556117876434872,
+      "learning_rate": 9.797385220718859e-06,
+      "loss": 0.7255,
+      "step": 1847
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.441372303733768,
+      "learning_rate": 9.797093049235034e-06,
+      "loss": 0.6524,
+      "step": 1848
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0590179812925162,
+      "learning_rate": 9.796800671609436e-06,
+      "loss": 0.606,
+      "step": 1849
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3853755330841069,
+      "learning_rate": 9.796508087854632e-06,
+      "loss": 0.7672,
+      "step": 1850
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5239656557729229,
+      "learning_rate": 9.796215297983193e-06,
+      "loss": 0.6584,
+      "step": 1851
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.256911512781908,
+      "learning_rate": 9.7959223020077e-06,
+      "loss": 0.7545,
+      "step": 1852
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5655837152379883,
+      "learning_rate": 9.795629099940746e-06,
+      "loss": 0.7406,
+      "step": 1853
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6580018466366606,
+      "learning_rate": 9.795335691794929e-06,
+      "loss": 0.7224,
+      "step": 1854
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5411778620866192,
+      "learning_rate": 9.795042077582856e-06,
+      "loss": 0.7289,
+      "step": 1855
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.266753864566481,
+      "learning_rate": 9.794748257317148e-06,
+      "loss": 0.6293,
+      "step": 1856
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6136210752690174,
+      "learning_rate": 9.794454231010427e-06,
+      "loss": 0.7333,
+      "step": 1857
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3479987135407498,
+      "learning_rate": 9.794159998675331e-06,
+      "loss": 0.6888,
+      "step": 1858
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.19682010807972,
+      "learning_rate": 9.793865560324503e-06,
+      "loss": 0.6259,
+      "step": 1859
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1944498047249905,
+      "learning_rate": 9.793570915970594e-06,
+      "loss": 0.7214,
+      "step": 1860
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5574350927096463,
+      "learning_rate": 9.793276065626267e-06,
+      "loss": 0.8108,
+      "step": 1861
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3910471579047876,
+      "learning_rate": 9.792981009304192e-06,
+      "loss": 0.755,
+      "step": 1862
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.510556371567547,
+      "learning_rate": 9.79268574701705e-06,
+      "loss": 0.6928,
+      "step": 1863
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4355538388427702,
+      "learning_rate": 9.792390278777527e-06,
+      "loss": 0.7146,
+      "step": 1864
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2402710254550355,
+      "learning_rate": 9.79209460459832e-06,
+      "loss": 0.74,
+      "step": 1865
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4442036367299391,
+      "learning_rate": 9.791798724492136e-06,
+      "loss": 0.7961,
+      "step": 1866
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4756016431766268,
+      "learning_rate": 9.791502638471689e-06,
+      "loss": 0.8248,
+      "step": 1867
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4810346186844292,
+      "learning_rate": 9.7912063465497e-06,
+      "loss": 0.7279,
+      "step": 1868
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4463490439069566,
+      "learning_rate": 9.790909848738907e-06,
+      "loss": 0.6892,
+      "step": 1869
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4644459363359048,
+      "learning_rate": 9.790613145052046e-06,
+      "loss": 0.7213,
+      "step": 1870
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.379919933460725,
+      "learning_rate": 9.790316235501869e-06,
+      "loss": 0.6874,
+      "step": 1871
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3988549402306314,
+      "learning_rate": 9.790019120101135e-06,
+      "loss": 0.7638,
+      "step": 1872
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3938779891963953,
+      "learning_rate": 9.789721798862612e-06,
+      "loss": 0.6801,
+      "step": 1873
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5220127918941817,
+      "learning_rate": 9.789424271799075e-06,
+      "loss": 0.7012,
+      "step": 1874
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4701294052286542,
+      "learning_rate": 9.789126538923312e-06,
+      "loss": 0.7644,
+      "step": 1875
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4744335096403551,
+      "learning_rate": 9.788828600248114e-06,
+      "loss": 0.7258,
+      "step": 1876
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3719658387294085,
+      "learning_rate": 9.788530455786288e-06,
+      "loss": 0.7312,
+      "step": 1877
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4609430480083614,
+      "learning_rate": 9.78823210555064e-06,
+      "loss": 0.7279,
+      "step": 1878
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2822442004718115,
+      "learning_rate": 9.787933549553996e-06,
+      "loss": 0.6871,
+      "step": 1879
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4215527371552605,
+      "learning_rate": 9.787634787809185e-06,
+      "loss": 0.766,
+      "step": 1880
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5936563493704698,
+      "learning_rate": 9.787335820329045e-06,
+      "loss": 0.669,
+      "step": 1881
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0289902426424864,
+      "learning_rate": 9.787036647126422e-06,
+      "loss": 0.7941,
+      "step": 1882
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4864233236380018,
+      "learning_rate": 9.786737268214172e-06,
+      "loss": 0.6943,
+      "step": 1883
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2383423952645285,
+      "learning_rate": 9.786437683605161e-06,
+      "loss": 0.6661,
+      "step": 1884
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3122559133761753,
+      "learning_rate": 9.786137893312265e-06,
+      "loss": 0.6801,
+      "step": 1885
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4809791652504465,
+      "learning_rate": 9.785837897348363e-06,
+      "loss": 0.793,
+      "step": 1886
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.391553468216565,
+      "learning_rate": 9.785537695726348e-06,
+      "loss": 0.8328,
+      "step": 1887
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8844344710946197,
+      "learning_rate": 9.78523728845912e-06,
+      "loss": 0.8189,
+      "step": 1888
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9410967259070377,
+      "learning_rate": 9.78493667555959e-06,
+      "loss": 0.8024,
+      "step": 1889
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5084049916047344,
+      "learning_rate": 9.784635857040672e-06,
+      "loss": 0.688,
+      "step": 1890
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.756003759638627,
+      "learning_rate": 9.784334832915297e-06,
+      "loss": 0.8083,
+      "step": 1891
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.674781576146402,
+      "learning_rate": 9.784033603196397e-06,
+      "loss": 0.7633,
+      "step": 1892
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.266695998484506,
+      "learning_rate": 9.78373216789692e-06,
+      "loss": 0.7451,
+      "step": 1893
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1854323077027529,
+      "learning_rate": 9.783430527029818e-06,
+      "loss": 0.6188,
+      "step": 1894
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.378993906359698,
+      "learning_rate": 9.783128680608052e-06,
+      "loss": 0.6888,
+      "step": 1895
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3835307812479902,
+      "learning_rate": 9.782826628644596e-06,
+      "loss": 0.775,
+      "step": 1896
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.396296872922543,
+      "learning_rate": 9.782524371152425e-06,
+      "loss": 0.7141,
+      "step": 1897
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.338313725786748,
+      "learning_rate": 9.782221908144531e-06,
+      "loss": 0.7296,
+      "step": 1898
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3733122485532232,
+      "learning_rate": 9.781919239633912e-06,
+      "loss": 0.7555,
+      "step": 1899
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3660415047927266,
+      "learning_rate": 9.781616365633574e-06,
+      "loss": 0.782,
+      "step": 1900
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6002646462505479,
+      "learning_rate": 9.78131328615653e-06,
+      "loss": 0.8896,
+      "step": 1901
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3860913892051065,
+      "learning_rate": 9.781010001215806e-06,
+      "loss": 0.7178,
+      "step": 1902
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4063952279728502,
+      "learning_rate": 9.780706510824436e-06,
+      "loss": 0.7682,
+      "step": 1903
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.667802014094716,
+      "learning_rate": 9.780402814995458e-06,
+      "loss": 0.8008,
+      "step": 1904
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6426046335956013,
+      "learning_rate": 9.780098913741926e-06,
+      "loss": 0.6751,
+      "step": 1905
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6045383870389796,
+      "learning_rate": 9.779794807076896e-06,
+      "loss": 0.7463,
+      "step": 1906
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.178044056443966,
+      "learning_rate": 9.77949049501344e-06,
+      "loss": 0.606,
+      "step": 1907
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5180348022675352,
+      "learning_rate": 9.779185977564633e-06,
+      "loss": 0.6726,
+      "step": 1908
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2545394508557248,
+      "learning_rate": 9.77888125474356e-06,
+      "loss": 0.7024,
+      "step": 1909
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3121538259669192,
+      "learning_rate": 9.778576326563315e-06,
+      "loss": 0.6475,
+      "step": 1910
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4889478891859502,
+      "learning_rate": 9.778271193037003e-06,
+      "loss": 0.7,
+      "step": 1911
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1677803839074068,
+      "learning_rate": 9.777965854177737e-06,
+      "loss": 0.6271,
+      "step": 1912
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9149982490794617,
+      "learning_rate": 9.777660309998638e-06,
+      "loss": 0.7438,
+      "step": 1913
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1701684007199862,
+      "learning_rate": 9.777354560512835e-06,
+      "loss": 0.6566,
+      "step": 1914
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1908601021051333,
+      "learning_rate": 9.777048605733465e-06,
+      "loss": 0.6769,
+      "step": 1915
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5099553304696556,
+      "learning_rate": 9.776742445673679e-06,
+      "loss": 0.729,
+      "step": 1916
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3878757997445414,
+      "learning_rate": 9.77643608034663e-06,
+      "loss": 0.7086,
+      "step": 1917
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4696213846492785,
+      "learning_rate": 9.776129509765487e-06,
+      "loss": 0.7998,
+      "step": 1918
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.601427010168563,
+      "learning_rate": 9.77582273394342e-06,
+      "loss": 0.7232,
+      "step": 1919
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1974257921557347,
+      "learning_rate": 9.775515752893613e-06,
+      "loss": 0.7142,
+      "step": 1920
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.076814879566647,
+      "learning_rate": 9.775208566629262e-06,
+      "loss": 0.6485,
+      "step": 1921
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0953723200808776,
+      "learning_rate": 9.774901175163562e-06,
+      "loss": 0.6357,
+      "step": 1922
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4593520482118714,
+      "learning_rate": 9.774593578509724e-06,
+      "loss": 0.7838,
+      "step": 1923
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2297271361017363,
+      "learning_rate": 9.774285776680967e-06,
+      "loss": 0.6595,
+      "step": 1924
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.277989700147221,
+      "learning_rate": 9.773977769690517e-06,
+      "loss": 0.6703,
+      "step": 1925
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6354444074749983,
+      "learning_rate": 9.773669557551609e-06,
+      "loss": 0.9246,
+      "step": 1926
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2773194134207981,
+      "learning_rate": 9.77336114027749e-06,
+      "loss": 0.6873,
+      "step": 1927
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.601119243308364,
+      "learning_rate": 9.773052517881412e-06,
+      "loss": 0.7341,
+      "step": 1928
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5255530596679987,
+      "learning_rate": 9.772743690376636e-06,
+      "loss": 0.8245,
+      "step": 1929
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6279130945615594,
+      "learning_rate": 9.772434657776435e-06,
+      "loss": 0.7885,
+      "step": 1930
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4136378962003853,
+      "learning_rate": 9.772125420094087e-06,
+      "loss": 0.7087,
+      "step": 1931
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2221840349229813,
+      "learning_rate": 9.771815977342882e-06,
+      "loss": 0.8014,
+      "step": 1932
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4040101185043288,
+      "learning_rate": 9.771506329536118e-06,
+      "loss": 0.7214,
+      "step": 1933
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1644466888287675,
+      "learning_rate": 9.7711964766871e-06,
+      "loss": 0.6894,
+      "step": 1934
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.039672067343357,
+      "learning_rate": 9.770886418809142e-06,
+      "loss": 0.7533,
+      "step": 1935
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4504170941195837,
+      "learning_rate": 9.77057615591557e-06,
+      "loss": 0.7902,
+      "step": 1936
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.494876176263219,
+      "learning_rate": 9.770265688019718e-06,
+      "loss": 0.7575,
+      "step": 1937
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5565569253962763,
+      "learning_rate": 9.769955015134923e-06,
+      "loss": 0.7143,
+      "step": 1938
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3787108082138888,
+      "learning_rate": 9.76964413727454e-06,
+      "loss": 0.7227,
+      "step": 1939
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5165981165935647,
+      "learning_rate": 9.769333054451923e-06,
+      "loss": 0.7165,
+      "step": 1940
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4398749058720028,
+      "learning_rate": 9.769021766680444e-06,
+      "loss": 0.7281,
+      "step": 1941
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4911177872000483,
+      "learning_rate": 9.768710273973479e-06,
+      "loss": 0.6622,
+      "step": 1942
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2002470374943537,
+      "learning_rate": 9.768398576344411e-06,
+      "loss": 0.7053,
+      "step": 1943
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2590740505946085,
+      "learning_rate": 9.768086673806638e-06,
+      "loss": 0.7302,
+      "step": 1944
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3277174739339732,
+      "learning_rate": 9.767774566373563e-06,
+      "loss": 0.7454,
+      "step": 1945
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4567360093679385,
+      "learning_rate": 9.767462254058593e-06,
+      "loss": 0.7696,
+      "step": 1946
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1217910822464707,
+      "learning_rate": 9.767149736875156e-06,
+      "loss": 0.6967,
+      "step": 1947
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4376108440250064,
+      "learning_rate": 9.766837014836675e-06,
+      "loss": 0.7334,
+      "step": 1948
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3276286389730703,
+      "learning_rate": 9.766524087956592e-06,
+      "loss": 0.7282,
+      "step": 1949
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3484304567530252,
+      "learning_rate": 9.766210956248354e-06,
+      "loss": 0.6539,
+      "step": 1950
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5285379248102726,
+      "learning_rate": 9.765897619725416e-06,
+      "loss": 0.7951,
+      "step": 1951
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6596007227271736,
+      "learning_rate": 9.765584078401242e-06,
+      "loss": 0.6977,
+      "step": 1952
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1068747981110454,
+      "learning_rate": 9.765270332289307e-06,
+      "loss": 0.6573,
+      "step": 1953
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4551060131705522,
+      "learning_rate": 9.764956381403095e-06,
+      "loss": 0.714,
+      "step": 1954
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.522243698994215,
+      "learning_rate": 9.764642225756093e-06,
+      "loss": 0.8164,
+      "step": 1955
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.504370000230076,
+      "learning_rate": 9.764327865361805e-06,
+      "loss": 0.6888,
+      "step": 1956
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3189237526000728,
+      "learning_rate": 9.764013300233736e-06,
+      "loss": 0.7651,
+      "step": 1957
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7008804743810826,
+      "learning_rate": 9.763698530385407e-06,
+      "loss": 0.7101,
+      "step": 1958
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2498159549252106,
+      "learning_rate": 9.76338355583034e-06,
+      "loss": 0.6977,
+      "step": 1959
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6087611449120198,
+      "learning_rate": 9.763068376582075e-06,
+      "loss": 0.7177,
+      "step": 1960
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.615266008582511,
+      "learning_rate": 9.762752992654155e-06,
+      "loss": 0.7853,
+      "step": 1961
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2811227412754869,
+      "learning_rate": 9.76243740406013e-06,
+      "loss": 0.7111,
+      "step": 1962
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6418845663138433,
+      "learning_rate": 9.762121610813564e-06,
+      "loss": 0.7928,
+      "step": 1963
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2708955166230838,
+      "learning_rate": 9.761805612928025e-06,
+      "loss": 0.7228,
+      "step": 1964
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1248277012604102,
+      "learning_rate": 9.761489410417095e-06,
+      "loss": 0.6448,
+      "step": 1965
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.360145550283168,
+      "learning_rate": 9.761173003294362e-06,
+      "loss": 0.6128,
+      "step": 1966
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6180466888927767,
+      "learning_rate": 9.76085639157342e-06,
+      "loss": 0.7973,
+      "step": 1967
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1747325398139674,
+      "learning_rate": 9.760539575267875e-06,
+      "loss": 0.6568,
+      "step": 1968
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3400430780177703,
+      "learning_rate": 9.760222554391343e-06,
+      "loss": 0.7199,
+      "step": 1969
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2856444395167905,
+      "learning_rate": 9.759905328957446e-06,
+      "loss": 0.7291,
+      "step": 1970
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.450261225483734,
+      "learning_rate": 9.759587898979816e-06,
+      "loss": 0.7166,
+      "step": 1971
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7577348807601747,
+      "learning_rate": 9.759270264472097e-06,
+      "loss": 0.7407,
+      "step": 1972
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3651818646712344,
+      "learning_rate": 9.75895242544793e-06,
+      "loss": 0.6721,
+      "step": 1973
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6297660293779095,
+      "learning_rate": 9.758634381920982e-06,
+      "loss": 0.7421,
+      "step": 1974
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.498264682567286,
+      "learning_rate": 9.758316133904914e-06,
+      "loss": 0.7459,
+      "step": 1975
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.285950585602542,
+      "learning_rate": 9.757997681413407e-06,
+      "loss": 0.625,
+      "step": 1976
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4485306145137316,
+      "learning_rate": 9.757679024460142e-06,
+      "loss": 0.7308,
+      "step": 1977
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3933975007623767,
+      "learning_rate": 9.757360163058814e-06,
+      "loss": 0.7152,
+      "step": 1978
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5302336185957546,
+      "learning_rate": 9.757041097223123e-06,
+      "loss": 0.7781,
+      "step": 1979
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.436618260363386,
+      "learning_rate": 9.756721826966783e-06,
+      "loss": 0.7758,
+      "step": 1980
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3412646463841902,
+      "learning_rate": 9.756402352303513e-06,
+      "loss": 0.6584,
+      "step": 1981
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.43660868386807,
+      "learning_rate": 9.756082673247039e-06,
+      "loss": 0.7716,
+      "step": 1982
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2812993424401187,
+      "learning_rate": 9.755762789811101e-06,
+      "loss": 0.7667,
+      "step": 1983
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4787789918687588,
+      "learning_rate": 9.755442702009443e-06,
+      "loss": 0.6697,
+      "step": 1984
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.301564860135352,
+      "learning_rate": 9.755122409855822e-06,
+      "loss": 0.6907,
+      "step": 1985
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5130886601971671,
+      "learning_rate": 9.754801913364002e-06,
+      "loss": 0.7352,
+      "step": 1986
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2068438858239676,
+      "learning_rate": 9.754481212547752e-06,
+      "loss": 0.6424,
+      "step": 1987
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4509438100404715,
+      "learning_rate": 9.754160307420858e-06,
+      "loss": 0.7409,
+      "step": 1988
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5008024279689067,
+      "learning_rate": 9.753839197997105e-06,
+      "loss": 0.7475,
+      "step": 1989
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2858573219168654,
+      "learning_rate": 9.753517884290298e-06,
+      "loss": 0.7009,
+      "step": 1990
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3781446607160437,
+      "learning_rate": 9.753196366314237e-06,
+      "loss": 0.7856,
+      "step": 1991
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5784851873085792,
+      "learning_rate": 9.752874644082743e-06,
+      "loss": 0.7397,
+      "step": 1992
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4211235248480767,
+      "learning_rate": 9.752552717609643e-06,
+      "loss": 0.8127,
+      "step": 1993
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2470802897671478,
+      "learning_rate": 9.752230586908767e-06,
+      "loss": 0.668,
+      "step": 1994
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.324698915464215,
+      "learning_rate": 9.751908251993956e-06,
+      "loss": 0.7327,
+      "step": 1995
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3074553760676526,
+      "learning_rate": 9.751585712879067e-06,
+      "loss": 0.6758,
+      "step": 1996
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3625893489569387,
+      "learning_rate": 9.751262969577958e-06,
+      "loss": 0.7934,
+      "step": 1997
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4589791970392385,
+      "learning_rate": 9.750940022104497e-06,
+      "loss": 0.8449,
+      "step": 1998
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4477726001643387,
+      "learning_rate": 9.75061687047256e-06,
+      "loss": 0.7896,
+      "step": 1999
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.530105491700277,
+      "learning_rate": 9.750293514696038e-06,
+      "loss": 0.6846,
+      "step": 2000
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.439284369727529,
+      "learning_rate": 9.749969954788825e-06,
+      "loss": 0.719,
+      "step": 2001
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4493045534813183,
+      "learning_rate": 9.749646190764823e-06,
+      "loss": 0.7234,
+      "step": 2002
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5304679843863345,
+      "learning_rate": 9.749322222637947e-06,
+      "loss": 0.8305,
+      "step": 2003
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.583750806393853,
+      "learning_rate": 9.748998050422117e-06,
+      "loss": 0.8108,
+      "step": 2004
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.568865005578827,
+      "learning_rate": 9.748673674131264e-06,
+      "loss": 0.7748,
+      "step": 2005
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4048955468343887,
+      "learning_rate": 9.748349093779327e-06,
+      "loss": 0.6684,
+      "step": 2006
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4416824397245391,
+      "learning_rate": 9.748024309380256e-06,
+      "loss": 0.6732,
+      "step": 2007
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2237990219850472,
+      "learning_rate": 9.747699320948006e-06,
+      "loss": 0.694,
+      "step": 2008
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3572977516585167,
+      "learning_rate": 9.747374128496541e-06,
+      "loss": 0.7802,
+      "step": 2009
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.343654384555796,
+      "learning_rate": 9.747048732039837e-06,
+      "loss": 0.6699,
+      "step": 2010
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9257069544979553,
+      "learning_rate": 9.746723131591878e-06,
+      "loss": 0.7939,
+      "step": 2011
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.238541838689657,
+      "learning_rate": 9.746397327166653e-06,
+      "loss": 0.6354,
+      "step": 2012
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4219674766330124,
+      "learning_rate": 9.746071318778165e-06,
+      "loss": 0.7951,
+      "step": 2013
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5136345312811246,
+      "learning_rate": 9.745745106440422e-06,
+      "loss": 0.9017,
+      "step": 2014
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2367457401702182,
+      "learning_rate": 9.745418690167444e-06,
+      "loss": 0.6923,
+      "step": 2015
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4678765346773321,
+      "learning_rate": 9.745092069973254e-06,
+      "loss": 0.6928,
+      "step": 2016
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.377859213441562,
+      "learning_rate": 9.74476524587189e-06,
+      "loss": 0.7377,
+      "step": 2017
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2733213691535807,
+      "learning_rate": 9.744438217877396e-06,
+      "loss": 0.635,
+      "step": 2018
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.343142890743768,
+      "learning_rate": 9.744110986003826e-06,
+      "loss": 0.6893,
+      "step": 2019
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.178776440732641,
+      "learning_rate": 9.743783550265242e-06,
+      "loss": 0.7098,
+      "step": 2020
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3765438290126355,
+      "learning_rate": 9.743455910675712e-06,
+      "loss": 0.7937,
+      "step": 2021
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1326767467480292,
+      "learning_rate": 9.74312806724932e-06,
+      "loss": 0.5826,
+      "step": 2022
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.134958513994432,
+      "learning_rate": 9.74280002000015e-06,
+      "loss": 0.6717,
+      "step": 2023
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3511537221587395,
+      "learning_rate": 9.742471768942299e-06,
+      "loss": 0.7328,
+      "step": 2024
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4525660558430264,
+      "learning_rate": 9.742143314089877e-06,
+      "loss": 0.7046,
+      "step": 2025
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.502267713253663,
+      "learning_rate": 9.741814655456994e-06,
+      "loss": 0.668,
+      "step": 2026
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3888914731988529,
+      "learning_rate": 9.741485793057774e-06,
+      "loss": 0.7753,
+      "step": 2027
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3917833255205765,
+      "learning_rate": 9.74115672690635e-06,
+      "loss": 0.732,
+      "step": 2028
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.418028939641859,
+      "learning_rate": 9.740827457016863e-06,
+      "loss": 0.8706,
+      "step": 2029
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1335976742412144,
+      "learning_rate": 9.74049798340346e-06,
+      "loss": 0.694,
+      "step": 2030
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3363141541340087,
+      "learning_rate": 9.740168306080303e-06,
+      "loss": 0.7168,
+      "step": 2031
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3594160142107548,
+      "learning_rate": 9.739838425061557e-06,
+      "loss": 0.6886,
+      "step": 2032
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3368762779067307,
+      "learning_rate": 9.739508340361397e-06,
+      "loss": 0.7907,
+      "step": 2033
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6363945250897882,
+      "learning_rate": 9.739178051994008e-06,
+      "loss": 0.8184,
+      "step": 2034
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1319095275319644,
+      "learning_rate": 9.738847559973584e-06,
+      "loss": 0.7747,
+      "step": 2035
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.388476005325695,
+      "learning_rate": 9.738516864314324e-06,
+      "loss": 0.6763,
+      "step": 2036
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3302766785657132,
+      "learning_rate": 9.738185965030444e-06,
+      "loss": 0.7202,
+      "step": 2037
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2726466500052973,
+      "learning_rate": 9.73785486213616e-06,
+      "loss": 0.67,
+      "step": 2038
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3505921969866446,
+      "learning_rate": 9.7375235556457e-06,
+      "loss": 0.7773,
+      "step": 2039
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3508454166850044,
+      "learning_rate": 9.7371920455733e-06,
+      "loss": 0.6871,
+      "step": 2040
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3388670091338304,
+      "learning_rate": 9.736860331933212e-06,
+      "loss": 0.7614,
+      "step": 2041
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.441089025221432,
+      "learning_rate": 9.736528414739681e-06,
+      "loss": 0.7768,
+      "step": 2042
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.319518572739601,
+      "learning_rate": 9.736196294006978e-06,
+      "loss": 0.7687,
+      "step": 2043
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3642369138110901,
+      "learning_rate": 9.735863969749373e-06,
+      "loss": 0.6641,
+      "step": 2044
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1604182216489614,
+      "learning_rate": 9.735531441981145e-06,
+      "loss": 0.6163,
+      "step": 2045
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4206895116061837,
+      "learning_rate": 9.735198710716585e-06,
+      "loss": 0.7347,
+      "step": 2046
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4141147029116459,
+      "learning_rate": 9.734865775969988e-06,
+      "loss": 0.8509,
+      "step": 2047
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5138362609321832,
+      "learning_rate": 9.734532637755665e-06,
+      "loss": 0.7931,
+      "step": 2048
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.252870149939792,
+      "learning_rate": 9.734199296087932e-06,
+      "loss": 0.6464,
+      "step": 2049
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3586604160783822,
+      "learning_rate": 9.73386575098111e-06,
+      "loss": 0.7039,
+      "step": 2050
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7224425584184548,
+      "learning_rate": 9.733532002449533e-06,
+      "loss": 0.8471,
+      "step": 2051
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6667845995677135,
+      "learning_rate": 9.733198050507544e-06,
+      "loss": 0.7656,
+      "step": 2052
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3385423754675865,
+      "learning_rate": 9.732863895169496e-06,
+      "loss": 0.6943,
+      "step": 2053
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.643952912417638,
+      "learning_rate": 9.732529536449741e-06,
+      "loss": 0.7912,
+      "step": 2054
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4050144195145295,
+      "learning_rate": 9.732194974362657e-06,
+      "loss": 0.7338,
+      "step": 2055
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1715518870570376,
+      "learning_rate": 9.731860208922612e-06,
+      "loss": 0.6914,
+      "step": 2056
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.416776520479726,
+      "learning_rate": 9.731525240143998e-06,
+      "loss": 0.6928,
+      "step": 2057
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4871003976916308,
+      "learning_rate": 9.731190068041205e-06,
+      "loss": 0.7226,
+      "step": 2058
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1283663774286343,
+      "learning_rate": 9.730854692628637e-06,
+      "loss": 0.7221,
+      "step": 2059
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2626461847684687,
+      "learning_rate": 9.730519113920708e-06,
+      "loss": 0.7045,
+      "step": 2060
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3937647910012372,
+      "learning_rate": 9.730183331931838e-06,
+      "loss": 0.7384,
+      "step": 2061
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4834508128738886,
+      "learning_rate": 9.729847346676455e-06,
+      "loss": 0.7901,
+      "step": 2062
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3629506853958282,
+      "learning_rate": 9.729511158168999e-06,
+      "loss": 0.6431,
+      "step": 2063
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8956458327859211,
+      "learning_rate": 9.729174766423912e-06,
+      "loss": 0.6579,
+      "step": 2064
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3838433817133615,
+      "learning_rate": 9.728838171455655e-06,
+      "loss": 0.6951,
+      "step": 2065
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2452024858836765,
+      "learning_rate": 9.72850137327869e-06,
+      "loss": 0.614,
+      "step": 2066
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.321857033401993,
+      "learning_rate": 9.728164371907489e-06,
+      "loss": 0.7252,
+      "step": 2067
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4620004847662509,
+      "learning_rate": 9.727827167356537e-06,
+      "loss": 0.8119,
+      "step": 2068
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5256976821778472,
+      "learning_rate": 9.72748975964032e-06,
+      "loss": 0.7709,
+      "step": 2069
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5931419550197627,
+      "learning_rate": 9.72715214877334e-06,
+      "loss": 0.8471,
+      "step": 2070
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4645032527943045,
+      "learning_rate": 9.726814334770106e-06,
+      "loss": 0.7784,
+      "step": 2071
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3854385428544438,
+      "learning_rate": 9.72647631764513e-06,
+      "loss": 0.7229,
+      "step": 2072
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2515102397551,
+      "learning_rate": 9.72613809741294e-06,
+      "loss": 0.7292,
+      "step": 2073
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3268616729171148,
+      "learning_rate": 9.725799674088072e-06,
+      "loss": 0.6417,
+      "step": 2074
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3719393642242506,
+      "learning_rate": 9.725461047685067e-06,
+      "loss": 0.7157,
+      "step": 2075
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3749245567691055,
+      "learning_rate": 9.725122218218477e-06,
+      "loss": 0.7386,
+      "step": 2076
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3211757042524874,
+      "learning_rate": 9.72478318570286e-06,
+      "loss": 0.6817,
+      "step": 2077
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4949615501484306,
+      "learning_rate": 9.724443950152789e-06,
+      "loss": 0.7538,
+      "step": 2078
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3579143827816267,
+      "learning_rate": 9.724104511582838e-06,
+      "loss": 0.7218,
+      "step": 2079
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4218840444270733,
+      "learning_rate": 9.723764870007596e-06,
+      "loss": 0.8252,
+      "step": 2080
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.480752758281894,
+      "learning_rate": 9.723425025441657e-06,
+      "loss": 0.7773,
+      "step": 2081
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2413083609615976,
+      "learning_rate": 9.723084977899624e-06,
+      "loss": 0.6915,
+      "step": 2082
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3434891458494709,
+      "learning_rate": 9.722744727396111e-06,
+      "loss": 0.8093,
+      "step": 2083
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4125719203761422,
+      "learning_rate": 9.72240427394574e-06,
+      "loss": 0.6673,
+      "step": 2084
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4645383198721649,
+      "learning_rate": 9.72206361756314e-06,
+      "loss": 0.7402,
+      "step": 2085
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3488534192362116,
+      "learning_rate": 9.721722758262948e-06,
+      "loss": 0.6404,
+      "step": 2086
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.090707189601462,
+      "learning_rate": 9.721381696059815e-06,
+      "loss": 0.7568,
+      "step": 2087
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2441883459207033,
+      "learning_rate": 9.721040430968393e-06,
+      "loss": 0.6286,
+      "step": 2088
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5541078782621158,
+      "learning_rate": 9.720698963003351e-06,
+      "loss": 0.7003,
+      "step": 2089
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4443439579911934,
+      "learning_rate": 9.720357292179363e-06,
+      "loss": 0.6955,
+      "step": 2090
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.130449948475184,
+      "learning_rate": 9.720015418511108e-06,
+      "loss": 0.7783,
+      "step": 2091
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4926896849567082,
+      "learning_rate": 9.719673342013277e-06,
+      "loss": 0.8035,
+      "step": 2092
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2500574364640271,
+      "learning_rate": 9.719331062700572e-06,
+      "loss": 0.6903,
+      "step": 2093
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9872685919945264,
+      "learning_rate": 9.7189885805877e-06,
+      "loss": 0.6569,
+      "step": 2094
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2321724209757248,
+      "learning_rate": 9.71864589568938e-06,
+      "loss": 0.6725,
+      "step": 2095
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4100233338643289,
+      "learning_rate": 9.718303008020338e-06,
+      "loss": 0.682,
+      "step": 2096
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4554605088131929,
+      "learning_rate": 9.717959917595305e-06,
+      "loss": 0.7985,
+      "step": 2097
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1013309845090828,
+      "learning_rate": 9.71761662442903e-06,
+      "loss": 0.6646,
+      "step": 2098
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.810164968039559,
+      "learning_rate": 9.717273128536259e-06,
+      "loss": 0.6412,
+      "step": 2099
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3983146979072087,
+      "learning_rate": 9.716929429931757e-06,
+      "loss": 0.7559,
+      "step": 2100
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2362992464613785,
+      "learning_rate": 9.716585528630291e-06,
+      "loss": 0.6768,
+      "step": 2101
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.242030099056653,
+      "learning_rate": 9.716241424646642e-06,
+      "loss": 0.7327,
+      "step": 2102
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3072065194060882,
+      "learning_rate": 9.715897117995594e-06,
+      "loss": 0.6382,
+      "step": 2103
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.248101199637205,
+      "learning_rate": 9.715552608691944e-06,
+      "loss": 0.6846,
+      "step": 2104
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.367403170719709,
+      "learning_rate": 9.715207896750499e-06,
+      "loss": 0.6757,
+      "step": 2105
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.340556146811684,
+      "learning_rate": 9.714862982186067e-06,
+      "loss": 0.6623,
+      "step": 2106
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.163406836688686,
+      "learning_rate": 9.714517865013473e-06,
+      "loss": 0.6326,
+      "step": 2107
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3078853877633336,
+      "learning_rate": 9.714172545247544e-06,
+      "loss": 0.659,
+      "step": 2108
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.388735161722494,
+      "learning_rate": 9.713827022903124e-06,
+      "loss": 0.6475,
+      "step": 2109
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.272344688131656,
+      "learning_rate": 9.71348129799506e-06,
+      "loss": 0.6763,
+      "step": 2110
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3210216452798138,
+      "learning_rate": 9.713135370538204e-06,
+      "loss": 0.7403,
+      "step": 2111
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.512285459469557,
+      "learning_rate": 9.712789240547426e-06,
+      "loss": 0.755,
+      "step": 2112
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.123550742533629,
+      "learning_rate": 9.712442908037599e-06,
+      "loss": 0.608,
+      "step": 2113
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7205003035196493,
+      "learning_rate": 9.712096373023603e-06,
+      "loss": 0.8525,
+      "step": 2114
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3596703625859476,
+      "learning_rate": 9.711749635520333e-06,
+      "loss": 0.677,
+      "step": 2115
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.343588629005641,
+      "learning_rate": 9.711402695542689e-06,
+      "loss": 0.7275,
+      "step": 2116
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3815561846617885,
+      "learning_rate": 9.711055553105575e-06,
+      "loss": 0.7477,
+      "step": 2117
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.435391015275193,
+      "learning_rate": 9.710708208223914e-06,
+      "loss": 0.7922,
+      "step": 2118
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.460707907873126,
+      "learning_rate": 9.710360660912629e-06,
+      "loss": 0.7988,
+      "step": 2119
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2184536024766452,
+      "learning_rate": 9.710012911186657e-06,
+      "loss": 0.6724,
+      "step": 2120
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2589992897943254,
+      "learning_rate": 9.70966495906094e-06,
+      "loss": 0.7356,
+      "step": 2121
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3450875016489898,
+      "learning_rate": 9.70931680455043e-06,
+      "loss": 0.7064,
+      "step": 2122
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4241917959677919,
+      "learning_rate": 9.708968447670088e-06,
+      "loss": 0.7235,
+      "step": 2123
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0953441079009265,
+      "learning_rate": 9.708619888434887e-06,
+      "loss": 0.5204,
+      "step": 2124
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.514504705083395,
+      "learning_rate": 9.708271126859801e-06,
+      "loss": 0.938,
+      "step": 2125
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2899814114245884,
+      "learning_rate": 9.70792216295982e-06,
+      "loss": 0.6707,
+      "step": 2126
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6641432571990542,
+      "learning_rate": 9.707572996749936e-06,
+      "loss": 0.7272,
+      "step": 2127
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.291411802101597,
+      "learning_rate": 9.707223628245157e-06,
+      "loss": 0.6454,
+      "step": 2128
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1526965255951893,
+      "learning_rate": 9.706874057460497e-06,
+      "loss": 0.5751,
+      "step": 2129
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2452085613049642,
+      "learning_rate": 9.706524284410973e-06,
+      "loss": 0.6961,
+      "step": 2130
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1979676236684609,
+      "learning_rate": 9.70617430911162e-06,
+      "loss": 0.6531,
+      "step": 2131
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.102203951363217,
+      "learning_rate": 9.705824131577476e-06,
+      "loss": 0.6145,
+      "step": 2132
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4212976020321957,
+      "learning_rate": 9.70547375182359e-06,
+      "loss": 0.6183,
+      "step": 2133
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0001744071467855,
+      "learning_rate": 9.705123169865016e-06,
+      "loss": 0.5796,
+      "step": 2134
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5308793457897698,
+      "learning_rate": 9.70477238571682e-06,
+      "loss": 0.8175,
+      "step": 2135
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3476423807624154,
+      "learning_rate": 9.704421399394077e-06,
+      "loss": 0.7601,
+      "step": 2136
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6599717655959412,
+      "learning_rate": 9.704070210911868e-06,
+      "loss": 0.7771,
+      "step": 2137
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1768727919502622,
+      "learning_rate": 9.703718820285288e-06,
+      "loss": 0.6911,
+      "step": 2138
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1630262455259792,
+      "learning_rate": 9.703367227529432e-06,
+      "loss": 0.5486,
+      "step": 2139
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.39184946056254,
+      "learning_rate": 9.703015432659413e-06,
+      "loss": 0.6803,
+      "step": 2140
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5560602296294819,
+      "learning_rate": 9.702663435690344e-06,
+      "loss": 0.8864,
+      "step": 2141
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.889460771549115,
+      "learning_rate": 9.702311236637357e-06,
+      "loss": 0.652,
+      "step": 2142
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3088886338512054,
+      "learning_rate": 9.701958835515582e-06,
+      "loss": 0.7727,
+      "step": 2143
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.263431653479495,
+      "learning_rate": 9.701606232340165e-06,
+      "loss": 0.6214,
+      "step": 2144
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2804802601264937,
+      "learning_rate": 9.701253427126257e-06,
+      "loss": 0.7861,
+      "step": 2145
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3763893818720145,
+      "learning_rate": 9.700900419889018e-06,
+      "loss": 0.6551,
+      "step": 2146
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3300159466996502,
+      "learning_rate": 9.700547210643618e-06,
+      "loss": 0.711,
+      "step": 2147
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4572814576067947,
+      "learning_rate": 9.700193799405238e-06,
+      "loss": 0.7737,
+      "step": 2148
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2097048534576145,
+      "learning_rate": 9.699840186189061e-06,
+      "loss": 0.6987,
+      "step": 2149
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6277289664139651,
+      "learning_rate": 9.699486371010285e-06,
+      "loss": 0.8054,
+      "step": 2150
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2780430775747431,
+      "learning_rate": 9.69913235388411e-06,
+      "loss": 0.7412,
+      "step": 2151
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4831013399551565,
+      "learning_rate": 9.698778134825755e-06,
+      "loss": 0.8117,
+      "step": 2152
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.385359750711335,
+      "learning_rate": 9.69842371385044e-06,
+      "loss": 0.7506,
+      "step": 2153
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4615303132444604,
+      "learning_rate": 9.698069090973391e-06,
+      "loss": 0.634,
+      "step": 2154
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3269714222886204,
+      "learning_rate": 9.697714266209853e-06,
+      "loss": 0.6321,
+      "step": 2155
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2764135548156283,
+      "learning_rate": 9.697359239575069e-06,
+      "loss": 0.7545,
+      "step": 2156
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3348778602203646,
+      "learning_rate": 9.697004011084298e-06,
+      "loss": 0.7598,
+      "step": 2157
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6129636406341399,
+      "learning_rate": 9.696648580752803e-06,
+      "loss": 0.7484,
+      "step": 2158
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4053194141681078,
+      "learning_rate": 9.696292948595857e-06,
+      "loss": 0.7562,
+      "step": 2159
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2648021812079524,
+      "learning_rate": 9.695937114628746e-06,
+      "loss": 0.6323,
+      "step": 2160
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5745567167574872,
+      "learning_rate": 9.695581078866759e-06,
+      "loss": 0.7222,
+      "step": 2161
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3165587479486276,
+      "learning_rate": 9.695224841325194e-06,
+      "loss": 0.7147,
+      "step": 2162
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2665289989117934,
+      "learning_rate": 9.694868402019362e-06,
+      "loss": 0.6042,
+      "step": 2163
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2960155345153597,
+      "learning_rate": 9.694511760964578e-06,
+      "loss": 0.6511,
+      "step": 2164
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5187550330243857,
+      "learning_rate": 9.694154918176168e-06,
+      "loss": 0.6867,
+      "step": 2165
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.552852428976057,
+      "learning_rate": 9.693797873669466e-06,
+      "loss": 0.7536,
+      "step": 2166
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4451504677663887,
+      "learning_rate": 9.693440627459817e-06,
+      "loss": 0.7479,
+      "step": 2167
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5875276456479803,
+      "learning_rate": 9.693083179562573e-06,
+      "loss": 0.8711,
+      "step": 2168
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4878094937994444,
+      "learning_rate": 9.69272552999309e-06,
+      "loss": 0.7585,
+      "step": 2169
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0963773441643316,
+      "learning_rate": 9.69236767876674e-06,
+      "loss": 0.5516,
+      "step": 2170
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4593163012667003,
+      "learning_rate": 9.692009625898902e-06,
+      "loss": 0.7448,
+      "step": 2171
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1589302932284153,
+      "learning_rate": 9.69165137140496e-06,
+      "loss": 0.6677,
+      "step": 2172
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.125352564796548,
+      "learning_rate": 9.69129291530031e-06,
+      "loss": 0.7798,
+      "step": 2173
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3068199847010278,
+      "learning_rate": 9.690934257600353e-06,
+      "loss": 0.6945,
+      "step": 2174
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.556727934919251,
+      "learning_rate": 9.690575398320504e-06,
+      "loss": 0.8108,
+      "step": 2175
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4016183648095744,
+      "learning_rate": 9.690216337476188e-06,
+      "loss": 0.7729,
+      "step": 2176
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2605783051283794,
+      "learning_rate": 9.689857075082828e-06,
+      "loss": 0.6745,
+      "step": 2177
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.315185640945685,
+      "learning_rate": 9.689497611155862e-06,
+      "loss": 0.6251,
+      "step": 2178
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.412925795136165,
+      "learning_rate": 9.689137945710742e-06,
+      "loss": 0.6879,
+      "step": 2179
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3064105451866659,
+      "learning_rate": 9.688778078762922e-06,
+      "loss": 0.764,
+      "step": 2180
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5166172784792988,
+      "learning_rate": 9.688418010327862e-06,
+      "loss": 0.7435,
+      "step": 2181
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3093983047135644,
+      "learning_rate": 9.688057740421042e-06,
+      "loss": 0.6743,
+      "step": 2182
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4981177151378535,
+      "learning_rate": 9.68769726905794e-06,
+      "loss": 0.6463,
+      "step": 2183
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.317278249630831,
+      "learning_rate": 9.687336596254045e-06,
+      "loss": 0.6375,
+      "step": 2184
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5163007701076725,
+      "learning_rate": 9.686975722024856e-06,
+      "loss": 0.8449,
+      "step": 2185
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4221074187384883,
+      "learning_rate": 9.686614646385883e-06,
+      "loss": 0.6795,
+      "step": 2186
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4942217252992578,
+      "learning_rate": 9.68625336935264e-06,
+      "loss": 0.8671,
+      "step": 2187
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2105223567765395,
+      "learning_rate": 9.685891890940655e-06,
+      "loss": 0.6977,
+      "step": 2188
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3126146894775699,
+      "learning_rate": 9.685530211165459e-06,
+      "loss": 0.6529,
+      "step": 2189
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.455007880668908,
+      "learning_rate": 9.685168330042594e-06,
+      "loss": 0.7599,
+      "step": 2190
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5043474087817539,
+      "learning_rate": 9.68480624758761e-06,
+      "loss": 0.8058,
+      "step": 2191
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2698451471789092,
+      "learning_rate": 9.68444396381607e-06,
+      "loss": 0.6949,
+      "step": 2192
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2288248490366764,
+      "learning_rate": 9.68408147874354e-06,
+      "loss": 0.7195,
+      "step": 2193
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3218652954388772,
+      "learning_rate": 9.683718792385595e-06,
+      "loss": 0.6939,
+      "step": 2194
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.253469635492087,
+      "learning_rate": 9.683355904757822e-06,
+      "loss": 0.7151,
+      "step": 2195
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3827215478711008,
+      "learning_rate": 9.682992815875818e-06,
+      "loss": 0.603,
+      "step": 2196
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9840639974496341,
+      "learning_rate": 9.682629525755182e-06,
+      "loss": 0.8083,
+      "step": 2197
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3230011239493078,
+      "learning_rate": 9.682266034411527e-06,
+      "loss": 0.7349,
+      "step": 2198
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4780803709620198,
+      "learning_rate": 9.681902341860471e-06,
+      "loss": 0.7141,
+      "step": 2199
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5426037747671353,
+      "learning_rate": 9.681538448117647e-06,
+      "loss": 0.6961,
+      "step": 2200
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4210317156546661,
+      "learning_rate": 9.681174353198687e-06,
+      "loss": 0.8599,
+      "step": 2201
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4570728268542528,
+      "learning_rate": 9.68081005711924e-06,
+      "loss": 0.8166,
+      "step": 2202
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3561395671823742,
+      "learning_rate": 9.68044555989496e-06,
+      "loss": 0.6553,
+      "step": 2203
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4380993190144884,
+      "learning_rate": 9.680080861541511e-06,
+      "loss": 0.8198,
+      "step": 2204
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3138403910479808,
+      "learning_rate": 9.679715962074566e-06,
+      "loss": 0.7433,
+      "step": 2205
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5305855700096493,
+      "learning_rate": 9.6793508615098e-06,
+      "loss": 0.8448,
+      "step": 2206
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6067025477877188,
+      "learning_rate": 9.678985559862909e-06,
+      "loss": 0.7967,
+      "step": 2207
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4745796951859074,
+      "learning_rate": 9.678620057149586e-06,
+      "loss": 0.6476,
+      "step": 2208
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3672414169666083,
+      "learning_rate": 9.678254353385538e-06,
+      "loss": 0.7406,
+      "step": 2209
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2787091251589753,
+      "learning_rate": 9.677888448586485e-06,
+      "loss": 0.6477,
+      "step": 2210
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2043144710770024,
+      "learning_rate": 9.677522342768145e-06,
+      "loss": 0.7613,
+      "step": 2211
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3165232304301542,
+      "learning_rate": 9.677156035946253e-06,
+      "loss": 0.8013,
+      "step": 2212
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6217769782874296,
+      "learning_rate": 9.67678952813655e-06,
+      "loss": 0.7886,
+      "step": 2213
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.642561272456916,
+      "learning_rate": 9.676422819354785e-06,
+      "loss": 0.7595,
+      "step": 2214
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.262324310051342,
+      "learning_rate": 9.676055909616716e-06,
+      "loss": 0.6585,
+      "step": 2215
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2787115575984302,
+      "learning_rate": 9.675688798938112e-06,
+      "loss": 0.7492,
+      "step": 2216
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4672774873294794,
+      "learning_rate": 9.675321487334744e-06,
+      "loss": 0.648,
+      "step": 2217
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5601417385406477,
+      "learning_rate": 9.674953974822402e-06,
+      "loss": 0.8727,
+      "step": 2218
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4532420391660805,
+      "learning_rate": 9.674586261416874e-06,
+      "loss": 0.6974,
+      "step": 2219
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3469167476563475,
+      "learning_rate": 9.674218347133965e-06,
+      "loss": 0.6727,
+      "step": 2220
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3964384147307938,
+      "learning_rate": 9.673850231989484e-06,
+      "loss": 0.745,
+      "step": 2221
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2461604334942586,
+      "learning_rate": 9.673481915999249e-06,
+      "loss": 0.6563,
+      "step": 2222
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4334991868283615,
+      "learning_rate": 9.673113399179085e-06,
+      "loss": 0.719,
+      "step": 2223
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.455840668941861,
+      "learning_rate": 9.672744681544834e-06,
+      "loss": 0.7652,
+      "step": 2224
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5113797292493734,
+      "learning_rate": 9.672375763112335e-06,
+      "loss": 0.7578,
+      "step": 2225
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.294058446462369,
+      "learning_rate": 9.672006643897444e-06,
+      "loss": 0.7218,
+      "step": 2226
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4285884303809662,
+      "learning_rate": 9.671637323916023e-06,
+      "loss": 0.7384,
+      "step": 2227
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.297002144910436,
+      "learning_rate": 9.671267803183942e-06,
+      "loss": 0.7482,
+      "step": 2228
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1029956249157788,
+      "learning_rate": 9.670898081717079e-06,
+      "loss": 0.6676,
+      "step": 2229
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3890382308527478,
+      "learning_rate": 9.670528159531325e-06,
+      "loss": 0.7307,
+      "step": 2230
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4053874413243972,
+      "learning_rate": 9.670158036642572e-06,
+      "loss": 0.9821,
+      "step": 2231
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.208928651765419,
+      "learning_rate": 9.669787713066728e-06,
+      "loss": 0.7685,
+      "step": 2232
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4125462716160646,
+      "learning_rate": 9.669417188819704e-06,
+      "loss": 0.7262,
+      "step": 2233
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4344827907809388,
+      "learning_rate": 9.669046463917427e-06,
+      "loss": 0.7471,
+      "step": 2234
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3652124293580232,
+      "learning_rate": 9.668675538375823e-06,
+      "loss": 0.6934,
+      "step": 2235
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.617250793437528,
+      "learning_rate": 9.668304412210835e-06,
+      "loss": 0.7044,
+      "step": 2236
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0871406908848529,
+      "learning_rate": 9.667933085438407e-06,
+      "loss": 0.6334,
+      "step": 2237
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.367157994562311,
+      "learning_rate": 9.6675615580745e-06,
+      "loss": 0.7425,
+      "step": 2238
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5833512534152672,
+      "learning_rate": 9.667189830135078e-06,
+      "loss": 0.8157,
+      "step": 2239
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1996893861549818,
+      "learning_rate": 9.666817901636115e-06,
+      "loss": 0.6965,
+      "step": 2240
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3770579507641916,
+      "learning_rate": 9.666445772593592e-06,
+      "loss": 0.6722,
+      "step": 2241
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5115929160913848,
+      "learning_rate": 9.666073443023502e-06,
+      "loss": 0.8226,
+      "step": 2242
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.457751679396775,
+      "learning_rate": 9.665700912941844e-06,
+      "loss": 0.7002,
+      "step": 2243
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5241978500905382,
+      "learning_rate": 9.665328182364627e-06,
+      "loss": 0.6835,
+      "step": 2244
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.55340845614395,
+      "learning_rate": 9.664955251307867e-06,
+      "loss": 0.8451,
+      "step": 2245
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.400699198095677,
+      "learning_rate": 9.66458211978759e-06,
+      "loss": 0.6413,
+      "step": 2246
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.381436563811886,
+      "learning_rate": 9.664208787819833e-06,
+      "loss": 0.7308,
+      "step": 2247
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.702346173374171,
+      "learning_rate": 9.663835255420635e-06,
+      "loss": 0.7722,
+      "step": 2248
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6302739712382561,
+      "learning_rate": 9.663461522606049e-06,
+      "loss": 0.7137,
+      "step": 2249
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2880654928648,
+      "learning_rate": 9.663087589392135e-06,
+      "loss": 0.721,
+      "step": 2250
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4724837117586997,
+      "learning_rate": 9.662713455794963e-06,
+      "loss": 0.7687,
+      "step": 2251
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.367927589398692,
+      "learning_rate": 9.66233912183061e-06,
+      "loss": 0.6747,
+      "step": 2252
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.457963945718696,
+      "learning_rate": 9.66196458751516e-06,
+      "loss": 0.8662,
+      "step": 2253
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4407073212261479,
+      "learning_rate": 9.66158985286471e-06,
+      "loss": 0.7043,
+      "step": 2254
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2950818756697164,
+      "learning_rate": 9.661214917895361e-06,
+      "loss": 1.0308,
+      "step": 2255
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2924049187723903,
+      "learning_rate": 9.660839782623226e-06,
+      "loss": 0.6069,
+      "step": 2256
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5811881838460655,
+      "learning_rate": 9.660464447064427e-06,
+      "loss": 0.805,
+      "step": 2257
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4311991209608466,
+      "learning_rate": 9.660088911235088e-06,
+      "loss": 0.6846,
+      "step": 2258
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3305639605783894,
+      "learning_rate": 9.659713175151352e-06,
+      "loss": 0.6365,
+      "step": 2259
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6766701655932343,
+      "learning_rate": 9.659337238829364e-06,
+      "loss": 0.7659,
+      "step": 2260
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4905966425214114,
+      "learning_rate": 9.658961102285276e-06,
+      "loss": 0.728,
+      "step": 2261
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3083585995893525,
+      "learning_rate": 9.658584765535254e-06,
+      "loss": 0.9092,
+      "step": 2262
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3123848662351913,
+      "learning_rate": 9.65820822859547e-06,
+      "loss": 0.6558,
+      "step": 2263
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4833042021301597,
+      "learning_rate": 9.657831491482103e-06,
+      "loss": 0.775,
+      "step": 2264
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2648657334305804,
+      "learning_rate": 9.657454554211344e-06,
+      "loss": 0.6939,
+      "step": 2265
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.57677861779668,
+      "learning_rate": 9.657077416799392e-06,
+      "loss": 0.7509,
+      "step": 2266
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3074224418090599,
+      "learning_rate": 9.656700079262448e-06,
+      "loss": 0.6773,
+      "step": 2267
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3591198982650323,
+      "learning_rate": 9.656322541616734e-06,
+      "loss": 0.6653,
+      "step": 2268
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3536342550421705,
+      "learning_rate": 9.655944803878467e-06,
+      "loss": 0.6778,
+      "step": 2269
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2386846898841586,
+      "learning_rate": 9.655566866063885e-06,
+      "loss": 0.724,
+      "step": 2270
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2436073646328671,
+      "learning_rate": 9.655188728189225e-06,
+      "loss": 0.7357,
+      "step": 2271
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3366176258519202,
+      "learning_rate": 9.654810390270737e-06,
+      "loss": 0.6444,
+      "step": 2272
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0888308376163796,
+      "learning_rate": 9.654431852324682e-06,
+      "loss": 0.6533,
+      "step": 2273
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4441886104123995,
+      "learning_rate": 9.654053114367321e-06,
+      "loss": 0.7393,
+      "step": 2274
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1240664370715616,
+      "learning_rate": 9.653674176414936e-06,
+      "loss": 0.6302,
+      "step": 2275
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4950327911321661,
+      "learning_rate": 9.653295038483806e-06,
+      "loss": 0.8143,
+      "step": 2276
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.355251927125555,
+      "learning_rate": 9.652915700590224e-06,
+      "loss": 0.709,
+      "step": 2277
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.241570949098876,
+      "learning_rate": 9.652536162750492e-06,
+      "loss": 0.7079,
+      "step": 2278
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4281199898711345,
+      "learning_rate": 9.65215642498092e-06,
+      "loss": 0.7088,
+      "step": 2279
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3142930266961088,
+      "learning_rate": 9.651776487297823e-06,
+      "loss": 0.664,
+      "step": 2280
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.25215204533141,
+      "learning_rate": 9.651396349717533e-06,
+      "loss": 0.6749,
+      "step": 2281
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.354681373331994,
+      "learning_rate": 9.651016012256382e-06,
+      "loss": 0.7301,
+      "step": 2282
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3104859656854522,
+      "learning_rate": 9.650635474930716e-06,
+      "loss": 0.7221,
+      "step": 2283
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5324462303800503,
+      "learning_rate": 9.650254737756883e-06,
+      "loss": 0.7095,
+      "step": 2284
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7725507744615125,
+      "learning_rate": 9.64987380075125e-06,
+      "loss": 0.6667,
+      "step": 2285
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1803465520581269,
+      "learning_rate": 9.649492663930182e-06,
+      "loss": 0.5664,
+      "step": 2286
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5219801808673303,
+      "learning_rate": 9.64911132731006e-06,
+      "loss": 0.6435,
+      "step": 2287
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7242492849114062,
+      "learning_rate": 9.648729790907271e-06,
+      "loss": 0.8453,
+      "step": 2288
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6519773391270627,
+      "learning_rate": 9.648348054738208e-06,
+      "loss": 0.8512,
+      "step": 2289
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3684089756402429,
+      "learning_rate": 9.647966118819278e-06,
+      "loss": 0.7718,
+      "step": 2290
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2217938449606833,
+      "learning_rate": 9.647583983166892e-06,
+      "loss": 0.6152,
+      "step": 2291
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.484162460911011,
+      "learning_rate": 9.647201647797472e-06,
+      "loss": 0.8316,
+      "step": 2292
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3832212639400643,
+      "learning_rate": 9.646819112727445e-06,
+      "loss": 0.7476,
+      "step": 2293
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2817185233510429,
+      "learning_rate": 9.646436377973253e-06,
+      "loss": 0.7248,
+      "step": 2294
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5627385122543895,
+      "learning_rate": 9.646053443551344e-06,
+      "loss": 0.7029,
+      "step": 2295
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2326105629775925,
+      "learning_rate": 9.64567030947817e-06,
+      "loss": 0.6584,
+      "step": 2296
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7948752951508269,
+      "learning_rate": 9.645286975770194e-06,
+      "loss": 0.7489,
+      "step": 2297
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4061296444240983,
+      "learning_rate": 9.644903442443894e-06,
+      "loss": 0.7367,
+      "step": 2298
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.158583547286112,
+      "learning_rate": 9.644519709515746e-06,
+      "loss": 0.6462,
+      "step": 2299
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4976327295514498,
+      "learning_rate": 9.644135777002244e-06,
+      "loss": 0.7585,
+      "step": 2300
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2232563822975355,
+      "learning_rate": 9.643751644919883e-06,
+      "loss": 0.6163,
+      "step": 2301
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4655901607494326,
+      "learning_rate": 9.643367313285172e-06,
+      "loss": 0.7398,
+      "step": 2302
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4875784585164795,
+      "learning_rate": 9.642982782114628e-06,
+      "loss": 0.7315,
+      "step": 2303
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2511543581606501,
+      "learning_rate": 9.642598051424772e-06,
+      "loss": 0.5813,
+      "step": 2304
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.285679145584096,
+      "learning_rate": 9.642213121232137e-06,
+      "loss": 0.6854,
+      "step": 2305
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2013941691083325,
+      "learning_rate": 9.641827991553267e-06,
+      "loss": 0.5801,
+      "step": 2306
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0657822279207845,
+      "learning_rate": 9.64144266240471e-06,
+      "loss": 0.6687,
+      "step": 2307
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4775947723407492,
+      "learning_rate": 9.641057133803024e-06,
+      "loss": 0.7384,
+      "step": 2308
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0348467302195457,
+      "learning_rate": 9.640671405764777e-06,
+      "loss": 0.5996,
+      "step": 2309
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3409286606343047,
+      "learning_rate": 9.640285478306546e-06,
+      "loss": 0.6757,
+      "step": 2310
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3652323030504345,
+      "learning_rate": 9.639899351444911e-06,
+      "loss": 0.665,
+      "step": 2311
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.380849179265484,
+      "learning_rate": 9.63951302519647e-06,
+      "loss": 0.6726,
+      "step": 2312
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4174795532952258,
+      "learning_rate": 9.639126499577818e-06,
+      "loss": 0.7553,
+      "step": 2313
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.962173046842687,
+      "learning_rate": 9.638739774605572e-06,
+      "loss": 0.7453,
+      "step": 2314
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4519417227744007,
+      "learning_rate": 9.638352850296344e-06,
+      "loss": 0.6808,
+      "step": 2315
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.593819565497597,
+      "learning_rate": 9.637965726666766e-06,
+      "loss": 0.6977,
+      "step": 2316
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.398246236020521,
+      "learning_rate": 9.63757840373347e-06,
+      "loss": 0.6503,
+      "step": 2317
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.369417303444511,
+      "learning_rate": 9.637190881513103e-06,
+      "loss": 0.7524,
+      "step": 2318
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2246446678779535,
+      "learning_rate": 9.636803160022314e-06,
+      "loss": 0.7854,
+      "step": 2319
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6113270395391124,
+      "learning_rate": 9.63641523927777e-06,
+      "loss": 0.7585,
+      "step": 2320
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.02329182331502,
+      "learning_rate": 9.636027119296133e-06,
+      "loss": 0.7972,
+      "step": 2321
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4290051182661196,
+      "learning_rate": 9.63563880009409e-06,
+      "loss": 0.6677,
+      "step": 2322
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3940993469318206,
+      "learning_rate": 9.63525028168832e-06,
+      "loss": 0.6899,
+      "step": 2323
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5524061086507241,
+      "learning_rate": 9.634861564095525e-06,
+      "loss": 0.7195,
+      "step": 2324
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4379706089482172,
+      "learning_rate": 9.634472647332405e-06,
+      "loss": 0.819,
+      "step": 2325
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3935338511587732,
+      "learning_rate": 9.634083531415672e-06,
+      "loss": 0.6856,
+      "step": 2326
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3928502919403196,
+      "learning_rate": 9.633694216362051e-06,
+      "loss": 0.6903,
+      "step": 2327
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6197498860095307,
+      "learning_rate": 9.633304702188271e-06,
+      "loss": 0.6969,
+      "step": 2328
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4757602129430907,
+      "learning_rate": 9.632914988911066e-06,
+      "loss": 0.8331,
+      "step": 2329
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3479423909318222,
+      "learning_rate": 9.632525076547187e-06,
+      "loss": 0.7401,
+      "step": 2330
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3051375857504801,
+      "learning_rate": 9.632134965113389e-06,
+      "loss": 0.7392,
+      "step": 2331
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3120925799622196,
+      "learning_rate": 9.631744654626435e-06,
+      "loss": 0.7283,
+      "step": 2332
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2710793024570073,
+      "learning_rate": 9.631354145103098e-06,
+      "loss": 0.6876,
+      "step": 2333
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3994512257036613,
+      "learning_rate": 9.63096343656016e-06,
+      "loss": 0.7134,
+      "step": 2334
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.566096364585771,
+      "learning_rate": 9.630572529014406e-06,
+      "loss": 0.7476,
+      "step": 2335
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2638398743329757,
+      "learning_rate": 9.630181422482639e-06,
+      "loss": 0.6529,
+      "step": 2336
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5576297164999953,
+      "learning_rate": 9.629790116981665e-06,
+      "loss": 0.8809,
+      "step": 2337
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0635675081037217,
+      "learning_rate": 9.629398612528299e-06,
+      "loss": 0.6992,
+      "step": 2338
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2351381487054345,
+      "learning_rate": 9.629006909139363e-06,
+      "loss": 0.7074,
+      "step": 2339
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5735620919927074,
+      "learning_rate": 9.628615006831692e-06,
+      "loss": 0.7119,
+      "step": 2340
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4560012287807338,
+      "learning_rate": 9.628222905622124e-06,
+      "loss": 0.8128,
+      "step": 2341
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6411658920182202,
+      "learning_rate": 9.627830605527513e-06,
+      "loss": 0.7735,
+      "step": 2342
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5368448887856723,
+      "learning_rate": 9.627438106564711e-06,
+      "loss": 0.6553,
+      "step": 2343
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2883393724480463,
+      "learning_rate": 9.62704540875059e-06,
+      "loss": 0.82,
+      "step": 2344
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5393995226377042,
+      "learning_rate": 9.626652512102021e-06,
+      "loss": 0.8458,
+      "step": 2345
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0826945215179558,
+      "learning_rate": 9.62625941663589e-06,
+      "loss": 0.6831,
+      "step": 2346
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1492034098472215,
+      "learning_rate": 9.625866122369089e-06,
+      "loss": 0.6653,
+      "step": 2347
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3904616509329597,
+      "learning_rate": 9.62547262931852e-06,
+      "loss": 0.7147,
+      "step": 2348
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5112642358139752,
+      "learning_rate": 9.625078937501089e-06,
+      "loss": 0.8202,
+      "step": 2349
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2603894471185912,
+      "learning_rate": 9.624685046933715e-06,
+      "loss": 0.6921,
+      "step": 2350
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7067931305212245,
+      "learning_rate": 9.624290957633325e-06,
+      "loss": 0.6618,
+      "step": 2351
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3718010013701412,
+      "learning_rate": 9.623896669616855e-06,
+      "loss": 0.7563,
+      "step": 2352
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.271960147939651,
+      "learning_rate": 9.623502182901245e-06,
+      "loss": 0.7628,
+      "step": 2353
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5271977521909517,
+      "learning_rate": 9.62310749750345e-06,
+      "loss": 0.7176,
+      "step": 2354
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2395202651193191,
+      "learning_rate": 9.62271261344043e-06,
+      "loss": 0.7062,
+      "step": 2355
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.063961258612239,
+      "learning_rate": 9.622317530729153e-06,
+      "loss": 0.7152,
+      "step": 2356
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2464111112014373,
+      "learning_rate": 9.621922249386598e-06,
+      "loss": 0.6747,
+      "step": 2357
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5939350945499988,
+      "learning_rate": 9.621526769429749e-06,
+      "loss": 0.7298,
+      "step": 2358
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.318765804788519,
+      "learning_rate": 9.621131090875603e-06,
+      "loss": 0.5884,
+      "step": 2359
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.330301464889375,
+      "learning_rate": 9.620735213741163e-06,
+      "loss": 0.7646,
+      "step": 2360
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4478099856743516,
+      "learning_rate": 9.62033913804344e-06,
+      "loss": 0.8138,
+      "step": 2361
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2662052883285038,
+      "learning_rate": 9.619942863799454e-06,
+      "loss": 0.7158,
+      "step": 2362
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.612247397881965,
+      "learning_rate": 9.619546391026234e-06,
+      "loss": 0.7307,
+      "step": 2363
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1984019191422266,
+      "learning_rate": 9.619149719740817e-06,
+      "loss": 0.6912,
+      "step": 2364
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7612799408470239,
+      "learning_rate": 9.618752849960248e-06,
+      "loss": 0.7098,
+      "step": 2365
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2879472807336036,
+      "learning_rate": 9.618355781701584e-06,
+      "loss": 0.7194,
+      "step": 2366
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0866892144657758,
+      "learning_rate": 9.617958514981885e-06,
+      "loss": 0.6347,
+      "step": 2367
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4262112306863903,
+      "learning_rate": 9.617561049818225e-06,
+      "loss": 0.7592,
+      "step": 2368
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4637617457525287,
+      "learning_rate": 9.617163386227683e-06,
+      "loss": 0.7062,
+      "step": 2369
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.419399739600774,
+      "learning_rate": 9.616765524227348e-06,
+      "loss": 0.7935,
+      "step": 2370
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4169750622284913,
+      "learning_rate": 9.616367463834313e-06,
+      "loss": 0.8394,
+      "step": 2371
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1579582650775373,
+      "learning_rate": 9.61596920506569e-06,
+      "loss": 0.5666,
+      "step": 2372
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3855749174247012,
+      "learning_rate": 9.61557074793859e-06,
+      "loss": 0.7778,
+      "step": 2373
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8258565204801012,
+      "learning_rate": 9.615172092470134e-06,
+      "loss": 0.8355,
+      "step": 2374
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4298874619816397,
+      "learning_rate": 9.614773238677455e-06,
+      "loss": 0.6596,
+      "step": 2375
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2827621317571203,
+      "learning_rate": 9.614374186577693e-06,
+      "loss": 0.7573,
+      "step": 2376
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2923754438641142,
+      "learning_rate": 9.613974936187993e-06,
+      "loss": 0.7468,
+      "step": 2377
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3758466754189773,
+      "learning_rate": 9.613575487525517e-06,
+      "loss": 0.7173,
+      "step": 2378
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3163814950466146,
+      "learning_rate": 9.613175840607428e-06,
+      "loss": 0.7134,
+      "step": 2379
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1925269249139991,
+      "learning_rate": 9.612775995450896e-06,
+      "loss": 0.6312,
+      "step": 2380
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6695729362567044,
+      "learning_rate": 9.612375952073108e-06,
+      "loss": 0.8745,
+      "step": 2381
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.341757384569493,
+      "learning_rate": 9.611975710491252e-06,
+      "loss": 0.6698,
+      "step": 2382
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1474854245464512,
+      "learning_rate": 9.611575270722529e-06,
+      "loss": 0.6666,
+      "step": 2383
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4382537206671202,
+      "learning_rate": 9.611174632784147e-06,
+      "loss": 0.5917,
+      "step": 2384
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2691814853423888,
+      "learning_rate": 9.61077379669332e-06,
+      "loss": 0.6515,
+      "step": 2385
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.163032688020882,
+      "learning_rate": 9.610372762467275e-06,
+      "loss": 0.6322,
+      "step": 2386
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3900376257621818,
+      "learning_rate": 9.609971530123243e-06,
+      "loss": 0.8593,
+      "step": 2387
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3978085068481676,
+      "learning_rate": 9.60957009967847e-06,
+      "loss": 0.7129,
+      "step": 2388
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6004223713635615,
+      "learning_rate": 9.609168471150202e-06,
+      "loss": 0.7536,
+      "step": 2389
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2157256972437602,
+      "learning_rate": 9.6087666445557e-06,
+      "loss": 0.669,
+      "step": 2390
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1999840598431915,
+      "learning_rate": 9.60836461991223e-06,
+      "loss": 0.7708,
+      "step": 2391
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.174588303561155,
+      "learning_rate": 9.607962397237071e-06,
+      "loss": 0.6414,
+      "step": 2392
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4816599666341985,
+      "learning_rate": 9.607559976547504e-06,
+      "loss": 0.7089,
+      "step": 2393
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.354990123263862,
+      "learning_rate": 9.607157357860823e-06,
+      "loss": 0.7647,
+      "step": 2394
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3339874166412893,
+      "learning_rate": 9.606754541194329e-06,
+      "loss": 0.7168,
+      "step": 2395
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7879646584102429,
+      "learning_rate": 9.606351526565336e-06,
+      "loss": 0.795,
+      "step": 2396
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4147330900543689,
+      "learning_rate": 9.605948313991156e-06,
+      "loss": 0.7597,
+      "step": 2397
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3074271405963755,
+      "learning_rate": 9.60554490348912e-06,
+      "loss": 0.6915,
+      "step": 2398
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2746027963682627,
+      "learning_rate": 9.605141295076561e-06,
+      "loss": 0.6008,
+      "step": 2399
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3906767341417732,
+      "learning_rate": 9.604737488770826e-06,
+      "loss": 0.7051,
+      "step": 2400
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2738344450725994,
+      "learning_rate": 9.604333484589266e-06,
+      "loss": 0.672,
+      "step": 2401
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0876631792278706,
+      "learning_rate": 9.60392928254924e-06,
+      "loss": 0.6988,
+      "step": 2402
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3066972711534408,
+      "learning_rate": 9.60352488266812e-06,
+      "loss": 0.7477,
+      "step": 2403
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.267454345570645,
+      "learning_rate": 9.603120284963284e-06,
+      "loss": 0.6892,
+      "step": 2404
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2157250063301441,
+      "learning_rate": 9.602715489452117e-06,
+      "loss": 0.6928,
+      "step": 2405
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.275107621101732,
+      "learning_rate": 9.602310496152014e-06,
+      "loss": 0.6082,
+      "step": 2406
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.36664290613971,
+      "learning_rate": 9.60190530508038e-06,
+      "loss": 0.7073,
+      "step": 2407
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2874867298437658,
+      "learning_rate": 9.601499916254626e-06,
+      "loss": 0.6938,
+      "step": 2408
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3359975926771532,
+      "learning_rate": 9.601094329692173e-06,
+      "loss": 0.7687,
+      "step": 2409
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4543355020428221,
+      "learning_rate": 9.600688545410449e-06,
+      "loss": 0.6862,
+      "step": 2410
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3443041927934658,
+      "learning_rate": 9.60028256342689e-06,
+      "loss": 0.6468,
+      "step": 2411
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3642928837986557,
+      "learning_rate": 9.599876383758946e-06,
+      "loss": 0.8265,
+      "step": 2412
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4634637941401931,
+      "learning_rate": 9.599470006424069e-06,
+      "loss": 0.749,
+      "step": 2413
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4014724798265819,
+      "learning_rate": 9.599063431439721e-06,
+      "loss": 0.7128,
+      "step": 2414
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2294519809922286,
+      "learning_rate": 9.598656658823378e-06,
+      "loss": 0.6707,
+      "step": 2415
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5430674514685319,
+      "learning_rate": 9.598249688592514e-06,
+      "loss": 0.676,
+      "step": 2416
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.282792833536899,
+      "learning_rate": 9.597842520764619e-06,
+      "loss": 0.7789,
+      "step": 2417
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3790858315149814,
+      "learning_rate": 9.597435155357192e-06,
+      "loss": 0.641,
+      "step": 2418
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.65152851474826,
+      "learning_rate": 9.597027592387739e-06,
+      "loss": 0.7044,
+      "step": 2419
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2774844318122518,
+      "learning_rate": 9.59661983187377e-06,
+      "loss": 0.6875,
+      "step": 2420
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4895774621808826,
+      "learning_rate": 9.59621187383281e-06,
+      "loss": 0.7795,
+      "step": 2421
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.288961882024136,
+      "learning_rate": 9.595803718282391e-06,
+      "loss": 0.753,
+      "step": 2422
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5885267238228693,
+      "learning_rate": 9.59539536524005e-06,
+      "loss": 0.6961,
+      "step": 2423
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3162745298144438,
+      "learning_rate": 9.594986814723335e-06,
+      "loss": 0.6865,
+      "step": 2424
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3973186386671363,
+      "learning_rate": 9.594578066749804e-06,
+      "loss": 0.628,
+      "step": 2425
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3549380686655361,
+      "learning_rate": 9.59416912133702e-06,
+      "loss": 0.6245,
+      "step": 2426
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.354047509334235,
+      "learning_rate": 9.593759978502556e-06,
+      "loss": 0.816,
+      "step": 2427
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.280980378325334,
+      "learning_rate": 9.593350638263998e-06,
+      "loss": 0.7251,
+      "step": 2428
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4337577007331113,
+      "learning_rate": 9.59294110063893e-06,
+      "loss": 0.739,
+      "step": 2429
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4721513252430287,
+      "learning_rate": 9.592531365644957e-06,
+      "loss": 0.7687,
+      "step": 2430
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2305441586930572,
+      "learning_rate": 9.592121433299682e-06,
+      "loss": 0.7837,
+      "step": 2431
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5023402963367687,
+      "learning_rate": 9.59171130362072e-06,
+      "loss": 0.7155,
+      "step": 2432
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4905991879039817,
+      "learning_rate": 9.591300976625697e-06,
+      "loss": 0.6961,
+      "step": 2433
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.090779850744525,
+      "learning_rate": 9.590890452332249e-06,
+      "loss": 0.5782,
+      "step": 2434
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4348159461862398,
+      "learning_rate": 9.59047973075801e-06,
+      "loss": 0.7413,
+      "step": 2435
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.160487339625738,
+      "learning_rate": 9.590068811920637e-06,
+      "loss": 0.6261,
+      "step": 2436
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6996322743748138,
+      "learning_rate": 9.589657695837781e-06,
+      "loss": 0.8357,
+      "step": 2437
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3959461891872786,
+      "learning_rate": 9.589246382527113e-06,
+      "loss": 0.7685,
+      "step": 2438
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5523277335839865,
+      "learning_rate": 9.588834872006308e-06,
+      "loss": 0.7852,
+      "step": 2439
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.67035639045566,
+      "learning_rate": 9.588423164293049e-06,
+      "loss": 0.7308,
+      "step": 2440
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3853924326916212,
+      "learning_rate": 9.588011259405028e-06,
+      "loss": 0.7428,
+      "step": 2441
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4228232086273893,
+      "learning_rate": 9.587599157359943e-06,
+      "loss": 0.7402,
+      "step": 2442
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4014291328873698,
+      "learning_rate": 9.587186858175507e-06,
+      "loss": 0.6541,
+      "step": 2443
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.214588474376003,
+      "learning_rate": 9.586774361869436e-06,
+      "loss": 0.6445,
+      "step": 2444
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.325461397350849,
+      "learning_rate": 9.586361668459452e-06,
+      "loss": 0.6909,
+      "step": 2445
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3169206714526458,
+      "learning_rate": 9.585948777963296e-06,
+      "loss": 0.6947,
+      "step": 2446
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3229004793889751,
+      "learning_rate": 9.585535690398706e-06,
+      "loss": 0.7331,
+      "step": 2447
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3622474492829808,
+      "learning_rate": 9.585122405783435e-06,
+      "loss": 0.7197,
+      "step": 2448
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7759588937460038,
+      "learning_rate": 9.584708924135245e-06,
+      "loss": 0.7579,
+      "step": 2449
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5250266643912653,
+      "learning_rate": 9.584295245471898e-06,
+      "loss": 0.7728,
+      "step": 2450
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6026054793441413,
+      "learning_rate": 9.583881369811176e-06,
+      "loss": 0.7607,
+      "step": 2451
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3190302269029253,
+      "learning_rate": 9.583467297170863e-06,
+      "loss": 0.7003,
+      "step": 2452
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2191843098043575,
+      "learning_rate": 9.583053027568754e-06,
+      "loss": 0.6795,
+      "step": 2453
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.001208310569608,
+      "learning_rate": 9.582638561022646e-06,
+      "loss": 0.697,
+      "step": 2454
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3007850099886393,
+      "learning_rate": 9.582223897550356e-06,
+      "loss": 0.9347,
+      "step": 2455
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3228721389753626,
+      "learning_rate": 9.581809037169698e-06,
+      "loss": 0.6736,
+      "step": 2456
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4220060992137296,
+      "learning_rate": 9.581393979898502e-06,
+      "loss": 0.7371,
+      "step": 2457
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9244780797685068,
+      "learning_rate": 9.580978725754604e-06,
+      "loss": 0.6463,
+      "step": 2458
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3795756763116076,
+      "learning_rate": 9.580563274755848e-06,
+      "loss": 0.7282,
+      "step": 2459
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4102785115117873,
+      "learning_rate": 9.580147626920084e-06,
+      "loss": 0.7417,
+      "step": 2460
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3353242705567965,
+      "learning_rate": 9.579731782265178e-06,
+      "loss": 0.7577,
+      "step": 2461
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.418598961429821,
+      "learning_rate": 9.579315740808998e-06,
+      "loss": 0.6592,
+      "step": 2462
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3546413233190255,
+      "learning_rate": 9.578899502569421e-06,
+      "loss": 0.7046,
+      "step": 2463
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5081481865418267,
+      "learning_rate": 9.578483067564335e-06,
+      "loss": 0.7179,
+      "step": 2464
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5257955339818638,
+      "learning_rate": 9.578066435811636e-06,
+      "loss": 0.6876,
+      "step": 2465
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4862769373166866,
+      "learning_rate": 9.577649607329225e-06,
+      "loss": 0.6929,
+      "step": 2466
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0848366934296487,
+      "learning_rate": 9.577232582135014e-06,
+      "loss": 0.6841,
+      "step": 2467
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5046847646081853,
+      "learning_rate": 9.576815360246926e-06,
+      "loss": 0.7282,
+      "step": 2468
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.326207208124238,
+      "learning_rate": 9.576397941682891e-06,
+      "loss": 0.7482,
+      "step": 2469
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3155471493358895,
+      "learning_rate": 9.575980326460841e-06,
+      "loss": 0.7224,
+      "step": 2470
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2111738747212668,
+      "learning_rate": 9.575562514598727e-06,
+      "loss": 0.7162,
+      "step": 2471
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5498335565151162,
+      "learning_rate": 9.575144506114501e-06,
+      "loss": 0.6419,
+      "step": 2472
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3104074425547603,
+      "learning_rate": 9.574726301026127e-06,
+      "loss": 0.7005,
+      "step": 2473
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5659477383768723,
+      "learning_rate": 9.574307899351574e-06,
+      "loss": 0.7517,
+      "step": 2474
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.571493986203944,
+      "learning_rate": 9.573889301108823e-06,
+      "loss": 0.7081,
+      "step": 2475
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4112628023664628,
+      "learning_rate": 9.573470506315864e-06,
+      "loss": 0.7277,
+      "step": 2476
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3792473302273527,
+      "learning_rate": 9.57305151499069e-06,
+      "loss": 0.6859,
+      "step": 2477
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1885442338145866,
+      "learning_rate": 9.572632327151309e-06,
+      "loss": 0.6728,
+      "step": 2478
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3124145510693412,
+      "learning_rate": 9.572212942815734e-06,
+      "loss": 0.7396,
+      "step": 2479
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5089077643653295,
+      "learning_rate": 9.571793362001983e-06,
+      "loss": 0.7935,
+      "step": 2480
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3423077796375054,
+      "learning_rate": 9.57137358472809e-06,
+      "loss": 0.7112,
+      "step": 2481
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1452955865731334,
+      "learning_rate": 9.570953611012092e-06,
+      "loss": 0.6925,
+      "step": 2482
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.299879506758181,
+      "learning_rate": 9.570533440872041e-06,
+      "loss": 0.7336,
+      "step": 2483
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5199858114335096,
+      "learning_rate": 9.570113074325986e-06,
+      "loss": 0.6735,
+      "step": 2484
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2335999900381207,
+      "learning_rate": 9.569692511391995e-06,
+      "loss": 0.65,
+      "step": 2485
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3508793102358208,
+      "learning_rate": 9.569271752088139e-06,
+      "loss": 0.7484,
+      "step": 2486
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1743848736207894,
+      "learning_rate": 9.5688507964325e-06,
+      "loss": 0.616,
+      "step": 2487
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4159756426134054,
+      "learning_rate": 9.568429644443167e-06,
+      "loss": 0.7666,
+      "step": 2488
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7343314750182928,
+      "learning_rate": 9.568008296138238e-06,
+      "loss": 0.8344,
+      "step": 2489
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3460872176855816,
+      "learning_rate": 9.567586751535818e-06,
+      "loss": 0.7234,
+      "step": 2490
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4468531266141327,
+      "learning_rate": 9.567165010654022e-06,
+      "loss": 0.717,
+      "step": 2491
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6165957663281147,
+      "learning_rate": 9.566743073510976e-06,
+      "loss": 0.8053,
+      "step": 2492
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0476816007244916,
+      "learning_rate": 9.566320940124809e-06,
+      "loss": 0.6116,
+      "step": 2493
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2204343156570063,
+      "learning_rate": 9.565898610513661e-06,
+      "loss": 0.6411,
+      "step": 2494
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2600778642881252,
+      "learning_rate": 9.565476084695681e-06,
+      "loss": 0.731,
+      "step": 2495
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.37697037815579,
+      "learning_rate": 9.565053362689026e-06,
+      "loss": 0.7849,
+      "step": 2496
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4007184757388083,
+      "learning_rate": 9.56463044451186e-06,
+      "loss": 0.6432,
+      "step": 2497
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3627765410410873,
+      "learning_rate": 9.564207330182359e-06,
+      "loss": 0.7954,
+      "step": 2498
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3220929616735637,
+      "learning_rate": 9.563784019718704e-06,
+      "loss": 0.6558,
+      "step": 2499
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3639346012284588,
+      "learning_rate": 9.563360513139085e-06,
+      "loss": 0.6845,
+      "step": 2500
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5255199037919653,
+      "learning_rate": 9.562936810461701e-06,
+      "loss": 0.6723,
+      "step": 2501
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2689359399377522,
+      "learning_rate": 9.562512911704761e-06,
+      "loss": 0.6322,
+      "step": 2502
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2869901674890907,
+      "learning_rate": 9.56208881688648e-06,
+      "loss": 0.7321,
+      "step": 2503
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3419816067369332,
+      "learning_rate": 9.561664526025082e-06,
+      "loss": 0.6381,
+      "step": 2504
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4884704969017262,
+      "learning_rate": 9.5612400391388e-06,
+      "loss": 0.816,
+      "step": 2505
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.078391192709676,
+      "learning_rate": 9.560815356245875e-06,
+      "loss": 0.5227,
+      "step": 2506
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.153457642569216,
+      "learning_rate": 9.560390477364555e-06,
+      "loss": 0.7517,
+      "step": 2507
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7191230927194285,
+      "learning_rate": 9.559965402513103e-06,
+      "loss": 0.675,
+      "step": 2508
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2417315894856311,
+      "learning_rate": 9.55954013170978e-06,
+      "loss": 0.6571,
+      "step": 2509
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.413253159474528,
+      "learning_rate": 9.559114664972864e-06,
+      "loss": 0.8249,
+      "step": 2510
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3156706642357816,
+      "learning_rate": 9.558689002320635e-06,
+      "loss": 0.6783,
+      "step": 2511
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3958866286197635,
+      "learning_rate": 9.55826314377139e-06,
+      "loss": 0.7837,
+      "step": 2512
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.245628108443086,
+      "learning_rate": 9.557837089343424e-06,
+      "loss": 0.6661,
+      "step": 2513
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.248295575846921,
+      "learning_rate": 9.557410839055047e-06,
+      "loss": 0.6523,
+      "step": 2514
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3362134346668082,
+      "learning_rate": 9.556984392924577e-06,
+      "loss": 0.7205,
+      "step": 2515
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.153665815569431,
+      "learning_rate": 9.55655775097034e-06,
+      "loss": 0.5604,
+      "step": 2516
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2297609985520817,
+      "learning_rate": 9.556130913210666e-06,
+      "loss": 0.6408,
+      "step": 2517
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3125513011303211,
+      "learning_rate": 9.5557038796639e-06,
+      "loss": 0.694,
+      "step": 2518
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4750326859596865,
+      "learning_rate": 9.555276650348393e-06,
+      "loss": 0.7808,
+      "step": 2519
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3338896628063734,
+      "learning_rate": 9.554849225282503e-06,
+      "loss": 0.6542,
+      "step": 2520
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3896536578181973,
+      "learning_rate": 9.554421604484599e-06,
+      "loss": 0.6399,
+      "step": 2521
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5018121695592053,
+      "learning_rate": 9.553993787973053e-06,
+      "loss": 0.7703,
+      "step": 2522
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.461528778337965,
+      "learning_rate": 9.553565775766253e-06,
+      "loss": 0.746,
+      "step": 2523
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.276148673633827,
+      "learning_rate": 9.55313756788259e-06,
+      "loss": 0.585,
+      "step": 2524
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3398920896469955,
+      "learning_rate": 9.552709164340466e-06,
+      "loss": 0.7328,
+      "step": 2525
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4126647017764131,
+      "learning_rate": 9.552280565158289e-06,
+      "loss": 0.7969,
+      "step": 2526
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1730260256615572,
+      "learning_rate": 9.551851770354477e-06,
+      "loss": 0.641,
+      "step": 2527
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.441069475326277,
+      "learning_rate": 9.551422779947458e-06,
+      "loss": 0.7483,
+      "step": 2528
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4679913535966023,
+      "learning_rate": 9.550993593955665e-06,
+      "loss": 0.7401,
+      "step": 2529
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8855857354623748,
+      "learning_rate": 9.550564212397541e-06,
+      "loss": 0.6861,
+      "step": 2530
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2120776593390667,
+      "learning_rate": 9.55013463529154e-06,
+      "loss": 0.6885,
+      "step": 2531
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.405221660273847,
+      "learning_rate": 9.54970486265612e-06,
+      "loss": 0.6747,
+      "step": 2532
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3508329029472304,
+      "learning_rate": 9.549274894509748e-06,
+      "loss": 0.737,
+      "step": 2533
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4482795434918112,
+      "learning_rate": 9.548844730870903e-06,
+      "loss": 0.8694,
+      "step": 2534
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5498368658975994,
+      "learning_rate": 9.548414371758068e-06,
+      "loss": 0.6356,
+      "step": 2535
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.338695629426084,
+      "learning_rate": 9.547983817189738e-06,
+      "loss": 0.7333,
+      "step": 2536
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2427880103251114,
+      "learning_rate": 9.547553067184413e-06,
+      "loss": 0.6779,
+      "step": 2537
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5125171296849391,
+      "learning_rate": 9.547122121760607e-06,
+      "loss": 0.6925,
+      "step": 2538
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5039991410784652,
+      "learning_rate": 9.546690980936836e-06,
+      "loss": 0.6868,
+      "step": 2539
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2583241297834498,
+      "learning_rate": 9.546259644731628e-06,
+      "loss": 0.6449,
+      "step": 2540
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4184829547308677,
+      "learning_rate": 9.545828113163516e-06,
+      "loss": 0.648,
+      "step": 2541
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3941873386985149,
+      "learning_rate": 9.545396386251048e-06,
+      "loss": 0.7857,
+      "step": 2542
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.212687369952324,
+      "learning_rate": 9.544964464012774e-06,
+      "loss": 0.688,
+      "step": 2543
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4886178117412527,
+      "learning_rate": 9.544532346467254e-06,
+      "loss": 0.7643,
+      "step": 2544
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.601335877211319,
+      "learning_rate": 9.544100033633057e-06,
+      "loss": 0.6994,
+      "step": 2545
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3576860159550228,
+      "learning_rate": 9.543667525528762e-06,
+      "loss": 0.7624,
+      "step": 2546
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.454048441572157,
+      "learning_rate": 9.543234822172956e-06,
+      "loss": 0.6951,
+      "step": 2547
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3546583793378961,
+      "learning_rate": 9.542801923584228e-06,
+      "loss": 0.7142,
+      "step": 2548
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3933015224101795,
+      "learning_rate": 9.542368829781186e-06,
+      "loss": 0.7941,
+      "step": 2549
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.367832279503824,
+      "learning_rate": 9.541935540782439e-06,
+      "loss": 0.7367,
+      "step": 2550
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3589562318202124,
+      "learning_rate": 9.541502056606606e-06,
+      "loss": 0.7164,
+      "step": 2551
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1261592796237314,
+      "learning_rate": 9.541068377272315e-06,
+      "loss": 0.6323,
+      "step": 2552
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5617539177878499,
+      "learning_rate": 9.540634502798203e-06,
+      "loss": 0.7298,
+      "step": 2553
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4523671988010745,
+      "learning_rate": 9.540200433202913e-06,
+      "loss": 0.7125,
+      "step": 2554
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4653788012944902,
+      "learning_rate": 9.5397661685051e-06,
+      "loss": 0.7753,
+      "step": 2555
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.484835791226125,
+      "learning_rate": 9.539331708723423e-06,
+      "loss": 0.7991,
+      "step": 2556
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.401653692327027,
+      "learning_rate": 9.538897053876553e-06,
+      "loss": 0.6846,
+      "step": 2557
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5680865700216575,
+      "learning_rate": 9.538462203983169e-06,
+      "loss": 0.645,
+      "step": 2558
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.260965303362877,
+      "learning_rate": 9.538027159061955e-06,
+      "loss": 0.6193,
+      "step": 2559
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4260610014383124,
+      "learning_rate": 9.537591919131608e-06,
+      "loss": 0.77,
+      "step": 2560
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.016826160940039,
+      "learning_rate": 9.53715648421083e-06,
+      "loss": 0.6249,
+      "step": 2561
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5232367626580785,
+      "learning_rate": 9.536720854318333e-06,
+      "loss": 0.6837,
+      "step": 2562
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2505452331378124,
+      "learning_rate": 9.536285029472838e-06,
+      "loss": 0.7235,
+      "step": 2563
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6254351970309422,
+      "learning_rate": 9.535849009693072e-06,
+      "loss": 0.7919,
+      "step": 2564
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3747345457478666,
+      "learning_rate": 9.535412794997772e-06,
+      "loss": 0.844,
+      "step": 2565
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.437376417454695,
+      "learning_rate": 9.534976385405683e-06,
+      "loss": 0.7097,
+      "step": 2566
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2570805175637358,
+      "learning_rate": 9.53453978093556e-06,
+      "loss": 0.7098,
+      "step": 2567
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.523000380358296,
+      "learning_rate": 9.534102981606162e-06,
+      "loss": 0.7104,
+      "step": 2568
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2995331060741317,
+      "learning_rate": 9.533665987436262e-06,
+      "loss": 0.7086,
+      "step": 2569
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0780594828610566,
+      "learning_rate": 9.533228798444637e-06,
+      "loss": 0.6487,
+      "step": 2570
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3863074784801146,
+      "learning_rate": 9.532791414650075e-06,
+      "loss": 0.6648,
+      "step": 2571
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1625049430292393,
+      "learning_rate": 9.53235383607137e-06,
+      "loss": 0.6513,
+      "step": 2572
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3460782703915486,
+      "learning_rate": 9.531916062727328e-06,
+      "loss": 0.7098,
+      "step": 2573
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.250811229130931,
+      "learning_rate": 9.531478094636758e-06,
+      "loss": 0.7575,
+      "step": 2574
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.495470525058453,
+      "learning_rate": 9.531039931818485e-06,
+      "loss": 0.7619,
+      "step": 2575
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.963641454793746,
+      "learning_rate": 9.530601574291331e-06,
+      "loss": 0.6544,
+      "step": 2576
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5680089219692939,
+      "learning_rate": 9.530163022074139e-06,
+      "loss": 0.7414,
+      "step": 2577
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2634304664725502,
+      "learning_rate": 9.529724275185753e-06,
+      "loss": 0.6595,
+      "step": 2578
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.8809713987514343,
+      "learning_rate": 9.529285333645027e-06,
+      "loss": 0.6656,
+      "step": 2579
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1658079311870193,
+      "learning_rate": 9.528846197470822e-06,
+      "loss": 0.5974,
+      "step": 2580
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.08382807228194,
+      "learning_rate": 9.528406866682009e-06,
+      "loss": 0.7346,
+      "step": 2581
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5259985682115296,
+      "learning_rate": 9.527967341297471e-06,
+      "loss": 0.7034,
+      "step": 2582
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2413923931641282,
+      "learning_rate": 9.527527621336087e-06,
+      "loss": 0.7645,
+      "step": 2583
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4980556548498363,
+      "learning_rate": 9.527087706816762e-06,
+      "loss": 0.7692,
+      "step": 2584
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3545089703190767,
+      "learning_rate": 9.526647597758394e-06,
+      "loss": 0.6911,
+      "step": 2585
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4692588886850648,
+      "learning_rate": 9.526207294179897e-06,
+      "loss": 0.7347,
+      "step": 2586
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5918233480725503,
+      "learning_rate": 9.525766796100194e-06,
+      "loss": 0.7358,
+      "step": 2587
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1593195890739234,
+      "learning_rate": 9.52532610353821e-06,
+      "loss": 0.7731,
+      "step": 2588
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5169715825656664,
+      "learning_rate": 9.524885216512887e-06,
+      "loss": 0.798,
+      "step": 2589
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3503884876923367,
+      "learning_rate": 9.524444135043168e-06,
+      "loss": 0.6524,
+      "step": 2590
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3683300827372413,
+      "learning_rate": 9.524002859148006e-06,
+      "loss": 0.6635,
+      "step": 2591
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5144322109184292,
+      "learning_rate": 9.523561388846368e-06,
+      "loss": 0.8082,
+      "step": 2592
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3329769949882757,
+      "learning_rate": 9.523119724157222e-06,
+      "loss": 0.7007,
+      "step": 2593
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2849434990485704,
+      "learning_rate": 9.522677865099548e-06,
+      "loss": 0.8077,
+      "step": 2594
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.135666248997524,
+      "learning_rate": 9.522235811692335e-06,
+      "loss": 0.7453,
+      "step": 2595
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4503041398705503,
+      "learning_rate": 9.521793563954575e-06,
+      "loss": 0.6471,
+      "step": 2596
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3331077739142774,
+      "learning_rate": 9.521351121905278e-06,
+      "loss": 0.7248,
+      "step": 2597
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.355540510711293,
+      "learning_rate": 9.520908485563452e-06,
+      "loss": 0.813,
+      "step": 2598
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2732757266881363,
+      "learning_rate": 9.520465654948119e-06,
+      "loss": 0.6615,
+      "step": 2599
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3950529861150578,
+      "learning_rate": 9.52002263007831e-06,
+      "loss": 0.786,
+      "step": 2600
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.434295544217669,
+      "learning_rate": 9.519579410973061e-06,
+      "loss": 0.5496,
+      "step": 2601
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5127317900537796,
+      "learning_rate": 9.51913599765142e-06,
+      "loss": 0.7809,
+      "step": 2602
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2866792233662312,
+      "learning_rate": 9.518692390132441e-06,
+      "loss": 0.6535,
+      "step": 2603
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3368656901422886,
+      "learning_rate": 9.518248588435185e-06,
+      "loss": 0.7283,
+      "step": 2604
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5032239252207795,
+      "learning_rate": 9.517804592578725e-06,
+      "loss": 0.7084,
+      "step": 2605
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6863412614255961,
+      "learning_rate": 9.517360402582141e-06,
+      "loss": 0.7285,
+      "step": 2606
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.453090562048841,
+      "learning_rate": 9.51691601846452e-06,
+      "loss": 0.7311,
+      "step": 2607
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2342157719740872,
+      "learning_rate": 9.516471440244956e-06,
+      "loss": 0.7274,
+      "step": 2608
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4622698926802522,
+      "learning_rate": 9.516026667942557e-06,
+      "loss": 0.728,
+      "step": 2609
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.46643727856658,
+      "learning_rate": 9.515581701576433e-06,
+      "loss": 0.6494,
+      "step": 2610
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3796030799829353,
+      "learning_rate": 9.515136541165708e-06,
+      "loss": 0.7878,
+      "step": 2611
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2384293812167033,
+      "learning_rate": 9.514691186729509e-06,
+      "loss": 0.7507,
+      "step": 2612
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8972891092700193,
+      "learning_rate": 9.514245638286976e-06,
+      "loss": 0.8064,
+      "step": 2613
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4617278679988468,
+      "learning_rate": 9.513799895857252e-06,
+      "loss": 0.7865,
+      "step": 2614
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3569312656676586,
+      "learning_rate": 9.513353959459496e-06,
+      "loss": 0.7552,
+      "step": 2615
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.580060466723354,
+      "learning_rate": 9.512907829112867e-06,
+      "loss": 0.7401,
+      "step": 2616
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.558754460539447,
+      "learning_rate": 9.512461504836538e-06,
+      "loss": 0.8203,
+      "step": 2617
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2648907833510366,
+      "learning_rate": 9.512014986649691e-06,
+      "loss": 0.66,
+      "step": 2618
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3396548027213457,
+      "learning_rate": 9.511568274571508e-06,
+      "loss": 0.6812,
+      "step": 2619
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4790338327262764,
+      "learning_rate": 9.51112136862119e-06,
+      "loss": 0.7444,
+      "step": 2620
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.354855230956673,
+      "learning_rate": 9.51067426881794e-06,
+      "loss": 0.6397,
+      "step": 2621
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.228083174452327,
+      "learning_rate": 9.510226975180969e-06,
+      "loss": 0.6446,
+      "step": 2622
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4586598157830954,
+      "learning_rate": 9.509779487729502e-06,
+      "loss": 0.6907,
+      "step": 2623
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1830743966405617,
+      "learning_rate": 9.509331806482767e-06,
+      "loss": 0.6343,
+      "step": 2624
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3472001688605422,
+      "learning_rate": 9.50888393146e-06,
+      "loss": 0.6371,
+      "step": 2625
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.262757207757021,
+      "learning_rate": 9.508435862680449e-06,
+      "loss": 0.7076,
+      "step": 2626
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7028514333041305,
+      "learning_rate": 9.507987600163369e-06,
+      "loss": 0.8141,
+      "step": 2627
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2766185256837406,
+      "learning_rate": 9.507539143928021e-06,
+      "loss": 0.6691,
+      "step": 2628
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4150721776099657,
+      "learning_rate": 9.507090493993677e-06,
+      "loss": 0.6814,
+      "step": 2629
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4647779963604988,
+      "learning_rate": 9.506641650379617e-06,
+      "loss": 0.786,
+      "step": 2630
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3802201248663137,
+      "learning_rate": 9.50619261310513e-06,
+      "loss": 0.6466,
+      "step": 2631
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3546307248967104,
+      "learning_rate": 9.50574338218951e-06,
+      "loss": 0.7505,
+      "step": 2632
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.591965574253377,
+      "learning_rate": 9.50529395765206e-06,
+      "loss": 0.7086,
+      "step": 2633
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3957597786607903,
+      "learning_rate": 9.504844339512096e-06,
+      "loss": 0.6869,
+      "step": 2634
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3324064007704077,
+      "learning_rate": 9.504394527788939e-06,
+      "loss": 0.6953,
+      "step": 2635
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.436204660295804,
+      "learning_rate": 9.503944522501915e-06,
+      "loss": 0.6954,
+      "step": 2636
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6008418506138293,
+      "learning_rate": 9.503494323670366e-06,
+      "loss": 0.838,
+      "step": 2637
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5168014465565576,
+      "learning_rate": 9.503043931313637e-06,
+      "loss": 0.8149,
+      "step": 2638
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3536107591586186,
+      "learning_rate": 9.502593345451078e-06,
+      "loss": 0.7127,
+      "step": 2639
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7466968205469144,
+      "learning_rate": 9.502142566102058e-06,
+      "loss": 0.6884,
+      "step": 2640
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3411914912394511,
+      "learning_rate": 9.501691593285943e-06,
+      "loss": 0.8356,
+      "step": 2641
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.192376754845205,
+      "learning_rate": 9.501240427022115e-06,
+      "loss": 0.6926,
+      "step": 2642
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.53725754842671,
+      "learning_rate": 9.500789067329963e-06,
+      "loss": 0.7669,
+      "step": 2643
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4417190570299423,
+      "learning_rate": 9.500337514228878e-06,
+      "loss": 0.8149,
+      "step": 2644
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5998289833741353,
+      "learning_rate": 9.49988576773827e-06,
+      "loss": 0.7679,
+      "step": 2645
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2055537000917462,
+      "learning_rate": 9.499433827877547e-06,
+      "loss": 0.6292,
+      "step": 2646
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4215148802945727,
+      "learning_rate": 9.498981694666133e-06,
+      "loss": 0.7268,
+      "step": 2647
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3602149573993683,
+      "learning_rate": 9.498529368123455e-06,
+      "loss": 0.7148,
+      "step": 2648
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.164579757405346,
+      "learning_rate": 9.49807684826895e-06,
+      "loss": 0.7598,
+      "step": 2649
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1670069240260306,
+      "learning_rate": 9.497624135122068e-06,
+      "loss": 0.6349,
+      "step": 2650
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2074122344315554,
+      "learning_rate": 9.497171228702257e-06,
+      "loss": 0.7226,
+      "step": 2651
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6077638059447947,
+      "learning_rate": 9.496718129028986e-06,
+      "loss": 0.6754,
+      "step": 2652
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2732369846484317,
+      "learning_rate": 9.49626483612172e-06,
+      "loss": 0.6282,
+      "step": 2653
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.540003325501775,
+      "learning_rate": 9.495811349999941e-06,
+      "loss": 0.6843,
+      "step": 2654
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.482008033213599,
+      "learning_rate": 9.495357670683135e-06,
+      "loss": 0.8544,
+      "step": 2655
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4175586910558278,
+      "learning_rate": 9.494903798190798e-06,
+      "loss": 0.6353,
+      "step": 2656
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2416855354987428,
+      "learning_rate": 9.494449732542435e-06,
+      "loss": 0.615,
+      "step": 2657
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.292509333414697,
+      "learning_rate": 9.493995473757558e-06,
+      "loss": 0.6097,
+      "step": 2658
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4119928953004541,
+      "learning_rate": 9.493541021855685e-06,
+      "loss": 0.714,
+      "step": 2659
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3224395883770506,
+      "learning_rate": 9.493086376856346e-06,
+      "loss": 0.821,
+      "step": 2660
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1819034894266292,
+      "learning_rate": 9.492631538779082e-06,
+      "loss": 0.6569,
+      "step": 2661
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4026448961635636,
+      "learning_rate": 9.492176507643433e-06,
+      "loss": 0.7992,
+      "step": 2662
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2878483614443603,
+      "learning_rate": 9.491721283468954e-06,
+      "loss": 0.7001,
+      "step": 2663
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3007967022705886,
+      "learning_rate": 9.49126586627521e-06,
+      "loss": 0.6651,
+      "step": 2664
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3859237269630198,
+      "learning_rate": 9.490810256081766e-06,
+      "loss": 0.7827,
+      "step": 2665
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2956094748226916,
+      "learning_rate": 9.490354452908208e-06,
+      "loss": 0.5656,
+      "step": 2666
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2210542758952247,
+      "learning_rate": 9.489898456774116e-06,
+      "loss": 0.6888,
+      "step": 2667
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4959853235915987,
+      "learning_rate": 9.489442267699087e-06,
+      "loss": 0.7431,
+      "step": 2668
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2344901398374546,
+      "learning_rate": 9.488985885702728e-06,
+      "loss": 0.5924,
+      "step": 2669
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0640863354102423,
+      "learning_rate": 9.488529310804645e-06,
+      "loss": 0.5327,
+      "step": 2670
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3022745618853988,
+      "learning_rate": 9.488072543024462e-06,
+      "loss": 0.6371,
+      "step": 2671
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4157389486661804,
+      "learning_rate": 9.487615582381809e-06,
+      "loss": 0.6644,
+      "step": 2672
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2191078288214041,
+      "learning_rate": 9.487158428896317e-06,
+      "loss": 0.672,
+      "step": 2673
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2981145675798829,
+      "learning_rate": 9.486701082587635e-06,
+      "loss": 0.6231,
+      "step": 2674
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.292562474391141,
+      "learning_rate": 9.486243543475414e-06,
+      "loss": 0.6567,
+      "step": 2675
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.211828782068359,
+      "learning_rate": 9.485785811579318e-06,
+      "loss": 0.7402,
+      "step": 2676
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0661364747501696,
+      "learning_rate": 9.485327886919013e-06,
+      "loss": 0.6851,
+      "step": 2677
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.609228619420828,
+      "learning_rate": 9.484869769514182e-06,
+      "loss": 0.8746,
+      "step": 2678
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4799443960702299,
+      "learning_rate": 9.484411459384508e-06,
+      "loss": 0.7331,
+      "step": 2679
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.083490596574181,
+      "learning_rate": 9.483952956549684e-06,
+      "loss": 0.6056,
+      "step": 2680
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2538854089634994,
+      "learning_rate": 9.483494261029418e-06,
+      "loss": 0.7284,
+      "step": 2681
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.458237519261043,
+      "learning_rate": 9.483035372843417e-06,
+      "loss": 0.6388,
+      "step": 2682
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4852078086782794,
+      "learning_rate": 9.482576292011402e-06,
+      "loss": 0.6666,
+      "step": 2683
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2793376884929804,
+      "learning_rate": 9.482117018553101e-06,
+      "loss": 0.6392,
+      "step": 2684
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3335213930331324,
+      "learning_rate": 9.481657552488248e-06,
+      "loss": 0.7482,
+      "step": 2685
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.649320012157539,
+      "learning_rate": 9.481197893836591e-06,
+      "loss": 0.6734,
+      "step": 2686
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2243211255443878,
+      "learning_rate": 9.480738042617878e-06,
+      "loss": 0.6603,
+      "step": 2687
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2883856728084357,
+      "learning_rate": 9.480277998851875e-06,
+      "loss": 0.7479,
+      "step": 2688
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3140059692934947,
+      "learning_rate": 9.479817762558345e-06,
+      "loss": 0.8152,
+      "step": 2689
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3528701038247748,
+      "learning_rate": 9.479357333757072e-06,
+      "loss": 0.7046,
+      "step": 2690
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4753835746736244,
+      "learning_rate": 9.478896712467836e-06,
+      "loss": 0.6003,
+      "step": 2691
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2423539775498367,
+      "learning_rate": 9.478435898710434e-06,
+      "loss": 0.665,
+      "step": 2692
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5541270622342318,
+      "learning_rate": 9.477974892504666e-06,
+      "loss": 0.7603,
+      "step": 2693
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.397236349709176,
+      "learning_rate": 9.477513693870347e-06,
+      "loss": 0.6758,
+      "step": 2694
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4308876457064434,
+      "learning_rate": 9.47705230282729e-06,
+      "loss": 0.7123,
+      "step": 2695
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4990080354213777,
+      "learning_rate": 9.476590719395324e-06,
+      "loss": 0.7321,
+      "step": 2696
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.462252077019707,
+      "learning_rate": 9.476128943594287e-06,
+      "loss": 0.8483,
+      "step": 2697
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4645489153338476,
+      "learning_rate": 9.47566697544402e-06,
+      "loss": 0.5837,
+      "step": 2698
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.350004027238919,
+      "learning_rate": 9.475204814964374e-06,
+      "loss": 0.7632,
+      "step": 2699
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0595796001556423,
+      "learning_rate": 9.474742462175212e-06,
+      "loss": 0.7056,
+      "step": 2700
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0198223434879272,
+      "learning_rate": 9.4742799170964e-06,
+      "loss": 0.689,
+      "step": 2701
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3233027381675044,
+      "learning_rate": 9.473817179747815e-06,
+      "loss": 0.6739,
+      "step": 2702
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5938704149326886,
+      "learning_rate": 9.473354250149342e-06,
+      "loss": 0.7416,
+      "step": 2703
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3584336863262032,
+      "learning_rate": 9.472891128320874e-06,
+      "loss": 0.6813,
+      "step": 2704
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3608048048966992,
+      "learning_rate": 9.472427814282311e-06,
+      "loss": 0.704,
+      "step": 2705
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3741153116572726,
+      "learning_rate": 9.471964308053567e-06,
+      "loss": 0.6371,
+      "step": 2706
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4042517472754614,
+      "learning_rate": 9.471500609654556e-06,
+      "loss": 0.6879,
+      "step": 2707
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.333176867965494,
+      "learning_rate": 9.471036719105205e-06,
+      "loss": 0.728,
+      "step": 2708
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5908032048914875,
+      "learning_rate": 9.470572636425451e-06,
+      "loss": 0.7736,
+      "step": 2709
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3514820814304527,
+      "learning_rate": 9.470108361635231e-06,
+      "loss": 0.6477,
+      "step": 2710
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.152872223059279,
+      "learning_rate": 9.469643894754502e-06,
+      "loss": 0.6633,
+      "step": 2711
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3283303438817147,
+      "learning_rate": 9.469179235803217e-06,
+      "loss": 0.7499,
+      "step": 2712
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5249621539346565,
+      "learning_rate": 9.46871438480135e-06,
+      "loss": 0.6424,
+      "step": 2713
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3618873415788924,
+      "learning_rate": 9.46824934176887e-06,
+      "loss": 0.7264,
+      "step": 2714
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5038723983021192,
+      "learning_rate": 9.467784106725767e-06,
+      "loss": 0.8605,
+      "step": 2715
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4514458196396292,
+      "learning_rate": 9.467318679692031e-06,
+      "loss": 0.7293,
+      "step": 2716
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3131097564552734,
+      "learning_rate": 9.466853060687662e-06,
+      "loss": 0.731,
+      "step": 2717
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.413146540741844,
+      "learning_rate": 9.466387249732668e-06,
+      "loss": 0.7552,
+      "step": 2718
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4498460701473344,
+      "learning_rate": 9.465921246847067e-06,
+      "loss": 0.7699,
+      "step": 2719
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.482511393917384,
+      "learning_rate": 9.465455052050884e-06,
+      "loss": 0.7875,
+      "step": 2720
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4706183339505337,
+      "learning_rate": 9.464988665364152e-06,
+      "loss": 0.7113,
+      "step": 2721
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3781472148164697,
+      "learning_rate": 9.464522086806914e-06,
+      "loss": 0.6309,
+      "step": 2722
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2657123912693227,
+      "learning_rate": 9.464055316399217e-06,
+      "loss": 0.6743,
+      "step": 2723
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.265322144257797,
+      "learning_rate": 9.463588354161122e-06,
+      "loss": 0.6923,
+      "step": 2724
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3454710135532582,
+      "learning_rate": 9.463121200112694e-06,
+      "loss": 0.7391,
+      "step": 2725
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3460135379007563,
+      "learning_rate": 9.462653854274008e-06,
+      "loss": 0.7967,
+      "step": 2726
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.321888323952072,
+      "learning_rate": 9.462186316665148e-06,
+      "loss": 0.7134,
+      "step": 2727
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5056389903325647,
+      "learning_rate": 9.461718587306205e-06,
+      "loss": 0.7822,
+      "step": 2728
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5417873489890792,
+      "learning_rate": 9.461250666217277e-06,
+      "loss": 0.7273,
+      "step": 2729
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1111171882260082,
+      "learning_rate": 9.46078255341847e-06,
+      "loss": 0.6031,
+      "step": 2730
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0853631998195463,
+      "learning_rate": 9.460314248929903e-06,
+      "loss": 0.7679,
+      "step": 2731
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.605098288555248,
+      "learning_rate": 9.4598457527717e-06,
+      "loss": 0.7452,
+      "step": 2732
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.388159467543237,
+      "learning_rate": 9.459377064963993e-06,
+      "loss": 0.724,
+      "step": 2733
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4718937172316016,
+      "learning_rate": 9.458908185526921e-06,
+      "loss": 0.7708,
+      "step": 2734
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3596965457962062,
+      "learning_rate": 9.458439114480634e-06,
+      "loss": 0.7707,
+      "step": 2735
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1206948293592007,
+      "learning_rate": 9.457969851845287e-06,
+      "loss": 0.5758,
+      "step": 2736
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3786785010994755,
+      "learning_rate": 9.457500397641049e-06,
+      "loss": 0.7233,
+      "step": 2737
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4294285970948428,
+      "learning_rate": 9.45703075188809e-06,
+      "loss": 0.6632,
+      "step": 2738
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6892019072429445,
+      "learning_rate": 9.456560914606594e-06,
+      "loss": 0.6812,
+      "step": 2739
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4877346979278823,
+      "learning_rate": 9.456090885816751e-06,
+      "loss": 0.7098,
+      "step": 2740
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.600419268559466,
+      "learning_rate": 9.455620665538758e-06,
+      "loss": 0.6927,
+      "step": 2741
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5373109170626031,
+      "learning_rate": 9.455150253792822e-06,
+      "loss": 0.8517,
+      "step": 2742
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.504329202859228,
+      "learning_rate": 9.454679650599156e-06,
+      "loss": 0.6539,
+      "step": 2743
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2548634865230899,
+      "learning_rate": 9.454208855977986e-06,
+      "loss": 0.6498,
+      "step": 2744
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.224546937565572,
+      "learning_rate": 9.45373786994954e-06,
+      "loss": 0.6451,
+      "step": 2745
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.311344278311618,
+      "learning_rate": 9.45326669253406e-06,
+      "loss": 0.6335,
+      "step": 2746
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5954681454524091,
+      "learning_rate": 9.452795323751793e-06,
+      "loss": 0.7083,
+      "step": 2747
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6870561136612472,
+      "learning_rate": 9.452323763622993e-06,
+      "loss": 0.6291,
+      "step": 2748
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2284487805072613,
+      "learning_rate": 9.451852012167924e-06,
+      "loss": 0.7939,
+      "step": 2749
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3323171132316818,
+      "learning_rate": 9.45138006940686e-06,
+      "loss": 0.7705,
+      "step": 2750
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2842560145798056,
+      "learning_rate": 9.450907935360081e-06,
+      "loss": 0.7326,
+      "step": 2751
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5361533481453271,
+      "learning_rate": 9.450435610047875e-06,
+      "loss": 0.6815,
+      "step": 2752
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2093375070700916,
+      "learning_rate": 9.44996309349054e-06,
+      "loss": 0.6952,
+      "step": 2753
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.477695998463955,
+      "learning_rate": 9.449490385708378e-06,
+      "loss": 0.7661,
+      "step": 2754
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3238302204297885,
+      "learning_rate": 9.449017486721705e-06,
+      "loss": 0.6186,
+      "step": 2755
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.575425399996806,
+      "learning_rate": 9.448544396550846e-06,
+      "loss": 0.7051,
+      "step": 2756
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.494822099649268,
+      "learning_rate": 9.448071115216123e-06,
+      "loss": 0.6996,
+      "step": 2757
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3796333402825074,
+      "learning_rate": 9.447597642737878e-06,
+      "loss": 0.6921,
+      "step": 2758
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.523547699225774,
+      "learning_rate": 9.447123979136457e-06,
+      "loss": 0.8002,
+      "step": 2759
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5392068227577183,
+      "learning_rate": 9.446650124432215e-06,
+      "loss": 0.6706,
+      "step": 2760
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3045181048688086,
+      "learning_rate": 9.446176078645515e-06,
+      "loss": 0.5929,
+      "step": 2761
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6039876501848627,
+      "learning_rate": 9.445701841796726e-06,
+      "loss": 0.8101,
+      "step": 2762
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.259767300153224,
+      "learning_rate": 9.445227413906227e-06,
+      "loss": 0.7208,
+      "step": 2763
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3847149388078068,
+      "learning_rate": 9.444752794994408e-06,
+      "loss": 0.6998,
+      "step": 2764
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3718915018164384,
+      "learning_rate": 9.44427798508166e-06,
+      "loss": 0.6729,
+      "step": 2765
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6122897819232676,
+      "learning_rate": 9.44380298418839e-06,
+      "loss": 0.7106,
+      "step": 2766
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6449779540459937,
+      "learning_rate": 9.443327792335009e-06,
+      "loss": 0.8148,
+      "step": 2767
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5805653795631074,
+      "learning_rate": 9.442852409541935e-06,
+      "loss": 0.7146,
+      "step": 2768
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.282480770705335,
+      "learning_rate": 9.4423768358296e-06,
+      "loss": 0.6551,
+      "step": 2769
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.10777891910441,
+      "learning_rate": 9.44190107121844e-06,
+      "loss": 0.6493,
+      "step": 2770
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4961648651976263,
+      "learning_rate": 9.441425115728898e-06,
+      "loss": 0.6773,
+      "step": 2771
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.581795186458153,
+      "learning_rate": 9.440948969381425e-06,
+      "loss": 0.7397,
+      "step": 2772
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2442637535885912,
+      "learning_rate": 9.440472632196485e-06,
+      "loss": 0.7665,
+      "step": 2773
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3117256383869043,
+      "learning_rate": 9.439996104194546e-06,
+      "loss": 0.6981,
+      "step": 2774
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2865532638053043,
+      "learning_rate": 9.439519385396088e-06,
+      "loss": 0.6891,
+      "step": 2775
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3018895965810482,
+      "learning_rate": 9.439042475821594e-06,
+      "loss": 0.6685,
+      "step": 2776
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0132862779887541,
+      "learning_rate": 9.438565375491558e-06,
+      "loss": 0.569,
+      "step": 2777
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2294458604346306,
+      "learning_rate": 9.438088084426482e-06,
+      "loss": 0.5814,
+      "step": 2778
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3127564457487573,
+      "learning_rate": 9.437610602646878e-06,
+      "loss": 0.6962,
+      "step": 2779
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2675846836551397,
+      "learning_rate": 9.437132930173264e-06,
+      "loss": 0.7035,
+      "step": 2780
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4589855477420082,
+      "learning_rate": 9.436655067026164e-06,
+      "loss": 0.6791,
+      "step": 2781
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4312551214929392,
+      "learning_rate": 9.436177013226116e-06,
+      "loss": 0.8323,
+      "step": 2782
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3996144833329816,
+      "learning_rate": 9.43569876879366e-06,
+      "loss": 0.7566,
+      "step": 2783
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3323610248086921,
+      "learning_rate": 9.43522033374935e-06,
+      "loss": 0.8597,
+      "step": 2784
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2236320106144918,
+      "learning_rate": 9.434741708113745e-06,
+      "loss": 0.6745,
+      "step": 2785
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4279714202614846,
+      "learning_rate": 9.434262891907413e-06,
+      "loss": 0.8935,
+      "step": 2786
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5111588829945857,
+      "learning_rate": 9.433783885150929e-06,
+      "loss": 0.7691,
+      "step": 2787
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3648945523029623,
+      "learning_rate": 9.433304687864876e-06,
+      "loss": 0.6743,
+      "step": 2788
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4805087465815574,
+      "learning_rate": 9.432825300069848e-06,
+      "loss": 0.7801,
+      "step": 2789
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5240117080986746,
+      "learning_rate": 9.432345721786442e-06,
+      "loss": 0.8044,
+      "step": 2790
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3596450747864919,
+      "learning_rate": 9.431865953035271e-06,
+      "loss": 0.71,
+      "step": 2791
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1498614823723037,
+      "learning_rate": 9.431385993836949e-06,
+      "loss": 0.6084,
+      "step": 2792
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2016573332428095,
+      "learning_rate": 9.430905844212102e-06,
+      "loss": 0.7059,
+      "step": 2793
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2890859955258116,
+      "learning_rate": 9.430425504181361e-06,
+      "loss": 0.628,
+      "step": 2794
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.330732409236103,
+      "learning_rate": 9.429944973765373e-06,
+      "loss": 0.6495,
+      "step": 2795
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.345239933926652,
+      "learning_rate": 9.429464252984779e-06,
+      "loss": 0.7192,
+      "step": 2796
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2851425947469837,
+      "learning_rate": 9.428983341860244e-06,
+      "loss": 0.6983,
+      "step": 2797
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1468151974134297,
+      "learning_rate": 9.42850224041243e-06,
+      "loss": 0.6899,
+      "step": 2798
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2320731960737499,
+      "learning_rate": 9.428020948662012e-06,
+      "loss": 0.6961,
+      "step": 2799
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.149237693053521,
+      "learning_rate": 9.427539466629672e-06,
+      "loss": 0.7317,
+      "step": 2800
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.528557420033599,
+      "learning_rate": 9.427057794336101e-06,
+      "loss": 0.7485,
+      "step": 2801
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4323399211707057,
+      "learning_rate": 9.426575931801996e-06,
+      "loss": 0.7152,
+      "step": 2802
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2296227647516491,
+      "learning_rate": 9.426093879048065e-06,
+      "loss": 0.6302,
+      "step": 2803
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.602700300328254,
+      "learning_rate": 9.425611636095023e-06,
+      "loss": 0.7237,
+      "step": 2804
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4217104485066376,
+      "learning_rate": 9.425129202963592e-06,
+      "loss": 0.7585,
+      "step": 2805
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3129896020187937,
+      "learning_rate": 9.424646579674504e-06,
+      "loss": 0.633,
+      "step": 2806
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3890930428164474,
+      "learning_rate": 9.424163766248499e-06,
+      "loss": 0.7213,
+      "step": 2807
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.563373327470476,
+      "learning_rate": 9.423680762706323e-06,
+      "loss": 0.7898,
+      "step": 2808
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2535709905272514,
+      "learning_rate": 9.423197569068733e-06,
+      "loss": 0.6619,
+      "step": 2809
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3920199715153379,
+      "learning_rate": 9.422714185356491e-06,
+      "loss": 0.7137,
+      "step": 2810
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2847791740893855,
+      "learning_rate": 9.422230611590373e-06,
+      "loss": 0.6488,
+      "step": 2811
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1973342062423242,
+      "learning_rate": 9.421746847791156e-06,
+      "loss": 0.6135,
+      "step": 2812
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4074779647950595,
+      "learning_rate": 9.421262893979628e-06,
+      "loss": 0.6637,
+      "step": 2813
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6200324526446042,
+      "learning_rate": 9.420778750176588e-06,
+      "loss": 0.7944,
+      "step": 2814
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3168023131464521,
+      "learning_rate": 9.420294416402839e-06,
+      "loss": 0.6905,
+      "step": 2815
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3725987933577133,
+      "learning_rate": 9.419809892679194e-06,
+      "loss": 0.7999,
+      "step": 2816
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.206371537654347,
+      "learning_rate": 9.419325179026475e-06,
+      "loss": 0.6769,
+      "step": 2817
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.319615961327452,
+      "learning_rate": 9.41884027546551e-06,
+      "loss": 0.7032,
+      "step": 2818
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.453104595397059,
+      "learning_rate": 9.418355182017138e-06,
+      "loss": 0.7791,
+      "step": 2819
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6995023099509494,
+      "learning_rate": 9.417869898702205e-06,
+      "loss": 0.711,
+      "step": 2820
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4747996300082287,
+      "learning_rate": 9.41738442554156e-06,
+      "loss": 0.6796,
+      "step": 2821
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2643038599852867,
+      "learning_rate": 9.41689876255607e-06,
+      "loss": 0.8294,
+      "step": 2822
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4224960637821769,
+      "learning_rate": 9.416412909766604e-06,
+      "loss": 0.7494,
+      "step": 2823
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.157022936273601,
+      "learning_rate": 9.41592686719404e-06,
+      "loss": 0.675,
+      "step": 2824
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.371345672934871,
+      "learning_rate": 9.41544063485926e-06,
+      "loss": 0.7505,
+      "step": 2825
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4599525779193412,
+      "learning_rate": 9.414954212783165e-06,
+      "loss": 0.921,
+      "step": 2826
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4469203907473016,
+      "learning_rate": 9.414467600986655e-06,
+      "loss": 0.7062,
+      "step": 2827
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7184282571321285,
+      "learning_rate": 9.41398079949064e-06,
+      "loss": 0.7855,
+      "step": 2828
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.677888726747674,
+      "learning_rate": 9.413493808316038e-06,
+      "loss": 0.9076,
+      "step": 2829
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5884225327886985,
+      "learning_rate": 9.41300662748378e-06,
+      "loss": 0.6623,
+      "step": 2830
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4181552460279159,
+      "learning_rate": 9.412519257014798e-06,
+      "loss": 0.6378,
+      "step": 2831
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.305575574492991,
+      "learning_rate": 9.412031696930036e-06,
+      "loss": 0.7172,
+      "step": 2832
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.359688111549324,
+      "learning_rate": 9.411543947250447e-06,
+      "loss": 0.7389,
+      "step": 2833
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.064772742363252,
+      "learning_rate": 9.411056007996989e-06,
+      "loss": 0.7791,
+      "step": 2834
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.432551509565343,
+      "learning_rate": 9.41056787919063e-06,
+      "loss": 0.6197,
+      "step": 2835
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3208379450032175,
+      "learning_rate": 9.410079560852347e-06,
+      "loss": 0.6936,
+      "step": 2836
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3107774810352004,
+      "learning_rate": 9.409591053003121e-06,
+      "loss": 0.7049,
+      "step": 2837
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.418163988441096,
+      "learning_rate": 9.40910235566395e-06,
+      "loss": 0.6943,
+      "step": 2838
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2213833436240007,
+      "learning_rate": 9.408613468855829e-06,
+      "loss": 0.6714,
+      "step": 2839
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0984818198676722,
+      "learning_rate": 9.40812439259977e-06,
+      "loss": 0.6326,
+      "step": 2840
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4178646448906271,
+      "learning_rate": 9.40763512691679e-06,
+      "loss": 0.7488,
+      "step": 2841
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.404288261146988,
+      "learning_rate": 9.407145671827909e-06,
+      "loss": 0.5713,
+      "step": 2842
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3709872379152181,
+      "learning_rate": 9.406656027354166e-06,
+      "loss": 0.6598,
+      "step": 2843
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.251709263029841,
+      "learning_rate": 9.406166193516596e-06,
+      "loss": 0.7165,
+      "step": 2844
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.244749848669076,
+      "learning_rate": 9.405676170336255e-06,
+      "loss": 0.6427,
+      "step": 2845
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.430335226375752,
+      "learning_rate": 9.405185957834196e-06,
+      "loss": 0.7436,
+      "step": 2846
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3914362622516085,
+      "learning_rate": 9.404695556031485e-06,
+      "loss": 0.7655,
+      "step": 2847
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8109331401076594,
+      "learning_rate": 9.404204964949196e-06,
+      "loss": 0.6876,
+      "step": 2848
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3017638303115124,
+      "learning_rate": 9.403714184608411e-06,
+      "loss": 0.6805,
+      "step": 2849
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.405002213092193,
+      "learning_rate": 9.403223215030221e-06,
+      "loss": 0.7455,
+      "step": 2850
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4106652116520255,
+      "learning_rate": 9.402732056235722e-06,
+      "loss": 0.5524,
+      "step": 2851
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1671432594918891,
+      "learning_rate": 9.402240708246022e-06,
+      "loss": 0.596,
+      "step": 2852
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4141203044993649,
+      "learning_rate": 9.401749171082233e-06,
+      "loss": 0.7472,
+      "step": 2853
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.085920858789821,
+      "learning_rate": 9.40125744476548e-06,
+      "loss": 0.6823,
+      "step": 2854
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3412048295315209,
+      "learning_rate": 9.400765529316892e-06,
+      "loss": 0.6647,
+      "step": 2855
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3196953299075371,
+      "learning_rate": 9.400273424757607e-06,
+      "loss": 0.6935,
+      "step": 2856
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2853313950944363,
+      "learning_rate": 9.399781131108775e-06,
+      "loss": 0.6733,
+      "step": 2857
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.266241185369157,
+      "learning_rate": 9.399288648391548e-06,
+      "loss": 0.6919,
+      "step": 2858
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.397787646229282,
+      "learning_rate": 9.398795976627091e-06,
+      "loss": 0.691,
+      "step": 2859
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4328593685122697,
+      "learning_rate": 9.398303115836573e-06,
+      "loss": 0.6877,
+      "step": 2860
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2458549814755615,
+      "learning_rate": 9.397810066041176e-06,
+      "loss": 0.745,
+      "step": 2861
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3633355939828615,
+      "learning_rate": 9.397316827262085e-06,
+      "loss": 0.7572,
+      "step": 2862
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.273204079235901,
+      "learning_rate": 9.396823399520495e-06,
+      "loss": 0.7345,
+      "step": 2863
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3156332303635367,
+      "learning_rate": 9.396329782837614e-06,
+      "loss": 0.6319,
+      "step": 2864
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5653260692165967,
+      "learning_rate": 9.39583597723465e-06,
+      "loss": 0.743,
+      "step": 2865
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3887405071738073,
+      "learning_rate": 9.395341982732825e-06,
+      "loss": 0.7406,
+      "step": 2866
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4557988005998708,
+      "learning_rate": 9.394847799353363e-06,
+      "loss": 0.6614,
+      "step": 2867
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6227865911189685,
+      "learning_rate": 9.394353427117506e-06,
+      "loss": 0.7959,
+      "step": 2868
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5561165949026279,
+      "learning_rate": 9.393858866046494e-06,
+      "loss": 0.8443,
+      "step": 2869
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.549932016986955,
+      "learning_rate": 9.393364116161582e-06,
+      "loss": 0.6928,
+      "step": 2870
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8130644970345604,
+      "learning_rate": 9.392869177484029e-06,
+      "loss": 0.7304,
+      "step": 2871
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.870203237753388,
+      "learning_rate": 9.392374050035104e-06,
+      "loss": 0.7756,
+      "step": 2872
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.401612313840594,
+      "learning_rate": 9.391878733836085e-06,
+      "loss": 0.7884,
+      "step": 2873
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5593691785868897,
+      "learning_rate": 9.391383228908253e-06,
+      "loss": 0.7714,
+      "step": 2874
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4083139056022076,
+      "learning_rate": 9.390887535272906e-06,
+      "loss": 0.7751,
+      "step": 2875
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5058900640485946,
+      "learning_rate": 9.39039165295134e-06,
+      "loss": 0.7446,
+      "step": 2876
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2691679773660636,
+      "learning_rate": 9.38989558196487e-06,
+      "loss": 0.6445,
+      "step": 2877
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3946044775737125,
+      "learning_rate": 9.389399322334806e-06,
+      "loss": 0.7504,
+      "step": 2878
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4367298012291072,
+      "learning_rate": 9.388902874082482e-06,
+      "loss": 0.6764,
+      "step": 2879
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.372681764036401,
+      "learning_rate": 9.388406237229223e-06,
+      "loss": 0.7192,
+      "step": 2880
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3981133726242674,
+      "learning_rate": 9.387909411796377e-06,
+      "loss": 0.7736,
+      "step": 2881
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.370088621702276,
+      "learning_rate": 9.387412397805289e-06,
+      "loss": 0.6848,
+      "step": 2882
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5315718551343804,
+      "learning_rate": 9.386915195277321e-06,
+      "loss": 0.7128,
+      "step": 2883
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.315449942997345,
+      "learning_rate": 9.386417804233836e-06,
+      "loss": 0.6404,
+      "step": 2884
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5302025304576823,
+      "learning_rate": 9.385920224696208e-06,
+      "loss": 0.7354,
+      "step": 2885
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4785851131681635,
+      "learning_rate": 9.385422456685822e-06,
+      "loss": 0.9289,
+      "step": 2886
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2803168395837465,
+      "learning_rate": 9.384924500224065e-06,
+      "loss": 0.5763,
+      "step": 2887
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3174233198555896,
+      "learning_rate": 9.384426355332336e-06,
+      "loss": 0.6933,
+      "step": 2888
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.188706601679293,
+      "learning_rate": 9.383928022032044e-06,
+      "loss": 0.5571,
+      "step": 2889
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3249974883858042,
+      "learning_rate": 9.383429500344598e-06,
+      "loss": 0.6453,
+      "step": 2890
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3297643956465124,
+      "learning_rate": 9.382930790291426e-06,
+      "loss": 0.7949,
+      "step": 2891
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2506907205722053,
+      "learning_rate": 9.382431891893957e-06,
+      "loss": 0.6956,
+      "step": 2892
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3219902288478163,
+      "learning_rate": 9.381932805173628e-06,
+      "loss": 0.7932,
+      "step": 2893
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3836225260623345,
+      "learning_rate": 9.381433530151887e-06,
+      "loss": 0.7237,
+      "step": 2894
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3066533188646627,
+      "learning_rate": 9.380934066850191e-06,
+      "loss": 0.6056,
+      "step": 2895
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3797636865199756,
+      "learning_rate": 9.38043441529e-06,
+      "loss": 0.6623,
+      "step": 2896
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2203520392413174,
+      "learning_rate": 9.379934575492788e-06,
+      "loss": 0.6816,
+      "step": 2897
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1655733760605016,
+      "learning_rate": 9.37943454748003e-06,
+      "loss": 0.6577,
+      "step": 2898
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0977261176214115,
+      "learning_rate": 9.37893433127322e-06,
+      "loss": 0.6163,
+      "step": 2899
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3010811212478959,
+      "learning_rate": 9.378433926893845e-06,
+      "loss": 0.6459,
+      "step": 2900
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4668955340041305,
+      "learning_rate": 9.377933334363416e-06,
+      "loss": 0.7916,
+      "step": 2901
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3332079933416794,
+      "learning_rate": 9.377432553703439e-06,
+      "loss": 0.7272,
+      "step": 2902
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3565479271969896,
+      "learning_rate": 9.376931584935439e-06,
+      "loss": 0.6994,
+      "step": 2903
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4808492637755637,
+      "learning_rate": 9.376430428080939e-06,
+      "loss": 0.693,
+      "step": 2904
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2335223340857944,
+      "learning_rate": 9.375929083161475e-06,
+      "loss": 0.6792,
+      "step": 2905
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1776781202184645,
+      "learning_rate": 9.375427550198595e-06,
+      "loss": 0.6148,
+      "step": 2906
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2723731286059576,
+      "learning_rate": 9.374925829213848e-06,
+      "loss": 0.6679,
+      "step": 2907
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.399935602767022,
+      "learning_rate": 9.374423920228795e-06,
+      "loss": 0.6572,
+      "step": 2908
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5399300640841775,
+      "learning_rate": 9.373921823265004e-06,
+      "loss": 0.7064,
+      "step": 2909
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2769128396706642,
+      "learning_rate": 9.373419538344049e-06,
+      "loss": 0.7192,
+      "step": 2910
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4020328962473565,
+      "learning_rate": 9.372917065487518e-06,
+      "loss": 0.8084,
+      "step": 2911
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1798532114221956,
+      "learning_rate": 9.372414404717001e-06,
+      "loss": 0.6009,
+      "step": 2912
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3857095760952018,
+      "learning_rate": 9.3719115560541e-06,
+      "loss": 0.7538,
+      "step": 2913
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.426648075338487,
+      "learning_rate": 9.371408519520421e-06,
+      "loss": 0.6803,
+      "step": 2914
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4752300688641593,
+      "learning_rate": 9.370905295137584e-06,
+      "loss": 0.7958,
+      "step": 2915
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5482000642142304,
+      "learning_rate": 9.370401882927212e-06,
+      "loss": 0.8417,
+      "step": 2916
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4886321790380028,
+      "learning_rate": 9.369898282910937e-06,
+      "loss": 0.817,
+      "step": 2917
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.225120164618549,
+      "learning_rate": 9.369394495110401e-06,
+      "loss": 0.6139,
+      "step": 2918
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.361935533842899,
+      "learning_rate": 9.36889051954725e-06,
+      "loss": 0.7562,
+      "step": 2919
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2389875084027513,
+      "learning_rate": 9.368386356243148e-06,
+      "loss": 0.6193,
+      "step": 2920
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3855234840431492,
+      "learning_rate": 9.367882005219752e-06,
+      "loss": 0.6818,
+      "step": 2921
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2388093208903501,
+      "learning_rate": 9.367377466498739e-06,
+      "loss": 0.6283,
+      "step": 2922
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4007822164942318,
+      "learning_rate": 9.36687274010179e-06,
+      "loss": 0.7448,
+      "step": 2923
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.380234656140066,
+      "learning_rate": 9.366367826050593e-06,
+      "loss": 0.7157,
+      "step": 2924
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3303164024033676,
+      "learning_rate": 9.365862724366848e-06,
+      "loss": 0.6982,
+      "step": 2925
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.439094058853707,
+      "learning_rate": 9.365357435072255e-06,
+      "loss": 0.8483,
+      "step": 2926
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5693518695149837,
+      "learning_rate": 9.364851958188534e-06,
+      "loss": 0.64,
+      "step": 2927
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7699479485468574,
+      "learning_rate": 9.364346293737403e-06,
+      "loss": 0.7517,
+      "step": 2928
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2951843612086735,
+      "learning_rate": 9.36384044174059e-06,
+      "loss": 0.7774,
+      "step": 2929
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4498607305844933,
+      "learning_rate": 9.363334402219836e-06,
+      "loss": 0.807,
+      "step": 2930
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3270894128967217,
+      "learning_rate": 9.362828175196884e-06,
+      "loss": 0.7357,
+      "step": 2931
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.547138162683318,
+      "learning_rate": 9.362321760693489e-06,
+      "loss": 0.7547,
+      "step": 2932
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1911662478241287,
+      "learning_rate": 9.361815158731413e-06,
+      "loss": 0.6941,
+      "step": 2933
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3076180363352947,
+      "learning_rate": 9.361308369332426e-06,
+      "loss": 0.7856,
+      "step": 2934
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6593268790198872,
+      "learning_rate": 9.360801392518303e-06,
+      "loss": 0.8095,
+      "step": 2935
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.365141219323556,
+      "learning_rate": 9.360294228310834e-06,
+      "loss": 0.8104,
+      "step": 2936
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3413233773996307,
+      "learning_rate": 9.359786876731811e-06,
+      "loss": 0.7194,
+      "step": 2937
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4838709007717004,
+      "learning_rate": 9.359279337803036e-06,
+      "loss": 0.8205,
+      "step": 2938
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3696851133454115,
+      "learning_rate": 9.358771611546319e-06,
+      "loss": 0.7346,
+      "step": 2939
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2113620359807042,
+      "learning_rate": 9.358263697983479e-06,
+      "loss": 0.7773,
+      "step": 2940
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3564886873297057,
+      "learning_rate": 9.35775559713634e-06,
+      "loss": 0.6683,
+      "step": 2941
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2180832823008583,
+      "learning_rate": 9.357247309026738e-06,
+      "loss": 0.6637,
+      "step": 2942
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2155547900881682,
+      "learning_rate": 9.356738833676517e-06,
+      "loss": 0.6425,
+      "step": 2943
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4989411820969627,
+      "learning_rate": 9.356230171107524e-06,
+      "loss": 0.7487,
+      "step": 2944
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4140857240505769,
+      "learning_rate": 9.355721321341618e-06,
+      "loss": 0.7609,
+      "step": 2945
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3462083058728007,
+      "learning_rate": 9.355212284400668e-06,
+      "loss": 0.6818,
+      "step": 2946
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3883685636459946,
+      "learning_rate": 9.354703060306546e-06,
+      "loss": 0.7985,
+      "step": 2947
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2793940372607795,
+      "learning_rate": 9.354193649081134e-06,
+      "loss": 0.689,
+      "step": 2948
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5778443774749407,
+      "learning_rate": 9.353684050746323e-06,
+      "loss": 0.7203,
+      "step": 2949
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.347323681698391,
+      "learning_rate": 9.353174265324014e-06,
+      "loss": 0.7099,
+      "step": 2950
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3697512453939171,
+      "learning_rate": 9.35266429283611e-06,
+      "loss": 0.737,
+      "step": 2951
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1799416350630145,
+      "learning_rate": 9.352154133304528e-06,
+      "loss": 0.6984,
+      "step": 2952
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.148997188612234,
+      "learning_rate": 9.351643786751191e-06,
+      "loss": 0.6893,
+      "step": 2953
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.406524991740243,
+      "learning_rate": 9.351133253198027e-06,
+      "loss": 0.7871,
+      "step": 2954
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.217940603886023,
+      "learning_rate": 9.350622532666979e-06,
+      "loss": 0.7306,
+      "step": 2955
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1671965569462655,
+      "learning_rate": 9.350111625179991e-06,
+      "loss": 0.8287,
+      "step": 2956
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2563173482567753,
+      "learning_rate": 9.349600530759018e-06,
+      "loss": 0.6718,
+      "step": 2957
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3547656339939995,
+      "learning_rate": 9.349089249426022e-06,
+      "loss": 0.6555,
+      "step": 2958
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1154471450354575,
+      "learning_rate": 9.348577781202976e-06,
+      "loss": 0.6107,
+      "step": 2959
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4028866257080634,
+      "learning_rate": 9.348066126111857e-06,
+      "loss": 0.7256,
+      "step": 2960
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.165260660930301,
+      "learning_rate": 9.347554284174654e-06,
+      "loss": 0.6347,
+      "step": 2961
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0717013775007838,
+      "learning_rate": 9.347042255413362e-06,
+      "loss": 0.687,
+      "step": 2962
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3260561476377766,
+      "learning_rate": 9.34653003984998e-06,
+      "loss": 0.7426,
+      "step": 2963
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.348483294958972,
+      "learning_rate": 9.346017637506523e-06,
+      "loss": 0.7381,
+      "step": 2964
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.322345401417197,
+      "learning_rate": 9.34550504840501e-06,
+      "loss": 0.69,
+      "step": 2965
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4381415732903846,
+      "learning_rate": 9.344992272567467e-06,
+      "loss": 0.751,
+      "step": 2966
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5240598796835407,
+      "learning_rate": 9.344479310015928e-06,
+      "loss": 0.8645,
+      "step": 2967
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2852957815279356,
+      "learning_rate": 9.343966160772438e-06,
+      "loss": 0.721,
+      "step": 2968
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4777067856190536,
+      "learning_rate": 9.343452824859048e-06,
+      "loss": 0.7308,
+      "step": 2969
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4103076710636597,
+      "learning_rate": 9.342939302297816e-06,
+      "loss": 0.6897,
+      "step": 2970
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.144158797266824,
+      "learning_rate": 9.34242559311081e-06,
+      "loss": 0.5939,
+      "step": 2971
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.39259921429325,
+      "learning_rate": 9.341911697320105e-06,
+      "loss": 0.7234,
+      "step": 2972
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5303848633117794,
+      "learning_rate": 9.341397614947785e-06,
+      "loss": 0.8191,
+      "step": 2973
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3087508701967536,
+      "learning_rate": 9.340883346015941e-06,
+      "loss": 0.7646,
+      "step": 2974
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0249263749467201,
+      "learning_rate": 9.340368890546672e-06,
+      "loss": 0.6361,
+      "step": 2975
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3942114768745406,
+      "learning_rate": 9.339854248562084e-06,
+      "loss": 0.6964,
+      "step": 2976
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6832449436726844,
+      "learning_rate": 9.339339420084295e-06,
+      "loss": 0.8441,
+      "step": 2977
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5255782168031358,
+      "learning_rate": 9.338824405135425e-06,
+      "loss": 0.7563,
+      "step": 2978
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3224987730432667,
+      "learning_rate": 9.338309203737609e-06,
+      "loss": 0.6508,
+      "step": 2979
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3943139809757936,
+      "learning_rate": 9.337793815912984e-06,
+      "loss": 0.6687,
+      "step": 2980
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.087191058850369,
+      "learning_rate": 9.337278241683698e-06,
+      "loss": 0.6136,
+      "step": 2981
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4320750822559702,
+      "learning_rate": 9.336762481071906e-06,
+      "loss": 0.9611,
+      "step": 2982
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.659531114091311,
+      "learning_rate": 9.336246534099772e-06,
+      "loss": 0.874,
+      "step": 2983
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.331457261442373,
+      "learning_rate": 9.335730400789466e-06,
+      "loss": 0.6447,
+      "step": 2984
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.32971881068208,
+      "learning_rate": 9.335214081163171e-06,
+      "loss": 0.7083,
+      "step": 2985
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2262242348910295,
+      "learning_rate": 9.33469757524307e-06,
+      "loss": 0.6664,
+      "step": 2986
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3127599763417261,
+      "learning_rate": 9.33418088305136e-06,
+      "loss": 0.8191,
+      "step": 2987
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.108654510163025,
+      "learning_rate": 9.333664004610246e-06,
+      "loss": 0.7364,
+      "step": 2988
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3545400542894974,
+      "learning_rate": 9.333146939941938e-06,
+      "loss": 0.7006,
+      "step": 2989
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6610442696814613,
+      "learning_rate": 9.332629689068655e-06,
+      "loss": 0.7816,
+      "step": 2990
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.431049896825617,
+      "learning_rate": 9.332112252012625e-06,
+      "loss": 0.6189,
+      "step": 2991
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2125233908747788,
+      "learning_rate": 9.331594628796083e-06,
+      "loss": 0.5664,
+      "step": 2992
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3128600183726131,
+      "learning_rate": 9.331076819441273e-06,
+      "loss": 0.6619,
+      "step": 2993
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4458374191308478,
+      "learning_rate": 9.330558823970448e-06,
+      "loss": 0.6864,
+      "step": 2994
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4022046601999947,
+      "learning_rate": 9.330040642405864e-06,
+      "loss": 0.7065,
+      "step": 2995
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7106547845889266,
+      "learning_rate": 9.329522274769791e-06,
+      "loss": 0.6198,
+      "step": 2996
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2803384893096974,
+      "learning_rate": 9.329003721084501e-06,
+      "loss": 0.6579,
+      "step": 2997
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2682859737217853,
+      "learning_rate": 9.328484981372283e-06,
+      "loss": 0.6832,
+      "step": 2998
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2572861082220763,
+      "learning_rate": 9.327966055655424e-06,
+      "loss": 0.6203,
+      "step": 2999
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4159685988586066,
+      "learning_rate": 9.327446943956224e-06,
+      "loss": 0.7947,
+      "step": 3000
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4399718914647444,
+      "learning_rate": 9.326927646296992e-06,
+      "loss": 0.6715,
+      "step": 3001
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3972179143559669,
+      "learning_rate": 9.326408162700043e-06,
+      "loss": 0.7283,
+      "step": 3002
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5043852691693045,
+      "learning_rate": 9.325888493187699e-06,
+      "loss": 0.7204,
+      "step": 3003
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3356757364647134,
+      "learning_rate": 9.325368637782292e-06,
+      "loss": 0.6275,
+      "step": 3004
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.403042643890403,
+      "learning_rate": 9.324848596506162e-06,
+      "loss": 0.7565,
+      "step": 3005
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.307291512988638,
+      "learning_rate": 9.324328369381654e-06,
+      "loss": 0.7982,
+      "step": 3006
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2196214848491123,
+      "learning_rate": 9.323807956431128e-06,
+      "loss": 0.6996,
+      "step": 3007
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3593811137082883,
+      "learning_rate": 9.323287357676944e-06,
+      "loss": 0.7596,
+      "step": 3008
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2583184587674383,
+      "learning_rate": 9.322766573141473e-06,
+      "loss": 0.7224,
+      "step": 3009
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2493323641080296,
+      "learning_rate": 9.322245602847094e-06,
+      "loss": 0.6876,
+      "step": 3010
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4041037058447698,
+      "learning_rate": 9.321724446816197e-06,
+      "loss": 0.7359,
+      "step": 3011
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2080863267318416,
+      "learning_rate": 9.321203105071176e-06,
+      "loss": 0.744,
+      "step": 3012
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.359895111033582,
+      "learning_rate": 9.320681577634432e-06,
+      "loss": 0.6879,
+      "step": 3013
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2880995165583518,
+      "learning_rate": 9.320159864528378e-06,
+      "loss": 0.7509,
+      "step": 3014
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2195653285123607,
+      "learning_rate": 9.319637965775432e-06,
+      "loss": 0.6686,
+      "step": 3015
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6322413744904947,
+      "learning_rate": 9.319115881398024e-06,
+      "loss": 0.6626,
+      "step": 3016
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5541616964583382,
+      "learning_rate": 9.31859361141859e-06,
+      "loss": 0.7781,
+      "step": 3017
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3360285787775548,
+      "learning_rate": 9.318071155859567e-06,
+      "loss": 0.6752,
+      "step": 3018
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2243012626282375,
+      "learning_rate": 9.31754851474341e-06,
+      "loss": 0.6232,
+      "step": 3019
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3606345334445802,
+      "learning_rate": 9.317025688092577e-06,
+      "loss": 0.7065,
+      "step": 3020
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.248664461137939,
+      "learning_rate": 9.316502675929537e-06,
+      "loss": 0.7317,
+      "step": 3021
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4748092108049502,
+      "learning_rate": 9.315979478276763e-06,
+      "loss": 0.7279,
+      "step": 3022
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1651790267630007,
+      "learning_rate": 9.315456095156739e-06,
+      "loss": 0.756,
+      "step": 3023
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.15306259251154,
+      "learning_rate": 9.314932526591956e-06,
+      "loss": 0.7422,
+      "step": 3024
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4472238440100036,
+      "learning_rate": 9.314408772604911e-06,
+      "loss": 0.6125,
+      "step": 3025
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.093998902707926,
+      "learning_rate": 9.313884833218115e-06,
+      "loss": 0.8199,
+      "step": 3026
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2681801408736568,
+      "learning_rate": 9.313360708454078e-06,
+      "loss": 0.694,
+      "step": 3027
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.208447731130712,
+      "learning_rate": 9.312836398335326e-06,
+      "loss": 0.6596,
+      "step": 3028
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1791799284170978,
+      "learning_rate": 9.312311902884388e-06,
+      "loss": 0.6736,
+      "step": 3029
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2548056558174725,
+      "learning_rate": 9.311787222123803e-06,
+      "loss": 0.6669,
+      "step": 3030
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4146732935257202,
+      "learning_rate": 9.311262356076118e-06,
+      "loss": 0.7208,
+      "step": 3031
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3761632976177818,
+      "learning_rate": 9.31073730476389e-06,
+      "loss": 0.7668,
+      "step": 3032
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1690649063996563,
+      "learning_rate": 9.310212068209678e-06,
+      "loss": 0.7116,
+      "step": 3033
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3766603229629903,
+      "learning_rate": 9.309686646436053e-06,
+      "loss": 0.6453,
+      "step": 3034
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3289142474210562,
+      "learning_rate": 9.309161039465597e-06,
+      "loss": 0.6404,
+      "step": 3035
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.23639261673957,
+      "learning_rate": 9.308635247320893e-06,
+      "loss": 0.7371,
+      "step": 3036
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9381241405940672,
+      "learning_rate": 9.308109270024537e-06,
+      "loss": 0.789,
+      "step": 3037
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1964329612392413,
+      "learning_rate": 9.30758310759913e-06,
+      "loss": 0.6636,
+      "step": 3038
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.356278496322358,
+      "learning_rate": 9.307056760067284e-06,
+      "loss": 0.6562,
+      "step": 3039
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4386384975962194,
+      "learning_rate": 9.306530227451616e-06,
+      "loss": 0.7635,
+      "step": 3040
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3758404684043992,
+      "learning_rate": 9.306003509774755e-06,
+      "loss": 0.6501,
+      "step": 3041
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5090150623063763,
+      "learning_rate": 9.305476607059332e-06,
+      "loss": 0.7028,
+      "step": 3042
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4400341846572344,
+      "learning_rate": 9.30494951932799e-06,
+      "loss": 0.7665,
+      "step": 3043
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3683390371027169,
+      "learning_rate": 9.30442224660338e-06,
+      "loss": 0.6581,
+      "step": 3044
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4453093736300076,
+      "learning_rate": 9.303894788908158e-06,
+      "loss": 0.65,
+      "step": 3045
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.207222121919483,
+      "learning_rate": 9.303367146264995e-06,
+      "loss": 0.5922,
+      "step": 3046
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4390454227303213,
+      "learning_rate": 9.302839318696558e-06,
+      "loss": 0.7327,
+      "step": 3047
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3375301899686334,
+      "learning_rate": 9.302311306225534e-06,
+      "loss": 0.7451,
+      "step": 3048
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2761302925687292,
+      "learning_rate": 9.301783108874611e-06,
+      "loss": 0.647,
+      "step": 3049
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.33273126319384,
+      "learning_rate": 9.301254726666488e-06,
+      "loss": 0.7983,
+      "step": 3050
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5185697385088044,
+      "learning_rate": 9.300726159623869e-06,
+      "loss": 0.786,
+      "step": 3051
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3233079494045885,
+      "learning_rate": 9.300197407769472e-06,
+      "loss": 0.6452,
+      "step": 3052
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2395482180154458,
+      "learning_rate": 9.299668471126011e-06,
+      "loss": 0.7075,
+      "step": 3053
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5989345233793104,
+      "learning_rate": 9.299139349716221e-06,
+      "loss": 0.6137,
+      "step": 3054
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4657538485983392,
+      "learning_rate": 9.29861004356284e-06,
+      "loss": 0.7382,
+      "step": 3055
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3864785899660617,
+      "learning_rate": 9.29808055268861e-06,
+      "loss": 0.6855,
+      "step": 3056
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5214449224094375,
+      "learning_rate": 9.297550877116288e-06,
+      "loss": 0.8028,
+      "step": 3057
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3155844976766775,
+      "learning_rate": 9.297021016868634e-06,
+      "loss": 0.7016,
+      "step": 3058
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2180569635984986,
+      "learning_rate": 9.296490971968416e-06,
+      "loss": 0.6354,
+      "step": 3059
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2465545233022397,
+      "learning_rate": 9.295960742438413e-06,
+      "loss": 0.6682,
+      "step": 3060
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2011682831149066,
+      "learning_rate": 9.295430328301406e-06,
+      "loss": 0.6042,
+      "step": 3061
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.228382911157058,
+      "learning_rate": 9.294899729580196e-06,
+      "loss": 0.6621,
+      "step": 3062
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2724730746437054,
+      "learning_rate": 9.294368946297576e-06,
+      "loss": 0.6145,
+      "step": 3063
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.225891663520107,
+      "learning_rate": 9.293837978476359e-06,
+      "loss": 0.6247,
+      "step": 3064
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.367225663025415,
+      "learning_rate": 9.293306826139361e-06,
+      "loss": 0.7632,
+      "step": 3065
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.160860262117841,
+      "learning_rate": 9.292775489309409e-06,
+      "loss": 0.5858,
+      "step": 3066
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3548351468184987,
+      "learning_rate": 9.292243968009332e-06,
+      "loss": 0.689,
+      "step": 3067
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.74115892389434,
+      "learning_rate": 9.291712262261972e-06,
+      "loss": 0.6776,
+      "step": 3068
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9335897867275345,
+      "learning_rate": 9.291180372090178e-06,
+      "loss": 0.6885,
+      "step": 3069
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2971156844388672,
+      "learning_rate": 9.290648297516806e-06,
+      "loss": 0.5956,
+      "step": 3070
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3999375227661492,
+      "learning_rate": 9.290116038564722e-06,
+      "loss": 0.63,
+      "step": 3071
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3866798931684805,
+      "learning_rate": 9.2895835952568e-06,
+      "loss": 0.7657,
+      "step": 3072
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0882721741544743,
+      "learning_rate": 9.289050967615914e-06,
+      "loss": 0.6267,
+      "step": 3073
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.274184131043298,
+      "learning_rate": 9.288518155664956e-06,
+      "loss": 0.7127,
+      "step": 3074
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2759887143749118,
+      "learning_rate": 9.28798515942682e-06,
+      "loss": 0.6543,
+      "step": 3075
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0940444380106407,
+      "learning_rate": 9.287451978924415e-06,
+      "loss": 0.6371,
+      "step": 3076
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6653157943001538,
+      "learning_rate": 9.28691861418065e-06,
+      "loss": 0.692,
+      "step": 3077
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.822924345104088,
+      "learning_rate": 9.286385065218442e-06,
+      "loss": 0.6359,
+      "step": 3078
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3520694090175225,
+      "learning_rate": 9.285851332060722e-06,
+      "loss": 0.762,
+      "step": 3079
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3328906716353668,
+      "learning_rate": 9.285317414730427e-06,
+      "loss": 0.6543,
+      "step": 3080
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2864368374163617,
+      "learning_rate": 9.284783313250499e-06,
+      "loss": 0.713,
+      "step": 3081
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3983688647063786,
+      "learning_rate": 9.284249027643887e-06,
+      "loss": 0.6948,
+      "step": 3082
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1267955162295291,
+      "learning_rate": 9.283714557933554e-06,
+      "loss": 0.6506,
+      "step": 3083
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.60900326771491,
+      "learning_rate": 9.283179904142465e-06,
+      "loss": 0.689,
+      "step": 3084
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.49516074090174,
+      "learning_rate": 9.282645066293597e-06,
+      "loss": 0.5953,
+      "step": 3085
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5710000889735571,
+      "learning_rate": 9.282110044409933e-06,
+      "loss": 0.6968,
+      "step": 3086
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0767247140557725,
+      "learning_rate": 9.281574838514464e-06,
+      "loss": 0.7623,
+      "step": 3087
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7024775156981564,
+      "learning_rate": 9.281039448630187e-06,
+      "loss": 0.6961,
+      "step": 3088
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3434255418443657,
+      "learning_rate": 9.280503874780112e-06,
+      "loss": 0.6972,
+      "step": 3089
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4076228589866788,
+      "learning_rate": 9.279968116987253e-06,
+      "loss": 0.7698,
+      "step": 3090
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2314536747863767,
+      "learning_rate": 9.27943217527463e-06,
+      "loss": 0.6567,
+      "step": 3091
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.353146479785923,
+      "learning_rate": 9.278896049665276e-06,
+      "loss": 0.7383,
+      "step": 3092
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.318942314072863,
+      "learning_rate": 9.27835974018223e-06,
+      "loss": 0.671,
+      "step": 3093
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3219428062499732,
+      "learning_rate": 9.277823246848537e-06,
+      "loss": 0.6616,
+      "step": 3094
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2615974903461276,
+      "learning_rate": 9.277286569687252e-06,
+      "loss": 0.7025,
+      "step": 3095
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2873067127334377,
+      "learning_rate": 9.276749708721438e-06,
+      "loss": 0.5946,
+      "step": 3096
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.425204625037313,
+      "learning_rate": 9.276212663974163e-06,
+      "loss": 0.7251,
+      "step": 3097
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.247350951711012,
+      "learning_rate": 9.275675435468505e-06,
+      "loss": 0.6763,
+      "step": 3098
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2321257971439847,
+      "learning_rate": 9.275138023227555e-06,
+      "loss": 0.7082,
+      "step": 3099
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.182445874674696,
+      "learning_rate": 9.274600427274399e-06,
+      "loss": 0.534,
+      "step": 3100
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0464625528724634,
+      "learning_rate": 9.274062647632144e-06,
+      "loss": 0.6372,
+      "step": 3101
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.416518671626807,
+      "learning_rate": 9.273524684323897e-06,
+      "loss": 0.6667,
+      "step": 3102
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0589609071642885,
+      "learning_rate": 9.27298653737278e-06,
+      "loss": 0.61,
+      "step": 3103
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3141909519005015,
+      "learning_rate": 9.272448206801912e-06,
+      "loss": 0.7253,
+      "step": 3104
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4071587674229664,
+      "learning_rate": 9.271909692634431e-06,
+      "loss": 0.5943,
+      "step": 3105
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4710912216473897,
+      "learning_rate": 9.271370994893475e-06,
+      "loss": 0.7481,
+      "step": 3106
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1877281776247681,
+      "learning_rate": 9.270832113602196e-06,
+      "loss": 0.8176,
+      "step": 3107
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.296166922468527,
+      "learning_rate": 9.270293048783747e-06,
+      "loss": 0.6703,
+      "step": 3108
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3455137855195014,
+      "learning_rate": 9.269753800461299e-06,
+      "loss": 0.6758,
+      "step": 3109
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3352130006847422,
+      "learning_rate": 9.269214368658016e-06,
+      "loss": 0.7512,
+      "step": 3110
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3843563509934464,
+      "learning_rate": 9.268674753397087e-06,
+      "loss": 0.6935,
+      "step": 3111
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.433426749228249,
+      "learning_rate": 9.268134954701697e-06,
+      "loss": 0.743,
+      "step": 3112
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6608210718821799,
+      "learning_rate": 9.267594972595042e-06,
+      "loss": 0.7617,
+      "step": 3113
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.468857763586006,
+      "learning_rate": 9.267054807100327e-06,
+      "loss": 0.8076,
+      "step": 3114
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.792720886052483,
+      "learning_rate": 9.266514458240762e-06,
+      "loss": 0.7088,
+      "step": 3115
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.383382132358956,
+      "learning_rate": 9.26597392603957e-06,
+      "loss": 0.758,
+      "step": 3116
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1884128544654633,
+      "learning_rate": 9.265433210519976e-06,
+      "loss": 0.6073,
+      "step": 3117
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1146672933990938,
+      "learning_rate": 9.26489231170522e-06,
+      "loss": 0.5493,
+      "step": 3118
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2596219707284881,
+      "learning_rate": 9.264351229618541e-06,
+      "loss": 0.5741,
+      "step": 3119
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4693966484876229,
+      "learning_rate": 9.263809964283192e-06,
+      "loss": 0.7896,
+      "step": 3120
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.326724181073313,
+      "learning_rate": 9.263268515722435e-06,
+      "loss": 0.7994,
+      "step": 3121
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3494609888835154,
+      "learning_rate": 9.262726883959535e-06,
+      "loss": 0.6536,
+      "step": 3122
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2179356326383264,
+      "learning_rate": 9.262185069017764e-06,
+      "loss": 0.7176,
+      "step": 3123
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.362407201919385,
+      "learning_rate": 9.261643070920409e-06,
+      "loss": 0.6325,
+      "step": 3124
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.261457191731673,
+      "learning_rate": 9.261100889690763e-06,
+      "loss": 0.6657,
+      "step": 3125
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.70127529692071,
+      "learning_rate": 9.26055852535212e-06,
+      "loss": 0.885,
+      "step": 3126
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3630935085708271,
+      "learning_rate": 9.260015977927788e-06,
+      "loss": 0.6693,
+      "step": 3127
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3096858827836768,
+      "learning_rate": 9.259473247441081e-06,
+      "loss": 0.7063,
+      "step": 3128
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1265414048350215,
+      "learning_rate": 9.258930333915325e-06,
+      "loss": 0.6951,
+      "step": 3129
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5282516258345513,
+      "learning_rate": 9.258387237373845e-06,
+      "loss": 0.7724,
+      "step": 3130
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1298308833527295,
+      "learning_rate": 9.257843957839982e-06,
+      "loss": 0.6232,
+      "step": 3131
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.289688117475435,
+      "learning_rate": 9.257300495337082e-06,
+      "loss": 0.6945,
+      "step": 3132
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.155896304215239,
+      "learning_rate": 9.256756849888499e-06,
+      "loss": 0.802,
+      "step": 3133
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2547940826822572,
+      "learning_rate": 9.256213021517593e-06,
+      "loss": 0.6821,
+      "step": 3134
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.278148983075184,
+      "learning_rate": 9.255669010247734e-06,
+      "loss": 0.7811,
+      "step": 3135
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2325948556229913,
+      "learning_rate": 9.2551248161023e-06,
+      "loss": 0.5913,
+      "step": 3136
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2535496993672801,
+      "learning_rate": 9.254580439104677e-06,
+      "loss": 0.6083,
+      "step": 3137
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0764896693365402,
+      "learning_rate": 9.254035879278255e-06,
+      "loss": 0.6767,
+      "step": 3138
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.411349251093626,
+      "learning_rate": 9.253491136646437e-06,
+      "loss": 0.6421,
+      "step": 3139
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1621813395421055,
+      "learning_rate": 9.252946211232634e-06,
+      "loss": 0.5721,
+      "step": 3140
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2000547797365941,
+      "learning_rate": 9.25240110306026e-06,
+      "loss": 0.6552,
+      "step": 3141
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3616141183524428,
+      "learning_rate": 9.25185581215274e-06,
+      "loss": 0.6803,
+      "step": 3142
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3186978341029825,
+      "learning_rate": 9.251310338533504e-06,
+      "loss": 0.7375,
+      "step": 3143
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4070116151303573,
+      "learning_rate": 9.250764682225997e-06,
+      "loss": 0.7075,
+      "step": 3144
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2895391677124912,
+      "learning_rate": 9.250218843253663e-06,
+      "loss": 0.7065,
+      "step": 3145
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.084860662735273,
+      "learning_rate": 9.24967282163996e-06,
+      "loss": 0.6637,
+      "step": 3146
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2869723905710293,
+      "learning_rate": 9.249126617408353e-06,
+      "loss": 0.5611,
+      "step": 3147
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2450777185492696,
+      "learning_rate": 9.248580230582311e-06,
+      "loss": 0.7005,
+      "step": 3148
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1883208043196607,
+      "learning_rate": 9.248033661185313e-06,
+      "loss": 0.7377,
+      "step": 3149
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5650465650950371,
+      "learning_rate": 9.247486909240849e-06,
+      "loss": 0.7969,
+      "step": 3150
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6344562850748228,
+      "learning_rate": 9.246939974772413e-06,
+      "loss": 0.7392,
+      "step": 3151
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.581615597727696,
+      "learning_rate": 9.246392857803508e-06,
+      "loss": 0.9398,
+      "step": 3152
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0959234294576392,
+      "learning_rate": 9.245845558357643e-06,
+      "loss": 0.7356,
+      "step": 3153
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.215077357712934,
+      "learning_rate": 9.24529807645834e-06,
+      "loss": 0.6371,
+      "step": 3154
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4315738580346158,
+      "learning_rate": 9.244750412129123e-06,
+      "loss": 0.6332,
+      "step": 3155
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1067884683817568,
+      "learning_rate": 9.244202565393528e-06,
+      "loss": 0.5756,
+      "step": 3156
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2149335042971574,
+      "learning_rate": 9.243654536275095e-06,
+      "loss": 0.6797,
+      "step": 3157
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3532617670604838,
+      "learning_rate": 9.243106324797377e-06,
+      "loss": 0.7658,
+      "step": 3158
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4143483738271017,
+      "learning_rate": 9.24255793098393e-06,
+      "loss": 0.6519,
+      "step": 3159
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2227956570465275,
+      "learning_rate": 9.24200935485832e-06,
+      "loss": 0.6025,
+      "step": 3160
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7270199250816307,
+      "learning_rate": 9.24146059644412e-06,
+      "loss": 0.823,
+      "step": 3161
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.362429977764866,
+      "learning_rate": 9.240911655764912e-06,
+      "loss": 0.5993,
+      "step": 3162
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1429998379684052,
+      "learning_rate": 9.240362532844287e-06,
+      "loss": 0.621,
+      "step": 3163
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.357662564289905,
+      "learning_rate": 9.23981322770584e-06,
+      "loss": 0.6667,
+      "step": 3164
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4274052622887226,
+      "learning_rate": 9.239263740373173e-06,
+      "loss": 0.8149,
+      "step": 3165
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3827323317522913,
+      "learning_rate": 9.238714070869903e-06,
+      "loss": 0.6153,
+      "step": 3166
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3071499024799784,
+      "learning_rate": 9.238164219219649e-06,
+      "loss": 0.8118,
+      "step": 3167
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.570721485526055,
+      "learning_rate": 9.23761418544604e-06,
+      "loss": 0.6836,
+      "step": 3168
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.243712748435271,
+      "learning_rate": 9.237063969572713e-06,
+      "loss": 0.6674,
+      "step": 3169
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2164262054039474,
+      "learning_rate": 9.23651357162331e-06,
+      "loss": 0.674,
+      "step": 3170
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3888244424122882,
+      "learning_rate": 9.235962991621484e-06,
+      "loss": 0.7228,
+      "step": 3171
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6969938553809123,
+      "learning_rate": 9.235412229590893e-06,
+      "loss": 0.6583,
+      "step": 3172
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4463324083686342,
+      "learning_rate": 9.234861285555207e-06,
+      "loss": 0.719,
+      "step": 3173
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.393808991739731,
+      "learning_rate": 9.2343101595381e-06,
+      "loss": 0.741,
+      "step": 3174
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2666502345829123,
+      "learning_rate": 9.233758851563254e-06,
+      "loss": 0.7002,
+      "step": 3175
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4827114062620375,
+      "learning_rate": 9.233207361654362e-06,
+      "loss": 0.8615,
+      "step": 3176
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2445062077542837,
+      "learning_rate": 9.23265568983512e-06,
+      "loss": 0.6468,
+      "step": 3177
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1261154730841652,
+      "learning_rate": 9.232103836129239e-06,
+      "loss": 0.6494,
+      "step": 3178
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4634231014992911,
+      "learning_rate": 9.23155180056043e-06,
+      "loss": 0.6184,
+      "step": 3179
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.449529278789235,
+      "learning_rate": 9.230999583152413e-06,
+      "loss": 0.7468,
+      "step": 3180
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.383358881481872,
+      "learning_rate": 9.230447183928926e-06,
+      "loss": 0.6565,
+      "step": 3181
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1620943723451804,
+      "learning_rate": 9.229894602913698e-06,
+      "loss": 0.6408,
+      "step": 3182
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2145217039654466,
+      "learning_rate": 9.22934184013048e-06,
+      "loss": 0.5739,
+      "step": 3183
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.026600017267716,
+      "learning_rate": 9.228788895603024e-06,
+      "loss": 0.627,
+      "step": 3184
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3906587354157005,
+      "learning_rate": 9.22823576935509e-06,
+      "loss": 0.7487,
+      "step": 3185
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.24481175033519,
+      "learning_rate": 9.22768246141045e-06,
+      "loss": 0.75,
+      "step": 3186
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3648483155510163,
+      "learning_rate": 9.227128971792877e-06,
+      "loss": 0.6222,
+      "step": 3187
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5901783637545224,
+      "learning_rate": 9.226575300526159e-06,
+      "loss": 0.8255,
+      "step": 3188
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2031395435581866,
+      "learning_rate": 9.226021447634085e-06,
+      "loss": 0.5829,
+      "step": 3189
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1287531666785136,
+      "learning_rate": 9.22546741314046e-06,
+      "loss": 0.5963,
+      "step": 3190
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6716843099402023,
+      "learning_rate": 9.22491319706909e-06,
+      "loss": 0.696,
+      "step": 3191
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3603693275820399,
+      "learning_rate": 9.224358799443791e-06,
+      "loss": 0.6952,
+      "step": 3192
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2076819843655415,
+      "learning_rate": 9.223804220288384e-06,
+      "loss": 0.704,
+      "step": 3193
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4855601887373666,
+      "learning_rate": 9.223249459626704e-06,
+      "loss": 0.769,
+      "step": 3194
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0538307239274305,
+      "learning_rate": 9.22269451748259e-06,
+      "loss": 0.5905,
+      "step": 3195
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5846549698362564,
+      "learning_rate": 9.222139393879885e-06,
+      "loss": 0.726,
+      "step": 3196
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4339190121527903,
+      "learning_rate": 9.22158408884245e-06,
+      "loss": 0.6433,
+      "step": 3197
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.459749946944265,
+      "learning_rate": 9.221028602394144e-06,
+      "loss": 0.7196,
+      "step": 3198
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4875636939764705,
+      "learning_rate": 9.220472934558838e-06,
+      "loss": 0.6676,
+      "step": 3199
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4777932418592106,
+      "learning_rate": 9.21991708536041e-06,
+      "loss": 0.6707,
+      "step": 3200
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.322137728368665,
+      "learning_rate": 9.219361054822746e-06,
+      "loss": 0.7599,
+      "step": 3201
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1290565360111442,
+      "learning_rate": 9.218804842969742e-06,
+      "loss": 0.5841,
+      "step": 3202
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4339619074688437,
+      "learning_rate": 9.218248449825297e-06,
+      "loss": 0.6541,
+      "step": 3203
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3789681381466974,
+      "learning_rate": 9.217691875413323e-06,
+      "loss": 0.8303,
+      "step": 3204
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3845370046461354,
+      "learning_rate": 9.217135119757734e-06,
+      "loss": 0.6473,
+      "step": 3205
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3170416459711232,
+      "learning_rate": 9.216578182882459e-06,
+      "loss": 0.6331,
+      "step": 3206
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3904984479339857,
+      "learning_rate": 9.216021064811428e-06,
+      "loss": 0.7489,
+      "step": 3207
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4749656019344577,
+      "learning_rate": 9.21546376556858e-06,
+      "loss": 0.6866,
+      "step": 3208
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1459325093993922,
+      "learning_rate": 9.214906285177867e-06,
+      "loss": 0.6022,
+      "step": 3209
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9581372762326251,
+      "learning_rate": 9.214348623663244e-06,
+      "loss": 0.5829,
+      "step": 3210
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.429977390208924,
+      "learning_rate": 9.213790781048677e-06,
+      "loss": 0.6967,
+      "step": 3211
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3235725505750866,
+      "learning_rate": 9.213232757358133e-06,
+      "loss": 0.7058,
+      "step": 3212
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3333344029158924,
+      "learning_rate": 9.212674552615594e-06,
+      "loss": 0.7852,
+      "step": 3213
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5612101695777176,
+      "learning_rate": 9.212116166845048e-06,
+      "loss": 0.658,
+      "step": 3214
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0102710730477265,
+      "learning_rate": 9.211557600070489e-06,
+      "loss": 0.63,
+      "step": 3215
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3887243525635227,
+      "learning_rate": 9.21099885231592e-06,
+      "loss": 0.7233,
+      "step": 3216
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4729215364406594,
+      "learning_rate": 9.210439923605352e-06,
+      "loss": 0.7633,
+      "step": 3217
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.271983866157681,
+      "learning_rate": 9.209880813962806e-06,
+      "loss": 0.7332,
+      "step": 3218
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2145077641398079,
+      "learning_rate": 9.209321523412303e-06,
+      "loss": 0.7164,
+      "step": 3219
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4922137296223275,
+      "learning_rate": 9.208762051977879e-06,
+      "loss": 0.6827,
+      "step": 3220
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5673742893740437,
+      "learning_rate": 9.208202399683577e-06,
+      "loss": 0.8274,
+      "step": 3221
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.447859867444778,
+      "learning_rate": 9.207642566553445e-06,
+      "loss": 0.6954,
+      "step": 3222
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.47966356946058,
+      "learning_rate": 9.207082552611541e-06,
+      "loss": 0.7129,
+      "step": 3223
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5299577063119987,
+      "learning_rate": 9.206522357881931e-06,
+      "loss": 0.6983,
+      "step": 3224
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5374216182589868,
+      "learning_rate": 9.205961982388686e-06,
+      "loss": 0.7518,
+      "step": 3225
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4663710971172919,
+      "learning_rate": 9.205401426155889e-06,
+      "loss": 0.6869,
+      "step": 3226
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.049972528360513,
+      "learning_rate": 9.204840689207626e-06,
+      "loss": 0.8316,
+      "step": 3227
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3634653324579744,
+      "learning_rate": 9.204279771567994e-06,
+      "loss": 0.7355,
+      "step": 3228
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1076762301396896,
+      "learning_rate": 9.203718673261098e-06,
+      "loss": 0.652,
+      "step": 3229
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2268028197840353,
+      "learning_rate": 9.203157394311045e-06,
+      "loss": 0.6197,
+      "step": 3230
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2952343718434103,
+      "learning_rate": 9.20259593474196e-06,
+      "loss": 0.7797,
+      "step": 3231
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2911043940333002,
+      "learning_rate": 9.202034294577968e-06,
+      "loss": 0.7083,
+      "step": 3232
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2477619698208078,
+      "learning_rate": 9.201472473843204e-06,
+      "loss": 0.7385,
+      "step": 3233
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.40051868036117,
+      "learning_rate": 9.20091047256181e-06,
+      "loss": 0.6936,
+      "step": 3234
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8900771674769973,
+      "learning_rate": 9.200348290757937e-06,
+      "loss": 0.5935,
+      "step": 3235
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3820013340134272,
+      "learning_rate": 9.199785928455745e-06,
+      "loss": 0.7135,
+      "step": 3236
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3893749753435194,
+      "learning_rate": 9.199223385679396e-06,
+      "loss": 0.7894,
+      "step": 3237
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4779347611285618,
+      "learning_rate": 9.198660662453066e-06,
+      "loss": 0.7801,
+      "step": 3238
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3853308658908643,
+      "learning_rate": 9.198097758800938e-06,
+      "loss": 0.6877,
+      "step": 3239
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2668080836884674,
+      "learning_rate": 9.197534674747199e-06,
+      "loss": 0.6652,
+      "step": 3240
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1168252082240915,
+      "learning_rate": 9.196971410316047e-06,
+      "loss": 0.6891,
+      "step": 3241
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1905160765561154,
+      "learning_rate": 9.196407965531686e-06,
+      "loss": 0.6301,
+      "step": 3242
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.293292489552429,
+      "learning_rate": 9.195844340418328e-06,
+      "loss": 0.6042,
+      "step": 3243
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4149963900225107,
+      "learning_rate": 9.195280535000196e-06,
+      "loss": 0.8028,
+      "step": 3244
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4639633490256374,
+      "learning_rate": 9.194716549301514e-06,
+      "loss": 0.7028,
+      "step": 3245
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.71799035556468,
+      "learning_rate": 9.19415238334652e-06,
+      "loss": 0.5674,
+      "step": 3246
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.866621331560199,
+      "learning_rate": 9.193588037159457e-06,
+      "loss": 0.7234,
+      "step": 3247
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3560043187801123,
+      "learning_rate": 9.193023510764578e-06,
+      "loss": 0.7221,
+      "step": 3248
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4277591830646057,
+      "learning_rate": 9.19245880418614e-06,
+      "loss": 0.7098,
+      "step": 3249
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.372959281283372,
+      "learning_rate": 9.191893917448409e-06,
+      "loss": 0.6369,
+      "step": 3250
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2504833856532778,
+      "learning_rate": 9.19132885057566e-06,
+      "loss": 0.7583,
+      "step": 3251
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1529205511840184,
+      "learning_rate": 9.190763603592177e-06,
+      "loss": 0.6834,
+      "step": 3252
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2372789390066457,
+      "learning_rate": 9.190198176522249e-06,
+      "loss": 0.6398,
+      "step": 3253
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2907230835335661,
+      "learning_rate": 9.189632569390172e-06,
+      "loss": 0.7938,
+      "step": 3254
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5305333872767424,
+      "learning_rate": 9.189066782220253e-06,
+      "loss": 0.8267,
+      "step": 3255
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5108206158578992,
+      "learning_rate": 9.188500815036806e-06,
+      "loss": 0.6809,
+      "step": 3256
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5240557944819526,
+      "learning_rate": 9.18793466786415e-06,
+      "loss": 0.781,
+      "step": 3257
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2101103356043943,
+      "learning_rate": 9.187368340726615e-06,
+      "loss": 0.7106,
+      "step": 3258
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.299838245471276,
+      "learning_rate": 9.186801833648535e-06,
+      "loss": 0.7315,
+      "step": 3259
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.345296261821039,
+      "learning_rate": 9.186235146654257e-06,
+      "loss": 0.6602,
+      "step": 3260
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3744375650486567,
+      "learning_rate": 9.18566827976813e-06,
+      "loss": 0.619,
+      "step": 3261
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3435194756126088,
+      "learning_rate": 9.185101233014516e-06,
+      "loss": 0.6874,
+      "step": 3262
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.232652660135615,
+      "learning_rate": 9.184534006417782e-06,
+      "loss": 0.7952,
+      "step": 3263
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6263153029928556,
+      "learning_rate": 9.183966600002301e-06,
+      "loss": 0.7673,
+      "step": 3264
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6635873788003779,
+      "learning_rate": 9.183399013792457e-06,
+      "loss": 0.7692,
+      "step": 3265
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3734045148492633,
+      "learning_rate": 9.18283124781264e-06,
+      "loss": 0.6492,
+      "step": 3266
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1838010718290153,
+      "learning_rate": 9.18226330208725e-06,
+      "loss": 0.6096,
+      "step": 3267
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1196272627927464,
+      "learning_rate": 9.181695176640693e-06,
+      "loss": 0.7815,
+      "step": 3268
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4318273021707135,
+      "learning_rate": 9.181126871497378e-06,
+      "loss": 0.7248,
+      "step": 3269
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.360520598485208,
+      "learning_rate": 9.180558386681731e-06,
+      "loss": 0.7978,
+      "step": 3270
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5208056163353045,
+      "learning_rate": 9.179989722218179e-06,
+      "loss": 0.6578,
+      "step": 3271
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3724643368876797,
+      "learning_rate": 9.17942087813116e-06,
+      "loss": 0.7762,
+      "step": 3272
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2532833020865024,
+      "learning_rate": 9.178851854445118e-06,
+      "loss": 0.7142,
+      "step": 3273
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8768076845986754,
+      "learning_rate": 9.178282651184506e-06,
+      "loss": 0.7521,
+      "step": 3274
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.278105409781011,
+      "learning_rate": 9.177713268373781e-06,
+      "loss": 0.6462,
+      "step": 3275
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3870854215994057,
+      "learning_rate": 9.177143706037411e-06,
+      "loss": 0.6741,
+      "step": 3276
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.360866837839546,
+      "learning_rate": 9.176573964199875e-06,
+      "loss": 0.6122,
+      "step": 3277
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4013663781492036,
+      "learning_rate": 9.176004042885653e-06,
+      "loss": 0.6924,
+      "step": 3278
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.282470783134264,
+      "learning_rate": 9.175433942119238e-06,
+      "loss": 0.6271,
+      "step": 3279
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4122520961199447,
+      "learning_rate": 9.174863661925128e-06,
+      "loss": 0.6898,
+      "step": 3280
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6660633680902348,
+      "learning_rate": 9.174293202327828e-06,
+      "loss": 0.6727,
+      "step": 3281
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3796480267136895,
+      "learning_rate": 9.173722563351852e-06,
+      "loss": 0.7164,
+      "step": 3282
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.571616358894164,
+      "learning_rate": 9.173151745021722e-06,
+      "loss": 0.7593,
+      "step": 3283
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3150108496867254,
+      "learning_rate": 9.172580747361968e-06,
+      "loss": 0.6602,
+      "step": 3284
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5540049513251584,
+      "learning_rate": 9.172009570397124e-06,
+      "loss": 0.8352,
+      "step": 3285
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2985133668938158,
+      "learning_rate": 9.17143821415174e-06,
+      "loss": 0.6501,
+      "step": 3286
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3790013246849553,
+      "learning_rate": 9.170866678650366e-06,
+      "loss": 0.7525,
+      "step": 3287
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.285229544700225,
+      "learning_rate": 9.17029496391756e-06,
+      "loss": 0.6559,
+      "step": 3288
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3856925906418869,
+      "learning_rate": 9.169723069977892e-06,
+      "loss": 0.6379,
+      "step": 3289
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3599221390063019,
+      "learning_rate": 9.169150996855939e-06,
+      "loss": 0.7953,
+      "step": 3290
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3906989555303513,
+      "learning_rate": 9.16857874457628e-06,
+      "loss": 0.6872,
+      "step": 3291
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5000393861032046,
+      "learning_rate": 9.16800631316351e-06,
+      "loss": 0.7536,
+      "step": 3292
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.382506290861645,
+      "learning_rate": 9.167433702642224e-06,
+      "loss": 0.7644,
+      "step": 3293
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2968791048095105,
+      "learning_rate": 9.166860913037032e-06,
+      "loss": 0.7485,
+      "step": 3294
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6207013757799162,
+      "learning_rate": 9.166287944372546e-06,
+      "loss": 0.6828,
+      "step": 3295
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3342762733511597,
+      "learning_rate": 9.165714796673389e-06,
+      "loss": 0.7105,
+      "step": 3296
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2966348545300852,
+      "learning_rate": 9.16514146996419e-06,
+      "loss": 0.7154,
+      "step": 3297
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4276771751444797,
+      "learning_rate": 9.164567964269583e-06,
+      "loss": 0.6395,
+      "step": 3298
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1155167440498341,
+      "learning_rate": 9.163994279614218e-06,
+      "loss": 0.6354,
+      "step": 3299
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.534832667188304,
+      "learning_rate": 9.163420416022745e-06,
+      "loss": 0.707,
+      "step": 3300
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.486567928003548,
+      "learning_rate": 9.162846373519824e-06,
+      "loss": 0.8257,
+      "step": 3301
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.290893560799339,
+      "learning_rate": 9.162272152130123e-06,
+      "loss": 0.71,
+      "step": 3302
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.545661612529355,
+      "learning_rate": 9.161697751878317e-06,
+      "loss": 0.7699,
+      "step": 3303
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2839064489958953,
+      "learning_rate": 9.161123172789091e-06,
+      "loss": 0.6443,
+      "step": 3304
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.287234016305,
+      "learning_rate": 9.160548414887135e-06,
+      "loss": 0.6803,
+      "step": 3305
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.348151316836439,
+      "learning_rate": 9.159973478197148e-06,
+      "loss": 0.724,
+      "step": 3306
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2270774885783426,
+      "learning_rate": 9.159398362743834e-06,
+      "loss": 0.6932,
+      "step": 3307
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1676606985788986,
+      "learning_rate": 9.15882306855191e-06,
+      "loss": 0.6491,
+      "step": 3308
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2489857635142798,
+      "learning_rate": 9.158247595646098e-06,
+      "loss": 0.6798,
+      "step": 3309
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1865390794729318,
+      "learning_rate": 9.157671944051125e-06,
+      "loss": 0.5393,
+      "step": 3310
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3292761153019026,
+      "learning_rate": 9.157096113791727e-06,
+      "loss": 0.6393,
+      "step": 3311
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4085083699598464,
+      "learning_rate": 9.156520104892653e-06,
+      "loss": 0.6655,
+      "step": 3312
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.478024226259933,
+      "learning_rate": 9.155943917378652e-06,
+      "loss": 0.8445,
+      "step": 3313
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3040063124861432,
+      "learning_rate": 9.155367551274485e-06,
+      "loss": 0.6162,
+      "step": 3314
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5723551413661885,
+      "learning_rate": 9.15479100660492e-06,
+      "loss": 0.7218,
+      "step": 3315
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1221385680476086,
+      "learning_rate": 9.154214283394733e-06,
+      "loss": 0.5816,
+      "step": 3316
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.298265006387059,
+      "learning_rate": 9.153637381668705e-06,
+      "loss": 0.6324,
+      "step": 3317
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5139437711018866,
+      "learning_rate": 9.153060301451629e-06,
+      "loss": 0.6407,
+      "step": 3318
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1716086979508784,
+      "learning_rate": 9.152483042768302e-06,
+      "loss": 0.5705,
+      "step": 3319
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.349230334245654,
+      "learning_rate": 9.151905605643531e-06,
+      "loss": 0.6168,
+      "step": 3320
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.556449497640018,
+      "learning_rate": 9.151327990102129e-06,
+      "loss": 0.7221,
+      "step": 3321
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4759683110081123,
+      "learning_rate": 9.150750196168918e-06,
+      "loss": 0.7192,
+      "step": 3322
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4138783681840807,
+      "learning_rate": 9.150172223868727e-06,
+      "loss": 0.7179,
+      "step": 3323
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4664189515547428,
+      "learning_rate": 9.149594073226391e-06,
+      "loss": 0.6711,
+      "step": 3324
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2276916970151408,
+      "learning_rate": 9.149015744266759e-06,
+      "loss": 0.7228,
+      "step": 3325
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3921929471881214,
+      "learning_rate": 9.148437237014677e-06,
+      "loss": 0.6347,
+      "step": 3326
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4138550674487502,
+      "learning_rate": 9.14785855149501e-06,
+      "loss": 0.6421,
+      "step": 3327
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4161678681565992,
+      "learning_rate": 9.147279687732622e-06,
+      "loss": 0.7366,
+      "step": 3328
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0232613778545938,
+      "learning_rate": 9.14670064575239e-06,
+      "loss": 0.7079,
+      "step": 3329
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4286436344053137,
+      "learning_rate": 9.146121425579197e-06,
+      "loss": 0.6553,
+      "step": 3330
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.30141662849788,
+      "learning_rate": 9.14554202723793e-06,
+      "loss": 0.6043,
+      "step": 3331
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5896030275021202,
+      "learning_rate": 9.144962450753491e-06,
+      "loss": 0.6002,
+      "step": 3332
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4678173987385983,
+      "learning_rate": 9.144382696150785e-06,
+      "loss": 0.6706,
+      "step": 3333
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3454743915632967,
+      "learning_rate": 9.143802763454723e-06,
+      "loss": 0.6977,
+      "step": 3334
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2555666602513478,
+      "learning_rate": 9.143222652690227e-06,
+      "loss": 0.6112,
+      "step": 3335
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1180641323848732,
+      "learning_rate": 9.142642363882228e-06,
+      "loss": 0.7505,
+      "step": 3336
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5858479921222672,
+      "learning_rate": 9.14206189705566e-06,
+      "loss": 0.7201,
+      "step": 3337
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2773144519873019,
+      "learning_rate": 9.141481252235467e-06,
+      "loss": 0.6651,
+      "step": 3338
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3730132737385021,
+      "learning_rate": 9.140900429446601e-06,
+      "loss": 0.7126,
+      "step": 3339
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.260454019022656,
+      "learning_rate": 9.140319428714022e-06,
+      "loss": 0.689,
+      "step": 3340
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3669834952198625,
+      "learning_rate": 9.139738250062695e-06,
+      "loss": 0.6703,
+      "step": 3341
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.345939884473173,
+      "learning_rate": 9.139156893517595e-06,
+      "loss": 0.6694,
+      "step": 3342
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.498355837498313,
+      "learning_rate": 9.138575359103707e-06,
+      "loss": 0.7096,
+      "step": 3343
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5646514819190225,
+      "learning_rate": 9.137993646846018e-06,
+      "loss": 0.6838,
+      "step": 3344
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.237663859211154,
+      "learning_rate": 9.137411756769525e-06,
+      "loss": 0.6897,
+      "step": 3345
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2797066316226773,
+      "learning_rate": 9.136829688899236e-06,
+      "loss": 0.7298,
+      "step": 3346
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1290592716008803,
+      "learning_rate": 9.136247443260162e-06,
+      "loss": 0.6085,
+      "step": 3347
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4211794388778323,
+      "learning_rate": 9.135665019877323e-06,
+      "loss": 0.7457,
+      "step": 3348
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4083142552280532,
+      "learning_rate": 9.135082418775746e-06,
+      "loss": 0.6936,
+      "step": 3349
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4893298275956544,
+      "learning_rate": 9.134499639980469e-06,
+      "loss": 0.733,
+      "step": 3350
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2933647064929867,
+      "learning_rate": 9.133916683516536e-06,
+      "loss": 0.6964,
+      "step": 3351
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3899403151696723,
+      "learning_rate": 9.133333549408997e-06,
+      "loss": 0.7809,
+      "step": 3352
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3591318552640022,
+      "learning_rate": 9.132750237682907e-06,
+      "loss": 0.737,
+      "step": 3353
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.348572796308381,
+      "learning_rate": 9.132166748363335e-06,
+      "loss": 0.7137,
+      "step": 3354
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3080274374748933,
+      "learning_rate": 9.131583081475356e-06,
+      "loss": 0.5917,
+      "step": 3355
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6015771063424868,
+      "learning_rate": 9.130999237044052e-06,
+      "loss": 0.9993,
+      "step": 3356
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5603521364903192,
+      "learning_rate": 9.13041521509451e-06,
+      "loss": 0.8151,
+      "step": 3357
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0422469129296112,
+      "learning_rate": 9.129831015651827e-06,
+      "loss": 0.7137,
+      "step": 3358
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3138287499351178,
+      "learning_rate": 9.129246638741108e-06,
+      "loss": 0.8021,
+      "step": 3359
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.522302108848306,
+      "learning_rate": 9.128662084387462e-06,
+      "loss": 0.7341,
+      "step": 3360
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3816298015217048,
+      "learning_rate": 9.128077352616013e-06,
+      "loss": 0.7147,
+      "step": 3361
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3517112144238312,
+      "learning_rate": 9.127492443451887e-06,
+      "loss": 0.7321,
+      "step": 3362
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.67731803602676,
+      "learning_rate": 9.126907356920219e-06,
+      "loss": 0.7437,
+      "step": 3363
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.203675764442819,
+      "learning_rate": 9.126322093046149e-06,
+      "loss": 0.5831,
+      "step": 3364
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4803134862706537,
+      "learning_rate": 9.12573665185483e-06,
+      "loss": 0.7595,
+      "step": 3365
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2243479963649333,
+      "learning_rate": 9.125151033371417e-06,
+      "loss": 0.6132,
+      "step": 3366
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2727208445378122,
+      "learning_rate": 9.12456523762108e-06,
+      "loss": 0.6735,
+      "step": 3367
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2788826851388224,
+      "learning_rate": 9.123979264628985e-06,
+      "loss": 0.695,
+      "step": 3368
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2288218660272479,
+      "learning_rate": 9.123393114420318e-06,
+      "loss": 0.7275,
+      "step": 3369
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4002836210217435,
+      "learning_rate": 9.122806787020266e-06,
+      "loss": 0.7798,
+      "step": 3370
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3538432262870372,
+      "learning_rate": 9.122220282454024e-06,
+      "loss": 0.7348,
+      "step": 3371
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.491676073792392,
+      "learning_rate": 9.121633600746796e-06,
+      "loss": 0.7282,
+      "step": 3372
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0697055232779964,
+      "learning_rate": 9.121046741923792e-06,
+      "loss": 0.64,
+      "step": 3373
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6824597912656862,
+      "learning_rate": 9.120459706010233e-06,
+      "loss": 0.7199,
+      "step": 3374
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.676892879596947,
+      "learning_rate": 9.119872493031343e-06,
+      "loss": 0.8434,
+      "step": 3375
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.19617071284011,
+      "learning_rate": 9.119285103012356e-06,
+      "loss": 0.7333,
+      "step": 3376
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3053228391626324,
+      "learning_rate": 9.118697535978513e-06,
+      "loss": 0.7404,
+      "step": 3377
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.320314856183731,
+      "learning_rate": 9.118109791955067e-06,
+      "loss": 0.6353,
+      "step": 3378
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5133447431214924,
+      "learning_rate": 9.11752187096727e-06,
+      "loss": 0.7477,
+      "step": 3379
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1505807159024413,
+      "learning_rate": 9.116933773040389e-06,
+      "loss": 0.7096,
+      "step": 3380
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2999697353255906,
+      "learning_rate": 9.116345498199693e-06,
+      "loss": 0.6088,
+      "step": 3381
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4985535589503958,
+      "learning_rate": 9.115757046470464e-06,
+      "loss": 0.8203,
+      "step": 3382
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3066294251364206,
+      "learning_rate": 9.115168417877988e-06,
+      "loss": 0.7038,
+      "step": 3383
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3089281194982156,
+      "learning_rate": 9.114579612447562e-06,
+      "loss": 0.8197,
+      "step": 3384
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3552715782822953,
+      "learning_rate": 9.113990630204485e-06,
+      "loss": 0.6735,
+      "step": 3385
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4193728649031074,
+      "learning_rate": 9.113401471174068e-06,
+      "loss": 0.6906,
+      "step": 3386
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3802364242249743,
+      "learning_rate": 9.11281213538163e-06,
+      "loss": 0.6753,
+      "step": 3387
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2405782389360562,
+      "learning_rate": 9.112222622852494e-06,
+      "loss": 0.6726,
+      "step": 3388
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.410693883828342,
+      "learning_rate": 9.111632933611993e-06,
+      "loss": 0.7126,
+      "step": 3389
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3358084163619703,
+      "learning_rate": 9.111043067685468e-06,
+      "loss": 0.6812,
+      "step": 3390
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4325939021044334,
+      "learning_rate": 9.110453025098267e-06,
+      "loss": 0.6336,
+      "step": 3391
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2818374029894086,
+      "learning_rate": 9.109862805875745e-06,
+      "loss": 0.6777,
+      "step": 3392
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2723805158248616,
+      "learning_rate": 9.109272410043265e-06,
+      "loss": 0.6162,
+      "step": 3393
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4206807215939992,
+      "learning_rate": 9.108681837626199e-06,
+      "loss": 0.6762,
+      "step": 3394
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2159710766396272,
+      "learning_rate": 9.108091088649922e-06,
+      "loss": 0.6867,
+      "step": 3395
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.406430702514209,
+      "learning_rate": 9.107500163139822e-06,
+      "loss": 0.7193,
+      "step": 3396
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.388391815532657,
+      "learning_rate": 9.106909061121291e-06,
+      "loss": 0.6626,
+      "step": 3397
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3935019016540782,
+      "learning_rate": 9.106317782619733e-06,
+      "loss": 0.6902,
+      "step": 3398
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4087879893854993,
+      "learning_rate": 9.105726327660556e-06,
+      "loss": 0.724,
+      "step": 3399
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3229337591719876,
+      "learning_rate": 9.105134696269172e-06,
+      "loss": 0.6957,
+      "step": 3400
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5419696731858448,
+      "learning_rate": 9.10454288847101e-06,
+      "loss": 0.7481,
+      "step": 3401
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.549525181099138,
+      "learning_rate": 9.103950904291496e-06,
+      "loss": 0.7125,
+      "step": 3402
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3940622795240751,
+      "learning_rate": 9.103358743756072e-06,
+      "loss": 0.6314,
+      "step": 3403
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.385597279837702,
+      "learning_rate": 9.102766406890185e-06,
+      "loss": 0.6622,
+      "step": 3404
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4259732494141122,
+      "learning_rate": 9.102173893719288e-06,
+      "loss": 0.6178,
+      "step": 3405
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4509271853419439,
+      "learning_rate": 9.101581204268843e-06,
+      "loss": 0.6844,
+      "step": 3406
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6898944306858759,
+      "learning_rate": 9.100988338564318e-06,
+      "loss": 0.7473,
+      "step": 3407
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3191936510838007,
+      "learning_rate": 9.100395296631192e-06,
+      "loss": 0.7629,
+      "step": 3408
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1722121169161281,
+      "learning_rate": 9.099802078494947e-06,
+      "loss": 0.6831,
+      "step": 3409
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0086697959040882,
+      "learning_rate": 9.099208684181078e-06,
+      "loss": 0.6458,
+      "step": 3410
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4034848039462993,
+      "learning_rate": 9.098615113715078e-06,
+      "loss": 0.7641,
+      "step": 3411
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2941264352569397,
+      "learning_rate": 9.098021367122462e-06,
+      "loss": 0.6286,
+      "step": 3412
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.413929104080721,
+      "learning_rate": 9.097427444428742e-06,
+      "loss": 0.7711,
+      "step": 3413
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.607640241721379,
+      "learning_rate": 9.096833345659437e-06,
+      "loss": 0.7039,
+      "step": 3414
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0556083512625363,
+      "learning_rate": 9.096239070840079e-06,
+      "loss": 0.7408,
+      "step": 3415
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3107916588508728,
+      "learning_rate": 9.095644619996206e-06,
+      "loss": 0.6886,
+      "step": 3416
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3935328211718179,
+      "learning_rate": 9.095049993153362e-06,
+      "loss": 0.774,
+      "step": 3417
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2787879884785371,
+      "learning_rate": 9.0944551903371e-06,
+      "loss": 0.7588,
+      "step": 3418
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2451533260077379,
+      "learning_rate": 9.09386021157298e-06,
+      "loss": 0.7115,
+      "step": 3419
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.179756437536037,
+      "learning_rate": 9.09326505688657e-06,
+      "loss": 0.6447,
+      "step": 3420
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.170116552797511,
+      "learning_rate": 9.092669726303441e-06,
+      "loss": 0.6784,
+      "step": 3421
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.260379483872233,
+      "learning_rate": 9.092074219849181e-06,
+      "loss": 0.6929,
+      "step": 3422
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1894950409558245,
+      "learning_rate": 9.09147853754938e-06,
+      "loss": 0.6639,
+      "step": 3423
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4256991245136985,
+      "learning_rate": 9.09088267942963e-06,
+      "loss": 0.7572,
+      "step": 3424
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2839308780674321,
+      "learning_rate": 9.090286645515542e-06,
+      "loss": 0.6472,
+      "step": 3425
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2607102485954864,
+      "learning_rate": 9.089690435832729e-06,
+      "loss": 0.574,
+      "step": 3426
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2017287452119647,
+      "learning_rate": 9.089094050406808e-06,
+      "loss": 0.707,
+      "step": 3427
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3844776960276692,
+      "learning_rate": 9.088497489263411e-06,
+      "loss": 0.6391,
+      "step": 3428
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3643196278033949,
+      "learning_rate": 9.087900752428168e-06,
+      "loss": 0.6778,
+      "step": 3429
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.167962235966637,
+      "learning_rate": 9.087303839926727e-06,
+      "loss": 0.6142,
+      "step": 3430
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3305808255025615,
+      "learning_rate": 9.086706751784738e-06,
+      "loss": 0.711,
+      "step": 3431
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.427042248457656,
+      "learning_rate": 9.086109488027857e-06,
+      "loss": 0.6256,
+      "step": 3432
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4222241225986298,
+      "learning_rate": 9.08551204868175e-06,
+      "loss": 0.7243,
+      "step": 3433
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5463771806153899,
+      "learning_rate": 9.084914433772094e-06,
+      "loss": 0.7791,
+      "step": 3434
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2438022817328542,
+      "learning_rate": 9.084316643324567e-06,
+      "loss": 0.6806,
+      "step": 3435
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3246013998839907,
+      "learning_rate": 9.083718677364855e-06,
+      "loss": 0.6815,
+      "step": 3436
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1903062071201151,
+      "learning_rate": 9.08312053591866e-06,
+      "loss": 0.631,
+      "step": 3437
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2877245568629385,
+      "learning_rate": 9.082522219011678e-06,
+      "loss": 0.7196,
+      "step": 3438
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5693742914358904,
+      "learning_rate": 9.081923726669626e-06,
+      "loss": 0.7377,
+      "step": 3439
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.222778403777818,
+      "learning_rate": 9.081325058918222e-06,
+      "loss": 0.6447,
+      "step": 3440
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0391298706458567,
+      "learning_rate": 9.080726215783189e-06,
+      "loss": 0.6792,
+      "step": 3441
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3462727819875475,
+      "learning_rate": 9.080127197290262e-06,
+      "loss": 0.6541,
+      "step": 3442
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2786276032408275,
+      "learning_rate": 9.079528003465183e-06,
+      "loss": 0.5574,
+      "step": 3443
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2412050132764403,
+      "learning_rate": 9.0789286343337e-06,
+      "loss": 0.6633,
+      "step": 3444
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.377016820738886,
+      "learning_rate": 9.07832908992157e-06,
+      "loss": 0.6901,
+      "step": 3445
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3428627585622306,
+      "learning_rate": 9.077729370254555e-06,
+      "loss": 0.5792,
+      "step": 3446
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3117631105471483,
+      "learning_rate": 9.077129475358426e-06,
+      "loss": 0.7566,
+      "step": 3447
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.12610176968136,
+      "learning_rate": 9.076529405258966e-06,
+      "loss": 0.6031,
+      "step": 3448
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2921130570815846,
+      "learning_rate": 9.075929159981957e-06,
+      "loss": 0.7203,
+      "step": 3449
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.246881560869362,
+      "learning_rate": 9.075328739553195e-06,
+      "loss": 0.6053,
+      "step": 3450
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4053585676509308,
+      "learning_rate": 9.07472814399848e-06,
+      "loss": 0.7857,
+      "step": 3451
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3202225151085338,
+      "learning_rate": 9.074127373343623e-06,
+      "loss": 0.6733,
+      "step": 3452
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6976299854382997,
+      "learning_rate": 9.07352642761444e-06,
+      "loss": 0.7465,
+      "step": 3453
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4571079699105318,
+      "learning_rate": 9.072925306836751e-06,
+      "loss": 0.7627,
+      "step": 3454
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3723504081264692,
+      "learning_rate": 9.072324011036392e-06,
+      "loss": 0.739,
+      "step": 3455
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.051224297750672,
+      "learning_rate": 9.071722540239203e-06,
+      "loss": 0.7274,
+      "step": 3456
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4627125665678573,
+      "learning_rate": 9.071120894471026e-06,
+      "loss": 0.7765,
+      "step": 3457
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4024309303692803,
+      "learning_rate": 9.070519073757717e-06,
+      "loss": 0.6253,
+      "step": 3458
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3572260682723283,
+      "learning_rate": 9.06991707812514e-06,
+      "loss": 0.8666,
+      "step": 3459
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1870187035871345,
+      "learning_rate": 9.069314907599159e-06,
+      "loss": 0.6861,
+      "step": 3460
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1139995214505902,
+      "learning_rate": 9.068712562205655e-06,
+      "loss": 0.6226,
+      "step": 3461
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3685625084904096,
+      "learning_rate": 9.06811004197051e-06,
+      "loss": 0.6523,
+      "step": 3462
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4006831027951545,
+      "learning_rate": 9.067507346919618e-06,
+      "loss": 0.6963,
+      "step": 3463
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0457091794894566,
+      "learning_rate": 9.066904477078875e-06,
+      "loss": 0.6436,
+      "step": 3464
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3617774315865006,
+      "learning_rate": 9.06630143247419e-06,
+      "loss": 0.7447,
+      "step": 3465
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0166260611207567,
+      "learning_rate": 9.065698213131477e-06,
+      "loss": 0.662,
+      "step": 3466
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.33001919229382,
+      "learning_rate": 9.065094819076655e-06,
+      "loss": 0.7278,
+      "step": 3467
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4725997855918243,
+      "learning_rate": 9.064491250335656e-06,
+      "loss": 0.6729,
+      "step": 3468
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2920374960128969,
+      "learning_rate": 9.063887506934417e-06,
+      "loss": 0.8204,
+      "step": 3469
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3332807099244524,
+      "learning_rate": 9.06328358889888e-06,
+      "loss": 0.5406,
+      "step": 3470
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.31545368508774,
+      "learning_rate": 9.062679496254997e-06,
+      "loss": 0.6456,
+      "step": 3471
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2492749100957197,
+      "learning_rate": 9.062075229028728e-06,
+      "loss": 0.693,
+      "step": 3472
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6688770982149177,
+      "learning_rate": 9.061470787246042e-06,
+      "loss": 0.7462,
+      "step": 3473
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.306028784815067,
+      "learning_rate": 9.06086617093291e-06,
+      "loss": 0.7441,
+      "step": 3474
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2385847643894583,
+      "learning_rate": 9.060261380115314e-06,
+      "loss": 0.6593,
+      "step": 3475
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.31735837056936,
+      "learning_rate": 9.059656414819243e-06,
+      "loss": 0.6993,
+      "step": 3476
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2070141600056048,
+      "learning_rate": 9.059051275070696e-06,
+      "loss": 0.631,
+      "step": 3477
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0491760989855312,
+      "learning_rate": 9.058445960895674e-06,
+      "loss": 0.6613,
+      "step": 3478
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1679996956045395,
+      "learning_rate": 9.057840472320192e-06,
+      "loss": 0.6177,
+      "step": 3479
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5150134557102684,
+      "learning_rate": 9.057234809370266e-06,
+      "loss": 0.6309,
+      "step": 3480
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3091600055279744,
+      "learning_rate": 9.056628972071926e-06,
+      "loss": 0.7244,
+      "step": 3481
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2699269977561993,
+      "learning_rate": 9.056022960451204e-06,
+      "loss": 0.7105,
+      "step": 3482
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3888061520731785,
+      "learning_rate": 9.055416774534142e-06,
+      "loss": 0.7611,
+      "step": 3483
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1919286006966752,
+      "learning_rate": 9.054810414346789e-06,
+      "loss": 0.6704,
+      "step": 3484
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.046421556286557,
+      "learning_rate": 9.054203879915202e-06,
+      "loss": 0.6515,
+      "step": 3485
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2376232865879697,
+      "learning_rate": 9.053597171265447e-06,
+      "loss": 0.6398,
+      "step": 3486
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1123458124697232,
+      "learning_rate": 9.052990288423591e-06,
+      "loss": 0.6867,
+      "step": 3487
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2353378653282254,
+      "learning_rate": 9.052383231415717e-06,
+      "loss": 0.7521,
+      "step": 3488
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1832761603528572,
+      "learning_rate": 9.05177600026791e-06,
+      "loss": 0.6111,
+      "step": 3489
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.450555994437402,
+      "learning_rate": 9.051168595006264e-06,
+      "loss": 0.537,
+      "step": 3490
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4456927063370693,
+      "learning_rate": 9.050561015656882e-06,
+      "loss": 0.6304,
+      "step": 3491
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5441846284340361,
+      "learning_rate": 9.049953262245872e-06,
+      "loss": 0.7052,
+      "step": 3492
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4636210755604566,
+      "learning_rate": 9.04934533479935e-06,
+      "loss": 0.8171,
+      "step": 3493
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2196698647437605,
+      "learning_rate": 9.048737233343442e-06,
+      "loss": 0.5915,
+      "step": 3494
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3697355361875403,
+      "learning_rate": 9.048128957904277e-06,
+      "loss": 0.9931,
+      "step": 3495
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.054019717451646,
+      "learning_rate": 9.047520508507994e-06,
+      "loss": 0.6127,
+      "step": 3496
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.261732888770693,
+      "learning_rate": 9.046911885180743e-06,
+      "loss": 0.5933,
+      "step": 3497
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8282355202411784,
+      "learning_rate": 9.046303087948674e-06,
+      "loss": 0.6659,
+      "step": 3498
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1657289159216475,
+      "learning_rate": 9.045694116837948e-06,
+      "loss": 0.7192,
+      "step": 3499
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3059423882923997,
+      "learning_rate": 9.045084971874738e-06,
+      "loss": 0.6056,
+      "step": 3500
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.264041014622054,
+      "learning_rate": 9.044475653085218e-06,
+      "loss": 0.6444,
+      "step": 3501
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4363541946396194,
+      "learning_rate": 9.04386616049557e-06,
+      "loss": 0.7611,
+      "step": 3502
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.173407136158236,
+      "learning_rate": 9.043256494131987e-06,
+      "loss": 0.6786,
+      "step": 3503
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2259007125287005,
+      "learning_rate": 9.042646654020667e-06,
+      "loss": 0.584,
+      "step": 3504
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2273434769371117,
+      "learning_rate": 9.04203664018782e-06,
+      "loss": 0.6466,
+      "step": 3505
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3782491146812181,
+      "learning_rate": 9.041426452659654e-06,
+      "loss": 0.7295,
+      "step": 3506
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1489501173681103,
+      "learning_rate": 9.040816091462393e-06,
+      "loss": 0.6705,
+      "step": 3507
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.244233130604873,
+      "learning_rate": 9.040205556622266e-06,
+      "loss": 0.7325,
+      "step": 3508
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4685505589226653,
+      "learning_rate": 9.039594848165507e-06,
+      "loss": 0.7459,
+      "step": 3509
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4061373901917573,
+      "learning_rate": 9.038983966118359e-06,
+      "loss": 0.6266,
+      "step": 3510
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9883273251803367,
+      "learning_rate": 9.038372910507079e-06,
+      "loss": 0.6694,
+      "step": 3511
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.293026727747141,
+      "learning_rate": 9.03776168135792e-06,
+      "loss": 0.7168,
+      "step": 3512
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1767380176433933,
+      "learning_rate": 9.037150278697147e-06,
+      "loss": 0.6958,
+      "step": 3513
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.16054519022143,
+      "learning_rate": 9.036538702551037e-06,
+      "loss": 0.7052,
+      "step": 3514
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3446934748427086,
+      "learning_rate": 9.035926952945868e-06,
+      "loss": 0.6511,
+      "step": 3515
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3762719244629429,
+      "learning_rate": 9.035315029907928e-06,
+      "loss": 0.6518,
+      "step": 3516
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2991619566070425,
+      "learning_rate": 9.034702933463516e-06,
+      "loss": 0.6646,
+      "step": 3517
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3364393058172124,
+      "learning_rate": 9.034090663638933e-06,
+      "loss": 0.828,
+      "step": 3518
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3384726739225237,
+      "learning_rate": 9.03347822046049e-06,
+      "loss": 0.7407,
+      "step": 3519
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4378313930412077,
+      "learning_rate": 9.032865603954504e-06,
+      "loss": 0.6568,
+      "step": 3520
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3916870613622263,
+      "learning_rate": 9.032252814147302e-06,
+      "loss": 0.8385,
+      "step": 3521
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.220728295578145,
+      "learning_rate": 9.031639851065217e-06,
+      "loss": 0.6616,
+      "step": 3522
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3711277381685834,
+      "learning_rate": 9.031026714734588e-06,
+      "loss": 0.7912,
+      "step": 3523
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1856302383378279,
+      "learning_rate": 9.03041340518176e-06,
+      "loss": 0.6343,
+      "step": 3524
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.393125643770941,
+      "learning_rate": 9.029799922433098e-06,
+      "loss": 0.6648,
+      "step": 3525
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5052416387206542,
+      "learning_rate": 9.029186266514956e-06,
+      "loss": 0.7597,
+      "step": 3526
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6337370702506913,
+      "learning_rate": 9.028572437453705e-06,
+      "loss": 0.8139,
+      "step": 3527
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4146771654824446,
+      "learning_rate": 9.027958435275726e-06,
+      "loss": 0.766,
+      "step": 3528
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3587024794690397,
+      "learning_rate": 9.027344260007401e-06,
+      "loss": 0.7037,
+      "step": 3529
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2214432008152682,
+      "learning_rate": 9.026729911675124e-06,
+      "loss": 0.6854,
+      "step": 3530
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3413323020291918,
+      "learning_rate": 9.026115390305295e-06,
+      "loss": 0.7066,
+      "step": 3531
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2304522239996043,
+      "learning_rate": 9.025500695924321e-06,
+      "loss": 0.6875,
+      "step": 3532
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3600687281591182,
+      "learning_rate": 9.024885828558618e-06,
+      "loss": 0.7205,
+      "step": 3533
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2966536194346387,
+      "learning_rate": 9.024270788234606e-06,
+      "loss": 0.7107,
+      "step": 3534
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.176466691970632,
+      "learning_rate": 9.023655574978716e-06,
+      "loss": 0.659,
+      "step": 3535
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1766106033146797,
+      "learning_rate": 9.023040188817385e-06,
+      "loss": 0.57,
+      "step": 3536
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2040604498346201,
+      "learning_rate": 9.022424629777056e-06,
+      "loss": 0.717,
+      "step": 3537
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1385368071228397,
+      "learning_rate": 9.021808897884183e-06,
+      "loss": 0.8272,
+      "step": 3538
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.155833857152688,
+      "learning_rate": 9.021192993165224e-06,
+      "loss": 0.6428,
+      "step": 3539
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2228313940840163,
+      "learning_rate": 9.020576915646649e-06,
+      "loss": 0.662,
+      "step": 3540
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9822977445883252,
+      "learning_rate": 9.019960665354926e-06,
+      "loss": 0.6388,
+      "step": 3541
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1553001763279063,
+      "learning_rate": 9.019344242316542e-06,
+      "loss": 0.7001,
+      "step": 3542
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2100287318205634,
+      "learning_rate": 9.018727646557985e-06,
+      "loss": 0.7439,
+      "step": 3543
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2741518241188552,
+      "learning_rate": 9.01811087810575e-06,
+      "loss": 0.7265,
+      "step": 3544
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1585114976693653,
+      "learning_rate": 9.01749393698634e-06,
+      "loss": 0.6446,
+      "step": 3545
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1273924333951613,
+      "learning_rate": 9.01687682322627e-06,
+      "loss": 0.7204,
+      "step": 3546
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.522959581412137,
+      "learning_rate": 9.016259536852059e-06,
+      "loss": 0.759,
+      "step": 3547
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4569503036511866,
+      "learning_rate": 9.015642077890228e-06,
+      "loss": 0.7702,
+      "step": 3548
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3249926719299077,
+      "learning_rate": 9.015024446367315e-06,
+      "loss": 0.6872,
+      "step": 3549
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2710450049259316,
+      "learning_rate": 9.014406642309859e-06,
+      "loss": 0.752,
+      "step": 3550
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.538450394048669,
+      "learning_rate": 9.013788665744409e-06,
+      "loss": 0.6971,
+      "step": 3551
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4517814508303715,
+      "learning_rate": 9.01317051669752e-06,
+      "loss": 0.6479,
+      "step": 3552
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2079377449808042,
+      "learning_rate": 9.012552195195759e-06,
+      "loss": 0.6708,
+      "step": 3553
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1898271509929799,
+      "learning_rate": 9.01193370126569e-06,
+      "loss": 0.5979,
+      "step": 3554
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.234904251362334,
+      "learning_rate": 9.011315034933896e-06,
+      "loss": 0.6115,
+      "step": 3555
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3550643985513373,
+      "learning_rate": 9.010696196226963e-06,
+      "loss": 0.7727,
+      "step": 3556
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.397739463292748,
+      "learning_rate": 9.010077185171481e-06,
+      "loss": 0.6328,
+      "step": 3557
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3713340508963132,
+      "learning_rate": 9.009458001794052e-06,
+      "loss": 0.7656,
+      "step": 3558
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3720649154014755,
+      "learning_rate": 9.008838646121282e-06,
+      "loss": 0.7109,
+      "step": 3559
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6722945812196126,
+      "learning_rate": 9.008219118179787e-06,
+      "loss": 0.6444,
+      "step": 3560
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3012495190980613,
+      "learning_rate": 9.007599417996191e-06,
+      "loss": 0.7151,
+      "step": 3561
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.265389964954274,
+      "learning_rate": 9.006979545597123e-06,
+      "loss": 0.604,
+      "step": 3562
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3143775064591368,
+      "learning_rate": 9.00635950100922e-06,
+      "loss": 0.6894,
+      "step": 3563
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.49989189039426,
+      "learning_rate": 9.005739284259123e-06,
+      "loss": 0.8084,
+      "step": 3564
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.478968634759133,
+      "learning_rate": 9.00511889537349e-06,
+      "loss": 0.7292,
+      "step": 3565
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1101966779425843,
+      "learning_rate": 9.00449833437898e-06,
+      "loss": 0.5803,
+      "step": 3566
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1150110830688196,
+      "learning_rate": 9.003877601302256e-06,
+      "loss": 0.6854,
+      "step": 3567
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.408573097450765,
+      "learning_rate": 9.003256696169995e-06,
+      "loss": 0.7591,
+      "step": 3568
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.197326457520928,
+      "learning_rate": 9.002635619008877e-06,
+      "loss": 0.6433,
+      "step": 3569
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.216272955961013,
+      "learning_rate": 9.002014369845592e-06,
+      "loss": 0.7045,
+      "step": 3570
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.273446038034203,
+      "learning_rate": 9.00139294870684e-06,
+      "loss": 0.6625,
+      "step": 3571
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3049580171250723,
+      "learning_rate": 9.000771355619318e-06,
+      "loss": 0.603,
+      "step": 3572
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3379152519075022,
+      "learning_rate": 9.000149590609741e-06,
+      "loss": 0.6399,
+      "step": 3573
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4404059973393764,
+      "learning_rate": 8.999527653704829e-06,
+      "loss": 0.7327,
+      "step": 3574
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3263590908650316,
+      "learning_rate": 8.998905544931304e-06,
+      "loss": 0.6669,
+      "step": 3575
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4755495570336197,
+      "learning_rate": 8.998283264315903e-06,
+      "loss": 0.6994,
+      "step": 3576
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2999241030644832,
+      "learning_rate": 8.997660811885367e-06,
+      "loss": 0.6687,
+      "step": 3577
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.237709156046751,
+      "learning_rate": 8.99703818766644e-06,
+      "loss": 0.6547,
+      "step": 3578
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0967906257087718,
+      "learning_rate": 8.996415391685882e-06,
+      "loss": 0.6062,
+      "step": 3579
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0599256117669205,
+      "learning_rate": 8.995792423970454e-06,
+      "loss": 0.6965,
+      "step": 3580
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0759777858180342,
+      "learning_rate": 8.995169284546926e-06,
+      "loss": 0.6854,
+      "step": 3581
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1695839720912096,
+      "learning_rate": 8.994545973442076e-06,
+      "loss": 0.626,
+      "step": 3582
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3850431655085789,
+      "learning_rate": 8.993922490682687e-06,
+      "loss": 0.6855,
+      "step": 3583
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.265512870758275,
+      "learning_rate": 8.993298836295556e-06,
+      "loss": 0.6949,
+      "step": 3584
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4657655682253912,
+      "learning_rate": 8.99267501030748e-06,
+      "loss": 0.7697,
+      "step": 3585
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3751988167582954,
+      "learning_rate": 8.992051012745268e-06,
+      "loss": 0.7178,
+      "step": 3586
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.345017100162311,
+      "learning_rate": 8.991426843635731e-06,
+      "loss": 0.6151,
+      "step": 3587
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4540550731913942,
+      "learning_rate": 8.990802503005695e-06,
+      "loss": 0.6489,
+      "step": 3588
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.333975878333601,
+      "learning_rate": 8.990177990881986e-06,
+      "loss": 0.6965,
+      "step": 3589
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5020664413000313,
+      "learning_rate": 8.989553307291443e-06,
+      "loss": 0.9094,
+      "step": 3590
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3689665581164983,
+      "learning_rate": 8.988928452260909e-06,
+      "loss": 0.7031,
+      "step": 3591
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.240780973592584,
+      "learning_rate": 8.988303425817235e-06,
+      "loss": 0.7159,
+      "step": 3592
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3139221781443808,
+      "learning_rate": 8.987678227987282e-06,
+      "loss": 0.7059,
+      "step": 3593
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.114114967790465,
+      "learning_rate": 8.987052858797914e-06,
+      "loss": 0.6311,
+      "step": 3594
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2139378022700034,
+      "learning_rate": 8.986427318276003e-06,
+      "loss": 0.6123,
+      "step": 3595
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3629090898933338,
+      "learning_rate": 8.985801606448435e-06,
+      "loss": 0.7208,
+      "step": 3596
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3751962499265202,
+      "learning_rate": 8.985175723342093e-06,
+      "loss": 0.7189,
+      "step": 3597
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2260822840894539,
+      "learning_rate": 8.984549668983875e-06,
+      "loss": 0.6718,
+      "step": 3598
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2790373155103878,
+      "learning_rate": 8.983923443400682e-06,
+      "loss": 0.7047,
+      "step": 3599
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4016954668708184,
+      "learning_rate": 8.983297046619427e-06,
+      "loss": 0.5802,
+      "step": 3600
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.453882049381042,
+      "learning_rate": 8.982670478667029e-06,
+      "loss": 0.6804,
+      "step": 3601
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4478891687572708,
+      "learning_rate": 8.98204373957041e-06,
+      "loss": 0.6667,
+      "step": 3602
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3568374577462752,
+      "learning_rate": 8.981416829356502e-06,
+      "loss": 0.7302,
+      "step": 3603
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1828542373403879,
+      "learning_rate": 8.980789748052245e-06,
+      "loss": 0.6217,
+      "step": 3604
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3974755269280426,
+      "learning_rate": 8.980162495684587e-06,
+      "loss": 0.7374,
+      "step": 3605
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3497766487018155,
+      "learning_rate": 8.979535072280483e-06,
+      "loss": 0.7292,
+      "step": 3606
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2107623143329922,
+      "learning_rate": 8.978907477866894e-06,
+      "loss": 0.6672,
+      "step": 3607
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2944105701291904,
+      "learning_rate": 8.978279712470789e-06,
+      "loss": 0.6422,
+      "step": 3608
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.204482355482109,
+      "learning_rate": 8.977651776119145e-06,
+      "loss": 0.6158,
+      "step": 3609
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2788781735528412,
+      "learning_rate": 8.977023668838945e-06,
+      "loss": 0.6989,
+      "step": 3610
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0906309926727098,
+      "learning_rate": 8.976395390657181e-06,
+      "loss": 0.6393,
+      "step": 3611
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3805743401597306,
+      "learning_rate": 8.975766941600852e-06,
+      "loss": 0.7012,
+      "step": 3612
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.194935612958415,
+      "learning_rate": 8.975138321696961e-06,
+      "loss": 0.6008,
+      "step": 3613
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1750686439833933,
+      "learning_rate": 8.974509530972523e-06,
+      "loss": 0.6142,
+      "step": 3614
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4350304779869854,
+      "learning_rate": 8.973880569454561e-06,
+      "loss": 0.7301,
+      "step": 3615
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.320279639706961,
+      "learning_rate": 8.9732514371701e-06,
+      "loss": 0.6919,
+      "step": 3616
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.398328624101428,
+      "learning_rate": 8.972622134146174e-06,
+      "loss": 0.7113,
+      "step": 3617
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.256888523774362,
+      "learning_rate": 8.97199266040983e-06,
+      "loss": 0.6999,
+      "step": 3618
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.000571028542316,
+      "learning_rate": 8.971363015988115e-06,
+      "loss": 0.6247,
+      "step": 3619
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3614075885233905,
+      "learning_rate": 8.970733200908084e-06,
+      "loss": 0.8073,
+      "step": 3620
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3149647759136893,
+      "learning_rate": 8.970103215196807e-06,
+      "loss": 0.8064,
+      "step": 3621
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3559466623290828,
+      "learning_rate": 8.96947305888135e-06,
+      "loss": 0.6552,
+      "step": 3622
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.502180963174721,
+      "learning_rate": 8.968842731988798e-06,
+      "loss": 0.7208,
+      "step": 3623
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4177712738778254,
+      "learning_rate": 8.968212234546235e-06,
+      "loss": 0.7695,
+      "step": 3624
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3572170356111188,
+      "learning_rate": 8.967581566580752e-06,
+      "loss": 0.6533,
+      "step": 3625
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.100331933034533,
+      "learning_rate": 8.966950728119453e-06,
+      "loss": 0.6491,
+      "step": 3626
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2496564906691983,
+      "learning_rate": 8.966319719189447e-06,
+      "loss": 0.6146,
+      "step": 3627
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5578066077030077,
+      "learning_rate": 8.96568853981785e-06,
+      "loss": 0.7548,
+      "step": 3628
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2447091694369299,
+      "learning_rate": 8.965057190031785e-06,
+      "loss": 0.7161,
+      "step": 3629
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4922229803867209,
+      "learning_rate": 8.96442566985838e-06,
+      "loss": 0.7059,
+      "step": 3630
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2058886807284164,
+      "learning_rate": 8.963793979324775e-06,
+      "loss": 0.6371,
+      "step": 3631
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3227679221752129,
+      "learning_rate": 8.963162118458115e-06,
+      "loss": 0.6749,
+      "step": 3632
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1932785989543608,
+      "learning_rate": 8.962530087285552e-06,
+      "loss": 0.5784,
+      "step": 3633
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.731071963859549,
+      "learning_rate": 8.961897885834247e-06,
+      "loss": 0.7026,
+      "step": 3634
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3647446849316087,
+      "learning_rate": 8.961265514131368e-06,
+      "loss": 0.823,
+      "step": 3635
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4318014243785473,
+      "learning_rate": 8.960632972204084e-06,
+      "loss": 0.7656,
+      "step": 3636
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3596101876840538,
+      "learning_rate": 8.960000260079582e-06,
+      "loss": 0.6993,
+      "step": 3637
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2730499400326136,
+      "learning_rate": 8.95936737778505e-06,
+      "loss": 0.7904,
+      "step": 3638
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2662082835626087,
+      "learning_rate": 8.958734325347684e-06,
+      "loss": 0.6456,
+      "step": 3639
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.331358956254048,
+      "learning_rate": 8.958101102794686e-06,
+      "loss": 0.6518,
+      "step": 3640
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4077511991111233,
+      "learning_rate": 8.95746771015327e-06,
+      "loss": 0.7101,
+      "step": 3641
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2202095825170094,
+      "learning_rate": 8.95683414745065e-06,
+      "loss": 0.5839,
+      "step": 3642
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2988782154728833,
+      "learning_rate": 8.956200414714057e-06,
+      "loss": 0.6383,
+      "step": 3643
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.088016240224964,
+      "learning_rate": 8.955566511970721e-06,
+      "loss": 0.6921,
+      "step": 3644
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4472040769534413,
+      "learning_rate": 8.954932439247882e-06,
+      "loss": 0.7138,
+      "step": 3645
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3866399897439747,
+      "learning_rate": 8.954298196572789e-06,
+      "loss": 0.7015,
+      "step": 3646
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3708997677818044,
+      "learning_rate": 8.953663783972692e-06,
+      "loss": 0.7351,
+      "step": 3647
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2954617759855245,
+      "learning_rate": 8.95302920147486e-06,
+      "loss": 0.7527,
+      "step": 3648
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3816158421091507,
+      "learning_rate": 8.95239444910656e-06,
+      "loss": 0.728,
+      "step": 3649
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3783283882513595,
+      "learning_rate": 8.951759526895067e-06,
+      "loss": 0.6466,
+      "step": 3650
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.357545689482643,
+      "learning_rate": 8.951124434867666e-06,
+      "loss": 0.7144,
+      "step": 3651
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.542551566089454,
+      "learning_rate": 8.950489173051648e-06,
+      "loss": 0.5842,
+      "step": 3652
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2963036426716965,
+      "learning_rate": 8.949853741474312e-06,
+      "loss": 0.7582,
+      "step": 3653
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.254180157604432,
+      "learning_rate": 8.949218140162965e-06,
+      "loss": 0.618,
+      "step": 3654
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.137271973348307,
+      "learning_rate": 8.948582369144919e-06,
+      "loss": 0.6888,
+      "step": 3655
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4727127783024587,
+      "learning_rate": 8.947946428447495e-06,
+      "loss": 0.7268,
+      "step": 3656
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.266760938139492,
+      "learning_rate": 8.94731031809802e-06,
+      "loss": 0.6888,
+      "step": 3657
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4836001205813758,
+      "learning_rate": 8.946674038123829e-06,
+      "loss": 0.7234,
+      "step": 3658
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2482433223524934,
+      "learning_rate": 8.946037588552266e-06,
+      "loss": 0.6915,
+      "step": 3659
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1000984810752894,
+      "learning_rate": 8.945400969410678e-06,
+      "loss": 0.5851,
+      "step": 3660
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4117550007529354,
+      "learning_rate": 8.944764180726423e-06,
+      "loss": 0.6823,
+      "step": 3661
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4921375522263804,
+      "learning_rate": 8.944127222526867e-06,
+      "loss": 0.7562,
+      "step": 3662
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.384864948894293,
+      "learning_rate": 8.94349009483938e-06,
+      "loss": 0.7644,
+      "step": 3663
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4126116428985924,
+      "learning_rate": 8.94285279769134e-06,
+      "loss": 0.7349,
+      "step": 3664
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4536877389491656,
+      "learning_rate": 8.942215331110136e-06,
+      "loss": 0.6628,
+      "step": 3665
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3499916836099262,
+      "learning_rate": 8.941577695123158e-06,
+      "loss": 0.6985,
+      "step": 3666
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.45918733023348,
+      "learning_rate": 8.940939889757809e-06,
+      "loss": 0.7817,
+      "step": 3667
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4799867653071794,
+      "learning_rate": 8.940301915041496e-06,
+      "loss": 0.7491,
+      "step": 3668
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4657717879019003,
+      "learning_rate": 8.939663771001632e-06,
+      "loss": 0.9546,
+      "step": 3669
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4045248936535366,
+      "learning_rate": 8.939025457665644e-06,
+      "loss": 0.6535,
+      "step": 3670
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3451367417739446,
+      "learning_rate": 8.93838697506096e-06,
+      "loss": 0.6252,
+      "step": 3671
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2628044236725389,
+      "learning_rate": 8.937748323215013e-06,
+      "loss": 0.6918,
+      "step": 3672
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2754401797339125,
+      "learning_rate": 8.937109502155254e-06,
+      "loss": 0.7158,
+      "step": 3673
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2694945789487286,
+      "learning_rate": 8.93647051190913e-06,
+      "loss": 0.6788,
+      "step": 3674
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2296001734752762,
+      "learning_rate": 8.935831352504103e-06,
+      "loss": 0.6748,
+      "step": 3675
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7889082219023755,
+      "learning_rate": 8.935192023967634e-06,
+      "loss": 0.8783,
+      "step": 3676
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.356263851619769,
+      "learning_rate": 8.934552526327201e-06,
+      "loss": 0.7241,
+      "step": 3677
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2479376935422941,
+      "learning_rate": 8.933912859610284e-06,
+      "loss": 0.7361,
+      "step": 3678
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.443893628252375,
+      "learning_rate": 8.93327302384437e-06,
+      "loss": 0.7897,
+      "step": 3679
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.359507941390039,
+      "learning_rate": 8.932633019056954e-06,
+      "loss": 0.6482,
+      "step": 3680
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2637178419580728,
+      "learning_rate": 8.931992845275539e-06,
+      "loss": 0.7879,
+      "step": 3681
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3198485711050447,
+      "learning_rate": 8.931352502527633e-06,
+      "loss": 0.7539,
+      "step": 3682
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.330935254610639,
+      "learning_rate": 8.930711990840755e-06,
+      "loss": 0.6986,
+      "step": 3683
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4246798285296502,
+      "learning_rate": 8.930071310242429e-06,
+      "loss": 0.7114,
+      "step": 3684
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.383146331833168,
+      "learning_rate": 8.929430460760187e-06,
+      "loss": 0.6737,
+      "step": 3685
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.261061015164984,
+      "learning_rate": 8.928789442421564e-06,
+      "loss": 0.6703,
+      "step": 3686
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3664305774954921,
+      "learning_rate": 8.928148255254112e-06,
+      "loss": 0.6993,
+      "step": 3687
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2990085119206416,
+      "learning_rate": 8.927506899285378e-06,
+      "loss": 0.6374,
+      "step": 3688
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.264171518885513,
+      "learning_rate": 8.926865374542928e-06,
+      "loss": 0.7727,
+      "step": 3689
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.204508628495879,
+      "learning_rate": 8.926223681054326e-06,
+      "loss": 0.6969,
+      "step": 3690
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4772323447189881,
+      "learning_rate": 8.92558181884715e-06,
+      "loss": 0.6795,
+      "step": 3691
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7498494256849917,
+      "learning_rate": 8.924939787948977e-06,
+      "loss": 0.824,
+      "step": 3692
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2600190291886686,
+      "learning_rate": 8.924297588387404e-06,
+      "loss": 0.6201,
+      "step": 3693
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0424306803243115,
+      "learning_rate": 8.92365522019002e-06,
+      "loss": 0.5915,
+      "step": 3694
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6700626780122299,
+      "learning_rate": 8.923012683384436e-06,
+      "loss": 0.7581,
+      "step": 3695
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5083359561378031,
+      "learning_rate": 8.922369977998257e-06,
+      "loss": 0.6437,
+      "step": 3696
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3177122106615833,
+      "learning_rate": 8.921727104059107e-06,
+      "loss": 0.6664,
+      "step": 3697
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6037904959192515,
+      "learning_rate": 8.921084061594608e-06,
+      "loss": 0.7626,
+      "step": 3698
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3800908984605427,
+      "learning_rate": 8.920440850632395e-06,
+      "loss": 0.6982,
+      "step": 3699
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2070619851876272,
+      "learning_rate": 8.919797471200106e-06,
+      "loss": 0.6288,
+      "step": 3700
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2968398192900668,
+      "learning_rate": 8.91915392332539e-06,
+      "loss": 0.7195,
+      "step": 3701
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3612051793283986,
+      "learning_rate": 8.918510207035901e-06,
+      "loss": 0.6877,
+      "step": 3702
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.190640456250583,
+      "learning_rate": 8.917866322359303e-06,
+      "loss": 0.5861,
+      "step": 3703
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0998143248140355,
+      "learning_rate": 8.917222269323263e-06,
+      "loss": 0.6442,
+      "step": 3704
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4499456132938584,
+      "learning_rate": 8.916578047955458e-06,
+      "loss": 0.8494,
+      "step": 3705
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5578197410509769,
+      "learning_rate": 8.915933658283572e-06,
+      "loss": 0.749,
+      "step": 3706
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2414222924484062,
+      "learning_rate": 8.915289100335295e-06,
+      "loss": 0.6283,
+      "step": 3707
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2947040185103496,
+      "learning_rate": 8.914644374138325e-06,
+      "loss": 0.7183,
+      "step": 3708
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0849966339937842,
+      "learning_rate": 8.91399947972037e-06,
+      "loss": 0.7246,
+      "step": 3709
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2962918952690878,
+      "learning_rate": 8.913354417109136e-06,
+      "loss": 0.7495,
+      "step": 3710
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4268757769865856,
+      "learning_rate": 8.91270918633235e-06,
+      "loss": 0.7814,
+      "step": 3711
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3371299551122193,
+      "learning_rate": 8.912063787417738e-06,
+      "loss": 0.7077,
+      "step": 3712
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3962618842642212,
+      "learning_rate": 8.911418220393031e-06,
+      "loss": 0.6656,
+      "step": 3713
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.481115702138401,
+      "learning_rate": 8.91077248528597e-06,
+      "loss": 0.6553,
+      "step": 3714
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8737254682766744,
+      "learning_rate": 8.910126582124307e-06,
+      "loss": 0.805,
+      "step": 3715
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2841072945938576,
+      "learning_rate": 8.909480510935797e-06,
+      "loss": 0.6355,
+      "step": 3716
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.07976719871456,
+      "learning_rate": 8.908834271748202e-06,
+      "loss": 0.6068,
+      "step": 3717
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3547458578334772,
+      "learning_rate": 8.908187864589292e-06,
+      "loss": 0.6797,
+      "step": 3718
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5904914753805524,
+      "learning_rate": 8.907541289486847e-06,
+      "loss": 0.692,
+      "step": 3719
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6521557504537878,
+      "learning_rate": 8.90689454646865e-06,
+      "loss": 0.7931,
+      "step": 3720
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5802703302657661,
+      "learning_rate": 8.906247635562493e-06,
+      "loss": 0.7818,
+      "step": 3721
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.426055883298087,
+      "learning_rate": 8.905600556796176e-06,
+      "loss": 0.7391,
+      "step": 3722
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4171613400879333,
+      "learning_rate": 8.904953310197504e-06,
+      "loss": 0.7706,
+      "step": 3723
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4518462600215347,
+      "learning_rate": 8.904305895794292e-06,
+      "loss": 0.6538,
+      "step": 3724
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3652968616672256,
+      "learning_rate": 8.903658313614358e-06,
+      "loss": 0.6675,
+      "step": 3725
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3464193635934205,
+      "learning_rate": 8.903010563685534e-06,
+      "loss": 0.6505,
+      "step": 3726
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.768469584175956,
+      "learning_rate": 8.902362646035655e-06,
+      "loss": 0.6466,
+      "step": 3727
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5022140062292457,
+      "learning_rate": 8.90171456069256e-06,
+      "loss": 0.7599,
+      "step": 3728
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2828352595967392,
+      "learning_rate": 8.901066307684102e-06,
+      "loss": 0.548,
+      "step": 3729
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.325796691754989,
+      "learning_rate": 8.900417887038136e-06,
+      "loss": 0.7875,
+      "step": 3730
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4880666513400185,
+      "learning_rate": 8.899769298782528e-06,
+      "loss": 0.6959,
+      "step": 3731
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2923102804283555,
+      "learning_rate": 8.899120542945147e-06,
+      "loss": 0.63,
+      "step": 3732
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3939978445594523,
+      "learning_rate": 8.898471619553872e-06,
+      "loss": 0.7574,
+      "step": 3733
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2770628551449132,
+      "learning_rate": 8.89782252863659e-06,
+      "loss": 0.7561,
+      "step": 3734
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3745854115661775,
+      "learning_rate": 8.89717327022119e-06,
+      "loss": 0.6988,
+      "step": 3735
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.396854775016122,
+      "learning_rate": 8.89652384433558e-06,
+      "loss": 0.7554,
+      "step": 3736
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.071043095599192,
+      "learning_rate": 8.89587425100766e-06,
+      "loss": 0.7071,
+      "step": 3737
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3366870528352055,
+      "learning_rate": 8.895224490265346e-06,
+      "loss": 0.7062,
+      "step": 3738
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3506237985165561,
+      "learning_rate": 8.894574562136561e-06,
+      "loss": 0.668,
+      "step": 3739
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.392194428822815,
+      "learning_rate": 8.893924466649235e-06,
+      "loss": 0.6808,
+      "step": 3740
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6524002443966341,
+      "learning_rate": 8.893274203831302e-06,
+      "loss": 0.6969,
+      "step": 3741
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.396365428608174,
+      "learning_rate": 8.892623773710704e-06,
+      "loss": 0.7201,
+      "step": 3742
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4241185672901293,
+      "learning_rate": 8.891973176315393e-06,
+      "loss": 0.6893,
+      "step": 3743
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1617631544510174,
+      "learning_rate": 8.89132241167333e-06,
+      "loss": 0.6325,
+      "step": 3744
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3597449849202556,
+      "learning_rate": 8.890671479812472e-06,
+      "loss": 0.7912,
+      "step": 3745
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3525726176642856,
+      "learning_rate": 8.890020380760798e-06,
+      "loss": 0.6279,
+      "step": 3746
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2480764642965534,
+      "learning_rate": 8.889369114546285e-06,
+      "loss": 0.5595,
+      "step": 3747
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3795602551613901,
+      "learning_rate": 8.888717681196918e-06,
+      "loss": 0.7001,
+      "step": 3748
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.346448976529747,
+      "learning_rate": 8.888066080740692e-06,
+      "loss": 0.7027,
+      "step": 3749
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3211007304127238,
+      "learning_rate": 8.887414313205608e-06,
+      "loss": 0.6847,
+      "step": 3750
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3656052328590098,
+      "learning_rate": 8.886762378619674e-06,
+      "loss": 0.7045,
+      "step": 3751
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6338624163710298,
+      "learning_rate": 8.886110277010902e-06,
+      "loss": 0.6915,
+      "step": 3752
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4157833553867987,
+      "learning_rate": 8.885458008407317e-06,
+      "loss": 0.6565,
+      "step": 3753
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4596672645017237,
+      "learning_rate": 8.88480557283695e-06,
+      "loss": 0.7766,
+      "step": 3754
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5684554350256095,
+      "learning_rate": 8.884152970327833e-06,
+      "loss": 0.7894,
+      "step": 3755
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.329893829510566,
+      "learning_rate": 8.883500200908013e-06,
+      "loss": 0.7858,
+      "step": 3756
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3857635821724743,
+      "learning_rate": 8.882847264605542e-06,
+      "loss": 0.7039,
+      "step": 3757
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3162597338286501,
+      "learning_rate": 8.882194161448476e-06,
+      "loss": 0.6909,
+      "step": 3758
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4281138331434071,
+      "learning_rate": 8.88154089146488e-06,
+      "loss": 0.6762,
+      "step": 3759
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5104956100179845,
+      "learning_rate": 8.880887454682827e-06,
+      "loss": 0.6693,
+      "step": 3760
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3211161539277319,
+      "learning_rate": 8.880233851130397e-06,
+      "loss": 0.658,
+      "step": 3761
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3822555641389969,
+      "learning_rate": 8.879580080835677e-06,
+      "loss": 0.7231,
+      "step": 3762
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3848791789217407,
+      "learning_rate": 8.87892614382676e-06,
+      "loss": 0.7124,
+      "step": 3763
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.457505796213476,
+      "learning_rate": 8.878272040131748e-06,
+      "loss": 0.6425,
+      "step": 3764
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4486321351365805,
+      "learning_rate": 8.877617769778751e-06,
+      "loss": 0.695,
+      "step": 3765
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0897661713005469,
+      "learning_rate": 8.876963332795881e-06,
+      "loss": 0.7532,
+      "step": 3766
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2423931848111893,
+      "learning_rate": 8.876308729211261e-06,
+      "loss": 0.6447,
+      "step": 3767
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1477171767190946,
+      "learning_rate": 8.875653959053023e-06,
+      "loss": 0.7161,
+      "step": 3768
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2020105719005614,
+      "learning_rate": 8.874999022349303e-06,
+      "loss": 0.5936,
+      "step": 3769
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5125038354740257,
+      "learning_rate": 8.874343919128245e-06,
+      "loss": 0.8249,
+      "step": 3770
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4032989721592577,
+      "learning_rate": 8.873688649417999e-06,
+      "loss": 0.6882,
+      "step": 3771
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3465492139883353,
+      "learning_rate": 8.873033213246726e-06,
+      "loss": 0.6257,
+      "step": 3772
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3419107832946866,
+      "learning_rate": 8.87237761064259e-06,
+      "loss": 0.8376,
+      "step": 3773
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.124767264460416,
+      "learning_rate": 8.871721841633762e-06,
+      "loss": 0.5949,
+      "step": 3774
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4604824937513279,
+      "learning_rate": 8.871065906248426e-06,
+      "loss": 0.7471,
+      "step": 3775
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1987381956257412,
+      "learning_rate": 8.870409804514765e-06,
+      "loss": 0.5505,
+      "step": 3776
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3201438142504685,
+      "learning_rate": 8.869753536460976e-06,
+      "loss": 0.7178,
+      "step": 3777
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3660409153371245,
+      "learning_rate": 8.86909710211526e-06,
+      "loss": 0.6961,
+      "step": 3778
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.217046165369111,
+      "learning_rate": 8.868440501505822e-06,
+      "loss": 0.7068,
+      "step": 3779
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.433207447365529,
+      "learning_rate": 8.867783734660883e-06,
+      "loss": 0.6969,
+      "step": 3780
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5777673586655694,
+      "learning_rate": 8.86712680160866e-06,
+      "loss": 0.7988,
+      "step": 3781
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.494279558984455,
+      "learning_rate": 8.866469702377386e-06,
+      "loss": 0.7006,
+      "step": 3782
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5087248955909183,
+      "learning_rate": 8.8658124369953e-06,
+      "loss": 0.6381,
+      "step": 3783
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4676144818817296,
+      "learning_rate": 8.865155005490643e-06,
+      "loss": 0.6983,
+      "step": 3784
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4874408607882035,
+      "learning_rate": 8.864497407891668e-06,
+      "loss": 0.7107,
+      "step": 3785
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1962846404480318,
+      "learning_rate": 8.86383964422663e-06,
+      "loss": 0.6287,
+      "step": 3786
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4782145179833965,
+      "learning_rate": 8.8631817145238e-06,
+      "loss": 0.7933,
+      "step": 3787
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1640911837475891,
+      "learning_rate": 8.862523618811448e-06,
+      "loss": 0.6588,
+      "step": 3788
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4299066199433452,
+      "learning_rate": 8.861865357117852e-06,
+      "loss": 0.66,
+      "step": 3789
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.498768260919709,
+      "learning_rate": 8.861206929471302e-06,
+      "loss": 0.7654,
+      "step": 3790
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2812645491207844,
+      "learning_rate": 8.86054833590009e-06,
+      "loss": 0.6806,
+      "step": 3791
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4075391924916625,
+      "learning_rate": 8.859889576432519e-06,
+      "loss": 0.8417,
+      "step": 3792
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2474589043335658,
+      "learning_rate": 8.859230651096896e-06,
+      "loss": 0.6394,
+      "step": 3793
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.301755200159064,
+      "learning_rate": 8.858571559921539e-06,
+      "loss": 0.7254,
+      "step": 3794
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.29516283630033,
+      "learning_rate": 8.857912302934766e-06,
+      "loss": 0.7548,
+      "step": 3795
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4312394938135846,
+      "learning_rate": 8.857252880164908e-06,
+      "loss": 0.687,
+      "step": 3796
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.232861204299373,
+      "learning_rate": 8.856593291640305e-06,
+      "loss": 0.7182,
+      "step": 3797
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1178675598366734,
+      "learning_rate": 8.8559335373893e-06,
+      "loss": 0.6814,
+      "step": 3798
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2576358739966487,
+      "learning_rate": 8.855273617440243e-06,
+      "loss": 0.7047,
+      "step": 3799
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.362045307414678,
+      "learning_rate": 8.854613531821491e-06,
+      "loss": 0.6921,
+      "step": 3800
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3702047760077074,
+      "learning_rate": 8.853953280561412e-06,
+      "loss": 0.6549,
+      "step": 3801
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4464085659880457,
+      "learning_rate": 8.853292863688378e-06,
+      "loss": 0.7636,
+      "step": 3802
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7268644208274606,
+      "learning_rate": 8.852632281230766e-06,
+      "loss": 0.7434,
+      "step": 3803
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4702755376198555,
+      "learning_rate": 8.851971533216968e-06,
+      "loss": 0.8713,
+      "step": 3804
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9769739204584389,
+      "learning_rate": 8.85131061967537e-06,
+      "loss": 0.6157,
+      "step": 3805
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.239082757136356,
+      "learning_rate": 8.85064954063438e-06,
+      "loss": 0.6698,
+      "step": 3806
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.458637234235418,
+      "learning_rate": 8.849988296122402e-06,
+      "loss": 0.6871,
+      "step": 3807
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3668632258651807,
+      "learning_rate": 8.849326886167854e-06,
+      "loss": 0.7791,
+      "step": 3808
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.242377250713748,
+      "learning_rate": 8.848665310799156e-06,
+      "loss": 0.7025,
+      "step": 3809
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3949644174095248,
+      "learning_rate": 8.84800357004474e-06,
+      "loss": 0.7276,
+      "step": 3810
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3854287988955096,
+      "learning_rate": 8.847341663933038e-06,
+      "loss": 0.7598,
+      "step": 3811
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3400358710114242,
+      "learning_rate": 8.846679592492498e-06,
+      "loss": 0.6537,
+      "step": 3812
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4959247110003973,
+      "learning_rate": 8.846017355751567e-06,
+      "loss": 0.7398,
+      "step": 3813
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5194533482131307,
+      "learning_rate": 8.845354953738706e-06,
+      "loss": 0.807,
+      "step": 3814
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6172728339263704,
+      "learning_rate": 8.844692386482379e-06,
+      "loss": 0.727,
+      "step": 3815
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.577120432461776,
+      "learning_rate": 8.844029654011055e-06,
+      "loss": 0.8155,
+      "step": 3816
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1186999939758437,
+      "learning_rate": 8.84336675635322e-06,
+      "loss": 0.6958,
+      "step": 3817
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.071005873252272,
+      "learning_rate": 8.842703693537352e-06,
+      "loss": 0.6305,
+      "step": 3818
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6031206703786243,
+      "learning_rate": 8.84204046559195e-06,
+      "loss": 0.7334,
+      "step": 3819
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4129394811850016,
+      "learning_rate": 8.841377072545513e-06,
+      "loss": 0.6772,
+      "step": 3820
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.199200841998088,
+      "learning_rate": 8.840713514426545e-06,
+      "loss": 0.6801,
+      "step": 3821
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2312917298631556,
+      "learning_rate": 8.840049791263567e-06,
+      "loss": 0.6854,
+      "step": 3822
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3804340455785566,
+      "learning_rate": 8.839385903085096e-06,
+      "loss": 0.611,
+      "step": 3823
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3395194187320807,
+      "learning_rate": 8.83872184991966e-06,
+      "loss": 0.6852,
+      "step": 3824
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4187690363830574,
+      "learning_rate": 8.8380576317958e-06,
+      "loss": 0.5995,
+      "step": 3825
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2405742959405885,
+      "learning_rate": 8.837393248742054e-06,
+      "loss": 0.6972,
+      "step": 3826
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.069766229571946,
+      "learning_rate": 8.836728700786974e-06,
+      "loss": 0.752,
+      "step": 3827
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3956178011803138,
+      "learning_rate": 8.836063987959116e-06,
+      "loss": 0.7989,
+      "step": 3828
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.008209252193002,
+      "learning_rate": 8.835399110287046e-06,
+      "loss": 0.6991,
+      "step": 3829
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3309087060495564,
+      "learning_rate": 8.834734067799336e-06,
+      "loss": 0.7787,
+      "step": 3830
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4542338880459502,
+      "learning_rate": 8.83406886052456e-06,
+      "loss": 0.7799,
+      "step": 3831
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.133197095894049,
+      "learning_rate": 8.833403488491308e-06,
+      "loss": 0.7247,
+      "step": 3832
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2172915977385508,
+      "learning_rate": 8.83273795172817e-06,
+      "loss": 0.6671,
+      "step": 3833
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.458475649669682,
+      "learning_rate": 8.832072250263746e-06,
+      "loss": 0.6871,
+      "step": 3834
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3379686820543697,
+      "learning_rate": 8.831406384126643e-06,
+      "loss": 0.8063,
+      "step": 3835
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2919947008016746,
+      "learning_rate": 8.830740353345475e-06,
+      "loss": 0.6492,
+      "step": 3836
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.23808996472165,
+      "learning_rate": 8.830074157948862e-06,
+      "loss": 0.6378,
+      "step": 3837
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.17804184088984,
+      "learning_rate": 8.829407797965433e-06,
+      "loss": 0.9022,
+      "step": 3838
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.386929164857071,
+      "learning_rate": 8.82874127342382e-06,
+      "loss": 0.7257,
+      "step": 3839
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3720268037035315,
+      "learning_rate": 8.82807458435267e-06,
+      "loss": 0.7025,
+      "step": 3840
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4881112515298218,
+      "learning_rate": 8.82740773078063e-06,
+      "loss": 0.6582,
+      "step": 3841
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5235461714690333,
+      "learning_rate": 8.826740712736355e-06,
+      "loss": 0.7801,
+      "step": 3842
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2206540703524595,
+      "learning_rate": 8.826073530248508e-06,
+      "loss": 0.7265,
+      "step": 3843
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1699173517244543,
+      "learning_rate": 8.82540618334576e-06,
+      "loss": 0.6451,
+      "step": 3844
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6805558497571051,
+      "learning_rate": 8.82473867205679e-06,
+      "loss": 0.6384,
+      "step": 3845
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4262808781729797,
+      "learning_rate": 8.824070996410281e-06,
+      "loss": 0.7671,
+      "step": 3846
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3571407297863756,
+      "learning_rate": 8.823403156434924e-06,
+      "loss": 0.6471,
+      "step": 3847
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3485902152213294,
+      "learning_rate": 8.822735152159418e-06,
+      "loss": 0.7729,
+      "step": 3848
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3843338615306067,
+      "learning_rate": 8.82206698361247e-06,
+      "loss": 0.6858,
+      "step": 3849
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.435694033090762,
+      "learning_rate": 8.82139865082279e-06,
+      "loss": 0.7892,
+      "step": 3850
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1926875804410493,
+      "learning_rate": 8.8207301538191e-06,
+      "loss": 0.7157,
+      "step": 3851
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4084989568980493,
+      "learning_rate": 8.820061492630127e-06,
+      "loss": 0.7152,
+      "step": 3852
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.327539464384595,
+      "learning_rate": 8.819392667284603e-06,
+      "loss": 0.6796,
+      "step": 3853
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.158511523677332,
+      "learning_rate": 8.818723677811269e-06,
+      "loss": 0.6157,
+      "step": 3854
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5191837471880547,
+      "learning_rate": 8.818054524238874e-06,
+      "loss": 0.6335,
+      "step": 3855
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1760251960733323,
+      "learning_rate": 8.817385206596174e-06,
+      "loss": 0.6683,
+      "step": 3856
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.266997823839113,
+      "learning_rate": 8.81671572491193e-06,
+      "loss": 0.7253,
+      "step": 3857
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1698431366986641,
+      "learning_rate": 8.81604607921491e-06,
+      "loss": 0.6758,
+      "step": 3858
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5202879972644625,
+      "learning_rate": 8.815376269533893e-06,
+      "loss": 0.8345,
+      "step": 3859
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2298308628328498,
+      "learning_rate": 8.814706295897659e-06,
+      "loss": 0.5794,
+      "step": 3860
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3913040469088436,
+      "learning_rate": 8.814036158335e-06,
+      "loss": 0.6312,
+      "step": 3861
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0507704722109887,
+      "learning_rate": 8.813365856874711e-06,
+      "loss": 0.5831,
+      "step": 3862
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6725811337955925,
+      "learning_rate": 8.8126953915456e-06,
+      "loss": 0.7784,
+      "step": 3863
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3561745555014033,
+      "learning_rate": 8.812024762376477e-06,
+      "loss": 0.7214,
+      "step": 3864
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5751621913312077,
+      "learning_rate": 8.81135396939616e-06,
+      "loss": 0.7951,
+      "step": 3865
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1153154391692313,
+      "learning_rate": 8.810683012633475e-06,
+      "loss": 0.6855,
+      "step": 3866
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5597506735410935,
+      "learning_rate": 8.810011892117254e-06,
+      "loss": 0.6959,
+      "step": 3867
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3385736739049485,
+      "learning_rate": 8.809340607876335e-06,
+      "loss": 0.7321,
+      "step": 3868
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2864004434109517,
+      "learning_rate": 8.808669159939568e-06,
+      "loss": 0.5891,
+      "step": 3869
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3393440333784423,
+      "learning_rate": 8.807997548335806e-06,
+      "loss": 0.7935,
+      "step": 3870
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3630756620548927,
+      "learning_rate": 8.807325773093904e-06,
+      "loss": 0.6628,
+      "step": 3871
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2617307179424182,
+      "learning_rate": 8.806653834242738e-06,
+      "loss": 0.7527,
+      "step": 3872
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.582158696211131,
+      "learning_rate": 8.805981731811177e-06,
+      "loss": 0.7513,
+      "step": 3873
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6224136761405623,
+      "learning_rate": 8.805309465828105e-06,
+      "loss": 0.7367,
+      "step": 3874
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4484183019707195,
+      "learning_rate": 8.804637036322411e-06,
+      "loss": 0.7811,
+      "step": 3875
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4422218757315846,
+      "learning_rate": 8.80396444332299e-06,
+      "loss": 0.6782,
+      "step": 3876
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4987321179001616,
+      "learning_rate": 8.803291686858745e-06,
+      "loss": 0.7699,
+      "step": 3877
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4930056872415487,
+      "learning_rate": 8.802618766958586e-06,
+      "loss": 0.6737,
+      "step": 3878
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5654886162789052,
+      "learning_rate": 8.80194568365143e-06,
+      "loss": 0.7228,
+      "step": 3879
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.270031453359261,
+      "learning_rate": 8.8012724369662e-06,
+      "loss": 0.7031,
+      "step": 3880
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4103986532736286,
+      "learning_rate": 8.800599026931827e-06,
+      "loss": 0.7284,
+      "step": 3881
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3273714271209722,
+      "learning_rate": 8.79992545357725e-06,
+      "loss": 0.6905,
+      "step": 3882
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4250571652392061,
+      "learning_rate": 8.799251716931412e-06,
+      "loss": 0.8052,
+      "step": 3883
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1343280798426887,
+      "learning_rate": 8.798577817023269e-06,
+      "loss": 0.5958,
+      "step": 3884
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.273953488425384,
+      "learning_rate": 8.797903753881775e-06,
+      "loss": 0.7275,
+      "step": 3885
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3235465194151372,
+      "learning_rate": 8.7972295275359e-06,
+      "loss": 0.6756,
+      "step": 3886
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1595057542247624,
+      "learning_rate": 8.796555138014613e-06,
+      "loss": 0.5584,
+      "step": 3887
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4476888003716564,
+      "learning_rate": 8.795880585346899e-06,
+      "loss": 0.7584,
+      "step": 3888
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3181104791207783,
+      "learning_rate": 8.795205869561742e-06,
+      "loss": 0.6149,
+      "step": 3889
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1101657660050792,
+      "learning_rate": 8.794530990688135e-06,
+      "loss": 0.6778,
+      "step": 3890
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.445286576248352,
+      "learning_rate": 8.793855948755081e-06,
+      "loss": 0.7838,
+      "step": 3891
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4530294736258333,
+      "learning_rate": 8.793180743791587e-06,
+      "loss": 0.8496,
+      "step": 3892
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.278641030861367,
+      "learning_rate": 8.79250537582667e-06,
+      "loss": 0.6225,
+      "step": 3893
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0610341537033177,
+      "learning_rate": 8.79182984488935e-06,
+      "loss": 0.7406,
+      "step": 3894
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5354446454852486,
+      "learning_rate": 8.791154151008657e-06,
+      "loss": 0.6917,
+      "step": 3895
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3269902726872569,
+      "learning_rate": 8.790478294213626e-06,
+      "loss": 0.6096,
+      "step": 3896
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0476036945728633,
+      "learning_rate": 8.789802274533301e-06,
+      "loss": 0.5989,
+      "step": 3897
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6769817019883135,
+      "learning_rate": 8.789126091996731e-06,
+      "loss": 0.7193,
+      "step": 3898
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4611172062146784,
+      "learning_rate": 8.788449746632976e-06,
+      "loss": 0.7115,
+      "step": 3899
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3954783155670876,
+      "learning_rate": 8.787773238471098e-06,
+      "loss": 0.5951,
+      "step": 3900
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3664411437377784,
+      "learning_rate": 8.787096567540166e-06,
+      "loss": 0.6774,
+      "step": 3901
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2552613250116742,
+      "learning_rate": 8.786419733869262e-06,
+      "loss": 0.5739,
+      "step": 3902
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.163207324336017,
+      "learning_rate": 8.785742737487468e-06,
+      "loss": 0.555,
+      "step": 3903
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3855104192378649,
+      "learning_rate": 8.78506557842388e-06,
+      "loss": 0.7187,
+      "step": 3904
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1514073127773992,
+      "learning_rate": 8.78438825670759e-06,
+      "loss": 0.6382,
+      "step": 3905
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3625307787554657,
+      "learning_rate": 8.78371077236771e-06,
+      "loss": 0.6275,
+      "step": 3906
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4785322227409363,
+      "learning_rate": 8.783033125433354e-06,
+      "loss": 0.7729,
+      "step": 3907
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4414094019889763,
+      "learning_rate": 8.782355315933638e-06,
+      "loss": 0.6911,
+      "step": 3908
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4393125391155766,
+      "learning_rate": 8.781677343897687e-06,
+      "loss": 0.7874,
+      "step": 3909
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4245584155278572,
+      "learning_rate": 8.780999209354642e-06,
+      "loss": 0.7323,
+      "step": 3910
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2829696396771526,
+      "learning_rate": 8.780320912333639e-06,
+      "loss": 0.7164,
+      "step": 3911
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3479916542303398,
+      "learning_rate": 8.779642452863827e-06,
+      "loss": 0.7075,
+      "step": 3912
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1378204474017002,
+      "learning_rate": 8.778963830974362e-06,
+      "loss": 0.5684,
+      "step": 3913
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7272491749285726,
+      "learning_rate": 8.778285046694403e-06,
+      "loss": 0.7796,
+      "step": 3914
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3745162613892625,
+      "learning_rate": 8.777606100053123e-06,
+      "loss": 0.8001,
+      "step": 3915
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0855110537171802,
+      "learning_rate": 8.776926991079696e-06,
+      "loss": 0.6609,
+      "step": 3916
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5676133937462222,
+      "learning_rate": 8.7762477198033e-06,
+      "loss": 0.7589,
+      "step": 3917
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2901544578142963,
+      "learning_rate": 8.775568286253135e-06,
+      "loss": 0.7045,
+      "step": 3918
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.512342745866291,
+      "learning_rate": 8.77488869045839e-06,
+      "loss": 0.7177,
+      "step": 3919
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.699096722199819,
+      "learning_rate": 8.77420893244827e-06,
+      "loss": 0.7803,
+      "step": 3920
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3314451286592648,
+      "learning_rate": 8.773529012251987e-06,
+      "loss": 0.6399,
+      "step": 3921
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1862676867825659,
+      "learning_rate": 8.772848929898758e-06,
+      "loss": 0.6527,
+      "step": 3922
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2870939496585414,
+      "learning_rate": 8.77216868541781e-06,
+      "loss": 0.6542,
+      "step": 3923
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4200520597999253,
+      "learning_rate": 8.771488278838368e-06,
+      "loss": 0.7273,
+      "step": 3924
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.468133558756688,
+      "learning_rate": 8.770807710189679e-06,
+      "loss": 0.7777,
+      "step": 3925
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1430475175079176,
+      "learning_rate": 8.770126979500984e-06,
+      "loss": 0.5483,
+      "step": 3926
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.124843092114356,
+      "learning_rate": 8.769446086801536e-06,
+      "loss": 0.7244,
+      "step": 3927
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1394640598417036,
+      "learning_rate": 8.768765032120595e-06,
+      "loss": 0.5649,
+      "step": 3928
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0364989545740944,
+      "learning_rate": 8.768083815487428e-06,
+      "loss": 0.687,
+      "step": 3929
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5150244431768531,
+      "learning_rate": 8.767402436931304e-06,
+      "loss": 0.7822,
+      "step": 3930
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3074734031912338,
+      "learning_rate": 8.76672089648151e-06,
+      "loss": 0.6689,
+      "step": 3931
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3064139719976704,
+      "learning_rate": 8.766039194167328e-06,
+      "loss": 0.6516,
+      "step": 3932
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2398291054232775,
+      "learning_rate": 8.765357330018056e-06,
+      "loss": 0.7404,
+      "step": 3933
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2243810184193693,
+      "learning_rate": 8.764675304062992e-06,
+      "loss": 0.7062,
+      "step": 3934
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.600118641289701,
+      "learning_rate": 8.763993116331448e-06,
+      "loss": 0.7783,
+      "step": 3935
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3277790264282734,
+      "learning_rate": 8.763310766852736e-06,
+      "loss": 0.7628,
+      "step": 3936
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.114678991440344,
+      "learning_rate": 8.762628255656178e-06,
+      "loss": 0.7279,
+      "step": 3937
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5042112391880953,
+      "learning_rate": 8.761945582771104e-06,
+      "loss": 0.7616,
+      "step": 3938
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.285055196515001,
+      "learning_rate": 8.76126274822685e-06,
+      "loss": 0.714,
+      "step": 3939
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4224927796884757,
+      "learning_rate": 8.76057975205276e-06,
+      "loss": 0.7629,
+      "step": 3940
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4564013460703304,
+      "learning_rate": 8.759896594278183e-06,
+      "loss": 0.6853,
+      "step": 3941
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.613061795755346,
+      "learning_rate": 8.759213274932476e-06,
+      "loss": 0.703,
+      "step": 3942
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5522075365146681,
+      "learning_rate": 8.758529794045e-06,
+      "loss": 0.7931,
+      "step": 3943
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2495344619332134,
+      "learning_rate": 8.75784615164513e-06,
+      "loss": 0.6298,
+      "step": 3944
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.271440124772513,
+      "learning_rate": 8.757162347762242e-06,
+      "loss": 0.5585,
+      "step": 3945
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8274579273342428,
+      "learning_rate": 8.75647838242572e-06,
+      "loss": 0.7193,
+      "step": 3946
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.628601510943832,
+      "learning_rate": 8.755794255664954e-06,
+      "loss": 0.8206,
+      "step": 3947
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4778531228223128,
+      "learning_rate": 8.755109967509345e-06,
+      "loss": 0.7046,
+      "step": 3948
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4202572076757745,
+      "learning_rate": 8.754425517988298e-06,
+      "loss": 0.7777,
+      "step": 3949
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.429025831251304,
+      "learning_rate": 8.753740907131226e-06,
+      "loss": 0.6295,
+      "step": 3950
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3367435654115014,
+      "learning_rate": 8.753056134967545e-06,
+      "loss": 0.8304,
+      "step": 3951
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3845331838912818,
+      "learning_rate": 8.752371201526685e-06,
+      "loss": 0.7567,
+      "step": 3952
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3711447111091375,
+      "learning_rate": 8.751686106838078e-06,
+      "loss": 0.8027,
+      "step": 3953
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2504405075562839,
+      "learning_rate": 8.751000850931162e-06,
+      "loss": 0.6304,
+      "step": 3954
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2857555089247457,
+      "learning_rate": 8.750315433835387e-06,
+      "loss": 0.6284,
+      "step": 3955
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.471534692833068,
+      "learning_rate": 8.749629855580204e-06,
+      "loss": 0.686,
+      "step": 3956
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5083260267510772,
+      "learning_rate": 8.748944116195076e-06,
+      "loss": 0.6347,
+      "step": 3957
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3409715968855427,
+      "learning_rate": 8.74825821570947e-06,
+      "loss": 0.6503,
+      "step": 3958
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.340518068240817,
+      "learning_rate": 8.74757215415286e-06,
+      "loss": 0.7542,
+      "step": 3959
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.394411855663737,
+      "learning_rate": 8.74688593155473e-06,
+      "loss": 0.7912,
+      "step": 3960
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5038474028146203,
+      "learning_rate": 8.746199547944565e-06,
+      "loss": 0.6963,
+      "step": 3961
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2043584530337488,
+      "learning_rate": 8.745513003351862e-06,
+      "loss": 0.6795,
+      "step": 3962
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.366577113938404,
+      "learning_rate": 8.744826297806124e-06,
+      "loss": 0.7224,
+      "step": 3963
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3740768376154595,
+      "learning_rate": 8.74413943133686e-06,
+      "loss": 0.6593,
+      "step": 3964
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.306173549117604,
+      "learning_rate": 8.743452403973586e-06,
+      "loss": 0.6845,
+      "step": 3965
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2979130026591585,
+      "learning_rate": 8.742765215745825e-06,
+      "loss": 0.7623,
+      "step": 3966
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4198825324589328,
+      "learning_rate": 8.742077866683108e-06,
+      "loss": 0.7723,
+      "step": 3967
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3957565120848938,
+      "learning_rate": 8.74139035681497e-06,
+      "loss": 0.7459,
+      "step": 3968
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.50157815531189,
+      "learning_rate": 8.740702686170955e-06,
+      "loss": 0.6455,
+      "step": 3969
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9083295406129455,
+      "learning_rate": 8.740014854780616e-06,
+      "loss": 0.6782,
+      "step": 3970
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5051982054457325,
+      "learning_rate": 8.739326862673508e-06,
+      "loss": 0.6922,
+      "step": 3971
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4036165307220105,
+      "learning_rate": 8.738638709879198e-06,
+      "loss": 0.7209,
+      "step": 3972
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.312883326470794,
+      "learning_rate": 8.737950396427255e-06,
+      "loss": 0.62,
+      "step": 3973
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3894762374916865,
+      "learning_rate": 8.73726192234726e-06,
+      "loss": 0.7301,
+      "step": 3974
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3626042874790143,
+      "learning_rate": 8.736573287668797e-06,
+      "loss": 0.7333,
+      "step": 3975
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5592864059308058,
+      "learning_rate": 8.735884492421457e-06,
+      "loss": 0.7378,
+      "step": 3976
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2936974676256061,
+      "learning_rate": 8.73519553663484e-06,
+      "loss": 0.6796,
+      "step": 3977
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3616836707847852,
+      "learning_rate": 8.734506420338554e-06,
+      "loss": 0.7109,
+      "step": 3978
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3575436113915826,
+      "learning_rate": 8.733817143562207e-06,
+      "loss": 0.6565,
+      "step": 3979
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2454163114526753,
+      "learning_rate": 8.733127706335423e-06,
+      "loss": 0.6551,
+      "step": 3980
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2860117443313765,
+      "learning_rate": 8.732438108687829e-06,
+      "loss": 0.8091,
+      "step": 3981
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3160413522754584,
+      "learning_rate": 8.731748350649054e-06,
+      "loss": 0.6823,
+      "step": 3982
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1558183879009098,
+      "learning_rate": 8.731058432248743e-06,
+      "loss": 0.6129,
+      "step": 3983
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3486330831993456,
+      "learning_rate": 8.73036835351654e-06,
+      "loss": 0.7159,
+      "step": 3984
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.151117401582231,
+      "learning_rate": 8.729678114482104e-06,
+      "loss": 0.7032,
+      "step": 3985
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4363446368405712,
+      "learning_rate": 8.72898771517509e-06,
+      "loss": 0.5775,
+      "step": 3986
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.187294743110652,
+      "learning_rate": 8.728297155625171e-06,
+      "loss": 0.5515,
+      "step": 3987
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4128653985828223,
+      "learning_rate": 8.727606435862018e-06,
+      "loss": 0.6734,
+      "step": 3988
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2646992473333805,
+      "learning_rate": 8.726915555915317e-06,
+      "loss": 0.6614,
+      "step": 3989
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7126009245818745,
+      "learning_rate": 8.726224515814752e-06,
+      "loss": 0.7566,
+      "step": 3990
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3151369346903221,
+      "learning_rate": 8.72553331559002e-06,
+      "loss": 0.6482,
+      "step": 3991
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.387655646094381,
+      "learning_rate": 8.724841955270827e-06,
+      "loss": 0.7195,
+      "step": 3992
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.345027016342766,
+      "learning_rate": 8.724150434886878e-06,
+      "loss": 0.7244,
+      "step": 3993
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3532510871834325,
+      "learning_rate": 8.723458754467893e-06,
+      "loss": 0.6949,
+      "step": 3994
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2277514707250134,
+      "learning_rate": 8.72276691404359e-06,
+      "loss": 0.6195,
+      "step": 3995
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4263970931878214,
+      "learning_rate": 8.722074913643703e-06,
+      "loss": 0.6486,
+      "step": 3996
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2364861097744657,
+      "learning_rate": 8.721382753297967e-06,
+      "loss": 0.59,
+      "step": 3997
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.286343627870259,
+      "learning_rate": 8.720690433036125e-06,
+      "loss": 0.669,
+      "step": 3998
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.627029025074474,
+      "learning_rate": 8.719997952887932e-06,
+      "loss": 0.6281,
+      "step": 3999
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4923092856212075,
+      "learning_rate": 8.719305312883137e-06,
+      "loss": 0.6599,
+      "step": 4000
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4228900322890028,
+      "learning_rate": 8.718612513051513e-06,
+      "loss": 0.6784,
+      "step": 4001
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.446795023812584,
+      "learning_rate": 8.717919553422827e-06,
+      "loss": 0.7716,
+      "step": 4002
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2245472308696215,
+      "learning_rate": 8.717226434026858e-06,
+      "loss": 0.6919,
+      "step": 4003
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.16599946019891,
+      "learning_rate": 8.71653315489339e-06,
+      "loss": 0.6242,
+      "step": 4004
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3723007874894244,
+      "learning_rate": 8.715839716052215e-06,
+      "loss": 0.7875,
+      "step": 4005
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3038599166489617,
+      "learning_rate": 8.715146117533131e-06,
+      "loss": 0.7048,
+      "step": 4006
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5240993534859482,
+      "learning_rate": 8.714452359365943e-06,
+      "loss": 0.7436,
+      "step": 4007
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3958936575042364,
+      "learning_rate": 8.713758441580467e-06,
+      "loss": 0.6614,
+      "step": 4008
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.128492961059908,
+      "learning_rate": 8.71306436420652e-06,
+      "loss": 0.5994,
+      "step": 4009
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9627838145520915,
+      "learning_rate": 8.712370127273927e-06,
+      "loss": 0.5289,
+      "step": 4010
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4859025643542163,
+      "learning_rate": 8.711675730812522e-06,
+      "loss": 0.5951,
+      "step": 4011
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1352469283299351,
+      "learning_rate": 8.710981174852144e-06,
+      "loss": 0.697,
+      "step": 4012
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3882734530150542,
+      "learning_rate": 8.710286459422641e-06,
+      "loss": 0.6084,
+      "step": 4013
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.375072645850066,
+      "learning_rate": 8.709591584553865e-06,
+      "loss": 0.6432,
+      "step": 4014
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.245696097498234,
+      "learning_rate": 8.708896550275675e-06,
+      "loss": 0.5858,
+      "step": 4015
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2201371524827127,
+      "learning_rate": 8.708201356617945e-06,
+      "loss": 0.7036,
+      "step": 4016
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4556003311179886,
+      "learning_rate": 8.70750600361054e-06,
+      "loss": 0.6989,
+      "step": 4017
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.26313381552746,
+      "learning_rate": 8.706810491283346e-06,
+      "loss": 0.6116,
+      "step": 4018
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.341407135930128,
+      "learning_rate": 8.706114819666249e-06,
+      "loss": 0.6811,
+      "step": 4019
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.259708228177282,
+      "learning_rate": 8.705418988789146e-06,
+      "loss": 0.7262,
+      "step": 4020
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.301894229321876,
+      "learning_rate": 8.704722998681937e-06,
+      "loss": 0.7344,
+      "step": 4021
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2030035616010089,
+      "learning_rate": 8.704026849374527e-06,
+      "loss": 0.6076,
+      "step": 4022
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4312579860211432,
+      "learning_rate": 8.703330540896836e-06,
+      "loss": 0.7201,
+      "step": 4023
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1398935391420701,
+      "learning_rate": 8.702634073278784e-06,
+      "loss": 0.6483,
+      "step": 4024
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2503353475927415,
+      "learning_rate": 8.701937446550298e-06,
+      "loss": 0.686,
+      "step": 4025
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.380272893585649,
+      "learning_rate": 8.701240660741317e-06,
+      "loss": 0.728,
+      "step": 4026
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3086618661767633,
+      "learning_rate": 8.700543715881781e-06,
+      "loss": 0.6361,
+      "step": 4027
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4288926594320799,
+      "learning_rate": 8.699846612001638e-06,
+      "loss": 0.7471,
+      "step": 4028
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.338248856657767,
+      "learning_rate": 8.699149349130848e-06,
+      "loss": 0.7697,
+      "step": 4029
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.553937367266149,
+      "learning_rate": 8.698451927299374e-06,
+      "loss": 0.8027,
+      "step": 4030
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6051627090331244,
+      "learning_rate": 8.697754346537182e-06,
+      "loss": 0.7163,
+      "step": 4031
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3534435215298206,
+      "learning_rate": 8.69705660687425e-06,
+      "loss": 0.6689,
+      "step": 4032
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5400486522744994,
+      "learning_rate": 8.696358708340562e-06,
+      "loss": 0.7367,
+      "step": 4033
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3578683022479008,
+      "learning_rate": 8.695660650966109e-06,
+      "loss": 0.6442,
+      "step": 4034
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2089133088498396,
+      "learning_rate": 8.694962434780885e-06,
+      "loss": 0.6748,
+      "step": 4035
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5086908396729537,
+      "learning_rate": 8.694264059814897e-06,
+      "loss": 0.8289,
+      "step": 4036
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4784831405176273,
+      "learning_rate": 8.693565526098156e-06,
+      "loss": 0.6914,
+      "step": 4037
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4244363887645428,
+      "learning_rate": 8.692866833660679e-06,
+      "loss": 0.7278,
+      "step": 4038
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1455339496909447,
+      "learning_rate": 8.692167982532487e-06,
+      "loss": 0.6299,
+      "step": 4039
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9018786373767151,
+      "learning_rate": 8.691468972743615e-06,
+      "loss": 0.6488,
+      "step": 4040
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2296138552405502,
+      "learning_rate": 8.6907698043241e-06,
+      "loss": 0.657,
+      "step": 4041
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2143825304660056,
+      "learning_rate": 8.690070477303987e-06,
+      "loss": 0.6629,
+      "step": 4042
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1983240594705233,
+      "learning_rate": 8.689370991713327e-06,
+      "loss": 0.6625,
+      "step": 4043
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4218621213935356,
+      "learning_rate": 8.688671347582178e-06,
+      "loss": 0.6951,
+      "step": 4044
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3630970044180335,
+      "learning_rate": 8.687971544940608e-06,
+      "loss": 0.7851,
+      "step": 4045
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2559866750064645,
+      "learning_rate": 8.687271583818687e-06,
+      "loss": 0.7433,
+      "step": 4046
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.8669076551870654,
+      "learning_rate": 8.686571464246491e-06,
+      "loss": 0.6933,
+      "step": 4047
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.387094804511501,
+      "learning_rate": 8.685871186254112e-06,
+      "loss": 0.7095,
+      "step": 4048
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4204972919828311,
+      "learning_rate": 8.685170749871638e-06,
+      "loss": 0.6697,
+      "step": 4049
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3732024143000037,
+      "learning_rate": 8.68447015512917e-06,
+      "loss": 0.6901,
+      "step": 4050
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4562885034672008,
+      "learning_rate": 8.683769402056814e-06,
+      "loss": 0.7064,
+      "step": 4051
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5782402772776936,
+      "learning_rate": 8.683068490684681e-06,
+      "loss": 0.6544,
+      "step": 4052
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3265149261063658,
+      "learning_rate": 8.682367421042895e-06,
+      "loss": 0.6585,
+      "step": 4053
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4663386407051413,
+      "learning_rate": 8.681666193161578e-06,
+      "loss": 0.7526,
+      "step": 4054
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3125192888387232,
+      "learning_rate": 8.680964807070865e-06,
+      "loss": 0.676,
+      "step": 4055
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.365746411885521,
+      "learning_rate": 8.680263262800897e-06,
+      "loss": 0.6685,
+      "step": 4056
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3581151911890545,
+      "learning_rate": 8.679561560381818e-06,
+      "loss": 0.6288,
+      "step": 4057
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.412166774132668,
+      "learning_rate": 8.678859699843787e-06,
+      "loss": 0.7104,
+      "step": 4058
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2246281800823116,
+      "learning_rate": 8.67815768121696e-06,
+      "loss": 0.6544,
+      "step": 4059
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5097643563704384,
+      "learning_rate": 8.677455504531507e-06,
+      "loss": 0.7968,
+      "step": 4060
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0620935246493262,
+      "learning_rate": 8.676753169817598e-06,
+      "loss": 0.7154,
+      "step": 4061
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5295255505606806,
+      "learning_rate": 8.676050677105419e-06,
+      "loss": 0.69,
+      "step": 4062
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3630737431288806,
+      "learning_rate": 8.675348026425154e-06,
+      "loss": 0.7051,
+      "step": 4063
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4161512381080226,
+      "learning_rate": 8.674645217807e-06,
+      "loss": 0.6788,
+      "step": 4064
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.189057073397432,
+      "learning_rate": 8.673942251281158e-06,
+      "loss": 0.657,
+      "step": 4065
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3696917420566237,
+      "learning_rate": 8.673239126877835e-06,
+      "loss": 0.7641,
+      "step": 4066
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.129071055508469,
+      "learning_rate": 8.672535844627243e-06,
+      "loss": 0.671,
+      "step": 4067
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2097878727839995,
+      "learning_rate": 8.67183240455961e-06,
+      "loss": 0.7407,
+      "step": 4068
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3698253202099917,
+      "learning_rate": 8.671128806705159e-06,
+      "loss": 0.735,
+      "step": 4069
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.403100732959251,
+      "learning_rate": 8.670425051094128e-06,
+      "loss": 0.6679,
+      "step": 4070
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4771457431613897,
+      "learning_rate": 8.669721137756756e-06,
+      "loss": 0.6058,
+      "step": 4071
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2832395525777092,
+      "learning_rate": 8.669017066723297e-06,
+      "loss": 0.6902,
+      "step": 4072
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.205742634359631,
+      "learning_rate": 8.668312838024002e-06,
+      "loss": 0.6502,
+      "step": 4073
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3629271344230613,
+      "learning_rate": 8.667608451689135e-06,
+      "loss": 0.7342,
+      "step": 4074
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.196016083983609,
+      "learning_rate": 8.666903907748963e-06,
+      "loss": 0.6455,
+      "step": 4075
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3451216710315534,
+      "learning_rate": 8.666199206233765e-06,
+      "loss": 0.6909,
+      "step": 4076
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3779863267054806,
+      "learning_rate": 8.665494347173822e-06,
+      "loss": 0.6556,
+      "step": 4077
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5134298880269197,
+      "learning_rate": 8.664789330599423e-06,
+      "loss": 0.7046,
+      "step": 4078
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3784625927758607,
+      "learning_rate": 8.664084156540864e-06,
+      "loss": 0.6832,
+      "step": 4079
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3855681928962102,
+      "learning_rate": 8.66337882502845e-06,
+      "loss": 0.733,
+      "step": 4080
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.290434581377569,
+      "learning_rate": 8.662673336092487e-06,
+      "loss": 0.7131,
+      "step": 4081
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3464429770355808,
+      "learning_rate": 8.661967689763296e-06,
+      "loss": 0.7191,
+      "step": 4082
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2630278132280466,
+      "learning_rate": 8.661261886071194e-06,
+      "loss": 0.712,
+      "step": 4083
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.291199825884452,
+      "learning_rate": 8.660555925046518e-06,
+      "loss": 0.6571,
+      "step": 4084
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4583684397117591,
+      "learning_rate": 8.659849806719602e-06,
+      "loss": 0.8173,
+      "step": 4085
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3891264490592956,
+      "learning_rate": 8.659143531120785e-06,
+      "loss": 0.6685,
+      "step": 4086
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5422719773815425,
+      "learning_rate": 8.658437098280426e-06,
+      "loss": 0.8355,
+      "step": 4087
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2062251616564796,
+      "learning_rate": 8.657730508228874e-06,
+      "loss": 0.6345,
+      "step": 4088
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5051195269183146,
+      "learning_rate": 8.657023760996497e-06,
+      "loss": 0.6827,
+      "step": 4089
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9404942453891991,
+      "learning_rate": 8.656316856613662e-06,
+      "loss": 0.6741,
+      "step": 4090
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.348020827648395,
+      "learning_rate": 8.655609795110751e-06,
+      "loss": 0.6637,
+      "step": 4091
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5155188404519768,
+      "learning_rate": 8.654902576518145e-06,
+      "loss": 0.7129,
+      "step": 4092
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5863796377857178,
+      "learning_rate": 8.654195200866236e-06,
+      "loss": 0.5869,
+      "step": 4093
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3976023439910834,
+      "learning_rate": 8.653487668185419e-06,
+      "loss": 0.7735,
+      "step": 4094
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2905584136804558,
+      "learning_rate": 8.652779978506103e-06,
+      "loss": 0.6939,
+      "step": 4095
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4302743002626506,
+      "learning_rate": 8.652072131858694e-06,
+      "loss": 0.7815,
+      "step": 4096
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.521207233706361,
+      "learning_rate": 8.651364128273612e-06,
+      "loss": 0.9088,
+      "step": 4097
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5523207207007153,
+      "learning_rate": 8.650655967781282e-06,
+      "loss": 0.7579,
+      "step": 4098
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2899758108155237,
+      "learning_rate": 8.649947650412135e-06,
+      "loss": 0.6572,
+      "step": 4099
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4523039682311107,
+      "learning_rate": 8.649239176196609e-06,
+      "loss": 0.7395,
+      "step": 4100
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5448759598545285,
+      "learning_rate": 8.648530545165146e-06,
+      "loss": 0.625,
+      "step": 4101
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5665902977326605,
+      "learning_rate": 8.647821757348202e-06,
+      "loss": 0.6922,
+      "step": 4102
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1430462596690065,
+      "learning_rate": 8.647112812776231e-06,
+      "loss": 0.6089,
+      "step": 4103
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3998214772296187,
+      "learning_rate": 8.646403711479702e-06,
+      "loss": 0.6818,
+      "step": 4104
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5097031649933166,
+      "learning_rate": 8.645694453489085e-06,
+      "loss": 0.6629,
+      "step": 4105
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2318723268697755,
+      "learning_rate": 8.644985038834855e-06,
+      "loss": 0.6303,
+      "step": 4106
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2622441535585458,
+      "learning_rate": 8.644275467547502e-06,
+      "loss": 0.5965,
+      "step": 4107
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6053925548900054,
+      "learning_rate": 8.643565739657515e-06,
+      "loss": 0.7249,
+      "step": 4108
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4209066577971952,
+      "learning_rate": 8.642855855195394e-06,
+      "loss": 0.6851,
+      "step": 4109
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2888861620066512,
+      "learning_rate": 8.642145814191643e-06,
+      "loss": 0.75,
+      "step": 4110
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.155662744504106,
+      "learning_rate": 8.641435616676776e-06,
+      "loss": 0.611,
+      "step": 4111
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3138371636008699,
+      "learning_rate": 8.64072526268131e-06,
+      "loss": 0.6167,
+      "step": 4112
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3567244592207135,
+      "learning_rate": 8.640014752235773e-06,
+      "loss": 0.6439,
+      "step": 4113
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.331558452539689,
+      "learning_rate": 8.639304085370692e-06,
+      "loss": 0.6665,
+      "step": 4114
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3169881158029901,
+      "learning_rate": 8.638593262116612e-06,
+      "loss": 0.7219,
+      "step": 4115
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1559635586901589,
+      "learning_rate": 8.637882282504075e-06,
+      "loss": 0.6613,
+      "step": 4116
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.409235200941238,
+      "learning_rate": 8.637171146563634e-06,
+      "loss": 0.6334,
+      "step": 4117
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3473212899791807,
+      "learning_rate": 8.636459854325849e-06,
+      "loss": 0.8222,
+      "step": 4118
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.362859488498707,
+      "learning_rate": 8.635748405821285e-06,
+      "loss": 0.6211,
+      "step": 4119
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.539672184089517,
+      "learning_rate": 8.635036801080513e-06,
+      "loss": 0.6958,
+      "step": 4120
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2485453109645417,
+      "learning_rate": 8.634325040134117e-06,
+      "loss": 0.7,
+      "step": 4121
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2804190210100153,
+      "learning_rate": 8.63361312301268e-06,
+      "loss": 0.7147,
+      "step": 4122
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4320493534176983,
+      "learning_rate": 8.632901049746793e-06,
+      "loss": 0.6119,
+      "step": 4123
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.343486856267334,
+      "learning_rate": 8.632188820367056e-06,
+      "loss": 0.7115,
+      "step": 4124
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2816058881985612,
+      "learning_rate": 8.631476434904077e-06,
+      "loss": 0.7418,
+      "step": 4125
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6034675228830237,
+      "learning_rate": 8.63076389338847e-06,
+      "loss": 0.7951,
+      "step": 4126
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.070201957158151,
+      "learning_rate": 8.630051195850851e-06,
+      "loss": 0.7088,
+      "step": 4127
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3151993215427864,
+      "learning_rate": 8.629338342321846e-06,
+      "loss": 0.7924,
+      "step": 4128
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.199689116289106,
+      "learning_rate": 8.62862533283209e-06,
+      "loss": 0.6199,
+      "step": 4129
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4403099850024639,
+      "learning_rate": 8.627912167412222e-06,
+      "loss": 0.6731,
+      "step": 4130
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3765062771682677,
+      "learning_rate": 8.62719884609289e-06,
+      "loss": 0.6331,
+      "step": 4131
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2332528347471179,
+      "learning_rate": 8.626485368904744e-06,
+      "loss": 0.6964,
+      "step": 4132
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3502272880209865,
+      "learning_rate": 8.625771735878445e-06,
+      "loss": 0.7126,
+      "step": 4133
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.050002855259659,
+      "learning_rate": 8.625057947044662e-06,
+      "loss": 0.6511,
+      "step": 4134
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3812566295707098,
+      "learning_rate": 8.624344002434061e-06,
+      "loss": 0.6715,
+      "step": 4135
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4493261265817245,
+      "learning_rate": 8.62362990207733e-06,
+      "loss": 0.6223,
+      "step": 4136
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9676056859134988,
+      "learning_rate": 8.622915646005152e-06,
+      "loss": 0.6429,
+      "step": 4137
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1242350961749759,
+      "learning_rate": 8.622201234248218e-06,
+      "loss": 0.5578,
+      "step": 4138
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2806299152048743,
+      "learning_rate": 8.62148666683723e-06,
+      "loss": 0.6003,
+      "step": 4139
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5337134402247248,
+      "learning_rate": 8.620771943802895e-06,
+      "loss": 0.7412,
+      "step": 4140
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4540597284879488,
+      "learning_rate": 8.620057065175926e-06,
+      "loss": 0.7668,
+      "step": 4141
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1283674731573088,
+      "learning_rate": 8.619342030987044e-06,
+      "loss": 0.687,
+      "step": 4142
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3725500291942447,
+      "learning_rate": 8.618626841266972e-06,
+      "loss": 0.6885,
+      "step": 4143
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4404254611365577,
+      "learning_rate": 8.617911496046446e-06,
+      "loss": 0.651,
+      "step": 4144
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.068540848335385,
+      "learning_rate": 8.617195995356206e-06,
+      "loss": 0.6389,
+      "step": 4145
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2244257634208542,
+      "learning_rate": 8.616480339226998e-06,
+      "loss": 0.6372,
+      "step": 4146
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3834588405621093,
+      "learning_rate": 8.615764527689574e-06,
+      "loss": 0.6835,
+      "step": 4147
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2605359062777495,
+      "learning_rate": 8.615048560774698e-06,
+      "loss": 0.6911,
+      "step": 4148
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4950431399039272,
+      "learning_rate": 8.614332438513132e-06,
+      "loss": 0.6504,
+      "step": 4149
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0913183723629605,
+      "learning_rate": 8.613616160935652e-06,
+      "loss": 0.6943,
+      "step": 4150
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2749465661852146,
+      "learning_rate": 8.612899728073039e-06,
+      "loss": 0.7385,
+      "step": 4151
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3337771209992442,
+      "learning_rate": 8.612183139956078e-06,
+      "loss": 0.6523,
+      "step": 4152
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4615254144833365,
+      "learning_rate": 8.611466396615562e-06,
+      "loss": 0.6706,
+      "step": 4153
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2602953976018385,
+      "learning_rate": 8.610749498082291e-06,
+      "loss": 0.5889,
+      "step": 4154
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3376975483551135,
+      "learning_rate": 8.610032444387074e-06,
+      "loss": 0.8125,
+      "step": 4155
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1082246004656815,
+      "learning_rate": 8.609315235560722e-06,
+      "loss": 0.6048,
+      "step": 4156
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3377509759908408,
+      "learning_rate": 8.608597871634056e-06,
+      "loss": 0.669,
+      "step": 4157
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2816209894914712,
+      "learning_rate": 8.607880352637905e-06,
+      "loss": 0.6914,
+      "step": 4158
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3776197171123774,
+      "learning_rate": 8.607162678603097e-06,
+      "loss": 0.6999,
+      "step": 4159
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0942946934662718,
+      "learning_rate": 8.606444849560476e-06,
+      "loss": 0.6946,
+      "step": 4160
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1200676958203648,
+      "learning_rate": 8.605726865540889e-06,
+      "loss": 0.5894,
+      "step": 4161
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2484894201754808,
+      "learning_rate": 8.605008726575186e-06,
+      "loss": 0.6661,
+      "step": 4162
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2519610118717959,
+      "learning_rate": 8.60429043269423e-06,
+      "loss": 0.6811,
+      "step": 4163
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5459724892793894,
+      "learning_rate": 8.603571983928888e-06,
+      "loss": 0.7177,
+      "step": 4164
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3742670441195681,
+      "learning_rate": 8.602853380310033e-06,
+      "loss": 0.6162,
+      "step": 4165
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3625908235759656,
+      "learning_rate": 8.602134621868542e-06,
+      "loss": 0.7416,
+      "step": 4166
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2026832857349612,
+      "learning_rate": 8.601415708635304e-06,
+      "loss": 0.7192,
+      "step": 4167
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1623292472404538,
+      "learning_rate": 8.600696640641213e-06,
+      "loss": 0.6821,
+      "step": 4168
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.095024388194761,
+      "learning_rate": 8.599977417917169e-06,
+      "loss": 0.67,
+      "step": 4169
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7767131054342107,
+      "learning_rate": 8.599258040494078e-06,
+      "loss": 0.6204,
+      "step": 4170
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5629591183734448,
+      "learning_rate": 8.59853850840285e-06,
+      "loss": 0.6776,
+      "step": 4171
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.424868516998497,
+      "learning_rate": 8.59781882167441e-06,
+      "loss": 0.7665,
+      "step": 4172
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.367120136366104,
+      "learning_rate": 8.597098980339683e-06,
+      "loss": 0.8224,
+      "step": 4173
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.394019697571849,
+      "learning_rate": 8.5963789844296e-06,
+      "loss": 0.9237,
+      "step": 4174
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.172477609141282,
+      "learning_rate": 8.595658833975104e-06,
+      "loss": 0.6272,
+      "step": 4175
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3786317487023236,
+      "learning_rate": 8.59493852900714e-06,
+      "loss": 0.6352,
+      "step": 4176
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4577066717246898,
+      "learning_rate": 8.59421806955666e-06,
+      "loss": 0.7178,
+      "step": 4177
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6060635117043036,
+      "learning_rate": 8.593497455654627e-06,
+      "loss": 0.6003,
+      "step": 4178
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3420135497176944,
+      "learning_rate": 8.592776687332003e-06,
+      "loss": 0.7235,
+      "step": 4179
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2394758734198617,
+      "learning_rate": 8.592055764619762e-06,
+      "loss": 0.6788,
+      "step": 4180
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8138747127526482,
+      "learning_rate": 8.591334687548888e-06,
+      "loss": 0.6004,
+      "step": 4181
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2271926722997872,
+      "learning_rate": 8.590613456150364e-06,
+      "loss": 0.5489,
+      "step": 4182
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.61645808384714,
+      "learning_rate": 8.58989207045518e-06,
+      "loss": 0.6487,
+      "step": 4183
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.334505559672788,
+      "learning_rate": 8.58917053049434e-06,
+      "loss": 0.6975,
+      "step": 4184
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1883453614957933,
+      "learning_rate": 8.58844883629885e-06,
+      "loss": 0.6203,
+      "step": 4185
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0450015848225704,
+      "learning_rate": 8.58772698789972e-06,
+      "loss": 0.6171,
+      "step": 4186
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2292223506208522,
+      "learning_rate": 8.587004985327971e-06,
+      "loss": 0.6692,
+      "step": 4187
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.349241079924775,
+      "learning_rate": 8.586282828614631e-06,
+      "loss": 0.7471,
+      "step": 4188
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0883197534502511,
+      "learning_rate": 8.58556051779073e-06,
+      "loss": 0.5624,
+      "step": 4189
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3690093874131217,
+      "learning_rate": 8.584838052887308e-06,
+      "loss": 0.8005,
+      "step": 4190
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.400213001312877,
+      "learning_rate": 8.58411543393541e-06,
+      "loss": 0.6392,
+      "step": 4191
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2430542430213363,
+      "learning_rate": 8.58339266096609e-06,
+      "loss": 0.5962,
+      "step": 4192
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4224015096966889,
+      "learning_rate": 8.582669734010407e-06,
+      "loss": 0.8635,
+      "step": 4193
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2633436613351074,
+      "learning_rate": 8.581946653099427e-06,
+      "loss": 0.6936,
+      "step": 4194
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4837472522400343,
+      "learning_rate": 8.581223418264222e-06,
+      "loss": 0.6959,
+      "step": 4195
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3512357878571815,
+      "learning_rate": 8.58050002953587e-06,
+      "loss": 0.7859,
+      "step": 4196
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.175209680973855,
+      "learning_rate": 8.579776486945457e-06,
+      "loss": 0.6075,
+      "step": 4197
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3287002698837178,
+      "learning_rate": 8.579052790524077e-06,
+      "loss": 0.6581,
+      "step": 4198
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4076734102063089,
+      "learning_rate": 8.578328940302827e-06,
+      "loss": 0.7339,
+      "step": 4199
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5130625572281824,
+      "learning_rate": 8.577604936312813e-06,
+      "loss": 0.8757,
+      "step": 4200
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3214568412845935,
+      "learning_rate": 8.576880778585148e-06,
+      "loss": 0.6595,
+      "step": 4201
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2264830111000777,
+      "learning_rate": 8.57615646715095e-06,
+      "loss": 0.6593,
+      "step": 4202
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3886784275246495,
+      "learning_rate": 8.575432002041341e-06,
+      "loss": 0.6081,
+      "step": 4203
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.326411602830511,
+      "learning_rate": 8.574707383287459e-06,
+      "loss": 0.6798,
+      "step": 4204
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4310639915899084,
+      "learning_rate": 8.57398261092044e-06,
+      "loss": 0.8275,
+      "step": 4205
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.321576856347319,
+      "learning_rate": 8.573257684971425e-06,
+      "loss": 0.5962,
+      "step": 4206
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3538952483380833,
+      "learning_rate": 8.572532605471572e-06,
+      "loss": 0.6926,
+      "step": 4207
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.527570525264118,
+      "learning_rate": 8.571807372452036e-06,
+      "loss": 0.7086,
+      "step": 4208
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.088372622127943,
+      "learning_rate": 8.571081985943984e-06,
+      "loss": 0.6088,
+      "step": 4209
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5241432896316764,
+      "learning_rate": 8.570356445978583e-06,
+      "loss": 0.832,
+      "step": 4210
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2896789797307993,
+      "learning_rate": 8.569630752587014e-06,
+      "loss": 0.7268,
+      "step": 4211
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3373668501060667,
+      "learning_rate": 8.568904905800464e-06,
+      "loss": 0.6468,
+      "step": 4212
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.342497945521129,
+      "learning_rate": 8.56817890565012e-06,
+      "loss": 0.7205,
+      "step": 4213
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3499494737421833,
+      "learning_rate": 8.567452752167183e-06,
+      "loss": 0.6733,
+      "step": 4214
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2030294018415089,
+      "learning_rate": 8.566726445382854e-06,
+      "loss": 0.7313,
+      "step": 4215
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5258275147990272,
+      "learning_rate": 8.565999985328348e-06,
+      "loss": 0.8168,
+      "step": 4216
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5602745273485377,
+      "learning_rate": 8.565273372034879e-06,
+      "loss": 0.7914,
+      "step": 4217
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3425468719317843,
+      "learning_rate": 8.564546605533674e-06,
+      "loss": 0.7392,
+      "step": 4218
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.380077347103825,
+      "learning_rate": 8.563819685855963e-06,
+      "loss": 0.7351,
+      "step": 4219
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6053968780210095,
+      "learning_rate": 8.563092613032981e-06,
+      "loss": 0.7641,
+      "step": 4220
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1110196876774818,
+      "learning_rate": 8.562365387095977e-06,
+      "loss": 0.6429,
+      "step": 4221
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1985526463245415,
+      "learning_rate": 8.561638008076197e-06,
+      "loss": 0.6562,
+      "step": 4222
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4455324351263126,
+      "learning_rate": 8.5609104760049e-06,
+      "loss": 0.672,
+      "step": 4223
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3548469830378302,
+      "learning_rate": 8.560182790913349e-06,
+      "loss": 0.7011,
+      "step": 4224
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2557098923579686,
+      "learning_rate": 8.559454952832815e-06,
+      "loss": 0.5869,
+      "step": 4225
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.166805289442246,
+      "learning_rate": 8.558726961794573e-06,
+      "loss": 0.5583,
+      "step": 4226
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9880555014314758,
+      "learning_rate": 8.557998817829909e-06,
+      "loss": 0.6189,
+      "step": 4227
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.372108768903103,
+      "learning_rate": 8.557270520970111e-06,
+      "loss": 0.7643,
+      "step": 4228
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3268188180585432,
+      "learning_rate": 8.556542071246476e-06,
+      "loss": 0.6658,
+      "step": 4229
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4082126064716185,
+      "learning_rate": 8.555813468690309e-06,
+      "loss": 0.7204,
+      "step": 4230
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7447947054512067,
+      "learning_rate": 8.555084713332917e-06,
+      "loss": 0.8188,
+      "step": 4231
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3189471894174627,
+      "learning_rate": 8.554355805205616e-06,
+      "loss": 0.7213,
+      "step": 4232
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0670153513946998,
+      "learning_rate": 8.553626744339732e-06,
+      "loss": 0.7318,
+      "step": 4233
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2962208534408646,
+      "learning_rate": 8.552897530766592e-06,
+      "loss": 0.6629,
+      "step": 4234
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3149176089534873,
+      "learning_rate": 8.552168164517532e-06,
+      "loss": 0.7087,
+      "step": 4235
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4128189997097402,
+      "learning_rate": 8.551438645623896e-06,
+      "loss": 0.6581,
+      "step": 4236
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2683901780484153,
+      "learning_rate": 8.55070897411703e-06,
+      "loss": 0.6397,
+      "step": 4237
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1172179363218238,
+      "learning_rate": 8.549979150028292e-06,
+      "loss": 0.7293,
+      "step": 4238
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4470086212513342,
+      "learning_rate": 8.549249173389045e-06,
+      "loss": 0.624,
+      "step": 4239
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1786928560654466,
+      "learning_rate": 8.548519044230654e-06,
+      "loss": 0.5956,
+      "step": 4240
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2166288398078111,
+      "learning_rate": 8.5477887625845e-06,
+      "loss": 0.7374,
+      "step": 4241
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1357914779959133,
+      "learning_rate": 8.547058328481959e-06,
+      "loss": 0.6839,
+      "step": 4242
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2671410555495624,
+      "learning_rate": 8.546327741954423e-06,
+      "loss": 0.6433,
+      "step": 4243
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5656287962633793,
+      "learning_rate": 8.545597003033286e-06,
+      "loss": 0.7042,
+      "step": 4244
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3789685709662982,
+      "learning_rate": 8.544866111749948e-06,
+      "loss": 0.6871,
+      "step": 4245
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1201539097718143,
+      "learning_rate": 8.544135068135819e-06,
+      "loss": 0.5353,
+      "step": 4246
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3586312370730667,
+      "learning_rate": 8.543403872222313e-06,
+      "loss": 0.6129,
+      "step": 4247
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4995240225074384,
+      "learning_rate": 8.54267252404085e-06,
+      "loss": 0.6754,
+      "step": 4248
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4616729292745971,
+      "learning_rate": 8.54194102362286e-06,
+      "loss": 0.6446,
+      "step": 4249
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2564092507140514,
+      "learning_rate": 8.541209370999777e-06,
+      "loss": 0.7425,
+      "step": 4250
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.308765075005006,
+      "learning_rate": 8.540477566203039e-06,
+      "loss": 0.6631,
+      "step": 4251
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0642665441405066,
+      "learning_rate": 8.539745609264094e-06,
+      "loss": 0.6354,
+      "step": 4252
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2411107883194352,
+      "learning_rate": 8.539013500214399e-06,
+      "loss": 0.6395,
+      "step": 4253
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4384416359892438,
+      "learning_rate": 8.538281239085411e-06,
+      "loss": 0.805,
+      "step": 4254
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5749047173546022,
+      "learning_rate": 8.537548825908597e-06,
+      "loss": 0.642,
+      "step": 4255
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4550355474208791,
+      "learning_rate": 8.536816260715433e-06,
+      "loss": 0.6984,
+      "step": 4256
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4518761374533111,
+      "learning_rate": 8.536083543537396e-06,
+      "loss": 0.6363,
+      "step": 4257
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.7834562625110695,
+      "learning_rate": 8.535350674405975e-06,
+      "loss": 0.7146,
+      "step": 4258
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.447743816523164,
+      "learning_rate": 8.534617653352661e-06,
+      "loss": 0.8305,
+      "step": 4259
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0133762141754483,
+      "learning_rate": 8.533884480408955e-06,
+      "loss": 0.6101,
+      "step": 4260
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2442130371887694,
+      "learning_rate": 8.533151155606364e-06,
+      "loss": 0.701,
+      "step": 4261
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3050765655101189,
+      "learning_rate": 8.532417678976398e-06,
+      "loss": 0.6324,
+      "step": 4262
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.378628784911102,
+      "learning_rate": 8.531684050550575e-06,
+      "loss": 0.7535,
+      "step": 4263
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3846407350241783,
+      "learning_rate": 8.530950270360425e-06,
+      "loss": 0.7318,
+      "step": 4264
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.477877673235753,
+      "learning_rate": 8.530216338437478e-06,
+      "loss": 0.7161,
+      "step": 4265
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1331705794602716,
+      "learning_rate": 8.529482254813272e-06,
+      "loss": 0.6802,
+      "step": 4266
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2093399752859808,
+      "learning_rate": 8.528748019519352e-06,
+      "loss": 0.6673,
+      "step": 4267
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.511313346095187,
+      "learning_rate": 8.528013632587272e-06,
+      "loss": 0.7182,
+      "step": 4268
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.182851076299292,
+      "learning_rate": 8.52727909404859e-06,
+      "loss": 0.6,
+      "step": 4269
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.420316131578891,
+      "learning_rate": 8.526544403934868e-06,
+      "loss": 0.7382,
+      "step": 4270
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.554832734634464,
+      "learning_rate": 8.52580956227768e-06,
+      "loss": 0.6864,
+      "step": 4271
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3412831938293663,
+      "learning_rate": 8.525074569108603e-06,
+      "loss": 0.6613,
+      "step": 4272
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4350887790458726,
+      "learning_rate": 8.524339424459219e-06,
+      "loss": 0.6382,
+      "step": 4273
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4023477337448427,
+      "learning_rate": 8.523604128361123e-06,
+      "loss": 0.7376,
+      "step": 4274
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4416408293726846,
+      "learning_rate": 8.522868680845908e-06,
+      "loss": 0.7296,
+      "step": 4275
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1881785157702534,
+      "learning_rate": 8.52213308194518e-06,
+      "loss": 0.6577,
+      "step": 4276
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1746777296421524,
+      "learning_rate": 8.521397331690551e-06,
+      "loss": 0.5885,
+      "step": 4277
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.664886897001362,
+      "learning_rate": 8.520661430113637e-06,
+      "loss": 0.6503,
+      "step": 4278
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3995094637873553,
+      "learning_rate": 8.519925377246057e-06,
+      "loss": 0.6935,
+      "step": 4279
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3307283737002022,
+      "learning_rate": 8.519189173119446e-06,
+      "loss": 0.7109,
+      "step": 4280
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3311938958476748,
+      "learning_rate": 8.518452817765438e-06,
+      "loss": 0.6731,
+      "step": 4281
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.326210237144696,
+      "learning_rate": 8.517716311215678e-06,
+      "loss": 0.6592,
+      "step": 4282
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2580398832050737,
+      "learning_rate": 8.516979653501813e-06,
+      "loss": 0.6561,
+      "step": 4283
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.206651322369004,
+      "learning_rate": 8.516242844655498e-06,
+      "loss": 0.5448,
+      "step": 4284
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4078818517761733,
+      "learning_rate": 8.515505884708399e-06,
+      "loss": 0.7031,
+      "step": 4285
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.212680104545325,
+      "learning_rate": 8.514768773692182e-06,
+      "loss": 0.5626,
+      "step": 4286
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3572098676723348,
+      "learning_rate": 8.514031511638524e-06,
+      "loss": 0.7497,
+      "step": 4287
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.409028470635537,
+      "learning_rate": 8.513294098579103e-06,
+      "loss": 0.6004,
+      "step": 4288
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1855058275508252,
+      "learning_rate": 8.512556534545612e-06,
+      "loss": 0.5546,
+      "step": 4289
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4210460657324393,
+      "learning_rate": 8.511818819569743e-06,
+      "loss": 0.6542,
+      "step": 4290
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2897962617217136,
+      "learning_rate": 8.5110809536832e-06,
+      "loss": 0.6769,
+      "step": 4291
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1894341374890267,
+      "learning_rate": 8.510342936917685e-06,
+      "loss": 0.6525,
+      "step": 4292
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3205490692090374,
+      "learning_rate": 8.50960476930492e-06,
+      "loss": 0.6628,
+      "step": 4293
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6890714655829442,
+      "learning_rate": 8.50886645087662e-06,
+      "loss": 0.7357,
+      "step": 4294
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9581814703868173,
+      "learning_rate": 8.508127981664514e-06,
+      "loss": 0.6339,
+      "step": 4295
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1486254871794204,
+      "learning_rate": 8.507389361700335e-06,
+      "loss": 0.6188,
+      "step": 4296
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.039330373610078,
+      "learning_rate": 8.506650591015825e-06,
+      "loss": 0.6751,
+      "step": 4297
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8094998576684136,
+      "learning_rate": 8.50591166964273e-06,
+      "loss": 0.8036,
+      "step": 4298
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3371916118390537,
+      "learning_rate": 8.5051725976128e-06,
+      "loss": 0.6493,
+      "step": 4299
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3909175437470571,
+      "learning_rate": 8.504433374957799e-06,
+      "loss": 0.7094,
+      "step": 4300
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4583359716129356,
+      "learning_rate": 8.50369400170949e-06,
+      "loss": 0.6951,
+      "step": 4301
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.225050571291483,
+      "learning_rate": 8.502954477899647e-06,
+      "loss": 0.718,
+      "step": 4302
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3648186167508853,
+      "learning_rate": 8.50221480356005e-06,
+      "loss": 0.7393,
+      "step": 4303
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5254672393260162,
+      "learning_rate": 8.50147497872248e-06,
+      "loss": 0.7011,
+      "step": 4304
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3167602677587056,
+      "learning_rate": 8.500735003418734e-06,
+      "loss": 0.69,
+      "step": 4305
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6533125211245632,
+      "learning_rate": 8.499994877680609e-06,
+      "loss": 0.662,
+      "step": 4306
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9868229600729054,
+      "learning_rate": 8.499254601539908e-06,
+      "loss": 0.6974,
+      "step": 4307
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0702046643567567,
+      "learning_rate": 8.498514175028442e-06,
+      "loss": 0.5802,
+      "step": 4308
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2050014928306427,
+      "learning_rate": 8.497773598178033e-06,
+      "loss": 0.6497,
+      "step": 4309
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2556813704534708,
+      "learning_rate": 8.497032871020501e-06,
+      "loss": 0.6069,
+      "step": 4310
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.343956854784069,
+      "learning_rate": 8.49629199358768e-06,
+      "loss": 0.7151,
+      "step": 4311
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2561835739233773,
+      "learning_rate": 8.495550965911403e-06,
+      "loss": 0.6396,
+      "step": 4312
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7334245379455864,
+      "learning_rate": 8.494809788023518e-06,
+      "loss": 0.7118,
+      "step": 4313
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2815992616901235,
+      "learning_rate": 8.494068459955871e-06,
+      "loss": 0.6038,
+      "step": 4314
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9735404998240998,
+      "learning_rate": 8.493326981740322e-06,
+      "loss": 0.6719,
+      "step": 4315
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4435181705216749,
+      "learning_rate": 8.492585353408732e-06,
+      "loss": 0.7598,
+      "step": 4316
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4215684600399718,
+      "learning_rate": 8.491843574992971e-06,
+      "loss": 0.6011,
+      "step": 4317
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1657418770927215,
+      "learning_rate": 8.491101646524916e-06,
+      "loss": 0.6126,
+      "step": 4318
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2509798366431855,
+      "learning_rate": 8.490359568036446e-06,
+      "loss": 0.7228,
+      "step": 4319
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.575603067255045,
+      "learning_rate": 8.489617339559455e-06,
+      "loss": 0.7575,
+      "step": 4320
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3143493750015462,
+      "learning_rate": 8.488874961125832e-06,
+      "loss": 0.6298,
+      "step": 4321
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4637893203101762,
+      "learning_rate": 8.488132432767483e-06,
+      "loss": 0.6992,
+      "step": 4322
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.352476592397541,
+      "learning_rate": 8.487389754516315e-06,
+      "loss": 0.6552,
+      "step": 4323
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2822842710592268,
+      "learning_rate": 8.486646926404243e-06,
+      "loss": 0.6646,
+      "step": 4324
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2809228441980989,
+      "learning_rate": 8.485903948463185e-06,
+      "loss": 0.6775,
+      "step": 4325
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.24517579443653,
+      "learning_rate": 8.485160820725073e-06,
+      "loss": 0.6777,
+      "step": 4326
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.377695197470166,
+      "learning_rate": 8.484417543221839e-06,
+      "loss": 0.637,
+      "step": 4327
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3833573899759286,
+      "learning_rate": 8.483674115985421e-06,
+      "loss": 0.7061,
+      "step": 4328
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2884331840900085,
+      "learning_rate": 8.48293053904777e-06,
+      "loss": 0.6733,
+      "step": 4329
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4724436677281454,
+      "learning_rate": 8.482186812440836e-06,
+      "loss": 0.7066,
+      "step": 4330
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4254924105696267,
+      "learning_rate": 8.481442936196578e-06,
+      "loss": 0.6505,
+      "step": 4331
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3766027178369529,
+      "learning_rate": 8.480698910346965e-06,
+      "loss": 0.5859,
+      "step": 4332
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5171684445153217,
+      "learning_rate": 8.479954734923967e-06,
+      "loss": 0.7299,
+      "step": 4333
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.308893265113104,
+      "learning_rate": 8.479210409959565e-06,
+      "loss": 0.6947,
+      "step": 4334
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.49212166910339,
+      "learning_rate": 8.478465935485741e-06,
+      "loss": 0.8108,
+      "step": 4335
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9432360331622553,
+      "learning_rate": 8.47772131153449e-06,
+      "loss": 0.6353,
+      "step": 4336
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4612282022923757,
+      "learning_rate": 8.476976538137809e-06,
+      "loss": 0.6921,
+      "step": 4337
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.464010722481263,
+      "learning_rate": 8.476231615327703e-06,
+      "loss": 0.6669,
+      "step": 4338
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2131088770967622,
+      "learning_rate": 8.475486543136181e-06,
+      "loss": 0.6232,
+      "step": 4339
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3269057566550946,
+      "learning_rate": 8.474741321595263e-06,
+      "loss": 0.6479,
+      "step": 4340
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3056913143640243,
+      "learning_rate": 8.473995950736973e-06,
+      "loss": 0.6852,
+      "step": 4341
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.35755453072713,
+      "learning_rate": 8.473250430593338e-06,
+      "loss": 0.7186,
+      "step": 4342
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2843347649185644,
+      "learning_rate": 8.472504761196397e-06,
+      "loss": 0.6539,
+      "step": 4343
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4362104345427948,
+      "learning_rate": 8.471758942578193e-06,
+      "loss": 0.7888,
+      "step": 4344
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.463413923118724,
+      "learning_rate": 8.471012974770776e-06,
+      "loss": 0.7083,
+      "step": 4345
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.472810493995596,
+      "learning_rate": 8.4702668578062e-06,
+      "loss": 0.6798,
+      "step": 4346
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3904191405621795,
+      "learning_rate": 8.46952059171653e-06,
+      "loss": 0.7187,
+      "step": 4347
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2736053114522516,
+      "learning_rate": 8.468774176533831e-06,
+      "loss": 0.6519,
+      "step": 4348
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4342003224061746,
+      "learning_rate": 8.46802761229018e-06,
+      "loss": 0.6487,
+      "step": 4349
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2935200944818244,
+      "learning_rate": 8.46728089901766e-06,
+      "loss": 0.7198,
+      "step": 4350
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4305145911917785,
+      "learning_rate": 8.466534036748359e-06,
+      "loss": 0.6805,
+      "step": 4351
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4264603320396327,
+      "learning_rate": 8.465787025514368e-06,
+      "loss": 0.6787,
+      "step": 4352
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9224918831744744,
+      "learning_rate": 8.465039865347791e-06,
+      "loss": 0.9568,
+      "step": 4353
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3741487279243405,
+      "learning_rate": 8.464292556280734e-06,
+      "loss": 0.7897,
+      "step": 4354
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2740754497563638,
+      "learning_rate": 8.463545098345308e-06,
+      "loss": 0.7382,
+      "step": 4355
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4580246676886366,
+      "learning_rate": 8.462797491573638e-06,
+      "loss": 0.6884,
+      "step": 4356
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1538810419415753,
+      "learning_rate": 8.462049735997848e-06,
+      "loss": 0.7206,
+      "step": 4357
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3264431319918768,
+      "learning_rate": 8.46130183165007e-06,
+      "loss": 0.6462,
+      "step": 4358
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3716323427760737,
+      "learning_rate": 8.46055377856244e-06,
+      "loss": 0.7893,
+      "step": 4359
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1398192078415474,
+      "learning_rate": 8.459805576767111e-06,
+      "loss": 0.5987,
+      "step": 4360
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4681605190385065,
+      "learning_rate": 8.459057226296232e-06,
+      "loss": 0.7226,
+      "step": 4361
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2554084642274268,
+      "learning_rate": 8.458308727181956e-06,
+      "loss": 0.6269,
+      "step": 4362
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2447310126462883,
+      "learning_rate": 8.457560079456455e-06,
+      "loss": 0.6176,
+      "step": 4363
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2497692278405528,
+      "learning_rate": 8.456811283151896e-06,
+      "loss": 0.6673,
+      "step": 4364
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1315173289553493,
+      "learning_rate": 8.456062338300458e-06,
+      "loss": 0.642,
+      "step": 4365
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9879165855001202,
+      "learning_rate": 8.455313244934324e-06,
+      "loss": 0.6816,
+      "step": 4366
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9786155802819638,
+      "learning_rate": 8.454564003085685e-06,
+      "loss": 0.7811,
+      "step": 4367
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2829836260918046,
+      "learning_rate": 8.453814612786736e-06,
+      "loss": 0.7404,
+      "step": 4368
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.365182504843494,
+      "learning_rate": 8.453065074069682e-06,
+      "loss": 0.7596,
+      "step": 4369
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.446460206347543,
+      "learning_rate": 8.45231538696673e-06,
+      "loss": 0.6245,
+      "step": 4370
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.32967942190852,
+      "learning_rate": 8.451565551510097e-06,
+      "loss": 0.6944,
+      "step": 4371
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8868533660923652,
+      "learning_rate": 8.450815567732007e-06,
+      "loss": 0.6822,
+      "step": 4372
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.213902179353957,
+      "learning_rate": 8.450065435664686e-06,
+      "loss": 0.6648,
+      "step": 4373
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2392409146665975,
+      "learning_rate": 8.449315155340369e-06,
+      "loss": 0.6788,
+      "step": 4374
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4594266100890503,
+      "learning_rate": 8.4485647267913e-06,
+      "loss": 0.7216,
+      "step": 4375
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3536610932854158,
+      "learning_rate": 8.44781415004972e-06,
+      "loss": 0.7051,
+      "step": 4376
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2423512751367787,
+      "learning_rate": 8.447063425147891e-06,
+      "loss": 0.7107,
+      "step": 4377
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4751878083925796,
+      "learning_rate": 8.446312552118068e-06,
+      "loss": 0.6466,
+      "step": 4378
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3340027637316472,
+      "learning_rate": 8.44556153099252e-06,
+      "loss": 0.6532,
+      "step": 4379
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.513824775104498,
+      "learning_rate": 8.44481036180352e-06,
+      "loss": 0.7006,
+      "step": 4380
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2850206764759426,
+      "learning_rate": 8.444059044583344e-06,
+      "loss": 0.7221,
+      "step": 4381
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1730139428237456,
+      "learning_rate": 8.443307579364282e-06,
+      "loss": 0.5962,
+      "step": 4382
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2732246874514075,
+      "learning_rate": 8.442555966178626e-06,
+      "loss": 0.6545,
+      "step": 4383
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3372979361626443,
+      "learning_rate": 8.441804205058672e-06,
+      "loss": 0.5245,
+      "step": 4384
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3781423032354934,
+      "learning_rate": 8.441052296036724e-06,
+      "loss": 0.7387,
+      "step": 4385
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3523717925830898,
+      "learning_rate": 8.440300239145098e-06,
+      "loss": 0.6577,
+      "step": 4386
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2595552962169192,
+      "learning_rate": 8.439548034416108e-06,
+      "loss": 0.5923,
+      "step": 4387
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4182689058520017,
+      "learning_rate": 8.43879568188208e-06,
+      "loss": 0.6549,
+      "step": 4388
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3437217806150024,
+      "learning_rate": 8.43804318157534e-06,
+      "loss": 0.6923,
+      "step": 4389
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3252694439145427,
+      "learning_rate": 8.437290533528231e-06,
+      "loss": 0.6366,
+      "step": 4390
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.461332803657591,
+      "learning_rate": 8.436537737773089e-06,
+      "loss": 0.7023,
+      "step": 4391
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0097831369067007,
+      "learning_rate": 8.43578479434227e-06,
+      "loss": 0.6903,
+      "step": 4392
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.372807784232749,
+      "learning_rate": 8.435031703268123e-06,
+      "loss": 0.6053,
+      "step": 4393
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2385486667200885,
+      "learning_rate": 8.434278464583018e-06,
+      "loss": 0.5995,
+      "step": 4394
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1984863520582638,
+      "learning_rate": 8.433525078319316e-06,
+      "loss": 0.6404,
+      "step": 4395
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3273657714078586,
+      "learning_rate": 8.432771544509395e-06,
+      "loss": 0.706,
+      "step": 4396
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3641077119349008,
+      "learning_rate": 8.432017863185635e-06,
+      "loss": 0.7431,
+      "step": 4397
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.557997356673863,
+      "learning_rate": 8.431264034380424e-06,
+      "loss": 0.7484,
+      "step": 4398
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1641349613852412,
+      "learning_rate": 8.430510058126156e-06,
+      "loss": 0.6016,
+      "step": 4399
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1955630334315956,
+      "learning_rate": 8.42975593445523e-06,
+      "loss": 0.6705,
+      "step": 4400
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3248965313420984,
+      "learning_rate": 8.429001663400054e-06,
+      "loss": 0.6725,
+      "step": 4401
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2413704491413458,
+      "learning_rate": 8.428247244993038e-06,
+      "loss": 0.7309,
+      "step": 4402
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3034351703506641,
+      "learning_rate": 8.427492679266605e-06,
+      "loss": 0.6727,
+      "step": 4403
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4132571702905012,
+      "learning_rate": 8.426737966253176e-06,
+      "loss": 0.6866,
+      "step": 4404
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2859317726686457,
+      "learning_rate": 8.425983105985188e-06,
+      "loss": 0.7111,
+      "step": 4405
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4300438088594811,
+      "learning_rate": 8.425228098495073e-06,
+      "loss": 0.7741,
+      "step": 4406
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1946230179952222,
+      "learning_rate": 8.424472943815278e-06,
+      "loss": 0.6524,
+      "step": 4407
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3059384331396595,
+      "learning_rate": 8.423717641978255e-06,
+      "loss": 0.7207,
+      "step": 4408
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2250124362371408,
+      "learning_rate": 8.422962193016459e-06,
+      "loss": 0.6863,
+      "step": 4409
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1344573503712223,
+      "learning_rate": 8.422206596962357e-06,
+      "loss": 0.5811,
+      "step": 4410
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3500575298385102,
+      "learning_rate": 8.421450853848414e-06,
+      "loss": 0.6522,
+      "step": 4411
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.471815073167405,
+      "learning_rate": 8.420694963707108e-06,
+      "loss": 0.7481,
+      "step": 4412
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1882977797886392,
+      "learning_rate": 8.419938926570921e-06,
+      "loss": 0.6006,
+      "step": 4413
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3987121866979122,
+      "learning_rate": 8.41918274247234e-06,
+      "loss": 0.7554,
+      "step": 4414
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2622127996688752,
+      "learning_rate": 8.418426411443864e-06,
+      "loss": 0.5818,
+      "step": 4415
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3588567291792188,
+      "learning_rate": 8.41766993351799e-06,
+      "loss": 0.6624,
+      "step": 4416
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.277540468272105,
+      "learning_rate": 8.416913308727229e-06,
+      "loss": 0.6509,
+      "step": 4417
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2597397271869022,
+      "learning_rate": 8.416156537104092e-06,
+      "loss": 0.614,
+      "step": 4418
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3647586493733277,
+      "learning_rate": 8.415399618681101e-06,
+      "loss": 0.6337,
+      "step": 4419
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4392153816471878,
+      "learning_rate": 8.414642553490783e-06,
+      "loss": 0.6471,
+      "step": 4420
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1941037326971267,
+      "learning_rate": 8.413885341565668e-06,
+      "loss": 0.6533,
+      "step": 4421
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3696239721031462,
+      "learning_rate": 8.413127982938299e-06,
+      "loss": 0.6984,
+      "step": 4422
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3162875261282792,
+      "learning_rate": 8.412370477641215e-06,
+      "loss": 0.671,
+      "step": 4423
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2857331746729854,
+      "learning_rate": 8.411612825706976e-06,
+      "loss": 0.6804,
+      "step": 4424
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2074401532806711,
+      "learning_rate": 8.410855027168134e-06,
+      "loss": 0.6864,
+      "step": 4425
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.48128738651036,
+      "learning_rate": 8.410097082057256e-06,
+      "loss": 0.7628,
+      "step": 4426
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4576290927258735,
+      "learning_rate": 8.40933899040691e-06,
+      "loss": 0.7125,
+      "step": 4427
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.44941886389722,
+      "learning_rate": 8.408580752249676e-06,
+      "loss": 0.6875,
+      "step": 4428
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6744500068747386,
+      "learning_rate": 8.407822367618135e-06,
+      "loss": 0.8753,
+      "step": 4429
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2269886678184718,
+      "learning_rate": 8.407063836544877e-06,
+      "loss": 0.7457,
+      "step": 4430
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2016937772224918,
+      "learning_rate": 8.4063051590625e-06,
+      "loss": 0.5656,
+      "step": 4431
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4136240512620295,
+      "learning_rate": 8.405546335203602e-06,
+      "loss": 0.738,
+      "step": 4432
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1547353394432693,
+      "learning_rate": 8.404787365000796e-06,
+      "loss": 0.6603,
+      "step": 4433
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5679300462571941,
+      "learning_rate": 8.40402824848669e-06,
+      "loss": 0.831,
+      "step": 4434
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3146945701060258,
+      "learning_rate": 8.403268985693912e-06,
+      "loss": 0.6369,
+      "step": 4435
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3041530417064047,
+      "learning_rate": 8.402509576655085e-06,
+      "loss": 0.6793,
+      "step": 4436
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.250569230829728,
+      "learning_rate": 8.401750021402843e-06,
+      "loss": 0.6519,
+      "step": 4437
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3655051937283456,
+      "learning_rate": 8.400990319969829e-06,
+      "loss": 0.7092,
+      "step": 4438
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1991948737452642,
+      "learning_rate": 8.400230472388684e-06,
+      "loss": 0.5519,
+      "step": 4439
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3791930854727212,
+      "learning_rate": 8.399470478692064e-06,
+      "loss": 0.6882,
+      "step": 4440
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0856326073515978,
+      "learning_rate": 8.398710338912626e-06,
+      "loss": 0.7427,
+      "step": 4441
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1006642718230688,
+      "learning_rate": 8.397950053083036e-06,
+      "loss": 0.7243,
+      "step": 4442
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.27061979886939,
+      "learning_rate": 8.397189621235964e-06,
+      "loss": 0.613,
+      "step": 4443
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3197130919548186,
+      "learning_rate": 8.396429043404088e-06,
+      "loss": 0.6969,
+      "step": 4444
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.217627196516449,
+      "learning_rate": 8.395668319620092e-06,
+      "loss": 0.6831,
+      "step": 4445
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4738764144888534,
+      "learning_rate": 8.394907449916667e-06,
+      "loss": 0.6744,
+      "step": 4446
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.236714386266411,
+      "learning_rate": 8.394146434326507e-06,
+      "loss": 0.7842,
+      "step": 4447
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.483474300854947,
+      "learning_rate": 8.393385272882315e-06,
+      "loss": 0.6503,
+      "step": 4448
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3718469529658548,
+      "learning_rate": 8.3926239656168e-06,
+      "loss": 0.6794,
+      "step": 4449
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4132384845602792,
+      "learning_rate": 8.391862512562679e-06,
+      "loss": 0.749,
+      "step": 4450
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3135982585590944,
+      "learning_rate": 8.39110091375267e-06,
+      "loss": 0.6526,
+      "step": 4451
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2616823925768716,
+      "learning_rate": 8.390339169219504e-06,
+      "loss": 0.728,
+      "step": 4452
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5539457597434596,
+      "learning_rate": 8.389577278995913e-06,
+      "loss": 0.6883,
+      "step": 4453
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8121504686576093,
+      "learning_rate": 8.388815243114637e-06,
+      "loss": 0.5622,
+      "step": 4454
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4558720712327498,
+      "learning_rate": 8.388053061608421e-06,
+      "loss": 0.7788,
+      "step": 4455
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.00653124396657,
+      "learning_rate": 8.387290734510022e-06,
+      "loss": 0.644,
+      "step": 4456
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3283395592514582,
+      "learning_rate": 8.386528261852196e-06,
+      "loss": 0.7363,
+      "step": 4457
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.497846266483922,
+      "learning_rate": 8.385765643667707e-06,
+      "loss": 0.8032,
+      "step": 4458
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1790749470279138,
+      "learning_rate": 8.385002879989328e-06,
+      "loss": 0.6406,
+      "step": 4459
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2774577554004858,
+      "learning_rate": 8.384239970849837e-06,
+      "loss": 0.654,
+      "step": 4460
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2189945716876898,
+      "learning_rate": 8.383476916282015e-06,
+      "loss": 0.5979,
+      "step": 4461
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3681939119817763,
+      "learning_rate": 8.382713716318658e-06,
+      "loss": 0.7731,
+      "step": 4462
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3693889346397177,
+      "learning_rate": 8.381950370992557e-06,
+      "loss": 0.6881,
+      "step": 4463
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3930573355039002,
+      "learning_rate": 8.381186880336518e-06,
+      "loss": 0.7318,
+      "step": 4464
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6922374206077655,
+      "learning_rate": 8.380423244383347e-06,
+      "loss": 0.6142,
+      "step": 4465
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3170406267943884,
+      "learning_rate": 8.37965946316586e-06,
+      "loss": 0.6395,
+      "step": 4466
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4134354371831757,
+      "learning_rate": 8.378895536716882e-06,
+      "loss": 0.6759,
+      "step": 4467
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4951496888442397,
+      "learning_rate": 8.378131465069235e-06,
+      "loss": 0.7277,
+      "step": 4468
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3048825869615335,
+      "learning_rate": 8.377367248255757e-06,
+      "loss": 0.5501,
+      "step": 4469
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3673386947315433,
+      "learning_rate": 8.376602886309285e-06,
+      "loss": 0.6808,
+      "step": 4470
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5529352238319758,
+      "learning_rate": 8.375838379262667e-06,
+      "loss": 0.5847,
+      "step": 4471
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.353894250351882,
+      "learning_rate": 8.375073727148756e-06,
+      "loss": 0.5952,
+      "step": 4472
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1703260854928867,
+      "learning_rate": 8.37430893000041e-06,
+      "loss": 0.6404,
+      "step": 4473
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3835764556331196,
+      "learning_rate": 8.373543987850494e-06,
+      "loss": 0.7547,
+      "step": 4474
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2598149007365986,
+      "learning_rate": 8.372778900731882e-06,
+      "loss": 0.5767,
+      "step": 4475
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3545026698687872,
+      "learning_rate": 8.372013668677446e-06,
+      "loss": 0.7629,
+      "step": 4476
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2589174232095313,
+      "learning_rate": 8.371248291720073e-06,
+      "loss": 0.6523,
+      "step": 4477
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.283837900380396,
+      "learning_rate": 8.370482769892654e-06,
+      "loss": 0.6572,
+      "step": 4478
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4322209439721065,
+      "learning_rate": 8.369717103228084e-06,
+      "loss": 0.6918,
+      "step": 4479
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2778014117248953,
+      "learning_rate": 8.368951291759264e-06,
+      "loss": 0.7221,
+      "step": 4480
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3112306016751059,
+      "learning_rate": 8.368185335519106e-06,
+      "loss": 0.6374,
+      "step": 4481
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.63270222772384,
+      "learning_rate": 8.367419234540522e-06,
+      "loss": 0.7357,
+      "step": 4482
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2734284912932328,
+      "learning_rate": 8.366652988856432e-06,
+      "loss": 0.6946,
+      "step": 4483
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0033104863347604,
+      "learning_rate": 8.365886598499766e-06,
+      "loss": 0.6672,
+      "step": 4484
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2752111718703203,
+      "learning_rate": 8.365120063503458e-06,
+      "loss": 0.663,
+      "step": 4485
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.312493265338354,
+      "learning_rate": 8.364353383900445e-06,
+      "loss": 0.6912,
+      "step": 4486
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3408610923699191,
+      "learning_rate": 8.363586559723675e-06,
+      "loss": 0.6537,
+      "step": 4487
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.324031227434583,
+      "learning_rate": 8.3628195910061e-06,
+      "loss": 0.7068,
+      "step": 4488
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4600480892986867,
+      "learning_rate": 8.362052477780677e-06,
+      "loss": 0.7894,
+      "step": 4489
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.12841708092308,
+      "learning_rate": 8.36128522008037e-06,
+      "loss": 0.5611,
+      "step": 4490
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3353838015285486,
+      "learning_rate": 8.360517817938154e-06,
+      "loss": 0.6812,
+      "step": 4491
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.45277100260119,
+      "learning_rate": 8.359750271386999e-06,
+      "loss": 0.6467,
+      "step": 4492
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2533637970753981,
+      "learning_rate": 8.358982580459896e-06,
+      "loss": 0.6368,
+      "step": 4493
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3867083205250301,
+      "learning_rate": 8.35821474518983e-06,
+      "loss": 0.7214,
+      "step": 4494
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2799210863178063,
+      "learning_rate": 8.357446765609796e-06,
+      "loss": 0.6222,
+      "step": 4495
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2635731067496694,
+      "learning_rate": 8.356678641752797e-06,
+      "loss": 0.5678,
+      "step": 4496
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3224067289836392,
+      "learning_rate": 8.355910373651844e-06,
+      "loss": 0.6914,
+      "step": 4497
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.310320050778353,
+      "learning_rate": 8.355141961339945e-06,
+      "loss": 0.6905,
+      "step": 4498
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1929827657668537,
+      "learning_rate": 8.354373404850124e-06,
+      "loss": 0.6566,
+      "step": 4499
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7383418760660427,
+      "learning_rate": 8.353604704215408e-06,
+      "loss": 0.6744,
+      "step": 4500
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4535192556244771,
+      "learning_rate": 8.352835859468829e-06,
+      "loss": 0.6669,
+      "step": 4501
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1202821627367983,
+      "learning_rate": 8.352066870643424e-06,
+      "loss": 0.5446,
+      "step": 4502
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6788908195577128,
+      "learning_rate": 8.351297737772244e-06,
+      "loss": 0.7887,
+      "step": 4503
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3781020127224657,
+      "learning_rate": 8.350528460888334e-06,
+      "loss": 0.6685,
+      "step": 4504
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.436402583258939,
+      "learning_rate": 8.349759040024753e-06,
+      "loss": 0.778,
+      "step": 4505
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.6379673867512463,
+      "learning_rate": 8.348989475214568e-06,
+      "loss": 0.6892,
+      "step": 4506
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.362218715286415,
+      "learning_rate": 8.348219766490845e-06,
+      "loss": 0.6637,
+      "step": 4507
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4372453836028631,
+      "learning_rate": 8.347449913886662e-06,
+      "loss": 0.6815,
+      "step": 4508
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.464981446863599,
+      "learning_rate": 8.346679917435104e-06,
+      "loss": 1.0281,
+      "step": 4509
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.543575205995568,
+      "learning_rate": 8.345909777169252e-06,
+      "loss": 0.7765,
+      "step": 4510
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4833059219286007,
+      "learning_rate": 8.345139493122208e-06,
+      "loss": 0.6541,
+      "step": 4511
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.135650588432162,
+      "learning_rate": 8.34436906532707e-06,
+      "loss": 0.5832,
+      "step": 4512
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.159644218605556,
+      "learning_rate": 8.343598493816944e-06,
+      "loss": 0.6427,
+      "step": 4513
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2231235003277914,
+      "learning_rate": 8.342827778624943e-06,
+      "loss": 0.626,
+      "step": 4514
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3834098960317631,
+      "learning_rate": 8.34205691978419e-06,
+      "loss": 0.7009,
+      "step": 4515
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2454721775092734,
+      "learning_rate": 8.341285917327807e-06,
+      "loss": 0.5826,
+      "step": 4516
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4835851238480755,
+      "learning_rate": 8.340514771288926e-06,
+      "loss": 0.65,
+      "step": 4517
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2743743286591265,
+      "learning_rate": 8.339743481700685e-06,
+      "loss": 0.619,
+      "step": 4518
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.618043183667192,
+      "learning_rate": 8.33897204859623e-06,
+      "loss": 0.7652,
+      "step": 4519
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5462754286163884,
+      "learning_rate": 8.338200472008708e-06,
+      "loss": 0.7586,
+      "step": 4520
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.092443614115244,
+      "learning_rate": 8.337428751971279e-06,
+      "loss": 0.535,
+      "step": 4521
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2115888038682459,
+      "learning_rate": 8.336656888517103e-06,
+      "loss": 0.6462,
+      "step": 4522
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3090350445805312,
+      "learning_rate": 8.33588488167935e-06,
+      "loss": 0.683,
+      "step": 4523
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.276911511045285,
+      "learning_rate": 8.335112731491192e-06,
+      "loss": 0.6194,
+      "step": 4524
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2748210075229895,
+      "learning_rate": 8.334340437985814e-06,
+      "loss": 0.6381,
+      "step": 4525
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2808211649336372,
+      "learning_rate": 8.333568001196402e-06,
+      "loss": 0.727,
+      "step": 4526
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2948384890072395,
+      "learning_rate": 8.332795421156147e-06,
+      "loss": 0.5857,
+      "step": 4527
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1367901335809154,
+      "learning_rate": 8.332022697898253e-06,
+      "loss": 0.6185,
+      "step": 4528
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4206721338646855,
+      "learning_rate": 8.331249831455921e-06,
+      "loss": 0.7459,
+      "step": 4529
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3017372282698727,
+      "learning_rate": 8.330476821862366e-06,
+      "loss": 0.6316,
+      "step": 4530
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3019507983618057,
+      "learning_rate": 8.329703669150801e-06,
+      "loss": 0.6477,
+      "step": 4531
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.484938484806245,
+      "learning_rate": 8.328930373354457e-06,
+      "loss": 0.762,
+      "step": 4532
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0412464115149358,
+      "learning_rate": 8.328156934506559e-06,
+      "loss": 0.5588,
+      "step": 4533
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.231112678374453,
+      "learning_rate": 8.327383352640347e-06,
+      "loss": 0.6746,
+      "step": 4534
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3587402006598301,
+      "learning_rate": 8.32660962778906e-06,
+      "loss": 0.7191,
+      "step": 4535
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3968549596465567,
+      "learning_rate": 8.325835759985951e-06,
+      "loss": 0.7462,
+      "step": 4536
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.318921267527922,
+      "learning_rate": 8.32506174926427e-06,
+      "loss": 0.7006,
+      "step": 4537
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.226388989660798,
+      "learning_rate": 8.324287595657284e-06,
+      "loss": 0.7545,
+      "step": 4538
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3733325203604891,
+      "learning_rate": 8.323513299198252e-06,
+      "loss": 0.6752,
+      "step": 4539
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.328944149001457,
+      "learning_rate": 8.322738859920453e-06,
+      "loss": 0.7141,
+      "step": 4540
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6082597927916316,
+      "learning_rate": 8.321964277857167e-06,
+      "loss": 0.6841,
+      "step": 4541
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4387552694371475,
+      "learning_rate": 8.321189553041675e-06,
+      "loss": 0.7497,
+      "step": 4542
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4110129566602347,
+      "learning_rate": 8.320414685507272e-06,
+      "loss": 0.7021,
+      "step": 4543
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3448001955155973,
+      "learning_rate": 8.319639675287255e-06,
+      "loss": 0.6929,
+      "step": 4544
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1482987867609615,
+      "learning_rate": 8.318864522414928e-06,
+      "loss": 0.599,
+      "step": 4545
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5254557082259181,
+      "learning_rate": 8.318089226923602e-06,
+      "loss": 0.6878,
+      "step": 4546
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4538895523782156,
+      "learning_rate": 8.317313788846591e-06,
+      "loss": 0.6712,
+      "step": 4547
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9940967341319739,
+      "learning_rate": 8.31653820821722e-06,
+      "loss": 0.636,
+      "step": 4548
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3239019468280318,
+      "learning_rate": 8.315762485068815e-06,
+      "loss": 0.7292,
+      "step": 4549
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5389311560962922,
+      "learning_rate": 8.31498661943471e-06,
+      "loss": 0.7453,
+      "step": 4550
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2433389164984945,
+      "learning_rate": 8.314210611348249e-06,
+      "loss": 0.6118,
+      "step": 4551
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4666984867320525,
+      "learning_rate": 8.313434460842775e-06,
+      "loss": 0.701,
+      "step": 4552
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.604306239922433,
+      "learning_rate": 8.312658167951644e-06,
+      "loss": 0.6691,
+      "step": 4553
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4071692302945669,
+      "learning_rate": 8.311881732708213e-06,
+      "loss": 0.7473,
+      "step": 4554
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4379944675821117,
+      "learning_rate": 8.311105155145849e-06,
+      "loss": 0.6696,
+      "step": 4555
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7932307799737057,
+      "learning_rate": 8.310328435297923e-06,
+      "loss": 0.6644,
+      "step": 4556
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4386048316197775,
+      "learning_rate": 8.309551573197809e-06,
+      "loss": 0.7172,
+      "step": 4557
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1004052203274088,
+      "learning_rate": 8.308774568878896e-06,
+      "loss": 0.5601,
+      "step": 4558
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3313520411402173,
+      "learning_rate": 8.307997422374569e-06,
+      "loss": 0.6424,
+      "step": 4559
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4409624234630134,
+      "learning_rate": 8.307220133718225e-06,
+      "loss": 0.7424,
+      "step": 4560
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4752848852163647,
+      "learning_rate": 8.306442702943268e-06,
+      "loss": 0.7508,
+      "step": 4561
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4277800946862336,
+      "learning_rate": 8.305665130083102e-06,
+      "loss": 0.7434,
+      "step": 4562
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2568917430162587,
+      "learning_rate": 8.304887415171146e-06,
+      "loss": 0.6668,
+      "step": 4563
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3544651260670677,
+      "learning_rate": 8.304109558240817e-06,
+      "loss": 0.7409,
+      "step": 4564
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4395002334291405,
+      "learning_rate": 8.30333155932554e-06,
+      "loss": 0.7291,
+      "step": 4565
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.41253762950922,
+      "learning_rate": 8.302553418458749e-06,
+      "loss": 0.7411,
+      "step": 4566
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0743798419675272,
+      "learning_rate": 8.301775135673884e-06,
+      "loss": 0.7052,
+      "step": 4567
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.392386463034661,
+      "learning_rate": 8.300996711004387e-06,
+      "loss": 0.6614,
+      "step": 4568
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.306745574756657,
+      "learning_rate": 8.300218144483709e-06,
+      "loss": 0.641,
+      "step": 4569
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0686644421719373,
+      "learning_rate": 8.299439436145309e-06,
+      "loss": 0.5289,
+      "step": 4570
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5215469925351464,
+      "learning_rate": 8.298660586022646e-06,
+      "loss": 0.6986,
+      "step": 4571
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.302162999173897,
+      "learning_rate": 8.297881594149193e-06,
+      "loss": 0.6993,
+      "step": 4572
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.29744112411865,
+      "learning_rate": 8.297102460558421e-06,
+      "loss": 0.6711,
+      "step": 4573
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2997408724007078,
+      "learning_rate": 8.296323185283816e-06,
+      "loss": 0.7566,
+      "step": 4574
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4063260299760016,
+      "learning_rate": 8.29554376835886e-06,
+      "loss": 0.6644,
+      "step": 4575
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3629666976483314,
+      "learning_rate": 8.29476420981705e-06,
+      "loss": 0.8257,
+      "step": 4576
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1226934857956383,
+      "learning_rate": 8.293984509691885e-06,
+      "loss": 0.7398,
+      "step": 4577
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1463557816679897,
+      "learning_rate": 8.293204668016867e-06,
+      "loss": 0.6624,
+      "step": 4578
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.252234299410147,
+      "learning_rate": 8.292424684825514e-06,
+      "loss": 0.6164,
+      "step": 4579
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4076334124441938,
+      "learning_rate": 8.291644560151335e-06,
+      "loss": 0.7523,
+      "step": 4580
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2600557770445304,
+      "learning_rate": 8.29086429402786e-06,
+      "loss": 0.7063,
+      "step": 4581
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3525155712334866,
+      "learning_rate": 8.290083886488618e-06,
+      "loss": 0.7058,
+      "step": 4582
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1962912554633431,
+      "learning_rate": 8.289303337567145e-06,
+      "loss": 0.5888,
+      "step": 4583
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4138577679451378,
+      "learning_rate": 8.28852264729698e-06,
+      "loss": 0.75,
+      "step": 4584
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4941507987807878,
+      "learning_rate": 8.287741815711674e-06,
+      "loss": 0.7071,
+      "step": 4585
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2523722871253657,
+      "learning_rate": 8.286960842844779e-06,
+      "loss": 0.6462,
+      "step": 4586
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3891948726110785,
+      "learning_rate": 8.286179728729856e-06,
+      "loss": 0.6234,
+      "step": 4587
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1369623192022718,
+      "learning_rate": 8.285398473400471e-06,
+      "loss": 0.5696,
+      "step": 4588
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4204519459762666,
+      "learning_rate": 8.284617076890199e-06,
+      "loss": 0.6328,
+      "step": 4589
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2681182824881685,
+      "learning_rate": 8.283835539232614e-06,
+      "loss": 0.6166,
+      "step": 4590
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3497587618738516,
+      "learning_rate": 8.283053860461303e-06,
+      "loss": 0.5694,
+      "step": 4591
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3830280904275576,
+      "learning_rate": 8.282272040609855e-06,
+      "loss": 0.6753,
+      "step": 4592
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3213826151914985,
+      "learning_rate": 8.281490079711866e-06,
+      "loss": 0.6407,
+      "step": 4593
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1361207553017305,
+      "learning_rate": 8.280707977800944e-06,
+      "loss": 0.4494,
+      "step": 4594
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.477178262173983,
+      "learning_rate": 8.279925734910691e-06,
+      "loss": 0.7249,
+      "step": 4595
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3795215390022162,
+      "learning_rate": 8.279143351074726e-06,
+      "loss": 0.7069,
+      "step": 4596
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.063886467811708,
+      "learning_rate": 8.278360826326669e-06,
+      "loss": 0.6944,
+      "step": 4597
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4548188161065878,
+      "learning_rate": 8.277578160700143e-06,
+      "loss": 0.5913,
+      "step": 4598
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5713121837668866,
+      "learning_rate": 8.276795354228785e-06,
+      "loss": 0.6319,
+      "step": 4599
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.341820666450519,
+      "learning_rate": 8.276012406946232e-06,
+      "loss": 0.7357,
+      "step": 4600
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6367108778091382,
+      "learning_rate": 8.275229318886132e-06,
+      "loss": 0.7149,
+      "step": 4601
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.734360411955157,
+      "learning_rate": 8.274446090082133e-06,
+      "loss": 0.6424,
+      "step": 4602
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3165411625929353,
+      "learning_rate": 8.273662720567892e-06,
+      "loss": 0.6197,
+      "step": 4603
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3862541044284555,
+      "learning_rate": 8.272879210377074e-06,
+      "loss": 0.6239,
+      "step": 4604
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5426648907762335,
+      "learning_rate": 8.272095559543346e-06,
+      "loss": 0.697,
+      "step": 4605
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9399238176292863,
+      "learning_rate": 8.271311768100386e-06,
+      "loss": 0.7044,
+      "step": 4606
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4143088840874771,
+      "learning_rate": 8.270527836081872e-06,
+      "loss": 0.6686,
+      "step": 4607
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1346976113234573,
+      "learning_rate": 8.269743763521495e-06,
+      "loss": 0.5708,
+      "step": 4608
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2600430194076644,
+      "learning_rate": 8.268959550452946e-06,
+      "loss": 0.771,
+      "step": 4609
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4333518289105542,
+      "learning_rate": 8.268175196909924e-06,
+      "loss": 0.7085,
+      "step": 4610
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4553596141344356,
+      "learning_rate": 8.267390702926135e-06,
+      "loss": 0.7609,
+      "step": 4611
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2465404077586446,
+      "learning_rate": 8.26660606853529e-06,
+      "loss": 0.7067,
+      "step": 4612
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.273778008939505,
+      "learning_rate": 8.26582129377111e-06,
+      "loss": 0.6405,
+      "step": 4613
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3168738646175167,
+      "learning_rate": 8.265036378667312e-06,
+      "loss": 0.7741,
+      "step": 4614
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2671879734707965,
+      "learning_rate": 8.264251323257633e-06,
+      "loss": 0.7258,
+      "step": 4615
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.266473836333213,
+      "learning_rate": 8.263466127575801e-06,
+      "loss": 0.7225,
+      "step": 4616
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9922249114475197,
+      "learning_rate": 8.262680791655565e-06,
+      "loss": 0.6681,
+      "step": 4617
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3264960649440913,
+      "learning_rate": 8.261895315530667e-06,
+      "loss": 0.6977,
+      "step": 4618
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1966973262933267,
+      "learning_rate": 8.261109699234862e-06,
+      "loss": 0.6114,
+      "step": 4619
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2632388849781018,
+      "learning_rate": 8.26032394280191e-06,
+      "loss": 0.6989,
+      "step": 4620
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5100854122656597,
+      "learning_rate": 8.259538046265578e-06,
+      "loss": 0.7572,
+      "step": 4621
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1551553606841807,
+      "learning_rate": 8.258752009659638e-06,
+      "loss": 0.6148,
+      "step": 4622
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.280548682922284,
+      "learning_rate": 8.257965833017864e-06,
+      "loss": 0.6304,
+      "step": 4623
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2993909643220785,
+      "learning_rate": 8.257179516374045e-06,
+      "loss": 0.6237,
+      "step": 4624
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.309002036124426,
+      "learning_rate": 8.256393059761966e-06,
+      "loss": 0.6643,
+      "step": 4625
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4021836095775264,
+      "learning_rate": 8.255606463215426e-06,
+      "loss": 0.7427,
+      "step": 4626
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2846777357216783,
+      "learning_rate": 8.254819726768224e-06,
+      "loss": 0.5941,
+      "step": 4627
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3811901622897562,
+      "learning_rate": 8.254032850454173e-06,
+      "loss": 0.5912,
+      "step": 4628
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4059664833354395,
+      "learning_rate": 8.253245834307079e-06,
+      "loss": 0.6588,
+      "step": 4629
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0902371758556293,
+      "learning_rate": 8.252458678360769e-06,
+      "loss": 0.6709,
+      "step": 4630
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.218750330668825,
+      "learning_rate": 8.251671382649065e-06,
+      "loss": 0.6559,
+      "step": 4631
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2312577978493213,
+      "learning_rate": 8.250883947205799e-06,
+      "loss": 0.6199,
+      "step": 4632
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1163515747430757,
+      "learning_rate": 8.25009637206481e-06,
+      "loss": 0.6673,
+      "step": 4633
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2763856327375642,
+      "learning_rate": 8.249308657259943e-06,
+      "loss": 0.6405,
+      "step": 4634
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4151747530222707,
+      "learning_rate": 8.248520802825046e-06,
+      "loss": 0.7625,
+      "step": 4635
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.230078899779973,
+      "learning_rate": 8.247732808793975e-06,
+      "loss": 0.6577,
+      "step": 4636
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3575914834081642,
+      "learning_rate": 8.246944675200592e-06,
+      "loss": 0.668,
+      "step": 4637
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2363002232635287,
+      "learning_rate": 8.246156402078765e-06,
+      "loss": 0.6616,
+      "step": 4638
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3171954201796088,
+      "learning_rate": 8.245367989462368e-06,
+      "loss": 0.6443,
+      "step": 4639
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3710226095892082,
+      "learning_rate": 8.244579437385281e-06,
+      "loss": 0.7202,
+      "step": 4640
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3526284754948243,
+      "learning_rate": 8.243790745881389e-06,
+      "loss": 0.6004,
+      "step": 4641
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2241661594765194,
+      "learning_rate": 8.243001914984583e-06,
+      "loss": 0.6956,
+      "step": 4642
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.409083408755373,
+      "learning_rate": 8.242212944728762e-06,
+      "loss": 0.7992,
+      "step": 4643
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4698887065104427,
+      "learning_rate": 8.241423835147833e-06,
+      "loss": 0.7337,
+      "step": 4644
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4114747416981333,
+      "learning_rate": 8.240634586275701e-06,
+      "loss": 0.7337,
+      "step": 4645
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4407264795688486,
+      "learning_rate": 8.239845198146284e-06,
+      "loss": 0.7252,
+      "step": 4646
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3453432941031438,
+      "learning_rate": 8.239055670793503e-06,
+      "loss": 0.6803,
+      "step": 4647
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2020387488414102,
+      "learning_rate": 8.238266004251284e-06,
+      "loss": 0.6116,
+      "step": 4648
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3436846785193945,
+      "learning_rate": 8.237476198553567e-06,
+      "loss": 0.7602,
+      "step": 4649
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3402723770408385,
+      "learning_rate": 8.236686253734285e-06,
+      "loss": 0.7215,
+      "step": 4650
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0651121438887836,
+      "learning_rate": 8.235896169827386e-06,
+      "loss": 0.6048,
+      "step": 4651
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5284337734568454,
+      "learning_rate": 8.235105946866823e-06,
+      "loss": 0.6927,
+      "step": 4652
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.097681653281794,
+      "learning_rate": 8.234315584886551e-06,
+      "loss": 0.62,
+      "step": 4653
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4770471389562136,
+      "learning_rate": 8.233525083920536e-06,
+      "loss": 0.6812,
+      "step": 4654
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1523262665809388,
+      "learning_rate": 8.232734444002748e-06,
+      "loss": 0.6739,
+      "step": 4655
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4851337706303138,
+      "learning_rate": 8.23194366516716e-06,
+      "loss": 0.715,
+      "step": 4656
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5925324789334638,
+      "learning_rate": 8.231152747447753e-06,
+      "loss": 0.699,
+      "step": 4657
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4369848617617627,
+      "learning_rate": 8.230361690878519e-06,
+      "loss": 0.6546,
+      "step": 4658
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.600160448767559,
+      "learning_rate": 8.229570495493447e-06,
+      "loss": 0.6907,
+      "step": 4659
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4278561799861962,
+      "learning_rate": 8.22877916132654e-06,
+      "loss": 0.6956,
+      "step": 4660
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5135120334872059,
+      "learning_rate": 8.2279876884118e-06,
+      "loss": 0.6784,
+      "step": 4661
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.354680287745225,
+      "learning_rate": 8.22719607678324e-06,
+      "loss": 0.728,
+      "step": 4662
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2092648944091822,
+      "learning_rate": 8.226404326474878e-06,
+      "loss": 0.6169,
+      "step": 4663
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3716761959362689,
+      "learning_rate": 8.225612437520736e-06,
+      "loss": 0.696,
+      "step": 4664
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2841172271580417,
+      "learning_rate": 8.224820409954844e-06,
+      "loss": 0.6089,
+      "step": 4665
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3640222501169028,
+      "learning_rate": 8.224028243811237e-06,
+      "loss": 0.6192,
+      "step": 4666
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7648683227982307,
+      "learning_rate": 8.223235939123955e-06,
+      "loss": 0.7049,
+      "step": 4667
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2842527040966156,
+      "learning_rate": 8.222443495927049e-06,
+      "loss": 0.6157,
+      "step": 4668
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2340336306689816,
+      "learning_rate": 8.221650914254566e-06,
+      "loss": 0.7169,
+      "step": 4669
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2992010280293202,
+      "learning_rate": 8.220858194140568e-06,
+      "loss": 0.6951,
+      "step": 4670
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1728314053145785,
+      "learning_rate": 8.220065335619123e-06,
+      "loss": 0.5966,
+      "step": 4671
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.177508758805864,
+      "learning_rate": 8.219272338724297e-06,
+      "loss": 0.6317,
+      "step": 4672
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.143978739326065,
+      "learning_rate": 8.21847920349017e-06,
+      "loss": 0.686,
+      "step": 4673
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2092807838243635,
+      "learning_rate": 8.217685929950823e-06,
+      "loss": 0.6158,
+      "step": 4674
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3298194381406103,
+      "learning_rate": 8.216892518140346e-06,
+      "loss": 0.7606,
+      "step": 4675
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5663576574986966,
+      "learning_rate": 8.216098968092833e-06,
+      "loss": 0.6368,
+      "step": 4676
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3738664605632083,
+      "learning_rate": 8.215305279842385e-06,
+      "loss": 0.7056,
+      "step": 4677
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4883216616452073,
+      "learning_rate": 8.214511453423107e-06,
+      "loss": 0.6295,
+      "step": 4678
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.274965877597899,
+      "learning_rate": 8.213717488869113e-06,
+      "loss": 0.7407,
+      "step": 4679
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.357705143017072,
+      "learning_rate": 8.212923386214522e-06,
+      "loss": 0.6968,
+      "step": 4680
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4744584678112294,
+      "learning_rate": 8.212129145493457e-06,
+      "loss": 0.7151,
+      "step": 4681
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.37535750746998,
+      "learning_rate": 8.211334766740048e-06,
+      "loss": 0.6438,
+      "step": 4682
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5551109484171701,
+      "learning_rate": 8.210540249988435e-06,
+      "loss": 0.6876,
+      "step": 4683
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3300636437811473,
+      "learning_rate": 8.209745595272755e-06,
+      "loss": 0.7686,
+      "step": 4684
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3744064395980158,
+      "learning_rate": 8.208950802627157e-06,
+      "loss": 0.5749,
+      "step": 4685
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3982357414903055,
+      "learning_rate": 8.208155872085799e-06,
+      "loss": 0.7586,
+      "step": 4686
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.378259134206478,
+      "learning_rate": 8.207360803682837e-06,
+      "loss": 0.6763,
+      "step": 4687
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.486533110408155,
+      "learning_rate": 8.206565597452438e-06,
+      "loss": 0.8258,
+      "step": 4688
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2470170733298722,
+      "learning_rate": 8.205770253428775e-06,
+      "loss": 0.7141,
+      "step": 4689
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5755422916952404,
+      "learning_rate": 8.204974771646023e-06,
+      "loss": 0.6649,
+      "step": 4690
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8047664407063126,
+      "learning_rate": 8.204179152138367e-06,
+      "loss": 0.6912,
+      "step": 4691
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4334226318100278,
+      "learning_rate": 8.203383394939998e-06,
+      "loss": 0.7595,
+      "step": 4692
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3330730303410907,
+      "learning_rate": 8.20258750008511e-06,
+      "loss": 0.6488,
+      "step": 4693
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3084066411857438,
+      "learning_rate": 8.201791467607905e-06,
+      "loss": 0.6817,
+      "step": 4694
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2892233100968413,
+      "learning_rate": 8.200995297542589e-06,
+      "loss": 0.5294,
+      "step": 4695
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1601647814852856,
+      "learning_rate": 8.200198989923376e-06,
+      "loss": 0.6628,
+      "step": 4696
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3285094133836504,
+      "learning_rate": 8.199402544784485e-06,
+      "loss": 0.727,
+      "step": 4697
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4120919432152486,
+      "learning_rate": 8.198605962160143e-06,
+      "loss": 0.6761,
+      "step": 4698
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2938225963111487,
+      "learning_rate": 8.197809242084575e-06,
+      "loss": 0.7051,
+      "step": 4699
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1391581287716575,
+      "learning_rate": 8.197012384592027e-06,
+      "loss": 0.6478,
+      "step": 4700
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2864230736201523,
+      "learning_rate": 8.196215389716732e-06,
+      "loss": 0.619,
+      "step": 4701
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2497694421162082,
+      "learning_rate": 8.195418257492948e-06,
+      "loss": 0.5993,
+      "step": 4702
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8798778550760968,
+      "learning_rate": 8.194620987954922e-06,
+      "loss": 0.7146,
+      "step": 4703
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3368317112819466,
+      "learning_rate": 8.193823581136919e-06,
+      "loss": 0.6738,
+      "step": 4704
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2209250792027022,
+      "learning_rate": 8.193026037073201e-06,
+      "loss": 0.6283,
+      "step": 4705
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2262055325098085,
+      "learning_rate": 8.192228355798045e-06,
+      "loss": 0.6922,
+      "step": 4706
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3827515093296427,
+      "learning_rate": 8.191430537345728e-06,
+      "loss": 0.7596,
+      "step": 4707
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2141836236461452,
+      "learning_rate": 8.19063258175053e-06,
+      "loss": 0.6787,
+      "step": 4708
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2414758894325364,
+      "learning_rate": 8.189834489046746e-06,
+      "loss": 0.6504,
+      "step": 4709
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.464076447534928,
+      "learning_rate": 8.18903625926867e-06,
+      "loss": 0.697,
+      "step": 4710
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5673937024837066,
+      "learning_rate": 8.188237892450603e-06,
+      "loss": 0.6795,
+      "step": 4711
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2349521710239084,
+      "learning_rate": 8.187439388626855e-06,
+      "loss": 0.6142,
+      "step": 4712
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4504298042864094,
+      "learning_rate": 8.186640747831735e-06,
+      "loss": 0.5866,
+      "step": 4713
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.573060577184645,
+      "learning_rate": 8.185841970099566e-06,
+      "loss": 0.6992,
+      "step": 4714
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4194680097384944,
+      "learning_rate": 8.185043055464673e-06,
+      "loss": 0.7299,
+      "step": 4715
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3531556194845544,
+      "learning_rate": 8.184244003961386e-06,
+      "loss": 0.6235,
+      "step": 4716
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3732766468342112,
+      "learning_rate": 8.183444815624041e-06,
+      "loss": 0.6605,
+      "step": 4717
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4470643973321633,
+      "learning_rate": 8.182645490486986e-06,
+      "loss": 0.8364,
+      "step": 4718
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5139184472511062,
+      "learning_rate": 8.181846028584563e-06,
+      "loss": 0.6766,
+      "step": 4719
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1367211938769037,
+      "learning_rate": 8.181046429951131e-06,
+      "loss": 0.5612,
+      "step": 4720
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1516172113093737,
+      "learning_rate": 8.180246694621048e-06,
+      "loss": 0.6197,
+      "step": 4721
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1205102410189482,
+      "learning_rate": 8.179446822628684e-06,
+      "loss": 0.6453,
+      "step": 4722
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3917879883652395,
+      "learning_rate": 8.178646814008407e-06,
+      "loss": 0.7584,
+      "step": 4723
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1743415352060067,
+      "learning_rate": 8.177846668794598e-06,
+      "loss": 0.6246,
+      "step": 4724
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6718116791211837,
+      "learning_rate": 8.177046387021641e-06,
+      "loss": 0.7073,
+      "step": 4725
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2640854854649821,
+      "learning_rate": 8.176245968723924e-06,
+      "loss": 0.6669,
+      "step": 4726
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.442124932197131,
+      "learning_rate": 8.175445413935844e-06,
+      "loss": 0.7051,
+      "step": 4727
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3514902519309384,
+      "learning_rate": 8.174644722691802e-06,
+      "loss": 0.7009,
+      "step": 4728
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.022304432308355,
+      "learning_rate": 8.173843895026207e-06,
+      "loss": 0.6803,
+      "step": 4729
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1045804269776445,
+      "learning_rate": 8.17304293097347e-06,
+      "loss": 0.6672,
+      "step": 4730
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.459675596133389,
+      "learning_rate": 8.172241830568012e-06,
+      "loss": 0.7891,
+      "step": 4731
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6161985243785557,
+      "learning_rate": 8.17144059384426e-06,
+      "loss": 0.6965,
+      "step": 4732
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4905892386862025,
+      "learning_rate": 8.170639220836639e-06,
+      "loss": 0.7091,
+      "step": 4733
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.453525977128991,
+      "learning_rate": 8.169837711579591e-06,
+      "loss": 0.7104,
+      "step": 4734
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2211852827923282,
+      "learning_rate": 8.169036066107556e-06,
+      "loss": 0.6828,
+      "step": 4735
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1957540087542176,
+      "learning_rate": 8.168234284454984e-06,
+      "loss": 0.6338,
+      "step": 4736
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5541804087748403,
+      "learning_rate": 8.16743236665633e-06,
+      "loss": 0.7092,
+      "step": 4737
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3615059927726318,
+      "learning_rate": 8.166630312746051e-06,
+      "loss": 0.7551,
+      "step": 4738
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2198244068161739,
+      "learning_rate": 8.165828122758615e-06,
+      "loss": 0.674,
+      "step": 4739
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8199951298561452,
+      "learning_rate": 8.165025796728495e-06,
+      "loss": 0.6968,
+      "step": 4740
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2641523188342947,
+      "learning_rate": 8.164223334690168e-06,
+      "loss": 0.6678,
+      "step": 4741
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4204587854079989,
+      "learning_rate": 8.163420736678115e-06,
+      "loss": 0.6838,
+      "step": 4742
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2144632548330412,
+      "learning_rate": 8.162618002726829e-06,
+      "loss": 0.7449,
+      "step": 4743
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2192859008185208,
+      "learning_rate": 8.161815132870806e-06,
+      "loss": 0.5697,
+      "step": 4744
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.287310456919279,
+      "learning_rate": 8.161012127144543e-06,
+      "loss": 0.7532,
+      "step": 4745
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3057060948081238,
+      "learning_rate": 8.160208985582547e-06,
+      "loss": 0.6092,
+      "step": 4746
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0046022963210097,
+      "learning_rate": 8.159405708219335e-06,
+      "loss": 0.6381,
+      "step": 4747
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.417638360385664,
+      "learning_rate": 8.158602295089423e-06,
+      "loss": 0.7118,
+      "step": 4748
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5352613042061227,
+      "learning_rate": 8.157798746227337e-06,
+      "loss": 0.8147,
+      "step": 4749
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5348855901707725,
+      "learning_rate": 8.156995061667605e-06,
+      "loss": 0.6119,
+      "step": 4750
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5727769020597637,
+      "learning_rate": 8.156191241444763e-06,
+      "loss": 0.5295,
+      "step": 4751
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6804128177232145,
+      "learning_rate": 8.155387285593356e-06,
+      "loss": 0.648,
+      "step": 4752
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2741434820757052,
+      "learning_rate": 8.154583194147929e-06,
+      "loss": 0.6768,
+      "step": 4753
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1829324132796557,
+      "learning_rate": 8.153778967143035e-06,
+      "loss": 0.6613,
+      "step": 4754
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2607926570786538,
+      "learning_rate": 8.152974604613238e-06,
+      "loss": 0.6127,
+      "step": 4755
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.51127681592346,
+      "learning_rate": 8.152170106593099e-06,
+      "loss": 0.666,
+      "step": 4756
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2953266933938263,
+      "learning_rate": 8.15136547311719e-06,
+      "loss": 0.6961,
+      "step": 4757
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2132932608747076,
+      "learning_rate": 8.150560704220087e-06,
+      "loss": 0.5891,
+      "step": 4758
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1603679131821099,
+      "learning_rate": 8.149755799936377e-06,
+      "loss": 0.6561,
+      "step": 4759
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2446380679976905,
+      "learning_rate": 8.148950760300642e-06,
+      "loss": 0.7299,
+      "step": 4760
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3989927319386564,
+      "learning_rate": 8.148145585347482e-06,
+      "loss": 0.6448,
+      "step": 4761
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.337407886579254,
+      "learning_rate": 8.147340275111492e-06,
+      "loss": 0.6953,
+      "step": 4762
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.32752037663147,
+      "learning_rate": 8.146534829627285e-06,
+      "loss": 0.7058,
+      "step": 4763
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.329328188078495,
+      "learning_rate": 8.145729248929466e-06,
+      "loss": 0.7937,
+      "step": 4764
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2506300825871344,
+      "learning_rate": 8.144923533052655e-06,
+      "loss": 0.7115,
+      "step": 4765
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2081980818744005,
+      "learning_rate": 8.144117682031478e-06,
+      "loss": 0.6376,
+      "step": 4766
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1155635332912655,
+      "learning_rate": 8.14331169590056e-06,
+      "loss": 0.6297,
+      "step": 4767
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2884709271783708,
+      "learning_rate": 8.142505574694539e-06,
+      "loss": 0.7142,
+      "step": 4768
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.294807852577825,
+      "learning_rate": 8.141699318448053e-06,
+      "loss": 0.6396,
+      "step": 4769
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1481796651152232,
+      "learning_rate": 8.14089292719575e-06,
+      "loss": 0.5956,
+      "step": 4770
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2073155022564463,
+      "learning_rate": 8.140086400972285e-06,
+      "loss": 0.602,
+      "step": 4771
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5112445976324969,
+      "learning_rate": 8.139279739812314e-06,
+      "loss": 0.6585,
+      "step": 4772
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3908168006387729,
+      "learning_rate": 8.1384729437505e-06,
+      "loss": 0.691,
+      "step": 4773
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6677905073657595,
+      "learning_rate": 8.137666012821514e-06,
+      "loss": 0.8336,
+      "step": 4774
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3171516679983146,
+      "learning_rate": 8.136858947060032e-06,
+      "loss": 0.7324,
+      "step": 4775
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2502866834273123,
+      "learning_rate": 8.136051746500735e-06,
+      "loss": 0.6484,
+      "step": 4776
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3296040002725136,
+      "learning_rate": 8.135244411178309e-06,
+      "loss": 0.6822,
+      "step": 4777
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3869650310031896,
+      "learning_rate": 8.134436941127448e-06,
+      "loss": 0.7819,
+      "step": 4778
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.45197294530881,
+      "learning_rate": 8.13362933638285e-06,
+      "loss": 0.6375,
+      "step": 4779
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.459529540631801,
+      "learning_rate": 8.132821596979223e-06,
+      "loss": 0.8373,
+      "step": 4780
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3084287995834278,
+      "learning_rate": 8.132013722951275e-06,
+      "loss": 0.7379,
+      "step": 4781
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.26533427522396,
+      "learning_rate": 8.13120571433372e-06,
+      "loss": 0.5666,
+      "step": 4782
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0081560383910544,
+      "learning_rate": 8.130397571161283e-06,
+      "loss": 0.7411,
+      "step": 4783
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.061075401876651,
+      "learning_rate": 8.129589293468689e-06,
+      "loss": 0.6268,
+      "step": 4784
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3867715625656638,
+      "learning_rate": 8.128780881290674e-06,
+      "loss": 0.7308,
+      "step": 4785
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.462825535411756,
+      "learning_rate": 8.127972334661978e-06,
+      "loss": 0.6635,
+      "step": 4786
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1733123777977057,
+      "learning_rate": 8.127163653617342e-06,
+      "loss": 0.6184,
+      "step": 4787
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2169491524575136,
+      "learning_rate": 8.12635483819152e-06,
+      "loss": 0.6616,
+      "step": 4788
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3534601918805933,
+      "learning_rate": 8.125545888419269e-06,
+      "loss": 0.6483,
+      "step": 4789
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0571229184089221,
+      "learning_rate": 8.124736804335348e-06,
+      "loss": 0.7012,
+      "step": 4790
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2689787062748217,
+      "learning_rate": 8.123927585974529e-06,
+      "loss": 0.6057,
+      "step": 4791
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2224672060481727,
+      "learning_rate": 8.123118233371585e-06,
+      "loss": 0.6775,
+      "step": 4792
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.292069817371602,
+      "learning_rate": 8.122308746561294e-06,
+      "loss": 0.67,
+      "step": 4793
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4700182744925097,
+      "learning_rate": 8.12149912557844e-06,
+      "loss": 0.6256,
+      "step": 4794
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3010973973191982,
+      "learning_rate": 8.12068937045782e-06,
+      "loss": 0.5934,
+      "step": 4795
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4341548229218928,
+      "learning_rate": 8.119879481234228e-06,
+      "loss": 0.7053,
+      "step": 4796
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.312789923017755,
+      "learning_rate": 8.119069457942464e-06,
+      "loss": 0.7214,
+      "step": 4797
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2179765269780678,
+      "learning_rate": 8.118259300617339e-06,
+      "loss": 0.633,
+      "step": 4798
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2318392944794954,
+      "learning_rate": 8.117449009293668e-06,
+      "loss": 0.591,
+      "step": 4799
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3910602519404613,
+      "learning_rate": 8.11663858400627e-06,
+      "loss": 0.6775,
+      "step": 4800
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.172853188743768,
+      "learning_rate": 8.115828024789972e-06,
+      "loss": 0.5816,
+      "step": 4801
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.317593314970492,
+      "learning_rate": 8.115017331679602e-06,
+      "loss": 0.6081,
+      "step": 4802
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.432784615940521,
+      "learning_rate": 8.114206504710002e-06,
+      "loss": 0.7009,
+      "step": 4803
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4122954796610236,
+      "learning_rate": 8.113395543916012e-06,
+      "loss": 0.6807,
+      "step": 4804
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2252087902160893,
+      "learning_rate": 8.112584449332481e-06,
+      "loss": 0.6593,
+      "step": 4805
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4045400788893032,
+      "learning_rate": 8.111773220994264e-06,
+      "loss": 0.6801,
+      "step": 4806
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.32302917448453,
+      "learning_rate": 8.110961858936222e-06,
+      "loss": 0.7022,
+      "step": 4807
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.45658835017669,
+      "learning_rate": 8.11015036319322e-06,
+      "loss": 0.6536,
+      "step": 4808
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.338025462104145,
+      "learning_rate": 8.109338733800132e-06,
+      "loss": 0.6959,
+      "step": 4809
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.353640275160754,
+      "learning_rate": 8.108526970791832e-06,
+      "loss": 0.6247,
+      "step": 4810
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.310281413868954,
+      "learning_rate": 8.107715074203205e-06,
+      "loss": 0.6938,
+      "step": 4811
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4210774374146469,
+      "learning_rate": 8.10690304406914e-06,
+      "loss": 0.6254,
+      "step": 4812
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.518432361905884,
+      "learning_rate": 8.106090880424534e-06,
+      "loss": 0.7056,
+      "step": 4813
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4235913988965292,
+      "learning_rate": 8.10527858330428e-06,
+      "loss": 0.6777,
+      "step": 4814
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4408807863706472,
+      "learning_rate": 8.104466152743295e-06,
+      "loss": 0.7719,
+      "step": 4815
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4832022347774565,
+      "learning_rate": 8.103653588776483e-06,
+      "loss": 0.6274,
+      "step": 4816
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2560165523339883,
+      "learning_rate": 8.102840891438764e-06,
+      "loss": 0.6428,
+      "step": 4817
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.37612624497732,
+      "learning_rate": 8.10202806076506e-06,
+      "loss": 0.8206,
+      "step": 4818
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2912476240081685,
+      "learning_rate": 8.101215096790305e-06,
+      "loss": 0.6479,
+      "step": 4819
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.315587167772025,
+      "learning_rate": 8.10040199954943e-06,
+      "loss": 0.6749,
+      "step": 4820
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0909617385567203,
+      "learning_rate": 8.099588769077374e-06,
+      "loss": 0.664,
+      "step": 4821
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3016176795005747,
+      "learning_rate": 8.098775405409087e-06,
+      "loss": 0.6537,
+      "step": 4822
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3116692637589464,
+      "learning_rate": 8.09796190857952e-06,
+      "loss": 0.74,
+      "step": 4823
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.604127184065974,
+      "learning_rate": 8.097148278623628e-06,
+      "loss": 0.7842,
+      "step": 4824
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.463205613863089,
+      "learning_rate": 8.096334515576379e-06,
+      "loss": 0.6792,
+      "step": 4825
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2916934121086692,
+      "learning_rate": 8.095520619472739e-06,
+      "loss": 0.624,
+      "step": 4826
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1934887165556476,
+      "learning_rate": 8.094706590347683e-06,
+      "loss": 0.6405,
+      "step": 4827
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3035592976617298,
+      "learning_rate": 8.093892428236192e-06,
+      "loss": 0.5987,
+      "step": 4828
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.660823842785949,
+      "learning_rate": 8.093078133173256e-06,
+      "loss": 0.7705,
+      "step": 4829
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.451307855952916,
+      "learning_rate": 8.09226370519386e-06,
+      "loss": 0.7028,
+      "step": 4830
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.254361221317256,
+      "learning_rate": 8.091449144333009e-06,
+      "loss": 0.5882,
+      "step": 4831
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.317374516509301,
+      "learning_rate": 8.090634450625704e-06,
+      "loss": 0.7227,
+      "step": 4832
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2838352577228402,
+      "learning_rate": 8.08981962410695e-06,
+      "loss": 0.6841,
+      "step": 4833
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3869052898827483,
+      "learning_rate": 8.089004664811767e-06,
+      "loss": 0.6496,
+      "step": 4834
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3303931089772236,
+      "learning_rate": 8.088189572775173e-06,
+      "loss": 0.6121,
+      "step": 4835
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2343354286088177,
+      "learning_rate": 8.087374348032198e-06,
+      "loss": 0.6004,
+      "step": 4836
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1997324284181214,
+      "learning_rate": 8.08655899061787e-06,
+      "loss": 0.6424,
+      "step": 4837
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3478592678464836,
+      "learning_rate": 8.085743500567228e-06,
+      "loss": 0.7348,
+      "step": 4838
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.272443141247995,
+      "learning_rate": 8.084927877915314e-06,
+      "loss": 0.599,
+      "step": 4839
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2279718058344797,
+      "learning_rate": 8.084112122697178e-06,
+      "loss": 0.7183,
+      "step": 4840
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.990275300585154,
+      "learning_rate": 8.083296234947878e-06,
+      "loss": 0.6918,
+      "step": 4841
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1516283022384999,
+      "learning_rate": 8.082480214702472e-06,
+      "loss": 0.6492,
+      "step": 4842
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.328356730439539,
+      "learning_rate": 8.081664061996024e-06,
+      "loss": 0.6728,
+      "step": 4843
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.276219091377177,
+      "learning_rate": 8.080847776863609e-06,
+      "loss": 0.6887,
+      "step": 4844
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4683944208626503,
+      "learning_rate": 8.080031359340303e-06,
+      "loss": 0.7756,
+      "step": 4845
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.188636726043177,
+      "learning_rate": 8.079214809461192e-06,
+      "loss": 0.6021,
+      "step": 4846
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5478046921169515,
+      "learning_rate": 8.07839812726136e-06,
+      "loss": 0.7618,
+      "step": 4847
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3977445090063156,
+      "learning_rate": 8.077581312775908e-06,
+      "loss": 0.6883,
+      "step": 4848
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1895662251304377,
+      "learning_rate": 8.07676436603993e-06,
+      "loss": 0.6472,
+      "step": 4849
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2905642188877793,
+      "learning_rate": 8.075947287088535e-06,
+      "loss": 0.6754,
+      "step": 4850
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1700843887055985,
+      "learning_rate": 8.075130075956836e-06,
+      "loss": 0.6374,
+      "step": 4851
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0990884760642128,
+      "learning_rate": 8.074312732679948e-06,
+      "loss": 0.5829,
+      "step": 4852
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4899403021895032,
+      "learning_rate": 8.073495257292996e-06,
+      "loss": 0.6973,
+      "step": 4853
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1382741465832233,
+      "learning_rate": 8.072677649831107e-06,
+      "loss": 0.6328,
+      "step": 4854
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4452245028414428,
+      "learning_rate": 8.071859910329416e-06,
+      "loss": 0.75,
+      "step": 4855
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2524058385571057,
+      "learning_rate": 8.071042038823063e-06,
+      "loss": 0.6027,
+      "step": 4856
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.576076002444789,
+      "learning_rate": 8.070224035347193e-06,
+      "loss": 0.6377,
+      "step": 4857
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.499829364813277,
+      "learning_rate": 8.069405899936961e-06,
+      "loss": 0.6253,
+      "step": 4858
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4005555757163406,
+      "learning_rate": 8.068587632627521e-06,
+      "loss": 0.6994,
+      "step": 4859
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3242999302123544,
+      "learning_rate": 8.067769233454037e-06,
+      "loss": 0.5839,
+      "step": 4860
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.1172070519069384,
+      "learning_rate": 8.066950702451673e-06,
+      "loss": 0.6875,
+      "step": 4861
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6928568287454016,
+      "learning_rate": 8.066132039655609e-06,
+      "loss": 0.6136,
+      "step": 4862
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4510050032037676,
+      "learning_rate": 8.065313245101024e-06,
+      "loss": 0.7231,
+      "step": 4863
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0191742753855886,
+      "learning_rate": 8.064494318823102e-06,
+      "loss": 0.6272,
+      "step": 4864
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.406079757873369,
+      "learning_rate": 8.06367526085703e-06,
+      "loss": 0.7173,
+      "step": 4865
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3548904861587752,
+      "learning_rate": 8.062856071238014e-06,
+      "loss": 0.7471,
+      "step": 4866
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4535703020977664,
+      "learning_rate": 8.062036750001249e-06,
+      "loss": 0.7429,
+      "step": 4867
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0074895917755,
+      "learning_rate": 8.061217297181946e-06,
+      "loss": 0.6392,
+      "step": 4868
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3307271995648378,
+      "learning_rate": 8.060397712815318e-06,
+      "loss": 0.7867,
+      "step": 4869
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3866921604155518,
+      "learning_rate": 8.059577996936583e-06,
+      "loss": 0.6948,
+      "step": 4870
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1552275416777988,
+      "learning_rate": 8.05875814958097e-06,
+      "loss": 0.5664,
+      "step": 4871
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.32487081037746,
+      "learning_rate": 8.057938170783704e-06,
+      "loss": 0.6989,
+      "step": 4872
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2344251758317772,
+      "learning_rate": 8.057118060580027e-06,
+      "loss": 0.6829,
+      "step": 4873
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4988687636597278,
+      "learning_rate": 8.056297819005177e-06,
+      "loss": 0.6842,
+      "step": 4874
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3873234031713977,
+      "learning_rate": 8.055477446094404e-06,
+      "loss": 0.7269,
+      "step": 4875
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2372515451627328,
+      "learning_rate": 8.054656941882959e-06,
+      "loss": 0.7049,
+      "step": 4876
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.329258890859748,
+      "learning_rate": 8.053836306406102e-06,
+      "loss": 0.6138,
+      "step": 4877
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4201498086830722,
+      "learning_rate": 8.053015539699099e-06,
+      "loss": 0.6214,
+      "step": 4878
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3952984113644131,
+      "learning_rate": 8.052194641797217e-06,
+      "loss": 0.6741,
+      "step": 4879
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2838032315068082,
+      "learning_rate": 8.051373612735736e-06,
+      "loss": 0.6019,
+      "step": 4880
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3797766390369919,
+      "learning_rate": 8.050552452549933e-06,
+      "loss": 0.726,
+      "step": 4881
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2663487658154784,
+      "learning_rate": 8.049731161275099e-06,
+      "loss": 0.7236,
+      "step": 4882
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3091026899637146,
+      "learning_rate": 8.048909738946523e-06,
+      "loss": 0.6758,
+      "step": 4883
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5695184725507731,
+      "learning_rate": 8.048088185599507e-06,
+      "loss": 0.6345,
+      "step": 4884
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2407414971219117,
+      "learning_rate": 8.047266501269352e-06,
+      "loss": 0.6203,
+      "step": 4885
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4448971703954858,
+      "learning_rate": 8.046444685991369e-06,
+      "loss": 0.6443,
+      "step": 4886
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3831172253298611,
+      "learning_rate": 8.045622739800873e-06,
+      "loss": 0.7167,
+      "step": 4887
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2437290646541752,
+      "learning_rate": 8.044800662733185e-06,
+      "loss": 0.6324,
+      "step": 4888
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2385389532550628,
+      "learning_rate": 8.043978454823632e-06,
+      "loss": 0.6132,
+      "step": 4889
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2675425455700546,
+      "learning_rate": 8.043156116107545e-06,
+      "loss": 0.6272,
+      "step": 4890
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5686871421685602,
+      "learning_rate": 8.042333646620262e-06,
+      "loss": 0.8112,
+      "step": 4891
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.305754429194889,
+      "learning_rate": 8.041511046397127e-06,
+      "loss": 0.6099,
+      "step": 4892
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5555363284677044,
+      "learning_rate": 8.040688315473489e-06,
+      "loss": 0.7047,
+      "step": 4893
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0770034758427227,
+      "learning_rate": 8.0398654538847e-06,
+      "loss": 0.6837,
+      "step": 4894
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5234972461530942,
+      "learning_rate": 8.039042461666125e-06,
+      "loss": 0.7644,
+      "step": 4895
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4075504453221435,
+      "learning_rate": 8.038219338853126e-06,
+      "loss": 0.6217,
+      "step": 4896
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3606517895391417,
+      "learning_rate": 8.037396085481075e-06,
+      "loss": 0.6644,
+      "step": 4897
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.316783600671757,
+      "learning_rate": 8.03657270158535e-06,
+      "loss": 0.6337,
+      "step": 4898
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2248528162502357,
+      "learning_rate": 8.035749187201333e-06,
+      "loss": 0.5847,
+      "step": 4899
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0797745220967476,
+      "learning_rate": 8.034925542364412e-06,
+      "loss": 0.6494,
+      "step": 4900
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4014479155856474,
+      "learning_rate": 8.034101767109982e-06,
+      "loss": 0.6366,
+      "step": 4901
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4386022611359413,
+      "learning_rate": 8.033277861473441e-06,
+      "loss": 0.6711,
+      "step": 4902
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.431258048550067,
+      "learning_rate": 8.032453825490197e-06,
+      "loss": 0.7397,
+      "step": 4903
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4125441437188435,
+      "learning_rate": 8.031629659195657e-06,
+      "loss": 0.6806,
+      "step": 4904
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2732554908982074,
+      "learning_rate": 8.030805362625238e-06,
+      "loss": 0.6036,
+      "step": 4905
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4162103299581776,
+      "learning_rate": 8.029980935814363e-06,
+      "loss": 0.7115,
+      "step": 4906
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.144417378995091,
+      "learning_rate": 8.029156378798459e-06,
+      "loss": 0.5741,
+      "step": 4907
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.393888923446136,
+      "learning_rate": 8.028331691612961e-06,
+      "loss": 0.7301,
+      "step": 4908
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1266967291346222,
+      "learning_rate": 8.027506874293304e-06,
+      "loss": 0.6863,
+      "step": 4909
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.218528168518841,
+      "learning_rate": 8.026681926874932e-06,
+      "loss": 0.667,
+      "step": 4910
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6623786168622974,
+      "learning_rate": 8.025856849393301e-06,
+      "loss": 0.7737,
+      "step": 4911
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5275153312645247,
+      "learning_rate": 8.02503164188386e-06,
+      "loss": 0.7658,
+      "step": 4912
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4343565415493356,
+      "learning_rate": 8.024206304382072e-06,
+      "loss": 0.757,
+      "step": 4913
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3115845717997472,
+      "learning_rate": 8.023380836923404e-06,
+      "loss": 0.6802,
+      "step": 4914
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2707060371522139,
+      "learning_rate": 8.022555239543328e-06,
+      "loss": 0.6869,
+      "step": 4915
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3825674881446297,
+      "learning_rate": 8.021729512277323e-06,
+      "loss": 0.6999,
+      "step": 4916
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3514512565295806,
+      "learning_rate": 8.02090365516087e-06,
+      "loss": 0.7333,
+      "step": 4917
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.541690029063588,
+      "learning_rate": 8.02007766822946e-06,
+      "loss": 0.7597,
+      "step": 4918
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3631739530348321,
+      "learning_rate": 8.019251551518585e-06,
+      "loss": 0.6771,
+      "step": 4919
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3108329610352283,
+      "learning_rate": 8.018425305063749e-06,
+      "loss": 0.7269,
+      "step": 4920
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3615209110351014,
+      "learning_rate": 8.017598928900452e-06,
+      "loss": 0.6943,
+      "step": 4921
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.123762594975722,
+      "learning_rate": 8.01677242306421e-06,
+      "loss": 0.6271,
+      "step": 4922
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3709987984516205,
+      "learning_rate": 8.015945787590537e-06,
+      "loss": 0.7167,
+      "step": 4923
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2771631800437429,
+      "learning_rate": 8.015119022514958e-06,
+      "loss": 0.6428,
+      "step": 4924
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4074275593754813,
+      "learning_rate": 8.014292127873002e-06,
+      "loss": 0.6975,
+      "step": 4925
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4764301272577314,
+      "learning_rate": 8.013465103700197e-06,
+      "loss": 0.7582,
+      "step": 4926
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3591090483989416,
+      "learning_rate": 8.012637950032086e-06,
+      "loss": 0.7153,
+      "step": 4927
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3816758845909771,
+      "learning_rate": 8.011810666904212e-06,
+      "loss": 0.7574,
+      "step": 4928
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0869457221348378,
+      "learning_rate": 8.010983254352127e-06,
+      "loss": 0.7627,
+      "step": 4929
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2962110249011858,
+      "learning_rate": 8.010155712411386e-06,
+      "loss": 0.5832,
+      "step": 4930
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3222175083260241,
+      "learning_rate": 8.009328041117547e-06,
+      "loss": 0.6255,
+      "step": 4931
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3477988688696045,
+      "learning_rate": 8.008500240506182e-06,
+      "loss": 0.6832,
+      "step": 4932
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.23035844724956,
+      "learning_rate": 8.007672310612863e-06,
+      "loss": 0.6424,
+      "step": 4933
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.303237592973441,
+      "learning_rate": 8.006844251473165e-06,
+      "loss": 0.6427,
+      "step": 4934
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.383674727135177,
+      "learning_rate": 8.006016063122672e-06,
+      "loss": 0.7247,
+      "step": 4935
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4202959443959144,
+      "learning_rate": 8.005187745596976e-06,
+      "loss": 0.6377,
+      "step": 4936
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.360269386096871,
+      "learning_rate": 8.004359298931668e-06,
+      "loss": 0.6898,
+      "step": 4937
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4579894211105762,
+      "learning_rate": 8.003530723162352e-06,
+      "loss": 0.7586,
+      "step": 4938
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.218758399467766,
+      "learning_rate": 8.002702018324629e-06,
+      "loss": 0.5729,
+      "step": 4939
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1214175339362866,
+      "learning_rate": 8.001873184454114e-06,
+      "loss": 0.6779,
+      "step": 4940
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2750504972523238,
+      "learning_rate": 8.001044221586423e-06,
+      "loss": 0.5814,
+      "step": 4941
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.078931000271361,
+      "learning_rate": 8.000215129757178e-06,
+      "loss": 0.5946,
+      "step": 4942
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2624400030598257,
+      "learning_rate": 7.999385909002007e-06,
+      "loss": 0.5929,
+      "step": 4943
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2449857882303654,
+      "learning_rate": 7.998556559356543e-06,
+      "loss": 0.6755,
+      "step": 4944
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1730155791554413,
+      "learning_rate": 7.997727080856425e-06,
+      "loss": 0.6729,
+      "step": 4945
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4976888042916328,
+      "learning_rate": 7.9968974735373e-06,
+      "loss": 0.7501,
+      "step": 4946
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3207053372685686,
+      "learning_rate": 7.996067737434815e-06,
+      "loss": 0.7218,
+      "step": 4947
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6087312336924746,
+      "learning_rate": 7.995237872584627e-06,
+      "loss": 0.5938,
+      "step": 4948
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3685792790430598,
+      "learning_rate": 7.994407879022397e-06,
+      "loss": 0.7159,
+      "step": 4949
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.034609534085348,
+      "learning_rate": 7.99357775678379e-06,
+      "loss": 0.7413,
+      "step": 4950
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4701901160681066,
+      "learning_rate": 7.99274750590448e-06,
+      "loss": 0.6973,
+      "step": 4951
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4409135875690864,
+      "learning_rate": 7.991917126420147e-06,
+      "loss": 0.6814,
+      "step": 4952
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.404094004013088,
+      "learning_rate": 7.99108661836647e-06,
+      "loss": 0.8092,
+      "step": 4953
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3865249877590693,
+      "learning_rate": 7.990255981779139e-06,
+      "loss": 0.7526,
+      "step": 4954
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.27796732573736,
+      "learning_rate": 7.989425216693849e-06,
+      "loss": 0.5487,
+      "step": 4955
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.424741007726846,
+      "learning_rate": 7.9885943231463e-06,
+      "loss": 0.7502,
+      "step": 4956
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2618511171437943,
+      "learning_rate": 7.987763301172197e-06,
+      "loss": 0.6373,
+      "step": 4957
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2203974340287627,
+      "learning_rate": 7.986932150807253e-06,
+      "loss": 0.6945,
+      "step": 4958
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.716733683158451,
+      "learning_rate": 7.986100872087177e-06,
+      "loss": 0.5937,
+      "step": 4959
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4319881159961803,
+      "learning_rate": 7.9852694650477e-06,
+      "loss": 0.7465,
+      "step": 4960
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.6151977851988275,
+      "learning_rate": 7.984437929724546e-06,
+      "loss": 0.7048,
+      "step": 4961
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.241664109799841,
+      "learning_rate": 7.983606266153445e-06,
+      "loss": 0.5919,
+      "step": 4962
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2184645402329057,
+      "learning_rate": 7.98277447437014e-06,
+      "loss": 0.7402,
+      "step": 4963
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4689357792702677,
+      "learning_rate": 7.981942554410371e-06,
+      "loss": 0.5537,
+      "step": 4964
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3158293797503982,
+      "learning_rate": 7.98111050630989e-06,
+      "loss": 0.7279,
+      "step": 4965
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9372994893421379,
+      "learning_rate": 7.980278330104452e-06,
+      "loss": 0.656,
+      "step": 4966
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.279068920486428,
+      "learning_rate": 7.979446025829815e-06,
+      "loss": 0.612,
+      "step": 4967
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.359679688142268,
+      "learning_rate": 7.978613593521748e-06,
+      "loss": 0.7161,
+      "step": 4968
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0854639105048687,
+      "learning_rate": 7.97778103321602e-06,
+      "loss": 0.713,
+      "step": 4969
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1594389141505919,
+      "learning_rate": 7.976948344948412e-06,
+      "loss": 0.6614,
+      "step": 4970
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3166007315285984,
+      "learning_rate": 7.976115528754702e-06,
+      "loss": 0.6292,
+      "step": 4971
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2440270364177093,
+      "learning_rate": 7.975282584670679e-06,
+      "loss": 0.6335,
+      "step": 4972
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.289816850053952,
+      "learning_rate": 7.974449512732137e-06,
+      "loss": 0.6343,
+      "step": 4973
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2099254722230033,
+      "learning_rate": 7.973616312974876e-06,
+      "loss": 0.7186,
+      "step": 4974
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5817369989908707,
+      "learning_rate": 7.972782985434697e-06,
+      "loss": 0.6441,
+      "step": 4975
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.278043508611519,
+      "learning_rate": 7.971949530147417e-06,
+      "loss": 0.6526,
+      "step": 4976
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3447330971043965,
+      "learning_rate": 7.971115947148842e-06,
+      "loss": 0.6262,
+      "step": 4977
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4018149037035288,
+      "learning_rate": 7.9702822364748e-06,
+      "loss": 0.6637,
+      "step": 4978
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1167451084004976,
+      "learning_rate": 7.969448398161115e-06,
+      "loss": 0.5967,
+      "step": 4979
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3443723804857606,
+      "learning_rate": 7.968614432243617e-06,
+      "loss": 0.6782,
+      "step": 4980
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4422557962698508,
+      "learning_rate": 7.96778033875815e-06,
+      "loss": 0.6684,
+      "step": 4981
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.731991366819159,
+      "learning_rate": 7.966946117740548e-06,
+      "loss": 0.7476,
+      "step": 4982
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5901263561749657,
+      "learning_rate": 7.966111769226666e-06,
+      "loss": 0.785,
+      "step": 4983
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4071791113630072,
+      "learning_rate": 7.965277293252354e-06,
+      "loss": 0.7168,
+      "step": 4984
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.255572763068824,
+      "learning_rate": 7.964442689853473e-06,
+      "loss": 0.6367,
+      "step": 4985
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5060491187704401,
+      "learning_rate": 7.963607959065887e-06,
+      "loss": 0.6856,
+      "step": 4986
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.377830447044763,
+      "learning_rate": 7.962773100925467e-06,
+      "loss": 0.7025,
+      "step": 4987
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.344802342769288,
+      "learning_rate": 7.961938115468088e-06,
+      "loss": 0.7619,
+      "step": 4988
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5508234040806648,
+      "learning_rate": 7.961103002729634e-06,
+      "loss": 0.7172,
+      "step": 4989
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1343576176045262,
+      "learning_rate": 7.960267762745985e-06,
+      "loss": 0.5861,
+      "step": 4990
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3709727361436794,
+      "learning_rate": 7.95943239555304e-06,
+      "loss": 0.6926,
+      "step": 4991
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1837716232658804,
+      "learning_rate": 7.958596901186695e-06,
+      "loss": 0.7018,
+      "step": 4992
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5996977426709824,
+      "learning_rate": 7.95776127968285e-06,
+      "loss": 0.6794,
+      "step": 4993
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.366519490285235,
+      "learning_rate": 7.956925531077417e-06,
+      "loss": 0.623,
+      "step": 4994
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4525649999367891,
+      "learning_rate": 7.956089655406307e-06,
+      "loss": 0.7168,
+      "step": 4995
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0815910032014093,
+      "learning_rate": 7.955253652705442e-06,
+      "loss": 0.6164,
+      "step": 4996
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4017849675700327,
+      "learning_rate": 7.954417523010747e-06,
+      "loss": 0.7627,
+      "step": 4997
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.367242228239053,
+      "learning_rate": 7.953581266358148e-06,
+      "loss": 0.6161,
+      "step": 4998
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2883022248987581,
+      "learning_rate": 7.952744882783587e-06,
+      "loss": 0.7154,
+      "step": 4999
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0521268146933107,
+      "learning_rate": 7.951908372323001e-06,
+      "loss": 0.6388,
+      "step": 5000
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6328251723901437,
+      "learning_rate": 7.951071735012337e-06,
+      "loss": 0.7022,
+      "step": 5001
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2074950911158826,
+      "learning_rate": 7.95023497088755e-06,
+      "loss": 0.6682,
+      "step": 5002
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3088082710150957,
+      "learning_rate": 7.949398079984594e-06,
+      "loss": 0.7448,
+      "step": 5003
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4061637938800646,
+      "learning_rate": 7.948561062339435e-06,
+      "loss": 0.7191,
+      "step": 5004
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.07289934920556,
+      "learning_rate": 7.94772391798804e-06,
+      "loss": 0.5954,
+      "step": 5005
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.675997303326753,
+      "learning_rate": 7.946886646966384e-06,
+      "loss": 0.7065,
+      "step": 5006
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1785858686916175,
+      "learning_rate": 7.946049249310445e-06,
+      "loss": 0.6512,
+      "step": 5007
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2801960545614497,
+      "learning_rate": 7.945211725056208e-06,
+      "loss": 0.6807,
+      "step": 5008
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2814398569048544,
+      "learning_rate": 7.944374074239665e-06,
+      "loss": 0.7928,
+      "step": 5009
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5787765561463245,
+      "learning_rate": 7.94353629689681e-06,
+      "loss": 0.6024,
+      "step": 5010
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5947038313617294,
+      "learning_rate": 7.942698393063646e-06,
+      "loss": 0.7416,
+      "step": 5011
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.522179579486184,
+      "learning_rate": 7.941860362776176e-06,
+      "loss": 0.7873,
+      "step": 5012
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2494705466966134,
+      "learning_rate": 7.941022206070415e-06,
+      "loss": 0.6428,
+      "step": 5013
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2339847845902399,
+      "learning_rate": 7.940183922982381e-06,
+      "loss": 0.6059,
+      "step": 5014
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4317128767416478,
+      "learning_rate": 7.939345513548095e-06,
+      "loss": 0.7111,
+      "step": 5015
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4151219642954032,
+      "learning_rate": 7.938506977803586e-06,
+      "loss": 0.6365,
+      "step": 5016
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2914049794251188,
+      "learning_rate": 7.937668315784888e-06,
+      "loss": 0.5855,
+      "step": 5017
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0267831417155695,
+      "learning_rate": 7.93682952752804e-06,
+      "loss": 0.6503,
+      "step": 5018
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2498910281223334,
+      "learning_rate": 7.935990613069087e-06,
+      "loss": 0.5996,
+      "step": 5019
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.32488726421259,
+      "learning_rate": 7.935151572444076e-06,
+      "loss": 0.7003,
+      "step": 5020
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4716488406252488,
+      "learning_rate": 7.934312405689067e-06,
+      "loss": 0.7349,
+      "step": 5021
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1260035314830776,
+      "learning_rate": 7.933473112840117e-06,
+      "loss": 0.5003,
+      "step": 5022
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3189708309066066,
+      "learning_rate": 7.932633693933298e-06,
+      "loss": 0.715,
+      "step": 5023
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1592705324879053,
+      "learning_rate": 7.931794149004675e-06,
+      "loss": 0.6642,
+      "step": 5024
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6571078571317208,
+      "learning_rate": 7.930954478090329e-06,
+      "loss": 0.7183,
+      "step": 5025
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3285016585121214,
+      "learning_rate": 7.930114681226341e-06,
+      "loss": 0.7328,
+      "step": 5026
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.688532165728546,
+      "learning_rate": 7.9292747584488e-06,
+      "loss": 0.6562,
+      "step": 5027
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4886744035640647,
+      "learning_rate": 7.928434709793798e-06,
+      "loss": 0.6579,
+      "step": 5028
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3661795932521366,
+      "learning_rate": 7.927594535297433e-06,
+      "loss": 0.6651,
+      "step": 5029
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3823495181737289,
+      "learning_rate": 7.926754234995814e-06,
+      "loss": 0.6459,
+      "step": 5030
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4114638176788246,
+      "learning_rate": 7.925913808925045e-06,
+      "loss": 0.6742,
+      "step": 5031
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2022380272574877,
+      "learning_rate": 7.925073257121245e-06,
+      "loss": 0.6019,
+      "step": 5032
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2984318470591005,
+      "learning_rate": 7.924232579620533e-06,
+      "loss": 0.611,
+      "step": 5033
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4358250645262953,
+      "learning_rate": 7.923391776459031e-06,
+      "loss": 0.6469,
+      "step": 5034
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.279874943910786,
+      "learning_rate": 7.922550847672876e-06,
+      "loss": 0.7419,
+      "step": 5035
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4012643152588258,
+      "learning_rate": 7.921709793298202e-06,
+      "loss": 0.8153,
+      "step": 5036
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.269487274507274,
+      "learning_rate": 7.920868613371151e-06,
+      "loss": 0.6615,
+      "step": 5037
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2688904274159056,
+      "learning_rate": 7.920027307927872e-06,
+      "loss": 0.6329,
+      "step": 5038
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4174113025186124,
+      "learning_rate": 7.919185877004515e-06,
+      "loss": 0.7289,
+      "step": 5039
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0591481619516174,
+      "learning_rate": 7.91834432063724e-06,
+      "loss": 0.637,
+      "step": 5040
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.381110846636413,
+      "learning_rate": 7.917502638862209e-06,
+      "loss": 0.7447,
+      "step": 5041
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.205753741978157,
+      "learning_rate": 7.916660831715592e-06,
+      "loss": 0.5663,
+      "step": 5042
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0308531486725125,
+      "learning_rate": 7.915818899233564e-06,
+      "loss": 0.5049,
+      "step": 5043
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3990605907559226,
+      "learning_rate": 7.914976841452304e-06,
+      "loss": 0.6916,
+      "step": 5044
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8812640319682454,
+      "learning_rate": 7.914134658407997e-06,
+      "loss": 0.6143,
+      "step": 5045
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.020985518477613,
+      "learning_rate": 7.913292350136834e-06,
+      "loss": 0.5964,
+      "step": 5046
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.225870575044792,
+      "learning_rate": 7.912449916675008e-06,
+      "loss": 0.5778,
+      "step": 5047
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2763643964477804,
+      "learning_rate": 7.911607358058726e-06,
+      "loss": 0.7043,
+      "step": 5048
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2901672734122516,
+      "learning_rate": 7.91076467432419e-06,
+      "loss": 0.6181,
+      "step": 5049
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4417739458267518,
+      "learning_rate": 7.909921865507614e-06,
+      "loss": 0.7115,
+      "step": 5050
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3227883710328159,
+      "learning_rate": 7.909078931645215e-06,
+      "loss": 0.7148,
+      "step": 5051
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5173670521851512,
+      "learning_rate": 7.908235872773214e-06,
+      "loss": 0.7987,
+      "step": 5052
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0170897020915677,
+      "learning_rate": 7.907392688927842e-06,
+      "loss": 0.6474,
+      "step": 5053
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.429485647788818,
+      "learning_rate": 7.90654938014533e-06,
+      "loss": 0.7753,
+      "step": 5054
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4808317678465648,
+      "learning_rate": 7.905705946461918e-06,
+      "loss": 0.7629,
+      "step": 5055
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4077206871316101,
+      "learning_rate": 7.90486238791385e-06,
+      "loss": 0.6614,
+      "step": 5056
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5114450436149414,
+      "learning_rate": 7.904018704537376e-06,
+      "loss": 0.7376,
+      "step": 5057
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4231508570543825,
+      "learning_rate": 7.90317489636875e-06,
+      "loss": 0.7351,
+      "step": 5058
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4377165980460136,
+      "learning_rate": 7.902330963444234e-06,
+      "loss": 0.7063,
+      "step": 5059
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.191149806908578,
+      "learning_rate": 7.901486905800092e-06,
+      "loss": 0.6793,
+      "step": 5060
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2987126224028325,
+      "learning_rate": 7.900642723472596e-06,
+      "loss": 0.6013,
+      "step": 5061
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5243959831346579,
+      "learning_rate": 7.899798416498023e-06,
+      "loss": 0.7622,
+      "step": 5062
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3524578814981687,
+      "learning_rate": 7.898953984912652e-06,
+      "loss": 0.7285,
+      "step": 5063
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4831765697160633,
+      "learning_rate": 7.898109428752773e-06,
+      "loss": 0.7367,
+      "step": 5064
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3186047842326496,
+      "learning_rate": 7.897264748054679e-06,
+      "loss": 0.7782,
+      "step": 5065
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.401728304710544,
+      "learning_rate": 7.896419942854662e-06,
+      "loss": 0.6751,
+      "step": 5066
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5403354887137797,
+      "learning_rate": 7.895575013189032e-06,
+      "loss": 0.6955,
+      "step": 5067
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4278868946205527,
+      "learning_rate": 7.894729959094097e-06,
+      "loss": 0.6312,
+      "step": 5068
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1969310177986647,
+      "learning_rate": 7.893884780606164e-06,
+      "loss": 0.6939,
+      "step": 5069
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3851434238207834,
+      "learning_rate": 7.893039477761562e-06,
+      "loss": 0.79,
+      "step": 5070
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4492255841949284,
+      "learning_rate": 7.892194050596606e-06,
+      "loss": 0.7668,
+      "step": 5071
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0591575354075593,
+      "learning_rate": 7.891348499147634e-06,
+      "loss": 0.5354,
+      "step": 5072
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2147010940599998,
+      "learning_rate": 7.890502823450975e-06,
+      "loss": 0.5476,
+      "step": 5073
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4323695815974422,
+      "learning_rate": 7.889657023542973e-06,
+      "loss": 0.6687,
+      "step": 5074
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2975076515574466,
+      "learning_rate": 7.888811099459974e-06,
+      "loss": 0.6173,
+      "step": 5075
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.139690188113205,
+      "learning_rate": 7.887965051238326e-06,
+      "loss": 0.6355,
+      "step": 5076
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2309849973716813,
+      "learning_rate": 7.88711887891439e-06,
+      "loss": 0.6212,
+      "step": 5077
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0918856331967033,
+      "learning_rate": 7.886272582524525e-06,
+      "loss": 0.6997,
+      "step": 5078
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5065310712761673,
+      "learning_rate": 7.885426162105101e-06,
+      "loss": 0.7549,
+      "step": 5079
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3471413281044429,
+      "learning_rate": 7.884579617692487e-06,
+      "loss": 0.7088,
+      "step": 5080
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.999023299494564,
+      "learning_rate": 7.883732949323061e-06,
+      "loss": 0.654,
+      "step": 5081
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3066290052913008,
+      "learning_rate": 7.882886157033209e-06,
+      "loss": 0.7225,
+      "step": 5082
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1071016335510933,
+      "learning_rate": 7.882039240859318e-06,
+      "loss": 0.7216,
+      "step": 5083
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.6359816340736435,
+      "learning_rate": 7.881192200837785e-06,
+      "loss": 0.7004,
+      "step": 5084
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.343261686394062,
+      "learning_rate": 7.880345037005002e-06,
+      "loss": 0.7085,
+      "step": 5085
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.552851500676158,
+      "learning_rate": 7.879497749397382e-06,
+      "loss": 0.6883,
+      "step": 5086
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5909877472196896,
+      "learning_rate": 7.878650338051328e-06,
+      "loss": 0.869,
+      "step": 5087
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.252495801287029,
+      "learning_rate": 7.877802803003258e-06,
+      "loss": 0.5582,
+      "step": 5088
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.376565500311921,
+      "learning_rate": 7.876955144289594e-06,
+      "loss": 0.6145,
+      "step": 5089
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.517208861552269,
+      "learning_rate": 7.876107361946758e-06,
+      "loss": 0.7668,
+      "step": 5090
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.355585030507422,
+      "learning_rate": 7.875259456011185e-06,
+      "loss": 0.6342,
+      "step": 5091
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2556716107430772,
+      "learning_rate": 7.874411426519308e-06,
+      "loss": 0.677,
+      "step": 5092
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4027614425725752,
+      "learning_rate": 7.873563273507571e-06,
+      "loss": 0.7128,
+      "step": 5093
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.427279190148912,
+      "learning_rate": 7.872714997012421e-06,
+      "loss": 0.7204,
+      "step": 5094
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1808690703374525,
+      "learning_rate": 7.87186659707031e-06,
+      "loss": 0.5889,
+      "step": 5095
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3914163588448134,
+      "learning_rate": 7.871018073717693e-06,
+      "loss": 0.701,
+      "step": 5096
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4383306633303725,
+      "learning_rate": 7.870169426991038e-06,
+      "loss": 0.6369,
+      "step": 5097
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2887720210517268,
+      "learning_rate": 7.86932065692681e-06,
+      "loss": 0.5319,
+      "step": 5098
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2960603875337404,
+      "learning_rate": 7.868471763561482e-06,
+      "loss": 0.643,
+      "step": 5099
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3205941771661331,
+      "learning_rate": 7.867622746931533e-06,
+      "loss": 0.7016,
+      "step": 5100
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1433889863339743,
+      "learning_rate": 7.86677360707345e-06,
+      "loss": 0.5913,
+      "step": 5101
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3220938006828602,
+      "learning_rate": 7.865924344023722e-06,
+      "loss": 0.6279,
+      "step": 5102
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.379753155121215,
+      "learning_rate": 7.865074957818839e-06,
+      "loss": 0.6466,
+      "step": 5103
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0986452167283436,
+      "learning_rate": 7.864225448495304e-06,
+      "loss": 0.687,
+      "step": 5104
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4301396613630981,
+      "learning_rate": 7.863375816089624e-06,
+      "loss": 0.6026,
+      "step": 5105
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3669638182634205,
+      "learning_rate": 7.862526060638309e-06,
+      "loss": 0.6157,
+      "step": 5106
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3525772834370862,
+      "learning_rate": 7.861676182177873e-06,
+      "loss": 0.6444,
+      "step": 5107
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2772548319170174,
+      "learning_rate": 7.860826180744837e-06,
+      "loss": 0.7453,
+      "step": 5108
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3464817018555741,
+      "learning_rate": 7.85997605637573e-06,
+      "loss": 0.5856,
+      "step": 5109
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.32176140264076,
+      "learning_rate": 7.859125809107082e-06,
+      "loss": 0.7495,
+      "step": 5110
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7593205687771936,
+      "learning_rate": 7.85827543897543e-06,
+      "loss": 0.7943,
+      "step": 5111
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3616014186770993,
+      "learning_rate": 7.857424946017317e-06,
+      "loss": 0.7312,
+      "step": 5112
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.308168183796249,
+      "learning_rate": 7.85657433026929e-06,
+      "loss": 0.6815,
+      "step": 5113
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5515525948627462,
+      "learning_rate": 7.855723591767903e-06,
+      "loss": 0.6515,
+      "step": 5114
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3249295158243153,
+      "learning_rate": 7.854872730549712e-06,
+      "loss": 0.705,
+      "step": 5115
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3383708089672088,
+      "learning_rate": 7.854021746651283e-06,
+      "loss": 0.7739,
+      "step": 5116
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2085909564203476,
+      "learning_rate": 7.853170640109182e-06,
+      "loss": 0.6549,
+      "step": 5117
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3585795195198376,
+      "learning_rate": 7.852319410959984e-06,
+      "loss": 0.6813,
+      "step": 5118
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5490157449080721,
+      "learning_rate": 7.85146805924027e-06,
+      "loss": 0.6829,
+      "step": 5119
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2826330759549562,
+      "learning_rate": 7.850616584986621e-06,
+      "loss": 0.6887,
+      "step": 5120
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3213953566823775,
+      "learning_rate": 7.84976498823563e-06,
+      "loss": 0.6209,
+      "step": 5121
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0745807381816357,
+      "learning_rate": 7.84891326902389e-06,
+      "loss": 0.6433,
+      "step": 5122
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3451869017192721,
+      "learning_rate": 7.848061427388002e-06,
+      "loss": 0.711,
+      "step": 5123
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.337363231790267,
+      "learning_rate": 7.847209463364574e-06,
+      "loss": 0.6395,
+      "step": 5124
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5910058406024226,
+      "learning_rate": 7.846357376990213e-06,
+      "loss": 0.743,
+      "step": 5125
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3546298580747849,
+      "learning_rate": 7.845505168301533e-06,
+      "loss": 0.6586,
+      "step": 5126
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2484848278576786,
+      "learning_rate": 7.844652837335162e-06,
+      "loss": 0.6389,
+      "step": 5127
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2332992627512402,
+      "learning_rate": 7.843800384127721e-06,
+      "loss": 0.7548,
+      "step": 5128
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9691300150213665,
+      "learning_rate": 7.842947808715848e-06,
+      "loss": 0.6217,
+      "step": 5129
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6284445576441613,
+      "learning_rate": 7.842095111136173e-06,
+      "loss": 0.6778,
+      "step": 5130
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.429163041820387,
+      "learning_rate": 7.841242291425342e-06,
+      "loss": 0.6144,
+      "step": 5131
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.294293996122812,
+      "learning_rate": 7.840389349620002e-06,
+      "loss": 0.6465,
+      "step": 5132
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3396174273921682,
+      "learning_rate": 7.839536285756804e-06,
+      "loss": 0.7517,
+      "step": 5133
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2839168924784812,
+      "learning_rate": 7.83868309987241e-06,
+      "loss": 0.6651,
+      "step": 5134
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.239552306666339,
+      "learning_rate": 7.837829792003479e-06,
+      "loss": 0.7474,
+      "step": 5135
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3451575012593648,
+      "learning_rate": 7.836976362186684e-06,
+      "loss": 0.682,
+      "step": 5136
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4724145273988265,
+      "learning_rate": 7.836122810458696e-06,
+      "loss": 0.6701,
+      "step": 5137
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.617521168938781,
+      "learning_rate": 7.835269136856194e-06,
+      "loss": 0.7807,
+      "step": 5138
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.217941927870993,
+      "learning_rate": 7.834415341415862e-06,
+      "loss": 0.6675,
+      "step": 5139
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.746007860765564,
+      "learning_rate": 7.833561424174392e-06,
+      "loss": 0.6053,
+      "step": 5140
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4389755521711565,
+      "learning_rate": 7.832707385168474e-06,
+      "loss": 0.5529,
+      "step": 5141
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2748275605539838,
+      "learning_rate": 7.831853224434815e-06,
+      "loss": 0.6166,
+      "step": 5142
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4047713053836215,
+      "learning_rate": 7.830998942010113e-06,
+      "loss": 0.6806,
+      "step": 5143
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.540154889374685,
+      "learning_rate": 7.830144537931082e-06,
+      "loss": 0.716,
+      "step": 5144
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3469488690490978,
+      "learning_rate": 7.829290012234438e-06,
+      "loss": 0.7381,
+      "step": 5145
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2450874770859697,
+      "learning_rate": 7.8284353649569e-06,
+      "loss": 0.5426,
+      "step": 5146
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2983297885902103,
+      "learning_rate": 7.827580596135196e-06,
+      "loss": 0.6529,
+      "step": 5147
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2792909788088034,
+      "learning_rate": 7.826725705806056e-06,
+      "loss": 0.7063,
+      "step": 5148
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0734891169004168,
+      "learning_rate": 7.825870694006217e-06,
+      "loss": 0.7333,
+      "step": 5149
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6131011830096187,
+      "learning_rate": 7.82501556077242e-06,
+      "loss": 0.612,
+      "step": 5150
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1336829860090127,
+      "learning_rate": 7.824160306141413e-06,
+      "loss": 0.5695,
+      "step": 5151
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3161728390557113,
+      "learning_rate": 7.823304930149949e-06,
+      "loss": 0.702,
+      "step": 5152
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4438291927218052,
+      "learning_rate": 7.822449432834784e-06,
+      "loss": 0.7419,
+      "step": 5153
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2841317776547925,
+      "learning_rate": 7.82159381423268e-06,
+      "loss": 0.6348,
+      "step": 5154
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3271276258889264,
+      "learning_rate": 7.820738074380404e-06,
+      "loss": 0.7215,
+      "step": 5155
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3145119232951468,
+      "learning_rate": 7.819882213314736e-06,
+      "loss": 0.6684,
+      "step": 5156
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0638800832215967,
+      "learning_rate": 7.819026231072445e-06,
+      "loss": 0.4903,
+      "step": 5157
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4205202253190865,
+      "learning_rate": 7.81817012769032e-06,
+      "loss": 0.5708,
+      "step": 5158
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5302815496406328,
+      "learning_rate": 7.817313903205148e-06,
+      "loss": 0.7404,
+      "step": 5159
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.567506453401413,
+      "learning_rate": 7.816457557653722e-06,
+      "loss": 0.7281,
+      "step": 5160
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3857706350840993,
+      "learning_rate": 7.815601091072845e-06,
+      "loss": 0.7622,
+      "step": 5161
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4765740064865254,
+      "learning_rate": 7.814744503499317e-06,
+      "loss": 0.7526,
+      "step": 5162
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4163987014238806,
+      "learning_rate": 7.813887794969948e-06,
+      "loss": 0.6396,
+      "step": 5163
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2872903473236976,
+      "learning_rate": 7.813030965521554e-06,
+      "loss": 0.6161,
+      "step": 5164
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9509212951142336,
+      "learning_rate": 7.812174015190955e-06,
+      "loss": 0.7314,
+      "step": 5165
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2372378383452285,
+      "learning_rate": 7.811316944014974e-06,
+      "loss": 0.6094,
+      "step": 5166
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2813338125052218,
+      "learning_rate": 7.810459752030444e-06,
+      "loss": 0.6288,
+      "step": 5167
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3194738721179335,
+      "learning_rate": 7.809602439274198e-06,
+      "loss": 0.6181,
+      "step": 5168
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5405131042123805,
+      "learning_rate": 7.80874500578308e-06,
+      "loss": 0.7696,
+      "step": 5169
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4037278305571848,
+      "learning_rate": 7.807887451593932e-06,
+      "loss": 0.73,
+      "step": 5170
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2021981424817538,
+      "learning_rate": 7.807029776743608e-06,
+      "loss": 0.704,
+      "step": 5171
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1696593618050959,
+      "learning_rate": 7.806171981268962e-06,
+      "loss": 0.6465,
+      "step": 5172
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1886768115652113,
+      "learning_rate": 7.805314065206857e-06,
+      "loss": 0.7261,
+      "step": 5173
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3002638402184807,
+      "learning_rate": 7.804456028594158e-06,
+      "loss": 0.6942,
+      "step": 5174
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1614107855645743,
+      "learning_rate": 7.803597871467738e-06,
+      "loss": 0.6568,
+      "step": 5175
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1227269197838334,
+      "learning_rate": 7.802739593864475e-06,
+      "loss": 0.7521,
+      "step": 5176
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0189860928117513,
+      "learning_rate": 7.801881195821246e-06,
+      "loss": 0.662,
+      "step": 5177
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1994998952019245,
+      "learning_rate": 7.801022677374945e-06,
+      "loss": 0.5132,
+      "step": 5178
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2853201763565134,
+      "learning_rate": 7.80016403856246e-06,
+      "loss": 0.6057,
+      "step": 5179
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3115997592430162,
+      "learning_rate": 7.799305279420691e-06,
+      "loss": 0.7039,
+      "step": 5180
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.187861349102758,
+      "learning_rate": 7.79844639998654e-06,
+      "loss": 0.649,
+      "step": 5181
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4365563776422183,
+      "learning_rate": 7.797587400296913e-06,
+      "loss": 0.635,
+      "step": 5182
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3943021685621335,
+      "learning_rate": 7.796728280388725e-06,
+      "loss": 0.6991,
+      "step": 5183
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2066399810492265,
+      "learning_rate": 7.795869040298895e-06,
+      "loss": 0.6755,
+      "step": 5184
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2834447056957574,
+      "learning_rate": 7.795009680064344e-06,
+      "loss": 0.5034,
+      "step": 5185
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.207246175401058,
+      "learning_rate": 7.794150199722004e-06,
+      "loss": 0.615,
+      "step": 5186
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4222067222485972,
+      "learning_rate": 7.793290599308807e-06,
+      "loss": 0.6849,
+      "step": 5187
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4346752468431132,
+      "learning_rate": 7.792430878861691e-06,
+      "loss": 0.6891,
+      "step": 5188
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3555089875883324,
+      "learning_rate": 7.791571038417602e-06,
+      "loss": 0.7006,
+      "step": 5189
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.382787978967613,
+      "learning_rate": 7.79071107801349e-06,
+      "loss": 0.6749,
+      "step": 5190
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5659991636481938,
+      "learning_rate": 7.789850997686303e-06,
+      "loss": 0.6378,
+      "step": 5191
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3373686040535036,
+      "learning_rate": 7.78899079747301e-06,
+      "loss": 0.6912,
+      "step": 5192
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.369215319226181,
+      "learning_rate": 7.788130477410567e-06,
+      "loss": 0.6068,
+      "step": 5193
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3166699578487153,
+      "learning_rate": 7.78727003753595e-06,
+      "loss": 0.7003,
+      "step": 5194
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2715399948712471,
+      "learning_rate": 7.786409477886133e-06,
+      "loss": 0.6603,
+      "step": 5195
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9243359745436168,
+      "learning_rate": 7.785548798498093e-06,
+      "loss": 0.5644,
+      "step": 5196
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5384218246532173,
+      "learning_rate": 7.784687999408817e-06,
+      "loss": 0.6777,
+      "step": 5197
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2329240595969457,
+      "learning_rate": 7.7838270806553e-06,
+      "loss": 0.6624,
+      "step": 5198
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6195260174005128,
+      "learning_rate": 7.782966042274529e-06,
+      "loss": 0.7608,
+      "step": 5199
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2869388210202972,
+      "learning_rate": 7.782104884303512e-06,
+      "loss": 0.6564,
+      "step": 5200
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5995060867037267,
+      "learning_rate": 7.78124360677925e-06,
+      "loss": 0.7169,
+      "step": 5201
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3604047298692135,
+      "learning_rate": 7.780382209738756e-06,
+      "loss": 0.6916,
+      "step": 5202
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3862046186101322,
+      "learning_rate": 7.779520693219047e-06,
+      "loss": 0.697,
+      "step": 5203
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5900884145088152,
+      "learning_rate": 7.778659057257144e-06,
+      "loss": 0.7543,
+      "step": 5204
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4457577693771306,
+      "learning_rate": 7.777797301890073e-06,
+      "loss": 0.7382,
+      "step": 5205
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.101331737917476,
+      "learning_rate": 7.776935427154864e-06,
+      "loss": 0.6148,
+      "step": 5206
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3618085420446295,
+      "learning_rate": 7.776073433088555e-06,
+      "loss": 0.6288,
+      "step": 5207
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2238432905768442,
+      "learning_rate": 7.775211319728191e-06,
+      "loss": 0.5932,
+      "step": 5208
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2341749180981552,
+      "learning_rate": 7.774349087110813e-06,
+      "loss": 0.6802,
+      "step": 5209
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3750731683362734,
+      "learning_rate": 7.773486735273476e-06,
+      "loss": 0.78,
+      "step": 5210
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3651682860250685,
+      "learning_rate": 7.772624264253237e-06,
+      "loss": 0.83,
+      "step": 5211
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1475941385271096,
+      "learning_rate": 7.77176167408716e-06,
+      "loss": 0.681,
+      "step": 5212
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6004001648730855,
+      "learning_rate": 7.77089896481231e-06,
+      "loss": 0.8007,
+      "step": 5213
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5486491964947298,
+      "learning_rate": 7.77003613646576e-06,
+      "loss": 0.7825,
+      "step": 5214
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2401224085113118,
+      "learning_rate": 7.769173189084589e-06,
+      "loss": 0.6969,
+      "step": 5215
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1558035972290548,
+      "learning_rate": 7.768310122705877e-06,
+      "loss": 0.5524,
+      "step": 5216
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.375208430895136,
+      "learning_rate": 7.767446937366716e-06,
+      "loss": 0.6803,
+      "step": 5217
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4959039519441406,
+      "learning_rate": 7.766583633104195e-06,
+      "loss": 0.6971,
+      "step": 5218
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1352821238246604,
+      "learning_rate": 7.765720209955414e-06,
+      "loss": 0.5938,
+      "step": 5219
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9904075536869366,
+      "learning_rate": 7.764856667957477e-06,
+      "loss": 0.6412,
+      "step": 5220
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.52175347145249,
+      "learning_rate": 7.76399300714749e-06,
+      "loss": 0.7238,
+      "step": 5221
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2580600840677996,
+      "learning_rate": 7.763129227562568e-06,
+      "loss": 0.7002,
+      "step": 5222
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5327927027264914,
+      "learning_rate": 7.762265329239829e-06,
+      "loss": 0.7043,
+      "step": 5223
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4991253342594317,
+      "learning_rate": 7.761401312216398e-06,
+      "loss": 0.749,
+      "step": 5224
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.147860141426618,
+      "learning_rate": 7.760537176529402e-06,
+      "loss": 0.6425,
+      "step": 5225
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3595079056687391,
+      "learning_rate": 7.759672922215976e-06,
+      "loss": 0.6229,
+      "step": 5226
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2079455273646065,
+      "learning_rate": 7.758808549313258e-06,
+      "loss": 0.6749,
+      "step": 5227
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4292918833080333,
+      "learning_rate": 7.757944057858392e-06,
+      "loss": 0.743,
+      "step": 5228
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4277772825313648,
+      "learning_rate": 7.757079447888529e-06,
+      "loss": 0.738,
+      "step": 5229
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3716100850324198,
+      "learning_rate": 7.75621471944082e-06,
+      "loss": 0.8566,
+      "step": 5230
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3508902139329648,
+      "learning_rate": 7.755349872552426e-06,
+      "loss": 0.6321,
+      "step": 5231
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4427254261765865,
+      "learning_rate": 7.754484907260513e-06,
+      "loss": 0.6233,
+      "step": 5232
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3278142149431618,
+      "learning_rate": 7.753619823602249e-06,
+      "loss": 0.7066,
+      "step": 5233
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2875359204692667,
+      "learning_rate": 7.752754621614807e-06,
+      "loss": 0.7408,
+      "step": 5234
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3267330910449868,
+      "learning_rate": 7.751889301335369e-06,
+      "loss": 0.6188,
+      "step": 5235
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3190903391168673,
+      "learning_rate": 7.75102386280112e-06,
+      "loss": 0.6984,
+      "step": 5236
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1895209999908871,
+      "learning_rate": 7.750158306049247e-06,
+      "loss": 0.6554,
+      "step": 5237
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6458885435097217,
+      "learning_rate": 7.749292631116946e-06,
+      "loss": 0.7734,
+      "step": 5238
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4406151626187234,
+      "learning_rate": 7.748426838041421e-06,
+      "loss": 0.6918,
+      "step": 5239
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4212418398834898,
+      "learning_rate": 7.747560926859872e-06,
+      "loss": 0.619,
+      "step": 5240
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9935612247318601,
+      "learning_rate": 7.74669489760951e-06,
+      "loss": 0.6237,
+      "step": 5241
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5605993942494438,
+      "learning_rate": 7.745828750327551e-06,
+      "loss": 0.6956,
+      "step": 5242
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3757552912669122,
+      "learning_rate": 7.744962485051217e-06,
+      "loss": 0.7365,
+      "step": 5243
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.417846478784282,
+      "learning_rate": 7.744096101817731e-06,
+      "loss": 0.755,
+      "step": 5244
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3486834349515509,
+      "learning_rate": 7.743229600664322e-06,
+      "loss": 0.6798,
+      "step": 5245
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3361288009758403,
+      "learning_rate": 7.742362981628229e-06,
+      "loss": 0.8296,
+      "step": 5246
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3551311410458589,
+      "learning_rate": 7.741496244746692e-06,
+      "loss": 0.6302,
+      "step": 5247
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.609104403443343,
+      "learning_rate": 7.740629390056955e-06,
+      "loss": 0.6966,
+      "step": 5248
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1815849218196066,
+      "learning_rate": 7.73976241759627e-06,
+      "loss": 0.5977,
+      "step": 5249
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3625899752726756,
+      "learning_rate": 7.738895327401891e-06,
+      "loss": 0.7263,
+      "step": 5250
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.45763700687492,
+      "learning_rate": 7.738028119511081e-06,
+      "loss": 0.6992,
+      "step": 5251
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2262434592570084,
+      "learning_rate": 7.737160793961103e-06,
+      "loss": 0.5183,
+      "step": 5252
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2104143493916868,
+      "learning_rate": 7.736293350789232e-06,
+      "loss": 0.5938,
+      "step": 5253
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2807319396586376,
+      "learning_rate": 7.73542579003274e-06,
+      "loss": 0.6122,
+      "step": 5254
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6909187237149337,
+      "learning_rate": 7.734558111728911e-06,
+      "loss": 0.6367,
+      "step": 5255
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3909224897887829,
+      "learning_rate": 7.733690315915028e-06,
+      "loss": 0.6613,
+      "step": 5256
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1977817999879694,
+      "learning_rate": 7.732822402628385e-06,
+      "loss": 0.6692,
+      "step": 5257
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3420454881117312,
+      "learning_rate": 7.731954371906279e-06,
+      "loss": 0.65,
+      "step": 5258
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.465600698416672,
+      "learning_rate": 7.731086223786006e-06,
+      "loss": 0.7046,
+      "step": 5259
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2547073209572843,
+      "learning_rate": 7.730217958304876e-06,
+      "loss": 0.7951,
+      "step": 5260
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1938118950693297,
+      "learning_rate": 7.729349575500202e-06,
+      "loss": 0.5962,
+      "step": 5261
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8954791779375517,
+      "learning_rate": 7.728481075409297e-06,
+      "loss": 0.6776,
+      "step": 5262
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1709116036312621,
+      "learning_rate": 7.727612458069482e-06,
+      "loss": 0.6982,
+      "step": 5263
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.285788099957416,
+      "learning_rate": 7.726743723518087e-06,
+      "loss": 0.62,
+      "step": 5264
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.119185216038814,
+      "learning_rate": 7.72587487179244e-06,
+      "loss": 0.6936,
+      "step": 5265
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.497128237635169,
+      "learning_rate": 7.72500590292988e-06,
+      "loss": 0.7185,
+      "step": 5266
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.661621015255879,
+      "learning_rate": 7.724136816967746e-06,
+      "loss": 0.8446,
+      "step": 5267
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5870278363533261,
+      "learning_rate": 7.723267613943388e-06,
+      "loss": 0.5778,
+      "step": 5268
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2917493769379615,
+      "learning_rate": 7.722398293894153e-06,
+      "loss": 0.7698,
+      "step": 5269
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3018253959430233,
+      "learning_rate": 7.721528856857403e-06,
+      "loss": 0.6677,
+      "step": 5270
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6834576327347741,
+      "learning_rate": 7.720659302870496e-06,
+      "loss": 0.7145,
+      "step": 5271
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2091641899824452,
+      "learning_rate": 7.719789631970798e-06,
+      "loss": 0.6874,
+      "step": 5272
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2054457863240207,
+      "learning_rate": 7.718919844195685e-06,
+      "loss": 0.6835,
+      "step": 5273
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.254985165502931,
+      "learning_rate": 7.718049939582529e-06,
+      "loss": 0.7978,
+      "step": 5274
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2281261414797537,
+      "learning_rate": 7.717179918168713e-06,
+      "loss": 0.7593,
+      "step": 5275
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.342277623987157,
+      "learning_rate": 7.716309779991625e-06,
+      "loss": 0.6316,
+      "step": 5276
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4756543799543582,
+      "learning_rate": 7.715439525088657e-06,
+      "loss": 0.8899,
+      "step": 5277
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.394084481384413,
+      "learning_rate": 7.714569153497204e-06,
+      "loss": 0.623,
+      "step": 5278
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1937393334992847,
+      "learning_rate": 7.713698665254669e-06,
+      "loss": 0.6185,
+      "step": 5279
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.282202688374508,
+      "learning_rate": 7.71282806039846e-06,
+      "loss": 0.6519,
+      "step": 5280
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4279663044873512,
+      "learning_rate": 7.711957338965986e-06,
+      "loss": 0.6554,
+      "step": 5281
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7116906050530143,
+      "learning_rate": 7.711086500994667e-06,
+      "loss": 0.6539,
+      "step": 5282
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3638767784973862,
+      "learning_rate": 7.710215546521921e-06,
+      "loss": 0.711,
+      "step": 5283
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.267202026877774,
+      "learning_rate": 7.70934447558518e-06,
+      "loss": 0.7493,
+      "step": 5284
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4326589092151416,
+      "learning_rate": 7.708473288221868e-06,
+      "loss": 0.6946,
+      "step": 5285
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3480276786447933,
+      "learning_rate": 7.70760198446943e-06,
+      "loss": 0.7322,
+      "step": 5286
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3546680853842459,
+      "learning_rate": 7.706730564365304e-06,
+      "loss": 0.6716,
+      "step": 5287
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3584473955727105,
+      "learning_rate": 7.705859027946939e-06,
+      "loss": 0.7004,
+      "step": 5288
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.417926960446653,
+      "learning_rate": 7.704987375251782e-06,
+      "loss": 0.7076,
+      "step": 5289
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3125996193597351,
+      "learning_rate": 7.704115606317296e-06,
+      "loss": 0.7722,
+      "step": 5290
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.230967505591116,
+      "learning_rate": 7.703243721180938e-06,
+      "loss": 0.6048,
+      "step": 5291
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0029991450618818,
+      "learning_rate": 7.702371719880178e-06,
+      "loss": 0.6132,
+      "step": 5292
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2165132060297872,
+      "learning_rate": 7.701499602452487e-06,
+      "loss": 0.7044,
+      "step": 5293
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1017109783193877,
+      "learning_rate": 7.70062736893534e-06,
+      "loss": 0.7998,
+      "step": 5294
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.134698053857538,
+      "learning_rate": 7.699755019366221e-06,
+      "loss": 0.6869,
+      "step": 5295
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5256466408257754,
+      "learning_rate": 7.698882553782617e-06,
+      "loss": 0.7282,
+      "step": 5296
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5437184700934938,
+      "learning_rate": 7.698009972222017e-06,
+      "loss": 0.7673,
+      "step": 5297
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3484743543404318,
+      "learning_rate": 7.697137274721922e-06,
+      "loss": 0.6129,
+      "step": 5298
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3076861689541275,
+      "learning_rate": 7.696264461319831e-06,
+      "loss": 0.6508,
+      "step": 5299
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4061833520244953,
+      "learning_rate": 7.69539153205325e-06,
+      "loss": 0.5865,
+      "step": 5300
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2006121878554985,
+      "learning_rate": 7.694518486959695e-06,
+      "loss": 0.5549,
+      "step": 5301
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4049707273530454,
+      "learning_rate": 7.693645326076677e-06,
+      "loss": 0.7718,
+      "step": 5302
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.523782167875408,
+      "learning_rate": 7.692772049441718e-06,
+      "loss": 0.72,
+      "step": 5303
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2063263722585063,
+      "learning_rate": 7.69189865709235e-06,
+      "loss": 0.5875,
+      "step": 5304
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.272666919806468,
+      "learning_rate": 7.691025149066103e-06,
+      "loss": 0.6561,
+      "step": 5305
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1919734858204662,
+      "learning_rate": 7.69015152540051e-06,
+      "loss": 0.6314,
+      "step": 5306
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.376202817835538,
+      "learning_rate": 7.689277786133113e-06,
+      "loss": 0.7212,
+      "step": 5307
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.212827293158639,
+      "learning_rate": 7.688403931301462e-06,
+      "loss": 0.6316,
+      "step": 5308
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2882904425553472,
+      "learning_rate": 7.687529960943107e-06,
+      "loss": 0.6557,
+      "step": 5309
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2318351225653479,
+      "learning_rate": 7.686655875095603e-06,
+      "loss": 0.6904,
+      "step": 5310
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3472537108403786,
+      "learning_rate": 7.685781673796515e-06,
+      "loss": 0.7916,
+      "step": 5311
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4451956660240837,
+      "learning_rate": 7.684907357083404e-06,
+      "loss": 0.675,
+      "step": 5312
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2058160031861977,
+      "learning_rate": 7.684032924993845e-06,
+      "loss": 0.6986,
+      "step": 5313
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8543313258367338,
+      "learning_rate": 7.683158377565415e-06,
+      "loss": 0.7117,
+      "step": 5314
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4259697485525848,
+      "learning_rate": 7.682283714835691e-06,
+      "loss": 0.7173,
+      "step": 5315
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1715381780459486,
+      "learning_rate": 7.681408936842266e-06,
+      "loss": 0.7554,
+      "step": 5316
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2733261610569047,
+      "learning_rate": 7.680534043622725e-06,
+      "loss": 0.6048,
+      "step": 5317
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0444621549859092,
+      "learning_rate": 7.679659035214666e-06,
+      "loss": 0.7341,
+      "step": 5318
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4408617534630699,
+      "learning_rate": 7.678783911655691e-06,
+      "loss": 0.7817,
+      "step": 5319
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4046417090849415,
+      "learning_rate": 7.677908672983404e-06,
+      "loss": 0.6315,
+      "step": 5320
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.354476860467369,
+      "learning_rate": 7.677033319235418e-06,
+      "loss": 0.6927,
+      "step": 5321
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2621237010276432,
+      "learning_rate": 7.676157850449348e-06,
+      "loss": 0.6399,
+      "step": 5322
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.1650346064472537,
+      "learning_rate": 7.675282266662816e-06,
+      "loss": 0.714,
+      "step": 5323
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3924254852045712,
+      "learning_rate": 7.674406567913447e-06,
+      "loss": 0.7349,
+      "step": 5324
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9780290141919613,
+      "learning_rate": 7.67353075423887e-06,
+      "loss": 0.5808,
+      "step": 5325
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5156917522776094,
+      "learning_rate": 7.672654825676724e-06,
+      "loss": 0.6786,
+      "step": 5326
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1905031745646713,
+      "learning_rate": 7.671778782264647e-06,
+      "loss": 0.6401,
+      "step": 5327
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.316536167152307,
+      "learning_rate": 7.670902624040286e-06,
+      "loss": 0.6449,
+      "step": 5328
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9814894641509682,
+      "learning_rate": 7.67002635104129e-06,
+      "loss": 0.7483,
+      "step": 5329
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.108339108297116,
+      "learning_rate": 7.669149963305314e-06,
+      "loss": 0.6099,
+      "step": 5330
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1963251029935544,
+      "learning_rate": 7.668273460870022e-06,
+      "loss": 0.674,
+      "step": 5331
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.701020999700932,
+      "learning_rate": 7.667396843773074e-06,
+      "loss": 0.6985,
+      "step": 5332
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2260339988866769,
+      "learning_rate": 7.666520112052145e-06,
+      "loss": 0.6956,
+      "step": 5333
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3527764939603046,
+      "learning_rate": 7.66564326574491e-06,
+      "loss": 0.7112,
+      "step": 5334
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5036328936844314,
+      "learning_rate": 7.664766304889044e-06,
+      "loss": 0.6307,
+      "step": 5335
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6506347710765816,
+      "learning_rate": 7.663889229522235e-06,
+      "loss": 0.7359,
+      "step": 5336
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.33144874365334,
+      "learning_rate": 7.663012039682174e-06,
+      "loss": 0.6342,
+      "step": 5337
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3835252793072652,
+      "learning_rate": 7.662134735406555e-06,
+      "loss": 0.615,
+      "step": 5338
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9730966709084176,
+      "learning_rate": 7.661257316733078e-06,
+      "loss": 0.583,
+      "step": 5339
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.435519654461649,
+      "learning_rate": 7.660379783699446e-06,
+      "loss": 0.8734,
+      "step": 5340
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8938367588128624,
+      "learning_rate": 7.65950213634337e-06,
+      "loss": 0.6855,
+      "step": 5341
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1611932465903652,
+      "learning_rate": 7.658624374702565e-06,
+      "loss": 0.659,
+      "step": 5342
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3803764614546168,
+      "learning_rate": 7.657746498814749e-06,
+      "loss": 0.6586,
+      "step": 5343
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4185093609487958,
+      "learning_rate": 7.656868508717648e-06,
+      "loss": 0.7068,
+      "step": 5344
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4097780484853124,
+      "learning_rate": 7.655990404448988e-06,
+      "loss": 0.6628,
+      "step": 5345
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4976899547969962,
+      "learning_rate": 7.655112186046508e-06,
+      "loss": 0.793,
+      "step": 5346
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.945457056800154,
+      "learning_rate": 7.65423385354794e-06,
+      "loss": 0.6019,
+      "step": 5347
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4976468823315296,
+      "learning_rate": 7.653355406991034e-06,
+      "loss": 0.637,
+      "step": 5348
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2209641607095358,
+      "learning_rate": 7.652476846413537e-06,
+      "loss": 0.6527,
+      "step": 5349
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.284000142796822,
+      "learning_rate": 7.651598171853203e-06,
+      "loss": 0.6428,
+      "step": 5350
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.219150296830671,
+      "learning_rate": 7.65071938334779e-06,
+      "loss": 0.6347,
+      "step": 5351
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7807870522484623,
+      "learning_rate": 7.64984048093506e-06,
+      "loss": 0.6282,
+      "step": 5352
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2700621831243124,
+      "learning_rate": 7.648961464652784e-06,
+      "loss": 0.7013,
+      "step": 5353
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.541627584731211,
+      "learning_rate": 7.648082334538735e-06,
+      "loss": 0.5763,
+      "step": 5354
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3399478939605285,
+      "learning_rate": 7.64720309063069e-06,
+      "loss": 0.6417,
+      "step": 5355
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.483405482864216,
+      "learning_rate": 7.646323732966433e-06,
+      "loss": 0.8437,
+      "step": 5356
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4184140414878321,
+      "learning_rate": 7.645444261583752e-06,
+      "loss": 0.6472,
+      "step": 5357
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.489203938515974,
+      "learning_rate": 7.644564676520439e-06,
+      "loss": 0.6581,
+      "step": 5358
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2710601433528805,
+      "learning_rate": 7.64368497781429e-06,
+      "loss": 0.7135,
+      "step": 5359
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2366942984263438,
+      "learning_rate": 7.642805165503112e-06,
+      "loss": 0.6037,
+      "step": 5360
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2039749349277493,
+      "learning_rate": 7.641925239624708e-06,
+      "loss": 0.6629,
+      "step": 5361
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9920607476722986,
+      "learning_rate": 7.641045200216896e-06,
+      "loss": 0.6714,
+      "step": 5362
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.118244184861973,
+      "learning_rate": 7.640165047317488e-06,
+      "loss": 0.7136,
+      "step": 5363
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3623581900755415,
+      "learning_rate": 7.639284780964307e-06,
+      "loss": 0.614,
+      "step": 5364
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3930888284344334,
+      "learning_rate": 7.638404401195182e-06,
+      "loss": 0.6301,
+      "step": 5365
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.390589408425436,
+      "learning_rate": 7.637523908047943e-06,
+      "loss": 0.7011,
+      "step": 5366
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.070319032525423,
+      "learning_rate": 7.63664330156043e-06,
+      "loss": 0.5308,
+      "step": 5367
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3815938240511296,
+      "learning_rate": 7.635762581770478e-06,
+      "loss": 0.6292,
+      "step": 5368
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.332667548861026,
+      "learning_rate": 7.634881748715941e-06,
+      "loss": 0.9432,
+      "step": 5369
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5114380413714792,
+      "learning_rate": 7.634000802434667e-06,
+      "loss": 0.7379,
+      "step": 5370
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4257881388255227,
+      "learning_rate": 7.633119742964511e-06,
+      "loss": 0.6754,
+      "step": 5371
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2580727447588085,
+      "learning_rate": 7.632238570343335e-06,
+      "loss": 0.6114,
+      "step": 5372
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0674927387684874,
+      "learning_rate": 7.631357284609006e-06,
+      "loss": 0.7299,
+      "step": 5373
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.669652366416178,
+      "learning_rate": 7.630475885799395e-06,
+      "loss": 0.6987,
+      "step": 5374
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.306042712176252,
+      "learning_rate": 7.629594373952375e-06,
+      "loss": 0.6854,
+      "step": 5375
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.308084616806511,
+      "learning_rate": 7.628712749105831e-06,
+      "loss": 0.702,
+      "step": 5376
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3974861707444055,
+      "learning_rate": 7.627831011297644e-06,
+      "loss": 0.7591,
+      "step": 5377
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3752244741859905,
+      "learning_rate": 7.626949160565706e-06,
+      "loss": 0.6321,
+      "step": 5378
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1635018676550337,
+      "learning_rate": 7.626067196947913e-06,
+      "loss": 0.64,
+      "step": 5379
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3557680696520502,
+      "learning_rate": 7.625185120482162e-06,
+      "loss": 0.6549,
+      "step": 5380
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5232183039606875,
+      "learning_rate": 7.624302931206362e-06,
+      "loss": 0.7462,
+      "step": 5381
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4007139318231592,
+      "learning_rate": 7.623420629158418e-06,
+      "loss": 0.7348,
+      "step": 5382
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1265468643330314,
+      "learning_rate": 7.622538214376248e-06,
+      "loss": 0.637,
+      "step": 5383
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3760287013755463,
+      "learning_rate": 7.621655686897771e-06,
+      "loss": 0.7016,
+      "step": 5384
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3722958161285408,
+      "learning_rate": 7.62077304676091e-06,
+      "loss": 0.7343,
+      "step": 5385
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.097735720741661,
+      "learning_rate": 7.619890294003593e-06,
+      "loss": 0.6001,
+      "step": 5386
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2875300402039147,
+      "learning_rate": 7.6190074286637575e-06,
+      "loss": 0.6342,
+      "step": 5387
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4749909249920894,
+      "learning_rate": 7.618124450779338e-06,
+      "loss": 0.6316,
+      "step": 5388
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.322829294124408,
+      "learning_rate": 7.617241360388282e-06,
+      "loss": 0.6509,
+      "step": 5389
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.435690770042427,
+      "learning_rate": 7.616358157528535e-06,
+      "loss": 0.7041,
+      "step": 5390
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1073015685164356,
+      "learning_rate": 7.615474842238051e-06,
+      "loss": 0.6585,
+      "step": 5391
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.245439761748477,
+      "learning_rate": 7.614591414554787e-06,
+      "loss": 0.6254,
+      "step": 5392
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0811258655894536,
+      "learning_rate": 7.613707874516708e-06,
+      "loss": 0.6654,
+      "step": 5393
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1145989870066315,
+      "learning_rate": 7.612824222161781e-06,
+      "loss": 0.5669,
+      "step": 5394
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3985713143200524,
+      "learning_rate": 7.611940457527976e-06,
+      "loss": 0.9038,
+      "step": 5395
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3152034095537428,
+      "learning_rate": 7.6110565806532745e-06,
+      "loss": 0.5317,
+      "step": 5396
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3178996229823519,
+      "learning_rate": 7.610172591575656e-06,
+      "loss": 0.5994,
+      "step": 5397
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3187805978501403,
+      "learning_rate": 7.609288490333109e-06,
+      "loss": 0.6542,
+      "step": 5398
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4815709020100734,
+      "learning_rate": 7.608404276963623e-06,
+      "loss": 0.6338,
+      "step": 5399
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2287343971309,
+      "learning_rate": 7.6075199515051955e-06,
+      "loss": 0.6624,
+      "step": 5400
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2331650387047333,
+      "learning_rate": 7.60663551399583e-06,
+      "loss": 0.7111,
+      "step": 5401
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1874218112141735,
+      "learning_rate": 7.605750964473531e-06,
+      "loss": 0.6341,
+      "step": 5402
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.350519667862822,
+      "learning_rate": 7.604866302976309e-06,
+      "loss": 0.6311,
+      "step": 5403
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.530848989178992,
+      "learning_rate": 7.60398152954218e-06,
+      "loss": 0.7476,
+      "step": 5404
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.616694641936085,
+      "learning_rate": 7.6030966442091665e-06,
+      "loss": 0.6772,
+      "step": 5405
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3874132395850227,
+      "learning_rate": 7.602211647015292e-06,
+      "loss": 0.7662,
+      "step": 5406
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2940354099826445,
+      "learning_rate": 7.601326537998589e-06,
+      "loss": 0.6747,
+      "step": 5407
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4118219364870586,
+      "learning_rate": 7.60044131719709e-06,
+      "loss": 0.7873,
+      "step": 5408
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2874602114708242,
+      "learning_rate": 7.599555984648836e-06,
+      "loss": 0.6371,
+      "step": 5409
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3554241636942943,
+      "learning_rate": 7.598670540391872e-06,
+      "loss": 0.756,
+      "step": 5410
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5198286859456132,
+      "learning_rate": 7.597784984464248e-06,
+      "loss": 0.5845,
+      "step": 5411
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4101466700475398,
+      "learning_rate": 7.596899316904018e-06,
+      "loss": 0.7016,
+      "step": 5412
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4380812141754786,
+      "learning_rate": 7.596013537749238e-06,
+      "loss": 0.6225,
+      "step": 5413
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.327883354919388,
+      "learning_rate": 7.595127647037976e-06,
+      "loss": 0.625,
+      "step": 5414
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3061855918028236,
+      "learning_rate": 7.5942416448083e-06,
+      "loss": 0.6532,
+      "step": 5415
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1919227047837229,
+      "learning_rate": 7.593355531098284e-06,
+      "loss": 0.5971,
+      "step": 5416
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3939932496961218,
+      "learning_rate": 7.592469305946003e-06,
+      "loss": 0.6492,
+      "step": 5417
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2387905912582033,
+      "learning_rate": 7.5915829693895435e-06,
+      "loss": 0.7329,
+      "step": 5418
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.369428527709501,
+      "learning_rate": 7.590696521466992e-06,
+      "loss": 0.6205,
+      "step": 5419
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.460470807807553,
+      "learning_rate": 7.589809962216439e-06,
+      "loss": 0.7936,
+      "step": 5420
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2136708406003947,
+      "learning_rate": 7.588923291675987e-06,
+      "loss": 0.682,
+      "step": 5421
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.657232407421497,
+      "learning_rate": 7.588036509883734e-06,
+      "loss": 0.6683,
+      "step": 5422
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2796924866038106,
+      "learning_rate": 7.587149616877789e-06,
+      "loss": 0.6583,
+      "step": 5423
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4975278349060235,
+      "learning_rate": 7.586262612696263e-06,
+      "loss": 0.7453,
+      "step": 5424
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4783099653031415,
+      "learning_rate": 7.585375497377271e-06,
+      "loss": 0.7776,
+      "step": 5425
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.263041513788647,
+      "learning_rate": 7.584488270958937e-06,
+      "loss": 0.7104,
+      "step": 5426
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.152117093776091,
+      "learning_rate": 7.583600933479387e-06,
+      "loss": 0.6396,
+      "step": 5427
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2909171395020376,
+      "learning_rate": 7.582713484976749e-06,
+      "loss": 0.7283,
+      "step": 5428
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1956674881487757,
+      "learning_rate": 7.5818259254891614e-06,
+      "loss": 0.6412,
+      "step": 5429
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4299344116117487,
+      "learning_rate": 7.580938255054765e-06,
+      "loss": 0.7021,
+      "step": 5430
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3490942894093114,
+      "learning_rate": 7.580050473711702e-06,
+      "loss": 0.6238,
+      "step": 5431
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.233538868069103,
+      "learning_rate": 7.579162581498125e-06,
+      "loss": 0.658,
+      "step": 5432
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5380415388264486,
+      "learning_rate": 7.5782745784521875e-06,
+      "loss": 0.701,
+      "step": 5433
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2980207160417596,
+      "learning_rate": 7.577386464612049e-06,
+      "loss": 0.6486,
+      "step": 5434
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2309890004530146,
+      "learning_rate": 7.576498240015875e-06,
+      "loss": 0.6982,
+      "step": 5435
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3384206437939188,
+      "learning_rate": 7.5756099047018315e-06,
+      "loss": 0.7301,
+      "step": 5436
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4654321499111198,
+      "learning_rate": 7.574721458708096e-06,
+      "loss": 0.6441,
+      "step": 5437
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2335381389106668,
+      "learning_rate": 7.573832902072844e-06,
+      "loss": 0.6159,
+      "step": 5438
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4843292666385246,
+      "learning_rate": 7.572944234834261e-06,
+      "loss": 0.7161,
+      "step": 5439
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4117501499448575,
+      "learning_rate": 7.572055457030533e-06,
+      "loss": 0.6639,
+      "step": 5440
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4126690977410767,
+      "learning_rate": 7.5711665686998545e-06,
+      "loss": 0.7678,
+      "step": 5441
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3312068410112317,
+      "learning_rate": 7.570277569880422e-06,
+      "loss": 0.6531,
+      "step": 5442
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6828454240286173,
+      "learning_rate": 7.569388460610437e-06,
+      "loss": 0.7069,
+      "step": 5443
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3685918446416179,
+      "learning_rate": 7.568499240928109e-06,
+      "loss": 0.6029,
+      "step": 5444
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3524278267662977,
+      "learning_rate": 7.5676099108716485e-06,
+      "loss": 0.5866,
+      "step": 5445
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2679206782991326,
+      "learning_rate": 7.5667204704792706e-06,
+      "loss": 0.6817,
+      "step": 5446
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1389546560905075,
+      "learning_rate": 7.565830919789198e-06,
+      "loss": 0.6069,
+      "step": 5447
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3315084938421895,
+      "learning_rate": 7.564941258839655e-06,
+      "loss": 0.6746,
+      "step": 5448
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2012257007035452,
+      "learning_rate": 7.5640514876688765e-06,
+      "loss": 0.6105,
+      "step": 5449
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2527042185164534,
+      "learning_rate": 7.563161606315095e-06,
+      "loss": 0.6427,
+      "step": 5450
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3528631518773806,
+      "learning_rate": 7.5622716148165475e-06,
+      "loss": 0.6397,
+      "step": 5451
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0959288539221244,
+      "learning_rate": 7.561381513211486e-06,
+      "loss": 0.6339,
+      "step": 5452
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.463921015927053,
+      "learning_rate": 7.5604913015381535e-06,
+      "loss": 0.6271,
+      "step": 5453
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3259423276687072,
+      "learning_rate": 7.559600979834809e-06,
+      "loss": 0.6048,
+      "step": 5454
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.15305256629453,
+      "learning_rate": 7.5587105481397095e-06,
+      "loss": 0.6198,
+      "step": 5455
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.110496684235668,
+      "learning_rate": 7.557820006491118e-06,
+      "loss": 0.7431,
+      "step": 5456
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3899340016528643,
+      "learning_rate": 7.5569293549273046e-06,
+      "loss": 0.7779,
+      "step": 5457
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3265228455709337,
+      "learning_rate": 7.556038593486543e-06,
+      "loss": 0.6273,
+      "step": 5458
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.147271760785399,
+      "learning_rate": 7.555147722207111e-06,
+      "loss": 0.5812,
+      "step": 5459
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3656685012517062,
+      "learning_rate": 7.554256741127291e-06,
+      "loss": 0.6271,
+      "step": 5460
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2293523602294,
+      "learning_rate": 7.553365650285368e-06,
+      "loss": 0.6495,
+      "step": 5461
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2380158137073605,
+      "learning_rate": 7.5524744497196376e-06,
+      "loss": 0.7713,
+      "step": 5462
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.333964609356168,
+      "learning_rate": 7.5515831394683954e-06,
+      "loss": 0.6289,
+      "step": 5463
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4601312045448283,
+      "learning_rate": 7.550691719569944e-06,
+      "loss": 0.7219,
+      "step": 5464
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.247219711707885,
+      "learning_rate": 7.549800190062586e-06,
+      "loss": 0.6639,
+      "step": 5465
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.09506264906139,
+      "learning_rate": 7.548908550984637e-06,
+      "loss": 0.5514,
+      "step": 5466
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0749138125448126,
+      "learning_rate": 7.548016802374412e-06,
+      "loss": 0.6052,
+      "step": 5467
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.493105914628131,
+      "learning_rate": 7.547124944270229e-06,
+      "loss": 0.7082,
+      "step": 5468
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2943054240539944,
+      "learning_rate": 7.546232976710413e-06,
+      "loss": 0.6639,
+      "step": 5469
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3522981772813125,
+      "learning_rate": 7.545340899733297e-06,
+      "loss": 0.7764,
+      "step": 5470
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2356744736984855,
+      "learning_rate": 7.544448713377213e-06,
+      "loss": 0.6562,
+      "step": 5471
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6808030606648183,
+      "learning_rate": 7.543556417680502e-06,
+      "loss": 0.7511,
+      "step": 5472
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1730009087726694,
+      "learning_rate": 7.542664012681507e-06,
+      "loss": 0.6002,
+      "step": 5473
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9580789543288842,
+      "learning_rate": 7.541771498418575e-06,
+      "loss": 0.603,
+      "step": 5474
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2531247272354646,
+      "learning_rate": 7.540878874930062e-06,
+      "loss": 0.667,
+      "step": 5475
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5413419630507208,
+      "learning_rate": 7.539986142254324e-06,
+      "loss": 0.768,
+      "step": 5476
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5234532105875198,
+      "learning_rate": 7.539093300429727e-06,
+      "loss": 0.6838,
+      "step": 5477
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3727545616047396,
+      "learning_rate": 7.538200349494633e-06,
+      "loss": 0.6182,
+      "step": 5478
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.39084603757103,
+      "learning_rate": 7.537307289487419e-06,
+      "loss": 0.7657,
+      "step": 5479
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5497931125540774,
+      "learning_rate": 7.536414120446458e-06,
+      "loss": 0.7135,
+      "step": 5480
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.551807925904203,
+      "learning_rate": 7.535520842410136e-06,
+      "loss": 0.7658,
+      "step": 5481
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3966951820873863,
+      "learning_rate": 7.534627455416835e-06,
+      "loss": 0.7709,
+      "step": 5482
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0050267683639909,
+      "learning_rate": 7.533733959504947e-06,
+      "loss": 0.4884,
+      "step": 5483
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6833832162158604,
+      "learning_rate": 7.532840354712868e-06,
+      "loss": 0.6639,
+      "step": 5484
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1287023812242465,
+      "learning_rate": 7.531946641078998e-06,
+      "loss": 0.6305,
+      "step": 5485
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1927902637521999,
+      "learning_rate": 7.531052818641743e-06,
+      "loss": 0.6093,
+      "step": 5486
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.346292444474493,
+      "learning_rate": 7.530158887439509e-06,
+      "loss": 0.6605,
+      "step": 5487
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2263069261492094,
+      "learning_rate": 7.529264847510715e-06,
+      "loss": 0.6281,
+      "step": 5488
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3164744232123209,
+      "learning_rate": 7.5283706988937765e-06,
+      "loss": 0.7158,
+      "step": 5489
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3791042062935737,
+      "learning_rate": 7.5274764416271185e-06,
+      "loss": 0.7045,
+      "step": 5490
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1854964480722898,
+      "learning_rate": 7.526582075749169e-06,
+      "loss": 0.666,
+      "step": 5491
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6655334061281628,
+      "learning_rate": 7.525687601298361e-06,
+      "loss": 0.8032,
+      "step": 5492
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4102656197987686,
+      "learning_rate": 7.5247930183131305e-06,
+      "loss": 0.6499,
+      "step": 5493
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4727064524602023,
+      "learning_rate": 7.523898326831921e-06,
+      "loss": 0.7443,
+      "step": 5494
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6556704072550783,
+      "learning_rate": 7.52300352689318e-06,
+      "loss": 0.7743,
+      "step": 5495
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0257299913207663,
+      "learning_rate": 7.522108618535359e-06,
+      "loss": 0.6502,
+      "step": 5496
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0865832605984407,
+      "learning_rate": 7.521213601796913e-06,
+      "loss": 0.6544,
+      "step": 5497
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3030202696516417,
+      "learning_rate": 7.520318476716303e-06,
+      "loss": 0.6063,
+      "step": 5498
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4860868629635093,
+      "learning_rate": 7.5194232433319955e-06,
+      "loss": 0.5709,
+      "step": 5499
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2783619608821013,
+      "learning_rate": 7.51852790168246e-06,
+      "loss": 0.7037,
+      "step": 5500
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5562769294571714,
+      "learning_rate": 7.517632451806172e-06,
+      "loss": 0.7351,
+      "step": 5501
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3710515207596348,
+      "learning_rate": 7.516736893741611e-06,
+      "loss": 0.7365,
+      "step": 5502
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0412340029495464,
+      "learning_rate": 7.515841227527258e-06,
+      "loss": 0.5458,
+      "step": 5503
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3414307885299115,
+      "learning_rate": 7.514945453201608e-06,
+      "loss": 0.6612,
+      "step": 5504
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3222041415208223,
+      "learning_rate": 7.514049570803148e-06,
+      "loss": 0.748,
+      "step": 5505
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2661464037677377,
+      "learning_rate": 7.5131535803703805e-06,
+      "loss": 0.6469,
+      "step": 5506
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3875850787956054,
+      "learning_rate": 7.512257481941805e-06,
+      "loss": 0.6564,
+      "step": 5507
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.281421797130802,
+      "learning_rate": 7.511361275555931e-06,
+      "loss": 0.7106,
+      "step": 5508
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.526531187941366,
+      "learning_rate": 7.510464961251271e-06,
+      "loss": 0.7007,
+      "step": 5509
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2316726302091079,
+      "learning_rate": 7.50956853906634e-06,
+      "loss": 0.7553,
+      "step": 5510
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2909405759808519,
+      "learning_rate": 7.508672009039659e-06,
+      "loss": 0.6713,
+      "step": 5511
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0266989454284827,
+      "learning_rate": 7.507775371209754e-06,
+      "loss": 0.6718,
+      "step": 5512
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1869006859587419,
+      "learning_rate": 7.5068786256151585e-06,
+      "loss": 0.6487,
+      "step": 5513
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3034742741203453,
+      "learning_rate": 7.505981772294404e-06,
+      "loss": 0.5771,
+      "step": 5514
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4211200512189794,
+      "learning_rate": 7.505084811286032e-06,
+      "loss": 0.7339,
+      "step": 5515
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.49058338126791,
+      "learning_rate": 7.5041877426285856e-06,
+      "loss": 0.6995,
+      "step": 5516
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.019827480472923,
+      "learning_rate": 7.503290566360615e-06,
+      "loss": 0.6426,
+      "step": 5517
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.150160880659869,
+      "learning_rate": 7.502393282520675e-06,
+      "loss": 0.7105,
+      "step": 5518
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.094097063826031,
+      "learning_rate": 7.501495891147322e-06,
+      "loss": 0.5795,
+      "step": 5519
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2026467017615508,
+      "learning_rate": 7.500598392279119e-06,
+      "loss": 0.65,
+      "step": 5520
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.30552592071088,
+      "learning_rate": 7.499700785954633e-06,
+      "loss": 0.6097,
+      "step": 5521
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2755549008584683,
+      "learning_rate": 7.498803072212439e-06,
+      "loss": 0.6119,
+      "step": 5522
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1391186012004428,
+      "learning_rate": 7.49790525109111e-06,
+      "loss": 0.5243,
+      "step": 5523
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3753792892309795,
+      "learning_rate": 7.497007322629231e-06,
+      "loss": 0.7544,
+      "step": 5524
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4119093468392927,
+      "learning_rate": 7.496109286865384e-06,
+      "loss": 0.7141,
+      "step": 5525
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2242673029372562,
+      "learning_rate": 7.495211143838164e-06,
+      "loss": 0.718,
+      "step": 5526
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4441154704714816,
+      "learning_rate": 7.494312893586163e-06,
+      "loss": 0.7513,
+      "step": 5527
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2365463967094987,
+      "learning_rate": 7.493414536147983e-06,
+      "loss": 0.6888,
+      "step": 5528
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1849250498257418,
+      "learning_rate": 7.492516071562226e-06,
+      "loss": 0.6887,
+      "step": 5529
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2268176002563542,
+      "learning_rate": 7.491617499867502e-06,
+      "loss": 0.6249,
+      "step": 5530
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3001246927322934,
+      "learning_rate": 7.4907188211024254e-06,
+      "loss": 0.7194,
+      "step": 5531
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3928647854941754,
+      "learning_rate": 7.489820035305613e-06,
+      "loss": 0.6175,
+      "step": 5532
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4033171825889623,
+      "learning_rate": 7.488921142515691e-06,
+      "loss": 0.6739,
+      "step": 5533
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9501556410033641,
+      "learning_rate": 7.488022142771282e-06,
+      "loss": 0.604,
+      "step": 5534
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0254761364204956,
+      "learning_rate": 7.4871230361110214e-06,
+      "loss": 0.5917,
+      "step": 5535
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2783310829384311,
+      "learning_rate": 7.486223822573545e-06,
+      "loss": 0.6495,
+      "step": 5536
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0605761191939327,
+      "learning_rate": 7.485324502197494e-06,
+      "loss": 0.6983,
+      "step": 5537
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1480288995696326,
+      "learning_rate": 7.484425075021513e-06,
+      "loss": 0.6935,
+      "step": 5538
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5422744784849918,
+      "learning_rate": 7.483525541084253e-06,
+      "loss": 0.7307,
+      "step": 5539
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4550306497176428,
+      "learning_rate": 7.482625900424372e-06,
+      "loss": 0.6946,
+      "step": 5540
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3269255528385073,
+      "learning_rate": 7.481726153080524e-06,
+      "loss": 0.6627,
+      "step": 5541
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.279726544314,
+      "learning_rate": 7.480826299091378e-06,
+      "loss": 0.7789,
+      "step": 5542
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3797666051546464,
+      "learning_rate": 7.4799263384956e-06,
+      "loss": 0.6901,
+      "step": 5543
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.340190971649125,
+      "learning_rate": 7.479026271331864e-06,
+      "loss": 0.6996,
+      "step": 5544
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9086178191419726,
+      "learning_rate": 7.478126097638848e-06,
+      "loss": 0.5621,
+      "step": 5545
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3139251928199962,
+      "learning_rate": 7.477225817455236e-06,
+      "loss": 0.5353,
+      "step": 5546
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.112783646711226,
+      "learning_rate": 7.476325430819713e-06,
+      "loss": 0.6471,
+      "step": 5547
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1772352015734389,
+      "learning_rate": 7.47542493777097e-06,
+      "loss": 0.6465,
+      "step": 5548
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.449380707528157,
+      "learning_rate": 7.4745243383477055e-06,
+      "loss": 0.7453,
+      "step": 5549
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2127109697978196,
+      "learning_rate": 7.473623632588618e-06,
+      "loss": 0.6396,
+      "step": 5550
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1559595157243086,
+      "learning_rate": 7.472722820532414e-06,
+      "loss": 0.6212,
+      "step": 5551
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3721340441907597,
+      "learning_rate": 7.471821902217804e-06,
+      "loss": 0.6888,
+      "step": 5552
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3047160479374917,
+      "learning_rate": 7.470920877683503e-06,
+      "loss": 0.6073,
+      "step": 5553
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4726111126620338,
+      "learning_rate": 7.470019746968226e-06,
+      "loss": 0.7404,
+      "step": 5554
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4089225175988387,
+      "learning_rate": 7.469118510110702e-06,
+      "loss": 0.7051,
+      "step": 5555
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2702553360829658,
+      "learning_rate": 7.468217167149656e-06,
+      "loss": 0.5774,
+      "step": 5556
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5823666847354403,
+      "learning_rate": 7.467315718123821e-06,
+      "loss": 0.8398,
+      "step": 5557
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6915713319532708,
+      "learning_rate": 7.466414163071934e-06,
+      "loss": 0.7419,
+      "step": 5558
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4246966956568756,
+      "learning_rate": 7.4655125020327376e-06,
+      "loss": 0.7314,
+      "step": 5559
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2378437971594773,
+      "learning_rate": 7.464610735044979e-06,
+      "loss": 0.5518,
+      "step": 5560
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0952414319904804,
+      "learning_rate": 7.463708862147408e-06,
+      "loss": 0.5752,
+      "step": 5561
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0276661820460067,
+      "learning_rate": 7.462806883378779e-06,
+      "loss": 0.6935,
+      "step": 5562
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6441716113461327,
+      "learning_rate": 7.4619047987778536e-06,
+      "loss": 0.6177,
+      "step": 5563
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2501308576313062,
+      "learning_rate": 7.461002608383396e-06,
+      "loss": 0.7044,
+      "step": 5564
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0763798179538258,
+      "learning_rate": 7.460100312234176e-06,
+      "loss": 0.5762,
+      "step": 5565
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4753784418587805,
+      "learning_rate": 7.4591979103689675e-06,
+      "loss": 0.6532,
+      "step": 5566
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.591391286238718,
+      "learning_rate": 7.458295402826546e-06,
+      "loss": 0.7808,
+      "step": 5567
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.211240168222406,
+      "learning_rate": 7.4573927896456966e-06,
+      "loss": 0.6375,
+      "step": 5568
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2945290713102773,
+      "learning_rate": 7.456490070865206e-06,
+      "loss": 0.6915,
+      "step": 5569
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4391477192548026,
+      "learning_rate": 7.455587246523868e-06,
+      "loss": 0.6853,
+      "step": 5570
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.305104511391397,
+      "learning_rate": 7.454684316660475e-06,
+      "loss": 0.6964,
+      "step": 5571
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.260863053695578,
+      "learning_rate": 7.453781281313831e-06,
+      "loss": 0.7979,
+      "step": 5572
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.242976193939855,
+      "learning_rate": 7.45287814052274e-06,
+      "loss": 0.6829,
+      "step": 5573
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4793632270308257,
+      "learning_rate": 7.4519748943260126e-06,
+      "loss": 0.7108,
+      "step": 5574
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7002503285615695,
+      "learning_rate": 7.4510715427624635e-06,
+      "loss": 0.6411,
+      "step": 5575
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3283420610971677,
+      "learning_rate": 7.450168085870911e-06,
+      "loss": 0.7272,
+      "step": 5576
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2929999339075002,
+      "learning_rate": 7.4492645236901795e-06,
+      "loss": 0.7748,
+      "step": 5577
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3354583891364142,
+      "learning_rate": 7.4483608562590955e-06,
+      "loss": 0.5999,
+      "step": 5578
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.185179822691518,
+      "learning_rate": 7.447457083616494e-06,
+      "loss": 0.7339,
+      "step": 5579
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.233456149195898,
+      "learning_rate": 7.4465532058012105e-06,
+      "loss": 0.6435,
+      "step": 5580
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1079765879851198,
+      "learning_rate": 7.445649222852086e-06,
+      "loss": 0.5919,
+      "step": 5581
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3875490114472775,
+      "learning_rate": 7.444745134807969e-06,
+      "loss": 0.7639,
+      "step": 5582
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3549345500778578,
+      "learning_rate": 7.443840941707709e-06,
+      "loss": 0.6593,
+      "step": 5583
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1686744513505933,
+      "learning_rate": 7.44293664359016e-06,
+      "loss": 0.7457,
+      "step": 5584
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3817171042030494,
+      "learning_rate": 7.442032240494184e-06,
+      "loss": 0.5925,
+      "step": 5585
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.385145519758471,
+      "learning_rate": 7.441127732458642e-06,
+      "loss": 0.6992,
+      "step": 5586
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2652713923677785,
+      "learning_rate": 7.440223119522405e-06,
+      "loss": 0.7676,
+      "step": 5587
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2369915525842279,
+      "learning_rate": 7.439318401724347e-06,
+      "loss": 0.6557,
+      "step": 5588
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3129690187261518,
+      "learning_rate": 7.438413579103344e-06,
+      "loss": 0.7312,
+      "step": 5589
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.375916473250459,
+      "learning_rate": 7.4375086516982775e-06,
+      "loss": 0.7224,
+      "step": 5590
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1570932868089745,
+      "learning_rate": 7.436603619548038e-06,
+      "loss": 0.6619,
+      "step": 5591
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2132714015396646,
+      "learning_rate": 7.4356984826915135e-06,
+      "loss": 0.6258,
+      "step": 5592
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3199196910874522,
+      "learning_rate": 7.434793241167601e-06,
+      "loss": 0.5898,
+      "step": 5593
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3761782730748282,
+      "learning_rate": 7.433887895015199e-06,
+      "loss": 0.7043,
+      "step": 5594
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3814092784169811,
+      "learning_rate": 7.432982444273215e-06,
+      "loss": 0.7093,
+      "step": 5595
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1943307608155183,
+      "learning_rate": 7.432076888980556e-06,
+      "loss": 0.6636,
+      "step": 5596
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1215318343214509,
+      "learning_rate": 7.431171229176136e-06,
+      "loss": 0.5849,
+      "step": 5597
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3241542044353871,
+      "learning_rate": 7.4302654648988755e-06,
+      "loss": 0.7116,
+      "step": 5598
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1763241784384597,
+      "learning_rate": 7.429359596187694e-06,
+      "loss": 0.6362,
+      "step": 5599
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.305644617586113,
+      "learning_rate": 7.428453623081522e-06,
+      "loss": 0.7736,
+      "step": 5600
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3806857064794789,
+      "learning_rate": 7.427547545619288e-06,
+      "loss": 0.7219,
+      "step": 5601
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.262266663597947,
+      "learning_rate": 7.4266413638399284e-06,
+      "loss": 0.5959,
+      "step": 5602
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3180393442253369,
+      "learning_rate": 7.425735077782387e-06,
+      "loss": 0.844,
+      "step": 5603
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2738385565842487,
+      "learning_rate": 7.424828687485606e-06,
+      "loss": 0.6156,
+      "step": 5604
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0816884356995304,
+      "learning_rate": 7.423922192988536e-06,
+      "loss": 0.5067,
+      "step": 5605
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5586236673432026,
+      "learning_rate": 7.423015594330131e-06,
+      "loss": 0.6909,
+      "step": 5606
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.175483270926632,
+      "learning_rate": 7.422108891549349e-06,
+      "loss": 0.6467,
+      "step": 5607
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0767810155155109,
+      "learning_rate": 7.421202084685155e-06,
+      "loss": 0.6796,
+      "step": 5608
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2984689047781954,
+      "learning_rate": 7.420295173776515e-06,
+      "loss": 0.7153,
+      "step": 5609
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.376801914291828,
+      "learning_rate": 7.4193881588624e-06,
+      "loss": 0.5983,
+      "step": 5610
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2942061238351932,
+      "learning_rate": 7.418481039981787e-06,
+      "loss": 0.5993,
+      "step": 5611
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4385839247636054,
+      "learning_rate": 7.41757381717366e-06,
+      "loss": 0.7708,
+      "step": 5612
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4180141100531525,
+      "learning_rate": 7.416666490476999e-06,
+      "loss": 0.6857,
+      "step": 5613
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3006291080746526,
+      "learning_rate": 7.415759059930799e-06,
+      "loss": 0.7048,
+      "step": 5614
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2865854331990902,
+      "learning_rate": 7.41485152557405e-06,
+      "loss": 0.6378,
+      "step": 5615
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.346074291321166,
+      "learning_rate": 7.4139438874457536e-06,
+      "loss": 0.7487,
+      "step": 5616
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.290724892563301,
+      "learning_rate": 7.413036145584913e-06,
+      "loss": 0.7229,
+      "step": 5617
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7303530592356307,
+      "learning_rate": 7.412128300030534e-06,
+      "loss": 0.7222,
+      "step": 5618
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4628483742016303,
+      "learning_rate": 7.411220350821631e-06,
+      "loss": 0.6086,
+      "step": 5619
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.439760865583589,
+      "learning_rate": 7.410312297997219e-06,
+      "loss": 0.7006,
+      "step": 5620
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5251878515846369,
+      "learning_rate": 7.409404141596319e-06,
+      "loss": 0.6507,
+      "step": 5621
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.451553344459204,
+      "learning_rate": 7.4084958816579575e-06,
+      "loss": 0.6981,
+      "step": 5622
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.342489088191663,
+      "learning_rate": 7.407587518221164e-06,
+      "loss": 0.6362,
+      "step": 5623
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2957008407186041,
+      "learning_rate": 7.406679051324972e-06,
+      "loss": 0.7019,
+      "step": 5624
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3345696232338986,
+      "learning_rate": 7.4057704810084215e-06,
+      "loss": 0.7085,
+      "step": 5625
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5745380796289066,
+      "learning_rate": 7.404861807310556e-06,
+      "loss": 0.7197,
+      "step": 5626
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2038786468862717,
+      "learning_rate": 7.4039530302704236e-06,
+      "loss": 0.6265,
+      "step": 5627
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4541207941953243,
+      "learning_rate": 7.403044149927074e-06,
+      "loss": 0.7355,
+      "step": 5628
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.09400376765448,
+      "learning_rate": 7.402135166319567e-06,
+      "loss": 0.726,
+      "step": 5629
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1513343060527728,
+      "learning_rate": 7.40122607948696e-06,
+      "loss": 0.6403,
+      "step": 5630
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1287757711959934,
+      "learning_rate": 7.400316889468322e-06,
+      "loss": 0.5817,
+      "step": 5631
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3882891244400082,
+      "learning_rate": 7.399407596302721e-06,
+      "loss": 0.723,
+      "step": 5632
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2322441929966288,
+      "learning_rate": 7.398498200029232e-06,
+      "loss": 0.6319,
+      "step": 5633
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.168846484833431,
+      "learning_rate": 7.397588700686933e-06,
+      "loss": 0.5948,
+      "step": 5634
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4686675903263744,
+      "learning_rate": 7.396679098314908e-06,
+      "loss": 0.7102,
+      "step": 5635
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.469140964843181,
+      "learning_rate": 7.395769392952246e-06,
+      "loss": 0.7384,
+      "step": 5636
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2607615081522825,
+      "learning_rate": 7.394859584638035e-06,
+      "loss": 0.6817,
+      "step": 5637
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3715724970665775,
+      "learning_rate": 7.393949673411375e-06,
+      "loss": 0.7515,
+      "step": 5638
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4977692111142114,
+      "learning_rate": 7.393039659311366e-06,
+      "loss": 0.7723,
+      "step": 5639
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.319174079145759,
+      "learning_rate": 7.392129542377114e-06,
+      "loss": 0.7036,
+      "step": 5640
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.331476865791734,
+      "learning_rate": 7.391219322647728e-06,
+      "loss": 0.7705,
+      "step": 5641
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4081488400034292,
+      "learning_rate": 7.390309000162321e-06,
+      "loss": 0.7285,
+      "step": 5642
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3150498956603547,
+      "learning_rate": 7.389398574960014e-06,
+      "loss": 0.5877,
+      "step": 5643
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5810463883178993,
+      "learning_rate": 7.388488047079927e-06,
+      "loss": 0.7025,
+      "step": 5644
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3881492286107608,
+      "learning_rate": 7.387577416561191e-06,
+      "loss": 0.5777,
+      "step": 5645
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5323556803149556,
+      "learning_rate": 7.386666683442936e-06,
+      "loss": 0.7081,
+      "step": 5646
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3268721715504923,
+      "learning_rate": 7.3857558477642966e-06,
+      "loss": 0.6317,
+      "step": 5647
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1860378748719262,
+      "learning_rate": 7.3848449095644155e-06,
+      "loss": 0.6836,
+      "step": 5648
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.468555809665193,
+      "learning_rate": 7.383933868882438e-06,
+      "loss": 0.8246,
+      "step": 5649
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0309204818023092,
+      "learning_rate": 7.383022725757514e-06,
+      "loss": 0.6494,
+      "step": 5650
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2067124173381851,
+      "learning_rate": 7.382111480228793e-06,
+      "loss": 0.6614,
+      "step": 5651
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4407793520969032,
+      "learning_rate": 7.381200132335438e-06,
+      "loss": 0.7104,
+      "step": 5652
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9006709619274917,
+      "learning_rate": 7.38028868211661e-06,
+      "loss": 0.6028,
+      "step": 5653
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2429645414623824,
+      "learning_rate": 7.379377129611478e-06,
+      "loss": 0.6058,
+      "step": 5654
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4344423417568486,
+      "learning_rate": 7.378465474859211e-06,
+      "loss": 0.5901,
+      "step": 5655
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2094333237256938,
+      "learning_rate": 7.377553717898983e-06,
+      "loss": 0.7247,
+      "step": 5656
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5612697427315556,
+      "learning_rate": 7.376641858769978e-06,
+      "loss": 0.7168,
+      "step": 5657
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8875547556709447,
+      "learning_rate": 7.375729897511379e-06,
+      "loss": 0.7768,
+      "step": 5658
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6092309763143262,
+      "learning_rate": 7.374817834162378e-06,
+      "loss": 0.7,
+      "step": 5659
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4286743139905222,
+      "learning_rate": 7.3739056687621616e-06,
+      "loss": 0.7448,
+      "step": 5660
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0698975927027576,
+      "learning_rate": 7.3729934013499335e-06,
+      "loss": 0.7716,
+      "step": 5661
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4721472020135118,
+      "learning_rate": 7.372081031964893e-06,
+      "loss": 0.6649,
+      "step": 5662
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5841554707155487,
+      "learning_rate": 7.37116856064625e-06,
+      "loss": 0.6344,
+      "step": 5663
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3612718671344781,
+      "learning_rate": 7.3702559874332125e-06,
+      "loss": 0.6059,
+      "step": 5664
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.137620726112591,
+      "learning_rate": 7.369343312364994e-06,
+      "loss": 0.6851,
+      "step": 5665
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4527815876305268,
+      "learning_rate": 7.368430535480818e-06,
+      "loss": 0.7125,
+      "step": 5666
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3714387852266283,
+      "learning_rate": 7.367517656819907e-06,
+      "loss": 0.6537,
+      "step": 5667
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4136482075566776,
+      "learning_rate": 7.366604676421491e-06,
+      "loss": 0.6578,
+      "step": 5668
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1543467563004115,
+      "learning_rate": 7.3656915943247984e-06,
+      "loss": 0.6028,
+      "step": 5669
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.524858262428567,
+      "learning_rate": 7.364778410569071e-06,
+      "loss": 0.6209,
+      "step": 5670
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0191271299530196,
+      "learning_rate": 7.363865125193549e-06,
+      "loss": 0.7529,
+      "step": 5671
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2528240012811316,
+      "learning_rate": 7.362951738237477e-06,
+      "loss": 0.6605,
+      "step": 5672
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5182254036891436,
+      "learning_rate": 7.362038249740108e-06,
+      "loss": 0.812,
+      "step": 5673
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4786104343817952,
+      "learning_rate": 7.3611246597406925e-06,
+      "loss": 0.7565,
+      "step": 5674
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3041021373118034,
+      "learning_rate": 7.360210968278494e-06,
+      "loss": 0.6036,
+      "step": 5675
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2286296318627605,
+      "learning_rate": 7.359297175392772e-06,
+      "loss": 0.5741,
+      "step": 5676
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.333498041170409,
+      "learning_rate": 7.358383281122797e-06,
+      "loss": 0.5681,
+      "step": 5677
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2626512792280231,
+      "learning_rate": 7.35746928550784e-06,
+      "loss": 0.6858,
+      "step": 5678
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2575016594374613,
+      "learning_rate": 7.356555188587178e-06,
+      "loss": 0.7555,
+      "step": 5679
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2933437970813773,
+      "learning_rate": 7.355640990400089e-06,
+      "loss": 0.6493,
+      "step": 5680
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.220721747041767,
+      "learning_rate": 7.354726690985862e-06,
+      "loss": 0.6941,
+      "step": 5681
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3184816861427069,
+      "learning_rate": 7.353812290383785e-06,
+      "loss": 0.6198,
+      "step": 5682
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5859321575549037,
+      "learning_rate": 7.352897788633151e-06,
+      "loss": 0.7735,
+      "step": 5683
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3534247264481427,
+      "learning_rate": 7.351983185773259e-06,
+      "loss": 0.7564,
+      "step": 5684
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2629664245957004,
+      "learning_rate": 7.351068481843411e-06,
+      "loss": 0.6369,
+      "step": 5685
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3505005188902146,
+      "learning_rate": 7.3501536768829154e-06,
+      "loss": 0.9603,
+      "step": 5686
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3318889706737163,
+      "learning_rate": 7.349238770931081e-06,
+      "loss": 0.6363,
+      "step": 5687
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1703250888732073,
+      "learning_rate": 7.348323764027225e-06,
+      "loss": 0.6657,
+      "step": 5688
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.222043673974498,
+      "learning_rate": 7.347408656210666e-06,
+      "loss": 0.6273,
+      "step": 5689
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2836591157903876,
+      "learning_rate": 7.346493447520731e-06,
+      "loss": 0.6606,
+      "step": 5690
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.408917982953239,
+      "learning_rate": 7.345578137996745e-06,
+      "loss": 0.5858,
+      "step": 5691
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4094575684417892,
+      "learning_rate": 7.344662727678043e-06,
+      "loss": 0.6327,
+      "step": 5692
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3577382440351453,
+      "learning_rate": 7.343747216603962e-06,
+      "loss": 0.7663,
+      "step": 5693
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2512081027912278,
+      "learning_rate": 7.342831604813844e-06,
+      "loss": 0.6888,
+      "step": 5694
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3963257048718716,
+      "learning_rate": 7.341915892347033e-06,
+      "loss": 0.689,
+      "step": 5695
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9973521207436392,
+      "learning_rate": 7.341000079242881e-06,
+      "loss": 0.6643,
+      "step": 5696
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.60469526032491,
+      "learning_rate": 7.340084165540742e-06,
+      "loss": 0.8206,
+      "step": 5697
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0072946401350187,
+      "learning_rate": 7.339168151279974e-06,
+      "loss": 0.6982,
+      "step": 5698
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3962978065226648,
+      "learning_rate": 7.338252036499941e-06,
+      "loss": 0.7059,
+      "step": 5699
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.928126997574276,
+      "learning_rate": 7.3373358212400104e-06,
+      "loss": 0.6895,
+      "step": 5700
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3514962930810077,
+      "learning_rate": 7.336419505539555e-06,
+      "loss": 0.7201,
+      "step": 5701
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2809110051959478,
+      "learning_rate": 7.33550308943795e-06,
+      "loss": 0.7392,
+      "step": 5702
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.410405660510508,
+      "learning_rate": 7.334586572974573e-06,
+      "loss": 0.7166,
+      "step": 5703
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1977281920855996,
+      "learning_rate": 7.333669956188815e-06,
+      "loss": 0.6516,
+      "step": 5704
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0927024044014138,
+      "learning_rate": 7.332753239120061e-06,
+      "loss": 0.7175,
+      "step": 5705
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4732408316870542,
+      "learning_rate": 7.331836421807704e-06,
+      "loss": 0.7542,
+      "step": 5706
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3973104855144627,
+      "learning_rate": 7.330919504291143e-06,
+      "loss": 0.621,
+      "step": 5707
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4796569321435522,
+      "learning_rate": 7.33000248660978e-06,
+      "loss": 0.703,
+      "step": 5708
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3108359346499778,
+      "learning_rate": 7.3290853688030196e-06,
+      "loss": 0.759,
+      "step": 5709
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2836637443831143,
+      "learning_rate": 7.328168150910275e-06,
+      "loss": 0.6324,
+      "step": 5710
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.348300100905622,
+      "learning_rate": 7.32725083297096e-06,
+      "loss": 0.6673,
+      "step": 5711
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.195872678132146,
+      "learning_rate": 7.326333415024494e-06,
+      "loss": 0.56,
+      "step": 5712
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5830730206123804,
+      "learning_rate": 7.325415897110299e-06,
+      "loss": 0.83,
+      "step": 5713
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6266931490851817,
+      "learning_rate": 7.324498279267803e-06,
+      "loss": 0.6584,
+      "step": 5714
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0041428841962377,
+      "learning_rate": 7.323580561536442e-06,
+      "loss": 0.6248,
+      "step": 5715
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.692785988695421,
+      "learning_rate": 7.3226627439556466e-06,
+      "loss": 0.738,
+      "step": 5716
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.023535904173149,
+      "learning_rate": 7.321744826564862e-06,
+      "loss": 0.6986,
+      "step": 5717
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3066084448698831,
+      "learning_rate": 7.3208268094035295e-06,
+      "loss": 0.63,
+      "step": 5718
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2439671829920096,
+      "learning_rate": 7.319908692511103e-06,
+      "loss": 0.5886,
+      "step": 5719
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3862236923271403,
+      "learning_rate": 7.318990475927033e-06,
+      "loss": 0.666,
+      "step": 5720
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3263136239997486,
+      "learning_rate": 7.318072159690775e-06,
+      "loss": 0.6945,
+      "step": 5721
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5057364484395517,
+      "learning_rate": 7.317153743841795e-06,
+      "loss": 0.6534,
+      "step": 5722
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1921886398476698,
+      "learning_rate": 7.316235228419557e-06,
+      "loss": 0.7417,
+      "step": 5723
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2805741257809202,
+      "learning_rate": 7.315316613463535e-06,
+      "loss": 0.6351,
+      "step": 5724
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1680651453924293,
+      "learning_rate": 7.314397899013199e-06,
+      "loss": 0.6435,
+      "step": 5725
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.579709725371114,
+      "learning_rate": 7.313479085108033e-06,
+      "loss": 0.8117,
+      "step": 5726
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.12867596268537,
+      "learning_rate": 7.312560171787518e-06,
+      "loss": 0.5755,
+      "step": 5727
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.460630916438581,
+      "learning_rate": 7.311641159091141e-06,
+      "loss": 0.7011,
+      "step": 5728
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3835136277707327,
+      "learning_rate": 7.310722047058396e-06,
+      "loss": 0.6453,
+      "step": 5729
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.367898887349231,
+      "learning_rate": 7.309802835728779e-06,
+      "loss": 0.5887,
+      "step": 5730
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.347383954671204,
+      "learning_rate": 7.308883525141788e-06,
+      "loss": 0.5457,
+      "step": 5731
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.752403281822804,
+      "learning_rate": 7.307964115336931e-06,
+      "loss": 0.7148,
+      "step": 5732
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3828783004993204,
+      "learning_rate": 7.307044606353715e-06,
+      "loss": 0.7022,
+      "step": 5733
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4771045090398756,
+      "learning_rate": 7.306124998231655e-06,
+      "loss": 0.6263,
+      "step": 5734
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9696100592115764,
+      "learning_rate": 7.305205291010269e-06,
+      "loss": 0.6928,
+      "step": 5735
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4278181883189487,
+      "learning_rate": 7.304285484729074e-06,
+      "loss": 0.6832,
+      "step": 5736
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4032868316243403,
+      "learning_rate": 7.303365579427603e-06,
+      "loss": 0.6308,
+      "step": 5737
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.308857327574184,
+      "learning_rate": 7.302445575145383e-06,
+      "loss": 0.6071,
+      "step": 5738
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2629335549474994,
+      "learning_rate": 7.301525471921949e-06,
+      "loss": 0.5981,
+      "step": 5739
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3519278225149673,
+      "learning_rate": 7.300605269796839e-06,
+      "loss": 0.7517,
+      "step": 5740
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0738902013009506,
+      "learning_rate": 7.299684968809595e-06,
+      "loss": 0.6679,
+      "step": 5741
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.252522923080774,
+      "learning_rate": 7.298764568999768e-06,
+      "loss": 0.5677,
+      "step": 5742
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.230715137534615,
+      "learning_rate": 7.297844070406908e-06,
+      "loss": 0.5717,
+      "step": 5743
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2643513990503148,
+      "learning_rate": 7.296923473070571e-06,
+      "loss": 0.6047,
+      "step": 5744
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.422909845813742,
+      "learning_rate": 7.296002777030316e-06,
+      "loss": 0.6575,
+      "step": 5745
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4094808762910096,
+      "learning_rate": 7.295081982325708e-06,
+      "loss": 0.7743,
+      "step": 5746
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5187582949616825,
+      "learning_rate": 7.2941610889963164e-06,
+      "loss": 0.6541,
+      "step": 5747
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3521696121058995,
+      "learning_rate": 7.293240097081714e-06,
+      "loss": 0.7378,
+      "step": 5748
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2449173773665083,
+      "learning_rate": 7.292319006621477e-06,
+      "loss": 0.6002,
+      "step": 5749
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4184962819860392,
+      "learning_rate": 7.291397817655186e-06,
+      "loss": 0.6442,
+      "step": 5750
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2616272505061137,
+      "learning_rate": 7.2904765302224276e-06,
+      "loss": 0.7093,
+      "step": 5751
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.274136833975844,
+      "learning_rate": 7.289555144362793e-06,
+      "loss": 0.6894,
+      "step": 5752
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9850279890002586,
+      "learning_rate": 7.288633660115874e-06,
+      "loss": 0.6766,
+      "step": 5753
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2733605367488647,
+      "learning_rate": 7.2877120775212685e-06,
+      "loss": 0.5691,
+      "step": 5754
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4781622497939897,
+      "learning_rate": 7.286790396618582e-06,
+      "loss": 0.7503,
+      "step": 5755
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.359817082463524,
+      "learning_rate": 7.285868617447418e-06,
+      "loss": 0.6231,
+      "step": 5756
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7006618309317183,
+      "learning_rate": 7.28494674004739e-06,
+      "loss": 0.7422,
+      "step": 5757
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2962597354495524,
+      "learning_rate": 7.28402476445811e-06,
+      "loss": 0.6093,
+      "step": 5758
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4057014200999307,
+      "learning_rate": 7.283102690719198e-06,
+      "loss": 0.5564,
+      "step": 5759
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2075237741348048,
+      "learning_rate": 7.282180518870281e-06,
+      "loss": 0.6241,
+      "step": 5760
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.538156355963023,
+      "learning_rate": 7.2812582489509844e-06,
+      "loss": 0.7072,
+      "step": 5761
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.619496513541743,
+      "learning_rate": 7.28033588100094e-06,
+      "loss": 0.6606,
+      "step": 5762
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1514278408627003,
+      "learning_rate": 7.279413415059782e-06,
+      "loss": 0.6472,
+      "step": 5763
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2690447506179339,
+      "learning_rate": 7.278490851167155e-06,
+      "loss": 0.579,
+      "step": 5764
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3544632736479323,
+      "learning_rate": 7.277568189362701e-06,
+      "loss": 0.6079,
+      "step": 5765
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.474338116814816,
+      "learning_rate": 7.276645429686071e-06,
+      "loss": 0.7317,
+      "step": 5766
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4615178554290025,
+      "learning_rate": 7.275722572176915e-06,
+      "loss": 0.7073,
+      "step": 5767
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2468164559879453,
+      "learning_rate": 7.2747996168748915e-06,
+      "loss": 0.5611,
+      "step": 5768
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3502205985315865,
+      "learning_rate": 7.2738765638196625e-06,
+      "loss": 0.7083,
+      "step": 5769
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3477476264653805,
+      "learning_rate": 7.272953413050894e-06,
+      "loss": 0.6481,
+      "step": 5770
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.395124992530006,
+      "learning_rate": 7.272030164608256e-06,
+      "loss": 0.7546,
+      "step": 5771
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3612977588525936,
+      "learning_rate": 7.27110681853142e-06,
+      "loss": 0.7435,
+      "step": 5772
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.28457409799206,
+      "learning_rate": 7.270183374860066e-06,
+      "loss": 0.6925,
+      "step": 5773
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3285275665379104,
+      "learning_rate": 7.269259833633877e-06,
+      "loss": 0.6838,
+      "step": 5774
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2912815745080455,
+      "learning_rate": 7.26833619489254e-06,
+      "loss": 0.6316,
+      "step": 5775
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4669334860900425,
+      "learning_rate": 7.267412458675744e-06,
+      "loss": 0.6367,
+      "step": 5776
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3072463175170006,
+      "learning_rate": 7.266488625023184e-06,
+      "loss": 0.7067,
+      "step": 5777
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2710661034284516,
+      "learning_rate": 7.265564693974561e-06,
+      "loss": 0.6205,
+      "step": 5778
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5783390569002544,
+      "learning_rate": 7.264640665569577e-06,
+      "loss": 0.7198,
+      "step": 5779
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5593824767132145,
+      "learning_rate": 7.2637165398479405e-06,
+      "loss": 0.669,
+      "step": 5780
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.235938294814157,
+      "learning_rate": 7.262792316849362e-06,
+      "loss": 0.614,
+      "step": 5781
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.266051149826196,
+      "learning_rate": 7.261867996613559e-06,
+      "loss": 0.6036,
+      "step": 5782
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2415247577616004,
+      "learning_rate": 7.26094357918025e-06,
+      "loss": 0.6547,
+      "step": 5783
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4164704550787455,
+      "learning_rate": 7.26001906458916e-06,
+      "loss": 0.6937,
+      "step": 5784
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1913501888500977,
+      "learning_rate": 7.259094452880019e-06,
+      "loss": 0.6512,
+      "step": 5785
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.309973172470814,
+      "learning_rate": 7.258169744092556e-06,
+      "loss": 0.6315,
+      "step": 5786
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4239174994508499,
+      "learning_rate": 7.257244938266511e-06,
+      "loss": 0.614,
+      "step": 5787
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3287820849891523,
+      "learning_rate": 7.256320035441623e-06,
+      "loss": 0.636,
+      "step": 5788
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.421616768402994,
+      "learning_rate": 7.255395035657639e-06,
+      "loss": 0.6811,
+      "step": 5789
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4777503049772003,
+      "learning_rate": 7.254469938954307e-06,
+      "loss": 0.6956,
+      "step": 5790
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4082081266741338,
+      "learning_rate": 7.25354474537138e-06,
+      "loss": 0.764,
+      "step": 5791
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.105990988576737,
+      "learning_rate": 7.252619454948617e-06,
+      "loss": 0.6881,
+      "step": 5792
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2393795313281446,
+      "learning_rate": 7.251694067725779e-06,
+      "loss": 0.5835,
+      "step": 5793
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5346144850064636,
+      "learning_rate": 7.250768583742634e-06,
+      "loss": 0.7264,
+      "step": 5794
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3285371203095682,
+      "learning_rate": 7.249843003038948e-06,
+      "loss": 0.6921,
+      "step": 5795
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.440230325006067,
+      "learning_rate": 7.2489173256544975e-06,
+      "loss": 0.7149,
+      "step": 5796
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6508923127287893,
+      "learning_rate": 7.247991551629061e-06,
+      "loss": 0.7548,
+      "step": 5797
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2683371424837095,
+      "learning_rate": 7.247065681002421e-06,
+      "loss": 0.7129,
+      "step": 5798
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4117248743247064,
+      "learning_rate": 7.246139713814365e-06,
+      "loss": 0.6539,
+      "step": 5799
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5375181635942645,
+      "learning_rate": 7.245213650104683e-06,
+      "loss": 0.7502,
+      "step": 5800
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.427960839719468,
+      "learning_rate": 7.2442874899131685e-06,
+      "loss": 0.6708,
+      "step": 5801
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2268875896576288,
+      "learning_rate": 7.2433612332796245e-06,
+      "loss": 0.6722,
+      "step": 5802
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.242880492927149,
+      "learning_rate": 7.242434880243851e-06,
+      "loss": 0.7177,
+      "step": 5803
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5273164210385173,
+      "learning_rate": 7.241508430845656e-06,
+      "loss": 0.819,
+      "step": 5804
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3463791813252286,
+      "learning_rate": 7.2405818851248535e-06,
+      "loss": 0.7133,
+      "step": 5805
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.289381203598552,
+      "learning_rate": 7.239655243121255e-06,
+      "loss": 0.5358,
+      "step": 5806
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4003873840130423,
+      "learning_rate": 7.238728504874683e-06,
+      "loss": 0.7294,
+      "step": 5807
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8919028362401733,
+      "learning_rate": 7.237801670424963e-06,
+      "loss": 0.8316,
+      "step": 5808
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6211896851037828,
+      "learning_rate": 7.236874739811921e-06,
+      "loss": 0.6799,
+      "step": 5809
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.115390109352899,
+      "learning_rate": 7.235947713075389e-06,
+      "loss": 0.6877,
+      "step": 5810
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1271713937719465,
+      "learning_rate": 7.235020590255204e-06,
+      "loss": 0.6022,
+      "step": 5811
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.248759818694233,
+      "learning_rate": 7.234093371391206e-06,
+      "loss": 0.6096,
+      "step": 5812
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2842234839560618,
+      "learning_rate": 7.233166056523242e-06,
+      "loss": 0.6014,
+      "step": 5813
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1565506132865184,
+      "learning_rate": 7.232238645691157e-06,
+      "loss": 0.5737,
+      "step": 5814
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3810795799580668,
+      "learning_rate": 7.231311138934807e-06,
+      "loss": 0.6657,
+      "step": 5815
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3582984429628218,
+      "learning_rate": 7.230383536294048e-06,
+      "loss": 0.6233,
+      "step": 5816
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8043935152042634,
+      "learning_rate": 7.229455837808741e-06,
+      "loss": 0.669,
+      "step": 5817
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1046390822099452,
+      "learning_rate": 7.228528043518752e-06,
+      "loss": 0.5937,
+      "step": 5818
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.304212670087652,
+      "learning_rate": 7.227600153463947e-06,
+      "loss": 0.7471,
+      "step": 5819
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3301975873184555,
+      "learning_rate": 7.226672167684205e-06,
+      "loss": 0.6381,
+      "step": 5820
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.583771309021479,
+      "learning_rate": 7.225744086219402e-06,
+      "loss": 0.6344,
+      "step": 5821
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4875480336899278,
+      "learning_rate": 7.224815909109416e-06,
+      "loss": 0.6971,
+      "step": 5822
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1944780656376464,
+      "learning_rate": 7.223887636394137e-06,
+      "loss": 0.6521,
+      "step": 5823
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3345170111930285,
+      "learning_rate": 7.222959268113452e-06,
+      "loss": 0.7298,
+      "step": 5824
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2874832104978102,
+      "learning_rate": 7.222030804307257e-06,
+      "loss": 0.5541,
+      "step": 5825
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2164736730731742,
+      "learning_rate": 7.22110224501545e-06,
+      "loss": 0.6121,
+      "step": 5826
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2242470188389458,
+      "learning_rate": 7.220173590277933e-06,
+      "loss": 0.6142,
+      "step": 5827
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3776362045298343,
+      "learning_rate": 7.219244840134612e-06,
+      "loss": 0.6013,
+      "step": 5828
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3604833238524943,
+      "learning_rate": 7.218315994625397e-06,
+      "loss": 0.6769,
+      "step": 5829
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4577814209646427,
+      "learning_rate": 7.217387053790204e-06,
+      "loss": 0.6825,
+      "step": 5830
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4921668535843606,
+      "learning_rate": 7.216458017668951e-06,
+      "loss": 0.6961,
+      "step": 5831
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.969978732581316,
+      "learning_rate": 7.21552888630156e-06,
+      "loss": 0.6797,
+      "step": 5832
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2736464103213332,
+      "learning_rate": 7.214599659727959e-06,
+      "loss": 0.622,
+      "step": 5833
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.123926274312551,
+      "learning_rate": 7.213670337988079e-06,
+      "loss": 0.6584,
+      "step": 5834
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1615162731364088,
+      "learning_rate": 7.2127409211218526e-06,
+      "loss": 0.7163,
+      "step": 5835
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.533272245884129,
+      "learning_rate": 7.211811409169222e-06,
+      "loss": 0.6109,
+      "step": 5836
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5994238830865577,
+      "learning_rate": 7.21088180217013e-06,
+      "loss": 0.7438,
+      "step": 5837
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.276127670669517,
+      "learning_rate": 7.2099521001645225e-06,
+      "loss": 0.614,
+      "step": 5838
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3222839716529209,
+      "learning_rate": 7.209022303192351e-06,
+      "loss": 0.6502,
+      "step": 5839
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4107785597632605,
+      "learning_rate": 7.208092411293572e-06,
+      "loss": 0.6797,
+      "step": 5840
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.154525525802328,
+      "learning_rate": 7.207162424508144e-06,
+      "loss": 0.6002,
+      "step": 5841
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3571365020771418,
+      "learning_rate": 7.206232342876031e-06,
+      "loss": 0.6605,
+      "step": 5842
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2193281552765902,
+      "learning_rate": 7.2053021664372e-06,
+      "loss": 0.5384,
+      "step": 5843
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1099434204490066,
+      "learning_rate": 7.204371895231623e-06,
+      "loss": 0.5979,
+      "step": 5844
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4613630111438998,
+      "learning_rate": 7.20344152929928e-06,
+      "loss": 0.6575,
+      "step": 5845
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3667393844816693,
+      "learning_rate": 7.202511068680143e-06,
+      "loss": 0.6608,
+      "step": 5846
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.88314691326305,
+      "learning_rate": 7.201580513414201e-06,
+      "loss": 0.5389,
+      "step": 5847
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2976628134495545,
+      "learning_rate": 7.200649863541442e-06,
+      "loss": 0.6076,
+      "step": 5848
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0789675958601137,
+      "learning_rate": 7.199719119101858e-06,
+      "loss": 0.5721,
+      "step": 5849
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4416996130134634,
+      "learning_rate": 7.198788280135443e-06,
+      "loss": 0.7628,
+      "step": 5850
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.515399632323569,
+      "learning_rate": 7.197857346682199e-06,
+      "loss": 0.6412,
+      "step": 5851
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.289163743605131,
+      "learning_rate": 7.19692631878213e-06,
+      "loss": 0.6434,
+      "step": 5852
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.237758199329402,
+      "learning_rate": 7.195995196475244e-06,
+      "loss": 0.7715,
+      "step": 5853
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2995322984133444,
+      "learning_rate": 7.195063979801554e-06,
+      "loss": 0.605,
+      "step": 5854
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2594243521310664,
+      "learning_rate": 7.194132668801077e-06,
+      "loss": 0.6659,
+      "step": 5855
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3449902390351698,
+      "learning_rate": 7.193201263513831e-06,
+      "loss": 0.7156,
+      "step": 5856
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3114031462131805,
+      "learning_rate": 7.192269763979844e-06,
+      "loss": 0.6392,
+      "step": 5857
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3969712188869927,
+      "learning_rate": 7.1913381702391405e-06,
+      "loss": 0.6176,
+      "step": 5858
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.738402590422815,
+      "learning_rate": 7.190406482331757e-06,
+      "loss": 0.6358,
+      "step": 5859
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2955382518231178,
+      "learning_rate": 7.189474700297728e-06,
+      "loss": 0.7672,
+      "step": 5860
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3647437873580721,
+      "learning_rate": 7.188542824177096e-06,
+      "loss": 0.5769,
+      "step": 5861
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4288143788199674,
+      "learning_rate": 7.187610854009903e-06,
+      "loss": 0.7161,
+      "step": 5862
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2117129717310613,
+      "learning_rate": 7.186678789836201e-06,
+      "loss": 0.6948,
+      "step": 5863
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3343375469837624,
+      "learning_rate": 7.18574663169604e-06,
+      "loss": 0.6577,
+      "step": 5864
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0387743550287787,
+      "learning_rate": 7.18481437962948e-06,
+      "loss": 0.5865,
+      "step": 5865
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2374413992130069,
+      "learning_rate": 7.183882033676579e-06,
+      "loss": 0.6844,
+      "step": 5866
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.306841391196018,
+      "learning_rate": 7.182949593877404e-06,
+      "loss": 0.736,
+      "step": 5867
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3230823482823253,
+      "learning_rate": 7.182017060272023e-06,
+      "loss": 0.8319,
+      "step": 5868
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.29997915238709,
+      "learning_rate": 7.1810844329005095e-06,
+      "loss": 0.6951,
+      "step": 5869
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.05208171756743,
+      "learning_rate": 7.180151711802941e-06,
+      "loss": 0.6021,
+      "step": 5870
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5257398578495813,
+      "learning_rate": 7.179218897019395e-06,
+      "loss": 0.7225,
+      "step": 5871
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.514663628457308,
+      "learning_rate": 7.178285988589963e-06,
+      "loss": 0.7441,
+      "step": 5872
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3087637910324548,
+      "learning_rate": 7.177352986554729e-06,
+      "loss": 0.6903,
+      "step": 5873
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4637641469435008,
+      "learning_rate": 7.176419890953788e-06,
+      "loss": 0.6795,
+      "step": 5874
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2071772222263706,
+      "learning_rate": 7.175486701827239e-06,
+      "loss": 0.5777,
+      "step": 5875
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4172080099009559,
+      "learning_rate": 7.174553419215179e-06,
+      "loss": 0.679,
+      "step": 5876
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3257389928351253,
+      "learning_rate": 7.173620043157716e-06,
+      "loss": 0.734,
+      "step": 5877
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4086382464444522,
+      "learning_rate": 7.17268657369496e-06,
+      "loss": 0.6441,
+      "step": 5878
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4064223179674387,
+      "learning_rate": 7.171753010867023e-06,
+      "loss": 0.6426,
+      "step": 5879
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6007327153070106,
+      "learning_rate": 7.1708193547140205e-06,
+      "loss": 0.7415,
+      "step": 5880
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.188437713054204,
+      "learning_rate": 7.169885605276077e-06,
+      "loss": 0.6962,
+      "step": 5881
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4550894480101255,
+      "learning_rate": 7.1689517625933165e-06,
+      "loss": 0.7289,
+      "step": 5882
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.209099647394263,
+      "learning_rate": 7.168017826705869e-06,
+      "loss": 0.6275,
+      "step": 5883
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4733825519031507,
+      "learning_rate": 7.167083797653866e-06,
+      "loss": 0.6934,
+      "step": 5884
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.307375996645806,
+      "learning_rate": 7.166149675477445e-06,
+      "loss": 0.7956,
+      "step": 5885
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3714364265836432,
+      "learning_rate": 7.16521546021675e-06,
+      "loss": 0.5979,
+      "step": 5886
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3950165585115446,
+      "learning_rate": 7.164281151911923e-06,
+      "loss": 0.7804,
+      "step": 5887
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0347906118504882,
+      "learning_rate": 7.163346750603117e-06,
+      "loss": 0.64,
+      "step": 5888
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5032155222955577,
+      "learning_rate": 7.162412256330481e-06,
+      "loss": 0.6265,
+      "step": 5889
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.065721990561345,
+      "learning_rate": 7.161477669134174e-06,
+      "loss": 0.6637,
+      "step": 5890
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.327661891932,
+      "learning_rate": 7.160542989054359e-06,
+      "loss": 0.7254,
+      "step": 5891
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.288425983692125,
+      "learning_rate": 7.1596082161312005e-06,
+      "loss": 0.7471,
+      "step": 5892
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1930911621347158,
+      "learning_rate": 7.158673350404866e-06,
+      "loss": 0.6993,
+      "step": 5893
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2544837104761521,
+      "learning_rate": 7.157738391915531e-06,
+      "loss": 0.6591,
+      "step": 5894
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.084929849618508,
+      "learning_rate": 7.156803340703371e-06,
+      "loss": 0.5888,
+      "step": 5895
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.37464710707723,
+      "learning_rate": 7.155868196808569e-06,
+      "loss": 0.63,
+      "step": 5896
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3296018137678114,
+      "learning_rate": 7.154932960271308e-06,
+      "loss": 0.7078,
+      "step": 5897
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1481988945614205,
+      "learning_rate": 7.153997631131778e-06,
+      "loss": 0.6789,
+      "step": 5898
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2763081283162852,
+      "learning_rate": 7.153062209430174e-06,
+      "loss": 0.7494,
+      "step": 5899
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4859223488362956,
+      "learning_rate": 7.152126695206689e-06,
+      "loss": 0.8341,
+      "step": 5900
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4290567863217811,
+      "learning_rate": 7.151191088501531e-06,
+      "loss": 0.6417,
+      "step": 5901
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4611094073527626,
+      "learning_rate": 7.150255389354899e-06,
+      "loss": 0.6793,
+      "step": 5902
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1873661249825744,
+      "learning_rate": 7.149319597807003e-06,
+      "loss": 0.5971,
+      "step": 5903
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2343613460333336,
+      "learning_rate": 7.148383713898058e-06,
+      "loss": 0.6384,
+      "step": 5904
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5054142668657309,
+      "learning_rate": 7.14744773766828e-06,
+      "loss": 0.5908,
+      "step": 5905
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.455268961566584,
+      "learning_rate": 7.146511669157892e-06,
+      "loss": 0.6128,
+      "step": 5906
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.284867639799029,
+      "learning_rate": 7.145575508407114e-06,
+      "loss": 0.7123,
+      "step": 5907
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.485782091928607,
+      "learning_rate": 7.14463925545618e-06,
+      "loss": 0.717,
+      "step": 5908
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6477433349944333,
+      "learning_rate": 7.143702910345318e-06,
+      "loss": 0.7472,
+      "step": 5909
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1549471463127325,
+      "learning_rate": 7.142766473114771e-06,
+      "loss": 0.567,
+      "step": 5910
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.429806839431787,
+      "learning_rate": 7.141829943804777e-06,
+      "loss": 0.6353,
+      "step": 5911
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2252675682983227,
+      "learning_rate": 7.1408933224555765e-06,
+      "loss": 0.6464,
+      "step": 5912
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1851418138536538,
+      "learning_rate": 7.139956609107425e-06,
+      "loss": 0.6326,
+      "step": 5913
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5211650584076015,
+      "learning_rate": 7.139019803800569e-06,
+      "loss": 0.7393,
+      "step": 5914
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3612996967764879,
+      "learning_rate": 7.138082906575271e-06,
+      "loss": 0.6287,
+      "step": 5915
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3614783387825435,
+      "learning_rate": 7.137145917471788e-06,
+      "loss": 0.7037,
+      "step": 5916
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3397623666255074,
+      "learning_rate": 7.136208836530385e-06,
+      "loss": 0.6707,
+      "step": 5917
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.891171662607715,
+      "learning_rate": 7.13527166379133e-06,
+      "loss": 0.7268,
+      "step": 5918
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2682732439090663,
+      "learning_rate": 7.134334399294897e-06,
+      "loss": 0.6458,
+      "step": 5919
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.311267602076328,
+      "learning_rate": 7.1333970430813615e-06,
+      "loss": 0.7258,
+      "step": 5920
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0287426401011899,
+      "learning_rate": 7.132459595191003e-06,
+      "loss": 0.613,
+      "step": 5921
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3601243053603989,
+      "learning_rate": 7.131522055664109e-06,
+      "loss": 0.6822,
+      "step": 5922
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2313290372301362,
+      "learning_rate": 7.13058442454096e-06,
+      "loss": 0.757,
+      "step": 5923
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.346341147873179,
+      "learning_rate": 7.129646701861858e-06,
+      "loss": 0.6114,
+      "step": 5924
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.429686312618615,
+      "learning_rate": 7.128708887667093e-06,
+      "loss": 0.7845,
+      "step": 5925
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5312874126484606,
+      "learning_rate": 7.127770981996966e-06,
+      "loss": 0.7279,
+      "step": 5926
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0518944416726996,
+      "learning_rate": 7.1268329848917795e-06,
+      "loss": 0.6758,
+      "step": 5927
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3901373609629037,
+      "learning_rate": 7.125894896391845e-06,
+      "loss": 0.671,
+      "step": 5928
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2483026131468202,
+      "learning_rate": 7.124956716537471e-06,
+      "loss": 0.6792,
+      "step": 5929
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3919962350654354,
+      "learning_rate": 7.124018445368975e-06,
+      "loss": 0.6575,
+      "step": 5930
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3012959790159615,
+      "learning_rate": 7.123080082926677e-06,
+      "loss": 0.6669,
+      "step": 5931
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3149507332781618,
+      "learning_rate": 7.122141629250896e-06,
+      "loss": 0.6871,
+      "step": 5932
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5125935284171594,
+      "learning_rate": 7.1212030843819656e-06,
+      "loss": 0.7504,
+      "step": 5933
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.254761467743061,
+      "learning_rate": 7.120264448360214e-06,
+      "loss": 0.56,
+      "step": 5934
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2785182358671259,
+      "learning_rate": 7.1193257212259784e-06,
+      "loss": 0.59,
+      "step": 5935
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2634791047225309,
+      "learning_rate": 7.118386903019594e-06,
+      "loss": 0.7148,
+      "step": 5936
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5374781084273883,
+      "learning_rate": 7.117447993781408e-06,
+      "loss": 0.7356,
+      "step": 5937
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1961886749128041,
+      "learning_rate": 7.116508993551765e-06,
+      "loss": 0.6826,
+      "step": 5938
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1499323336195284,
+      "learning_rate": 7.115569902371018e-06,
+      "loss": 0.5745,
+      "step": 5939
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1256470536856829,
+      "learning_rate": 7.114630720279521e-06,
+      "loss": 0.6238,
+      "step": 5940
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2920356650765226,
+      "learning_rate": 7.11369144731763e-06,
+      "loss": 0.6662,
+      "step": 5941
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.615772954734369,
+      "learning_rate": 7.1127520835257125e-06,
+      "loss": 0.6709,
+      "step": 5942
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.32347372240801,
+      "learning_rate": 7.111812628944132e-06,
+      "loss": 0.567,
+      "step": 5943
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2825993035978476,
+      "learning_rate": 7.110873083613259e-06,
+      "loss": 0.5999,
+      "step": 5944
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2357688708416628,
+      "learning_rate": 7.109933447573469e-06,
+      "loss": 0.706,
+      "step": 5945
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8208195022291407,
+      "learning_rate": 7.108993720865138e-06,
+      "loss": 0.6646,
+      "step": 5946
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.324934833663141,
+      "learning_rate": 7.108053903528652e-06,
+      "loss": 0.7141,
+      "step": 5947
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3824411389976605,
+      "learning_rate": 7.107113995604395e-06,
+      "loss": 0.7147,
+      "step": 5948
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4177806721112873,
+      "learning_rate": 7.106173997132755e-06,
+      "loss": 0.6834,
+      "step": 5949
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3741778434483303,
+      "learning_rate": 7.105233908154128e-06,
+      "loss": 0.7524,
+      "step": 5950
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1686613619111454,
+      "learning_rate": 7.104293728708912e-06,
+      "loss": 0.4931,
+      "step": 5951
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6122912723878227,
+      "learning_rate": 7.103353458837509e-06,
+      "loss": 0.7435,
+      "step": 5952
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4022297258192369,
+      "learning_rate": 7.102413098580322e-06,
+      "loss": 0.7153,
+      "step": 5953
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5525243926826555,
+      "learning_rate": 7.101472647977761e-06,
+      "loss": 0.6445,
+      "step": 5954
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3547060504146786,
+      "learning_rate": 7.100532107070242e-06,
+      "loss": 0.7479,
+      "step": 5955
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2046849767751335,
+      "learning_rate": 7.099591475898178e-06,
+      "loss": 0.7097,
+      "step": 5956
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3742195766740772,
+      "learning_rate": 7.098650754501994e-06,
+      "loss": 0.7078,
+      "step": 5957
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4367876822348389,
+      "learning_rate": 7.0977099429221145e-06,
+      "loss": 0.7718,
+      "step": 5958
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.35769122562761,
+      "learning_rate": 7.096769041198964e-06,
+      "loss": 0.746,
+      "step": 5959
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6684101842141588,
+      "learning_rate": 7.095828049372979e-06,
+      "loss": 0.6119,
+      "step": 5960
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.184288284522403,
+      "learning_rate": 7.0948869674845955e-06,
+      "loss": 0.6561,
+      "step": 5961
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.617618574391751,
+      "learning_rate": 7.093945795574254e-06,
+      "loss": 0.747,
+      "step": 5962
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.403164259344188,
+      "learning_rate": 7.093004533682397e-06,
+      "loss": 0.6326,
+      "step": 5963
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4152955746039229,
+      "learning_rate": 7.0920631818494745e-06,
+      "loss": 0.6116,
+      "step": 5964
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3926761052051804,
+      "learning_rate": 7.091121740115936e-06,
+      "loss": 0.7846,
+      "step": 5965
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.603019940374699,
+      "learning_rate": 7.090180208522241e-06,
+      "loss": 0.6797,
+      "step": 5966
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.30261354349122,
+      "learning_rate": 7.089238587108849e-06,
+      "loss": 0.6608,
+      "step": 5967
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3198522998044293,
+      "learning_rate": 7.08829687591622e-06,
+      "loss": 0.6545,
+      "step": 5968
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.334557465783978,
+      "learning_rate": 7.087355074984823e-06,
+      "loss": 0.6876,
+      "step": 5969
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3103874274288063,
+      "learning_rate": 7.086413184355129e-06,
+      "loss": 0.6694,
+      "step": 5970
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2860360569124505,
+      "learning_rate": 7.085471204067616e-06,
+      "loss": 0.6559,
+      "step": 5971
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.536424744625984,
+      "learning_rate": 7.084529134162759e-06,
+      "loss": 0.6716,
+      "step": 5972
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3188731581163595,
+      "learning_rate": 7.083586974681042e-06,
+      "loss": 0.6572,
+      "step": 5973
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2819686887878998,
+      "learning_rate": 7.082644725662954e-06,
+      "loss": 0.7175,
+      "step": 5974
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3460370422095795,
+      "learning_rate": 7.0817023871489835e-06,
+      "loss": 0.7108,
+      "step": 5975
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6250912341674653,
+      "learning_rate": 7.080759959179626e-06,
+      "loss": 0.7905,
+      "step": 5976
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2168013538388869,
+      "learning_rate": 7.079817441795378e-06,
+      "loss": 0.6005,
+      "step": 5977
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2160058858576814,
+      "learning_rate": 7.078874835036742e-06,
+      "loss": 0.5669,
+      "step": 5978
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3333747582651796,
+      "learning_rate": 7.077932138944225e-06,
+      "loss": 0.6643,
+      "step": 5979
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.475877732941908,
+      "learning_rate": 7.076989353558335e-06,
+      "loss": 0.75,
+      "step": 5980
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2204066996876537,
+      "learning_rate": 7.076046478919589e-06,
+      "loss": 0.5966,
+      "step": 5981
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4058290037460044,
+      "learning_rate": 7.075103515068501e-06,
+      "loss": 0.6864,
+      "step": 5982
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0758689527268905,
+      "learning_rate": 7.074160462045593e-06,
+      "loss": 0.6084,
+      "step": 5983
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2325641693719573,
+      "learning_rate": 7.073217319891391e-06,
+      "loss": 0.6128,
+      "step": 5984
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.348784117468302,
+      "learning_rate": 7.072274088646425e-06,
+      "loss": 0.7141,
+      "step": 5985
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1995277886134852,
+      "learning_rate": 7.071330768351223e-06,
+      "loss": 0.6316,
+      "step": 5986
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2920703183884255,
+      "learning_rate": 7.070387359046328e-06,
+      "loss": 0.6007,
+      "step": 5987
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.418700828024544,
+      "learning_rate": 7.0694438607722746e-06,
+      "loss": 0.7224,
+      "step": 5988
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2872392739403506,
+      "learning_rate": 7.068500273569612e-06,
+      "loss": 0.6983,
+      "step": 5989
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4386690618975528,
+      "learning_rate": 7.067556597478884e-06,
+      "loss": 0.6647,
+      "step": 5990
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4570133323451602,
+      "learning_rate": 7.066612832540644e-06,
+      "loss": 0.6138,
+      "step": 5991
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2110548063716973,
+      "learning_rate": 7.065668978795449e-06,
+      "loss": 0.5959,
+      "step": 5992
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3410171553514159,
+      "learning_rate": 7.064725036283857e-06,
+      "loss": 0.6749,
+      "step": 5993
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6021419268763577,
+      "learning_rate": 7.063781005046433e-06,
+      "loss": 0.6838,
+      "step": 5994
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5275621125552945,
+      "learning_rate": 7.0628368851237424e-06,
+      "loss": 0.6152,
+      "step": 5995
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3203675383429148,
+      "learning_rate": 7.061892676556357e-06,
+      "loss": 0.6331,
+      "step": 5996
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3926486702818335,
+      "learning_rate": 7.06094837938485e-06,
+      "loss": 0.7182,
+      "step": 5997
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1779260913416856,
+      "learning_rate": 7.060003993649802e-06,
+      "loss": 0.6585,
+      "step": 5998
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4618947318414985,
+      "learning_rate": 7.059059519391794e-06,
+      "loss": 0.7217,
+      "step": 5999
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2573434591338273,
+      "learning_rate": 7.058114956651413e-06,
+      "loss": 0.6306,
+      "step": 6000
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5246797831451289,
+      "learning_rate": 7.057170305469248e-06,
+      "loss": 0.7419,
+      "step": 6001
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2902158918286017,
+      "learning_rate": 7.056225565885895e-06,
+      "loss": 0.6167,
+      "step": 6002
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2916057757486545,
+      "learning_rate": 7.055280737941949e-06,
+      "loss": 0.7207,
+      "step": 6003
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4165977949246367,
+      "learning_rate": 7.054335821678012e-06,
+      "loss": 0.6517,
+      "step": 6004
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3233375207106859,
+      "learning_rate": 7.05339081713469e-06,
+      "loss": 0.6383,
+      "step": 6005
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4557157505065885,
+      "learning_rate": 7.05244572435259e-06,
+      "loss": 0.7076,
+      "step": 6006
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6362149445107488,
+      "learning_rate": 7.051500543372328e-06,
+      "loss": 0.5099,
+      "step": 6007
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2299807658311235,
+      "learning_rate": 7.050555274234518e-06,
+      "loss": 0.617,
+      "step": 6008
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2140616060338143,
+      "learning_rate": 7.049609916979782e-06,
+      "loss": 0.6211,
+      "step": 6009
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2086683335329704,
+      "learning_rate": 7.04866447164874e-06,
+      "loss": 0.611,
+      "step": 6010
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3353728558283444,
+      "learning_rate": 7.0477189382820266e-06,
+      "loss": 0.6738,
+      "step": 6011
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4498318286163019,
+      "learning_rate": 7.046773316920269e-06,
+      "loss": 0.6384,
+      "step": 6012
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.275207719524183,
+      "learning_rate": 7.045827607604103e-06,
+      "loss": 0.6339,
+      "step": 6013
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8127613965852258,
+      "learning_rate": 7.044881810374169e-06,
+      "loss": 0.7029,
+      "step": 6014
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0143787312485755,
+      "learning_rate": 7.043935925271108e-06,
+      "loss": 0.6424,
+      "step": 6015
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1801706671371133,
+      "learning_rate": 7.042989952335571e-06,
+      "loss": 0.534,
+      "step": 6016
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4763358411080658,
+      "learning_rate": 7.0420438916082024e-06,
+      "loss": 0.7671,
+      "step": 6017
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.477372487801573,
+      "learning_rate": 7.041097743129662e-06,
+      "loss": 0.6034,
+      "step": 6018
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.059592494812258,
+      "learning_rate": 7.040151506940605e-06,
+      "loss": 0.5132,
+      "step": 6019
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.365706624764385,
+      "learning_rate": 7.039205183081694e-06,
+      "loss": 0.6814,
+      "step": 6020
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3428909627662178,
+      "learning_rate": 7.038258771593596e-06,
+      "loss": 0.8891,
+      "step": 6021
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1867286986777763,
+      "learning_rate": 7.037312272516977e-06,
+      "loss": 0.7883,
+      "step": 6022
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5852409020161335,
+      "learning_rate": 7.036365685892515e-06,
+      "loss": 0.7266,
+      "step": 6023
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0981295875992476,
+      "learning_rate": 7.035419011760882e-06,
+      "loss": 0.5471,
+      "step": 6024
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2912317274563747,
+      "learning_rate": 7.0344722501627625e-06,
+      "loss": 0.6323,
+      "step": 6025
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0863262524248636,
+      "learning_rate": 7.033525401138838e-06,
+      "loss": 0.6325,
+      "step": 6026
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3956403277392988,
+      "learning_rate": 7.0325784647298e-06,
+      "loss": 0.7083,
+      "step": 6027
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5820543972682788,
+      "learning_rate": 7.031631440976336e-06,
+      "loss": 0.6684,
+      "step": 6028
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2027624268843344,
+      "learning_rate": 7.0306843299191465e-06,
+      "loss": 0.5995,
+      "step": 6029
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2274539761456607,
+      "learning_rate": 7.0297371315989284e-06,
+      "loss": 0.699,
+      "step": 6030
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3613212401624553,
+      "learning_rate": 7.028789846056383e-06,
+      "loss": 0.6139,
+      "step": 6031
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.595663215323718,
+      "learning_rate": 7.0278424733322225e-06,
+      "loss": 0.7305,
+      "step": 6032
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.333494576946721,
+      "learning_rate": 7.026895013467153e-06,
+      "loss": 0.7013,
+      "step": 6033
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.395730553364316,
+      "learning_rate": 7.0259474665018915e-06,
+      "loss": 0.7437,
+      "step": 6034
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1522341680160761,
+      "learning_rate": 7.024999832477154e-06,
+      "loss": 0.6036,
+      "step": 6035
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1736083287693218,
+      "learning_rate": 7.0240521114336655e-06,
+      "loss": 0.6608,
+      "step": 6036
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1962363161655203,
+      "learning_rate": 7.023104303412149e-06,
+      "loss": 0.7031,
+      "step": 6037
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3263383169585943,
+      "learning_rate": 7.022156408453336e-06,
+      "loss": 0.7008,
+      "step": 6038
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4185056469040833,
+      "learning_rate": 7.0212084265979575e-06,
+      "loss": 0.6453,
+      "step": 6039
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4066757592253258,
+      "learning_rate": 7.02026035788675e-06,
+      "loss": 0.7481,
+      "step": 6040
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4592032500530658,
+      "learning_rate": 7.019312202360457e-06,
+      "loss": 0.7205,
+      "step": 6041
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4374089358853719,
+      "learning_rate": 7.018363960059823e-06,
+      "loss": 0.6339,
+      "step": 6042
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3354945188580352,
+      "learning_rate": 7.017415631025593e-06,
+      "loss": 0.6156,
+      "step": 6043
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4062295400750948,
+      "learning_rate": 7.016467215298519e-06,
+      "loss": 0.7317,
+      "step": 6044
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.30229046876227,
+      "learning_rate": 7.01551871291936e-06,
+      "loss": 0.6073,
+      "step": 6045
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3765690141736324,
+      "learning_rate": 7.014570123928872e-06,
+      "loss": 0.6659,
+      "step": 6046
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.138017769165743,
+      "learning_rate": 7.01362144836782e-06,
+      "loss": 0.5926,
+      "step": 6047
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5728293120211703,
+      "learning_rate": 7.012672686276969e-06,
+      "loss": 0.7827,
+      "step": 6048
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3058184010396123,
+      "learning_rate": 7.011723837697091e-06,
+      "loss": 0.6651,
+      "step": 6049
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3871253498900284,
+      "learning_rate": 7.010774902668958e-06,
+      "loss": 0.7338,
+      "step": 6050
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2502854178710168,
+      "learning_rate": 7.009825881233351e-06,
+      "loss": 0.6487,
+      "step": 6051
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1667783505594949,
+      "learning_rate": 7.008876773431048e-06,
+      "loss": 0.6365,
+      "step": 6052
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3389447250102036,
+      "learning_rate": 7.007927579302837e-06,
+      "loss": 0.6177,
+      "step": 6053
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2718506250300572,
+      "learning_rate": 7.0069782988895056e-06,
+      "loss": 0.646,
+      "step": 6054
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3454582035468432,
+      "learning_rate": 7.006028932231847e-06,
+      "loss": 0.5949,
+      "step": 6055
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4110779450312718,
+      "learning_rate": 7.005079479370658e-06,
+      "loss": 0.722,
+      "step": 6056
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4436893480519062,
+      "learning_rate": 7.0041299403467365e-06,
+      "loss": 0.7127,
+      "step": 6057
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1719930181332325,
+      "learning_rate": 7.003180315200891e-06,
+      "loss": 0.7542,
+      "step": 6058
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3957655903422195,
+      "learning_rate": 7.002230603973924e-06,
+      "loss": 0.6949,
+      "step": 6059
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3400522536475947,
+      "learning_rate": 7.001280806706649e-06,
+      "loss": 0.6868,
+      "step": 6060
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4811432862877358,
+      "learning_rate": 7.0003309234398805e-06,
+      "loss": 0.6946,
+      "step": 6061
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3253249017169155,
+      "learning_rate": 6.999380954214438e-06,
+      "loss": 0.6122,
+      "step": 6062
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4922827504545497,
+      "learning_rate": 6.9984308990711425e-06,
+      "loss": 0.6753,
+      "step": 6063
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.606559127401534,
+      "learning_rate": 6.9974807580508205e-06,
+      "loss": 0.6234,
+      "step": 6064
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1751361016023867,
+      "learning_rate": 6.996530531194302e-06,
+      "loss": 0.6192,
+      "step": 6065
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2208277857120522,
+      "learning_rate": 6.995580218542419e-06,
+      "loss": 0.6875,
+      "step": 6066
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0456155340520412,
+      "learning_rate": 6.994629820136011e-06,
+      "loss": 0.5622,
+      "step": 6067
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3170861979633837,
+      "learning_rate": 6.9936793360159175e-06,
+      "loss": 0.6452,
+      "step": 6068
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3656600280698612,
+      "learning_rate": 6.992728766222982e-06,
+      "loss": 0.6756,
+      "step": 6069
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0797727600686113,
+      "learning_rate": 6.991778110798053e-06,
+      "loss": 0.692,
+      "step": 6070
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3557722195374737,
+      "learning_rate": 6.990827369781983e-06,
+      "loss": 0.4827,
+      "step": 6071
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2148129618662207,
+      "learning_rate": 6.9898765432156266e-06,
+      "loss": 0.6715,
+      "step": 6072
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1109709505648795,
+      "learning_rate": 6.988925631139843e-06,
+      "loss": 0.6214,
+      "step": 6073
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8479404357405906,
+      "learning_rate": 6.987974633595498e-06,
+      "loss": 0.6322,
+      "step": 6074
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2253588728786111,
+      "learning_rate": 6.987023550623453e-06,
+      "loss": 0.668,
+      "step": 6075
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5336745829885896,
+      "learning_rate": 6.9860723822645825e-06,
+      "loss": 0.781,
+      "step": 6076
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1939212164231303,
+      "learning_rate": 6.985121128559757e-06,
+      "loss": 0.5767,
+      "step": 6077
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3472137154725852,
+      "learning_rate": 6.984169789549857e-06,
+      "loss": 0.6495,
+      "step": 6078
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5331537610404011,
+      "learning_rate": 6.9832183652757625e-06,
+      "loss": 0.6323,
+      "step": 6079
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6209527050734245,
+      "learning_rate": 6.982266855778356e-06,
+      "loss": 0.6694,
+      "step": 6080
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3567417454775035,
+      "learning_rate": 6.981315261098529e-06,
+      "loss": 0.7253,
+      "step": 6081
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6281740810379361,
+      "learning_rate": 6.980363581277172e-06,
+      "loss": 0.6188,
+      "step": 6082
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8220895968334148,
+      "learning_rate": 6.979411816355183e-06,
+      "loss": 0.6274,
+      "step": 6083
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2662594522003474,
+      "learning_rate": 6.978459966373458e-06,
+      "loss": 0.648,
+      "step": 6084
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.499251491450255,
+      "learning_rate": 6.9775080313729026e-06,
+      "loss": 0.8189,
+      "step": 6085
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4605302687915587,
+      "learning_rate": 6.9765560113944235e-06,
+      "loss": 0.6687,
+      "step": 6086
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5538514949630868,
+      "learning_rate": 6.9756039064789295e-06,
+      "loss": 0.6253,
+      "step": 6087
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2536222570918503,
+      "learning_rate": 6.9746517166673375e-06,
+      "loss": 0.7017,
+      "step": 6088
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1931738311710305,
+      "learning_rate": 6.973699442000561e-06,
+      "loss": 0.672,
+      "step": 6089
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2519734369286015,
+      "learning_rate": 6.972747082519526e-06,
+      "loss": 0.6514,
+      "step": 6090
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8637631147814677,
+      "learning_rate": 6.971794638265153e-06,
+      "loss": 0.6653,
+      "step": 6091
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.336436874208002,
+      "learning_rate": 6.970842109278375e-06,
+      "loss": 0.641,
+      "step": 6092
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2089909322749395,
+      "learning_rate": 6.969889495600122e-06,
+      "loss": 0.5864,
+      "step": 6093
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2619711124514152,
+      "learning_rate": 6.96893679727133e-06,
+      "loss": 0.6492,
+      "step": 6094
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0107079399158725,
+      "learning_rate": 6.967984014332938e-06,
+      "loss": 0.6651,
+      "step": 6095
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4729041669000758,
+      "learning_rate": 6.96703114682589e-06,
+      "loss": 0.6656,
+      "step": 6096
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4372700623601669,
+      "learning_rate": 6.966078194791133e-06,
+      "loss": 0.7463,
+      "step": 6097
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3085274788600019,
+      "learning_rate": 6.965125158269619e-06,
+      "loss": 0.6697,
+      "step": 6098
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4016666899599357,
+      "learning_rate": 6.9641720373022996e-06,
+      "loss": 0.6822,
+      "step": 6099
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3722988642017537,
+      "learning_rate": 6.963218831930132e-06,
+      "loss": 0.7708,
+      "step": 6100
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0441799498306608,
+      "learning_rate": 6.96226554219408e-06,
+      "loss": 0.7,
+      "step": 6101
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8150582181959136,
+      "learning_rate": 6.961312168135107e-06,
+      "loss": 0.7483,
+      "step": 6102
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.150325370046868,
+      "learning_rate": 6.960358709794182e-06,
+      "loss": 0.6311,
+      "step": 6103
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3180406836219387,
+      "learning_rate": 6.959405167212278e-06,
+      "loss": 0.7616,
+      "step": 6104
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0836150365990533,
+      "learning_rate": 6.9584515404303685e-06,
+      "loss": 0.6155,
+      "step": 6105
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.078548636024581,
+      "learning_rate": 6.9574978294894355e-06,
+      "loss": 0.5819,
+      "step": 6106
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1872689750458236,
+      "learning_rate": 6.956544034430463e-06,
+      "loss": 0.812,
+      "step": 6107
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1413486017378032,
+      "learning_rate": 6.955590155294434e-06,
+      "loss": 0.5858,
+      "step": 6108
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3623187875905205,
+      "learning_rate": 6.954636192122339e-06,
+      "loss": 0.6428,
+      "step": 6109
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3616074677317638,
+      "learning_rate": 6.953682144955176e-06,
+      "loss": 0.6075,
+      "step": 6110
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.199657616624627,
+      "learning_rate": 6.952728013833941e-06,
+      "loss": 0.6,
+      "step": 6111
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2482811816473205,
+      "learning_rate": 6.951773798799632e-06,
+      "loss": 0.6273,
+      "step": 6112
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4057683351065084,
+      "learning_rate": 6.950819499893257e-06,
+      "loss": 0.6917,
+      "step": 6113
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4744477306513817,
+      "learning_rate": 6.949865117155823e-06,
+      "loss": 0.7085,
+      "step": 6114
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.561447457593074,
+      "learning_rate": 6.948910650628343e-06,
+      "loss": 0.7609,
+      "step": 6115
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1460340362098291,
+      "learning_rate": 6.947956100351833e-06,
+      "loss": 0.6062,
+      "step": 6116
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2743809254784273,
+      "learning_rate": 6.947001466367309e-06,
+      "loss": 0.5378,
+      "step": 6117
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3551571709141574,
+      "learning_rate": 6.946046748715796e-06,
+      "loss": 0.6792,
+      "step": 6118
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.591669938661219,
+      "learning_rate": 6.94509194743832e-06,
+      "loss": 0.5456,
+      "step": 6119
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.799479002970984,
+      "learning_rate": 6.944137062575913e-06,
+      "loss": 0.6668,
+      "step": 6120
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2992879377417874,
+      "learning_rate": 6.943182094169605e-06,
+      "loss": 0.6998,
+      "step": 6121
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.551416638367394,
+      "learning_rate": 6.942227042260435e-06,
+      "loss": 0.773,
+      "step": 6122
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.599590556234692,
+      "learning_rate": 6.941271906889443e-06,
+      "loss": 0.6823,
+      "step": 6123
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.497495831856004,
+      "learning_rate": 6.940316688097675e-06,
+      "loss": 0.6826,
+      "step": 6124
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3838686544687406,
+      "learning_rate": 6.9393613859261755e-06,
+      "loss": 0.5822,
+      "step": 6125
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3201599479268067,
+      "learning_rate": 6.938406000416e-06,
+      "loss": 0.6923,
+      "step": 6126
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6949975214524338,
+      "learning_rate": 6.9374505316082e-06,
+      "loss": 0.7567,
+      "step": 6127
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3140620941531245,
+      "learning_rate": 6.936494979543836e-06,
+      "loss": 0.646,
+      "step": 6128
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1401828533619094,
+      "learning_rate": 6.935539344263971e-06,
+      "loss": 0.7511,
+      "step": 6129
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6832727554378772,
+      "learning_rate": 6.93458362580967e-06,
+      "loss": 0.8157,
+      "step": 6130
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2786954866252314,
+      "learning_rate": 6.9336278242220015e-06,
+      "loss": 0.6345,
+      "step": 6131
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3227610204792841,
+      "learning_rate": 6.932671939542037e-06,
+      "loss": 0.6083,
+      "step": 6132
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.305845567376167,
+      "learning_rate": 6.9317159718108575e-06,
+      "loss": 0.5339,
+      "step": 6133
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3918838239454947,
+      "learning_rate": 6.93075992106954e-06,
+      "loss": 0.6973,
+      "step": 6134
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3259333919555205,
+      "learning_rate": 6.929803787359169e-06,
+      "loss": 0.6127,
+      "step": 6135
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3660717218951035,
+      "learning_rate": 6.928847570720829e-06,
+      "loss": 0.6215,
+      "step": 6136
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4568558063599357,
+      "learning_rate": 6.927891271195616e-06,
+      "loss": 0.799,
+      "step": 6137
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2559697778194883,
+      "learning_rate": 6.926934888824619e-06,
+      "loss": 0.6845,
+      "step": 6138
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3051371269856522,
+      "learning_rate": 6.925978423648941e-06,
+      "loss": 0.6701,
+      "step": 6139
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.267770676625802,
+      "learning_rate": 6.92502187570968e-06,
+      "loss": 0.6581,
+      "step": 6140
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2424536004476585,
+      "learning_rate": 6.92406524504794e-06,
+      "loss": 0.6914,
+      "step": 6141
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0976093337215802,
+      "learning_rate": 6.923108531704832e-06,
+      "loss": 0.6207,
+      "step": 6142
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2079767500936347,
+      "learning_rate": 6.9221517357214674e-06,
+      "loss": 0.6078,
+      "step": 6143
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4630586718526009,
+      "learning_rate": 6.921194857138963e-06,
+      "loss": 0.602,
+      "step": 6144
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4851803650965771,
+      "learning_rate": 6.920237895998436e-06,
+      "loss": 0.6165,
+      "step": 6145
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5393513204485054,
+      "learning_rate": 6.919280852341011e-06,
+      "loss": 0.6298,
+      "step": 6146
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.908605170856812,
+      "learning_rate": 6.91832372620781e-06,
+      "loss": 0.599,
+      "step": 6147
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3101202838977961,
+      "learning_rate": 6.917366517639969e-06,
+      "loss": 0.6648,
+      "step": 6148
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3093602731468301,
+      "learning_rate": 6.91640922667862e-06,
+      "loss": 0.666,
+      "step": 6149
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4355690093801596,
+      "learning_rate": 6.9154518533648966e-06,
+      "loss": 0.8167,
+      "step": 6150
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3734313480975366,
+      "learning_rate": 6.91449439773994e-06,
+      "loss": 0.6856,
+      "step": 6151
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4603044366072992,
+      "learning_rate": 6.913536859844896e-06,
+      "loss": 0.5688,
+      "step": 6152
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1363051871228425,
+      "learning_rate": 6.912579239720913e-06,
+      "loss": 0.6225,
+      "step": 6153
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.224056727399907,
+      "learning_rate": 6.911621537409139e-06,
+      "loss": 0.9078,
+      "step": 6154
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3762194550657079,
+      "learning_rate": 6.9106637529507305e-06,
+      "loss": 0.642,
+      "step": 6155
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3383495083693018,
+      "learning_rate": 6.909705886386844e-06,
+      "loss": 0.6916,
+      "step": 6156
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5635240162222823,
+      "learning_rate": 6.908747937758644e-06,
+      "loss": 0.74,
+      "step": 6157
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.182584372483705,
+      "learning_rate": 6.9077899071072954e-06,
+      "loss": 0.5406,
+      "step": 6158
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2452290268678967,
+      "learning_rate": 6.906831794473963e-06,
+      "loss": 0.6981,
+      "step": 6159
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.254947564224067,
+      "learning_rate": 6.9058735998998224e-06,
+      "loss": 0.67,
+      "step": 6160
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3353344583545133,
+      "learning_rate": 6.9049153234260495e-06,
+      "loss": 0.8269,
+      "step": 6161
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4697093913971218,
+      "learning_rate": 6.903956965093821e-06,
+      "loss": 0.6235,
+      "step": 6162
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3032509251547448,
+      "learning_rate": 6.902998524944322e-06,
+      "loss": 0.7,
+      "step": 6163
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4171996806473752,
+      "learning_rate": 6.9020400030187394e-06,
+      "loss": 0.6372,
+      "step": 6164
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.002994964373694,
+      "learning_rate": 6.90108139935826e-06,
+      "loss": 0.749,
+      "step": 6165
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.399421335189733,
+      "learning_rate": 6.90012271400408e-06,
+      "loss": 0.6585,
+      "step": 6166
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3295066540090843,
+      "learning_rate": 6.899163946997396e-06,
+      "loss": 0.6349,
+      "step": 6167
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.338675347854111,
+      "learning_rate": 6.898205098379407e-06,
+      "loss": 0.728,
+      "step": 6168
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2175436666465014,
+      "learning_rate": 6.897246168191317e-06,
+      "loss": 0.592,
+      "step": 6169
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.323710549430772,
+      "learning_rate": 6.8962871564743325e-06,
+      "loss": 0.5959,
+      "step": 6170
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1202175162656343,
+      "learning_rate": 6.895328063269667e-06,
+      "loss": 0.7362,
+      "step": 6171
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4565079800279512,
+      "learning_rate": 6.894368888618534e-06,
+      "loss": 0.7024,
+      "step": 6172
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.181975039866851,
+      "learning_rate": 6.8934096325621505e-06,
+      "loss": 0.7684,
+      "step": 6173
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4620164971394936,
+      "learning_rate": 6.892450295141737e-06,
+      "loss": 0.5541,
+      "step": 6174
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4487621997162676,
+      "learning_rate": 6.891490876398521e-06,
+      "loss": 0.6072,
+      "step": 6175
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.430136038073101,
+      "learning_rate": 6.890531376373729e-06,
+      "loss": 0.7543,
+      "step": 6176
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3512312989099113,
+      "learning_rate": 6.889571795108594e-06,
+      "loss": 0.6579,
+      "step": 6177
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4169470848911239,
+      "learning_rate": 6.88861213264435e-06,
+      "loss": 0.5731,
+      "step": 6178
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.299594756198441,
+      "learning_rate": 6.887652389022236e-06,
+      "loss": 0.6479,
+      "step": 6179
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5378589907984523,
+      "learning_rate": 6.886692564283496e-06,
+      "loss": 0.6919,
+      "step": 6180
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.241582444961384,
+      "learning_rate": 6.885732658469374e-06,
+      "loss": 0.5487,
+      "step": 6181
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2153263916515569,
+      "learning_rate": 6.88477267162112e-06,
+      "loss": 0.6268,
+      "step": 6182
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4183669701541908,
+      "learning_rate": 6.883812603779985e-06,
+      "loss": 0.6687,
+      "step": 6183
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.293130890798889,
+      "learning_rate": 6.88285245498723e-06,
+      "loss": 0.7,
+      "step": 6184
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2883541907542004,
+      "learning_rate": 6.88189222528411e-06,
+      "loss": 0.6325,
+      "step": 6185
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.401237998686181,
+      "learning_rate": 6.8809319147118905e-06,
+      "loss": 0.9336,
+      "step": 6186
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1931075749298232,
+      "learning_rate": 6.879971523311837e-06,
+      "loss": 0.5656,
+      "step": 6187
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3727891440309916,
+      "learning_rate": 6.8790110511252195e-06,
+      "loss": 0.6942,
+      "step": 6188
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5602071017697539,
+      "learning_rate": 6.878050498193314e-06,
+      "loss": 0.7326,
+      "step": 6189
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5396324875485492,
+      "learning_rate": 6.877089864557395e-06,
+      "loss": 0.7352,
+      "step": 6190
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6395241603053146,
+      "learning_rate": 6.876129150258744e-06,
+      "loss": 0.6593,
+      "step": 6191
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4232169003114576,
+      "learning_rate": 6.875168355338645e-06,
+      "loss": 0.6948,
+      "step": 6192
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3698343708626677,
+      "learning_rate": 6.874207479838386e-06,
+      "loss": 0.6408,
+      "step": 6193
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5519856609596718,
+      "learning_rate": 6.873246523799256e-06,
+      "loss": 0.7072,
+      "step": 6194
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.338584752395911,
+      "learning_rate": 6.872285487262555e-06,
+      "loss": 0.7614,
+      "step": 6195
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6032480523570127,
+      "learning_rate": 6.871324370269573e-06,
+      "loss": 0.7559,
+      "step": 6196
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.33078197623551,
+      "learning_rate": 6.870363172861616e-06,
+      "loss": 0.5948,
+      "step": 6197
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2461572495900723,
+      "learning_rate": 6.869401895079988e-06,
+      "loss": 0.6757,
+      "step": 6198
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2649334105346868,
+      "learning_rate": 6.868440536965997e-06,
+      "loss": 0.7739,
+      "step": 6199
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4742517348954147,
+      "learning_rate": 6.867479098560956e-06,
+      "loss": 0.7076,
+      "step": 6200
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5091499204101355,
+      "learning_rate": 6.866517579906178e-06,
+      "loss": 0.6795,
+      "step": 6201
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.210885672714905,
+      "learning_rate": 6.865555981042983e-06,
+      "loss": 0.6295,
+      "step": 6202
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2203133086645763,
+      "learning_rate": 6.864594302012691e-06,
+      "loss": 0.7823,
+      "step": 6203
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1990118300714967,
+      "learning_rate": 6.863632542856632e-06,
+      "loss": 0.5949,
+      "step": 6204
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5488810357024132,
+      "learning_rate": 6.8626707036161314e-06,
+      "loss": 0.793,
+      "step": 6205
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6051805511716435,
+      "learning_rate": 6.86170878433252e-06,
+      "loss": 0.6707,
+      "step": 6206
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9668304376984388,
+      "learning_rate": 6.860746785047139e-06,
+      "loss": 0.6165,
+      "step": 6207
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2987779686936178,
+      "learning_rate": 6.859784705801323e-06,
+      "loss": 0.6506,
+      "step": 6208
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.127703977077069,
+      "learning_rate": 6.858822546636417e-06,
+      "loss": 0.6328,
+      "step": 6209
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4180191165258051,
+      "learning_rate": 6.857860307593764e-06,
+      "loss": 0.6293,
+      "step": 6210
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5504826205783087,
+      "learning_rate": 6.856897988714718e-06,
+      "loss": 0.6263,
+      "step": 6211
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2370429463247044,
+      "learning_rate": 6.855935590040629e-06,
+      "loss": 0.5877,
+      "step": 6212
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.455322705434293,
+      "learning_rate": 6.854973111612855e-06,
+      "loss": 0.7992,
+      "step": 6213
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3121275341956926,
+      "learning_rate": 6.854010553472757e-06,
+      "loss": 0.6551,
+      "step": 6214
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2642517554587092,
+      "learning_rate": 6.8530479156616925e-06,
+      "loss": 0.6383,
+      "step": 6215
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3775523510923504,
+      "learning_rate": 6.852085198221035e-06,
+      "loss": 0.7,
+      "step": 6216
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2017824020177221,
+      "learning_rate": 6.851122401192149e-06,
+      "loss": 0.6486,
+      "step": 6217
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4202940693518247,
+      "learning_rate": 6.850159524616413e-06,
+      "loss": 0.7302,
+      "step": 6218
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.245117781293269,
+      "learning_rate": 6.849196568535201e-06,
+      "loss": 0.6459,
+      "step": 6219
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6783090024632124,
+      "learning_rate": 6.848233532989894e-06,
+      "loss": 0.7071,
+      "step": 6220
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3820364827110674,
+      "learning_rate": 6.847270418021876e-06,
+      "loss": 0.6809,
+      "step": 6221
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4940628263055902,
+      "learning_rate": 6.8463072236725345e-06,
+      "loss": 0.6991,
+      "step": 6222
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.258787001426038,
+      "learning_rate": 6.845343949983258e-06,
+      "loss": 0.6676,
+      "step": 6223
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4183265439656414,
+      "learning_rate": 6.8443805969954445e-06,
+      "loss": 0.745,
+      "step": 6224
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4381209485387791,
+      "learning_rate": 6.8434171647504885e-06,
+      "loss": 0.6645,
+      "step": 6225
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3771149232069935,
+      "learning_rate": 6.84245365328979e-06,
+      "loss": 0.6373,
+      "step": 6226
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.137345397783944,
+      "learning_rate": 6.841490062654757e-06,
+      "loss": 0.8112,
+      "step": 6227
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2991096519246934,
+      "learning_rate": 6.840526392886793e-06,
+      "loss": 0.6696,
+      "step": 6228
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2995097873748416,
+      "learning_rate": 6.839562644027311e-06,
+      "loss": 0.7157,
+      "step": 6229
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2839994852813736,
+      "learning_rate": 6.838598816117725e-06,
+      "loss": 0.553,
+      "step": 6230
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4026615605586876,
+      "learning_rate": 6.837634909199453e-06,
+      "loss": 0.6381,
+      "step": 6231
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.441420381360402,
+      "learning_rate": 6.836670923313917e-06,
+      "loss": 0.594,
+      "step": 6232
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3542138007486233,
+      "learning_rate": 6.835706858502542e-06,
+      "loss": 0.5993,
+      "step": 6233
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1497372429604504,
+      "learning_rate": 6.834742714806754e-06,
+      "loss": 0.6266,
+      "step": 6234
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4799425526257,
+      "learning_rate": 6.833778492267984e-06,
+      "loss": 0.7585,
+      "step": 6235
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.378849171057668,
+      "learning_rate": 6.832814190927669e-06,
+      "loss": 0.6048,
+      "step": 6236
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2509319584756802,
+      "learning_rate": 6.831849810827247e-06,
+      "loss": 0.6139,
+      "step": 6237
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2541686559175715,
+      "learning_rate": 6.830885352008159e-06,
+      "loss": 0.6605,
+      "step": 6238
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3556575505214226,
+      "learning_rate": 6.8299208145118475e-06,
+      "loss": 0.6192,
+      "step": 6239
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2797267601223299,
+      "learning_rate": 6.828956198379766e-06,
+      "loss": 0.6043,
+      "step": 6240
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3183518054730308,
+      "learning_rate": 6.827991503653363e-06,
+      "loss": 0.7164,
+      "step": 6241
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1864135458948446,
+      "learning_rate": 6.827026730374093e-06,
+      "loss": 0.7364,
+      "step": 6242
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1596660910836654,
+      "learning_rate": 6.826061878583416e-06,
+      "loss": 0.6514,
+      "step": 6243
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4486222643339768,
+      "learning_rate": 6.825096948322791e-06,
+      "loss": 0.704,
+      "step": 6244
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1720351124442268,
+      "learning_rate": 6.824131939633688e-06,
+      "loss": 0.654,
+      "step": 6245
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4609279311248,
+      "learning_rate": 6.823166852557572e-06,
+      "loss": 0.6349,
+      "step": 6246
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2971847322088368,
+      "learning_rate": 6.822201687135916e-06,
+      "loss": 0.5983,
+      "step": 6247
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1589159137868505,
+      "learning_rate": 6.821236443410195e-06,
+      "loss": 0.5779,
+      "step": 6248
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2849003726492314,
+      "learning_rate": 6.820271121421889e-06,
+      "loss": 0.6267,
+      "step": 6249
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.283495722076771,
+      "learning_rate": 6.819305721212479e-06,
+      "loss": 0.5773,
+      "step": 6250
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.315628065265775,
+      "learning_rate": 6.818340242823449e-06,
+      "loss": 0.6341,
+      "step": 6251
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0260455105498307,
+      "learning_rate": 6.8173746862962904e-06,
+      "loss": 0.6451,
+      "step": 6252
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5385633581017302,
+      "learning_rate": 6.816409051672492e-06,
+      "loss": 0.6698,
+      "step": 6253
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3049187439716128,
+      "learning_rate": 6.815443338993554e-06,
+      "loss": 0.5927,
+      "step": 6254
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2189239679816648,
+      "learning_rate": 6.81447754830097e-06,
+      "loss": 0.6951,
+      "step": 6255
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3197441022846217,
+      "learning_rate": 6.813511679636247e-06,
+      "loss": 0.6545,
+      "step": 6256
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.276831873493978,
+      "learning_rate": 6.8125457330408865e-06,
+      "loss": 0.691,
+      "step": 6257
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1289897805698363,
+      "learning_rate": 6.8115797085564e-06,
+      "loss": 0.5566,
+      "step": 6258
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1731151274682106,
+      "learning_rate": 6.810613606224299e-06,
+      "loss": 0.6191,
+      "step": 6259
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2982156102611768,
+      "learning_rate": 6.809647426086098e-06,
+      "loss": 0.647,
+      "step": 6260
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3231968120202873,
+      "learning_rate": 6.808681168183317e-06,
+      "loss": 0.7402,
+      "step": 6261
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4172391208837583,
+      "learning_rate": 6.807714832557479e-06,
+      "loss": 0.5652,
+      "step": 6262
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2639470560541461,
+      "learning_rate": 6.806748419250107e-06,
+      "loss": 0.5996,
+      "step": 6263
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5464311580556491,
+      "learning_rate": 6.805781928302732e-06,
+      "loss": 0.7558,
+      "step": 6264
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2905378050815057,
+      "learning_rate": 6.804815359756887e-06,
+      "loss": 0.61,
+      "step": 6265
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5391473624893757,
+      "learning_rate": 6.803848713654105e-06,
+      "loss": 0.7836,
+      "step": 6266
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3863154352013258,
+      "learning_rate": 6.8028819900359276e-06,
+      "loss": 0.6045,
+      "step": 6267
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4198925585935958,
+      "learning_rate": 6.801915188943895e-06,
+      "loss": 0.6001,
+      "step": 6268
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5858714712568789,
+      "learning_rate": 6.800948310419554e-06,
+      "loss": 0.7738,
+      "step": 6269
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1641153991441087,
+      "learning_rate": 6.799981354504455e-06,
+      "loss": 0.6053,
+      "step": 6270
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4397116388902498,
+      "learning_rate": 6.799014321240145e-06,
+      "loss": 0.6883,
+      "step": 6271
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2427639874988354,
+      "learning_rate": 6.798047210668185e-06,
+      "loss": 0.5584,
+      "step": 6272
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4679280466261941,
+      "learning_rate": 6.797080022830131e-06,
+      "loss": 0.6012,
+      "step": 6273
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4353277733405998,
+      "learning_rate": 6.796112757767547e-06,
+      "loss": 0.6112,
+      "step": 6274
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.393322572311512,
+      "learning_rate": 6.795145415521997e-06,
+      "loss": 0.7384,
+      "step": 6275
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2486157063290704,
+      "learning_rate": 6.79417799613505e-06,
+      "loss": 0.6323,
+      "step": 6276
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3412580275441108,
+      "learning_rate": 6.793210499648281e-06,
+      "loss": 0.703,
+      "step": 6277
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5903846692752575,
+      "learning_rate": 6.792242926103261e-06,
+      "loss": 0.6553,
+      "step": 6278
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2797920974227235,
+      "learning_rate": 6.7912752755415716e-06,
+      "loss": 0.6924,
+      "step": 6279
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.541524034083124,
+      "learning_rate": 6.790307548004794e-06,
+      "loss": 0.7223,
+      "step": 6280
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0911869322162022,
+      "learning_rate": 6.7893397435345144e-06,
+      "loss": 0.6046,
+      "step": 6281
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2675947776135306,
+      "learning_rate": 6.788371862172319e-06,
+      "loss": 0.6317,
+      "step": 6282
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4368813020396913,
+      "learning_rate": 6.787403903959804e-06,
+      "loss": 0.678,
+      "step": 6283
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3406693706870918,
+      "learning_rate": 6.786435868938561e-06,
+      "loss": 0.5736,
+      "step": 6284
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4935561576240115,
+      "learning_rate": 6.785467757150193e-06,
+      "loss": 0.6612,
+      "step": 6285
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7284169103746128,
+      "learning_rate": 6.7844995686362955e-06,
+      "loss": 0.66,
+      "step": 6286
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3582416207900767,
+      "learning_rate": 6.7835313034384786e-06,
+      "loss": 0.6581,
+      "step": 6287
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.357309068366409,
+      "learning_rate": 6.782562961598349e-06,
+      "loss": 0.5707,
+      "step": 6288
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2747620255566696,
+      "learning_rate": 6.78159454315752e-06,
+      "loss": 0.7395,
+      "step": 6289
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3377637904291395,
+      "learning_rate": 6.780626048157606e-06,
+      "loss": 0.9698,
+      "step": 6290
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.089796793687096,
+      "learning_rate": 6.779657476640222e-06,
+      "loss": 0.6747,
+      "step": 6291
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3778098541680854,
+      "learning_rate": 6.778688828646996e-06,
+      "loss": 0.6687,
+      "step": 6292
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2057210415431665,
+      "learning_rate": 6.777720104219548e-06,
+      "loss": 0.6406,
+      "step": 6293
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2227949606387762,
+      "learning_rate": 6.776751303399509e-06,
+      "loss": 0.6763,
+      "step": 6294
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0762966034803145,
+      "learning_rate": 6.775782426228509e-06,
+      "loss": 0.648,
+      "step": 6295
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.271782426179737,
+      "learning_rate": 6.774813472748183e-06,
+      "loss": 0.6457,
+      "step": 6296
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.311647951909346,
+      "learning_rate": 6.773844443000169e-06,
+      "loss": 0.611,
+      "step": 6297
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.624956758526164,
+      "learning_rate": 6.77287533702611e-06,
+      "loss": 0.6934,
+      "step": 6298
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3697951941827564,
+      "learning_rate": 6.771906154867649e-06,
+      "loss": 0.6879,
+      "step": 6299
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3172294590369809,
+      "learning_rate": 6.770936896566434e-06,
+      "loss": 0.6292,
+      "step": 6300
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2972475571479645,
+      "learning_rate": 6.769967562164116e-06,
+      "loss": 0.7285,
+      "step": 6301
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0077090405042084,
+      "learning_rate": 6.768998151702351e-06,
+      "loss": 0.6494,
+      "step": 6302
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1060576114708378,
+      "learning_rate": 6.768028665222795e-06,
+      "loss": 0.7105,
+      "step": 6303
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.178359837635333,
+      "learning_rate": 6.767059102767109e-06,
+      "loss": 0.5936,
+      "step": 6304
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4238638454852528,
+      "learning_rate": 6.766089464376958e-06,
+      "loss": 0.8379,
+      "step": 6305
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3202645229904477,
+      "learning_rate": 6.76511975009401e-06,
+      "loss": 0.5957,
+      "step": 6306
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1289610392350182,
+      "learning_rate": 6.7641499599599355e-06,
+      "loss": 0.5855,
+      "step": 6307
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.8588522258260753,
+      "learning_rate": 6.763180094016409e-06,
+      "loss": 0.6075,
+      "step": 6308
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5048825258580725,
+      "learning_rate": 6.7622101523051045e-06,
+      "loss": 0.7963,
+      "step": 6309
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3132411368950039,
+      "learning_rate": 6.761240134867706e-06,
+      "loss": 0.6627,
+      "step": 6310
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2921226570358664,
+      "learning_rate": 6.760270041745898e-06,
+      "loss": 0.6273,
+      "step": 6311
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.284730887442815,
+      "learning_rate": 6.759299872981364e-06,
+      "loss": 0.6214,
+      "step": 6312
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4285536913997376,
+      "learning_rate": 6.758329628615796e-06,
+      "loss": 0.7302,
+      "step": 6313
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3923490077340048,
+      "learning_rate": 6.757359308690889e-06,
+      "loss": 0.6145,
+      "step": 6314
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1791171892388554,
+      "learning_rate": 6.756388913248339e-06,
+      "loss": 0.6616,
+      "step": 6315
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2940110799561195,
+      "learning_rate": 6.755418442329846e-06,
+      "loss": 0.6106,
+      "step": 6316
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.515558101198112,
+      "learning_rate": 6.754447895977113e-06,
+      "loss": 0.6349,
+      "step": 6317
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.425533736442929,
+      "learning_rate": 6.753477274231846e-06,
+      "loss": 0.4964,
+      "step": 6318
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3893106715243961,
+      "learning_rate": 6.7525065771357546e-06,
+      "loss": 0.7409,
+      "step": 6319
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.542051136447836,
+      "learning_rate": 6.751535804730553e-06,
+      "loss": 0.7413,
+      "step": 6320
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1925840739325244,
+      "learning_rate": 6.750564957057958e-06,
+      "loss": 0.6415,
+      "step": 6321
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5093978915445205,
+      "learning_rate": 6.749594034159687e-06,
+      "loss": 0.7051,
+      "step": 6322
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7980446216140202,
+      "learning_rate": 6.748623036077464e-06,
+      "loss": 0.7659,
+      "step": 6323
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4612523333795109,
+      "learning_rate": 6.7476519628530145e-06,
+      "loss": 0.6703,
+      "step": 6324
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.404226356784962,
+      "learning_rate": 6.7466808145280685e-06,
+      "loss": 0.5823,
+      "step": 6325
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3984652884559348,
+      "learning_rate": 6.745709591144358e-06,
+      "loss": 0.642,
+      "step": 6326
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.531400920468216,
+      "learning_rate": 6.744738292743617e-06,
+      "loss": 0.6191,
+      "step": 6327
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2084001938116757,
+      "learning_rate": 6.743766919367588e-06,
+      "loss": 0.6778,
+      "step": 6328
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2130186109383674,
+      "learning_rate": 6.742795471058009e-06,
+      "loss": 0.7332,
+      "step": 6329
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2959002508774204,
+      "learning_rate": 6.741823947856627e-06,
+      "loss": 0.7546,
+      "step": 6330
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.495912352322573,
+      "learning_rate": 6.740852349805193e-06,
+      "loss": 0.6433,
+      "step": 6331
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3442367318959554,
+      "learning_rate": 6.739880676945455e-06,
+      "loss": 0.6114,
+      "step": 6332
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3166338650560547,
+      "learning_rate": 6.73890892931917e-06,
+      "loss": 0.7039,
+      "step": 6333
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2929268091513657,
+      "learning_rate": 6.737937106968094e-06,
+      "loss": 0.7111,
+      "step": 6334
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7062155578498441,
+      "learning_rate": 6.736965209933992e-06,
+      "loss": 0.6934,
+      "step": 6335
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0754637833868357,
+      "learning_rate": 6.735993238258624e-06,
+      "loss": 0.6093,
+      "step": 6336
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3076129748188534,
+      "learning_rate": 6.735021191983763e-06,
+      "loss": 0.5719,
+      "step": 6337
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4332918029604587,
+      "learning_rate": 6.734049071151175e-06,
+      "loss": 0.8119,
+      "step": 6338
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.470562813788377,
+      "learning_rate": 6.7330768758026374e-06,
+      "loss": 0.7287,
+      "step": 6339
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4081839707571546,
+      "learning_rate": 6.732104605979928e-06,
+      "loss": 0.7266,
+      "step": 6340
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.486676663506164,
+      "learning_rate": 6.7311322617248246e-06,
+      "loss": 0.7504,
+      "step": 6341
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1239213678117297,
+      "learning_rate": 6.730159843079113e-06,
+      "loss": 0.4974,
+      "step": 6342
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.287897885450683,
+      "learning_rate": 6.729187350084579e-06,
+      "loss": 0.7057,
+      "step": 6343
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4051005601739748,
+      "learning_rate": 6.728214782783013e-06,
+      "loss": 0.5962,
+      "step": 6344
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.289105873372394,
+      "learning_rate": 6.72724214121621e-06,
+      "loss": 0.6826,
+      "step": 6345
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1119155049355165,
+      "learning_rate": 6.726269425425967e-06,
+      "loss": 0.5667,
+      "step": 6346
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3231269911870134,
+      "learning_rate": 6.725296635454079e-06,
+      "loss": 0.6542,
+      "step": 6347
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3707776780515015,
+      "learning_rate": 6.724323771342354e-06,
+      "loss": 0.6951,
+      "step": 6348
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3616158757330585,
+      "learning_rate": 6.723350833132596e-06,
+      "loss": 0.6978,
+      "step": 6349
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6069831151429697,
+      "learning_rate": 6.722377820866614e-06,
+      "loss": 0.6985,
+      "step": 6350
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3429718523498264,
+      "learning_rate": 6.721404734586221e-06,
+      "loss": 0.6679,
+      "step": 6351
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1869170958479085,
+      "learning_rate": 6.7204315743332325e-06,
+      "loss": 0.6698,
+      "step": 6352
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.472388477090318,
+      "learning_rate": 6.71945834014947e-06,
+      "loss": 0.646,
+      "step": 6353
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.204317470488074,
+      "learning_rate": 6.7184850320767505e-06,
+      "loss": 0.7198,
+      "step": 6354
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8313691460200174,
+      "learning_rate": 6.717511650156903e-06,
+      "loss": 0.6666,
+      "step": 6355
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.068954642064932,
+      "learning_rate": 6.716538194431754e-06,
+      "loss": 0.5244,
+      "step": 6356
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.422748345256986,
+      "learning_rate": 6.715564664943138e-06,
+      "loss": 0.6829,
+      "step": 6357
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.591079682739026,
+      "learning_rate": 6.714591061732886e-06,
+      "loss": 0.7864,
+      "step": 6358
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.421273977806412,
+      "learning_rate": 6.7136173848428375e-06,
+      "loss": 0.6972,
+      "step": 6359
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5728950336734093,
+      "learning_rate": 6.712643634314834e-06,
+      "loss": 0.7401,
+      "step": 6360
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2635626964109024,
+      "learning_rate": 6.711669810190717e-06,
+      "loss": 0.7278,
+      "step": 6361
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5555924630684286,
+      "learning_rate": 6.710695912512339e-06,
+      "loss": 0.6457,
+      "step": 6362
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2868891144966947,
+      "learning_rate": 6.7097219413215474e-06,
+      "loss": 0.654,
+      "step": 6363
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2767409151394278,
+      "learning_rate": 6.708747896660196e-06,
+      "loss": 0.6498,
+      "step": 6364
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3722240526317082,
+      "learning_rate": 6.70777377857014e-06,
+      "loss": 0.5686,
+      "step": 6365
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.112264777153277,
+      "learning_rate": 6.706799587093243e-06,
+      "loss": 0.599,
+      "step": 6366
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5805929487671004,
+      "learning_rate": 6.705825322271367e-06,
+      "loss": 0.6295,
+      "step": 6367
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5603766787302131,
+      "learning_rate": 6.7048509841463775e-06,
+      "loss": 0.6003,
+      "step": 6368
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.476414500469196,
+      "learning_rate": 6.703876572760144e-06,
+      "loss": 0.7167,
+      "step": 6369
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3758485515827692,
+      "learning_rate": 6.702902088154539e-06,
+      "loss": 0.6967,
+      "step": 6370
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7673613181048546,
+      "learning_rate": 6.701927530371439e-06,
+      "loss": 0.6984,
+      "step": 6371
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.250092720824177,
+      "learning_rate": 6.700952899452723e-06,
+      "loss": 0.5485,
+      "step": 6372
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2827736548158786,
+      "learning_rate": 6.699978195440273e-06,
+      "loss": 0.7542,
+      "step": 6373
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4580777336918538,
+      "learning_rate": 6.6990034183759726e-06,
+      "loss": 0.6828,
+      "step": 6374
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.556013178488352,
+      "learning_rate": 6.698028568301712e-06,
+      "loss": 0.711,
+      "step": 6375
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.310068572547958,
+      "learning_rate": 6.697053645259382e-06,
+      "loss": 0.6558,
+      "step": 6376
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.196991499921567,
+      "learning_rate": 6.696078649290878e-06,
+      "loss": 0.6132,
+      "step": 6377
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4211226517724307,
+      "learning_rate": 6.695103580438097e-06,
+      "loss": 0.7017,
+      "step": 6378
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2965778583129894,
+      "learning_rate": 6.694128438742939e-06,
+      "loss": 0.6518,
+      "step": 6379
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3902844729599313,
+      "learning_rate": 6.693153224247309e-06,
+      "loss": 0.7389,
+      "step": 6380
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3121203982271699,
+      "learning_rate": 6.692177936993114e-06,
+      "loss": 0.603,
+      "step": 6381
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3419237132713848,
+      "learning_rate": 6.691202577022267e-06,
+      "loss": 0.7941,
+      "step": 6382
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4301185075407528,
+      "learning_rate": 6.690227144376676e-06,
+      "loss": 0.6328,
+      "step": 6383
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4610255988191705,
+      "learning_rate": 6.689251639098261e-06,
+      "loss": 0.7735,
+      "step": 6384
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5240979363482894,
+      "learning_rate": 6.68827606122894e-06,
+      "loss": 0.5784,
+      "step": 6385
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.290674245950422,
+      "learning_rate": 6.687300410810638e-06,
+      "loss": 0.5359,
+      "step": 6386
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3408905067269934,
+      "learning_rate": 6.686324687885281e-06,
+      "loss": 0.761,
+      "step": 6387
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5572434050769821,
+      "learning_rate": 6.685348892494793e-06,
+      "loss": 0.6683,
+      "step": 6388
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4941976547480786,
+      "learning_rate": 6.684373024681112e-06,
+      "loss": 0.7329,
+      "step": 6389
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1287092371644263,
+      "learning_rate": 6.683397084486169e-06,
+      "loss": 0.7725,
+      "step": 6390
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5418220384766914,
+      "learning_rate": 6.682421071951907e-06,
+      "loss": 0.7101,
+      "step": 6391
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1381435989776372,
+      "learning_rate": 6.681444987120262e-06,
+      "loss": 0.7988,
+      "step": 6392
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3004296396278072,
+      "learning_rate": 6.680468830033182e-06,
+      "loss": 0.5971,
+      "step": 6393
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.407729817527568,
+      "learning_rate": 6.679492600732614e-06,
+      "loss": 0.7194,
+      "step": 6394
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3945190455946535,
+      "learning_rate": 6.678516299260507e-06,
+      "loss": 0.6041,
+      "step": 6395
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4486203509060385,
+      "learning_rate": 6.677539925658819e-06,
+      "loss": 0.7861,
+      "step": 6396
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.410859732086959,
+      "learning_rate": 6.676563479969502e-06,
+      "loss": 0.7576,
+      "step": 6397
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.757740936287316,
+      "learning_rate": 6.67558696223452e-06,
+      "loss": 0.641,
+      "step": 6398
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5070725099088096,
+      "learning_rate": 6.674610372495832e-06,
+      "loss": 0.667,
+      "step": 6399
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4549836565536267,
+      "learning_rate": 6.673633710795409e-06,
+      "loss": 0.6997,
+      "step": 6400
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3996898867080125,
+      "learning_rate": 6.672656977175218e-06,
+      "loss": 0.6468,
+      "step": 6401
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.117034254647557,
+      "learning_rate": 6.671680171677232e-06,
+      "loss": 0.6471,
+      "step": 6402
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2311284151814204,
+      "learning_rate": 6.670703294343423e-06,
+      "loss": 0.6258,
+      "step": 6403
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3735298828746763,
+      "learning_rate": 6.669726345215776e-06,
+      "loss": 0.6145,
+      "step": 6404
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3024717946351967,
+      "learning_rate": 6.668749324336268e-06,
+      "loss": 0.6541,
+      "step": 6405
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4206888841106067,
+      "learning_rate": 6.667772231746885e-06,
+      "loss": 0.5972,
+      "step": 6406
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4960545016467295,
+      "learning_rate": 6.666795067489615e-06,
+      "loss": 0.6675,
+      "step": 6407
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3704823001096242,
+      "learning_rate": 6.665817831606447e-06,
+      "loss": 0.6442,
+      "step": 6408
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4047239388743251,
+      "learning_rate": 6.66484052413938e-06,
+      "loss": 0.6671,
+      "step": 6409
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1188784157341607,
+      "learning_rate": 6.663863145130406e-06,
+      "loss": 0.6673,
+      "step": 6410
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3791296051944077,
+      "learning_rate": 6.662885694621527e-06,
+      "loss": 0.6351,
+      "step": 6411
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.180905453808124,
+      "learning_rate": 6.661908172654746e-06,
+      "loss": 0.6863,
+      "step": 6412
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3970173692497339,
+      "learning_rate": 6.6609305792720705e-06,
+      "loss": 0.6715,
+      "step": 6413
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.547865074590845,
+      "learning_rate": 6.659952914515508e-06,
+      "loss": 0.8415,
+      "step": 6414
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.599773884943437,
+      "learning_rate": 6.658975178427073e-06,
+      "loss": 0.6953,
+      "step": 6415
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3123799283055577,
+      "learning_rate": 6.65799737104878e-06,
+      "loss": 0.7655,
+      "step": 6416
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3407608824719097,
+      "learning_rate": 6.657019492422645e-06,
+      "loss": 0.7289,
+      "step": 6417
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6205907303423885,
+      "learning_rate": 6.656041542590694e-06,
+      "loss": 0.7424,
+      "step": 6418
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3569173442714257,
+      "learning_rate": 6.65506352159495e-06,
+      "loss": 0.705,
+      "step": 6419
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.449295336372914,
+      "learning_rate": 6.65408542947744e-06,
+      "loss": 0.6404,
+      "step": 6420
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4224339097311292,
+      "learning_rate": 6.653107266280194e-06,
+      "loss": 0.7045,
+      "step": 6421
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1153171062497613,
+      "learning_rate": 6.652129032045249e-06,
+      "loss": 0.6041,
+      "step": 6422
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4424908478873295,
+      "learning_rate": 6.651150726814639e-06,
+      "loss": 0.6476,
+      "step": 6423
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2581874605997645,
+      "learning_rate": 6.650172350630406e-06,
+      "loss": 0.7188,
+      "step": 6424
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.329286405706879,
+      "learning_rate": 6.649193903534591e-06,
+      "loss": 0.6177,
+      "step": 6425
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.456155963576878,
+      "learning_rate": 6.6482153855692395e-06,
+      "loss": 0.6164,
+      "step": 6426
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3979286349726137,
+      "learning_rate": 6.647236796776405e-06,
+      "loss": 0.662,
+      "step": 6427
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6255142711339867,
+      "learning_rate": 6.646258137198136e-06,
+      "loss": 0.7303,
+      "step": 6428
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4326710080344975,
+      "learning_rate": 6.645279406876488e-06,
+      "loss": 0.6942,
+      "step": 6429
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.598671809912353,
+      "learning_rate": 6.644300605853518e-06,
+      "loss": 0.7163,
+      "step": 6430
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4893649864366292,
+      "learning_rate": 6.643321734171291e-06,
+      "loss": 0.639,
+      "step": 6431
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3084011118198702,
+      "learning_rate": 6.642342791871868e-06,
+      "loss": 0.6434,
+      "step": 6432
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3294850327238508,
+      "learning_rate": 6.64136377899732e-06,
+      "loss": 0.7107,
+      "step": 6433
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0353351260431545,
+      "learning_rate": 6.640384695589714e-06,
+      "loss": 0.6515,
+      "step": 6434
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3145570067236985,
+      "learning_rate": 6.639405541691122e-06,
+      "loss": 0.6588,
+      "step": 6435
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2899833757233639,
+      "learning_rate": 6.638426317343624e-06,
+      "loss": 0.6556,
+      "step": 6436
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4941038355904928,
+      "learning_rate": 6.6374470225892985e-06,
+      "loss": 0.8019,
+      "step": 6437
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0946207995608173,
+      "learning_rate": 6.6364676574702284e-06,
+      "loss": 0.6991,
+      "step": 6438
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1962321956450137,
+      "learning_rate": 6.635488222028497e-06,
+      "loss": 0.6573,
+      "step": 6439
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.13928249314855,
+      "learning_rate": 6.6345087163061935e-06,
+      "loss": 0.6239,
+      "step": 6440
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0215659228469236,
+      "learning_rate": 6.633529140345412e-06,
+      "loss": 0.7042,
+      "step": 6441
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4339934455102306,
+      "learning_rate": 6.632549494188245e-06,
+      "loss": 0.7331,
+      "step": 6442
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.410227069061665,
+      "learning_rate": 6.631569777876791e-06,
+      "loss": 0.6745,
+      "step": 6443
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.396139921218717,
+      "learning_rate": 6.630589991453148e-06,
+      "loss": 0.705,
+      "step": 6444
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.37301848761108,
+      "learning_rate": 6.629610134959423e-06,
+      "loss": 0.6521,
+      "step": 6445
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.322816710743683,
+      "learning_rate": 6.628630208437719e-06,
+      "loss": 0.6853,
+      "step": 6446
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3990182635948263,
+      "learning_rate": 6.627650211930152e-06,
+      "loss": 0.809,
+      "step": 6447
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.370466928689467,
+      "learning_rate": 6.626670145478826e-06,
+      "loss": 0.6491,
+      "step": 6448
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2824736029566404,
+      "learning_rate": 6.6256900091258644e-06,
+      "loss": 0.7459,
+      "step": 6449
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0844959980118118,
+      "learning_rate": 6.62470980291338e-06,
+      "loss": 0.6138,
+      "step": 6450
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.356296991728247,
+      "learning_rate": 6.623729526883498e-06,
+      "loss": 0.7132,
+      "step": 6451
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.322764151398057,
+      "learning_rate": 6.622749181078344e-06,
+      "loss": 0.6501,
+      "step": 6452
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8840742142825808,
+      "learning_rate": 6.621768765540041e-06,
+      "loss": 0.6783,
+      "step": 6453
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4026850415541756,
+      "learning_rate": 6.620788280310722e-06,
+      "loss": 0.7162,
+      "step": 6454
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.361167515396772,
+      "learning_rate": 6.619807725432522e-06,
+      "loss": 0.6873,
+      "step": 6455
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.366308615853848,
+      "learning_rate": 6.618827100947577e-06,
+      "loss": 0.6149,
+      "step": 6456
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4572770244413524,
+      "learning_rate": 6.6178464068980256e-06,
+      "loss": 0.6752,
+      "step": 6457
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.424033255505512,
+      "learning_rate": 6.616865643326011e-06,
+      "loss": 0.6031,
+      "step": 6458
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3154071983710784,
+      "learning_rate": 6.615884810273678e-06,
+      "loss": 0.6741,
+      "step": 6459
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5477175985014682,
+      "learning_rate": 6.614903907783177e-06,
+      "loss": 0.6504,
+      "step": 6460
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3464270600271415,
+      "learning_rate": 6.613922935896659e-06,
+      "loss": 0.6201,
+      "step": 6461
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2447902167365077,
+      "learning_rate": 6.612941894656277e-06,
+      "loss": 0.5419,
+      "step": 6462
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4131387290296664,
+      "learning_rate": 6.61196078410419e-06,
+      "loss": 0.7385,
+      "step": 6463
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.348831861550452,
+      "learning_rate": 6.610979604282557e-06,
+      "loss": 0.702,
+      "step": 6464
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4497774277200732,
+      "learning_rate": 6.609998355233543e-06,
+      "loss": 0.6278,
+      "step": 6465
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5202001089791253,
+      "learning_rate": 6.609017036999314e-06,
+      "loss": 0.6843,
+      "step": 6466
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3867512566477214,
+      "learning_rate": 6.608035649622041e-06,
+      "loss": 0.7355,
+      "step": 6467
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.442233439316492,
+      "learning_rate": 6.607054193143894e-06,
+      "loss": 0.7301,
+      "step": 6468
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.881921726028489,
+      "learning_rate": 6.606072667607048e-06,
+      "loss": 0.7838,
+      "step": 6469
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3716622337191893,
+      "learning_rate": 6.605091073053685e-06,
+      "loss": 0.7141,
+      "step": 6470
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0727681210671776,
+      "learning_rate": 6.604109409525982e-06,
+      "loss": 0.6049,
+      "step": 6471
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4233260142273718,
+      "learning_rate": 6.603127677066126e-06,
+      "loss": 0.6859,
+      "step": 6472
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7435766030312823,
+      "learning_rate": 6.602145875716301e-06,
+      "loss": 0.8287,
+      "step": 6473
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.345179245373518,
+      "learning_rate": 6.601164005518702e-06,
+      "loss": 0.6697,
+      "step": 6474
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3345358120600872,
+      "learning_rate": 6.600182066515519e-06,
+      "loss": 0.742,
+      "step": 6475
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.410074862318654,
+      "learning_rate": 6.599200058748949e-06,
+      "loss": 0.6344,
+      "step": 6476
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4560290432909078,
+      "learning_rate": 6.59821798226119e-06,
+      "loss": 0.6442,
+      "step": 6477
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9829502963014012,
+      "learning_rate": 6.597235837094446e-06,
+      "loss": 0.609,
+      "step": 6478
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5470255120228065,
+      "learning_rate": 6.59625362329092e-06,
+      "loss": 0.7416,
+      "step": 6479
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3665869795029013,
+      "learning_rate": 6.595271340892823e-06,
+      "loss": 0.6615,
+      "step": 6480
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2842307775084592,
+      "learning_rate": 6.594288989942361e-06,
+      "loss": 0.6709,
+      "step": 6481
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.912208433813456,
+      "learning_rate": 6.593306570481751e-06,
+      "loss": 0.5426,
+      "step": 6482
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.270734921442222,
+      "learning_rate": 6.592324082553209e-06,
+      "loss": 0.5918,
+      "step": 6483
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.315162835490268,
+      "learning_rate": 6.591341526198955e-06,
+      "loss": 0.7004,
+      "step": 6484
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8482567095622415,
+      "learning_rate": 6.590358901461213e-06,
+      "loss": 0.6492,
+      "step": 6485
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.287453400926302,
+      "learning_rate": 6.589376208382206e-06,
+      "loss": 0.6217,
+      "step": 6486
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.6755039736058768,
+      "learning_rate": 6.588393447004163e-06,
+      "loss": 0.6521,
+      "step": 6487
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5117145208229903,
+      "learning_rate": 6.587410617369316e-06,
+      "loss": 0.7307,
+      "step": 6488
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3581291522285563,
+      "learning_rate": 6.586427719519901e-06,
+      "loss": 0.7074,
+      "step": 6489
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1572266764330965,
+      "learning_rate": 6.585444753498153e-06,
+      "loss": 0.6848,
+      "step": 6490
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4368116039522991,
+      "learning_rate": 6.584461719346312e-06,
+      "loss": 0.727,
+      "step": 6491
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4414195440584872,
+      "learning_rate": 6.583478617106623e-06,
+      "loss": 0.6413,
+      "step": 6492
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3168999260819352,
+      "learning_rate": 6.582495446821331e-06,
+      "loss": 0.6712,
+      "step": 6493
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4605391427943133,
+      "learning_rate": 6.581512208532685e-06,
+      "loss": 0.6213,
+      "step": 6494
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.468551966175783,
+      "learning_rate": 6.580528902282938e-06,
+      "loss": 0.6256,
+      "step": 6495
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2401315604679755,
+      "learning_rate": 6.579545528114344e-06,
+      "loss": 0.5854,
+      "step": 6496
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2363126372087145,
+      "learning_rate": 6.57856208606916e-06,
+      "loss": 0.6404,
+      "step": 6497
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.000306553421967,
+      "learning_rate": 6.577578576189648e-06,
+      "loss": 0.7183,
+      "step": 6498
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2729663252449461,
+      "learning_rate": 6.576594998518071e-06,
+      "loss": 0.6633,
+      "step": 6499
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2584155967234718,
+      "learning_rate": 6.575611353096695e-06,
+      "loss": 0.6236,
+      "step": 6500
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2871570321860277,
+      "learning_rate": 6.574627639967791e-06,
+      "loss": 0.581,
+      "step": 6501
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4533758465794675,
+      "learning_rate": 6.57364385917363e-06,
+      "loss": 0.6734,
+      "step": 6502
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5518646416027957,
+      "learning_rate": 6.572660010756489e-06,
+      "loss": 0.7874,
+      "step": 6503
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3117987184957403,
+      "learning_rate": 6.5716760947586425e-06,
+      "loss": 0.6719,
+      "step": 6504
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1847872096604726,
+      "learning_rate": 6.570692111222375e-06,
+      "loss": 0.6282,
+      "step": 6505
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1094633346564495,
+      "learning_rate": 6.56970806018997e-06,
+      "loss": 0.724,
+      "step": 6506
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.301787830561529,
+      "learning_rate": 6.568723941703712e-06,
+      "loss": 0.7362,
+      "step": 6507
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2404807598271255,
+      "learning_rate": 6.567739755805895e-06,
+      "loss": 0.6431,
+      "step": 6508
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0922823064523939,
+      "learning_rate": 6.566755502538806e-06,
+      "loss": 0.7118,
+      "step": 6509
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3875201885297892,
+      "learning_rate": 6.565771181944747e-06,
+      "loss": 0.6525,
+      "step": 6510
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2949472477610058,
+      "learning_rate": 6.5647867940660105e-06,
+      "loss": 0.5852,
+      "step": 6511
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3931765059792802,
+      "learning_rate": 6.563802338944902e-06,
+      "loss": 0.6186,
+      "step": 6512
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1780555413134974,
+      "learning_rate": 6.562817816623724e-06,
+      "loss": 0.6414,
+      "step": 6513
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4398264670947676,
+      "learning_rate": 6.561833227144784e-06,
+      "loss": 0.6227,
+      "step": 6514
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2179035999407253,
+      "learning_rate": 6.560848570550391e-06,
+      "loss": 0.6179,
+      "step": 6515
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2677051292948258,
+      "learning_rate": 6.559863846882859e-06,
+      "loss": 0.5703,
+      "step": 6516
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1627338489846137,
+      "learning_rate": 6.558879056184505e-06,
+      "loss": 0.5849,
+      "step": 6517
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4091039501338727,
+      "learning_rate": 6.557894198497644e-06,
+      "loss": 0.6107,
+      "step": 6518
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3621795045221947,
+      "learning_rate": 6.556909273864601e-06,
+      "loss": 0.6205,
+      "step": 6519
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3438196714969202,
+      "learning_rate": 6.555924282327698e-06,
+      "loss": 0.6588,
+      "step": 6520
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.561402313847184,
+      "learning_rate": 6.554939223929264e-06,
+      "loss": 0.6082,
+      "step": 6521
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0793908200253333,
+      "learning_rate": 6.553954098711629e-06,
+      "loss": 0.6105,
+      "step": 6522
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5242150380426491,
+      "learning_rate": 6.5529689067171255e-06,
+      "loss": 0.6833,
+      "step": 6523
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6419518456982,
+      "learning_rate": 6.551983647988089e-06,
+      "loss": 0.7282,
+      "step": 6524
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7783110756593068,
+      "learning_rate": 6.550998322566857e-06,
+      "loss": 0.7121,
+      "step": 6525
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3817206952429324,
+      "learning_rate": 6.550012930495775e-06,
+      "loss": 0.7115,
+      "step": 6526
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3221729859356655,
+      "learning_rate": 6.549027471817186e-06,
+      "loss": 0.6737,
+      "step": 6527
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.354418848556309,
+      "learning_rate": 6.5480419465734355e-06,
+      "loss": 0.6803,
+      "step": 6528
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1512532433385936,
+      "learning_rate": 6.547056354806874e-06,
+      "loss": 0.5582,
+      "step": 6529
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.929621961586048,
+      "learning_rate": 6.546070696559856e-06,
+      "loss": 0.6281,
+      "step": 6530
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2433796684760525,
+      "learning_rate": 6.545084971874738e-06,
+      "loss": 0.733,
+      "step": 6531
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4952214506055819,
+      "learning_rate": 6.544099180793877e-06,
+      "loss": 0.7026,
+      "step": 6532
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2794568893980713,
+      "learning_rate": 6.543113323359635e-06,
+      "loss": 0.5908,
+      "step": 6533
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4266326688583186,
+      "learning_rate": 6.542127399614376e-06,
+      "loss": 0.7176,
+      "step": 6534
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2569793573705885,
+      "learning_rate": 6.54114140960047e-06,
+      "loss": 0.617,
+      "step": 6535
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0319605470360513,
+      "learning_rate": 6.540155353360284e-06,
+      "loss": 0.6734,
+      "step": 6536
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1951239142174763,
+      "learning_rate": 6.539169230936193e-06,
+      "loss": 0.6365,
+      "step": 6537
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3646340106943098,
+      "learning_rate": 6.5381830423705714e-06,
+      "loss": 0.7379,
+      "step": 6538
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2750651706784635,
+      "learning_rate": 6.5371967877058e-06,
+      "loss": 0.6422,
+      "step": 6539
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8513622348536698,
+      "learning_rate": 6.536210466984259e-06,
+      "loss": 0.6537,
+      "step": 6540
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1371485950808327,
+      "learning_rate": 6.535224080248332e-06,
+      "loss": 0.6473,
+      "step": 6541
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4749079872881328,
+      "learning_rate": 6.534237627540408e-06,
+      "loss": 0.5959,
+      "step": 6542
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2088449480519163,
+      "learning_rate": 6.533251108902875e-06,
+      "loss": 0.6294,
+      "step": 6543
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8402671877658026,
+      "learning_rate": 6.532264524378128e-06,
+      "loss": 0.7662,
+      "step": 6544
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1357367603724615,
+      "learning_rate": 6.531277874008562e-06,
+      "loss": 0.5755,
+      "step": 6545
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6184059989722408,
+      "learning_rate": 6.5302911578365755e-06,
+      "loss": 0.7275,
+      "step": 6546
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1354494008987157,
+      "learning_rate": 6.529304375904568e-06,
+      "loss": 0.5338,
+      "step": 6547
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5821609729543946,
+      "learning_rate": 6.528317528254948e-06,
+      "loss": 0.7446,
+      "step": 6548
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.386250822855528,
+      "learning_rate": 6.52733061493012e-06,
+      "loss": 0.5979,
+      "step": 6549
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4859423262049583,
+      "learning_rate": 6.526343635972492e-06,
+      "loss": 0.6877,
+      "step": 6550
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5219179238465719,
+      "learning_rate": 6.525356591424479e-06,
+      "loss": 0.6578,
+      "step": 6551
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8416413039522093,
+      "learning_rate": 6.5243694813284975e-06,
+      "loss": 0.6431,
+      "step": 6552
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4448873029002471,
+      "learning_rate": 6.523382305726964e-06,
+      "loss": 0.6478,
+      "step": 6553
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5718045702376937,
+      "learning_rate": 6.522395064662299e-06,
+      "loss": 0.6225,
+      "step": 6554
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3633733510653196,
+      "learning_rate": 6.521407758176929e-06,
+      "loss": 0.6542,
+      "step": 6555
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2942515183053795,
+      "learning_rate": 6.520420386313277e-06,
+      "loss": 0.7284,
+      "step": 6556
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1748556259067495,
+      "learning_rate": 6.5194329491137775e-06,
+      "loss": 0.6226,
+      "step": 6557
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3040878646488465,
+      "learning_rate": 6.518445446620858e-06,
+      "loss": 0.6435,
+      "step": 6558
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2898323199706616,
+      "learning_rate": 6.517457878876958e-06,
+      "loss": 0.6342,
+      "step": 6559
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3485270545224264,
+      "learning_rate": 6.516470245924511e-06,
+      "loss": 0.611,
+      "step": 6560
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3193073702693299,
+      "learning_rate": 6.515482547805962e-06,
+      "loss": 0.7236,
+      "step": 6561
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0580446356985416,
+      "learning_rate": 6.514494784563753e-06,
+      "loss": 0.5826,
+      "step": 6562
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3752910897785022,
+      "learning_rate": 6.5135069562403285e-06,
+      "loss": 0.7401,
+      "step": 6563
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4575392056592582,
+      "learning_rate": 6.512519062878142e-06,
+      "loss": 0.6539,
+      "step": 6564
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.33716729214082,
+      "learning_rate": 6.511531104519639e-06,
+      "loss": 0.6391,
+      "step": 6565
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4418069240957219,
+      "learning_rate": 6.510543081207281e-06,
+      "loss": 0.6799,
+      "step": 6566
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3869183601320345,
+      "learning_rate": 6.509554992983521e-06,
+      "loss": 0.5726,
+      "step": 6567
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3354492825440356,
+      "learning_rate": 6.508566839890823e-06,
+      "loss": 0.5787,
+      "step": 6568
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.502054634414479,
+      "learning_rate": 6.507578621971646e-06,
+      "loss": 0.6949,
+      "step": 6569
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4498816513049937,
+      "learning_rate": 6.506590339268459e-06,
+      "loss": 0.7592,
+      "step": 6570
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3070898182976298,
+      "learning_rate": 6.50560199182373e-06,
+      "loss": 0.6681,
+      "step": 6571
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.277196328646991,
+      "learning_rate": 6.504613579679929e-06,
+      "loss": 0.6588,
+      "step": 6572
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4141852760443774,
+      "learning_rate": 6.503625102879534e-06,
+      "loss": 0.6672,
+      "step": 6573
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3234237182067865,
+      "learning_rate": 6.502636561465018e-06,
+      "loss": 0.6317,
+      "step": 6574
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7300192484561279,
+      "learning_rate": 6.501647955478863e-06,
+      "loss": 0.7217,
+      "step": 6575
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4849931453806158,
+      "learning_rate": 6.500659284963549e-06,
+      "loss": 0.6768,
+      "step": 6576
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.530200626279419,
+      "learning_rate": 6.499670549961565e-06,
+      "loss": 0.6772,
+      "step": 6577
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4857542251876863,
+      "learning_rate": 6.498681750515398e-06,
+      "loss": 0.7667,
+      "step": 6578
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8856206408774248,
+      "learning_rate": 6.497692886667537e-06,
+      "loss": 0.6887,
+      "step": 6579
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4485677914994182,
+      "learning_rate": 6.496703958460479e-06,
+      "loss": 0.7197,
+      "step": 6580
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3309912829988868,
+      "learning_rate": 6.495714965936717e-06,
+      "loss": 0.7534,
+      "step": 6581
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8911881717960943,
+      "learning_rate": 6.494725909138751e-06,
+      "loss": 0.496,
+      "step": 6582
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2604067438089321,
+      "learning_rate": 6.493736788109086e-06,
+      "loss": 0.732,
+      "step": 6583
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3683628514166248,
+      "learning_rate": 6.492747602890223e-06,
+      "loss": 0.6485,
+      "step": 6584
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.107530712206856,
+      "learning_rate": 6.491758353524672e-06,
+      "loss": 0.5392,
+      "step": 6585
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0808075595717093,
+      "learning_rate": 6.490769040054942e-06,
+      "loss": 0.6931,
+      "step": 6586
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3765828939573337,
+      "learning_rate": 6.489779662523545e-06,
+      "loss": 0.7091,
+      "step": 6587
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.231777759171698,
+      "learning_rate": 6.4887902209730004e-06,
+      "loss": 0.6454,
+      "step": 6588
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2542736000173385,
+      "learning_rate": 6.487800715445822e-06,
+      "loss": 0.672,
+      "step": 6589
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6446357847690394,
+      "learning_rate": 6.486811145984534e-06,
+      "loss": 0.6477,
+      "step": 6590
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2684550332319917,
+      "learning_rate": 6.48582151263166e-06,
+      "loss": 0.6287,
+      "step": 6591
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4283863133712411,
+      "learning_rate": 6.484831815429726e-06,
+      "loss": 0.6895,
+      "step": 6592
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5564873066019438,
+      "learning_rate": 6.483842054421263e-06,
+      "loss": 0.7447,
+      "step": 6593
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2344634466347226,
+      "learning_rate": 6.4828522296488014e-06,
+      "loss": 0.705,
+      "step": 6594
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2639818799715148,
+      "learning_rate": 6.481862341154878e-06,
+      "loss": 0.6149,
+      "step": 6595
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6246923558808812,
+      "learning_rate": 6.480872388982028e-06,
+      "loss": 0.6304,
+      "step": 6596
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4583992589089256,
+      "learning_rate": 6.479882373172795e-06,
+      "loss": 0.6696,
+      "step": 6597
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3366810697590523,
+      "learning_rate": 6.47889229376972e-06,
+      "loss": 0.6427,
+      "step": 6598
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4839809137797952,
+      "learning_rate": 6.477902150815347e-06,
+      "loss": 0.602,
+      "step": 6599
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4394149042727369,
+      "learning_rate": 6.476911944352231e-06,
+      "loss": 0.7023,
+      "step": 6600
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3185793264398928,
+      "learning_rate": 6.475921674422917e-06,
+      "loss": 0.5636,
+      "step": 6601
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.283738401271852,
+      "learning_rate": 6.474931341069963e-06,
+      "loss": 0.6582,
+      "step": 6602
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1032341131054144,
+      "learning_rate": 6.473940944335923e-06,
+      "loss": 0.6812,
+      "step": 6603
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4186297917376707,
+      "learning_rate": 6.472950484263359e-06,
+      "loss": 0.7336,
+      "step": 6604
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3217764445939912,
+      "learning_rate": 6.471959960894832e-06,
+      "loss": 0.6518,
+      "step": 6605
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1861134257534383,
+      "learning_rate": 6.470969374272908e-06,
+      "loss": 0.6467,
+      "step": 6606
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.321091662337821,
+      "learning_rate": 6.469978724440153e-06,
+      "loss": 0.6941,
+      "step": 6607
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6951553407907156,
+      "learning_rate": 6.4689880114391375e-06,
+      "loss": 0.6645,
+      "step": 6608
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2384843191902206,
+      "learning_rate": 6.467997235312437e-06,
+      "loss": 0.6573,
+      "step": 6609
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.265072755441111,
+      "learning_rate": 6.4670063961026265e-06,
+      "loss": 0.6639,
+      "step": 6610
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2031236896416457,
+      "learning_rate": 6.466015493852283e-06,
+      "loss": 0.6779,
+      "step": 6611
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4125820054130527,
+      "learning_rate": 6.465024528603987e-06,
+      "loss": 0.7576,
+      "step": 6612
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.581523895956334,
+      "learning_rate": 6.464033500400327e-06,
+      "loss": 0.6496,
+      "step": 6613
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2777251572587354,
+      "learning_rate": 6.463042409283885e-06,
+      "loss": 0.6893,
+      "step": 6614
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5739002195510774,
+      "learning_rate": 6.462051255297255e-06,
+      "loss": 0.7114,
+      "step": 6615
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9954544519987386,
+      "learning_rate": 6.461060038483025e-06,
+      "loss": 0.7121,
+      "step": 6616
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5855829088321447,
+      "learning_rate": 6.460068758883789e-06,
+      "loss": 0.6651,
+      "step": 6617
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1901650352313966,
+      "learning_rate": 6.4590774165421485e-06,
+      "loss": 0.6116,
+      "step": 6618
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.066710868625082,
+      "learning_rate": 6.458086011500703e-06,
+      "loss": 0.759,
+      "step": 6619
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.479777163720274,
+      "learning_rate": 6.457094543802053e-06,
+      "loss": 0.7783,
+      "step": 6620
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.152016154063115,
+      "learning_rate": 6.456103013488804e-06,
+      "loss": 0.5895,
+      "step": 6621
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5224760052024071,
+      "learning_rate": 6.455111420603568e-06,
+      "loss": 0.7097,
+      "step": 6622
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2974401141113654,
+      "learning_rate": 6.454119765188951e-06,
+      "loss": 0.6247,
+      "step": 6623
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4971890005066584,
+      "learning_rate": 6.453128047287573e-06,
+      "loss": 0.6759,
+      "step": 6624
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5563463567124618,
+      "learning_rate": 6.452136266942044e-06,
+      "loss": 0.7566,
+      "step": 6625
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.675635297624497,
+      "learning_rate": 6.451144424194985e-06,
+      "loss": 0.6751,
+      "step": 6626
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.318941719365279,
+      "learning_rate": 6.450152519089019e-06,
+      "loss": 0.7979,
+      "step": 6627
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3320478491540644,
+      "learning_rate": 6.449160551666769e-06,
+      "loss": 0.6838,
+      "step": 6628
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3325557067201281,
+      "learning_rate": 6.448168521970865e-06,
+      "loss": 0.8094,
+      "step": 6629
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2058551112129146,
+      "learning_rate": 6.447176430043931e-06,
+      "loss": 0.684,
+      "step": 6630
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2411547834474541,
+      "learning_rate": 6.446184275928604e-06,
+      "loss": 0.58,
+      "step": 6631
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5867919353535613,
+      "learning_rate": 6.445192059667518e-06,
+      "loss": 0.7664,
+      "step": 6632
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3568405477225303,
+      "learning_rate": 6.444199781303311e-06,
+      "loss": 0.7488,
+      "step": 6633
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1431600962912274,
+      "learning_rate": 6.443207440878624e-06,
+      "loss": 0.6271,
+      "step": 6634
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1905296543334773,
+      "learning_rate": 6.442215038436097e-06,
+      "loss": 0.6973,
+      "step": 6635
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4504975097754174,
+      "learning_rate": 6.441222574018378e-06,
+      "loss": 0.7837,
+      "step": 6636
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.268533260049299,
+      "learning_rate": 6.440230047668116e-06,
+      "loss": 0.6951,
+      "step": 6637
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5471884100826674,
+      "learning_rate": 6.4392374594279614e-06,
+      "loss": 0.6435,
+      "step": 6638
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4422367717246551,
+      "learning_rate": 6.438244809340568e-06,
+      "loss": 0.6944,
+      "step": 6639
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6211494791511671,
+      "learning_rate": 6.4372520974485916e-06,
+      "loss": 0.5991,
+      "step": 6640
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.504627096886663,
+      "learning_rate": 6.436259323794691e-06,
+      "loss": 0.709,
+      "step": 6641
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3014105926302757,
+      "learning_rate": 6.43526648842153e-06,
+      "loss": 0.6125,
+      "step": 6642
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5129613517203664,
+      "learning_rate": 6.434273591371771e-06,
+      "loss": 0.6475,
+      "step": 6643
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9257292621053868,
+      "learning_rate": 6.43328063268808e-06,
+      "loss": 0.6023,
+      "step": 6644
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2384144468620113,
+      "learning_rate": 6.432287612413131e-06,
+      "loss": 0.589,
+      "step": 6645
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.364655138044537,
+      "learning_rate": 6.431294530589589e-06,
+      "loss": 0.7245,
+      "step": 6646
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3515461583355615,
+      "learning_rate": 6.4303013872601375e-06,
+      "loss": 0.6126,
+      "step": 6647
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9923935646943955,
+      "learning_rate": 6.4293081824674485e-06,
+      "loss": 0.7722,
+      "step": 6648
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.243804927027931,
+      "learning_rate": 6.428314916254203e-06,
+      "loss": 0.5762,
+      "step": 6649
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.368618489692791,
+      "learning_rate": 6.427321588663085e-06,
+      "loss": 0.5891,
+      "step": 6650
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3741803412827147,
+      "learning_rate": 6.42632819973678e-06,
+      "loss": 0.6604,
+      "step": 6651
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5801137778056598,
+      "learning_rate": 6.425334749517975e-06,
+      "loss": 0.6318,
+      "step": 6652
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4809423183280788,
+      "learning_rate": 6.424341238049362e-06,
+      "loss": 0.6783,
+      "step": 6653
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.530921241497288,
+      "learning_rate": 6.423347665373633e-06,
+      "loss": 0.7722,
+      "step": 6654
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3959307808135712,
+      "learning_rate": 6.422354031533485e-06,
+      "loss": 0.6884,
+      "step": 6655
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3395509760713973,
+      "learning_rate": 6.421360336571618e-06,
+      "loss": 0.5747,
+      "step": 6656
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4664637981318687,
+      "learning_rate": 6.42036658053073e-06,
+      "loss": 0.6354,
+      "step": 6657
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3435697208761437,
+      "learning_rate": 6.419372763453529e-06,
+      "loss": 0.6464,
+      "step": 6658
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2364348882448153,
+      "learning_rate": 6.418378885382716e-06,
+      "loss": 0.7074,
+      "step": 6659
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1930656174151615,
+      "learning_rate": 6.417384946361005e-06,
+      "loss": 0.6458,
+      "step": 6660
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4107869116238243,
+      "learning_rate": 6.416390946431107e-06,
+      "loss": 0.5879,
+      "step": 6661
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0994903634637505,
+      "learning_rate": 6.4153968856357355e-06,
+      "loss": 0.6312,
+      "step": 6662
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1533830057805066,
+      "learning_rate": 6.414402764017607e-06,
+      "loss": 0.59,
+      "step": 6663
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.25234638843396,
+      "learning_rate": 6.41340858161944e-06,
+      "loss": 0.6506,
+      "step": 6664
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4753974223005317,
+      "learning_rate": 6.412414338483961e-06,
+      "loss": 0.6632,
+      "step": 6665
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5433215043691542,
+      "learning_rate": 6.411420034653891e-06,
+      "loss": 0.6625,
+      "step": 6666
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2923875181391324,
+      "learning_rate": 6.410425670171959e-06,
+      "loss": 0.6798,
+      "step": 6667
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.405242068492109,
+      "learning_rate": 6.409431245080894e-06,
+      "loss": 0.5482,
+      "step": 6668
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.372331954576652,
+      "learning_rate": 6.408436759423431e-06,
+      "loss": 0.6439,
+      "step": 6669
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4186803262459133,
+      "learning_rate": 6.4074422132423035e-06,
+      "loss": 0.6624,
+      "step": 6670
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2790596747293015,
+      "learning_rate": 6.406447606580248e-06,
+      "loss": 0.6089,
+      "step": 6671
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.436741858199504,
+      "learning_rate": 6.405452939480008e-06,
+      "loss": 0.6682,
+      "step": 6672
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3131510672912374,
+      "learning_rate": 6.404458211984324e-06,
+      "loss": 0.7713,
+      "step": 6673
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5129457461139444,
+      "learning_rate": 6.403463424135943e-06,
+      "loss": 0.6407,
+      "step": 6674
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5185962666963277,
+      "learning_rate": 6.402468575977614e-06,
+      "loss": 0.6369,
+      "step": 6675
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2225375218159111,
+      "learning_rate": 6.401473667552085e-06,
+      "loss": 0.6337,
+      "step": 6676
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5372278966407231,
+      "learning_rate": 6.400478698902112e-06,
+      "loss": 0.63,
+      "step": 6677
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5537883090421667,
+      "learning_rate": 6.399483670070451e-06,
+      "loss": 0.6546,
+      "step": 6678
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1185870456485763,
+      "learning_rate": 6.398488581099859e-06,
+      "loss": 0.7412,
+      "step": 6679
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2572563556798089,
+      "learning_rate": 6.3974934320330996e-06,
+      "loss": 0.5823,
+      "step": 6680
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.208068492571065,
+      "learning_rate": 6.396498222912933e-06,
+      "loss": 0.6337,
+      "step": 6681
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2117316763965196,
+      "learning_rate": 6.395502953782129e-06,
+      "loss": 0.6276,
+      "step": 6682
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.643857601623222,
+      "learning_rate": 6.394507624683455e-06,
+      "loss": 0.7073,
+      "step": 6683
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1474500122114935,
+      "learning_rate": 6.393512235659681e-06,
+      "loss": 0.5627,
+      "step": 6684
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2446472389283583,
+      "learning_rate": 6.392516786753586e-06,
+      "loss": 0.6262,
+      "step": 6685
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.370957591183377,
+      "learning_rate": 6.391521278007939e-06,
+      "loss": 0.7282,
+      "step": 6686
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.138116533160141,
+      "learning_rate": 6.390525709465527e-06,
+      "loss": 0.6188,
+      "step": 6687
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0723790986542374,
+      "learning_rate": 6.389530081169126e-06,
+      "loss": 0.6324,
+      "step": 6688
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3276320791432943,
+      "learning_rate": 6.388534393161525e-06,
+      "loss": 0.6662,
+      "step": 6689
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1674063396639691,
+      "learning_rate": 6.38753864548551e-06,
+      "loss": 0.6507,
+      "step": 6690
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9328376345159664,
+      "learning_rate": 6.386542838183865e-06,
+      "loss": 0.5964,
+      "step": 6691
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6975217780245564,
+      "learning_rate": 6.385546971299389e-06,
+      "loss": 0.7712,
+      "step": 6692
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.156247054389509,
+      "learning_rate": 6.384551044874871e-06,
+      "loss": 0.6228,
+      "step": 6693
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.696496409682809,
+      "learning_rate": 6.383555058953115e-06,
+      "loss": 0.6945,
+      "step": 6694
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2607880082046625,
+      "learning_rate": 6.382559013576914e-06,
+      "loss": 0.6922,
+      "step": 6695
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1104957428312057,
+      "learning_rate": 6.381562908789075e-06,
+      "loss": 0.6476,
+      "step": 6696
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2428925195096203,
+      "learning_rate": 6.380566744632399e-06,
+      "loss": 0.7959,
+      "step": 6697
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.578002810783373,
+      "learning_rate": 6.379570521149696e-06,
+      "loss": 0.7135,
+      "step": 6698
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3378147761320465,
+      "learning_rate": 6.378574238383776e-06,
+      "loss": 0.6376,
+      "step": 6699
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.460891591184007,
+      "learning_rate": 6.377577896377451e-06,
+      "loss": 0.7012,
+      "step": 6700
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4525277806155394,
+      "learning_rate": 6.3765814951735346e-06,
+      "loss": 0.6151,
+      "step": 6701
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6001077552211065,
+      "learning_rate": 6.375585034814845e-06,
+      "loss": 0.6741,
+      "step": 6702
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4024595285446606,
+      "learning_rate": 6.374588515344204e-06,
+      "loss": 0.648,
+      "step": 6703
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.631825772628772,
+      "learning_rate": 6.373591936804433e-06,
+      "loss": 0.649,
+      "step": 6704
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3278009405514832,
+      "learning_rate": 6.372595299238357e-06,
+      "loss": 0.6239,
+      "step": 6705
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5589102296191293,
+      "learning_rate": 6.3715986026888046e-06,
+      "loss": 0.6475,
+      "step": 6706
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3701985461566208,
+      "learning_rate": 6.370601847198606e-06,
+      "loss": 0.6815,
+      "step": 6707
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4141008887672242,
+      "learning_rate": 6.3696050328105915e-06,
+      "loss": 0.6654,
+      "step": 6708
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.099620738147407,
+      "learning_rate": 6.3686081595676e-06,
+      "loss": 0.7434,
+      "step": 6709
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5329009258327229,
+      "learning_rate": 6.367611227512468e-06,
+      "loss": 0.7768,
+      "step": 6710
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2896358578610112,
+      "learning_rate": 6.366614236688034e-06,
+      "loss": 0.6616,
+      "step": 6711
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2659323865665726,
+      "learning_rate": 6.365617187137144e-06,
+      "loss": 0.6345,
+      "step": 6712
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3073072118106648,
+      "learning_rate": 6.3646200789026426e-06,
+      "loss": 0.6195,
+      "step": 6713
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2986373503516586,
+      "learning_rate": 6.3636229120273766e-06,
+      "loss": 0.6903,
+      "step": 6714
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.438211589790649,
+      "learning_rate": 6.362625686554196e-06,
+      "loss": 0.6994,
+      "step": 6715
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5501553961445225,
+      "learning_rate": 6.361628402525956e-06,
+      "loss": 0.7067,
+      "step": 6716
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4233928787659866,
+      "learning_rate": 6.360631059985512e-06,
+      "loss": 0.664,
+      "step": 6717
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3072137075065764,
+      "learning_rate": 6.35963365897572e-06,
+      "loss": 0.6153,
+      "step": 6718
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3841193986984432,
+      "learning_rate": 6.3586361995394415e-06,
+      "loss": 0.6639,
+      "step": 6719
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.599021146518122,
+      "learning_rate": 6.35763868171954e-06,
+      "loss": 0.7707,
+      "step": 6720
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8873374243372891,
+      "learning_rate": 6.3566411055588806e-06,
+      "loss": 0.5826,
+      "step": 6721
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4246711186390533,
+      "learning_rate": 6.355643471100332e-06,
+      "loss": 0.7172,
+      "step": 6722
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3844393476705115,
+      "learning_rate": 6.354645778386763e-06,
+      "loss": 0.6701,
+      "step": 6723
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3704962734962216,
+      "learning_rate": 6.353648027461048e-06,
+      "loss": 0.7776,
+      "step": 6724
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3412419200149772,
+      "learning_rate": 6.3526502183660646e-06,
+      "loss": 0.7438,
+      "step": 6725
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3912627687780645,
+      "learning_rate": 6.351652351144687e-06,
+      "loss": 0.7477,
+      "step": 6726
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2120296729211621,
+      "learning_rate": 6.350654425839799e-06,
+      "loss": 0.5504,
+      "step": 6727
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3734235965887094,
+      "learning_rate": 6.349656442494281e-06,
+      "loss": 0.6224,
+      "step": 6728
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.331100550341406,
+      "learning_rate": 6.348658401151018e-06,
+      "loss": 0.668,
+      "step": 6729
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3339623851304616,
+      "learning_rate": 6.347660301852903e-06,
+      "loss": 0.6473,
+      "step": 6730
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1131348590727221,
+      "learning_rate": 6.346662144642823e-06,
+      "loss": 0.5266,
+      "step": 6731
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.31065571575867,
+      "learning_rate": 6.345663929563672e-06,
+      "loss": 0.6136,
+      "step": 6732
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1729982498078402,
+      "learning_rate": 6.344665656658342e-06,
+      "loss": 0.6194,
+      "step": 6733
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5806240909141411,
+      "learning_rate": 6.343667325969736e-06,
+      "loss": 0.714,
+      "step": 6734
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1012920160768882,
+      "learning_rate": 6.342668937540752e-06,
+      "loss": 0.7755,
+      "step": 6735
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3023812758497155,
+      "learning_rate": 6.341670491414294e-06,
+      "loss": 0.5706,
+      "step": 6736
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3369088216283014,
+      "learning_rate": 6.3406719876332675e-06,
+      "loss": 0.5876,
+      "step": 6737
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2300337789136984,
+      "learning_rate": 6.339673426240578e-06,
+      "loss": 0.5669,
+      "step": 6738
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7989690718368576,
+      "learning_rate": 6.3386748072791395e-06,
+      "loss": 0.7461,
+      "step": 6739
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2558171304567582,
+      "learning_rate": 6.337676130791861e-06,
+      "loss": 0.5544,
+      "step": 6740
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6985791985545677,
+      "learning_rate": 6.336677396821663e-06,
+      "loss": 0.6871,
+      "step": 6741
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3869021110891115,
+      "learning_rate": 6.3356786054114585e-06,
+      "loss": 0.6806,
+      "step": 6742
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0884422844378134,
+      "learning_rate": 6.3346797566041695e-06,
+      "loss": 0.5733,
+      "step": 6743
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.864489620016627,
+      "learning_rate": 6.33368085044272e-06,
+      "loss": 0.7017,
+      "step": 6744
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2836952176590914,
+      "learning_rate": 6.332681886970032e-06,
+      "loss": 0.7181,
+      "step": 6745
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9392596762098067,
+      "learning_rate": 6.331682866229039e-06,
+      "loss": 0.6994,
+      "step": 6746
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5847428707233457,
+      "learning_rate": 6.3306837882626635e-06,
+      "loss": 0.6833,
+      "step": 6747
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6450477738163247,
+      "learning_rate": 6.3296846531138445e-06,
+      "loss": 0.7006,
+      "step": 6748
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4991512494768775,
+      "learning_rate": 6.328685460825512e-06,
+      "loss": 0.6766,
+      "step": 6749
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.512922011036466,
+      "learning_rate": 6.327686211440609e-06,
+      "loss": 0.6866,
+      "step": 6750
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2428872874508774,
+      "learning_rate": 6.32668690500207e-06,
+      "loss": 0.6309,
+      "step": 6751
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4047564855659425,
+      "learning_rate": 6.325687541552843e-06,
+      "loss": 0.6233,
+      "step": 6752
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3930192342915275,
+      "learning_rate": 6.324688121135868e-06,
+      "loss": 0.7843,
+      "step": 6753
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4292322461974842,
+      "learning_rate": 6.323688643794094e-06,
+      "loss": 0.7095,
+      "step": 6754
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5374500299540506,
+      "learning_rate": 6.322689109570472e-06,
+      "loss": 0.6947,
+      "step": 6755
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2756209881988858,
+      "learning_rate": 6.3216895185079515e-06,
+      "loss": 0.6079,
+      "step": 6756
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3236652020280686,
+      "learning_rate": 6.3206898706494915e-06,
+      "loss": 0.7595,
+      "step": 6757
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5271758320671462,
+      "learning_rate": 6.3196901660380436e-06,
+      "loss": 0.713,
+      "step": 6758
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.204212504344179,
+      "learning_rate": 6.318690404716572e-06,
+      "loss": 0.6976,
+      "step": 6759
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4943855279567733,
+      "learning_rate": 6.317690586728036e-06,
+      "loss": 0.7005,
+      "step": 6760
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4478503894196517,
+      "learning_rate": 6.316690712115401e-06,
+      "loss": 0.7731,
+      "step": 6761
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2215161383202777,
+      "learning_rate": 6.315690780921634e-06,
+      "loss": 0.6403,
+      "step": 6762
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4997526920815927,
+      "learning_rate": 6.314690793189703e-06,
+      "loss": 0.8103,
+      "step": 6763
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3041302682545688,
+      "learning_rate": 6.313690748962582e-06,
+      "loss": 0.7033,
+      "step": 6764
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.165924714194685,
+      "learning_rate": 6.312690648283243e-06,
+      "loss": 0.6492,
+      "step": 6765
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1493825889679798,
+      "learning_rate": 6.311690491194664e-06,
+      "loss": 0.6631,
+      "step": 6766
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5017597907367886,
+      "learning_rate": 6.310690277739821e-06,
+      "loss": 0.791,
+      "step": 6767
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7614206722734966,
+      "learning_rate": 6.3096900079617e-06,
+      "loss": 0.6977,
+      "step": 6768
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.669825390921468,
+      "learning_rate": 6.3086896819032814e-06,
+      "loss": 0.6748,
+      "step": 6769
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6577339870465313,
+      "learning_rate": 6.307689299607555e-06,
+      "loss": 0.756,
+      "step": 6770
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5633615184275975,
+      "learning_rate": 6.306688861117505e-06,
+      "loss": 0.6639,
+      "step": 6771
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3422730964789884,
+      "learning_rate": 6.305688366476124e-06,
+      "loss": 0.6848,
+      "step": 6772
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3230736602860111,
+      "learning_rate": 6.304687815726406e-06,
+      "loss": 0.6624,
+      "step": 6773
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.435755776305495,
+      "learning_rate": 6.303687208911348e-06,
+      "loss": 0.6884,
+      "step": 6774
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5083913975431416,
+      "learning_rate": 6.302686546073946e-06,
+      "loss": 0.6531,
+      "step": 6775
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5531223617242027,
+      "learning_rate": 6.301685827257202e-06,
+      "loss": 0.5982,
+      "step": 6776
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3533567528193133,
+      "learning_rate": 6.300685052504119e-06,
+      "loss": 0.6842,
+      "step": 6777
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.30718180008148,
+      "learning_rate": 6.299684221857705e-06,
+      "loss": 0.5594,
+      "step": 6778
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1987055477750819,
+      "learning_rate": 6.298683335360962e-06,
+      "loss": 0.6543,
+      "step": 6779
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4503861729250729,
+      "learning_rate": 6.297682393056906e-06,
+      "loss": 0.72,
+      "step": 6780
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6208234239764023,
+      "learning_rate": 6.296681394988546e-06,
+      "loss": 0.7745,
+      "step": 6781
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3581320759206943,
+      "learning_rate": 6.2956803411989e-06,
+      "loss": 0.6523,
+      "step": 6782
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2699082638879688,
+      "learning_rate": 6.294679231730983e-06,
+      "loss": 0.5926,
+      "step": 6783
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0921374987309878,
+      "learning_rate": 6.293678066627816e-06,
+      "loss": 0.7126,
+      "step": 6784
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0384224777720632,
+      "learning_rate": 6.292676845932422e-06,
+      "loss": 0.4759,
+      "step": 6785
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9166426759935171,
+      "learning_rate": 6.291675569687824e-06,
+      "loss": 0.6476,
+      "step": 6786
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5651260444536097,
+      "learning_rate": 6.290674237937051e-06,
+      "loss": 0.779,
+      "step": 6787
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5312489307050379,
+      "learning_rate": 6.289672850723132e-06,
+      "loss": 0.6146,
+      "step": 6788
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4862904869619595,
+      "learning_rate": 6.288671408089098e-06,
+      "loss": 0.6665,
+      "step": 6789
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2218947696789206,
+      "learning_rate": 6.2876699100779815e-06,
+      "loss": 0.5712,
+      "step": 6790
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3181001765863656,
+      "learning_rate": 6.2866683567328235e-06,
+      "loss": 0.6933,
+      "step": 6791
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1503518128229895,
+      "learning_rate": 6.285666748096659e-06,
+      "loss": 0.7548,
+      "step": 6792
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4302947810975457,
+      "learning_rate": 6.284665084212533e-06,
+      "loss": 0.6567,
+      "step": 6793
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3868936042150932,
+      "learning_rate": 6.283663365123486e-06,
+      "loss": 0.6859,
+      "step": 6794
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2794320867348612,
+      "learning_rate": 6.282661590872565e-06,
+      "loss": 0.6043,
+      "step": 6795
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3958295212606713,
+      "learning_rate": 6.281659761502819e-06,
+      "loss": 0.6876,
+      "step": 6796
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5864951610765903,
+      "learning_rate": 6.2806578770573e-06,
+      "loss": 0.7404,
+      "step": 6797
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.420021275417359,
+      "learning_rate": 6.279655937579058e-06,
+      "loss": 0.5796,
+      "step": 6798
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2749992084711863,
+      "learning_rate": 6.278653943111152e-06,
+      "loss": 0.663,
+      "step": 6799
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.27859273686327,
+      "learning_rate": 6.277651893696637e-06,
+      "loss": 0.6389,
+      "step": 6800
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0944300006568275,
+      "learning_rate": 6.276649789378575e-06,
+      "loss": 0.6402,
+      "step": 6801
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4838540656412906,
+      "learning_rate": 6.275647630200029e-06,
+      "loss": 0.6576,
+      "step": 6802
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3132126710451344,
+      "learning_rate": 6.274645416204062e-06,
+      "loss": 0.6781,
+      "step": 6803
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3107676325554978,
+      "learning_rate": 6.273643147433743e-06,
+      "loss": 0.6549,
+      "step": 6804
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4104929800227353,
+      "learning_rate": 6.272640823932141e-06,
+      "loss": 0.6564,
+      "step": 6805
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0012939596503827,
+      "learning_rate": 6.271638445742332e-06,
+      "loss": 0.6118,
+      "step": 6806
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.347259465844794,
+      "learning_rate": 6.270636012907383e-06,
+      "loss": 0.6506,
+      "step": 6807
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3745591015290395,
+      "learning_rate": 6.269633525470376e-06,
+      "loss": 0.5783,
+      "step": 6808
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3688597075824975,
+      "learning_rate": 6.268630983474388e-06,
+      "loss": 0.7463,
+      "step": 6809
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6001161383862832,
+      "learning_rate": 6.267628386962502e-06,
+      "loss": 0.8849,
+      "step": 6810
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3535355012720722,
+      "learning_rate": 6.266625735977802e-06,
+      "loss": 0.6922,
+      "step": 6811
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3194219371936375,
+      "learning_rate": 6.265623030563372e-06,
+      "loss": 0.6403,
+      "step": 6812
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3906117809865808,
+      "learning_rate": 6.264620270762302e-06,
+      "loss": 0.7018,
+      "step": 6813
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2177789674606807,
+      "learning_rate": 6.263617456617681e-06,
+      "loss": 0.5395,
+      "step": 6814
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2912390465124617,
+      "learning_rate": 6.262614588172607e-06,
+      "loss": 0.5928,
+      "step": 6815
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2971867531677432,
+      "learning_rate": 6.261611665470168e-06,
+      "loss": 0.6705,
+      "step": 6816
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3715755890145072,
+      "learning_rate": 6.260608688553469e-06,
+      "loss": 0.6039,
+      "step": 6817
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7525903893623476,
+      "learning_rate": 6.259605657465607e-06,
+      "loss": 0.7095,
+      "step": 6818
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3429868125015858,
+      "learning_rate": 6.258602572249683e-06,
+      "loss": 0.6531,
+      "step": 6819
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4951181779277427,
+      "learning_rate": 6.2575994329488045e-06,
+      "loss": 0.6613,
+      "step": 6820
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6484948521460872,
+      "learning_rate": 6.2565962396060774e-06,
+      "loss": 0.6736,
+      "step": 6821
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.374428496740465,
+      "learning_rate": 6.255592992264612e-06,
+      "loss": 0.6613,
+      "step": 6822
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1494912966461253,
+      "learning_rate": 6.254589690967517e-06,
+      "loss": 0.5748,
+      "step": 6823
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3965238177417312,
+      "learning_rate": 6.2535863357579105e-06,
+      "loss": 0.7829,
+      "step": 6824
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3574100940874079,
+      "learning_rate": 6.252582926678908e-06,
+      "loss": 0.7527,
+      "step": 6825
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2804300291490882,
+      "learning_rate": 6.2515794637736275e-06,
+      "loss": 0.7662,
+      "step": 6826
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3214437310506213,
+      "learning_rate": 6.250575947085189e-06,
+      "loss": 0.6329,
+      "step": 6827
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.484266684379491,
+      "learning_rate": 6.2495723766567165e-06,
+      "loss": 0.7315,
+      "step": 6828
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0731285031004165,
+      "learning_rate": 6.248568752531337e-06,
+      "loss": 0.6806,
+      "step": 6829
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.365303675765666,
+      "learning_rate": 6.247565074752177e-06,
+      "loss": 0.6858,
+      "step": 6830
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4673644950886762,
+      "learning_rate": 6.246561343362368e-06,
+      "loss": 0.6218,
+      "step": 6831
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.321099193091917,
+      "learning_rate": 6.24555755840504e-06,
+      "loss": 0.6943,
+      "step": 6832
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9255080097680393,
+      "learning_rate": 6.244553719923332e-06,
+      "loss": 0.5905,
+      "step": 6833
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.369366320394769,
+      "learning_rate": 6.243549827960378e-06,
+      "loss": 0.6368,
+      "step": 6834
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.396519546009984,
+      "learning_rate": 6.242545882559318e-06,
+      "loss": 0.7409,
+      "step": 6835
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.259305926379241,
+      "learning_rate": 6.241541883763294e-06,
+      "loss": 0.6296,
+      "step": 6836
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.320373250627707,
+      "learning_rate": 6.240537831615451e-06,
+      "loss": 0.5977,
+      "step": 6837
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3760613017678671,
+      "learning_rate": 6.239533726158934e-06,
+      "loss": 0.6122,
+      "step": 6838
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.097793078757008,
+      "learning_rate": 6.238529567436892e-06,
+      "loss": 0.6884,
+      "step": 6839
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.73722317254124,
+      "learning_rate": 6.2375253554924766e-06,
+      "loss": 0.7391,
+      "step": 6840
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0329853508440332,
+      "learning_rate": 6.23652109036884e-06,
+      "loss": 0.6434,
+      "step": 6841
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.175758501383523,
+      "learning_rate": 6.235516772109138e-06,
+      "loss": 0.6551,
+      "step": 6842
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.583856961235008,
+      "learning_rate": 6.23451240075653e-06,
+      "loss": 0.6504,
+      "step": 6843
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.433337794686723,
+      "learning_rate": 6.233507976354174e-06,
+      "loss": 0.6958,
+      "step": 6844
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.35558288748534,
+      "learning_rate": 6.2325034989452335e-06,
+      "loss": 0.703,
+      "step": 6845
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.937682537896577,
+      "learning_rate": 6.231498968572872e-06,
+      "loss": 0.6397,
+      "step": 6846
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2428050008414466,
+      "learning_rate": 6.230494385280257e-06,
+      "loss": 0.6272,
+      "step": 6847
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2813275144816019,
+      "learning_rate": 6.229489749110559e-06,
+      "loss": 0.662,
+      "step": 6848
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5428480988041184,
+      "learning_rate": 6.228485060106948e-06,
+      "loss": 0.6546,
+      "step": 6849
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3638794168474293,
+      "learning_rate": 6.227480318312597e-06,
+      "loss": 0.7901,
+      "step": 6850
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.382615719871637,
+      "learning_rate": 6.226475523770683e-06,
+      "loss": 0.6928,
+      "step": 6851
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.699945126297881,
+      "learning_rate": 6.225470676524385e-06,
+      "loss": 0.824,
+      "step": 6852
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1507909940448933,
+      "learning_rate": 6.224465776616883e-06,
+      "loss": 0.6352,
+      "step": 6853
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6185116980156158,
+      "learning_rate": 6.223460824091358e-06,
+      "loss": 0.69,
+      "step": 6854
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.483889449464025,
+      "learning_rate": 6.222455818990998e-06,
+      "loss": 0.8317,
+      "step": 6855
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4881515350816188,
+      "learning_rate": 6.221450761358988e-06,
+      "loss": 0.881,
+      "step": 6856
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1649777544463786,
+      "learning_rate": 6.220445651238519e-06,
+      "loss": 0.6634,
+      "step": 6857
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4635345956891441,
+      "learning_rate": 6.2194404886727835e-06,
+      "loss": 0.6928,
+      "step": 6858
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3576494102832208,
+      "learning_rate": 6.218435273704973e-06,
+      "loss": 0.6731,
+      "step": 6859
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1455548981742054,
+      "learning_rate": 6.217430006378285e-06,
+      "loss": 0.6119,
+      "step": 6860
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3093706566279075,
+      "learning_rate": 6.216424686735917e-06,
+      "loss": 0.6528,
+      "step": 6861
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.606886535394172,
+      "learning_rate": 6.215419314821074e-06,
+      "loss": 0.7054,
+      "step": 6862
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5905614747606494,
+      "learning_rate": 6.214413890676956e-06,
+      "loss": 0.7118,
+      "step": 6863
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3199383586821165,
+      "learning_rate": 6.213408414346765e-06,
+      "loss": 0.5635,
+      "step": 6864
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3505777320182566,
+      "learning_rate": 6.212402885873715e-06,
+      "loss": 0.7178,
+      "step": 6865
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6080458353394744,
+      "learning_rate": 6.211397305301011e-06,
+      "loss": 0.699,
+      "step": 6866
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3650990461588213,
+      "learning_rate": 6.210391672671869e-06,
+      "loss": 0.6535,
+      "step": 6867
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3147803966875864,
+      "learning_rate": 6.2093859880295e-06,
+      "loss": 0.5685,
+      "step": 6868
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9624864033843433,
+      "learning_rate": 6.208380251417122e-06,
+      "loss": 0.5534,
+      "step": 6869
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1462973372785308,
+      "learning_rate": 6.207374462877953e-06,
+      "loss": 0.57,
+      "step": 6870
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8593915504766811,
+      "learning_rate": 6.2063686224552144e-06,
+      "loss": 0.7726,
+      "step": 6871
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4244183098434118,
+      "learning_rate": 6.205362730192131e-06,
+      "loss": 0.7977,
+      "step": 6872
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4393304741578836,
+      "learning_rate": 6.2043567861319245e-06,
+      "loss": 0.7017,
+      "step": 6873
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3190389422471025,
+      "learning_rate": 6.203350790317825e-06,
+      "loss": 0.6531,
+      "step": 6874
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4163998885099411,
+      "learning_rate": 6.2023447427930615e-06,
+      "loss": 0.689,
+      "step": 6875
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4242741865102708,
+      "learning_rate": 6.201338643600869e-06,
+      "loss": 0.6676,
+      "step": 6876
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3613539957821303,
+      "learning_rate": 6.2003324927844765e-06,
+      "loss": 0.6954,
+      "step": 6877
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1244084489156643,
+      "learning_rate": 6.199326290387126e-06,
+      "loss": 0.6073,
+      "step": 6878
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6580080951249918,
+      "learning_rate": 6.198320036452051e-06,
+      "loss": 0.7332,
+      "step": 6879
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.181815782494273,
+      "learning_rate": 6.197313731022497e-06,
+      "loss": 0.6609,
+      "step": 6880
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5808913908012587,
+      "learning_rate": 6.196307374141707e-06,
+      "loss": 0.7078,
+      "step": 6881
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5430905466386013,
+      "learning_rate": 6.195300965852923e-06,
+      "loss": 0.6519,
+      "step": 6882
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9583223578321983,
+      "learning_rate": 6.194294506199394e-06,
+      "loss": 0.6296,
+      "step": 6883
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3425075964210078,
+      "learning_rate": 6.193287995224371e-06,
+      "loss": 0.6343,
+      "step": 6884
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3674936981670414,
+      "learning_rate": 6.192281432971104e-06,
+      "loss": 0.6898,
+      "step": 6885
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3493033948313926,
+      "learning_rate": 6.19127481948285e-06,
+      "loss": 0.6618,
+      "step": 6886
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3706563819241877,
+      "learning_rate": 6.1902681548028646e-06,
+      "loss": 0.6452,
+      "step": 6887
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4297628658999697,
+      "learning_rate": 6.189261438974403e-06,
+      "loss": 0.6388,
+      "step": 6888
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.243566352386944,
+      "learning_rate": 6.18825467204073e-06,
+      "loss": 0.6591,
+      "step": 6889
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2153533023286187,
+      "learning_rate": 6.187247854045109e-06,
+      "loss": 0.5531,
+      "step": 6890
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0591081156706637,
+      "learning_rate": 6.186240985030802e-06,
+      "loss": 0.5982,
+      "step": 6891
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1342641653637806,
+      "learning_rate": 6.1852340650410784e-06,
+      "loss": 0.6532,
+      "step": 6892
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3561337633173833,
+      "learning_rate": 6.184227094119207e-06,
+      "loss": 0.6738,
+      "step": 6893
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0354678815069602,
+      "learning_rate": 6.183220072308459e-06,
+      "loss": 0.5871,
+      "step": 6894
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2762247219581846,
+      "learning_rate": 6.1822129996521105e-06,
+      "loss": 0.6453,
+      "step": 6895
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.416571590040315,
+      "learning_rate": 6.181205876193437e-06,
+      "loss": 0.6706,
+      "step": 6896
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4113016362368898,
+      "learning_rate": 6.180198701975715e-06,
+      "loss": 0.6189,
+      "step": 6897
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3060555208310982,
+      "learning_rate": 6.179191477042227e-06,
+      "loss": 0.6543,
+      "step": 6898
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7446650996491109,
+      "learning_rate": 6.178184201436256e-06,
+      "loss": 0.7674,
+      "step": 6899
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3006555158363504,
+      "learning_rate": 6.177176875201086e-06,
+      "loss": 0.6395,
+      "step": 6900
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3520397421612085,
+      "learning_rate": 6.176169498380001e-06,
+      "loss": 0.6589,
+      "step": 6901
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.301080396154864,
+      "learning_rate": 6.175162071016295e-06,
+      "loss": 0.621,
+      "step": 6902
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3041328473671003,
+      "learning_rate": 6.174154593153258e-06,
+      "loss": 0.6795,
+      "step": 6903
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3292455211245053,
+      "learning_rate": 6.173147064834183e-06,
+      "loss": 0.6392,
+      "step": 6904
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.291743686786193,
+      "learning_rate": 6.1721394861023665e-06,
+      "loss": 0.6855,
+      "step": 6905
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3503539936234352,
+      "learning_rate": 6.171131857001104e-06,
+      "loss": 0.6055,
+      "step": 6906
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0918895261501436,
+      "learning_rate": 6.170124177573698e-06,
+      "loss": 0.6085,
+      "step": 6907
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1698035642622162,
+      "learning_rate": 6.169116447863451e-06,
+      "loss": 0.6407,
+      "step": 6908
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2275873127146941,
+      "learning_rate": 6.168108667913666e-06,
+      "loss": 0.6467,
+      "step": 6909
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1750875453463177,
+      "learning_rate": 6.16710083776765e-06,
+      "loss": 0.6398,
+      "step": 6910
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1067953902372984,
+      "learning_rate": 6.1660929574687115e-06,
+      "loss": 0.5936,
+      "step": 6911
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4358337592275696,
+      "learning_rate": 6.165085027060162e-06,
+      "loss": 0.7067,
+      "step": 6912
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4019522168856968,
+      "learning_rate": 6.164077046585314e-06,
+      "loss": 0.6423,
+      "step": 6913
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4658076675496272,
+      "learning_rate": 6.163069016087483e-06,
+      "loss": 0.7444,
+      "step": 6914
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3286179948896337,
+      "learning_rate": 6.162060935609985e-06,
+      "loss": 0.6292,
+      "step": 6915
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6343708557750667,
+      "learning_rate": 6.161052805196141e-06,
+      "loss": 0.6336,
+      "step": 6916
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.377573857701949,
+      "learning_rate": 6.160044624889272e-06,
+      "loss": 0.5573,
+      "step": 6917
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5887674114410522,
+      "learning_rate": 6.1590363947327025e-06,
+      "loss": 0.7085,
+      "step": 6918
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7715511806139108,
+      "learning_rate": 6.158028114769758e-06,
+      "loss": 0.6242,
+      "step": 6919
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6190330895882357,
+      "learning_rate": 6.157019785043765e-06,
+      "loss": 0.7932,
+      "step": 6920
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3344428073031063,
+      "learning_rate": 6.156011405598057e-06,
+      "loss": 0.8013,
+      "step": 6921
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4295174153870227,
+      "learning_rate": 6.155002976475962e-06,
+      "loss": 0.634,
+      "step": 6922
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2249557358329461,
+      "learning_rate": 6.15399449772082e-06,
+      "loss": 0.6589,
+      "step": 6923
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4034266288506068,
+      "learning_rate": 6.152985969375962e-06,
+      "loss": 0.6813,
+      "step": 6924
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6596915855938998,
+      "learning_rate": 6.151977391484729e-06,
+      "loss": 0.5977,
+      "step": 6925
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4682350025400415,
+      "learning_rate": 6.150968764090461e-06,
+      "loss": 0.7314,
+      "step": 6926
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.270884681852607,
+      "learning_rate": 6.149960087236505e-06,
+      "loss": 0.602,
+      "step": 6927
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3642677006602428,
+      "learning_rate": 6.1489513609662025e-06,
+      "loss": 0.6624,
+      "step": 6928
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3378657977498871,
+      "learning_rate": 6.147942585322898e-06,
+      "loss": 0.6495,
+      "step": 6929
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.227266128543334,
+      "learning_rate": 6.146933760349947e-06,
+      "loss": 0.7119,
+      "step": 6930
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4167340079921555,
+      "learning_rate": 6.145924886090695e-06,
+      "loss": 0.542,
+      "step": 6931
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3407720574636584,
+      "learning_rate": 6.144915962588503e-06,
+      "loss": 0.6437,
+      "step": 6932
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3070589526372316,
+      "learning_rate": 6.143906989886718e-06,
+      "loss": 0.6407,
+      "step": 6933
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1722501341382547,
+      "learning_rate": 6.142897968028704e-06,
+      "loss": 0.6196,
+      "step": 6934
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1734865307745297,
+      "learning_rate": 6.141888897057817e-06,
+      "loss": 0.5355,
+      "step": 6935
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.41695862304438,
+      "learning_rate": 6.140879777017423e-06,
+      "loss": 0.6424,
+      "step": 6936
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2441355151065638,
+      "learning_rate": 6.139870607950885e-06,
+      "loss": 0.5483,
+      "step": 6937
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4067386031704905,
+      "learning_rate": 6.138861389901566e-06,
+      "loss": 0.6766,
+      "step": 6938
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.385440849062308,
+      "learning_rate": 6.137852122912839e-06,
+      "loss": 0.7906,
+      "step": 6939
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.416499078100972,
+      "learning_rate": 6.136842807028069e-06,
+      "loss": 0.7168,
+      "step": 6940
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.144023369033625,
+      "learning_rate": 6.135833442290634e-06,
+      "loss": 0.6213,
+      "step": 6941
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4487685767986391,
+      "learning_rate": 6.134824028743906e-06,
+      "loss": 0.6529,
+      "step": 6942
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3697394704838424,
+      "learning_rate": 6.133814566431262e-06,
+      "loss": 0.6364,
+      "step": 6943
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.762376379621403,
+      "learning_rate": 6.1328050553960804e-06,
+      "loss": 0.6688,
+      "step": 6944
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3315858707530028,
+      "learning_rate": 6.131795495681743e-06,
+      "loss": 0.6204,
+      "step": 6945
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4953799062700242,
+      "learning_rate": 6.130785887331633e-06,
+      "loss": 0.6409,
+      "step": 6946
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4656186280309886,
+      "learning_rate": 6.129776230389135e-06,
+      "loss": 0.7029,
+      "step": 6947
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5420388619194516,
+      "learning_rate": 6.128766524897636e-06,
+      "loss": 0.6638,
+      "step": 6948
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0017723453995586,
+      "learning_rate": 6.1277567709005245e-06,
+      "loss": 0.6299,
+      "step": 6949
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6461247684129103,
+      "learning_rate": 6.126746968441195e-06,
+      "loss": 0.7412,
+      "step": 6950
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.308189044473393,
+      "learning_rate": 6.1257371175630375e-06,
+      "loss": 0.6611,
+      "step": 6951
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6348006425259598,
+      "learning_rate": 6.124727218309449e-06,
+      "loss": 0.5315,
+      "step": 6952
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2797580315251877,
+      "learning_rate": 6.123717270723827e-06,
+      "loss": 0.6742,
+      "step": 6953
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.620281482026334,
+      "learning_rate": 6.122707274849572e-06,
+      "loss": 0.6987,
+      "step": 6954
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2021043321954463,
+      "learning_rate": 6.121697230730084e-06,
+      "loss": 0.6344,
+      "step": 6955
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2056958632058665,
+      "learning_rate": 6.120687138408768e-06,
+      "loss": 0.5773,
+      "step": 6956
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3572067817625721,
+      "learning_rate": 6.11967699792903e-06,
+      "loss": 0.6809,
+      "step": 6957
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0581880374810595,
+      "learning_rate": 6.118666809334277e-06,
+      "loss": 0.6435,
+      "step": 6958
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4016005155490219,
+      "learning_rate": 6.117656572667921e-06,
+      "loss": 0.6675,
+      "step": 6959
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3464304891462733,
+      "learning_rate": 6.116646287973374e-06,
+      "loss": 0.6457,
+      "step": 6960
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3722025906810549,
+      "learning_rate": 6.115635955294047e-06,
+      "loss": 0.6713,
+      "step": 6961
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5745556859891185,
+      "learning_rate": 6.114625574673359e-06,
+      "loss": 0.6732,
+      "step": 6962
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4926779552036527,
+      "learning_rate": 6.1136151461547266e-06,
+      "loss": 0.7606,
+      "step": 6963
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.364078636965625,
+      "learning_rate": 6.112604669781572e-06,
+      "loss": 0.648,
+      "step": 6964
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4302172879692945,
+      "learning_rate": 6.111594145597319e-06,
+      "loss": 0.5034,
+      "step": 6965
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4440235182003747,
+      "learning_rate": 6.1105835736453874e-06,
+      "loss": 0.7947,
+      "step": 6966
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6216899795332125,
+      "learning_rate": 6.109572953969206e-06,
+      "loss": 0.7147,
+      "step": 6967
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1784507155201775,
+      "learning_rate": 6.108562286612204e-06,
+      "loss": 0.5582,
+      "step": 6968
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5740761732187922,
+      "learning_rate": 6.107551571617813e-06,
+      "loss": 0.6996,
+      "step": 6969
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4757429259412094,
+      "learning_rate": 6.106540809029463e-06,
+      "loss": 0.7582,
+      "step": 6970
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.920407508119366,
+      "learning_rate": 6.105529998890589e-06,
+      "loss": 0.5893,
+      "step": 6971
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1678165535012883,
+      "learning_rate": 6.104519141244631e-06,
+      "loss": 0.5606,
+      "step": 6972
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4772799836143704,
+      "learning_rate": 6.103508236135025e-06,
+      "loss": 0.6411,
+      "step": 6973
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7966920193297864,
+      "learning_rate": 6.1024972836052135e-06,
+      "loss": 0.6547,
+      "step": 6974
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.227604760292804,
+      "learning_rate": 6.1014862836986364e-06,
+      "loss": 0.7354,
+      "step": 6975
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2083476961283943,
+      "learning_rate": 6.10047523645874e-06,
+      "loss": 0.6596,
+      "step": 6976
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5786647139162069,
+      "learning_rate": 6.099464141928973e-06,
+      "loss": 0.6281,
+      "step": 6977
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5268315756684125,
+      "learning_rate": 6.098453000152784e-06,
+      "loss": 0.6808,
+      "step": 6978
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3501898883985102,
+      "learning_rate": 6.0974418111736235e-06,
+      "loss": 0.6673,
+      "step": 6979
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1642146032127085,
+      "learning_rate": 6.096430575034942e-06,
+      "loss": 0.7011,
+      "step": 6980
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2717595181838912,
+      "learning_rate": 6.095419291780198e-06,
+      "loss": 0.7317,
+      "step": 6981
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.20569734327623,
+      "learning_rate": 6.0944079614528475e-06,
+      "loss": 0.6618,
+      "step": 6982
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4222571816595728,
+      "learning_rate": 6.0933965840963494e-06,
+      "loss": 0.6417,
+      "step": 6983
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5092794027991263,
+      "learning_rate": 6.092385159754165e-06,
+      "loss": 0.6852,
+      "step": 6984
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3301033044248591,
+      "learning_rate": 6.091373688469757e-06,
+      "loss": 0.6989,
+      "step": 6985
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7553563952779938,
+      "learning_rate": 6.090362170286591e-06,
+      "loss": 0.7445,
+      "step": 6986
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.194031486757901,
+      "learning_rate": 6.089350605248134e-06,
+      "loss": 0.5694,
+      "step": 6987
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4475061483382337,
+      "learning_rate": 6.088338993397857e-06,
+      "loss": 0.685,
+      "step": 6988
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3915384237240331,
+      "learning_rate": 6.0873273347792275e-06,
+      "loss": 0.6312,
+      "step": 6989
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.365172630087313,
+      "learning_rate": 6.086315629435722e-06,
+      "loss": 0.6803,
+      "step": 6990
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.655979956113953,
+      "learning_rate": 6.085303877410816e-06,
+      "loss": 0.7727,
+      "step": 6991
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3920370455938935,
+      "learning_rate": 6.084292078747982e-06,
+      "loss": 0.7037,
+      "step": 6992
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4464193141229085,
+      "learning_rate": 6.0832802334907044e-06,
+      "loss": 0.5868,
+      "step": 6993
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4044728590410738,
+      "learning_rate": 6.0822683416824625e-06,
+      "loss": 0.6156,
+      "step": 6994
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3673476275897596,
+      "learning_rate": 6.08125640336674e-06,
+      "loss": 0.7596,
+      "step": 6995
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.508209154825498,
+      "learning_rate": 6.08024441858702e-06,
+      "loss": 0.7139,
+      "step": 6996
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5317814119345712,
+      "learning_rate": 6.079232387386793e-06,
+      "loss": 0.7688,
+      "step": 6997
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7329642835832373,
+      "learning_rate": 6.078220309809546e-06,
+      "loss": 0.73,
+      "step": 6998
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4262452555906313,
+      "learning_rate": 6.077208185898772e-06,
+      "loss": 0.6393,
+      "step": 6999
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1660694835115901,
+      "learning_rate": 6.076196015697963e-06,
+      "loss": 0.4843,
+      "step": 7000
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2539718190323728,
+      "learning_rate": 6.0751837992506126e-06,
+      "loss": 0.6176,
+      "step": 7001
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4766400415563778,
+      "learning_rate": 6.074171536600222e-06,
+      "loss": 0.6835,
+      "step": 7002
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.157458667798011,
+      "learning_rate": 6.073159227790287e-06,
+      "loss": 0.7126,
+      "step": 7003
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3958183678594083,
+      "learning_rate": 6.07214687286431e-06,
+      "loss": 0.6351,
+      "step": 7004
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5840429806485412,
+      "learning_rate": 6.0711344718657934e-06,
+      "loss": 0.7493,
+      "step": 7005
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4434642147999204,
+      "learning_rate": 6.070122024838245e-06,
+      "loss": 0.6732,
+      "step": 7006
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2934297222534865,
+      "learning_rate": 6.069109531825169e-06,
+      "loss": 0.6729,
+      "step": 7007
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3460199364727858,
+      "learning_rate": 6.068096992870075e-06,
+      "loss": 0.6809,
+      "step": 7008
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.206198385602857,
+      "learning_rate": 6.067084408016475e-06,
+      "loss": 0.5391,
+      "step": 7009
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1736802651465674,
+      "learning_rate": 6.06607177730788e-06,
+      "loss": 0.6126,
+      "step": 7010
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.08349686913667,
+      "learning_rate": 6.065059100787808e-06,
+      "loss": 0.5771,
+      "step": 7011
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4272721711859406,
+      "learning_rate": 6.064046378499775e-06,
+      "loss": 0.7839,
+      "step": 7012
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.527319236130746,
+      "learning_rate": 6.063033610487298e-06,
+      "loss": 0.612,
+      "step": 7013
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.350483229855132,
+      "learning_rate": 6.0620207967939e-06,
+      "loss": 0.6098,
+      "step": 7014
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3926755897772862,
+      "learning_rate": 6.061007937463101e-06,
+      "loss": 0.625,
+      "step": 7015
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3217948097132644,
+      "learning_rate": 6.05999503253843e-06,
+      "loss": 0.652,
+      "step": 7016
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3902314157838584,
+      "learning_rate": 6.058982082063413e-06,
+      "loss": 0.6692,
+      "step": 7017
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1118693022590473,
+      "learning_rate": 6.057969086081575e-06,
+      "loss": 0.6072,
+      "step": 7018
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3487115461663228,
+      "learning_rate": 6.0569560446364495e-06,
+      "loss": 0.6497,
+      "step": 7019
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.7659769421627076,
+      "learning_rate": 6.05594295777157e-06,
+      "loss": 0.774,
+      "step": 7020
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4696688946452974,
+      "learning_rate": 6.054929825530469e-06,
+      "loss": 0.6974,
+      "step": 7021
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3659739171587446,
+      "learning_rate": 6.0539166479566844e-06,
+      "loss": 0.7659,
+      "step": 7022
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6205457019639247,
+      "learning_rate": 6.052903425093752e-06,
+      "loss": 0.702,
+      "step": 7023
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4842940736281283,
+      "learning_rate": 6.051890156985217e-06,
+      "loss": 0.6225,
+      "step": 7024
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0918667875470927,
+      "learning_rate": 6.050876843674619e-06,
+      "loss": 0.6198,
+      "step": 7025
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.343256120976173,
+      "learning_rate": 6.049863485205503e-06,
+      "loss": 0.7322,
+      "step": 7026
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3595248986007975,
+      "learning_rate": 6.048850081621415e-06,
+      "loss": 0.6169,
+      "step": 7027
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2628959124426673,
+      "learning_rate": 6.047836632965901e-06,
+      "loss": 0.6998,
+      "step": 7028
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4593489162144457,
+      "learning_rate": 6.046823139282515e-06,
+      "loss": 0.6584,
+      "step": 7029
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2969638912777655,
+      "learning_rate": 6.045809600614807e-06,
+      "loss": 0.7312,
+      "step": 7030
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4019061639806039,
+      "learning_rate": 6.044796017006331e-06,
+      "loss": 0.6712,
+      "step": 7031
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1983434084312452,
+      "learning_rate": 6.043782388500642e-06,
+      "loss": 0.5986,
+      "step": 7032
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2949717407768104,
+      "learning_rate": 6.0427687151413004e-06,
+      "loss": 0.6664,
+      "step": 7033
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2074867365613442,
+      "learning_rate": 6.041754996971866e-06,
+      "loss": 0.5894,
+      "step": 7034
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6449789720668901,
+      "learning_rate": 6.040741234035898e-06,
+      "loss": 0.6415,
+      "step": 7035
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2785047420567588,
+      "learning_rate": 6.039727426376961e-06,
+      "loss": 0.6936,
+      "step": 7036
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9622859523889564,
+      "learning_rate": 6.038713574038622e-06,
+      "loss": 0.6201,
+      "step": 7037
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.283908104929152,
+      "learning_rate": 6.037699677064446e-06,
+      "loss": 0.6396,
+      "step": 7038
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3927115868387885,
+      "learning_rate": 6.036685735498004e-06,
+      "loss": 0.669,
+      "step": 7039
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4028633777088422,
+      "learning_rate": 6.035671749382869e-06,
+      "loss": 0.6585,
+      "step": 7040
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5120456282392796,
+      "learning_rate": 6.03465771876261e-06,
+      "loss": 0.7452,
+      "step": 7041
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1600624214251272,
+      "learning_rate": 6.0336436436808054e-06,
+      "loss": 0.5964,
+      "step": 7042
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3426648297424864,
+      "learning_rate": 6.032629524181031e-06,
+      "loss": 0.5987,
+      "step": 7043
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.295108611351601,
+      "learning_rate": 6.031615360306867e-06,
+      "loss": 0.6551,
+      "step": 7044
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5113579053901236,
+      "learning_rate": 6.030601152101892e-06,
+      "loss": 0.6076,
+      "step": 7045
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.49215376357375,
+      "learning_rate": 6.029586899609692e-06,
+      "loss": 0.7935,
+      "step": 7046
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3697305212557767,
+      "learning_rate": 6.02857260287385e-06,
+      "loss": 0.5698,
+      "step": 7047
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.551237094103736,
+      "learning_rate": 6.02755826193795e-06,
+      "loss": 0.5651,
+      "step": 7048
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5083882838217135,
+      "learning_rate": 6.026543876845586e-06,
+      "loss": 0.6795,
+      "step": 7049
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2567692103460215,
+      "learning_rate": 6.025529447640343e-06,
+      "loss": 0.6297,
+      "step": 7050
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5068894487620863,
+      "learning_rate": 6.024514974365818e-06,
+      "loss": 0.7189,
+      "step": 7051
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3542783147781017,
+      "learning_rate": 6.0235004570656005e-06,
+      "loss": 0.626,
+      "step": 7052
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3222214449027747,
+      "learning_rate": 6.02248589578329e-06,
+      "loss": 0.6535,
+      "step": 7053
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5069605213310437,
+      "learning_rate": 6.021471290562484e-06,
+      "loss": 0.7432,
+      "step": 7054
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9773005533390118,
+      "learning_rate": 6.020456641446781e-06,
+      "loss": 0.7505,
+      "step": 7055
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4243303118030133,
+      "learning_rate": 6.019441948479784e-06,
+      "loss": 0.7515,
+      "step": 7056
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.37375974903803,
+      "learning_rate": 6.018427211705094e-06,
+      "loss": 0.6203,
+      "step": 7057
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.28195501050273,
+      "learning_rate": 6.017412431166322e-06,
+      "loss": 0.7051,
+      "step": 7058
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1876183194860686,
+      "learning_rate": 6.016397606907069e-06,
+      "loss": 0.6074,
+      "step": 7059
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.296218626085181,
+      "learning_rate": 6.015382738970949e-06,
+      "loss": 0.5905,
+      "step": 7060
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.16517250538346,
+      "learning_rate": 6.01436782740157e-06,
+      "loss": 0.6426,
+      "step": 7061
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.415686489184929,
+      "learning_rate": 6.013352872242546e-06,
+      "loss": 0.7518,
+      "step": 7062
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2419694423273582,
+      "learning_rate": 6.012337873537494e-06,
+      "loss": 0.5629,
+      "step": 7063
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3183879346803031,
+      "learning_rate": 6.011322831330028e-06,
+      "loss": 0.6625,
+      "step": 7064
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3346294885053887,
+      "learning_rate": 6.010307745663768e-06,
+      "loss": 0.641,
+      "step": 7065
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1945645241915634,
+      "learning_rate": 6.009292616582333e-06,
+      "loss": 0.702,
+      "step": 7066
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3761670648352071,
+      "learning_rate": 6.008277444129347e-06,
+      "loss": 0.5899,
+      "step": 7067
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1957100271977266,
+      "learning_rate": 6.007262228348434e-06,
+      "loss": 0.6446,
+      "step": 7068
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4694336895922728,
+      "learning_rate": 6.0062469692832205e-06,
+      "loss": 0.7045,
+      "step": 7069
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1221961444050639,
+      "learning_rate": 6.005231666977331e-06,
+      "loss": 0.5952,
+      "step": 7070
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2406167789141604,
+      "learning_rate": 6.0042163214744006e-06,
+      "loss": 0.6699,
+      "step": 7071
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3950059090374445,
+      "learning_rate": 6.003200932818058e-06,
+      "loss": 0.7003,
+      "step": 7072
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.479904430328958,
+      "learning_rate": 6.002185501051937e-06,
+      "loss": 0.6033,
+      "step": 7073
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.187243723816199,
+      "learning_rate": 6.001170026219673e-06,
+      "loss": 0.7506,
+      "step": 7074
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1272983813774202,
+      "learning_rate": 6.000154508364901e-06,
+      "loss": 0.6219,
+      "step": 7075
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.463605009110526,
+      "learning_rate": 5.999138947531265e-06,
+      "loss": 0.6655,
+      "step": 7076
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2653859835886754,
+      "learning_rate": 5.998123343762403e-06,
+      "loss": 0.5949,
+      "step": 7077
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3504549786729032,
+      "learning_rate": 5.997107697101957e-06,
+      "loss": 0.6582,
+      "step": 7078
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7079717837257682,
+      "learning_rate": 5.996092007593572e-06,
+      "loss": 0.668,
+      "step": 7079
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2214210761410038,
+      "learning_rate": 5.9950762752808965e-06,
+      "loss": 0.7255,
+      "step": 7080
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.374735301942726,
+      "learning_rate": 5.994060500207577e-06,
+      "loss": 0.6544,
+      "step": 7081
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1599210701187883,
+      "learning_rate": 5.993044682417264e-06,
+      "loss": 0.7839,
+      "step": 7082
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2435934719007824,
+      "learning_rate": 5.992028821953609e-06,
+      "loss": 0.6235,
+      "step": 7083
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4020508213096465,
+      "learning_rate": 5.9910129188602665e-06,
+      "loss": 0.6804,
+      "step": 7084
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1230678297224395,
+      "learning_rate": 5.989996973180892e-06,
+      "loss": 0.575,
+      "step": 7085
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.070084837772088,
+      "learning_rate": 5.988980984959142e-06,
+      "loss": 0.6626,
+      "step": 7086
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9582777246919292,
+      "learning_rate": 5.987964954238677e-06,
+      "loss": 0.5502,
+      "step": 7087
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1429518600926036,
+      "learning_rate": 5.986948881063157e-06,
+      "loss": 0.7031,
+      "step": 7088
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2523563925266277,
+      "learning_rate": 5.985932765476246e-06,
+      "loss": 0.5845,
+      "step": 7089
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3338414475444205,
+      "learning_rate": 5.984916607521608e-06,
+      "loss": 0.6395,
+      "step": 7090
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3929961360090937,
+      "learning_rate": 5.983900407242911e-06,
+      "loss": 0.7522,
+      "step": 7091
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.318459252156488,
+      "learning_rate": 5.982884164683821e-06,
+      "loss": 0.5613,
+      "step": 7092
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1777548038420542,
+      "learning_rate": 5.9818678798880085e-06,
+      "loss": 0.5845,
+      "step": 7093
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.194791110844647,
+      "learning_rate": 5.9808515528991486e-06,
+      "loss": 0.591,
+      "step": 7094
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5653439309105217,
+      "learning_rate": 5.979835183760911e-06,
+      "loss": 0.6534,
+      "step": 7095
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.31328911456368,
+      "learning_rate": 5.978818772516976e-06,
+      "loss": 0.6254,
+      "step": 7096
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.260693913415879,
+      "learning_rate": 5.977802319211016e-06,
+      "loss": 0.6484,
+      "step": 7097
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4948685321335993,
+      "learning_rate": 5.976785823886713e-06,
+      "loss": 0.6342,
+      "step": 7098
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4528575327473947,
+      "learning_rate": 5.975769286587747e-06,
+      "loss": 0.643,
+      "step": 7099
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2148679124175374,
+      "learning_rate": 5.974752707357805e-06,
+      "loss": 0.656,
+      "step": 7100
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3204992044897956,
+      "learning_rate": 5.973736086240566e-06,
+      "loss": 0.5813,
+      "step": 7101
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3524138801241252,
+      "learning_rate": 5.9727194232797185e-06,
+      "loss": 0.7256,
+      "step": 7102
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5526416465576058,
+      "learning_rate": 5.97170271851895e-06,
+      "loss": 0.6899,
+      "step": 7103
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.097745712703655,
+      "learning_rate": 5.970685972001953e-06,
+      "loss": 0.6659,
+      "step": 7104
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.374786295866741,
+      "learning_rate": 5.96966918377242e-06,
+      "loss": 0.6377,
+      "step": 7105
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9065593384079198,
+      "learning_rate": 5.9686523538740385e-06,
+      "loss": 0.6836,
+      "step": 7106
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1137873550059099,
+      "learning_rate": 5.96763548235051e-06,
+      "loss": 0.6284,
+      "step": 7107
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3114012882851793,
+      "learning_rate": 5.966618569245529e-06,
+      "loss": 0.64,
+      "step": 7108
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.00626469840752,
+      "learning_rate": 5.965601614602798e-06,
+      "loss": 0.738,
+      "step": 7109
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.427746253494117,
+      "learning_rate": 5.964584618466014e-06,
+      "loss": 0.7101,
+      "step": 7110
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3655085376807747,
+      "learning_rate": 5.9635675808788785e-06,
+      "loss": 0.7215,
+      "step": 7111
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2086158619719198,
+      "learning_rate": 5.9625505018851e-06,
+      "loss": 0.6978,
+      "step": 7112
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3659729395181752,
+      "learning_rate": 5.961533381528382e-06,
+      "loss": 0.7229,
+      "step": 7113
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2521033035436753,
+      "learning_rate": 5.960516219852433e-06,
+      "loss": 0.6334,
+      "step": 7114
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3874904925959235,
+      "learning_rate": 5.959499016900963e-06,
+      "loss": 0.7243,
+      "step": 7115
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.144026636756326,
+      "learning_rate": 5.9584817727176836e-06,
+      "loss": 0.6859,
+      "step": 7116
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3235137612921497,
+      "learning_rate": 5.9574644873463075e-06,
+      "loss": 0.6199,
+      "step": 7117
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3687764055216356,
+      "learning_rate": 5.95644716083055e-06,
+      "loss": 0.7152,
+      "step": 7118
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4158610995245502,
+      "learning_rate": 5.955429793214129e-06,
+      "loss": 0.5554,
+      "step": 7119
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3781104341418577,
+      "learning_rate": 5.954412384540761e-06,
+      "loss": 0.5851,
+      "step": 7120
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4517178700448459,
+      "learning_rate": 5.953394934854167e-06,
+      "loss": 0.7175,
+      "step": 7121
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4253585864579001,
+      "learning_rate": 5.952377444198068e-06,
+      "loss": 0.6804,
+      "step": 7122
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.291626154389515,
+      "learning_rate": 5.951359912616192e-06,
+      "loss": 0.6404,
+      "step": 7123
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5356933235625052,
+      "learning_rate": 5.950342340152261e-06,
+      "loss": 0.6164,
+      "step": 7124
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3408875949082606,
+      "learning_rate": 5.949324726850002e-06,
+      "loss": 0.7347,
+      "step": 7125
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.38729193322036,
+      "learning_rate": 5.948307072753146e-06,
+      "loss": 0.6186,
+      "step": 7126
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7178985391512442,
+      "learning_rate": 5.947289377905422e-06,
+      "loss": 0.8937,
+      "step": 7127
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5118458557380712,
+      "learning_rate": 5.946271642350565e-06,
+      "loss": 0.6576,
+      "step": 7128
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3436844262066046,
+      "learning_rate": 5.945253866132308e-06,
+      "loss": 0.7229,
+      "step": 7129
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4271425294423115,
+      "learning_rate": 5.9442360492943876e-06,
+      "loss": 0.6768,
+      "step": 7130
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3750766000671812,
+      "learning_rate": 5.94321819188054e-06,
+      "loss": 0.7091,
+      "step": 7131
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1856258821952963,
+      "learning_rate": 5.942200293934506e-06,
+      "loss": 0.6197,
+      "step": 7132
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5503519475811693,
+      "learning_rate": 5.941182355500028e-06,
+      "loss": 0.6057,
+      "step": 7133
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3168761179955648,
+      "learning_rate": 5.940164376620847e-06,
+      "loss": 0.534,
+      "step": 7134
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5363936830973823,
+      "learning_rate": 5.93914635734071e-06,
+      "loss": 0.76,
+      "step": 7135
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5962515125171128,
+      "learning_rate": 5.938128297703361e-06,
+      "loss": 0.6642,
+      "step": 7136
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2207481917429093,
+      "learning_rate": 5.937110197752551e-06,
+      "loss": 0.6307,
+      "step": 7137
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1614500258293627,
+      "learning_rate": 5.936092057532029e-06,
+      "loss": 0.6471,
+      "step": 7138
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.385273251789624,
+      "learning_rate": 5.935073877085546e-06,
+      "loss": 0.6814,
+      "step": 7139
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4620294472586965,
+      "learning_rate": 5.934055656456855e-06,
+      "loss": 0.6175,
+      "step": 7140
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2982653586257276,
+      "learning_rate": 5.9330373956897135e-06,
+      "loss": 0.6636,
+      "step": 7141
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5206650216276327,
+      "learning_rate": 5.932019094827877e-06,
+      "loss": 0.6753,
+      "step": 7142
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.475908048511423,
+      "learning_rate": 5.931000753915104e-06,
+      "loss": 0.7025,
+      "step": 7143
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5565092962058291,
+      "learning_rate": 5.9299823729951544e-06,
+      "loss": 0.6925,
+      "step": 7144
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2414677668356702,
+      "learning_rate": 5.928963952111791e-06,
+      "loss": 0.6129,
+      "step": 7145
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2467171190780901,
+      "learning_rate": 5.9279454913087795e-06,
+      "loss": 0.6614,
+      "step": 7146
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.656473690757349,
+      "learning_rate": 5.926926990629883e-06,
+      "loss": 0.551,
+      "step": 7147
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1634942598380251,
+      "learning_rate": 5.9259084501188695e-06,
+      "loss": 0.5746,
+      "step": 7148
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5672446473776438,
+      "learning_rate": 5.9248898698195054e-06,
+      "loss": 0.7306,
+      "step": 7149
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.542695464603821,
+      "learning_rate": 5.923871249775566e-06,
+      "loss": 0.7644,
+      "step": 7150
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4172499334479387,
+      "learning_rate": 5.922852590030821e-06,
+      "loss": 0.7594,
+      "step": 7151
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3230922037590307,
+      "learning_rate": 5.921833890629045e-06,
+      "loss": 0.6623,
+      "step": 7152
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2324049490492666,
+      "learning_rate": 5.920815151614012e-06,
+      "loss": 0.5993,
+      "step": 7153
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4803915355342379,
+      "learning_rate": 5.919796373029504e-06,
+      "loss": 0.7196,
+      "step": 7154
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2100820776737198,
+      "learning_rate": 5.918777554919296e-06,
+      "loss": 0.709,
+      "step": 7155
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1023966678398867,
+      "learning_rate": 5.917758697327171e-06,
+      "loss": 0.684,
+      "step": 7156
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3560909501758684,
+      "learning_rate": 5.9167398002969104e-06,
+      "loss": 0.6819,
+      "step": 7157
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2536592472854882,
+      "learning_rate": 5.915720863872297e-06,
+      "loss": 0.6149,
+      "step": 7158
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.179204554652368,
+      "learning_rate": 5.914701888097121e-06,
+      "loss": 0.6719,
+      "step": 7159
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2427305911667654,
+      "learning_rate": 5.913682873015166e-06,
+      "loss": 0.6133,
+      "step": 7160
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.382019826558232,
+      "learning_rate": 5.912663818670224e-06,
+      "loss": 0.6073,
+      "step": 7161
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4736084018489959,
+      "learning_rate": 5.911644725106085e-06,
+      "loss": 0.6064,
+      "step": 7162
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3995587171233739,
+      "learning_rate": 5.910625592366542e-06,
+      "loss": 0.728,
+      "step": 7163
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3853104986104623,
+      "learning_rate": 5.90960642049539e-06,
+      "loss": 0.7232,
+      "step": 7164
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4772263893758864,
+      "learning_rate": 5.9085872095364236e-06,
+      "loss": 0.6717,
+      "step": 7165
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3977686593680432,
+      "learning_rate": 5.907567959533441e-06,
+      "loss": 0.5901,
+      "step": 7166
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2847690057916579,
+      "learning_rate": 5.906548670530241e-06,
+      "loss": 0.6097,
+      "step": 7167
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6426075501650315,
+      "learning_rate": 5.905529342570627e-06,
+      "loss": 0.7063,
+      "step": 7168
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3119525463147135,
+      "learning_rate": 5.904509975698399e-06,
+      "loss": 0.638,
+      "step": 7169
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2864399232547943,
+      "learning_rate": 5.903490569957366e-06,
+      "loss": 0.627,
+      "step": 7170
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3390900903125544,
+      "learning_rate": 5.902471125391328e-06,
+      "loss": 0.645,
+      "step": 7171
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0227922321352678,
+      "learning_rate": 5.901451642044097e-06,
+      "loss": 0.722,
+      "step": 7172
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4623896411383095,
+      "learning_rate": 5.90043211995948e-06,
+      "loss": 0.6357,
+      "step": 7173
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5336213900679851,
+      "learning_rate": 5.8994125591812914e-06,
+      "loss": 1.0025,
+      "step": 7174
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5357276609684356,
+      "learning_rate": 5.898392959753343e-06,
+      "loss": 0.7224,
+      "step": 7175
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.455537152496798,
+      "learning_rate": 5.897373321719445e-06,
+      "loss": 0.5897,
+      "step": 7176
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7698545393853766,
+      "learning_rate": 5.896353645123418e-06,
+      "loss": 0.7939,
+      "step": 7177
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.59085203516472,
+      "learning_rate": 5.895333930009079e-06,
+      "loss": 0.7354,
+      "step": 7178
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3386004531539477,
+      "learning_rate": 5.894314176420247e-06,
+      "loss": 0.6256,
+      "step": 7179
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2866124647587278,
+      "learning_rate": 5.893294384400743e-06,
+      "loss": 0.6668,
+      "step": 7180
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9453940561164815,
+      "learning_rate": 5.89227455399439e-06,
+      "loss": 0.5942,
+      "step": 7181
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4557355119434703,
+      "learning_rate": 5.8912546852450116e-06,
+      "loss": 0.6448,
+      "step": 7182
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4293851231217256,
+      "learning_rate": 5.890234778196435e-06,
+      "loss": 0.7388,
+      "step": 7183
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4835652561356,
+      "learning_rate": 5.889214832892489e-06,
+      "loss": 0.6631,
+      "step": 7184
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3011930576289932,
+      "learning_rate": 5.888194849376998e-06,
+      "loss": 0.6116,
+      "step": 7185
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2067902973663756,
+      "learning_rate": 5.8871748276937975e-06,
+      "loss": 0.8095,
+      "step": 7186
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.302788319441171,
+      "learning_rate": 5.886154767886719e-06,
+      "loss": 0.7206,
+      "step": 7187
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3778826357276381,
+      "learning_rate": 5.885134669999597e-06,
+      "loss": 0.6866,
+      "step": 7188
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.322261599537146,
+      "learning_rate": 5.8841145340762665e-06,
+      "loss": 0.6794,
+      "step": 7189
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5591157443184083,
+      "learning_rate": 5.8830943601605665e-06,
+      "loss": 0.5978,
+      "step": 7190
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.451681624164651,
+      "learning_rate": 5.882074148296332e-06,
+      "loss": 0.5637,
+      "step": 7191
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.643123252890569,
+      "learning_rate": 5.881053898527409e-06,
+      "loss": 0.8225,
+      "step": 7192
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.370499479071422,
+      "learning_rate": 5.880033610897638e-06,
+      "loss": 0.6703,
+      "step": 7193
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.454465158476384,
+      "learning_rate": 5.879013285450863e-06,
+      "loss": 0.6407,
+      "step": 7194
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5464664069534482,
+      "learning_rate": 5.877992922230928e-06,
+      "loss": 0.699,
+      "step": 7195
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2447126130512889,
+      "learning_rate": 5.876972521281683e-06,
+      "loss": 0.628,
+      "step": 7196
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4804421300112722,
+      "learning_rate": 5.875952082646975e-06,
+      "loss": 0.6001,
+      "step": 7197
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6077061822528083,
+      "learning_rate": 5.8749316063706565e-06,
+      "loss": 0.6703,
+      "step": 7198
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.292901271514413,
+      "learning_rate": 5.873911092496577e-06,
+      "loss": 0.6048,
+      "step": 7199
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4707759601948216,
+      "learning_rate": 5.872890541068591e-06,
+      "loss": 0.5912,
+      "step": 7200
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.366702890541019,
+      "learning_rate": 5.871869952130556e-06,
+      "loss": 0.6407,
+      "step": 7201
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2721721372169332,
+      "learning_rate": 5.870849325726326e-06,
+      "loss": 0.5879,
+      "step": 7202
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.343695625310671,
+      "learning_rate": 5.869828661899761e-06,
+      "loss": 0.6442,
+      "step": 7203
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.26090619076882,
+      "learning_rate": 5.8688079606947226e-06,
+      "loss": 0.6075,
+      "step": 7204
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4773058171469369,
+      "learning_rate": 5.86778722215507e-06,
+      "loss": 0.7004,
+      "step": 7205
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.299437260235512,
+      "learning_rate": 5.866766446324668e-06,
+      "loss": 0.6357,
+      "step": 7206
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4338975223268946,
+      "learning_rate": 5.865745633247381e-06,
+      "loss": 0.6488,
+      "step": 7207
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1450937616143544,
+      "learning_rate": 5.864724782967077e-06,
+      "loss": 0.538,
+      "step": 7208
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6793106733360104,
+      "learning_rate": 5.8637038955276225e-06,
+      "loss": 0.6505,
+      "step": 7209
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2992877737231008,
+      "learning_rate": 5.862682970972888e-06,
+      "loss": 0.6842,
+      "step": 7210
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4060177488072194,
+      "learning_rate": 5.861662009346747e-06,
+      "loss": 0.6509,
+      "step": 7211
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0866047055642876,
+      "learning_rate": 5.8606410106930675e-06,
+      "loss": 0.5262,
+      "step": 7212
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.142344813428512,
+      "learning_rate": 5.85961997505573e-06,
+      "loss": 0.6992,
+      "step": 7213
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5117677199333046,
+      "learning_rate": 5.858598902478604e-06,
+      "loss": 0.657,
+      "step": 7214
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3010467556955299,
+      "learning_rate": 5.857577793005573e-06,
+      "loss": 0.5395,
+      "step": 7215
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3273286621800315,
+      "learning_rate": 5.856556646680514e-06,
+      "loss": 0.5531,
+      "step": 7216
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0726273097071144,
+      "learning_rate": 5.855535463547309e-06,
+      "loss": 0.7298,
+      "step": 7217
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.338082146973492,
+      "learning_rate": 5.854514243649839e-06,
+      "loss": 0.6189,
+      "step": 7218
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2614311976715626,
+      "learning_rate": 5.853492987031989e-06,
+      "loss": 0.6826,
+      "step": 7219
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3744842082863553,
+      "learning_rate": 5.8524716937376446e-06,
+      "loss": 0.6438,
+      "step": 7220
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4330271901156597,
+      "learning_rate": 5.851450363810694e-06,
+      "loss": 0.5751,
+      "step": 7221
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7377213082858878,
+      "learning_rate": 5.850428997295023e-06,
+      "loss": 0.7508,
+      "step": 7222
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6248869984169594,
+      "learning_rate": 5.8494075942345244e-06,
+      "loss": 0.6487,
+      "step": 7223
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2059430709936223,
+      "learning_rate": 5.8483861546730915e-06,
+      "loss": 0.724,
+      "step": 7224
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.408331597066521,
+      "learning_rate": 5.8473646786546134e-06,
+      "loss": 0.6631,
+      "step": 7225
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6388338158466371,
+      "learning_rate": 5.846343166222991e-06,
+      "loss": 0.7819,
+      "step": 7226
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3351627285732466,
+      "learning_rate": 5.845321617422115e-06,
+      "loss": 0.605,
+      "step": 7227
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3698047002178915,
+      "learning_rate": 5.8443000322958875e-06,
+      "loss": 0.6141,
+      "step": 7228
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1914013347147143,
+      "learning_rate": 5.843278410888208e-06,
+      "loss": 0.5928,
+      "step": 7229
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3357295481593896,
+      "learning_rate": 5.842256753242975e-06,
+      "loss": 0.6373,
+      "step": 7230
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5883142155467433,
+      "learning_rate": 5.841235059404097e-06,
+      "loss": 0.6376,
+      "step": 7231
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4869995811059253,
+      "learning_rate": 5.840213329415471e-06,
+      "loss": 0.648,
+      "step": 7232
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5089098693869738,
+      "learning_rate": 5.839191563321009e-06,
+      "loss": 0.7254,
+      "step": 7233
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1016054521880134,
+      "learning_rate": 5.838169761164616e-06,
+      "loss": 0.5393,
+      "step": 7234
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.396589376648525,
+      "learning_rate": 5.837147922990204e-06,
+      "loss": 0.6377,
+      "step": 7235
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2414782080276419,
+      "learning_rate": 5.836126048841678e-06,
+      "loss": 0.5879,
+      "step": 7236
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.206729179141129,
+      "learning_rate": 5.8351041387629546e-06,
+      "loss": 0.6507,
+      "step": 7237
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.170154614154385,
+      "learning_rate": 5.834082192797948e-06,
+      "loss": 0.627,
+      "step": 7238
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1362838622164637,
+      "learning_rate": 5.83306021099057e-06,
+      "loss": 0.6861,
+      "step": 7239
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2261927338175527,
+      "learning_rate": 5.832038193384741e-06,
+      "loss": 0.646,
+      "step": 7240
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.37619491305248,
+      "learning_rate": 5.831016140024377e-06,
+      "loss": 0.7575,
+      "step": 7241
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1293456970163938,
+      "learning_rate": 5.829994050953398e-06,
+      "loss": 0.5998,
+      "step": 7242
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.492231675027839,
+      "learning_rate": 5.828971926215726e-06,
+      "loss": 0.7079,
+      "step": 7243
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2827998485607575,
+      "learning_rate": 5.827949765855285e-06,
+      "loss": 0.6547,
+      "step": 7244
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.351923560354576,
+      "learning_rate": 5.826927569915999e-06,
+      "loss": 0.6719,
+      "step": 7245
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3041699401093274,
+      "learning_rate": 5.825905338441794e-06,
+      "loss": 0.6641,
+      "step": 7246
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3104895296726846,
+      "learning_rate": 5.824883071476597e-06,
+      "loss": 0.7042,
+      "step": 7247
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6235692646632138,
+      "learning_rate": 5.823860769064335e-06,
+      "loss": 0.7151,
+      "step": 7248
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4421443422488842,
+      "learning_rate": 5.822838431248943e-06,
+      "loss": 0.6398,
+      "step": 7249
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5671018127549643,
+      "learning_rate": 5.821816058074351e-06,
+      "loss": 0.6528,
+      "step": 7250
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5929583257599103,
+      "learning_rate": 5.820793649584493e-06,
+      "loss": 0.6376,
+      "step": 7251
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4638829213123496,
+      "learning_rate": 5.819771205823303e-06,
+      "loss": 0.6532,
+      "step": 7252
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6307165792281957,
+      "learning_rate": 5.818748726834718e-06,
+      "loss": 0.8375,
+      "step": 7253
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3917608428825539,
+      "learning_rate": 5.817726212662678e-06,
+      "loss": 0.5932,
+      "step": 7254
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.324902972478672,
+      "learning_rate": 5.8167036633511206e-06,
+      "loss": 0.727,
+      "step": 7255
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3438008866062428,
+      "learning_rate": 5.815681078943989e-06,
+      "loss": 0.608,
+      "step": 7256
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.011209452080815,
+      "learning_rate": 5.814658459485223e-06,
+      "loss": 0.6848,
+      "step": 7257
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2893712097030454,
+      "learning_rate": 5.81363580501877e-06,
+      "loss": 0.5872,
+      "step": 7258
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5158280113952602,
+      "learning_rate": 5.812613115588575e-06,
+      "loss": 0.6091,
+      "step": 7259
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4106310269274789,
+      "learning_rate": 5.811590391238584e-06,
+      "loss": 0.7638,
+      "step": 7260
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2513871830880887,
+      "learning_rate": 5.8105676320127455e-06,
+      "loss": 0.6597,
+      "step": 7261
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 11.76427044484105,
+      "learning_rate": 5.809544837955011e-06,
+      "loss": 0.7246,
+      "step": 7262
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5021066851378662,
+      "learning_rate": 5.808522009109333e-06,
+      "loss": 0.6872,
+      "step": 7263
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4175833789593337,
+      "learning_rate": 5.807499145519663e-06,
+      "loss": 0.6318,
+      "step": 7264
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7596807175117295,
+      "learning_rate": 5.806476247229956e-06,
+      "loss": 0.728,
+      "step": 7265
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.20526946851165,
+      "learning_rate": 5.805453314284168e-06,
+      "loss": 0.6376,
+      "step": 7266
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4393606664310483,
+      "learning_rate": 5.804430346726258e-06,
+      "loss": 0.6276,
+      "step": 7267
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.103436082005873,
+      "learning_rate": 5.803407344600185e-06,
+      "loss": 0.657,
+      "step": 7268
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3333160402060664,
+      "learning_rate": 5.802384307949909e-06,
+      "loss": 0.7057,
+      "step": 7269
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3684198108149557,
+      "learning_rate": 5.8013612368193905e-06,
+      "loss": 0.7502,
+      "step": 7270
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6610837290402596,
+      "learning_rate": 5.800338131252597e-06,
+      "loss": 0.7103,
+      "step": 7271
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3315071451256901,
+      "learning_rate": 5.79931499129349e-06,
+      "loss": 0.6138,
+      "step": 7272
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5613786820344353,
+      "learning_rate": 5.7982918169860395e-06,
+      "loss": 0.6331,
+      "step": 7273
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.38545437863561,
+      "learning_rate": 5.79726860837421e-06,
+      "loss": 0.5146,
+      "step": 7274
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4640481874376625,
+      "learning_rate": 5.796245365501973e-06,
+      "loss": 0.6328,
+      "step": 7275
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3635086003772183,
+      "learning_rate": 5.795222088413299e-06,
+      "loss": 0.6553,
+      "step": 7276
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6838999430430497,
+      "learning_rate": 5.794198777152162e-06,
+      "loss": 0.6735,
+      "step": 7277
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3676851384809066,
+      "learning_rate": 5.793175431762534e-06,
+      "loss": 0.7034,
+      "step": 7278
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.578031625602269,
+      "learning_rate": 5.792152052288391e-06,
+      "loss": 0.6571,
+      "step": 7279
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4425100558055945,
+      "learning_rate": 5.791128638773711e-06,
+      "loss": 0.7911,
+      "step": 7280
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1332034345604656,
+      "learning_rate": 5.790105191262471e-06,
+      "loss": 0.5805,
+      "step": 7281
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5870700800503945,
+      "learning_rate": 5.789081709798652e-06,
+      "loss": 0.7151,
+      "step": 7282
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3367570036747665,
+      "learning_rate": 5.788058194426234e-06,
+      "loss": 0.6418,
+      "step": 7283
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4193472992246678,
+      "learning_rate": 5.787034645189199e-06,
+      "loss": 0.6006,
+      "step": 7284
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3668498071219222,
+      "learning_rate": 5.786011062131535e-06,
+      "loss": 0.6322,
+      "step": 7285
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.307204245492043,
+      "learning_rate": 5.784987445297222e-06,
+      "loss": 0.6401,
+      "step": 7286
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3042136429966429,
+      "learning_rate": 5.783963794730254e-06,
+      "loss": 0.6079,
+      "step": 7287
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4634909759068147,
+      "learning_rate": 5.782940110474612e-06,
+      "loss": 0.587,
+      "step": 7288
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2098455702051154,
+      "learning_rate": 5.7819163925742915e-06,
+      "loss": 0.6851,
+      "step": 7289
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3104391533040605,
+      "learning_rate": 5.780892641073281e-06,
+      "loss": 0.5857,
+      "step": 7290
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.462368615598109,
+      "learning_rate": 5.779868856015578e-06,
+      "loss": 0.6759,
+      "step": 7291
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4089211384773124,
+      "learning_rate": 5.778845037445171e-06,
+      "loss": 0.6994,
+      "step": 7292
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.318964168687437,
+      "learning_rate": 5.7778211854060575e-06,
+      "loss": 0.6029,
+      "step": 7293
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6396915346291654,
+      "learning_rate": 5.776797299942236e-06,
+      "loss": 0.6484,
+      "step": 7294
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7097149630981738,
+      "learning_rate": 5.775773381097705e-06,
+      "loss": 0.6207,
+      "step": 7295
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2352160572649027,
+      "learning_rate": 5.774749428916464e-06,
+      "loss": 0.6511,
+      "step": 7296
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1075303503027114,
+      "learning_rate": 5.773725443442514e-06,
+      "loss": 0.5733,
+      "step": 7297
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2547575524250671,
+      "learning_rate": 5.7727014247198585e-06,
+      "loss": 0.5963,
+      "step": 7298
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6255466580165538,
+      "learning_rate": 5.771677372792502e-06,
+      "loss": 0.6525,
+      "step": 7299
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.284063075229078,
+      "learning_rate": 5.770653287704451e-06,
+      "loss": 0.6011,
+      "step": 7300
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2909150193763579,
+      "learning_rate": 5.7696291694997105e-06,
+      "loss": 0.6936,
+      "step": 7301
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.278458537259758,
+      "learning_rate": 5.768605018222292e-06,
+      "loss": 0.743,
+      "step": 7302
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2274540172046566,
+      "learning_rate": 5.767580833916203e-06,
+      "loss": 0.6327,
+      "step": 7303
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2922383206701507,
+      "learning_rate": 5.766556616625456e-06,
+      "loss": 0.7812,
+      "step": 7304
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.312444831784848,
+      "learning_rate": 5.765532366394064e-06,
+      "loss": 0.6244,
+      "step": 7305
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1765263976555604,
+      "learning_rate": 5.764508083266043e-06,
+      "loss": 0.6742,
+      "step": 7306
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3902868030322193,
+      "learning_rate": 5.763483767285405e-06,
+      "loss": 0.5518,
+      "step": 7307
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3227904341326984,
+      "learning_rate": 5.762459418496169e-06,
+      "loss": 0.5884,
+      "step": 7308
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1478363758323264,
+      "learning_rate": 5.7614350369423555e-06,
+      "loss": 0.5965,
+      "step": 7309
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.414889803058578,
+      "learning_rate": 5.760410622667981e-06,
+      "loss": 0.6357,
+      "step": 7310
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2012586054548666,
+      "learning_rate": 5.75938617571707e-06,
+      "loss": 0.6007,
+      "step": 7311
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5948822031183856,
+      "learning_rate": 5.758361696133643e-06,
+      "loss": 0.6528,
+      "step": 7312
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3440308376783814,
+      "learning_rate": 5.757337183961724e-06,
+      "loss": 0.6327,
+      "step": 7313
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4404350272069777,
+      "learning_rate": 5.7563126392453415e-06,
+      "loss": 0.7771,
+      "step": 7314
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2424253745626122,
+      "learning_rate": 5.755288062028519e-06,
+      "loss": 0.7171,
+      "step": 7315
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5026420997149066,
+      "learning_rate": 5.754263452355287e-06,
+      "loss": 0.6767,
+      "step": 7316
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5244383787112135,
+      "learning_rate": 5.753238810269674e-06,
+      "loss": 0.815,
+      "step": 7317
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4881767229231895,
+      "learning_rate": 5.752214135815712e-06,
+      "loss": 0.7221,
+      "step": 7318
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.296833745841554,
+      "learning_rate": 5.751189429037435e-06,
+      "loss": 0.6856,
+      "step": 7319
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4084390765252033,
+      "learning_rate": 5.750164689978873e-06,
+      "loss": 0.703,
+      "step": 7320
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.396506267952014,
+      "learning_rate": 5.7491399186840646e-06,
+      "loss": 0.6362,
+      "step": 7321
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2786535459309625,
+      "learning_rate": 5.748115115197045e-06,
+      "loss": 0.6309,
+      "step": 7322
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2476451783996423,
+      "learning_rate": 5.7470902795618535e-06,
+      "loss": 0.5967,
+      "step": 7323
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4231652896971987,
+      "learning_rate": 5.746065411822528e-06,
+      "loss": 0.5825,
+      "step": 7324
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4465727235111865,
+      "learning_rate": 5.745040512023111e-06,
+      "loss": 0.6649,
+      "step": 7325
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.621298861189749,
+      "learning_rate": 5.744015580207642e-06,
+      "loss": 0.8172,
+      "step": 7326
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5829580370580147,
+      "learning_rate": 5.742990616420169e-06,
+      "loss": 0.6809,
+      "step": 7327
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4115566760409932,
+      "learning_rate": 5.741965620704734e-06,
+      "loss": 0.6111,
+      "step": 7328
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1098155058385162,
+      "learning_rate": 5.740940593105383e-06,
+      "loss": 0.6283,
+      "step": 7329
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2441667771358642,
+      "learning_rate": 5.7399155336661645e-06,
+      "loss": 0.6244,
+      "step": 7330
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.292853551729411,
+      "learning_rate": 5.738890442431128e-06,
+      "loss": 0.6061,
+      "step": 7331
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3560529683769136,
+      "learning_rate": 5.737865319444324e-06,
+      "loss": 0.7416,
+      "step": 7332
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2190474735458205,
+      "learning_rate": 5.736840164749803e-06,
+      "loss": 0.6757,
+      "step": 7333
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3034225827317498,
+      "learning_rate": 5.73581497839162e-06,
+      "loss": 0.67,
+      "step": 7334
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4328431213331765,
+      "learning_rate": 5.734789760413827e-06,
+      "loss": 0.6838,
+      "step": 7335
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.160081198971644,
+      "learning_rate": 5.733764510860482e-06,
+      "loss": 0.7172,
+      "step": 7336
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4028928187531204,
+      "learning_rate": 5.7327392297756426e-06,
+      "loss": 0.7256,
+      "step": 7337
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2986780649027818,
+      "learning_rate": 5.731713917203368e-06,
+      "loss": 0.6221,
+      "step": 7338
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4067136742114636,
+      "learning_rate": 5.730688573187715e-06,
+      "loss": 0.6867,
+      "step": 7339
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3214472126721502,
+      "learning_rate": 5.729663197772745e-06,
+      "loss": 0.6856,
+      "step": 7340
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2030299349226696,
+      "learning_rate": 5.728637791002525e-06,
+      "loss": 0.6312,
+      "step": 7341
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4403055655916626,
+      "learning_rate": 5.727612352921116e-06,
+      "loss": 0.7026,
+      "step": 7342
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5539960512847801,
+      "learning_rate": 5.726586883572584e-06,
+      "loss": 0.8169,
+      "step": 7343
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3692158677632713,
+      "learning_rate": 5.725561383000994e-06,
+      "loss": 0.701,
+      "step": 7344
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1436582811058027,
+      "learning_rate": 5.724535851250416e-06,
+      "loss": 0.6792,
+      "step": 7345
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.263639283820271,
+      "learning_rate": 5.723510288364918e-06,
+      "loss": 0.6994,
+      "step": 7346
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.463300904528656,
+      "learning_rate": 5.722484694388573e-06,
+      "loss": 0.5504,
+      "step": 7347
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3170927723748735,
+      "learning_rate": 5.721459069365452e-06,
+      "loss": 0.6324,
+      "step": 7348
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3735279838087313,
+      "learning_rate": 5.720433413339627e-06,
+      "loss": 0.6388,
+      "step": 7349
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4652162339535335,
+      "learning_rate": 5.719407726355174e-06,
+      "loss": 0.7324,
+      "step": 7350
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4713310760064089,
+      "learning_rate": 5.718382008456168e-06,
+      "loss": 0.6699,
+      "step": 7351
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4487956289572448,
+      "learning_rate": 5.717356259686689e-06,
+      "loss": 0.6474,
+      "step": 7352
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3691183696577562,
+      "learning_rate": 5.716330480090812e-06,
+      "loss": 0.6913,
+      "step": 7353
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4478521273290461,
+      "learning_rate": 5.71530466971262e-06,
+      "loss": 0.5998,
+      "step": 7354
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.499183287179133,
+      "learning_rate": 5.714278828596192e-06,
+      "loss": 0.659,
+      "step": 7355
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2984241288122613,
+      "learning_rate": 5.713252956785613e-06,
+      "loss": 0.6525,
+      "step": 7356
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4259529546824163,
+      "learning_rate": 5.712227054324968e-06,
+      "loss": 0.567,
+      "step": 7357
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2245077421954733,
+      "learning_rate": 5.711201121258337e-06,
+      "loss": 0.5819,
+      "step": 7358
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.2452673014201654,
+      "learning_rate": 5.710175157629812e-06,
+      "loss": 0.7262,
+      "step": 7359
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.443348971142059,
+      "learning_rate": 5.709149163483476e-06,
+      "loss": 0.787,
+      "step": 7360
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4583371250884662,
+      "learning_rate": 5.708123138863424e-06,
+      "loss": 0.6365,
+      "step": 7361
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.506177072693425,
+      "learning_rate": 5.707097083813743e-06,
+      "loss": 0.7371,
+      "step": 7362
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4682070590491698,
+      "learning_rate": 5.706070998378524e-06,
+      "loss": 0.6601,
+      "step": 7363
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3822428962442737,
+      "learning_rate": 5.705044882601862e-06,
+      "loss": 0.6337,
+      "step": 7364
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2285173482407719,
+      "learning_rate": 5.704018736527853e-06,
+      "loss": 0.5668,
+      "step": 7365
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1363957687803945,
+      "learning_rate": 5.7029925602005905e-06,
+      "loss": 0.551,
+      "step": 7366
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.422079319057327,
+      "learning_rate": 5.701966353664171e-06,
+      "loss": 0.7667,
+      "step": 7367
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1999317570495294,
+      "learning_rate": 5.7009401169626955e-06,
+      "loss": 0.6492,
+      "step": 7368
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0064727929997115,
+      "learning_rate": 5.69991385014026e-06,
+      "loss": 0.6575,
+      "step": 7369
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3031315629986717,
+      "learning_rate": 5.698887553240968e-06,
+      "loss": 0.6283,
+      "step": 7370
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7186992227535502,
+      "learning_rate": 5.697861226308923e-06,
+      "loss": 0.8054,
+      "step": 7371
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.294599075223805,
+      "learning_rate": 5.696834869388226e-06,
+      "loss": 0.6328,
+      "step": 7372
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2862180073503935,
+      "learning_rate": 5.695808482522982e-06,
+      "loss": 0.64,
+      "step": 7373
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4566231354049304,
+      "learning_rate": 5.694782065757298e-06,
+      "loss": 0.6283,
+      "step": 7374
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.544774430532192,
+      "learning_rate": 5.693755619135282e-06,
+      "loss": 0.7515,
+      "step": 7375
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3538301887235888,
+      "learning_rate": 5.692729142701042e-06,
+      "loss": 0.6627,
+      "step": 7376
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3018395724621432,
+      "learning_rate": 5.691702636498688e-06,
+      "loss": 0.5694,
+      "step": 7377
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.305922715354227,
+      "learning_rate": 5.69067610057233e-06,
+      "loss": 0.6512,
+      "step": 7378
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0381749323141174,
+      "learning_rate": 5.689649534966083e-06,
+      "loss": 0.7085,
+      "step": 7379
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3501342935248215,
+      "learning_rate": 5.6886229397240585e-06,
+      "loss": 0.6917,
+      "step": 7380
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5383338699810043,
+      "learning_rate": 5.6875963148903735e-06,
+      "loss": 0.7698,
+      "step": 7381
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2451791934408283,
+      "learning_rate": 5.6865696605091426e-06,
+      "loss": 0.6042,
+      "step": 7382
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5391132787762778,
+      "learning_rate": 5.685542976624485e-06,
+      "loss": 0.7976,
+      "step": 7383
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.45190529378847,
+      "learning_rate": 5.684516263280519e-06,
+      "loss": 0.6658,
+      "step": 7384
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3522676490274002,
+      "learning_rate": 5.683489520521365e-06,
+      "loss": 0.6161,
+      "step": 7385
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1607065637327085,
+      "learning_rate": 5.682462748391144e-06,
+      "loss": 0.5342,
+      "step": 7386
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.49996939589224,
+      "learning_rate": 5.681435946933978e-06,
+      "loss": 0.7317,
+      "step": 7387
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3087022136668567,
+      "learning_rate": 5.680409116193991e-06,
+      "loss": 0.6688,
+      "step": 7388
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0598754339573118,
+      "learning_rate": 5.679382256215311e-06,
+      "loss": 0.7496,
+      "step": 7389
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.189432619653814,
+      "learning_rate": 5.678355367042061e-06,
+      "loss": 0.5387,
+      "step": 7390
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.17456807624717,
+      "learning_rate": 5.67732844871837e-06,
+      "loss": 0.5626,
+      "step": 7391
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4059659871231929,
+      "learning_rate": 5.6763015012883686e-06,
+      "loss": 0.7257,
+      "step": 7392
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.374683915045708,
+      "learning_rate": 5.675274524796184e-06,
+      "loss": 0.6677,
+      "step": 7393
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3485281202969666,
+      "learning_rate": 5.674247519285951e-06,
+      "loss": 0.7022,
+      "step": 7394
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3854825227415446,
+      "learning_rate": 5.673220484801798e-06,
+      "loss": 0.7169,
+      "step": 7395
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.368727682547861,
+      "learning_rate": 5.672193421387862e-06,
+      "loss": 0.7142,
+      "step": 7396
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1226288652930947,
+      "learning_rate": 5.671166329088278e-06,
+      "loss": 0.599,
+      "step": 7397
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4007992658336954,
+      "learning_rate": 5.6701392079471825e-06,
+      "loss": 0.6289,
+      "step": 7398
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.493155395192425,
+      "learning_rate": 5.6691120580087126e-06,
+      "loss": 0.6021,
+      "step": 7399
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.246651560015247,
+      "learning_rate": 5.668084879317006e-06,
+      "loss": 0.5824,
+      "step": 7400
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4243754655166803,
+      "learning_rate": 5.6670576719162065e-06,
+      "loss": 0.664,
+      "step": 7401
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2025457232976142,
+      "learning_rate": 5.6660304358504524e-06,
+      "loss": 0.709,
+      "step": 7402
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1190512483058166,
+      "learning_rate": 5.665003171163886e-06,
+      "loss": 0.5152,
+      "step": 7403
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5595819106773856,
+      "learning_rate": 5.6639758779006535e-06,
+      "loss": 0.7041,
+      "step": 7404
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3109562783685775,
+      "learning_rate": 5.662948556104898e-06,
+      "loss": 0.6048,
+      "step": 7405
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.299684340786798,
+      "learning_rate": 5.661921205820767e-06,
+      "loss": 0.7565,
+      "step": 7406
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5346160255292145,
+      "learning_rate": 5.660893827092407e-06,
+      "loss": 0.6585,
+      "step": 7407
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.487651265968238,
+      "learning_rate": 5.659866419963969e-06,
+      "loss": 0.6146,
+      "step": 7408
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.392970561050957,
+      "learning_rate": 5.6588389844796e-06,
+      "loss": 0.6508,
+      "step": 7409
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4009610267703216,
+      "learning_rate": 5.657811520683454e-06,
+      "loss": 0.716,
+      "step": 7410
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.369317506549727,
+      "learning_rate": 5.656784028619681e-06,
+      "loss": 0.601,
+      "step": 7411
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1291809724481106,
+      "learning_rate": 5.655756508332434e-06,
+      "loss": 0.6429,
+      "step": 7412
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2925745539919,
+      "learning_rate": 5.654728959865872e-06,
+      "loss": 0.6629,
+      "step": 7413
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1777508630080271,
+      "learning_rate": 5.653701383264147e-06,
+      "loss": 0.5829,
+      "step": 7414
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5341782550914929,
+      "learning_rate": 5.652673778571418e-06,
+      "loss": 0.6786,
+      "step": 7415
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4222403272704927,
+      "learning_rate": 5.651646145831842e-06,
+      "loss": 0.7003,
+      "step": 7416
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2546322282193063,
+      "learning_rate": 5.650618485089582e-06,
+      "loss": 0.6391,
+      "step": 7417
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3045597277758996,
+      "learning_rate": 5.649590796388794e-06,
+      "loss": 0.6015,
+      "step": 7418
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1587245210630879,
+      "learning_rate": 5.648563079773646e-06,
+      "loss": 0.6871,
+      "step": 7419
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2953700525787357,
+      "learning_rate": 5.647535335288296e-06,
+      "loss": 0.626,
+      "step": 7420
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4288822586506662,
+      "learning_rate": 5.6465075629769095e-06,
+      "loss": 0.661,
+      "step": 7421
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3713174727819084,
+      "learning_rate": 5.645479762883657e-06,
+      "loss": 0.6522,
+      "step": 7422
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4516591555493532,
+      "learning_rate": 5.644451935052697e-06,
+      "loss": 0.7173,
+      "step": 7423
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3306496689450638,
+      "learning_rate": 5.6434240795282045e-06,
+      "loss": 0.5893,
+      "step": 7424
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1918233751402356,
+      "learning_rate": 5.642396196354345e-06,
+      "loss": 0.6772,
+      "step": 7425
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.267316757055061,
+      "learning_rate": 5.641368285575292e-06,
+      "loss": 0.7298,
+      "step": 7426
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2444211945940422,
+      "learning_rate": 5.640340347235215e-06,
+      "loss": 0.66,
+      "step": 7427
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.420836998948893,
+      "learning_rate": 5.639312381378286e-06,
+      "loss": 0.7087,
+      "step": 7428
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2563039892201564,
+      "learning_rate": 5.63828438804868e-06,
+      "loss": 0.5586,
+      "step": 7429
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3519451973268115,
+      "learning_rate": 5.637256367290573e-06,
+      "loss": 0.6452,
+      "step": 7430
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2546853794905386,
+      "learning_rate": 5.636228319148141e-06,
+      "loss": 0.6469,
+      "step": 7431
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3020716314313576,
+      "learning_rate": 5.635200243665562e-06,
+      "loss": 0.5965,
+      "step": 7432
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3992662471505246,
+      "learning_rate": 5.634172140887013e-06,
+      "loss": 0.6617,
+      "step": 7433
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5459976475510118,
+      "learning_rate": 5.6331440108566735e-06,
+      "loss": 0.7571,
+      "step": 7434
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6500109929792606,
+      "learning_rate": 5.632115853618727e-06,
+      "loss": 0.7279,
+      "step": 7435
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5214284908305913,
+      "learning_rate": 5.631087669217356e-06,
+      "loss": 0.6133,
+      "step": 7436
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3522985378877486,
+      "learning_rate": 5.630059457696741e-06,
+      "loss": 0.6827,
+      "step": 7437
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.311821612875022,
+      "learning_rate": 5.629031219101065e-06,
+      "loss": 0.7026,
+      "step": 7438
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.303452089769175,
+      "learning_rate": 5.628002953474521e-06,
+      "loss": 0.6162,
+      "step": 7439
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5616986242782762,
+      "learning_rate": 5.626974660861288e-06,
+      "loss": 0.6736,
+      "step": 7440
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4025989489573738,
+      "learning_rate": 5.6259463413055604e-06,
+      "loss": 0.7311,
+      "step": 7441
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7940568546964248,
+      "learning_rate": 5.624917994851523e-06,
+      "loss": 0.6348,
+      "step": 7442
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.319458510309299,
+      "learning_rate": 5.6238896215433666e-06,
+      "loss": 0.6456,
+      "step": 7443
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3207820725170425,
+      "learning_rate": 5.622861221425286e-06,
+      "loss": 0.6607,
+      "step": 7444
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4923386217361463,
+      "learning_rate": 5.6218327945414695e-06,
+      "loss": 0.7212,
+      "step": 7445
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0507572422453775,
+      "learning_rate": 5.620804340936114e-06,
+      "loss": 0.6575,
+      "step": 7446
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3667691683498318,
+      "learning_rate": 5.619775860653412e-06,
+      "loss": 0.7054,
+      "step": 7447
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4846998486703795,
+      "learning_rate": 5.6187473537375635e-06,
+      "loss": 0.7911,
+      "step": 7448
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4139104048928839,
+      "learning_rate": 5.617718820232762e-06,
+      "loss": 0.8095,
+      "step": 7449
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4954643174651914,
+      "learning_rate": 5.616690260183206e-06,
+      "loss": 0.7164,
+      "step": 7450
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4449291071398889,
+      "learning_rate": 5.615661673633098e-06,
+      "loss": 0.6915,
+      "step": 7451
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4569404253877367,
+      "learning_rate": 5.614633060626634e-06,
+      "loss": 0.708,
+      "step": 7452
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.685336308408025,
+      "learning_rate": 5.613604421208021e-06,
+      "loss": 0.6217,
+      "step": 7453
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3253585156326604,
+      "learning_rate": 5.612575755421459e-06,
+      "loss": 0.6618,
+      "step": 7454
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3206269024335404,
+      "learning_rate": 5.611547063311152e-06,
+      "loss": 0.66,
+      "step": 7455
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2787376472863765,
+      "learning_rate": 5.610518344921305e-06,
+      "loss": 0.6892,
+      "step": 7456
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1736588776984707,
+      "learning_rate": 5.609489600296127e-06,
+      "loss": 0.5891,
+      "step": 7457
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7457092915299168,
+      "learning_rate": 5.608460829479821e-06,
+      "loss": 0.6196,
+      "step": 7458
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3857412954246666,
+      "learning_rate": 5.607432032516601e-06,
+      "loss": 0.6956,
+      "step": 7459
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4368093080758861,
+      "learning_rate": 5.606403209450672e-06,
+      "loss": 0.6418,
+      "step": 7460
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2354086885054303,
+      "learning_rate": 5.605374360326246e-06,
+      "loss": 0.5508,
+      "step": 7461
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5352896503038396,
+      "learning_rate": 5.604345485187535e-06,
+      "loss": 0.7249,
+      "step": 7462
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.245166350245015,
+      "learning_rate": 5.603316584078754e-06,
+      "loss": 0.6289,
+      "step": 7463
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.28637195541019,
+      "learning_rate": 5.602287657044116e-06,
+      "loss": 0.6756,
+      "step": 7464
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2316772645717373,
+      "learning_rate": 5.601258704127835e-06,
+      "loss": 0.6379,
+      "step": 7465
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4646053854171754,
+      "learning_rate": 5.600229725374129e-06,
+      "loss": 0.7973,
+      "step": 7466
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.330619940733785,
+      "learning_rate": 5.599200720827215e-06,
+      "loss": 0.6415,
+      "step": 7467
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3194118791076435,
+      "learning_rate": 5.598171690531312e-06,
+      "loss": 0.7276,
+      "step": 7468
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4890886416489308,
+      "learning_rate": 5.597142634530639e-06,
+      "loss": 0.7484,
+      "step": 7469
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4877736299035285,
+      "learning_rate": 5.5961135528694155e-06,
+      "loss": 0.6855,
+      "step": 7470
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.47787135412234,
+      "learning_rate": 5.595084445591866e-06,
+      "loss": 0.6695,
+      "step": 7471
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2526147607979936,
+      "learning_rate": 5.594055312742213e-06,
+      "loss": 0.6413,
+      "step": 7472
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2103067955149802,
+      "learning_rate": 5.59302615436468e-06,
+      "loss": 0.6494,
+      "step": 7473
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3637872127415753,
+      "learning_rate": 5.5919969705034914e-06,
+      "loss": 0.6744,
+      "step": 7474
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4391044744457,
+      "learning_rate": 5.590967761202876e-06,
+      "loss": 0.6266,
+      "step": 7475
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1491689519861545,
+      "learning_rate": 5.589938526507059e-06,
+      "loss": 0.5958,
+      "step": 7476
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1806706675233747,
+      "learning_rate": 5.5889092664602696e-06,
+      "loss": 0.6983,
+      "step": 7477
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.366580682375656,
+      "learning_rate": 5.5878799811067384e-06,
+      "loss": 0.6569,
+      "step": 7478
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1680833674718651,
+      "learning_rate": 5.586850670490694e-06,
+      "loss": 0.5925,
+      "step": 7479
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4850460632712243,
+      "learning_rate": 5.58582133465637e-06,
+      "loss": 0.7872,
+      "step": 7480
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3508276639011632,
+      "learning_rate": 5.584791973647997e-06,
+      "loss": 0.7637,
+      "step": 7481
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.408968558660001,
+      "learning_rate": 5.583762587509812e-06,
+      "loss": 0.679,
+      "step": 7482
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.06531858930643,
+      "learning_rate": 5.582733176286048e-06,
+      "loss": 0.7233,
+      "step": 7483
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.476459350846206,
+      "learning_rate": 5.581703740020943e-06,
+      "loss": 0.7342,
+      "step": 7484
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2881700496697062,
+      "learning_rate": 5.580674278758732e-06,
+      "loss": 0.6769,
+      "step": 7485
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3213836377025066,
+      "learning_rate": 5.579644792543653e-06,
+      "loss": 0.6874,
+      "step": 7486
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.244106031751188,
+      "learning_rate": 5.578615281419947e-06,
+      "loss": 0.5528,
+      "step": 7487
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3992540921944605,
+      "learning_rate": 5.577585745431854e-06,
+      "loss": 0.7236,
+      "step": 7488
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.272605335838474,
+      "learning_rate": 5.576556184623615e-06,
+      "loss": 0.5805,
+      "step": 7489
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5292825346589587,
+      "learning_rate": 5.575526599039472e-06,
+      "loss": 0.618,
+      "step": 7490
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6281231868498698,
+      "learning_rate": 5.574496988723669e-06,
+      "loss": 0.7946,
+      "step": 7491
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3647810223043357,
+      "learning_rate": 5.573467353720452e-06,
+      "loss": 0.6312,
+      "step": 7492
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.37049581031457,
+      "learning_rate": 5.572437694074065e-06,
+      "loss": 0.6909,
+      "step": 7493
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5377205527912834,
+      "learning_rate": 5.571408009828757e-06,
+      "loss": 0.5747,
+      "step": 7494
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4283896500864939,
+      "learning_rate": 5.570378301028771e-06,
+      "loss": 0.7792,
+      "step": 7495
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2553109760351087,
+      "learning_rate": 5.569348567718363e-06,
+      "loss": 0.6743,
+      "step": 7496
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2580800033419823,
+      "learning_rate": 5.568318809941777e-06,
+      "loss": 0.6695,
+      "step": 7497
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3928826239922107,
+      "learning_rate": 5.567289027743266e-06,
+      "loss": 0.6819,
+      "step": 7498
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3851209232409538,
+      "learning_rate": 5.56625922116708e-06,
+      "loss": 0.6157,
+      "step": 7499
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4786631375357762,
+      "learning_rate": 5.565229390257478e-06,
+      "loss": 0.6414,
+      "step": 7500
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6341723940590698,
+      "learning_rate": 5.564199535058708e-06,
+      "loss": 0.6223,
+      "step": 7501
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4334506803872795,
+      "learning_rate": 5.563169655615029e-06,
+      "loss": 0.7508,
+      "step": 7502
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3011843787462054,
+      "learning_rate": 5.562139751970694e-06,
+      "loss": 0.5946,
+      "step": 7503
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1389569848978338,
+      "learning_rate": 5.561109824169962e-06,
+      "loss": 0.6027,
+      "step": 7504
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.370983545292414,
+      "learning_rate": 5.5600798722570925e-06,
+      "loss": 0.6333,
+      "step": 7505
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5219074685341178,
+      "learning_rate": 5.559049896276343e-06,
+      "loss": 0.9236,
+      "step": 7506
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5174899350572582,
+      "learning_rate": 5.558019896271975e-06,
+      "loss": 0.727,
+      "step": 7507
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.277241471926554,
+      "learning_rate": 5.556989872288248e-06,
+      "loss": 0.5905,
+      "step": 7508
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5836535817484596,
+      "learning_rate": 5.555959824369426e-06,
+      "loss": 0.7837,
+      "step": 7509
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.25439002374614,
+      "learning_rate": 5.554929752559772e-06,
+      "loss": 0.5548,
+      "step": 7510
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.481357563634828,
+      "learning_rate": 5.553899656903552e-06,
+      "loss": 0.6783,
+      "step": 7511
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3143096657274074,
+      "learning_rate": 5.552869537445029e-06,
+      "loss": 0.8516,
+      "step": 7512
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.310824195811083,
+      "learning_rate": 5.55183939422847e-06,
+      "loss": 0.6593,
+      "step": 7513
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3816274727354247,
+      "learning_rate": 5.550809227298144e-06,
+      "loss": 0.5886,
+      "step": 7514
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.338159006136659,
+      "learning_rate": 5.54977903669832e-06,
+      "loss": 0.6552,
+      "step": 7515
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3295886365569147,
+      "learning_rate": 5.548748822473265e-06,
+      "loss": 0.6941,
+      "step": 7516
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3831414233039474,
+      "learning_rate": 5.54771858466725e-06,
+      "loss": 0.6636,
+      "step": 7517
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3471766078836274,
+      "learning_rate": 5.546688323324548e-06,
+      "loss": 0.673,
+      "step": 7518
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3277834423827177,
+      "learning_rate": 5.545658038489433e-06,
+      "loss": 0.6738,
+      "step": 7519
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1834625857904215,
+      "learning_rate": 5.544627730206176e-06,
+      "loss": 0.5853,
+      "step": 7520
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2946579697313971,
+      "learning_rate": 5.543597398519053e-06,
+      "loss": 0.6678,
+      "step": 7521
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4626120736828587,
+      "learning_rate": 5.542567043472337e-06,
+      "loss": 0.6315,
+      "step": 7522
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.729605933493393,
+      "learning_rate": 5.541536665110309e-06,
+      "loss": 0.6753,
+      "step": 7523
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7590198542682045,
+      "learning_rate": 5.540506263477243e-06,
+      "loss": 0.6623,
+      "step": 7524
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5595009505662496,
+      "learning_rate": 5.53947583861742e-06,
+      "loss": 0.6524,
+      "step": 7525
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9304928827695786,
+      "learning_rate": 5.5384453905751175e-06,
+      "loss": 0.663,
+      "step": 7526
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2187844776767645,
+      "learning_rate": 5.537414919394618e-06,
+      "loss": 0.5694,
+      "step": 7527
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2543633841307371,
+      "learning_rate": 5.536384425120202e-06,
+      "loss": 0.6475,
+      "step": 7528
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4312335966180119,
+      "learning_rate": 5.535353907796155e-06,
+      "loss": 0.768,
+      "step": 7529
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4776546665358932,
+      "learning_rate": 5.534323367466758e-06,
+      "loss": 0.6459,
+      "step": 7530
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2394524762970653,
+      "learning_rate": 5.5332928041762936e-06,
+      "loss": 0.5401,
+      "step": 7531
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6894405040313722,
+      "learning_rate": 5.5322622179690514e-06,
+      "loss": 0.6871,
+      "step": 7532
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3372951019756678,
+      "learning_rate": 5.531231608889313e-06,
+      "loss": 0.683,
+      "step": 7533
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2483978197905605,
+      "learning_rate": 5.530200976981375e-06,
+      "loss": 0.575,
+      "step": 7534
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.323620231471372,
+      "learning_rate": 5.529170322289515e-06,
+      "loss": 0.7375,
+      "step": 7535
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2787314236751877,
+      "learning_rate": 5.528139644858031e-06,
+      "loss": 0.6698,
+      "step": 7536
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2531645415423205,
+      "learning_rate": 5.527108944731207e-06,
+      "loss": 0.664,
+      "step": 7537
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4648357688826312,
+      "learning_rate": 5.526078221953341e-06,
+      "loss": 0.6607,
+      "step": 7538
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1805227658428523,
+      "learning_rate": 5.525047476568722e-06,
+      "loss": 0.5651,
+      "step": 7539
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5380346060500916,
+      "learning_rate": 5.524016708621641e-06,
+      "loss": 0.6338,
+      "step": 7540
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3010837690848276,
+      "learning_rate": 5.522985918156397e-06,
+      "loss": 0.698,
+      "step": 7541
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5206345457867505,
+      "learning_rate": 5.5219551052172825e-06,
+      "loss": 0.7663,
+      "step": 7542
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.091110692906463,
+      "learning_rate": 5.5209242698485955e-06,
+      "loss": 0.5362,
+      "step": 7543
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2633821569928547,
+      "learning_rate": 5.519893412094631e-06,
+      "loss": 0.5554,
+      "step": 7544
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2572400804880368,
+      "learning_rate": 5.518862531999691e-06,
+      "loss": 0.5815,
+      "step": 7545
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.308538067879664,
+      "learning_rate": 5.51783162960807e-06,
+      "loss": 0.5751,
+      "step": 7546
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.212010521190663,
+      "learning_rate": 5.516800704964071e-06,
+      "loss": 0.506,
+      "step": 7547
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3870063795659866,
+      "learning_rate": 5.515769758111996e-06,
+      "loss": 0.7124,
+      "step": 7548
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5345544043807222,
+      "learning_rate": 5.514738789096146e-06,
+      "loss": 0.7797,
+      "step": 7549
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3186350177659054,
+      "learning_rate": 5.513707797960823e-06,
+      "loss": 0.6551,
+      "step": 7550
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3859071308981878,
+      "learning_rate": 5.512676784750332e-06,
+      "loss": 0.5993,
+      "step": 7551
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2879660924079228,
+      "learning_rate": 5.511645749508977e-06,
+      "loss": 0.6422,
+      "step": 7552
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.142649162726335,
+      "learning_rate": 5.5106146922810664e-06,
+      "loss": 0.4638,
+      "step": 7553
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8234064176364333,
+      "learning_rate": 5.509583613110904e-06,
+      "loss": 0.6483,
+      "step": 7554
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2804834613687268,
+      "learning_rate": 5.508552512042799e-06,
+      "loss": 0.6662,
+      "step": 7555
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0757544503065584,
+      "learning_rate": 5.50752138912106e-06,
+      "loss": 0.6233,
+      "step": 7556
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3523056262801385,
+      "learning_rate": 5.506490244389997e-06,
+      "loss": 0.7063,
+      "step": 7557
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2724929728811083,
+      "learning_rate": 5.50545907789392e-06,
+      "loss": 0.6557,
+      "step": 7558
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6912001865589237,
+      "learning_rate": 5.504427889677141e-06,
+      "loss": 0.704,
+      "step": 7559
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.315052731797768,
+      "learning_rate": 5.50339667978397e-06,
+      "loss": 0.6473,
+      "step": 7560
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4309266012008115,
+      "learning_rate": 5.502365448258724e-06,
+      "loss": 0.7038,
+      "step": 7561
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3286605653729482,
+      "learning_rate": 5.5013341951457166e-06,
+      "loss": 0.651,
+      "step": 7562
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3136707235131424,
+      "learning_rate": 5.500302920489262e-06,
+      "loss": 0.6751,
+      "step": 7563
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2183340948756234,
+      "learning_rate": 5.499271624333676e-06,
+      "loss": 0.5967,
+      "step": 7564
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1562757615568584,
+      "learning_rate": 5.498240306723277e-06,
+      "loss": 0.6137,
+      "step": 7565
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0135102123328477,
+      "learning_rate": 5.497208967702382e-06,
+      "loss": 0.611,
+      "step": 7566
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2265857521858716,
+      "learning_rate": 5.496177607315312e-06,
+      "loss": 0.5862,
+      "step": 7567
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3048747492143105,
+      "learning_rate": 5.495146225606383e-06,
+      "loss": 0.623,
+      "step": 7568
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0411818321272424,
+      "learning_rate": 5.494114822619918e-06,
+      "loss": 0.645,
+      "step": 7569
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4541549440184462,
+      "learning_rate": 5.493083398400239e-06,
+      "loss": 0.7591,
+      "step": 7570
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2888119738718626,
+      "learning_rate": 5.492051952991669e-06,
+      "loss": 0.6207,
+      "step": 7571
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4017421197314766,
+      "learning_rate": 5.4910204864385306e-06,
+      "loss": 0.6776,
+      "step": 7572
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3119388150921094,
+      "learning_rate": 5.489988998785147e-06,
+      "loss": 0.6126,
+      "step": 7573
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5049370508426458,
+      "learning_rate": 5.488957490075846e-06,
+      "loss": 0.6806,
+      "step": 7574
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5534866409637937,
+      "learning_rate": 5.487925960354953e-06,
+      "loss": 0.6708,
+      "step": 7575
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2153101328900524,
+      "learning_rate": 5.486894409666794e-06,
+      "loss": 0.5825,
+      "step": 7576
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2687858344709224,
+      "learning_rate": 5.485862838055699e-06,
+      "loss": 0.6713,
+      "step": 7577
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2748616824829386,
+      "learning_rate": 5.4848312455659935e-06,
+      "loss": 0.5931,
+      "step": 7578
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.253850210773842,
+      "learning_rate": 5.483799632242012e-06,
+      "loss": 0.6959,
+      "step": 7579
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3861207136156304,
+      "learning_rate": 5.482767998128081e-06,
+      "loss": 0.664,
+      "step": 7580
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5248281772903827,
+      "learning_rate": 5.4817363432685355e-06,
+      "loss": 0.7277,
+      "step": 7581
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3661500869365508,
+      "learning_rate": 5.4807046677077045e-06,
+      "loss": 0.8189,
+      "step": 7582
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4357964399984406,
+      "learning_rate": 5.479672971489924e-06,
+      "loss": 0.6212,
+      "step": 7583
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4882135650764168,
+      "learning_rate": 5.478641254659528e-06,
+      "loss": 0.7715,
+      "step": 7584
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.241003843729453,
+      "learning_rate": 5.4776095172608535e-06,
+      "loss": 0.6721,
+      "step": 7585
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9653824720067455,
+      "learning_rate": 5.476577759338232e-06,
+      "loss": 0.6273,
+      "step": 7586
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1825559142488085,
+      "learning_rate": 5.475545980936003e-06,
+      "loss": 0.5989,
+      "step": 7587
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3080968959406623,
+      "learning_rate": 5.474514182098504e-06,
+      "loss": 0.7633,
+      "step": 7588
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.196081046535042,
+      "learning_rate": 5.473482362870073e-06,
+      "loss": 0.6338,
+      "step": 7589
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.146689784266956,
+      "learning_rate": 5.472450523295052e-06,
+      "loss": 0.6066,
+      "step": 7590
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7406428931406424,
+      "learning_rate": 5.471418663417779e-06,
+      "loss": 0.8965,
+      "step": 7591
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.285453588973629,
+      "learning_rate": 5.470386783282597e-06,
+      "loss": 0.5737,
+      "step": 7592
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.275064919717669,
+      "learning_rate": 5.469354882933845e-06,
+      "loss": 0.6215,
+      "step": 7593
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2903284214821755,
+      "learning_rate": 5.468322962415871e-06,
+      "loss": 0.5849,
+      "step": 7594
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4347735388517535,
+      "learning_rate": 5.4672910217730155e-06,
+      "loss": 0.6785,
+      "step": 7595
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0153397776427777,
+      "learning_rate": 5.4662590610496235e-06,
+      "loss": 0.6491,
+      "step": 7596
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3861229259813372,
+      "learning_rate": 5.465227080290042e-06,
+      "loss": 0.6298,
+      "step": 7597
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5740576529439643,
+      "learning_rate": 5.464195079538615e-06,
+      "loss": 0.687,
+      "step": 7598
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3960723213257644,
+      "learning_rate": 5.463163058839694e-06,
+      "loss": 0.7049,
+      "step": 7599
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3790670187776877,
+      "learning_rate": 5.462131018237623e-06,
+      "loss": 0.6084,
+      "step": 7600
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2996446334067133,
+      "learning_rate": 5.461098957776755e-06,
+      "loss": 0.6277,
+      "step": 7601
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2803882746934254,
+      "learning_rate": 5.4600668775014355e-06,
+      "loss": 0.5953,
+      "step": 7602
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3411386875738092,
+      "learning_rate": 5.459034777456018e-06,
+      "loss": 0.5523,
+      "step": 7603
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0050231184070486,
+      "learning_rate": 5.4580026576848565e-06,
+      "loss": 0.6623,
+      "step": 7604
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.5557724019502084,
+      "learning_rate": 5.4569705182322975e-06,
+      "loss": 0.6032,
+      "step": 7605
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.312403203278246,
+      "learning_rate": 5.455938359142699e-06,
+      "loss": 0.6338,
+      "step": 7606
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1818041489611113,
+      "learning_rate": 5.454906180460413e-06,
+      "loss": 0.7688,
+      "step": 7607
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.375477980669854,
+      "learning_rate": 5.453873982229797e-06,
+      "loss": 0.6308,
+      "step": 7608
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2382379155707501,
+      "learning_rate": 5.452841764495203e-06,
+      "loss": 0.6486,
+      "step": 7609
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3358638747660627,
+      "learning_rate": 5.451809527300992e-06,
+      "loss": 0.6706,
+      "step": 7610
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.418582553838702,
+      "learning_rate": 5.450777270691517e-06,
+      "loss": 0.7059,
+      "step": 7611
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8567867924751822,
+      "learning_rate": 5.449744994711141e-06,
+      "loss": 0.6488,
+      "step": 7612
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2605553807049052,
+      "learning_rate": 5.448712699404221e-06,
+      "loss": 0.5869,
+      "step": 7613
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.495620517677001,
+      "learning_rate": 5.4476803848151146e-06,
+      "loss": 0.6867,
+      "step": 7614
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.360164930627099,
+      "learning_rate": 5.446648050988187e-06,
+      "loss": 0.6671,
+      "step": 7615
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.194191570641183,
+      "learning_rate": 5.445615697967797e-06,
+      "loss": 0.6782,
+      "step": 7616
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2856455330938261,
+      "learning_rate": 5.444583325798308e-06,
+      "loss": 0.6022,
+      "step": 7617
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3160088745223621,
+      "learning_rate": 5.443550934524085e-06,
+      "loss": 0.6901,
+      "step": 7618
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4118332926124797,
+      "learning_rate": 5.442518524189489e-06,
+      "loss": 0.6238,
+      "step": 7619
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5461159561607503,
+      "learning_rate": 5.441486094838886e-06,
+      "loss": 0.6064,
+      "step": 7620
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.628891229044501,
+      "learning_rate": 5.440453646516642e-06,
+      "loss": 0.7022,
+      "step": 7621
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4959051534361443,
+      "learning_rate": 5.439421179267126e-06,
+      "loss": 0.7439,
+      "step": 7622
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.409039798781433,
+      "learning_rate": 5.438388693134702e-06,
+      "loss": 0.6311,
+      "step": 7623
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3820080719598493,
+      "learning_rate": 5.4373561881637405e-06,
+      "loss": 0.6715,
+      "step": 7624
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.073712006094816,
+      "learning_rate": 5.436323664398607e-06,
+      "loss": 0.5288,
+      "step": 7625
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6362311646032388,
+      "learning_rate": 5.435291121883675e-06,
+      "loss": 0.6872,
+      "step": 7626
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2180828862231143,
+      "learning_rate": 5.434258560663316e-06,
+      "loss": 0.6478,
+      "step": 7627
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4511549723028414,
+      "learning_rate": 5.433225980781898e-06,
+      "loss": 0.6299,
+      "step": 7628
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2413185966828948,
+      "learning_rate": 5.432193382283794e-06,
+      "loss": 0.6121,
+      "step": 7629
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0215631877523759,
+      "learning_rate": 5.431160765213379e-06,
+      "loss": 0.6552,
+      "step": 7630
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.038836132824243,
+      "learning_rate": 5.430128129615025e-06,
+      "loss": 0.6679,
+      "step": 7631
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5648520331898574,
+      "learning_rate": 5.429095475533108e-06,
+      "loss": 0.6835,
+      "step": 7632
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4234120769522645,
+      "learning_rate": 5.428062803012003e-06,
+      "loss": 0.7693,
+      "step": 7633
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4907988117781275,
+      "learning_rate": 5.4270301120960856e-06,
+      "loss": 0.7671,
+      "step": 7634
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.233065913991723,
+      "learning_rate": 5.425997402829732e-06,
+      "loss": 0.5937,
+      "step": 7635
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3743735495368792,
+      "learning_rate": 5.424964675257324e-06,
+      "loss": 0.6389,
+      "step": 7636
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4774046990617458,
+      "learning_rate": 5.423931929423235e-06,
+      "loss": 0.7145,
+      "step": 7637
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4643874245040887,
+      "learning_rate": 5.422899165371846e-06,
+      "loss": 0.6708,
+      "step": 7638
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2162790940867745,
+      "learning_rate": 5.421866383147541e-06,
+      "loss": 0.5743,
+      "step": 7639
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4047047147201661,
+      "learning_rate": 5.420833582794696e-06,
+      "loss": 0.7185,
+      "step": 7640
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.480180074056718,
+      "learning_rate": 5.419800764357697e-06,
+      "loss": 0.6319,
+      "step": 7641
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2898078616081372,
+      "learning_rate": 5.418767927880923e-06,
+      "loss": 0.602,
+      "step": 7642
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3195939683289912,
+      "learning_rate": 5.417735073408756e-06,
+      "loss": 0.684,
+      "step": 7643
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5022091617900453,
+      "learning_rate": 5.416702200985585e-06,
+      "loss": 0.6157,
+      "step": 7644
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3758502793946372,
+      "learning_rate": 5.415669310655791e-06,
+      "loss": 0.6437,
+      "step": 7645
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.315423542083344,
+      "learning_rate": 5.414636402463764e-06,
+      "loss": 0.688,
+      "step": 7646
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2349696934757026,
+      "learning_rate": 5.413603476453884e-06,
+      "loss": 0.5671,
+      "step": 7647
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2841525507375278,
+      "learning_rate": 5.4125705326705425e-06,
+      "loss": 0.6317,
+      "step": 7648
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3337356994423173,
+      "learning_rate": 5.411537571158127e-06,
+      "loss": 0.7514,
+      "step": 7649
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4241057889217614,
+      "learning_rate": 5.410504591961025e-06,
+      "loss": 0.6333,
+      "step": 7650
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4513837226403241,
+      "learning_rate": 5.409471595123628e-06,
+      "loss": 0.7215,
+      "step": 7651
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4907187645023874,
+      "learning_rate": 5.408438580690322e-06,
+      "loss": 0.6465,
+      "step": 7652
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.34787452927799,
+      "learning_rate": 5.4074055487055025e-06,
+      "loss": 0.6675,
+      "step": 7653
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5114751739787249,
+      "learning_rate": 5.406372499213557e-06,
+      "loss": 0.7205,
+      "step": 7654
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3105985313632091,
+      "learning_rate": 5.4053394322588825e-06,
+      "loss": 0.6455,
+      "step": 7655
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.333271473431849,
+      "learning_rate": 5.404306347885868e-06,
+      "loss": 0.6848,
+      "step": 7656
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3096230408650515,
+      "learning_rate": 5.403273246138912e-06,
+      "loss": 0.6091,
+      "step": 7657
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4127252577102791,
+      "learning_rate": 5.4022401270624036e-06,
+      "loss": 0.7004,
+      "step": 7658
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2257051171618234,
+      "learning_rate": 5.401206990700741e-06,
+      "loss": 0.6413,
+      "step": 7659
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3141504194459253,
+      "learning_rate": 5.400173837098324e-06,
+      "loss": 0.6157,
+      "step": 7660
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.321053227294633,
+      "learning_rate": 5.399140666299543e-06,
+      "loss": 0.5847,
+      "step": 7661
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3809212412274146,
+      "learning_rate": 5.3981074783488e-06,
+      "loss": 0.6521,
+      "step": 7662
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7372267656027054,
+      "learning_rate": 5.3970742732904905e-06,
+      "loss": 0.8133,
+      "step": 7663
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.560993094223503,
+      "learning_rate": 5.396041051169016e-06,
+      "loss": 0.6344,
+      "step": 7664
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3651663583215168,
+      "learning_rate": 5.395007812028775e-06,
+      "loss": 0.6674,
+      "step": 7665
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0833335028731754,
+      "learning_rate": 5.39397455591417e-06,
+      "loss": 0.5896,
+      "step": 7666
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2920705236353627,
+      "learning_rate": 5.3929412828696006e-06,
+      "loss": 0.6806,
+      "step": 7667
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3599778538084828,
+      "learning_rate": 5.3919079929394665e-06,
+      "loss": 0.6788,
+      "step": 7668
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4342002696175618,
+      "learning_rate": 5.390874686168176e-06,
+      "loss": 0.6187,
+      "step": 7669
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4129408794506617,
+      "learning_rate": 5.389841362600129e-06,
+      "loss": 0.6737,
+      "step": 7670
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.4412097959122154,
+      "learning_rate": 5.388808022279731e-06,
+      "loss": 0.5864,
+      "step": 7671
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.70972172967744,
+      "learning_rate": 5.387774665251385e-06,
+      "loss": 0.6258,
+      "step": 7672
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.403758635041916,
+      "learning_rate": 5.386741291559499e-06,
+      "loss": 0.59,
+      "step": 7673
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.694838304301629,
+      "learning_rate": 5.385707901248478e-06,
+      "loss": 0.7374,
+      "step": 7674
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.188113527672625,
+      "learning_rate": 5.3846744943627295e-06,
+      "loss": 0.6318,
+      "step": 7675
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4342046244372972,
+      "learning_rate": 5.38364107094666e-06,
+      "loss": 0.6779,
+      "step": 7676
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.36579303616192,
+      "learning_rate": 5.382607631044682e-06,
+      "loss": 0.6192,
+      "step": 7677
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5699546095553483,
+      "learning_rate": 5.381574174701201e-06,
+      "loss": 0.7332,
+      "step": 7678
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.361955584661753,
+      "learning_rate": 5.380540701960627e-06,
+      "loss": 0.7091,
+      "step": 7679
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3858020903621922,
+      "learning_rate": 5.379507212867372e-06,
+      "loss": 0.7481,
+      "step": 7680
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.546928967600681,
+      "learning_rate": 5.378473707465847e-06,
+      "loss": 0.7393,
+      "step": 7681
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5275216093038495,
+      "learning_rate": 5.377440185800463e-06,
+      "loss": 0.6369,
+      "step": 7682
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7486063792288862,
+      "learning_rate": 5.376406647915635e-06,
+      "loss": 0.7072,
+      "step": 7683
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3238783943016414,
+      "learning_rate": 5.375373093855774e-06,
+      "loss": 0.6232,
+      "step": 7684
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5115769962017744,
+      "learning_rate": 5.374339523665295e-06,
+      "loss": 0.6371,
+      "step": 7685
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0528472276650365,
+      "learning_rate": 5.373305937388613e-06,
+      "loss": 0.6707,
+      "step": 7686
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5134432188379636,
+      "learning_rate": 5.372272335070145e-06,
+      "loss": 0.591,
+      "step": 7687
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4168928579433104,
+      "learning_rate": 5.371238716754305e-06,
+      "loss": 0.6297,
+      "step": 7688
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3775681300614906,
+      "learning_rate": 5.37020508248551e-06,
+      "loss": 0.7162,
+      "step": 7689
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2836560471903462,
+      "learning_rate": 5.369171432308178e-06,
+      "loss": 0.7152,
+      "step": 7690
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5002627978185388,
+      "learning_rate": 5.368137766266728e-06,
+      "loss": 0.5787,
+      "step": 7691
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4926119925761345,
+      "learning_rate": 5.367104084405578e-06,
+      "loss": 0.7646,
+      "step": 7692
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0419115350321377,
+      "learning_rate": 5.3660703867691475e-06,
+      "loss": 0.5318,
+      "step": 7693
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4808755321957636,
+      "learning_rate": 5.365036673401857e-06,
+      "loss": 0.5693,
+      "step": 7694
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4212257429843747,
+      "learning_rate": 5.364002944348129e-06,
+      "loss": 0.6599,
+      "step": 7695
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4125364533681144,
+      "learning_rate": 5.362969199652382e-06,
+      "loss": 0.6638,
+      "step": 7696
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.59327252348538,
+      "learning_rate": 5.361935439359042e-06,
+      "loss": 0.6521,
+      "step": 7697
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2756322453070736,
+      "learning_rate": 5.36090166351253e-06,
+      "loss": 0.6077,
+      "step": 7698
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3971896507359416,
+      "learning_rate": 5.359867872157267e-06,
+      "loss": 0.6756,
+      "step": 7699
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3477001163005686,
+      "learning_rate": 5.358834065337684e-06,
+      "loss": 0.6681,
+      "step": 7700
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5001455943631459,
+      "learning_rate": 5.357800243098199e-06,
+      "loss": 0.6417,
+      "step": 7701
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4133790375886919,
+      "learning_rate": 5.3567664054832426e-06,
+      "loss": 0.6251,
+      "step": 7702
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3714551507971529,
+      "learning_rate": 5.355732552537237e-06,
+      "loss": 0.6298,
+      "step": 7703
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.335390258337809,
+      "learning_rate": 5.354698684304613e-06,
+      "loss": 0.6386,
+      "step": 7704
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4545921658220988,
+      "learning_rate": 5.353664800829796e-06,
+      "loss": 0.6752,
+      "step": 7705
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.409399220074993,
+      "learning_rate": 5.352630902157217e-06,
+      "loss": 0.7401,
+      "step": 7706
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1673959343930684,
+      "learning_rate": 5.3515969883313e-06,
+      "loss": 0.5609,
+      "step": 7707
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.56039852550189,
+      "learning_rate": 5.350563059396478e-06,
+      "loss": 0.6122,
+      "step": 7708
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5042608256844536,
+      "learning_rate": 5.3495291153971806e-06,
+      "loss": 0.8709,
+      "step": 7709
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0668121587765769,
+      "learning_rate": 5.348495156377839e-06,
+      "loss": 0.638,
+      "step": 7710
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.69776269447896,
+      "learning_rate": 5.347461182382886e-06,
+      "loss": 0.6325,
+      "step": 7711
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3280494928719877,
+      "learning_rate": 5.346427193456749e-06,
+      "loss": 0.6482,
+      "step": 7712
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.360118972381458,
+      "learning_rate": 5.345393189643866e-06,
+      "loss": 0.6543,
+      "step": 7713
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4209679285086194,
+      "learning_rate": 5.344359170988668e-06,
+      "loss": 0.5958,
+      "step": 7714
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9695162563868223,
+      "learning_rate": 5.3433251375355884e-06,
+      "loss": 0.5605,
+      "step": 7715
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.203845849673395,
+      "learning_rate": 5.342291089329066e-06,
+      "loss": 0.5875,
+      "step": 7716
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.204596840622313,
+      "learning_rate": 5.34125702641353e-06,
+      "loss": 0.6776,
+      "step": 7717
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6411133676240925,
+      "learning_rate": 5.340222948833421e-06,
+      "loss": 0.6601,
+      "step": 7718
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5266872439459125,
+      "learning_rate": 5.339188856633173e-06,
+      "loss": 0.6923,
+      "step": 7719
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3525146734411486,
+      "learning_rate": 5.338154749857225e-06,
+      "loss": 0.6978,
+      "step": 7720
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.46907934829957,
+      "learning_rate": 5.337120628550016e-06,
+      "loss": 0.5889,
+      "step": 7721
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.6452544164730565,
+      "learning_rate": 5.3360864927559805e-06,
+      "loss": 0.7275,
+      "step": 7722
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4120002328633385,
+      "learning_rate": 5.335052342519561e-06,
+      "loss": 0.6671,
+      "step": 7723
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4190773522611813,
+      "learning_rate": 5.3340181778851954e-06,
+      "loss": 0.6122,
+      "step": 7724
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4423230002286542,
+      "learning_rate": 5.332983998897327e-06,
+      "loss": 0.6115,
+      "step": 7725
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.111435819988678,
+      "learning_rate": 5.331949805600391e-06,
+      "loss": 0.5661,
+      "step": 7726
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2580098152540613,
+      "learning_rate": 5.330915598038835e-06,
+      "loss": 0.6389,
+      "step": 7727
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4111757140456809,
+      "learning_rate": 5.329881376257098e-06,
+      "loss": 0.6075,
+      "step": 7728
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4385857233317134,
+      "learning_rate": 5.328847140299624e-06,
+      "loss": 0.6504,
+      "step": 7729
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4042644950599914,
+      "learning_rate": 5.327812890210856e-06,
+      "loss": 0.5383,
+      "step": 7730
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.287275268067156,
+      "learning_rate": 5.326778626035238e-06,
+      "loss": 0.6482,
+      "step": 7731
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3512377824964428,
+      "learning_rate": 5.3257443478172146e-06,
+      "loss": 0.6221,
+      "step": 7732
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5596979348481452,
+      "learning_rate": 5.32471005560123e-06,
+      "loss": 0.5238,
+      "step": 7733
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3194282589189439,
+      "learning_rate": 5.323675749431732e-06,
+      "loss": 0.6464,
+      "step": 7734
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2331613266722274,
+      "learning_rate": 5.322641429353167e-06,
+      "loss": 0.5555,
+      "step": 7735
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4038621226294563,
+      "learning_rate": 5.321607095409981e-06,
+      "loss": 0.6098,
+      "step": 7736
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4698970863212106,
+      "learning_rate": 5.3205727476466206e-06,
+      "loss": 0.7493,
+      "step": 7737
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5022186806537081,
+      "learning_rate": 5.3195383861075365e-06,
+      "loss": 0.6962,
+      "step": 7738
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3811046286589859,
+      "learning_rate": 5.318504010837175e-06,
+      "loss": 0.5499,
+      "step": 7739
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4140958026081905,
+      "learning_rate": 5.3174696218799894e-06,
+      "loss": 0.6065,
+      "step": 7740
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3080966519413812,
+      "learning_rate": 5.316435219280425e-06,
+      "loss": 0.6367,
+      "step": 7741
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6092009585713682,
+      "learning_rate": 5.315400803082934e-06,
+      "loss": 0.6516,
+      "step": 7742
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3748559266175726,
+      "learning_rate": 5.314366373331969e-06,
+      "loss": 0.7537,
+      "step": 7743
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5761320976047601,
+      "learning_rate": 5.313331930071981e-06,
+      "loss": 0.6863,
+      "step": 7744
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.310250558757194,
+      "learning_rate": 5.312297473347422e-06,
+      "loss": 0.5982,
+      "step": 7745
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2975364838062138,
+      "learning_rate": 5.311263003202745e-06,
+      "loss": 0.6333,
+      "step": 7746
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2985994297916117,
+      "learning_rate": 5.310228519682403e-06,
+      "loss": 0.6752,
+      "step": 7747
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.486287531883658,
+      "learning_rate": 5.309194022830852e-06,
+      "loss": 0.6835,
+      "step": 7748
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1370253859897672,
+      "learning_rate": 5.308159512692544e-06,
+      "loss": 0.7022,
+      "step": 7749
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2204838164665572,
+      "learning_rate": 5.307124989311936e-06,
+      "loss": 0.7031,
+      "step": 7750
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3891308175176929,
+      "learning_rate": 5.306090452733482e-06,
+      "loss": 0.7576,
+      "step": 7751
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4232116141876274,
+      "learning_rate": 5.305055903001642e-06,
+      "loss": 0.5643,
+      "step": 7752
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.671729159550873,
+      "learning_rate": 5.304021340160868e-06,
+      "loss": 0.6364,
+      "step": 7753
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6458941635110291,
+      "learning_rate": 5.302986764255621e-06,
+      "loss": 0.7516,
+      "step": 7754
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1364020637368528,
+      "learning_rate": 5.301952175330357e-06,
+      "loss": 0.6061,
+      "step": 7755
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1251465954512263,
+      "learning_rate": 5.300917573429536e-06,
+      "loss": 0.6612,
+      "step": 7756
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2074241036935098,
+      "learning_rate": 5.299882958597617e-06,
+      "loss": 0.5719,
+      "step": 7757
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4933641892245626,
+      "learning_rate": 5.298848330879061e-06,
+      "loss": 0.6352,
+      "step": 7758
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4830459697232714,
+      "learning_rate": 5.297813690318325e-06,
+      "loss": 0.6394,
+      "step": 7759
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5198031722300382,
+      "learning_rate": 5.29677903695987e-06,
+      "loss": 0.5996,
+      "step": 7760
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7628880203997936,
+      "learning_rate": 5.295744370848162e-06,
+      "loss": 0.7649,
+      "step": 7761
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.176236181875708,
+      "learning_rate": 5.294709692027658e-06,
+      "loss": 0.6295,
+      "step": 7762
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3912791125050876,
+      "learning_rate": 5.293675000542822e-06,
+      "loss": 0.6531,
+      "step": 7763
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2857373585063256,
+      "learning_rate": 5.292640296438116e-06,
+      "loss": 0.667,
+      "step": 7764
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.336508584915425,
+      "learning_rate": 5.291605579758007e-06,
+      "loss": 0.6258,
+      "step": 7765
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2734855067378856,
+      "learning_rate": 5.290570850546954e-06,
+      "loss": 0.6425,
+      "step": 7766
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0670173726321892,
+      "learning_rate": 5.289536108849429e-06,
+      "loss": 0.6403,
+      "step": 7767
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.342571103220208,
+      "learning_rate": 5.28850135470989e-06,
+      "loss": 0.6248,
+      "step": 7768
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0605616798974051,
+      "learning_rate": 5.287466588172804e-06,
+      "loss": 0.5659,
+      "step": 7769
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3147506178726993,
+      "learning_rate": 5.286431809282639e-06,
+      "loss": 0.6798,
+      "step": 7770
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0955540917538213,
+      "learning_rate": 5.2853970180838624e-06,
+      "loss": 0.7498,
+      "step": 7771
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.623148909218714,
+      "learning_rate": 5.2843622146209405e-06,
+      "loss": 0.6676,
+      "step": 7772
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6109084389224073,
+      "learning_rate": 5.283327398938339e-06,
+      "loss": 0.6836,
+      "step": 7773
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1574577720910524,
+      "learning_rate": 5.2822925710805305e-06,
+      "loss": 0.6096,
+      "step": 7774
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2780711837355831,
+      "learning_rate": 5.28125773109198e-06,
+      "loss": 0.7194,
+      "step": 7775
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4101159823900353,
+      "learning_rate": 5.2802228790171596e-06,
+      "loss": 0.7076,
+      "step": 7776
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0213208995706986,
+      "learning_rate": 5.279188014900537e-06,
+      "loss": 0.6888,
+      "step": 7777
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2329956098911716,
+      "learning_rate": 5.278153138786583e-06,
+      "loss": 0.7503,
+      "step": 7778
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4524381345408597,
+      "learning_rate": 5.27711825071977e-06,
+      "loss": 0.6615,
+      "step": 7779
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6392146766113949,
+      "learning_rate": 5.276083350744568e-06,
+      "loss": 0.7575,
+      "step": 7780
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.12460548218819,
+      "learning_rate": 5.275048438905452e-06,
+      "loss": 0.626,
+      "step": 7781
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4722221589162454,
+      "learning_rate": 5.27401351524689e-06,
+      "loss": 0.7205,
+      "step": 7782
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.290579876488239,
+      "learning_rate": 5.272978579813357e-06,
+      "loss": 0.6231,
+      "step": 7783
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3073599149928343,
+      "learning_rate": 5.2719436326493255e-06,
+      "loss": 0.6706,
+      "step": 7784
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.057969999440984,
+      "learning_rate": 5.270908673799272e-06,
+      "loss": 0.6228,
+      "step": 7785
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3555748153845268,
+      "learning_rate": 5.26987370330767e-06,
+      "loss": 0.6447,
+      "step": 7786
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.521015275897371,
+      "learning_rate": 5.2688387212189915e-06,
+      "loss": 0.6592,
+      "step": 7787
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3716674135023208,
+      "learning_rate": 5.267803727577717e-06,
+      "loss": 0.6402,
+      "step": 7788
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2725350120133505,
+      "learning_rate": 5.266768722428318e-06,
+      "loss": 0.6294,
+      "step": 7789
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3000735776665076,
+      "learning_rate": 5.2657337058152735e-06,
+      "loss": 0.7222,
+      "step": 7790
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4672454680788012,
+      "learning_rate": 5.26469867778306e-06,
+      "loss": 0.6514,
+      "step": 7791
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4387291303034238,
+      "learning_rate": 5.263663638376155e-06,
+      "loss": 0.6546,
+      "step": 7792
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4514370073576344,
+      "learning_rate": 5.262628587639036e-06,
+      "loss": 0.6076,
+      "step": 7793
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.432786759925151,
+      "learning_rate": 5.261593525616181e-06,
+      "loss": 0.8042,
+      "step": 7794
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5370087714366703,
+      "learning_rate": 5.2605584523520715e-06,
+      "loss": 0.63,
+      "step": 7795
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.187172353099644,
+      "learning_rate": 5.259523367891184e-06,
+      "loss": 0.5511,
+      "step": 7796
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0168497691362,
+      "learning_rate": 5.258488272277998e-06,
+      "loss": 0.6298,
+      "step": 7797
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6804297241949127,
+      "learning_rate": 5.257453165556996e-06,
+      "loss": 0.616,
+      "step": 7798
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4274213134105107,
+      "learning_rate": 5.256418047772659e-06,
+      "loss": 0.6939,
+      "step": 7799
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.280123380171981,
+      "learning_rate": 5.255382918969467e-06,
+      "loss": 0.607,
+      "step": 7800
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2534776928709201,
+      "learning_rate": 5.2543477791919025e-06,
+      "loss": 0.6588,
+      "step": 7801
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2720326990053992,
+      "learning_rate": 5.253312628484447e-06,
+      "loss": 0.6752,
+      "step": 7802
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4393118075161313,
+      "learning_rate": 5.252277466891583e-06,
+      "loss": 0.6121,
+      "step": 7803
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3307714240145567,
+      "learning_rate": 5.251242294457796e-06,
+      "loss": 0.7641,
+      "step": 7804
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3509775414921994,
+      "learning_rate": 5.2502071112275675e-06,
+      "loss": 0.6111,
+      "step": 7805
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3964316099733671,
+      "learning_rate": 5.249171917245382e-06,
+      "loss": 0.6947,
+      "step": 7806
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.873590979396924,
+      "learning_rate": 5.248136712555724e-06,
+      "loss": 0.735,
+      "step": 7807
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0989884940218044,
+      "learning_rate": 5.247101497203079e-06,
+      "loss": 0.6884,
+      "step": 7808
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.55898296550815,
+      "learning_rate": 5.2460662712319335e-06,
+      "loss": 0.7562,
+      "step": 7809
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.317341012008643,
+      "learning_rate": 5.24503103468677e-06,
+      "loss": 0.665,
+      "step": 7810
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.597174318223604,
+      "learning_rate": 5.243995787612078e-06,
+      "loss": 0.6845,
+      "step": 7811
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4647697737494665,
+      "learning_rate": 5.242960530052344e-06,
+      "loss": 0.6933,
+      "step": 7812
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5238894883446477,
+      "learning_rate": 5.241925262052055e-06,
+      "loss": 0.7016,
+      "step": 7813
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5128342647427089,
+      "learning_rate": 5.240889983655701e-06,
+      "loss": 0.6156,
+      "step": 7814
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.457622010375306,
+      "learning_rate": 5.239854694907764e-06,
+      "loss": 0.6105,
+      "step": 7815
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3252879705856124,
+      "learning_rate": 5.238819395852738e-06,
+      "loss": 0.5732,
+      "step": 7816
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.112641173563034,
+      "learning_rate": 5.2377840865351115e-06,
+      "loss": 0.6287,
+      "step": 7817
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8931060260655086,
+      "learning_rate": 5.236748766999372e-06,
+      "loss": 0.697,
+      "step": 7818
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2292424445828039,
+      "learning_rate": 5.235713437290012e-06,
+      "loss": 0.6593,
+      "step": 7819
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.242989707885668,
+      "learning_rate": 5.234678097451519e-06,
+      "loss": 0.5679,
+      "step": 7820
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4570837563589745,
+      "learning_rate": 5.233642747528387e-06,
+      "loss": 0.6669,
+      "step": 7821
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3300774010479386,
+      "learning_rate": 5.232607387565104e-06,
+      "loss": 0.6356,
+      "step": 7822
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.204824000508039,
+      "learning_rate": 5.231572017606166e-06,
+      "loss": 0.6363,
+      "step": 7823
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4766006598193602,
+      "learning_rate": 5.230536637696062e-06,
+      "loss": 0.7308,
+      "step": 7824
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4221455539084782,
+      "learning_rate": 5.229501247879284e-06,
+      "loss": 0.6415,
+      "step": 7825
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2856957008852015,
+      "learning_rate": 5.228465848200327e-06,
+      "loss": 0.5806,
+      "step": 7826
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.172715036647704,
+      "learning_rate": 5.227430438703683e-06,
+      "loss": 0.678,
+      "step": 7827
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2171560054600876,
+      "learning_rate": 5.226395019433847e-06,
+      "loss": 0.6675,
+      "step": 7828
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5663974874875166,
+      "learning_rate": 5.225359590435312e-06,
+      "loss": 0.6362,
+      "step": 7829
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2093865653779954,
+      "learning_rate": 5.224324151752575e-06,
+      "loss": 0.5545,
+      "step": 7830
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.310396279472042,
+      "learning_rate": 5.223288703430128e-06,
+      "loss": 0.7119,
+      "step": 7831
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3315382989826836,
+      "learning_rate": 5.222253245512471e-06,
+      "loss": 0.5851,
+      "step": 7832
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3243399010371417,
+      "learning_rate": 5.221217778044096e-06,
+      "loss": 0.621,
+      "step": 7833
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.392366758075126,
+      "learning_rate": 5.220182301069499e-06,
+      "loss": 0.6119,
+      "step": 7834
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2249413664989306,
+      "learning_rate": 5.219146814633178e-06,
+      "loss": 0.6128,
+      "step": 7835
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2651815389640804,
+      "learning_rate": 5.218111318779631e-06,
+      "loss": 0.527,
+      "step": 7836
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2333103083622627,
+      "learning_rate": 5.2170758135533554e-06,
+      "loss": 0.5657,
+      "step": 7837
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7765325692665375,
+      "learning_rate": 5.2160402989988475e-06,
+      "loss": 0.7982,
+      "step": 7838
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2706937152663709,
+      "learning_rate": 5.215004775160608e-06,
+      "loss": 0.6657,
+      "step": 7839
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9399023147780294,
+      "learning_rate": 5.2139692420831325e-06,
+      "loss": 0.7181,
+      "step": 7840
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3295596171466322,
+      "learning_rate": 5.212933699810924e-06,
+      "loss": 0.6596,
+      "step": 7841
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8020367050424935,
+      "learning_rate": 5.211898148388481e-06,
+      "loss": 0.7264,
+      "step": 7842
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2639584093254783,
+      "learning_rate": 5.2108625878603e-06,
+      "loss": 0.6671,
+      "step": 7843
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2537531177880668,
+      "learning_rate": 5.209827018270886e-06,
+      "loss": 0.5912,
+      "step": 7844
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3662976707498706,
+      "learning_rate": 5.208791439664736e-06,
+      "loss": 0.6219,
+      "step": 7845
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2530341875241366,
+      "learning_rate": 5.207755852086355e-06,
+      "loss": 0.6346,
+      "step": 7846
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4523235028413441,
+      "learning_rate": 5.206720255580241e-06,
+      "loss": 0.6526,
+      "step": 7847
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1145220013229928,
+      "learning_rate": 5.2056846501908975e-06,
+      "loss": 0.6441,
+      "step": 7848
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3111996700669148,
+      "learning_rate": 5.204649035962825e-06,
+      "loss": 0.655,
+      "step": 7849
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3031707736552716,
+      "learning_rate": 5.2036134129405305e-06,
+      "loss": 0.663,
+      "step": 7850
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4534155298210527,
+      "learning_rate": 5.202577781168513e-06,
+      "loss": 0.5694,
+      "step": 7851
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1704764857072587,
+      "learning_rate": 5.201542140691277e-06,
+      "loss": 0.6916,
+      "step": 7852
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.564121764048131,
+      "learning_rate": 5.200506491553326e-06,
+      "loss": 0.7531,
+      "step": 7853
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5682972733628502,
+      "learning_rate": 5.199470833799164e-06,
+      "loss": 0.7237,
+      "step": 7854
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.084815531007893,
+      "learning_rate": 5.198435167473297e-06,
+      "loss": 0.5795,
+      "step": 7855
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2441870157242416,
+      "learning_rate": 5.1973994926202296e-06,
+      "loss": 0.6393,
+      "step": 7856
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.291367459496978,
+      "learning_rate": 5.196363809284466e-06,
+      "loss": 0.6494,
+      "step": 7857
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4392866304990024,
+      "learning_rate": 5.19532811751051e-06,
+      "loss": 0.6795,
+      "step": 7858
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1036746686732464,
+      "learning_rate": 5.1942924173428725e-06,
+      "loss": 0.5779,
+      "step": 7859
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5390700771222638,
+      "learning_rate": 5.193256708826055e-06,
+      "loss": 0.7143,
+      "step": 7860
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3787784483330083,
+      "learning_rate": 5.192220992004569e-06,
+      "loss": 0.6168,
+      "step": 7861
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2855691921914738,
+      "learning_rate": 5.191185266922917e-06,
+      "loss": 0.7186,
+      "step": 7862
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3260745216752932,
+      "learning_rate": 5.190149533625607e-06,
+      "loss": 0.7059,
+      "step": 7863
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5028918019300883,
+      "learning_rate": 5.18911379215715e-06,
+      "loss": 0.6635,
+      "step": 7864
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2633826392514798,
+      "learning_rate": 5.188078042562052e-06,
+      "loss": 0.5856,
+      "step": 7865
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3884739361615308,
+      "learning_rate": 5.187042284884821e-06,
+      "loss": 0.6399,
+      "step": 7866
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.706082194428201,
+      "learning_rate": 5.186006519169967e-06,
+      "loss": 0.5241,
+      "step": 7867
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2500117128122028,
+      "learning_rate": 5.184970745461998e-06,
+      "loss": 0.5274,
+      "step": 7868
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.51439981993534,
+      "learning_rate": 5.1839349638054245e-06,
+      "loss": 0.707,
+      "step": 7869
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.409026211226792,
+      "learning_rate": 5.182899174244757e-06,
+      "loss": 0.6603,
+      "step": 7870
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1990502146046567,
+      "learning_rate": 5.181863376824504e-06,
+      "loss": 0.65,
+      "step": 7871
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4183357986967264,
+      "learning_rate": 5.1808275715891756e-06,
+      "loss": 0.6741,
+      "step": 7872
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3470080720135584,
+      "learning_rate": 5.1797917585832846e-06,
+      "loss": 0.6267,
+      "step": 7873
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5738303941479859,
+      "learning_rate": 5.178755937851341e-06,
+      "loss": 0.6664,
+      "step": 7874
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2970756622772148,
+      "learning_rate": 5.177720109437857e-06,
+      "loss": 0.5644,
+      "step": 7875
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3142460907515943,
+      "learning_rate": 5.176684273387344e-06,
+      "loss": 0.6738,
+      "step": 7876
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1837801095641212,
+      "learning_rate": 5.1756484297443135e-06,
+      "loss": 0.5645,
+      "step": 7877
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.076521704721044,
+      "learning_rate": 5.1746125785532795e-06,
+      "loss": 0.7331,
+      "step": 7878
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2649638954666937,
+      "learning_rate": 5.173576719858755e-06,
+      "loss": 0.664,
+      "step": 7879
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2720165465153113,
+      "learning_rate": 5.172540853705251e-06,
+      "loss": 0.6645,
+      "step": 7880
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3483433789379822,
+      "learning_rate": 5.171504980137282e-06,
+      "loss": 0.6956,
+      "step": 7881
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2519660658041682,
+      "learning_rate": 5.170469099199363e-06,
+      "loss": 0.6563,
+      "step": 7882
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3455496863834815,
+      "learning_rate": 5.169433210936007e-06,
+      "loss": 0.553,
+      "step": 7883
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2298207208344047,
+      "learning_rate": 5.168397315391729e-06,
+      "loss": 0.6464,
+      "step": 7884
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.181560968006825,
+      "learning_rate": 5.167361412611042e-06,
+      "loss": 0.5338,
+      "step": 7885
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3240377552375142,
+      "learning_rate": 5.1663255026384625e-06,
+      "loss": 0.6435,
+      "step": 7886
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3660175231906204,
+      "learning_rate": 5.165289585518507e-06,
+      "loss": 0.6814,
+      "step": 7887
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3374106496693752,
+      "learning_rate": 5.164253661295689e-06,
+      "loss": 0.6407,
+      "step": 7888
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2915512927099175,
+      "learning_rate": 5.1632177300145255e-06,
+      "loss": 0.6473,
+      "step": 7889
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3226889609358166,
+      "learning_rate": 5.162181791719532e-06,
+      "loss": 0.6616,
+      "step": 7890
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3401206705733928,
+      "learning_rate": 5.161145846455227e-06,
+      "loss": 0.7355,
+      "step": 7891
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1540968500426392,
+      "learning_rate": 5.160109894266125e-06,
+      "loss": 0.6103,
+      "step": 7892
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2490738026607722,
+      "learning_rate": 5.159073935196746e-06,
+      "loss": 0.7536,
+      "step": 7893
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2536821804579659,
+      "learning_rate": 5.1580379692916025e-06,
+      "loss": 0.506,
+      "step": 7894
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4294226810552488,
+      "learning_rate": 5.157001996595218e-06,
+      "loss": 0.7411,
+      "step": 7895
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5993639463143563,
+      "learning_rate": 5.155966017152108e-06,
+      "loss": 0.6078,
+      "step": 7896
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.424260323840232,
+      "learning_rate": 5.1549300310067896e-06,
+      "loss": 0.6876,
+      "step": 7897
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.366633451778539,
+      "learning_rate": 5.153894038203785e-06,
+      "loss": 0.6703,
+      "step": 7898
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.426747458424628,
+      "learning_rate": 5.152858038787608e-06,
+      "loss": 0.7849,
+      "step": 7899
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5071294005211906,
+      "learning_rate": 5.151822032802782e-06,
+      "loss": 0.6373,
+      "step": 7900
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.164710513916161,
+      "learning_rate": 5.150786020293824e-06,
+      "loss": 0.5627,
+      "step": 7901
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.174864675156024,
+      "learning_rate": 5.149750001305257e-06,
+      "loss": 0.7744,
+      "step": 7902
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3293855090436315,
+      "learning_rate": 5.148713975881598e-06,
+      "loss": 0.709,
+      "step": 7903
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.237716903459832,
+      "learning_rate": 5.147677944067368e-06,
+      "loss": 0.5824,
+      "step": 7904
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3907189273714602,
+      "learning_rate": 5.146641905907088e-06,
+      "loss": 0.6434,
+      "step": 7905
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.309594936412421,
+      "learning_rate": 5.145605861445279e-06,
+      "loss": 0.6072,
+      "step": 7906
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3671177514813941,
+      "learning_rate": 5.144569810726462e-06,
+      "loss": 0.611,
+      "step": 7907
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5348525365948533,
+      "learning_rate": 5.143533753795157e-06,
+      "loss": 0.6327,
+      "step": 7908
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2715892499771315,
+      "learning_rate": 5.142497690695888e-06,
+      "loss": 0.7316,
+      "step": 7909
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4380116725599978,
+      "learning_rate": 5.141461621473175e-06,
+      "loss": 0.644,
+      "step": 7910
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4128282745664713,
+      "learning_rate": 5.140425546171542e-06,
+      "loss": 0.6613,
+      "step": 7911
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5260553795020086,
+      "learning_rate": 5.13938946483551e-06,
+      "loss": 0.6646,
+      "step": 7912
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.67158767398759,
+      "learning_rate": 5.1383533775096025e-06,
+      "loss": 0.5868,
+      "step": 7913
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.856982083163053,
+      "learning_rate": 5.137317284238344e-06,
+      "loss": 0.6624,
+      "step": 7914
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.407244811541438,
+      "learning_rate": 5.136281185066253e-06,
+      "loss": 0.6536,
+      "step": 7915
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0104135492890864,
+      "learning_rate": 5.135245080037857e-06,
+      "loss": 0.6374,
+      "step": 7916
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.442111155592104,
+      "learning_rate": 5.1342089691976794e-06,
+      "loss": 0.6797,
+      "step": 7917
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6868308652181734,
+      "learning_rate": 5.1331728525902435e-06,
+      "loss": 0.6537,
+      "step": 7918
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2976801210357485,
+      "learning_rate": 5.1321367302600726e-06,
+      "loss": 0.5789,
+      "step": 7919
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4170412672140758,
+      "learning_rate": 5.131100602251692e-06,
+      "loss": 0.6639,
+      "step": 7920
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3344852453953697,
+      "learning_rate": 5.130064468609627e-06,
+      "loss": 0.6442,
+      "step": 7921
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2815044919793868,
+      "learning_rate": 5.129028329378403e-06,
+      "loss": 0.6229,
+      "step": 7922
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1186292320222806,
+      "learning_rate": 5.127992184602542e-06,
+      "loss": 0.6347,
+      "step": 7923
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4360024714981503,
+      "learning_rate": 5.126956034326573e-06,
+      "loss": 0.8792,
+      "step": 7924
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3690569953142657,
+      "learning_rate": 5.125919878595022e-06,
+      "loss": 0.6265,
+      "step": 7925
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1476297483952966,
+      "learning_rate": 5.124883717452412e-06,
+      "loss": 0.6696,
+      "step": 7926
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5293922128407504,
+      "learning_rate": 5.123847550943272e-06,
+      "loss": 0.6287,
+      "step": 7927
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3932530074812168,
+      "learning_rate": 5.122811379112124e-06,
+      "loss": 0.6818,
+      "step": 7928
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1168196191789237,
+      "learning_rate": 5.121775202003499e-06,
+      "loss": 0.5616,
+      "step": 7929
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4292376369097963,
+      "learning_rate": 5.120739019661923e-06,
+      "loss": 0.756,
+      "step": 7930
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7595457350618942,
+      "learning_rate": 5.119702832131922e-06,
+      "loss": 0.639,
+      "step": 7931
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4087812687735635,
+      "learning_rate": 5.118666639458023e-06,
+      "loss": 0.6575,
+      "step": 7932
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6463128298234198,
+      "learning_rate": 5.117630441684755e-06,
+      "loss": 0.7842,
+      "step": 7933
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3591717732455153,
+      "learning_rate": 5.116594238856645e-06,
+      "loss": 0.8169,
+      "step": 7934
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6152910070156097,
+      "learning_rate": 5.115558031018221e-06,
+      "loss": 0.7202,
+      "step": 7935
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.320572720865813,
+      "learning_rate": 5.114521818214011e-06,
+      "loss": 0.6712,
+      "step": 7936
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3812063232653007,
+      "learning_rate": 5.113485600488541e-06,
+      "loss": 0.5665,
+      "step": 7937
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2507852890254338,
+      "learning_rate": 5.112449377886345e-06,
+      "loss": 0.6425,
+      "step": 7938
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.299123474021115,
+      "learning_rate": 5.111413150451948e-06,
+      "loss": 0.6603,
+      "step": 7939
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2329512723748468,
+      "learning_rate": 5.11037691822988e-06,
+      "loss": 0.6463,
+      "step": 7940
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5018144674023988,
+      "learning_rate": 5.10934068126467e-06,
+      "loss": 0.6544,
+      "step": 7941
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3495973269634518,
+      "learning_rate": 5.108304439600847e-06,
+      "loss": 0.7059,
+      "step": 7942
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4271436209913848,
+      "learning_rate": 5.107268193282941e-06,
+      "loss": 0.7598,
+      "step": 7943
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1549698962552326,
+      "learning_rate": 5.1062319423554815e-06,
+      "loss": 0.6231,
+      "step": 7944
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.417615047581873,
+      "learning_rate": 5.105195686863e-06,
+      "loss": 0.671,
+      "step": 7945
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.513355679696008,
+      "learning_rate": 5.104159426850024e-06,
+      "loss": 0.78,
+      "step": 7946
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2665327874486345,
+      "learning_rate": 5.103123162361087e-06,
+      "loss": 0.6499,
+      "step": 7947
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3459237442253444,
+      "learning_rate": 5.102086893440717e-06,
+      "loss": 0.6786,
+      "step": 7948
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4628117608953075,
+      "learning_rate": 5.101050620133447e-06,
+      "loss": 0.6862,
+      "step": 7949
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.290079533578596,
+      "learning_rate": 5.100014342483804e-06,
+      "loss": 0.8003,
+      "step": 7950
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5510014709641549,
+      "learning_rate": 5.098978060536325e-06,
+      "loss": 0.7421,
+      "step": 7951
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.065026078741823,
+      "learning_rate": 5.097941774335538e-06,
+      "loss": 0.6692,
+      "step": 7952
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2413676101055888,
+      "learning_rate": 5.096905483925972e-06,
+      "loss": 0.6141,
+      "step": 7953
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4970404169507114,
+      "learning_rate": 5.095869189352166e-06,
+      "loss": 0.681,
+      "step": 7954
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1873961331848888,
+      "learning_rate": 5.094832890658644e-06,
+      "loss": 0.6257,
+      "step": 7955
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5904733994414102,
+      "learning_rate": 5.0937965878899435e-06,
+      "loss": 0.67,
+      "step": 7956
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.073370038553461,
+      "learning_rate": 5.092760281090592e-06,
+      "loss": 0.6054,
+      "step": 7957
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3515479931872605,
+      "learning_rate": 5.091723970305128e-06,
+      "loss": 0.547,
+      "step": 7958
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4167314011165493,
+      "learning_rate": 5.090687655578078e-06,
+      "loss": 0.643,
+      "step": 7959
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.302481977446522,
+      "learning_rate": 5.089651336953979e-06,
+      "loss": 0.7028,
+      "step": 7960
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0106085425481097,
+      "learning_rate": 5.088615014477362e-06,
+      "loss": 0.4414,
+      "step": 7961
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5377065465007012,
+      "learning_rate": 5.087578688192759e-06,
+      "loss": 0.6541,
+      "step": 7962
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0554225454257204,
+      "learning_rate": 5.086542358144708e-06,
+      "loss": 0.65,
+      "step": 7963
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2686344941580006,
+      "learning_rate": 5.0855060243777366e-06,
+      "loss": 0.6225,
+      "step": 7964
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0452710191635928,
+      "learning_rate": 5.084469686936382e-06,
+      "loss": 0.7219,
+      "step": 7965
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.721590495174163,
+      "learning_rate": 5.083433345865175e-06,
+      "loss": 0.7613,
+      "step": 7966
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9216238590078143,
+      "learning_rate": 5.082397001208653e-06,
+      "loss": 0.8915,
+      "step": 7967
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4958094340659176,
+      "learning_rate": 5.081360653011349e-06,
+      "loss": 0.6472,
+      "step": 7968
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.408512929255127,
+      "learning_rate": 5.080324301317795e-06,
+      "loss": 0.7977,
+      "step": 7969
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.024653126726591,
+      "learning_rate": 5.079287946172529e-06,
+      "loss": 0.6485,
+      "step": 7970
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3591435459704895,
+      "learning_rate": 5.07825158762008e-06,
+      "loss": 0.6394,
+      "step": 7971
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5913250967484982,
+      "learning_rate": 5.077215225704988e-06,
+      "loss": 0.7779,
+      "step": 7972
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8650766423781979,
+      "learning_rate": 5.076178860471787e-06,
+      "loss": 0.6087,
+      "step": 7973
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3242162283221246,
+      "learning_rate": 5.0751424919650085e-06,
+      "loss": 0.5594,
+      "step": 7974
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5023742613064561,
+      "learning_rate": 5.07410612022919e-06,
+      "loss": 0.7552,
+      "step": 7975
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.119140945851636,
+      "learning_rate": 5.073069745308868e-06,
+      "loss": 0.8531,
+      "step": 7976
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1521223263307812,
+      "learning_rate": 5.072033367248574e-06,
+      "loss": 0.5979,
+      "step": 7977
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3865361708159607,
+      "learning_rate": 5.0709969860928485e-06,
+      "loss": 0.7132,
+      "step": 7978
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3582111735684894,
+      "learning_rate": 5.069960601886224e-06,
+      "loss": 0.6586,
+      "step": 7979
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1671088501378546,
+      "learning_rate": 5.068924214673234e-06,
+      "loss": 0.5937,
+      "step": 7980
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4045445242866037,
+      "learning_rate": 5.067887824498419e-06,
+      "loss": 0.5945,
+      "step": 7981
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0400866525643224,
+      "learning_rate": 5.0668514314063135e-06,
+      "loss": 0.632,
+      "step": 7982
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3162896867334655,
+      "learning_rate": 5.065815035441452e-06,
+      "loss": 0.5586,
+      "step": 7983
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3561793751952653,
+      "learning_rate": 5.064778636648371e-06,
+      "loss": 0.6928,
+      "step": 7984
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3162766525488738,
+      "learning_rate": 5.063742235071609e-06,
+      "loss": 0.5876,
+      "step": 7985
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4648889017655795,
+      "learning_rate": 5.062705830755702e-06,
+      "loss": 0.6693,
+      "step": 7986
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.128774918643345,
+      "learning_rate": 5.061669423745185e-06,
+      "loss": 0.5129,
+      "step": 7987
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5043356048628596,
+      "learning_rate": 5.060633014084595e-06,
+      "loss": 0.6695,
+      "step": 7988
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.38649987169792,
+      "learning_rate": 5.05959660181847e-06,
+      "loss": 0.6714,
+      "step": 7989
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1237572209561677,
+      "learning_rate": 5.058560186991346e-06,
+      "loss": 0.6478,
+      "step": 7990
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3207539447243184,
+      "learning_rate": 5.0575237696477596e-06,
+      "loss": 0.5623,
+      "step": 7991
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6711754323378234,
+      "learning_rate": 5.05648734983225e-06,
+      "loss": 0.5099,
+      "step": 7992
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4102457859931583,
+      "learning_rate": 5.055450927589352e-06,
+      "loss": 0.6246,
+      "step": 7993
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6049865894821829,
+      "learning_rate": 5.054414502963605e-06,
+      "loss": 0.7974,
+      "step": 7994
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1960726450508108,
+      "learning_rate": 5.053378075999545e-06,
+      "loss": 0.6715,
+      "step": 7995
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1752592330113036,
+      "learning_rate": 5.052341646741711e-06,
+      "loss": 0.5841,
+      "step": 7996
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3031860264674393,
+      "learning_rate": 5.05130521523464e-06,
+      "loss": 0.7001,
+      "step": 7997
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.100982612199065,
+      "learning_rate": 5.050268781522868e-06,
+      "loss": 0.6017,
+      "step": 7998
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2024984145698006,
+      "learning_rate": 5.049232345650936e-06,
+      "loss": 0.5537,
+      "step": 7999
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.19270178346535,
+      "learning_rate": 5.04819590766338e-06,
+      "loss": 0.5781,
+      "step": 8000
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3444071643692892,
+      "learning_rate": 5.0471594676047385e-06,
+      "loss": 0.6346,
+      "step": 8001
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2023231493460407,
+      "learning_rate": 5.0461230255195495e-06,
+      "loss": 0.6054,
+      "step": 8002
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2211265870861092,
+      "learning_rate": 5.045086581452352e-06,
+      "loss": 0.6251,
+      "step": 8003
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2090247082698766,
+      "learning_rate": 5.044050135447682e-06,
+      "loss": 0.6072,
+      "step": 8004
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.248249489667846,
+      "learning_rate": 5.043013687550083e-06,
+      "loss": 0.7098,
+      "step": 8005
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1588419622499857,
+      "learning_rate": 5.041977237804088e-06,
+      "loss": 0.6901,
+      "step": 8006
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2398183685241835,
+      "learning_rate": 5.040940786254236e-06,
+      "loss": 0.7157,
+      "step": 8007
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2869237839823766,
+      "learning_rate": 5.039904332945069e-06,
+      "loss": 0.5783,
+      "step": 8008
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2419867041148214,
+      "learning_rate": 5.038867877921124e-06,
+      "loss": 0.6034,
+      "step": 8009
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5992156603070735,
+      "learning_rate": 5.037831421226939e-06,
+      "loss": 0.6461,
+      "step": 8010
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2547072600803868,
+      "learning_rate": 5.036794962907054e-06,
+      "loss": 0.679,
+      "step": 8011
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8067691751026953,
+      "learning_rate": 5.035758503006007e-06,
+      "loss": 0.8357,
+      "step": 8012
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2230700493202555,
+      "learning_rate": 5.034722041568337e-06,
+      "loss": 0.6575,
+      "step": 8013
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4435210093931554,
+      "learning_rate": 5.033685578638586e-06,
+      "loss": 0.7234,
+      "step": 8014
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3650108826169824,
+      "learning_rate": 5.03264911426129e-06,
+      "loss": 0.6693,
+      "step": 8015
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4839475430781575,
+      "learning_rate": 5.031612648480987e-06,
+      "loss": 0.6599,
+      "step": 8016
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.40427627162294,
+      "learning_rate": 5.03057618134222e-06,
+      "loss": 0.5954,
+      "step": 8017
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3329293234521102,
+      "learning_rate": 5.029539712889524e-06,
+      "loss": 0.6588,
+      "step": 8018
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3305927690425137,
+      "learning_rate": 5.028503243167443e-06,
+      "loss": 0.677,
+      "step": 8019
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5426251526002643,
+      "learning_rate": 5.027466772220513e-06,
+      "loss": 0.6367,
+      "step": 8020
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3814873708753501,
+      "learning_rate": 5.026430300093275e-06,
+      "loss": 0.6798,
+      "step": 8021
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.142998392905113,
+      "learning_rate": 5.025393826830267e-06,
+      "loss": 0.6119,
+      "step": 8022
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.357073723758426,
+      "learning_rate": 5.024357352476032e-06,
+      "loss": 0.674,
+      "step": 8023
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2196103256539348,
+      "learning_rate": 5.023320877075107e-06,
+      "loss": 0.6486,
+      "step": 8024
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6451440487580382,
+      "learning_rate": 5.02228440067203e-06,
+      "loss": 0.6379,
+      "step": 8025
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3083916363627153,
+      "learning_rate": 5.0212479233113445e-06,
+      "loss": 0.5597,
+      "step": 8026
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3937831580326308,
+      "learning_rate": 5.020211445037587e-06,
+      "loss": 0.7188,
+      "step": 8027
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3378575033592865,
+      "learning_rate": 5.019174965895299e-06,
+      "loss": 0.6443,
+      "step": 8028
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4379159826652577,
+      "learning_rate": 5.0181384859290215e-06,
+      "loss": 0.6143,
+      "step": 8029
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.506471061170633,
+      "learning_rate": 5.017102005183292e-06,
+      "loss": 0.609,
+      "step": 8030
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4832958023510614,
+      "learning_rate": 5.01606552370265e-06,
+      "loss": 0.6088,
+      "step": 8031
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3832356372778578,
+      "learning_rate": 5.015029041531639e-06,
+      "loss": 0.6016,
+      "step": 8032
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4000487720160582,
+      "learning_rate": 5.013992558714795e-06,
+      "loss": 0.763,
+      "step": 8033
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5075123366775076,
+      "learning_rate": 5.01295607529666e-06,
+      "loss": 0.6217,
+      "step": 8034
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5215723491071607,
+      "learning_rate": 5.011919591321773e-06,
+      "loss": 0.6232,
+      "step": 8035
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3322020754509352,
+      "learning_rate": 5.010883106834676e-06,
+      "loss": 0.6103,
+      "step": 8036
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6373766044153908,
+      "learning_rate": 5.009846621879906e-06,
+      "loss": 0.7267,
+      "step": 8037
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5521550013040533,
+      "learning_rate": 5.008810136502006e-06,
+      "loss": 0.6225,
+      "step": 8038
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4693161170934643,
+      "learning_rate": 5.007773650745514e-06,
+      "loss": 0.6717,
+      "step": 8039
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4777565162704067,
+      "learning_rate": 5.00673716465497e-06,
+      "loss": 0.631,
+      "step": 8040
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3616079516971826,
+      "learning_rate": 5.005700678274918e-06,
+      "loss": 0.6254,
+      "step": 8041
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1366741549004185,
+      "learning_rate": 5.004664191649893e-06,
+      "loss": 0.6188,
+      "step": 8042
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2976733348048815,
+      "learning_rate": 5.003627704824438e-06,
+      "loss": 0.616,
+      "step": 8043
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4402872496754817,
+      "learning_rate": 5.0025912178430925e-06,
+      "loss": 0.6327,
+      "step": 8044
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4899267129492924,
+      "learning_rate": 5.001554730750395e-06,
+      "loss": 0.662,
+      "step": 8045
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.189723033500461,
+      "learning_rate": 5.0005182435908885e-06,
+      "loss": 0.6069,
+      "step": 8046
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3665605863813837,
+      "learning_rate": 4.999481756409112e-06,
+      "loss": 0.6054,
+      "step": 8047
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1819598498527493,
+      "learning_rate": 4.998445269249606e-06,
+      "loss": 0.4947,
+      "step": 8048
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5566925443735036,
+      "learning_rate": 4.997408782156909e-06,
+      "loss": 0.748,
+      "step": 8049
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2716283327105626,
+      "learning_rate": 4.996372295175563e-06,
+      "loss": 0.6017,
+      "step": 8050
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.27935315342144,
+      "learning_rate": 4.995335808350109e-06,
+      "loss": 0.606,
+      "step": 8051
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5424679713852063,
+      "learning_rate": 4.994299321725085e-06,
+      "loss": 0.6672,
+      "step": 8052
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3691920126555133,
+      "learning_rate": 4.99326283534503e-06,
+      "loss": 0.6366,
+      "step": 8053
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2623719445111197,
+      "learning_rate": 4.9922263492544885e-06,
+      "loss": 0.6279,
+      "step": 8054
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4667381145030087,
+      "learning_rate": 4.991189863497996e-06,
+      "loss": 0.6854,
+      "step": 8055
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3154569421150484,
+      "learning_rate": 4.990153378120096e-06,
+      "loss": 0.6578,
+      "step": 8056
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4117872645731249,
+      "learning_rate": 4.989116893165325e-06,
+      "loss": 0.7107,
+      "step": 8057
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.049375508723282,
+      "learning_rate": 4.988080408678229e-06,
+      "loss": 0.6035,
+      "step": 8058
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.343823533516718,
+      "learning_rate": 4.987043924703342e-06,
+      "loss": 0.7358,
+      "step": 8059
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.191075881550418,
+      "learning_rate": 4.986007441285206e-06,
+      "loss": 0.7325,
+      "step": 8060
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3968260170612719,
+      "learning_rate": 4.984970958468364e-06,
+      "loss": 0.6428,
+      "step": 8061
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3827326142273242,
+      "learning_rate": 4.983934476297351e-06,
+      "loss": 0.7618,
+      "step": 8062
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8475161580590456,
+      "learning_rate": 4.9828979948167095e-06,
+      "loss": 0.6927,
+      "step": 8063
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.249853413854311,
+      "learning_rate": 4.981861514070979e-06,
+      "loss": 0.7701,
+      "step": 8064
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2640800386442421,
+      "learning_rate": 4.980825034104702e-06,
+      "loss": 0.62,
+      "step": 8065
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0984057652364314,
+      "learning_rate": 4.979788554962415e-06,
+      "loss": 0.5373,
+      "step": 8066
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7693832435858814,
+      "learning_rate": 4.978752076688657e-06,
+      "loss": 0.7373,
+      "step": 8067
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3067967761504957,
+      "learning_rate": 4.97771559932797e-06,
+      "loss": 0.7127,
+      "step": 8068
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3399555464696562,
+      "learning_rate": 4.976679122924896e-06,
+      "loss": 0.8153,
+      "step": 8069
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.418286474930391,
+      "learning_rate": 4.97564264752397e-06,
+      "loss": 0.6389,
+      "step": 8070
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4268457760726092,
+      "learning_rate": 4.974606173169733e-06,
+      "loss": 0.6981,
+      "step": 8071
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4726132580256435,
+      "learning_rate": 4.973569699906727e-06,
+      "loss": 0.7073,
+      "step": 8072
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3133546291155929,
+      "learning_rate": 4.972533227779489e-06,
+      "loss": 0.7123,
+      "step": 8073
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2819640372311318,
+      "learning_rate": 4.971496756832557e-06,
+      "loss": 0.6779,
+      "step": 8074
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3646500639761785,
+      "learning_rate": 4.9704602871104765e-06,
+      "loss": 0.6815,
+      "step": 8075
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4873930431607385,
+      "learning_rate": 4.969423818657783e-06,
+      "loss": 0.7092,
+      "step": 8076
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.594104569860747,
+      "learning_rate": 4.9683873515190145e-06,
+      "loss": 0.7866,
+      "step": 8077
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3793775697280768,
+      "learning_rate": 4.9673508857387115e-06,
+      "loss": 0.6013,
+      "step": 8078
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3674529349048534,
+      "learning_rate": 4.966314421361416e-06,
+      "loss": 0.6976,
+      "step": 8079
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1456329774376,
+      "learning_rate": 4.965277958431664e-06,
+      "loss": 0.5147,
+      "step": 8080
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2574226608815384,
+      "learning_rate": 4.964241496993994e-06,
+      "loss": 0.5535,
+      "step": 8081
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3518168337354937,
+      "learning_rate": 4.963205037092947e-06,
+      "loss": 0.6119,
+      "step": 8082
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4375001185479448,
+      "learning_rate": 4.962168578773063e-06,
+      "loss": 0.6577,
+      "step": 8083
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4173368011497434,
+      "learning_rate": 4.9611321220788775e-06,
+      "loss": 0.6714,
+      "step": 8084
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.315561290502196,
+      "learning_rate": 4.9600956670549324e-06,
+      "loss": 0.6032,
+      "step": 8085
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4505196622662295,
+      "learning_rate": 4.959059213745764e-06,
+      "loss": 0.7003,
+      "step": 8086
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4165584483148799,
+      "learning_rate": 4.9580227621959145e-06,
+      "loss": 0.6483,
+      "step": 8087
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.29593452627279,
+      "learning_rate": 4.956986312449919e-06,
+      "loss": 0.6305,
+      "step": 8088
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2516381020182556,
+      "learning_rate": 4.955949864552318e-06,
+      "loss": 0.6185,
+      "step": 8089
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.312879452203872,
+      "learning_rate": 4.954913418547651e-06,
+      "loss": 0.6308,
+      "step": 8090
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2256613302963109,
+      "learning_rate": 4.953876974480452e-06,
+      "loss": 0.5304,
+      "step": 8091
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2774532594035437,
+      "learning_rate": 4.952840532395262e-06,
+      "loss": 0.5147,
+      "step": 8092
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2331771471962663,
+      "learning_rate": 4.951804092336621e-06,
+      "loss": 0.7088,
+      "step": 8093
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4875476841177628,
+      "learning_rate": 4.950767654349067e-06,
+      "loss": 0.6783,
+      "step": 8094
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3937735379146214,
+      "learning_rate": 4.949731218477133e-06,
+      "loss": 0.706,
+      "step": 8095
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2173121423993418,
+      "learning_rate": 4.9486947847653615e-06,
+      "loss": 0.5678,
+      "step": 8096
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1350486032423033,
+      "learning_rate": 4.947658353258291e-06,
+      "loss": 0.7641,
+      "step": 8097
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2417563202858786,
+      "learning_rate": 4.946621924000456e-06,
+      "loss": 0.6744,
+      "step": 8098
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3167227575039893,
+      "learning_rate": 4.945585497036396e-06,
+      "loss": 0.6874,
+      "step": 8099
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.390380512172467,
+      "learning_rate": 4.944549072410648e-06,
+      "loss": 0.7341,
+      "step": 8100
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7518656237209775,
+      "learning_rate": 4.9435126501677525e-06,
+      "loss": 0.8288,
+      "step": 8101
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3485988426869873,
+      "learning_rate": 4.942476230352241e-06,
+      "loss": 0.644,
+      "step": 8102
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.113577921918832,
+      "learning_rate": 4.941439813008656e-06,
+      "loss": 0.517,
+      "step": 8103
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4014975814280253,
+      "learning_rate": 4.940403398181531e-06,
+      "loss": 0.7324,
+      "step": 8104
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9470654792064204,
+      "learning_rate": 4.939366985915408e-06,
+      "loss": 0.685,
+      "step": 8105
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6453872043708184,
+      "learning_rate": 4.938330576254817e-06,
+      "loss": 0.7549,
+      "step": 8106
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3251195098174329,
+      "learning_rate": 4.9372941692443e-06,
+      "loss": 0.6117,
+      "step": 8107
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6445578529433738,
+      "learning_rate": 4.936257764928392e-06,
+      "loss": 0.7549,
+      "step": 8108
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.433631823126446,
+      "learning_rate": 4.935221363351631e-06,
+      "loss": 0.7247,
+      "step": 8109
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.523004607911338,
+      "learning_rate": 4.934184964558549e-06,
+      "loss": 0.7234,
+      "step": 8110
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3389175700427112,
+      "learning_rate": 4.933148568593687e-06,
+      "loss": 0.591,
+      "step": 8111
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3856016582802932,
+      "learning_rate": 4.932112175501583e-06,
+      "loss": 0.6762,
+      "step": 8112
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4412207903976872,
+      "learning_rate": 4.931075785326767e-06,
+      "loss": 0.7353,
+      "step": 8113
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6637095357972915,
+      "learning_rate": 4.930039398113779e-06,
+      "loss": 0.726,
+      "step": 8114
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.424277739161314,
+      "learning_rate": 4.929003013907152e-06,
+      "loss": 0.6238,
+      "step": 8115
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.505969149390795,
+      "learning_rate": 4.927966632751427e-06,
+      "loss": 0.6802,
+      "step": 8116
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4082002159660465,
+      "learning_rate": 4.926930254691134e-06,
+      "loss": 0.67,
+      "step": 8117
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.104116001185997,
+      "learning_rate": 4.92589387977081e-06,
+      "loss": 0.5668,
+      "step": 8118
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6573060145803928,
+      "learning_rate": 4.924857508034994e-06,
+      "loss": 0.6956,
+      "step": 8119
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.389801050857735,
+      "learning_rate": 4.9238211395282156e-06,
+      "loss": 0.6942,
+      "step": 8120
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4493517859412919,
+      "learning_rate": 4.922784774295013e-06,
+      "loss": 0.6387,
+      "step": 8121
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3963002605792714,
+      "learning_rate": 4.92174841237992e-06,
+      "loss": 0.7301,
+      "step": 8122
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0647638150221006,
+      "learning_rate": 4.920712053827475e-06,
+      "loss": 0.6757,
+      "step": 8123
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2953632588682322,
+      "learning_rate": 4.919675698682206e-06,
+      "loss": 0.6415,
+      "step": 8124
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3150845809670142,
+      "learning_rate": 4.918639346988652e-06,
+      "loss": 0.6061,
+      "step": 8125
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1537043974485284,
+      "learning_rate": 4.917602998791348e-06,
+      "loss": 0.5871,
+      "step": 8126
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3444288717991808,
+      "learning_rate": 4.9165666541348265e-06,
+      "loss": 0.618,
+      "step": 8127
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8491401857921501,
+      "learning_rate": 4.91553031306362e-06,
+      "loss": 0.6462,
+      "step": 8128
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5760397145872695,
+      "learning_rate": 4.914493975622263e-06,
+      "loss": 0.698,
+      "step": 8129
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6002180417313165,
+      "learning_rate": 4.9134576418552945e-06,
+      "loss": 0.6559,
+      "step": 8130
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2574484816416298,
+      "learning_rate": 4.9124213118072415e-06,
+      "loss": 0.6607,
+      "step": 8131
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0718694779368259,
+      "learning_rate": 4.911384985522639e-06,
+      "loss": 0.7255,
+      "step": 8132
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0893438008103915,
+      "learning_rate": 4.910348663046021e-06,
+      "loss": 0.511,
+      "step": 8133
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8573615595056128,
+      "learning_rate": 4.909312344421923e-06,
+      "loss": 0.6968,
+      "step": 8134
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2126121967366723,
+      "learning_rate": 4.908276029694873e-06,
+      "loss": 0.5821,
+      "step": 8135
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.370976337686449,
+      "learning_rate": 4.907239718909408e-06,
+      "loss": 0.5566,
+      "step": 8136
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4040562218073542,
+      "learning_rate": 4.90620341211006e-06,
+      "loss": 0.6983,
+      "step": 8137
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.20340127125035,
+      "learning_rate": 4.905167109341357e-06,
+      "loss": 0.6079,
+      "step": 8138
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2507979679021037,
+      "learning_rate": 4.904130810647836e-06,
+      "loss": 0.5942,
+      "step": 8139
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4481940160607667,
+      "learning_rate": 4.903094516074027e-06,
+      "loss": 0.7435,
+      "step": 8140
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2796737021584061,
+      "learning_rate": 4.902058225664465e-06,
+      "loss": 0.6265,
+      "step": 8141
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1957837594428522,
+      "learning_rate": 4.901021939463676e-06,
+      "loss": 0.6019,
+      "step": 8142
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.418385866274468,
+      "learning_rate": 4.899985657516195e-06,
+      "loss": 0.6681,
+      "step": 8143
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.327612024424129,
+      "learning_rate": 4.898949379866556e-06,
+      "loss": 0.6853,
+      "step": 8144
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3870999472225751,
+      "learning_rate": 4.897913106559285e-06,
+      "loss": 0.6154,
+      "step": 8145
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3831279151079567,
+      "learning_rate": 4.896876837638915e-06,
+      "loss": 0.6998,
+      "step": 8146
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4285741698887118,
+      "learning_rate": 4.895840573149977e-06,
+      "loss": 0.6871,
+      "step": 8147
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.310403040141803,
+      "learning_rate": 4.8948043131370025e-06,
+      "loss": 0.704,
+      "step": 8148
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3448914811959194,
+      "learning_rate": 4.893768057644519e-06,
+      "loss": 0.7018,
+      "step": 8149
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2951575336595451,
+      "learning_rate": 4.89273180671706e-06,
+      "loss": 0.5408,
+      "step": 8150
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5274812371005977,
+      "learning_rate": 4.891695560399154e-06,
+      "loss": 0.9236,
+      "step": 8151
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7818116275553355,
+      "learning_rate": 4.890659318735333e-06,
+      "loss": 0.6067,
+      "step": 8152
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.216174380200408,
+      "learning_rate": 4.8896230817701214e-06,
+      "loss": 0.5963,
+      "step": 8153
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5122884294630325,
+      "learning_rate": 4.888586849548053e-06,
+      "loss": 0.678,
+      "step": 8154
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0982958380626868,
+      "learning_rate": 4.887550622113657e-06,
+      "loss": 0.6808,
+      "step": 8155
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4085984693775966,
+      "learning_rate": 4.88651439951146e-06,
+      "loss": 0.7183,
+      "step": 8156
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3404960945176798,
+      "learning_rate": 4.885478181785991e-06,
+      "loss": 0.6603,
+      "step": 8157
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3433806010757874,
+      "learning_rate": 4.884441968981781e-06,
+      "loss": 0.6716,
+      "step": 8158
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3474124208637674,
+      "learning_rate": 4.883405761143357e-06,
+      "loss": 0.6992,
+      "step": 8159
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1730877775775592,
+      "learning_rate": 4.882369558315246e-06,
+      "loss": 0.598,
+      "step": 8160
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.173141644097892,
+      "learning_rate": 4.8813333605419775e-06,
+      "loss": 0.5418,
+      "step": 8161
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.318318465197464,
+      "learning_rate": 4.88029716786808e-06,
+      "loss": 0.5984,
+      "step": 8162
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0434593234792935,
+      "learning_rate": 4.879260980338078e-06,
+      "loss": 0.611,
+      "step": 8163
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8471062896602963,
+      "learning_rate": 4.878224797996502e-06,
+      "loss": 0.7248,
+      "step": 8164
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.746035224440482,
+      "learning_rate": 4.877188620887876e-06,
+      "loss": 0.6178,
+      "step": 8165
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1132848223547025,
+      "learning_rate": 4.8761524490567315e-06,
+      "loss": 0.577,
+      "step": 8166
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.05681666553053,
+      "learning_rate": 4.875116282547589e-06,
+      "loss": 0.6508,
+      "step": 8167
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3336719053269335,
+      "learning_rate": 4.874080121404979e-06,
+      "loss": 0.7135,
+      "step": 8168
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3720578927531737,
+      "learning_rate": 4.873043965673427e-06,
+      "loss": 0.694,
+      "step": 8169
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6198951337088743,
+      "learning_rate": 4.872007815397459e-06,
+      "loss": 0.6782,
+      "step": 8170
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4453433953211436,
+      "learning_rate": 4.870971670621599e-06,
+      "loss": 0.6428,
+      "step": 8171
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4146715523545637,
+      "learning_rate": 4.869935531390374e-06,
+      "loss": 0.7336,
+      "step": 8172
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3042369594189627,
+      "learning_rate": 4.86889939774831e-06,
+      "loss": 0.6785,
+      "step": 8173
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2907204657678506,
+      "learning_rate": 4.86786326973993e-06,
+      "loss": 0.6741,
+      "step": 8174
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4026923780986378,
+      "learning_rate": 4.866827147409758e-06,
+      "loss": 0.5497,
+      "step": 8175
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9819240999667385,
+      "learning_rate": 4.8657910308023205e-06,
+      "loss": 0.5854,
+      "step": 8176
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2448331056688677,
+      "learning_rate": 4.864754919962144e-06,
+      "loss": 0.6221,
+      "step": 8177
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9860736640041744,
+      "learning_rate": 4.863718814933749e-06,
+      "loss": 0.6546,
+      "step": 8178
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3735812066652822,
+      "learning_rate": 4.862682715761658e-06,
+      "loss": 0.6188,
+      "step": 8179
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1651261693304469,
+      "learning_rate": 4.861646622490399e-06,
+      "loss": 0.7259,
+      "step": 8180
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.415782385188748,
+      "learning_rate": 4.860610535164491e-06,
+      "loss": 0.6155,
+      "step": 8181
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.533040557053244,
+      "learning_rate": 4.85957445382846e-06,
+      "loss": 0.6802,
+      "step": 8182
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5180232072445352,
+      "learning_rate": 4.858538378526825e-06,
+      "loss": 0.6361,
+      "step": 8183
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.142423778524322,
+      "learning_rate": 4.857502309304114e-06,
+      "loss": 0.8265,
+      "step": 8184
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9886446715212484,
+      "learning_rate": 4.856466246204844e-06,
+      "loss": 0.599,
+      "step": 8185
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2150794065843775,
+      "learning_rate": 4.8554301892735385e-06,
+      "loss": 0.5982,
+      "step": 8186
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2525416107742309,
+      "learning_rate": 4.854394138554722e-06,
+      "loss": 0.721,
+      "step": 8187
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2441188441871034,
+      "learning_rate": 4.853358094092913e-06,
+      "loss": 0.5513,
+      "step": 8188
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0634556191870523,
+      "learning_rate": 4.852322055932633e-06,
+      "loss": 0.6532,
+      "step": 8189
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6894585782301992,
+      "learning_rate": 4.851286024118402e-06,
+      "loss": 0.6981,
+      "step": 8190
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.422365919868552,
+      "learning_rate": 4.850249998694745e-06,
+      "loss": 0.6358,
+      "step": 8191
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3830378319561434,
+      "learning_rate": 4.849213979706177e-06,
+      "loss": 0.6222,
+      "step": 8192
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4473565578623055,
+      "learning_rate": 4.848177967197219e-06,
+      "loss": 0.9773,
+      "step": 8193
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2535495845551061,
+      "learning_rate": 4.8471419612123925e-06,
+      "loss": 0.6625,
+      "step": 8194
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.219976283754778,
+      "learning_rate": 4.846105961796218e-06,
+      "loss": 0.6303,
+      "step": 8195
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3115304972052293,
+      "learning_rate": 4.845069968993212e-06,
+      "loss": 0.6623,
+      "step": 8196
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.485686766033905,
+      "learning_rate": 4.844033982847893e-06,
+      "loss": 0.582,
+      "step": 8197
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6123009314791046,
+      "learning_rate": 4.842998003404784e-06,
+      "loss": 0.7962,
+      "step": 8198
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3975935714392844,
+      "learning_rate": 4.841962030708398e-06,
+      "loss": 0.7006,
+      "step": 8199
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1373970891158518,
+      "learning_rate": 4.840926064803256e-06,
+      "loss": 0.6186,
+      "step": 8200
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2802228713427128,
+      "learning_rate": 4.839890105733876e-06,
+      "loss": 0.5986,
+      "step": 8201
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.110473802593007,
+      "learning_rate": 4.8388541535447756e-06,
+      "loss": 0.6121,
+      "step": 8202
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.34622864685992,
+      "learning_rate": 4.837818208280469e-06,
+      "loss": 0.8926,
+      "step": 8203
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.337511233402135,
+      "learning_rate": 4.836782269985475e-06,
+      "loss": 0.707,
+      "step": 8204
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5599892107654483,
+      "learning_rate": 4.8357463387043124e-06,
+      "loss": 0.6031,
+      "step": 8205
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3818697328748724,
+      "learning_rate": 4.834710414481496e-06,
+      "loss": 0.6606,
+      "step": 8206
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2065957646838357,
+      "learning_rate": 4.833674497361539e-06,
+      "loss": 0.5776,
+      "step": 8207
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2876906581369738,
+      "learning_rate": 4.832638587388959e-06,
+      "loss": 0.6345,
+      "step": 8208
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3910888263703733,
+      "learning_rate": 4.831602684608274e-06,
+      "loss": 0.6699,
+      "step": 8209
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5253489488213563,
+      "learning_rate": 4.830566789063995e-06,
+      "loss": 0.7456,
+      "step": 8210
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.959875664118391,
+      "learning_rate": 4.829530900800638e-06,
+      "loss": 0.5756,
+      "step": 8211
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5550247304145242,
+      "learning_rate": 4.8284950198627186e-06,
+      "loss": 0.678,
+      "step": 8212
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.093265907839395,
+      "learning_rate": 4.827459146294751e-06,
+      "loss": 0.5029,
+      "step": 8213
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5538892821584651,
+      "learning_rate": 4.826423280141247e-06,
+      "loss": 0.5694,
+      "step": 8214
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3923569922028642,
+      "learning_rate": 4.825387421446721e-06,
+      "loss": 0.6013,
+      "step": 8215
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5246166035723965,
+      "learning_rate": 4.824351570255689e-06,
+      "loss": 0.6384,
+      "step": 8216
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6900824006168174,
+      "learning_rate": 4.823315726612659e-06,
+      "loss": 0.7909,
+      "step": 8217
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3728618958565042,
+      "learning_rate": 4.8222798905621445e-06,
+      "loss": 0.6205,
+      "step": 8218
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4059867901802678,
+      "learning_rate": 4.82124406214866e-06,
+      "loss": 0.6302,
+      "step": 8219
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7777600822628357,
+      "learning_rate": 4.820208241416719e-06,
+      "loss": 0.6825,
+      "step": 8220
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3286624798329694,
+      "learning_rate": 4.819172428410826e-06,
+      "loss": 0.6339,
+      "step": 8221
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3687247192519307,
+      "learning_rate": 4.818136623175498e-06,
+      "loss": 0.6242,
+      "step": 8222
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5404217420122317,
+      "learning_rate": 4.817100825755244e-06,
+      "loss": 0.6898,
+      "step": 8223
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6276387430554717,
+      "learning_rate": 4.816065036194576e-06,
+      "loss": 0.7546,
+      "step": 8224
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3506518015308009,
+      "learning_rate": 4.815029254538003e-06,
+      "loss": 0.6056,
+      "step": 8225
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1517139184232006,
+      "learning_rate": 4.813993480830034e-06,
+      "loss": 0.5364,
+      "step": 8226
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3615277441584517,
+      "learning_rate": 4.812957715115181e-06,
+      "loss": 0.6584,
+      "step": 8227
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2390330882321388,
+      "learning_rate": 4.8119219574379495e-06,
+      "loss": 0.6808,
+      "step": 8228
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.504340119795104,
+      "learning_rate": 4.810886207842852e-06,
+      "loss": 0.7174,
+      "step": 8229
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.477794614760091,
+      "learning_rate": 4.809850466374393e-06,
+      "loss": 0.5477,
+      "step": 8230
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4285066326960014,
+      "learning_rate": 4.8088147330770855e-06,
+      "loss": 0.6058,
+      "step": 8231
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3115656099624011,
+      "learning_rate": 4.807779007995434e-06,
+      "loss": 0.6935,
+      "step": 8232
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.612638548143309,
+      "learning_rate": 4.8067432911739455e-06,
+      "loss": 0.6778,
+      "step": 8233
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.220874168044264,
+      "learning_rate": 4.80570758265713e-06,
+      "loss": 0.5941,
+      "step": 8234
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5076520708329837,
+      "learning_rate": 4.804671882489491e-06,
+      "loss": 0.6742,
+      "step": 8235
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3017070039284304,
+      "learning_rate": 4.803636190715536e-06,
+      "loss": 0.6396,
+      "step": 8236
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3139986276649516,
+      "learning_rate": 4.802600507379771e-06,
+      "loss": 0.5847,
+      "step": 8237
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3637400541985136,
+      "learning_rate": 4.801564832526704e-06,
+      "loss": 0.7074,
+      "step": 8238
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2040407178468717,
+      "learning_rate": 4.800529166200837e-06,
+      "loss": 0.6394,
+      "step": 8239
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6351362451039393,
+      "learning_rate": 4.7994935084466745e-06,
+      "loss": 0.6444,
+      "step": 8240
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1066863508819489,
+      "learning_rate": 4.798457859308724e-06,
+      "loss": 0.6983,
+      "step": 8241
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.464271829760359,
+      "learning_rate": 4.797422218831488e-06,
+      "loss": 0.6421,
+      "step": 8242
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2900931578558914,
+      "learning_rate": 4.796386587059471e-06,
+      "loss": 0.5774,
+      "step": 8243
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.472145205978801,
+      "learning_rate": 4.795350964037174e-06,
+      "loss": 0.6483,
+      "step": 8244
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3635117888111603,
+      "learning_rate": 4.794315349809105e-06,
+      "loss": 0.7192,
+      "step": 8245
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2358084241901657,
+      "learning_rate": 4.7932797444197604e-06,
+      "loss": 0.6091,
+      "step": 8246
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.787364313886238,
+      "learning_rate": 4.792244147913647e-06,
+      "loss": 0.7762,
+      "step": 8247
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3508382734007107,
+      "learning_rate": 4.791208560335264e-06,
+      "loss": 0.6328,
+      "step": 8248
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4161473662264503,
+      "learning_rate": 4.790172981729116e-06,
+      "loss": 0.6614,
+      "step": 8249
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3949145710401867,
+      "learning_rate": 4.789137412139701e-06,
+      "loss": 0.6331,
+      "step": 8250
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.453069720577004,
+      "learning_rate": 4.7881018516115205e-06,
+      "loss": 0.6852,
+      "step": 8251
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.033357028652473,
+      "learning_rate": 4.787066300189077e-06,
+      "loss": 0.6638,
+      "step": 8252
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7353881035265137,
+      "learning_rate": 4.786030757916868e-06,
+      "loss": 0.6285,
+      "step": 8253
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3228379227193892,
+      "learning_rate": 4.784995224839394e-06,
+      "loss": 0.6987,
+      "step": 8254
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2688385333444037,
+      "learning_rate": 4.783959701001153e-06,
+      "loss": 0.6095,
+      "step": 8255
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4411494401835383,
+      "learning_rate": 4.782924186446647e-06,
+      "loss": 0.5727,
+      "step": 8256
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2193127523177814,
+      "learning_rate": 4.7818886812203715e-06,
+      "loss": 0.5917,
+      "step": 8257
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.29774859854206,
+      "learning_rate": 4.780853185366823e-06,
+      "loss": 0.5676,
+      "step": 8258
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5388127477587652,
+      "learning_rate": 4.779817698930502e-06,
+      "loss": 0.6246,
+      "step": 8259
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4649100602636127,
+      "learning_rate": 4.778782221955907e-06,
+      "loss": 0.7134,
+      "step": 8260
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.536166257885664,
+      "learning_rate": 4.77774675448753e-06,
+      "loss": 0.647,
+      "step": 8261
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2681947809320613,
+      "learning_rate": 4.776711296569872e-06,
+      "loss": 0.5511,
+      "step": 8262
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2463619491557574,
+      "learning_rate": 4.775675848247427e-06,
+      "loss": 0.5266,
+      "step": 8263
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.665452615986885,
+      "learning_rate": 4.774640409564688e-06,
+      "loss": 0.7971,
+      "step": 8264
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8130539055822523,
+      "learning_rate": 4.773604980566154e-06,
+      "loss": 0.6198,
+      "step": 8265
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4263100805032969,
+      "learning_rate": 4.772569561296318e-06,
+      "loss": 0.6502,
+      "step": 8266
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1887161252959306,
+      "learning_rate": 4.771534151799676e-06,
+      "loss": 0.5265,
+      "step": 8267
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.267513487215964,
+      "learning_rate": 4.770498752120718e-06,
+      "loss": 0.6383,
+      "step": 8268
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.527682815499234,
+      "learning_rate": 4.76946336230394e-06,
+      "loss": 0.6044,
+      "step": 8269
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5632998561924825,
+      "learning_rate": 4.768427982393836e-06,
+      "loss": 0.7216,
+      "step": 8270
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3433589830612727,
+      "learning_rate": 4.767392612434897e-06,
+      "loss": 0.5568,
+      "step": 8271
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2122312622931861,
+      "learning_rate": 4.766357252471615e-06,
+      "loss": 0.5345,
+      "step": 8272
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.528105618418981,
+      "learning_rate": 4.765321902548482e-06,
+      "loss": 0.6494,
+      "step": 8273
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4308933502397774,
+      "learning_rate": 4.76428656270999e-06,
+      "loss": 0.6562,
+      "step": 8274
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4728693277879839,
+      "learning_rate": 4.7632512330006296e-06,
+      "loss": 0.7518,
+      "step": 8275
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3576657875138711,
+      "learning_rate": 4.76221591346489e-06,
+      "loss": 0.6607,
+      "step": 8276
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.276746909120404,
+      "learning_rate": 4.761180604147262e-06,
+      "loss": 0.6902,
+      "step": 8277
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.524438205300818,
+      "learning_rate": 4.760145305092238e-06,
+      "loss": 0.5857,
+      "step": 8278
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0577446790823433,
+      "learning_rate": 4.759110016344302e-06,
+      "loss": 0.5031,
+      "step": 8279
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0772026571852518,
+      "learning_rate": 4.7580747379479455e-06,
+      "loss": 0.7513,
+      "step": 8280
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3270645108412555,
+      "learning_rate": 4.757039469947658e-06,
+      "loss": 0.5656,
+      "step": 8281
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3769384074794193,
+      "learning_rate": 4.756004212387923e-06,
+      "loss": 0.6946,
+      "step": 8282
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.2992002938573304,
+      "learning_rate": 4.7549689653132304e-06,
+      "loss": 0.6783,
+      "step": 8283
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5584100007960218,
+      "learning_rate": 4.753933728768069e-06,
+      "loss": 0.7131,
+      "step": 8284
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2526544357171143,
+      "learning_rate": 4.752898502796922e-06,
+      "loss": 0.5998,
+      "step": 8285
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0722967156771461,
+      "learning_rate": 4.751863287444278e-06,
+      "loss": 0.6186,
+      "step": 8286
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2071779078140292,
+      "learning_rate": 4.750828082754619e-06,
+      "loss": 0.6908,
+      "step": 8287
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1533484724117793,
+      "learning_rate": 4.7497928887724325e-06,
+      "loss": 0.5487,
+      "step": 8288
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2855482836815617,
+      "learning_rate": 4.748757705542205e-06,
+      "loss": 0.5915,
+      "step": 8289
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.260388341253322,
+      "learning_rate": 4.7477225331084174e-06,
+      "loss": 0.7209,
+      "step": 8290
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2388438470706344,
+      "learning_rate": 4.746687371515554e-06,
+      "loss": 0.5148,
+      "step": 8291
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2961202304276627,
+      "learning_rate": 4.7456522208081e-06,
+      "loss": 0.6639,
+      "step": 8292
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.50424362736603,
+      "learning_rate": 4.7446170810305346e-06,
+      "loss": 0.6804,
+      "step": 8293
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3830818143129548,
+      "learning_rate": 4.743581952227342e-06,
+      "loss": 0.6938,
+      "step": 8294
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3787549477341405,
+      "learning_rate": 4.7425468344430035e-06,
+      "loss": 0.5684,
+      "step": 8295
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4948953520636332,
+      "learning_rate": 4.741511727722003e-06,
+      "loss": 0.6816,
+      "step": 8296
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.42882706980322,
+      "learning_rate": 4.740476632108818e-06,
+      "loss": 0.6976,
+      "step": 8297
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3664413386045855,
+      "learning_rate": 4.73944154764793e-06,
+      "loss": 0.5224,
+      "step": 8298
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5660172232008522,
+      "learning_rate": 4.73840647438382e-06,
+      "loss": 0.594,
+      "step": 8299
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.345562255592215,
+      "learning_rate": 4.737371412360966e-06,
+      "loss": 0.5521,
+      "step": 8300
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9728360976679433,
+      "learning_rate": 4.7363363616238465e-06,
+      "loss": 0.5283,
+      "step": 8301
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9705643516492826,
+      "learning_rate": 4.73530132221694e-06,
+      "loss": 0.7734,
+      "step": 8302
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2561115699208594,
+      "learning_rate": 4.734266294184728e-06,
+      "loss": 0.5674,
+      "step": 8303
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3987870891333427,
+      "learning_rate": 4.733231277571683e-06,
+      "loss": 0.6392,
+      "step": 8304
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1715705586130052,
+      "learning_rate": 4.732196272422285e-06,
+      "loss": 0.6511,
+      "step": 8305
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.430215237255479,
+      "learning_rate": 4.731161278781008e-06,
+      "loss": 0.7852,
+      "step": 8306
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2193956777965793,
+      "learning_rate": 4.730126296692332e-06,
+      "loss": 0.6176,
+      "step": 8307
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4776004768406734,
+      "learning_rate": 4.72909132620073e-06,
+      "loss": 0.6778,
+      "step": 8308
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.333897162328494,
+      "learning_rate": 4.7280563673506745e-06,
+      "loss": 0.7201,
+      "step": 8309
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5811165897542345,
+      "learning_rate": 4.727021420186646e-06,
+      "loss": 0.7064,
+      "step": 8310
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.244944728082174,
+      "learning_rate": 4.725986484753112e-06,
+      "loss": 0.6276,
+      "step": 8311
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4498662102167363,
+      "learning_rate": 4.72495156109455e-06,
+      "loss": 0.6398,
+      "step": 8312
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4888634986550902,
+      "learning_rate": 4.723916649255432e-06,
+      "loss": 0.6927,
+      "step": 8313
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6586724089290132,
+      "learning_rate": 4.722881749280232e-06,
+      "loss": 0.644,
+      "step": 8314
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.601982040344599,
+      "learning_rate": 4.7218468612134175e-06,
+      "loss": 0.5803,
+      "step": 8315
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.194693863411821,
+      "learning_rate": 4.720811985099464e-06,
+      "loss": 0.615,
+      "step": 8316
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8067828168654314,
+      "learning_rate": 4.719777120982843e-06,
+      "loss": 0.6106,
+      "step": 8317
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.307837071230348,
+      "learning_rate": 4.718742268908022e-06,
+      "loss": 0.5949,
+      "step": 8318
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4126554174288064,
+      "learning_rate": 4.717707428919471e-06,
+      "loss": 0.6896,
+      "step": 8319
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5953523929956592,
+      "learning_rate": 4.716672601061661e-06,
+      "loss": 0.6601,
+      "step": 8320
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3148370097174795,
+      "learning_rate": 4.715637785379062e-06,
+      "loss": 0.6734,
+      "step": 8321
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4023474037843742,
+      "learning_rate": 4.714602981916139e-06,
+      "loss": 0.6283,
+      "step": 8322
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.366378185550712,
+      "learning_rate": 4.713568190717362e-06,
+      "loss": 0.7059,
+      "step": 8323
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.110859805227456,
+      "learning_rate": 4.712533411827197e-06,
+      "loss": 0.6208,
+      "step": 8324
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4406774348848828,
+      "learning_rate": 4.711498645290113e-06,
+      "loss": 0.6875,
+      "step": 8325
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.225493375125535,
+      "learning_rate": 4.710463891150573e-06,
+      "loss": 0.7073,
+      "step": 8326
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4536164250098969,
+      "learning_rate": 4.709429149453046e-06,
+      "loss": 0.5236,
+      "step": 8327
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.256472927402162,
+      "learning_rate": 4.708394420241996e-06,
+      "loss": 0.674,
+      "step": 8328
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.230398878972196,
+      "learning_rate": 4.707359703561885e-06,
+      "loss": 0.6089,
+      "step": 8329
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2868427419659618,
+      "learning_rate": 4.70632499945718e-06,
+      "loss": 0.5889,
+      "step": 8330
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.301347944251827,
+      "learning_rate": 4.705290307972344e-06,
+      "loss": 0.6601,
+      "step": 8331
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1378475493136424,
+      "learning_rate": 4.7042556291518415e-06,
+      "loss": 0.6219,
+      "step": 8332
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3578569109309353,
+      "learning_rate": 4.703220963040131e-06,
+      "loss": 0.7224,
+      "step": 8333
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8894297399699853,
+      "learning_rate": 4.702186309681677e-06,
+      "loss": 0.7706,
+      "step": 8334
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4077084799658117,
+      "learning_rate": 4.701151669120942e-06,
+      "loss": 0.5992,
+      "step": 8335
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3982037748892429,
+      "learning_rate": 4.700117041402384e-06,
+      "loss": 0.7325,
+      "step": 8336
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3230815381852112,
+      "learning_rate": 4.699082426570465e-06,
+      "loss": 0.6048,
+      "step": 8337
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.246521574857396,
+      "learning_rate": 4.6980478246696435e-06,
+      "loss": 0.5701,
+      "step": 8338
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1626439728375086,
+      "learning_rate": 4.697013235744382e-06,
+      "loss": 0.5092,
+      "step": 8339
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3214349616542922,
+      "learning_rate": 4.695978659839133e-06,
+      "loss": 0.7281,
+      "step": 8340
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.364852005932907,
+      "learning_rate": 4.694944096998361e-06,
+      "loss": 0.6073,
+      "step": 8341
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3015705804346565,
+      "learning_rate": 4.693909547266518e-06,
+      "loss": 0.5623,
+      "step": 8342
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4267117877337485,
+      "learning_rate": 4.692875010688066e-06,
+      "loss": 0.6425,
+      "step": 8343
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2142046201710959,
+      "learning_rate": 4.6918404873074574e-06,
+      "loss": 0.6096,
+      "step": 8344
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.100304603898331,
+      "learning_rate": 4.69080597716915e-06,
+      "loss": 0.7909,
+      "step": 8345
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.27421260047085,
+      "learning_rate": 4.6897714803175995e-06,
+      "loss": 0.5834,
+      "step": 8346
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3496377562254158,
+      "learning_rate": 4.688736996797257e-06,
+      "loss": 0.5495,
+      "step": 8347
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5730624015990402,
+      "learning_rate": 4.687702526652579e-06,
+      "loss": 0.6966,
+      "step": 8348
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4088426219550545,
+      "learning_rate": 4.68666806992802e-06,
+      "loss": 0.7409,
+      "step": 8349
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8017700257221982,
+      "learning_rate": 4.685633626668032e-06,
+      "loss": 0.5455,
+      "step": 8350
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2087998839444347,
+      "learning_rate": 4.684599196917067e-06,
+      "loss": 0.6962,
+      "step": 8351
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4987518836552012,
+      "learning_rate": 4.683564780719576e-06,
+      "loss": 0.6743,
+      "step": 8352
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.532514984662589,
+      "learning_rate": 4.682530378120014e-06,
+      "loss": 0.5736,
+      "step": 8353
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.239821249696437,
+      "learning_rate": 4.681495989162826e-06,
+      "loss": 0.6068,
+      "step": 8354
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5391713284790096,
+      "learning_rate": 4.680461613892465e-06,
+      "loss": 0.6942,
+      "step": 8355
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1296485664479252,
+      "learning_rate": 4.679427252353379e-06,
+      "loss": 0.5293,
+      "step": 8356
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2459519498851517,
+      "learning_rate": 4.678392904590021e-06,
+      "loss": 0.5621,
+      "step": 8357
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.320938085136134,
+      "learning_rate": 4.677358570646834e-06,
+      "loss": 0.6238,
+      "step": 8358
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5251312291602293,
+      "learning_rate": 4.676324250568269e-06,
+      "loss": 0.7895,
+      "step": 8359
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0281495419183184,
+      "learning_rate": 4.6752899443987694e-06,
+      "loss": 0.579,
+      "step": 8360
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.541788642604086,
+      "learning_rate": 4.674255652182788e-06,
+      "loss": 0.7753,
+      "step": 8361
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.49915525737044,
+      "learning_rate": 4.673221373964764e-06,
+      "loss": 0.5625,
+      "step": 8362
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7144403945247966,
+      "learning_rate": 4.672187109789144e-06,
+      "loss": 0.5935,
+      "step": 8363
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1306040596371119,
+      "learning_rate": 4.671152859700377e-06,
+      "loss": 0.6251,
+      "step": 8364
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3155540299494368,
+      "learning_rate": 4.670118623742904e-06,
+      "loss": 0.7362,
+      "step": 8365
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1656433094033254,
+      "learning_rate": 4.669084401961166e-06,
+      "loss": 0.5651,
+      "step": 8366
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2158645481090924,
+      "learning_rate": 4.668050194399609e-06,
+      "loss": 0.674,
+      "step": 8367
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.325823358475401,
+      "learning_rate": 4.667016001102675e-06,
+      "loss": 0.6562,
+      "step": 8368
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.746436890004541,
+      "learning_rate": 4.665981822114805e-06,
+      "loss": 0.7498,
+      "step": 8369
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6915753258844657,
+      "learning_rate": 4.66494765748044e-06,
+      "loss": 0.6504,
+      "step": 8370
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5369081991944393,
+      "learning_rate": 4.6639135072440195e-06,
+      "loss": 0.6301,
+      "step": 8371
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2202504671406038,
+      "learning_rate": 4.662879371449987e-06,
+      "loss": 0.5664,
+      "step": 8372
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.231255958088168,
+      "learning_rate": 4.6618452501427755e-06,
+      "loss": 0.7123,
+      "step": 8373
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5824775425636435,
+      "learning_rate": 4.660811143366828e-06,
+      "loss": 0.7041,
+      "step": 8374
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7362284037549067,
+      "learning_rate": 4.6597770511665814e-06,
+      "loss": 0.6681,
+      "step": 8375
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.321259092445271,
+      "learning_rate": 4.658742973586471e-06,
+      "loss": 0.6007,
+      "step": 8376
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0060665380962432,
+      "learning_rate": 4.657708910670936e-06,
+      "loss": 0.6586,
+      "step": 8377
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1310579592480108,
+      "learning_rate": 4.656674862464412e-06,
+      "loss": 0.5103,
+      "step": 8378
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3885635174090156,
+      "learning_rate": 4.655640829011335e-06,
+      "loss": 0.7343,
+      "step": 8379
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3836509300847442,
+      "learning_rate": 4.654606810356135e-06,
+      "loss": 0.7176,
+      "step": 8380
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1369134961577378,
+      "learning_rate": 4.653572806543251e-06,
+      "loss": 0.4774,
+      "step": 8381
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.169176542598694,
+      "learning_rate": 4.652538817617117e-06,
+      "loss": 0.598,
+      "step": 8382
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1091428914758685,
+      "learning_rate": 4.651504843622163e-06,
+      "loss": 0.7109,
+      "step": 8383
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1813906344269407,
+      "learning_rate": 4.65047088460282e-06,
+      "loss": 0.5906,
+      "step": 8384
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4136526314125903,
+      "learning_rate": 4.6494369406035225e-06,
+      "loss": 0.7036,
+      "step": 8385
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5790032431392802,
+      "learning_rate": 4.6484030116687014e-06,
+      "loss": 0.6411,
+      "step": 8386
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1574413482865908,
+      "learning_rate": 4.647369097842785e-06,
+      "loss": 0.5616,
+      "step": 8387
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3369277120790208,
+      "learning_rate": 4.646335199170205e-06,
+      "loss": 0.6503,
+      "step": 8388
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.43348101828361,
+      "learning_rate": 4.645301315695387e-06,
+      "loss": 0.6644,
+      "step": 8389
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2969472032231864,
+      "learning_rate": 4.6442674474627645e-06,
+      "loss": 0.5963,
+      "step": 8390
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.139400811011833,
+      "learning_rate": 4.643233594516759e-06,
+      "loss": 0.6114,
+      "step": 8391
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.171301220192883,
+      "learning_rate": 4.642199756901802e-06,
+      "loss": 0.6569,
+      "step": 8392
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2288044807645415,
+      "learning_rate": 4.64116593466232e-06,
+      "loss": 0.6137,
+      "step": 8393
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4088238264538566,
+      "learning_rate": 4.6401321278427334e-06,
+      "loss": 0.5648,
+      "step": 8394
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3378347522871281,
+      "learning_rate": 4.639098336487472e-06,
+      "loss": 0.7237,
+      "step": 8395
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7790280024108833,
+      "learning_rate": 4.638064560640959e-06,
+      "loss": 0.725,
+      "step": 8396
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5539080826690854,
+      "learning_rate": 4.63703080034762e-06,
+      "loss": 0.6792,
+      "step": 8397
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.440197513090281,
+      "learning_rate": 4.635997055651873e-06,
+      "loss": 0.5377,
+      "step": 8398
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5926633364740694,
+      "learning_rate": 4.634963326598143e-06,
+      "loss": 0.7402,
+      "step": 8399
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3677246417387112,
+      "learning_rate": 4.633929613230855e-06,
+      "loss": 0.7304,
+      "step": 8400
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4530853795574528,
+      "learning_rate": 4.632895915594424e-06,
+      "loss": 0.692,
+      "step": 8401
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1092278038725047,
+      "learning_rate": 4.631862233733274e-06,
+      "loss": 0.6294,
+      "step": 8402
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.193687789167538,
+      "learning_rate": 4.630828567691823e-06,
+      "loss": 0.6442,
+      "step": 8403
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.218024556867895,
+      "learning_rate": 4.629794917514492e-06,
+      "loss": 0.6975,
+      "step": 8404
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2044261238382932,
+      "learning_rate": 4.628761283245697e-06,
+      "loss": 0.6491,
+      "step": 8405
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2923301862922498,
+      "learning_rate": 4.627727664929856e-06,
+      "loss": 0.5299,
+      "step": 8406
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4149187649883244,
+      "learning_rate": 4.626694062611387e-06,
+      "loss": 0.6461,
+      "step": 8407
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2397239881528455,
+      "learning_rate": 4.6256604763347066e-06,
+      "loss": 0.5936,
+      "step": 8408
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2390375021256665,
+      "learning_rate": 4.624626906144227e-06,
+      "loss": 0.6843,
+      "step": 8409
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4518018168006785,
+      "learning_rate": 4.623593352084367e-06,
+      "loss": 0.6785,
+      "step": 8410
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.425307635541342,
+      "learning_rate": 4.622559814199538e-06,
+      "loss": 0.7214,
+      "step": 8411
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3703569950413828,
+      "learning_rate": 4.621526292534156e-06,
+      "loss": 0.7069,
+      "step": 8412
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4134534807636416,
+      "learning_rate": 4.6204927871326295e-06,
+      "loss": 0.6464,
+      "step": 8413
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2833172668286055,
+      "learning_rate": 4.619459298039373e-06,
+      "loss": 0.7601,
+      "step": 8414
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2008766703176643,
+      "learning_rate": 4.6184258252988016e-06,
+      "loss": 0.654,
+      "step": 8415
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4707575833513828,
+      "learning_rate": 4.6173923689553205e-06,
+      "loss": 0.5371,
+      "step": 8416
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0784219767308614,
+      "learning_rate": 4.616358929053339e-06,
+      "loss": 0.6415,
+      "step": 8417
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3586124567980478,
+      "learning_rate": 4.615325505637272e-06,
+      "loss": 0.6358,
+      "step": 8418
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5989396866028964,
+      "learning_rate": 4.614292098751524e-06,
+      "loss": 0.6056,
+      "step": 8419
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2301932802803026,
+      "learning_rate": 4.613258708440503e-06,
+      "loss": 0.6683,
+      "step": 8420
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0113790361190989,
+      "learning_rate": 4.612225334748616e-06,
+      "loss": 0.6824,
+      "step": 8421
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5055545883127268,
+      "learning_rate": 4.611191977720272e-06,
+      "loss": 0.6955,
+      "step": 8422
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3491884636101774,
+      "learning_rate": 4.610158637399872e-06,
+      "loss": 0.6411,
+      "step": 8423
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.393386771910131,
+      "learning_rate": 4.609125313831826e-06,
+      "loss": 0.6701,
+      "step": 8424
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2600870342271984,
+      "learning_rate": 4.608092007060533e-06,
+      "loss": 0.5886,
+      "step": 8425
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4894026383874475,
+      "learning_rate": 4.607058717130403e-06,
+      "loss": 0.6387,
+      "step": 8426
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4381930938179612,
+      "learning_rate": 4.6060254440858315e-06,
+      "loss": 0.6367,
+      "step": 8427
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.500945793786105,
+      "learning_rate": 4.6049921879712254e-06,
+      "loss": 0.6897,
+      "step": 8428
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3497386596966876,
+      "learning_rate": 4.603958948830985e-06,
+      "loss": 0.712,
+      "step": 8429
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3629851934199693,
+      "learning_rate": 4.602925726709512e-06,
+      "loss": 0.651,
+      "step": 8430
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3988398037791676,
+      "learning_rate": 4.6018925216512025e-06,
+      "loss": 0.5748,
+      "step": 8431
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4494505574669372,
+      "learning_rate": 4.600859333700457e-06,
+      "loss": 0.6775,
+      "step": 8432
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4395405090130535,
+      "learning_rate": 4.599826162901679e-06,
+      "loss": 0.5525,
+      "step": 8433
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3628703735745162,
+      "learning_rate": 4.5987930092992596e-06,
+      "loss": 0.6194,
+      "step": 8434
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3657991159099283,
+      "learning_rate": 4.597759872937597e-06,
+      "loss": 0.7485,
+      "step": 8435
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5205629514770296,
+      "learning_rate": 4.5967267538610915e-06,
+      "loss": 0.868,
+      "step": 8436
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1372659613577125,
+      "learning_rate": 4.595693652114133e-06,
+      "loss": 0.5595,
+      "step": 8437
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3048651688520319,
+      "learning_rate": 4.594660567741118e-06,
+      "loss": 0.5687,
+      "step": 8438
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3097483030042245,
+      "learning_rate": 4.593627500786444e-06,
+      "loss": 0.6224,
+      "step": 8439
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.287693464330295,
+      "learning_rate": 4.592594451294501e-06,
+      "loss": 0.5305,
+      "step": 8440
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.129016201057602,
+      "learning_rate": 4.59156141930968e-06,
+      "loss": 0.5363,
+      "step": 8441
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.40512715843003,
+      "learning_rate": 4.590528404876374e-06,
+      "loss": 0.657,
+      "step": 8442
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7450067789457073,
+      "learning_rate": 4.5894954080389755e-06,
+      "loss": 0.6769,
+      "step": 8443
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3544969295666096,
+      "learning_rate": 4.588462428841875e-06,
+      "loss": 0.6602,
+      "step": 8444
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4461658752801925,
+      "learning_rate": 4.587429467329458e-06,
+      "loss": 0.6544,
+      "step": 8445
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3775160464573253,
+      "learning_rate": 4.586396523546116e-06,
+      "loss": 0.6326,
+      "step": 8446
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.518753832148509,
+      "learning_rate": 4.585363597536239e-06,
+      "loss": 0.6957,
+      "step": 8447
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3752593021156485,
+      "learning_rate": 4.584330689344211e-06,
+      "loss": 0.6027,
+      "step": 8448
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3544672950630545,
+      "learning_rate": 4.5832977990144165e-06,
+      "loss": 0.6754,
+      "step": 8449
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4472719441643445,
+      "learning_rate": 4.582264926591244e-06,
+      "loss": 0.8268,
+      "step": 8450
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.401362484955057,
+      "learning_rate": 4.581232072119081e-06,
+      "loss": 0.6301,
+      "step": 8451
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3593899700924201,
+      "learning_rate": 4.580199235642306e-06,
+      "loss": 0.5909,
+      "step": 8452
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4119596895299844,
+      "learning_rate": 4.5791664172053044e-06,
+      "loss": 0.7102,
+      "step": 8453
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2441588701716377,
+      "learning_rate": 4.578133616852462e-06,
+      "loss": 0.624,
+      "step": 8454
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2168839291963556,
+      "learning_rate": 4.577100834628155e-06,
+      "loss": 0.5199,
+      "step": 8455
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3787075652772345,
+      "learning_rate": 4.5760680705767665e-06,
+      "loss": 0.5411,
+      "step": 8456
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3756227764259796,
+      "learning_rate": 4.5750353247426785e-06,
+      "loss": 0.6928,
+      "step": 8457
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7255093812334648,
+      "learning_rate": 4.5740025971702695e-06,
+      "loss": 0.7023,
+      "step": 8458
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3386111206361984,
+      "learning_rate": 4.572969887903916e-06,
+      "loss": 0.5827,
+      "step": 8459
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.342116041000349,
+      "learning_rate": 4.571937196987998e-06,
+      "loss": 0.6348,
+      "step": 8460
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5239715953481587,
+      "learning_rate": 4.570904524466893e-06,
+      "loss": 0.6812,
+      "step": 8461
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4762510032015284,
+      "learning_rate": 4.5698718703849755e-06,
+      "loss": 0.6916,
+      "step": 8462
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0615508252220676,
+      "learning_rate": 4.5688392347866226e-06,
+      "loss": 0.6271,
+      "step": 8463
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3886513021152016,
+      "learning_rate": 4.5678066177162065e-06,
+      "loss": 0.6637,
+      "step": 8464
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.224483973262663,
+      "learning_rate": 4.566774019218104e-06,
+      "loss": 0.7258,
+      "step": 8465
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3367852835938165,
+      "learning_rate": 4.565741439336686e-06,
+      "loss": 0.6763,
+      "step": 8466
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4739703369203232,
+      "learning_rate": 4.5647088781163255e-06,
+      "loss": 0.665,
+      "step": 8467
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7887930523101097,
+      "learning_rate": 4.563676335601393e-06,
+      "loss": 0.648,
+      "step": 8468
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4369548963262768,
+      "learning_rate": 4.562643811836263e-06,
+      "loss": 0.7021,
+      "step": 8469
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3812155689611902,
+      "learning_rate": 4.561611306865299e-06,
+      "loss": 0.6286,
+      "step": 8470
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.470011769196712,
+      "learning_rate": 4.560578820732876e-06,
+      "loss": 0.6773,
+      "step": 8471
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6152244079400953,
+      "learning_rate": 4.559546353483359e-06,
+      "loss": 0.6496,
+      "step": 8472
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5006321266754805,
+      "learning_rate": 4.558513905161116e-06,
+      "loss": 0.6722,
+      "step": 8473
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4052550052504307,
+      "learning_rate": 4.557481475810512e-06,
+      "loss": 0.7361,
+      "step": 8474
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.325892544478158,
+      "learning_rate": 4.556449065475917e-06,
+      "loss": 0.611,
+      "step": 8475
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0297716519902873,
+      "learning_rate": 4.555416674201693e-06,
+      "loss": 0.6306,
+      "step": 8476
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4258601727033569,
+      "learning_rate": 4.554384302032204e-06,
+      "loss": 0.7829,
+      "step": 8477
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.133205589328371,
+      "learning_rate": 4.553351949011814e-06,
+      "loss": 0.8132,
+      "step": 8478
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3813451687332046,
+      "learning_rate": 4.5523196151848846e-06,
+      "loss": 0.6689,
+      "step": 8479
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.154289859383127,
+      "learning_rate": 4.551287300595781e-06,
+      "loss": 0.6687,
+      "step": 8480
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3251699877657612,
+      "learning_rate": 4.550255005288861e-06,
+      "loss": 0.6538,
+      "step": 8481
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.558178441700343,
+      "learning_rate": 4.549222729308483e-06,
+      "loss": 0.7138,
+      "step": 8482
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4193449390206716,
+      "learning_rate": 4.548190472699011e-06,
+      "loss": 0.7282,
+      "step": 8483
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2077968303730549,
+      "learning_rate": 4.547158235504797e-06,
+      "loss": 0.6131,
+      "step": 8484
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1712867832954337,
+      "learning_rate": 4.546126017770205e-06,
+      "loss": 0.6358,
+      "step": 8485
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3457180464340874,
+      "learning_rate": 4.5450938195395875e-06,
+      "loss": 0.5888,
+      "step": 8486
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2125153791129117,
+      "learning_rate": 4.544061640857303e-06,
+      "loss": 0.7046,
+      "step": 8487
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7734587570074731,
+      "learning_rate": 4.543029481767703e-06,
+      "loss": 0.6291,
+      "step": 8488
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0752775620447714,
+      "learning_rate": 4.541997342315145e-06,
+      "loss": 0.5436,
+      "step": 8489
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4904337505792264,
+      "learning_rate": 4.540965222543981e-06,
+      "loss": 0.7503,
+      "step": 8490
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2165757446730832,
+      "learning_rate": 4.539933122498566e-06,
+      "loss": 0.6735,
+      "step": 8491
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4149964820839538,
+      "learning_rate": 4.5389010422232474e-06,
+      "loss": 0.6047,
+      "step": 8492
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.392477020590218,
+      "learning_rate": 4.5378689817623765e-06,
+      "loss": 0.6811,
+      "step": 8493
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.747943617208049,
+      "learning_rate": 4.536836941160308e-06,
+      "loss": 0.6729,
+      "step": 8494
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.409721100913299,
+      "learning_rate": 4.535804920461386e-06,
+      "loss": 0.6849,
+      "step": 8495
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5138543137515068,
+      "learning_rate": 4.53477291970996e-06,
+      "loss": 0.7865,
+      "step": 8496
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6970757024538723,
+      "learning_rate": 4.5337409389503764e-06,
+      "loss": 0.631,
+      "step": 8497
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1695499509977447,
+      "learning_rate": 4.532708978226987e-06,
+      "loss": 0.6474,
+      "step": 8498
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2858163890522178,
+      "learning_rate": 4.5316770375841315e-06,
+      "loss": 0.7441,
+      "step": 8499
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2447286263065933,
+      "learning_rate": 4.530645117066155e-06,
+      "loss": 0.5564,
+      "step": 8500
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.609148238789212,
+      "learning_rate": 4.529613216717406e-06,
+      "loss": 0.6701,
+      "step": 8501
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3386350940741072,
+      "learning_rate": 4.528581336582223e-06,
+      "loss": 0.5808,
+      "step": 8502
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.0382062673689054,
+      "learning_rate": 4.527549476704949e-06,
+      "loss": 0.5761,
+      "step": 8503
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3047868233680828,
+      "learning_rate": 4.526517637129927e-06,
+      "loss": 0.658,
+      "step": 8504
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4806155196264217,
+      "learning_rate": 4.525485817901499e-06,
+      "loss": 0.6246,
+      "step": 8505
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2994550361309787,
+      "learning_rate": 4.524454019063999e-06,
+      "loss": 0.6132,
+      "step": 8506
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4114721622693744,
+      "learning_rate": 4.52342224066177e-06,
+      "loss": 0.6128,
+      "step": 8507
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5453867120944427,
+      "learning_rate": 4.522390482739148e-06,
+      "loss": 0.6328,
+      "step": 8508
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.341931948461782,
+      "learning_rate": 4.5213587453404736e-06,
+      "loss": 0.7041,
+      "step": 8509
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.315421532698708,
+      "learning_rate": 4.520327028510076e-06,
+      "loss": 0.5468,
+      "step": 8510
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3917906506665716,
+      "learning_rate": 4.5192953322922955e-06,
+      "loss": 0.5893,
+      "step": 8511
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2881131193564548,
+      "learning_rate": 4.518263656731468e-06,
+      "loss": 0.643,
+      "step": 8512
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.568893338511838,
+      "learning_rate": 4.5172320018719205e-06,
+      "loss": 0.6951,
+      "step": 8513
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3540227830402234,
+      "learning_rate": 4.5162003677579905e-06,
+      "loss": 0.5581,
+      "step": 8514
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6234470662817442,
+      "learning_rate": 4.5151687544340065e-06,
+      "loss": 0.686,
+      "step": 8515
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1881623202293237,
+      "learning_rate": 4.514137161944304e-06,
+      "loss": 0.7399,
+      "step": 8516
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1562670310772591,
+      "learning_rate": 4.513105590333207e-06,
+      "loss": 0.5966,
+      "step": 8517
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5448896793085813,
+      "learning_rate": 4.512074039645049e-06,
+      "loss": 0.6758,
+      "step": 8518
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1908214020969863,
+      "learning_rate": 4.511042509924157e-06,
+      "loss": 0.5546,
+      "step": 8519
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5052316299029966,
+      "learning_rate": 4.5100110012148546e-06,
+      "loss": 0.6354,
+      "step": 8520
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2373959373345105,
+      "learning_rate": 4.508979513561471e-06,
+      "loss": 0.6879,
+      "step": 8521
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4511721780683315,
+      "learning_rate": 4.507948047008332e-06,
+      "loss": 0.7002,
+      "step": 8522
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1974909190638066,
+      "learning_rate": 4.506916601599763e-06,
+      "loss": 0.5979,
+      "step": 8523
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2932968652779688,
+      "learning_rate": 4.505885177380083e-06,
+      "loss": 0.6339,
+      "step": 8524
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2347158588187337,
+      "learning_rate": 4.504853774393618e-06,
+      "loss": 0.6841,
+      "step": 8525
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4567653520840604,
+      "learning_rate": 4.5038223926846905e-06,
+      "loss": 0.7721,
+      "step": 8526
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5001122366444681,
+      "learning_rate": 4.5027910322976186e-06,
+      "loss": 0.6533,
+      "step": 8527
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3498300439814654,
+      "learning_rate": 4.501759693276724e-06,
+      "loss": 0.6468,
+      "step": 8528
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2816471889594383,
+      "learning_rate": 4.5007283756663245e-06,
+      "loss": 0.8095,
+      "step": 8529
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.372207127048134,
+      "learning_rate": 4.49969707951074e-06,
+      "loss": 0.6801,
+      "step": 8530
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1514916586417474,
+      "learning_rate": 4.498665804854285e-06,
+      "loss": 0.649,
+      "step": 8531
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1492261924393163,
+      "learning_rate": 4.497634551741277e-06,
+      "loss": 0.6651,
+      "step": 8532
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4064021966933393,
+      "learning_rate": 4.49660332021603e-06,
+      "loss": 0.5861,
+      "step": 8533
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3932173291808834,
+      "learning_rate": 4.495572110322862e-06,
+      "loss": 0.6428,
+      "step": 8534
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2212756018952697,
+      "learning_rate": 4.494540922106082e-06,
+      "loss": 0.7079,
+      "step": 8535
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3842762886046904,
+      "learning_rate": 4.4935097556100045e-06,
+      "loss": 0.6512,
+      "step": 8536
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2403202042529766,
+      "learning_rate": 4.492478610878942e-06,
+      "loss": 0.5877,
+      "step": 8537
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3397406575475856,
+      "learning_rate": 4.491447487957203e-06,
+      "loss": 0.7317,
+      "step": 8538
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.445385937726481,
+      "learning_rate": 4.490416386889097e-06,
+      "loss": 0.645,
+      "step": 8539
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9762761355139595,
+      "learning_rate": 4.489385307718934e-06,
+      "loss": 0.5389,
+      "step": 8540
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5913048338574203,
+      "learning_rate": 4.488354250491024e-06,
+      "loss": 0.6929,
+      "step": 8541
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3524909013239472,
+      "learning_rate": 4.48732321524967e-06,
+      "loss": 0.6349,
+      "step": 8542
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2545736369796794,
+      "learning_rate": 4.486292202039178e-06,
+      "loss": 0.6676,
+      "step": 8543
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5184472551815018,
+      "learning_rate": 4.485261210903854e-06,
+      "loss": 0.6422,
+      "step": 8544
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2546617350201708,
+      "learning_rate": 4.484230241888005e-06,
+      "loss": 0.5922,
+      "step": 8545
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.284556878957551,
+      "learning_rate": 4.48319929503593e-06,
+      "loss": 0.6357,
+      "step": 8546
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.203934690813163,
+      "learning_rate": 4.482168370391931e-06,
+      "loss": 0.5777,
+      "step": 8547
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2090549054793616,
+      "learning_rate": 4.481137468000312e-06,
+      "loss": 0.6826,
+      "step": 8548
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2373675541980083,
+      "learning_rate": 4.48010658790537e-06,
+      "loss": 0.7231,
+      "step": 8549
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1735999319400094,
+      "learning_rate": 4.479075730151406e-06,
+      "loss": 0.6131,
+      "step": 8550
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5758659720203405,
+      "learning_rate": 4.478044894782718e-06,
+      "loss": 0.6367,
+      "step": 8551
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.20258385142419,
+      "learning_rate": 4.477014081843605e-06,
+      "loss": 0.4848,
+      "step": 8552
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2324431020502977,
+      "learning_rate": 4.47598329137836e-06,
+      "loss": 0.6966,
+      "step": 8553
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2663536992900217,
+      "learning_rate": 4.47495252343128e-06,
+      "loss": 0.606,
+      "step": 8554
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.196933173551853,
+      "learning_rate": 4.473921778046661e-06,
+      "loss": 0.6252,
+      "step": 8555
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6641579471255212,
+      "learning_rate": 4.4728910552687935e-06,
+      "loss": 0.5793,
+      "step": 8556
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.408237572594407,
+      "learning_rate": 4.471860355141971e-06,
+      "loss": 0.682,
+      "step": 8557
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.342011940739351,
+      "learning_rate": 4.470829677710485e-06,
+      "loss": 0.5768,
+      "step": 8558
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.474854995393856,
+      "learning_rate": 4.469799023018628e-06,
+      "loss": 0.5863,
+      "step": 8559
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.27353311549484,
+      "learning_rate": 4.468768391110687e-06,
+      "loss": 0.6266,
+      "step": 8560
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.265127406927414,
+      "learning_rate": 4.467737782030951e-06,
+      "loss": 0.564,
+      "step": 8561
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4468200936140256,
+      "learning_rate": 4.466707195823707e-06,
+      "loss": 0.6819,
+      "step": 8562
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1371880184183847,
+      "learning_rate": 4.465676632533245e-06,
+      "loss": 0.5729,
+      "step": 8563
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5516247540854653,
+      "learning_rate": 4.464646092203846e-06,
+      "loss": 0.632,
+      "step": 8564
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4599108158075418,
+      "learning_rate": 4.463615574879798e-06,
+      "loss": 0.6466,
+      "step": 8565
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2097392694290794,
+      "learning_rate": 4.462585080605384e-06,
+      "loss": 0.5878,
+      "step": 8566
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1167701036170476,
+      "learning_rate": 4.461554609424884e-06,
+      "loss": 0.6403,
+      "step": 8567
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1019549823435215,
+      "learning_rate": 4.460524161382582e-06,
+      "loss": 0.6236,
+      "step": 8568
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.571778550906292,
+      "learning_rate": 4.459493736522759e-06,
+      "loss": 0.6308,
+      "step": 8569
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4987088531871486,
+      "learning_rate": 4.4584633348896945e-06,
+      "loss": 0.6943,
+      "step": 8570
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.529676267163338,
+      "learning_rate": 4.457432956527665e-06,
+      "loss": 0.6566,
+      "step": 8571
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3649694429784736,
+      "learning_rate": 4.456402601480949e-06,
+      "loss": 0.5972,
+      "step": 8572
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1889487224902735,
+      "learning_rate": 4.455372269793826e-06,
+      "loss": 0.687,
+      "step": 8573
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2880596072004658,
+      "learning_rate": 4.4543419615105685e-06,
+      "loss": 0.714,
+      "step": 8574
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5989115091193955,
+      "learning_rate": 4.453311676675453e-06,
+      "loss": 0.627,
+      "step": 8575
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4787601810625501,
+      "learning_rate": 4.452281415332751e-06,
+      "loss": 0.7012,
+      "step": 8576
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3410379673058999,
+      "learning_rate": 4.451251177526738e-06,
+      "loss": 0.5614,
+      "step": 8577
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3384624320467244,
+      "learning_rate": 4.450220963301683e-06,
+      "loss": 0.727,
+      "step": 8578
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.407819082571223,
+      "learning_rate": 4.449190772701857e-06,
+      "loss": 0.6408,
+      "step": 8579
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3109429373806598,
+      "learning_rate": 4.44816060577153e-06,
+      "loss": 0.6206,
+      "step": 8580
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.232984923281101,
+      "learning_rate": 4.447130462554974e-06,
+      "loss": 0.6584,
+      "step": 8581
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9150458095123852,
+      "learning_rate": 4.44610034309645e-06,
+      "loss": 0.6704,
+      "step": 8582
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3141239214459448,
+      "learning_rate": 4.4450702474402295e-06,
+      "loss": 0.6237,
+      "step": 8583
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.277456818324174,
+      "learning_rate": 4.444040175630577e-06,
+      "loss": 0.7749,
+      "step": 8584
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2656672530032465,
+      "learning_rate": 4.443010127711755e-06,
+      "loss": 0.6357,
+      "step": 8585
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3659876084962042,
+      "learning_rate": 4.441980103728027e-06,
+      "loss": 0.6815,
+      "step": 8586
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3979013552676383,
+      "learning_rate": 4.440950103723658e-06,
+      "loss": 0.6009,
+      "step": 8587
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6097773964570359,
+      "learning_rate": 4.439920127742909e-06,
+      "loss": 0.6369,
+      "step": 8588
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.521802989649919,
+      "learning_rate": 4.438890175830039e-06,
+      "loss": 0.6937,
+      "step": 8589
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2917901448295122,
+      "learning_rate": 4.437860248029307e-06,
+      "loss": 0.6516,
+      "step": 8590
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4150340811076587,
+      "learning_rate": 4.4368303443849735e-06,
+      "loss": 0.6816,
+      "step": 8591
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4029321048694943,
+      "learning_rate": 4.435800464941292e-06,
+      "loss": 0.6487,
+      "step": 8592
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4590502054721874,
+      "learning_rate": 4.434770609742523e-06,
+      "loss": 0.6958,
+      "step": 8593
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.276965484052214,
+      "learning_rate": 4.433740778832919e-06,
+      "loss": 0.7379,
+      "step": 8594
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4074055606177824,
+      "learning_rate": 4.432710972256737e-06,
+      "loss": 0.582,
+      "step": 8595
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.458705664593947,
+      "learning_rate": 4.431681190058224e-06,
+      "loss": 0.6821,
+      "step": 8596
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2942634378439517,
+      "learning_rate": 4.430651432281639e-06,
+      "loss": 0.6145,
+      "step": 8597
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1294168503776707,
+      "learning_rate": 4.429621698971228e-06,
+      "loss": 0.6329,
+      "step": 8598
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3323519841694276,
+      "learning_rate": 4.428591990171246e-06,
+      "loss": 0.7404,
+      "step": 8599
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2789334834695931,
+      "learning_rate": 4.4275623059259355e-06,
+      "loss": 0.6461,
+      "step": 8600
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.809477791815784,
+      "learning_rate": 4.426532646279548e-06,
+      "loss": 0.5791,
+      "step": 8601
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2270810085030996,
+      "learning_rate": 4.425503011276332e-06,
+      "loss": 0.6109,
+      "step": 8602
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3138080257791664,
+      "learning_rate": 4.42447340096053e-06,
+      "loss": 0.636,
+      "step": 8603
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5565788346930696,
+      "learning_rate": 4.423443815376387e-06,
+      "loss": 0.6089,
+      "step": 8604
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2138322951463612,
+      "learning_rate": 4.422414254568147e-06,
+      "loss": 0.6612,
+      "step": 8605
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.454582518686615,
+      "learning_rate": 4.421384718580055e-06,
+      "loss": 0.6676,
+      "step": 8606
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3295045392930824,
+      "learning_rate": 4.420355207456349e-06,
+      "loss": 0.6307,
+      "step": 8607
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4251460614781717,
+      "learning_rate": 4.4193257212412695e-06,
+      "loss": 0.5966,
+      "step": 8608
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.017757494500382,
+      "learning_rate": 4.41829625997906e-06,
+      "loss": 0.726,
+      "step": 8609
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4093172669415739,
+      "learning_rate": 4.417266823713953e-06,
+      "loss": 0.6984,
+      "step": 8610
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2226928359830591,
+      "learning_rate": 4.416237412490189e-06,
+      "loss": 0.697,
+      "step": 8611
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.79436523508238,
+      "learning_rate": 4.415208026352003e-06,
+      "loss": 0.5761,
+      "step": 8612
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4930843849448863,
+      "learning_rate": 4.414178665343633e-06,
+      "loss": 0.6471,
+      "step": 8613
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4360966345842754,
+      "learning_rate": 4.413149329509307e-06,
+      "loss": 0.649,
+      "step": 8614
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3865098061039285,
+      "learning_rate": 4.412120018893263e-06,
+      "loss": 0.6787,
+      "step": 8615
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1412118510777947,
+      "learning_rate": 4.411090733539731e-06,
+      "loss": 0.7148,
+      "step": 8616
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2258835502828025,
+      "learning_rate": 4.410061473492943e-06,
+      "loss": 0.6313,
+      "step": 8617
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4996383626912548,
+      "learning_rate": 4.409032238797125e-06,
+      "loss": 0.7091,
+      "step": 8618
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7310130925462486,
+      "learning_rate": 4.4080030294965085e-06,
+      "loss": 0.6699,
+      "step": 8619
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4382101727065262,
+      "learning_rate": 4.406973845635322e-06,
+      "loss": 0.675,
+      "step": 8620
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5322269861483662,
+      "learning_rate": 4.405944687257789e-06,
+      "loss": 0.6841,
+      "step": 8621
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5415147211445284,
+      "learning_rate": 4.4049155544081355e-06,
+      "loss": 0.8052,
+      "step": 8622
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4128782668816768,
+      "learning_rate": 4.4038864471305845e-06,
+      "loss": 0.7962,
+      "step": 8623
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.402649344842548,
+      "learning_rate": 4.402857365469364e-06,
+      "loss": 0.7491,
+      "step": 8624
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2760732174073919,
+      "learning_rate": 4.40182830946869e-06,
+      "loss": 0.5967,
+      "step": 8625
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3275823186529456,
+      "learning_rate": 4.400799279172786e-06,
+      "loss": 0.6978,
+      "step": 8626
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5533374283677541,
+      "learning_rate": 4.399770274625871e-06,
+      "loss": 0.7951,
+      "step": 8627
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.611718723867381,
+      "learning_rate": 4.3987412958721664e-06,
+      "loss": 0.6395,
+      "step": 8628
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.268935954440164,
+      "learning_rate": 4.397712342955885e-06,
+      "loss": 0.5847,
+      "step": 8629
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4519665185221937,
+      "learning_rate": 4.396683415921247e-06,
+      "loss": 0.6106,
+      "step": 8630
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.43602103958184,
+      "learning_rate": 4.3956545148124665e-06,
+      "loss": 0.521,
+      "step": 8631
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6071648367801703,
+      "learning_rate": 4.394625639673756e-06,
+      "loss": 0.6403,
+      "step": 8632
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.405950216276389,
+      "learning_rate": 4.39359679054933e-06,
+      "loss": 0.7599,
+      "step": 8633
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0764106558440494,
+      "learning_rate": 4.392567967483401e-06,
+      "loss": 0.589,
+      "step": 8634
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7318358711351394,
+      "learning_rate": 4.3915391705201805e-06,
+      "loss": 0.6896,
+      "step": 8635
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5172470610531665,
+      "learning_rate": 4.390510399703875e-06,
+      "loss": 0.6755,
+      "step": 8636
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4917999222238538,
+      "learning_rate": 4.389481655078695e-06,
+      "loss": 0.621,
+      "step": 8637
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3722310608871202,
+      "learning_rate": 4.38845293668885e-06,
+      "loss": 0.5946,
+      "step": 8638
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3641956315936665,
+      "learning_rate": 4.387424244578543e-06,
+      "loss": 0.6764,
+      "step": 8639
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5813282785657399,
+      "learning_rate": 4.386395578791981e-06,
+      "loss": 0.8163,
+      "step": 8640
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5817351642322792,
+      "learning_rate": 4.385366939373365e-06,
+      "loss": 0.6822,
+      "step": 8641
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.383994169502398,
+      "learning_rate": 4.384338326366905e-06,
+      "loss": 0.6513,
+      "step": 8642
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4210221513026509,
+      "learning_rate": 4.383309739816795e-06,
+      "loss": 0.61,
+      "step": 8643
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4385626847581894,
+      "learning_rate": 4.38228117976724e-06,
+      "loss": 0.6445,
+      "step": 8644
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0290813273278407,
+      "learning_rate": 4.381252646262437e-06,
+      "loss": 0.7353,
+      "step": 8645
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2363846205300086,
+      "learning_rate": 4.3802241393465885e-06,
+      "loss": 0.6988,
+      "step": 8646
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5358104715708938,
+      "learning_rate": 4.3791956590638866e-06,
+      "loss": 0.6847,
+      "step": 8647
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.073722186880252,
+      "learning_rate": 4.378167205458531e-06,
+      "loss": 0.4747,
+      "step": 8648
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4066062344250283,
+      "learning_rate": 4.377138778574716e-06,
+      "loss": 0.6568,
+      "step": 8649
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4326472222200868,
+      "learning_rate": 4.376110378456634e-06,
+      "loss": 0.7069,
+      "step": 8650
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3848297443946602,
+      "learning_rate": 4.375082005148479e-06,
+      "loss": 0.7765,
+      "step": 8651
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4591769000778816,
+      "learning_rate": 4.37405365869444e-06,
+      "loss": 0.7262,
+      "step": 8652
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.41616814404164,
+      "learning_rate": 4.373025339138713e-06,
+      "loss": 0.6013,
+      "step": 8653
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6561500356202496,
+      "learning_rate": 4.371997046525481e-06,
+      "loss": 0.6557,
+      "step": 8654
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.284252294078411,
+      "learning_rate": 4.370968780898935e-06,
+      "loss": 0.7586,
+      "step": 8655
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4431970852285934,
+      "learning_rate": 4.3699405423032625e-06,
+      "loss": 0.6074,
+      "step": 8656
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.492277197207456,
+      "learning_rate": 4.368912330782647e-06,
+      "loss": 0.655,
+      "step": 8657
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0310448937035428,
+      "learning_rate": 4.367884146381274e-06,
+      "loss": 0.5972,
+      "step": 8658
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.329536827153917,
+      "learning_rate": 4.366855989143326e-06,
+      "loss": 0.5819,
+      "step": 8659
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5000574028149867,
+      "learning_rate": 4.365827859112989e-06,
+      "loss": 0.6961,
+      "step": 8660
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4783621595364824,
+      "learning_rate": 4.364799756334439e-06,
+      "loss": 0.5653,
+      "step": 8661
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.229369570534812,
+      "learning_rate": 4.3637716808518596e-06,
+      "loss": 0.5312,
+      "step": 8662
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2503845911882965,
+      "learning_rate": 4.362743632709426e-06,
+      "loss": 0.5976,
+      "step": 8663
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.451310884065438,
+      "learning_rate": 4.3617156119513206e-06,
+      "loss": 0.6361,
+      "step": 8664
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.381021602918778,
+      "learning_rate": 4.360687618621715e-06,
+      "loss": 0.6895,
+      "step": 8665
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3033210010714784,
+      "learning_rate": 4.359659652764786e-06,
+      "loss": 0.617,
+      "step": 8666
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.514178080958876,
+      "learning_rate": 4.35863171442471e-06,
+      "loss": 0.68,
+      "step": 8667
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.274250109592793,
+      "learning_rate": 4.357603803645657e-06,
+      "loss": 0.5811,
+      "step": 8668
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3535109682439748,
+      "learning_rate": 4.356575920471796e-06,
+      "loss": 0.6183,
+      "step": 8669
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4918006799738404,
+      "learning_rate": 4.355548064947303e-06,
+      "loss": 0.714,
+      "step": 8670
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3151215736565978,
+      "learning_rate": 4.354520237116347e-06,
+      "loss": 0.6212,
+      "step": 8671
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2212887518625726,
+      "learning_rate": 4.353492437023092e-06,
+      "loss": 0.6082,
+      "step": 8672
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5092981590730017,
+      "learning_rate": 4.352464664711706e-06,
+      "loss": 0.6229,
+      "step": 8673
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.237543643155562,
+      "learning_rate": 4.351436920226357e-06,
+      "loss": 0.6019,
+      "step": 8674
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1432558102349168,
+      "learning_rate": 4.350409203611207e-06,
+      "loss": 0.6533,
+      "step": 8675
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1824685128068408,
+      "learning_rate": 4.34938151491042e-06,
+      "loss": 0.6616,
+      "step": 8676
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4360831154408005,
+      "learning_rate": 4.348353854168158e-06,
+      "loss": 0.5907,
+      "step": 8677
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4688380648261499,
+      "learning_rate": 4.347326221428585e-06,
+      "loss": 0.6277,
+      "step": 8678
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4745662021047206,
+      "learning_rate": 4.346298616735855e-06,
+      "loss": 0.7168,
+      "step": 8679
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.266635446959506,
+      "learning_rate": 4.345271040134129e-06,
+      "loss": 0.5084,
+      "step": 8680
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3178196076745712,
+      "learning_rate": 4.344243491667566e-06,
+      "loss": 0.6707,
+      "step": 8681
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6914229549314446,
+      "learning_rate": 4.343215971380323e-06,
+      "loss": 0.671,
+      "step": 8682
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.273286585909244,
+      "learning_rate": 4.342188479316548e-06,
+      "loss": 0.7242,
+      "step": 8683
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9495536008420438,
+      "learning_rate": 4.3411610155204e-06,
+      "loss": 0.6003,
+      "step": 8684
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4771901812960726,
+      "learning_rate": 4.340133580036033e-06,
+      "loss": 0.6602,
+      "step": 8685
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.291509223382206,
+      "learning_rate": 4.339106172907594e-06,
+      "loss": 0.7046,
+      "step": 8686
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2636703480338747,
+      "learning_rate": 4.338078794179234e-06,
+      "loss": 0.6465,
+      "step": 8687
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3262877605374523,
+      "learning_rate": 4.337051443895102e-06,
+      "loss": 0.7716,
+      "step": 8688
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2652870381303,
+      "learning_rate": 4.336024122099348e-06,
+      "loss": 0.5945,
+      "step": 8689
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4139547868017324,
+      "learning_rate": 4.334996828836115e-06,
+      "loss": 0.7557,
+      "step": 8690
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.336326914178329,
+      "learning_rate": 4.333969564149549e-06,
+      "loss": 0.7733,
+      "step": 8691
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2729059644172873,
+      "learning_rate": 4.332942328083796e-06,
+      "loss": 0.5993,
+      "step": 8692
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2008956313788155,
+      "learning_rate": 4.3319151206829955e-06,
+      "loss": 0.6287,
+      "step": 8693
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5024571975142154,
+      "learning_rate": 4.330887941991288e-06,
+      "loss": 0.6059,
+      "step": 8694
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3778817480816794,
+      "learning_rate": 4.329860792052819e-06,
+      "loss": 0.5827,
+      "step": 8695
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3837742433998699,
+      "learning_rate": 4.3288336709117246e-06,
+      "loss": 0.6448,
+      "step": 8696
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4822084700951388,
+      "learning_rate": 4.32780657861214e-06,
+      "loss": 0.6875,
+      "step": 8697
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.384908644730678,
+      "learning_rate": 4.326779515198203e-06,
+      "loss": 0.6478,
+      "step": 8698
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.531463335588062,
+      "learning_rate": 4.325752480714052e-06,
+      "loss": 0.6287,
+      "step": 8699
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6363296041156563,
+      "learning_rate": 4.324725475203818e-06,
+      "loss": 0.8415,
+      "step": 8700
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4101411767565264,
+      "learning_rate": 4.323698498711634e-06,
+      "loss": 0.6045,
+      "step": 8701
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3502287661363093,
+      "learning_rate": 4.32267155128163e-06,
+      "loss": 0.6933,
+      "step": 8702
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4878288374609334,
+      "learning_rate": 4.321644632957941e-06,
+      "loss": 0.6778,
+      "step": 8703
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7010371002124383,
+      "learning_rate": 4.320617743784691e-06,
+      "loss": 0.7733,
+      "step": 8704
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3118518847636484,
+      "learning_rate": 4.31959088380601e-06,
+      "loss": 0.6501,
+      "step": 8705
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3484311914658975,
+      "learning_rate": 4.318564053066023e-06,
+      "loss": 0.6178,
+      "step": 8706
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3374500275179577,
+      "learning_rate": 4.317537251608859e-06,
+      "loss": 0.5617,
+      "step": 8707
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2184765001936508,
+      "learning_rate": 4.316510479478636e-06,
+      "loss": 0.7142,
+      "step": 8708
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5561408867864914,
+      "learning_rate": 4.315483736719482e-06,
+      "loss": 0.7615,
+      "step": 8709
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2837268290824657,
+      "learning_rate": 4.314457023375517e-06,
+      "loss": 0.6211,
+      "step": 8710
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3754293469827727,
+      "learning_rate": 4.313430339490859e-06,
+      "loss": 0.6405,
+      "step": 8711
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0722230862431583,
+      "learning_rate": 4.312403685109627e-06,
+      "loss": 0.6326,
+      "step": 8712
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4142017108540845,
+      "learning_rate": 4.311377060275942e-06,
+      "loss": 0.5847,
+      "step": 8713
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.595234163276579,
+      "learning_rate": 4.310350465033919e-06,
+      "loss": 0.5745,
+      "step": 8714
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6181766411900205,
+      "learning_rate": 4.309323899427671e-06,
+      "loss": 0.735,
+      "step": 8715
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3775678150282196,
+      "learning_rate": 4.308297363501314e-06,
+      "loss": 0.712,
+      "step": 8716
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6520668925353557,
+      "learning_rate": 4.3072708572989585e-06,
+      "loss": 0.5747,
+      "step": 8717
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3240287600736957,
+      "learning_rate": 4.306244380864719e-06,
+      "loss": 0.6309,
+      "step": 8718
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2497796682844318,
+      "learning_rate": 4.305217934242703e-06,
+      "loss": 0.588,
+      "step": 8719
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3967110987005353,
+      "learning_rate": 4.304191517477019e-06,
+      "loss": 0.6564,
+      "step": 8720
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3173516865872652,
+      "learning_rate": 4.303165130611776e-06,
+      "loss": 0.6361,
+      "step": 8721
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.298899384299054,
+      "learning_rate": 4.302138773691079e-06,
+      "loss": 0.6459,
+      "step": 8722
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6303367544540142,
+      "learning_rate": 4.301112446759033e-06,
+      "loss": 0.6826,
+      "step": 8723
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0659628153154945,
+      "learning_rate": 4.30008614985974e-06,
+      "loss": 0.6702,
+      "step": 8724
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2108790211130553,
+      "learning_rate": 4.299059883037308e-06,
+      "loss": 0.6094,
+      "step": 8725
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4081424949861163,
+      "learning_rate": 4.2980336463358296e-06,
+      "loss": 0.677,
+      "step": 8726
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1825300311483338,
+      "learning_rate": 4.29700743979941e-06,
+      "loss": 0.4853,
+      "step": 8727
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2337588413897382,
+      "learning_rate": 4.295981263472148e-06,
+      "loss": 0.6302,
+      "step": 8728
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3324376254017078,
+      "learning_rate": 4.294955117398139e-06,
+      "loss": 0.493,
+      "step": 8729
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5298734649480632,
+      "learning_rate": 4.293929001621477e-06,
+      "loss": 0.7276,
+      "step": 8730
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2886847824283323,
+      "learning_rate": 4.2929029161862575e-06,
+      "loss": 0.557,
+      "step": 8731
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1595226624511983,
+      "learning_rate": 4.291876861136578e-06,
+      "loss": 0.6584,
+      "step": 8732
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.771761288933658,
+      "learning_rate": 4.290850836516526e-06,
+      "loss": 0.7111,
+      "step": 8733
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4458925310336905,
+      "learning_rate": 4.28982484237019e-06,
+      "loss": 0.6483,
+      "step": 8734
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.335056436622656,
+      "learning_rate": 4.288798878741664e-06,
+      "loss": 0.6788,
+      "step": 8735
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.449871139123005,
+      "learning_rate": 4.287772945675035e-06,
+      "loss": 0.9212,
+      "step": 8736
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4913329499442909,
+      "learning_rate": 4.286747043214388e-06,
+      "loss": 0.7358,
+      "step": 8737
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4215187490283878,
+      "learning_rate": 4.285721171403809e-06,
+      "loss": 0.7708,
+      "step": 8738
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2942215655152374,
+      "learning_rate": 4.284695330287383e-06,
+      "loss": 0.6342,
+      "step": 8739
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3058498715773295,
+      "learning_rate": 4.283669519909189e-06,
+      "loss": 0.6756,
+      "step": 8740
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4314598422864544,
+      "learning_rate": 4.282643740313312e-06,
+      "loss": 0.6153,
+      "step": 8741
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2819523383718405,
+      "learning_rate": 4.281617991543832e-06,
+      "loss": 0.6159,
+      "step": 8742
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4282028339157231,
+      "learning_rate": 4.280592273644829e-06,
+      "loss": 0.6567,
+      "step": 8743
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.106438601385419,
+      "learning_rate": 4.279566586660375e-06,
+      "loss": 0.5647,
+      "step": 8744
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4957670651087687,
+      "learning_rate": 4.278540930634549e-06,
+      "loss": 0.7262,
+      "step": 8745
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7699622082367064,
+      "learning_rate": 4.277515305611427e-06,
+      "loss": 0.6838,
+      "step": 8746
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3954823533460916,
+      "learning_rate": 4.276489711635083e-06,
+      "loss": 0.7582,
+      "step": 8747
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3290893346899793,
+      "learning_rate": 4.275464148749585e-06,
+      "loss": 0.6426,
+      "step": 8748
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0374550092957884,
+      "learning_rate": 4.274438616999007e-06,
+      "loss": 0.6955,
+      "step": 8749
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6696604762033809,
+      "learning_rate": 4.273413116427419e-06,
+      "loss": 0.6181,
+      "step": 8750
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2071287560915898,
+      "learning_rate": 4.272387647078885e-06,
+      "loss": 0.5626,
+      "step": 8751
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5018053369453048,
+      "learning_rate": 4.271362208997476e-06,
+      "loss": 0.6949,
+      "step": 8752
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.268560981872239,
+      "learning_rate": 4.270336802227255e-06,
+      "loss": 0.7078,
+      "step": 8753
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4399201530920045,
+      "learning_rate": 4.269311426812287e-06,
+      "loss": 0.672,
+      "step": 8754
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6681606745536448,
+      "learning_rate": 4.268286082796634e-06,
+      "loss": 0.5537,
+      "step": 8755
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.473420655153086,
+      "learning_rate": 4.267260770224358e-06,
+      "loss": 0.6116,
+      "step": 8756
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3498677655056157,
+      "learning_rate": 4.26623548913952e-06,
+      "loss": 0.6565,
+      "step": 8757
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.283445507196875,
+      "learning_rate": 4.2652102395861746e-06,
+      "loss": 0.5614,
+      "step": 8758
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6806179696550134,
+      "learning_rate": 4.264185021608382e-06,
+      "loss": 0.7853,
+      "step": 8759
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.579421698605534,
+      "learning_rate": 4.263159835250199e-06,
+      "loss": 0.6268,
+      "step": 8760
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2090934549312844,
+      "learning_rate": 4.26213468055568e-06,
+      "loss": 0.5725,
+      "step": 8761
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5491856151269199,
+      "learning_rate": 4.2611095575688745e-06,
+      "loss": 0.6841,
+      "step": 8762
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.888442010046794,
+      "learning_rate": 4.260084466333837e-06,
+      "loss": 0.8088,
+      "step": 8763
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2576039222300817,
+      "learning_rate": 4.259059406894619e-06,
+      "loss": 0.5743,
+      "step": 8764
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.084133582505406,
+      "learning_rate": 4.258034379295268e-06,
+      "loss": 0.5909,
+      "step": 8765
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.631498686555072,
+      "learning_rate": 4.257009383579833e-06,
+      "loss": 0.6474,
+      "step": 8766
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.984981141008236,
+      "learning_rate": 4.255984419792358e-06,
+      "loss": 0.6393,
+      "step": 8767
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1114868600875025,
+      "learning_rate": 4.254959487976892e-06,
+      "loss": 0.5882,
+      "step": 8768
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5210015536869217,
+      "learning_rate": 4.253934588177473e-06,
+      "loss": 0.673,
+      "step": 8769
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3555146795314117,
+      "learning_rate": 4.252909720438149e-06,
+      "loss": 0.6642,
+      "step": 8770
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3126521623145104,
+      "learning_rate": 4.251884884802956e-06,
+      "loss": 0.5599,
+      "step": 8771
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.17415946592407,
+      "learning_rate": 4.250860081315937e-06,
+      "loss": 0.565,
+      "step": 8772
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4748441573804103,
+      "learning_rate": 4.2498353100211275e-06,
+      "loss": 0.6906,
+      "step": 8773
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.22268307089488,
+      "learning_rate": 4.248810570962567e-06,
+      "loss": 0.55,
+      "step": 8774
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0819426012319149,
+      "learning_rate": 4.247785864184289e-06,
+      "loss": 0.586,
+      "step": 8775
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2624170040682687,
+      "learning_rate": 4.246761189730327e-06,
+      "loss": 0.5924,
+      "step": 8776
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5223093885987335,
+      "learning_rate": 4.245736547644714e-06,
+      "loss": 0.5892,
+      "step": 8777
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5142143038168272,
+      "learning_rate": 4.2447119379714805e-06,
+      "loss": 0.7347,
+      "step": 8778
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5719469692872892,
+      "learning_rate": 4.24368736075466e-06,
+      "loss": 0.6199,
+      "step": 8779
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4941945273773356,
+      "learning_rate": 4.242662816038276e-06,
+      "loss": 0.7575,
+      "step": 8780
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.069991206929444,
+      "learning_rate": 4.241638303866358e-06,
+      "loss": 0.7628,
+      "step": 8781
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.379366302984491,
+      "learning_rate": 4.24061382428293e-06,
+      "loss": 0.6473,
+      "step": 8782
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0342948355922696,
+      "learning_rate": 4.23958937733202e-06,
+      "loss": 0.7069,
+      "step": 8783
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7900827793782566,
+      "learning_rate": 4.238564963057646e-06,
+      "loss": 0.7238,
+      "step": 8784
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1027628845411326,
+      "learning_rate": 4.237540581503831e-06,
+      "loss": 0.6255,
+      "step": 8785
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.238470360542041,
+      "learning_rate": 4.236516232714597e-06,
+      "loss": 0.5473,
+      "step": 8786
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4503751376647736,
+      "learning_rate": 4.235491916733959e-06,
+      "loss": 0.8528,
+      "step": 8787
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5982069255634799,
+      "learning_rate": 4.234467633605937e-06,
+      "loss": 0.5839,
+      "step": 8788
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6092636820001027,
+      "learning_rate": 4.233443383374545e-06,
+      "loss": 0.9133,
+      "step": 8789
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5755311573072102,
+      "learning_rate": 4.232419166083799e-06,
+      "loss": 0.7417,
+      "step": 8790
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4518529186035345,
+      "learning_rate": 4.23139498177771e-06,
+      "loss": 0.6396,
+      "step": 8791
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6398824491436692,
+      "learning_rate": 4.23037083050029e-06,
+      "loss": 0.638,
+      "step": 8792
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2382645277463735,
+      "learning_rate": 4.229346712295551e-06,
+      "loss": 0.6254,
+      "step": 8793
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4366830830416737,
+      "learning_rate": 4.228322627207499e-06,
+      "loss": 0.7012,
+      "step": 8794
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.095420877674918,
+      "learning_rate": 4.227298575280142e-06,
+      "loss": 0.6852,
+      "step": 8795
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3838715163180406,
+      "learning_rate": 4.226274556557487e-06,
+      "loss": 0.6493,
+      "step": 8796
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5898422133309733,
+      "learning_rate": 4.225250571083538e-06,
+      "loss": 0.6488,
+      "step": 8797
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4186033712883823,
+      "learning_rate": 4.2242266189022975e-06,
+      "loss": 0.6625,
+      "step": 8798
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4331041941515066,
+      "learning_rate": 4.223202700057765e-06,
+      "loss": 0.657,
+      "step": 8799
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.678148327199827,
+      "learning_rate": 4.2221788145939425e-06,
+      "loss": 0.6969,
+      "step": 8800
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4515265074284662,
+      "learning_rate": 4.221154962554831e-06,
+      "loss": 0.6189,
+      "step": 8801
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4776672603842658,
+      "learning_rate": 4.220131143984424e-06,
+      "loss": 0.6868,
+      "step": 8802
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4874632710627358,
+      "learning_rate": 4.219107358926718e-06,
+      "loss": 0.6422,
+      "step": 8803
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5754386264841844,
+      "learning_rate": 4.21808360742571e-06,
+      "loss": 0.6831,
+      "step": 8804
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.382812586279162,
+      "learning_rate": 4.217059889525389e-06,
+      "loss": 0.6069,
+      "step": 8805
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4767113127098495,
+      "learning_rate": 4.216036205269748e-06,
+      "loss": 0.5551,
+      "step": 8806
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.426845720926234,
+      "learning_rate": 4.215012554702778e-06,
+      "loss": 0.6964,
+      "step": 8807
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.188575905854017,
+      "learning_rate": 4.213988937868469e-06,
+      "loss": 0.6355,
+      "step": 8808
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4926623840125295,
+      "learning_rate": 4.212965354810802e-06,
+      "loss": 0.717,
+      "step": 8809
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.524051511931508,
+      "learning_rate": 4.211941805573767e-06,
+      "loss": 0.7011,
+      "step": 8810
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6577364427701897,
+      "learning_rate": 4.21091829020135e-06,
+      "loss": 0.6075,
+      "step": 8811
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0588657457648352,
+      "learning_rate": 4.209894808737531e-06,
+      "loss": 0.5898,
+      "step": 8812
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4031454889575947,
+      "learning_rate": 4.20887136122629e-06,
+      "loss": 0.6701,
+      "step": 8813
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.213006118159163,
+      "learning_rate": 4.207847947711609e-06,
+      "loss": 0.6744,
+      "step": 8814
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4058892491299728,
+      "learning_rate": 4.206824568237468e-06,
+      "loss": 0.6606,
+      "step": 8815
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6112888806953602,
+      "learning_rate": 4.205801222847839e-06,
+      "loss": 0.6938,
+      "step": 8816
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3966199645816344,
+      "learning_rate": 4.204777911586702e-06,
+      "loss": 0.5903,
+      "step": 8817
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.233658725494542,
+      "learning_rate": 4.203754634498027e-06,
+      "loss": 0.6355,
+      "step": 8818
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4190941927973033,
+      "learning_rate": 4.202731391625793e-06,
+      "loss": 0.7869,
+      "step": 8819
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4970704038503333,
+      "learning_rate": 4.201708183013963e-06,
+      "loss": 0.6366,
+      "step": 8820
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3969102791887158,
+      "learning_rate": 4.200685008706511e-06,
+      "loss": 0.6172,
+      "step": 8821
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2947034945456837,
+      "learning_rate": 4.199661868747406e-06,
+      "loss": 0.7039,
+      "step": 8822
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.301830925305037,
+      "learning_rate": 4.198638763180611e-06,
+      "loss": 0.5223,
+      "step": 8823
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1336704231785277,
+      "learning_rate": 4.1976156920500935e-06,
+      "loss": 0.5476,
+      "step": 8824
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3116288787947181,
+      "learning_rate": 4.196592655399816e-06,
+      "loss": 0.7209,
+      "step": 8825
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2950948157870559,
+      "learning_rate": 4.195569653273743e-06,
+      "loss": 0.6564,
+      "step": 8826
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1858786822111402,
+      "learning_rate": 4.1945466857158336e-06,
+      "loss": 0.6138,
+      "step": 8827
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4278602455335774,
+      "learning_rate": 4.193523752770045e-06,
+      "loss": 0.6011,
+      "step": 8828
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3549738300458252,
+      "learning_rate": 4.19250085448034e-06,
+      "loss": 0.6616,
+      "step": 8829
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3223974964026706,
+      "learning_rate": 4.191477990890668e-06,
+      "loss": 0.5086,
+      "step": 8830
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.452162500691464,
+      "learning_rate": 4.19045516204499e-06,
+      "loss": 0.6499,
+      "step": 8831
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.395376931807339,
+      "learning_rate": 4.189432367987255e-06,
+      "loss": 0.7066,
+      "step": 8832
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.272408896996976,
+      "learning_rate": 4.188409608761419e-06,
+      "loss": 0.6391,
+      "step": 8833
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4103338412384943,
+      "learning_rate": 4.187386884411426e-06,
+      "loss": 0.7595,
+      "step": 8834
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3185299298093682,
+      "learning_rate": 4.186364194981231e-06,
+      "loss": 0.6289,
+      "step": 8835
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2982649771747128,
+      "learning_rate": 4.1853415405147765e-06,
+      "loss": 0.6095,
+      "step": 8836
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3760369308566291,
+      "learning_rate": 4.184318921056013e-06,
+      "loss": 0.6678,
+      "step": 8837
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.424476294163743,
+      "learning_rate": 4.18329633664888e-06,
+      "loss": 0.5686,
+      "step": 8838
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2655074189401607,
+      "learning_rate": 4.182273787337323e-06,
+      "loss": 0.5877,
+      "step": 8839
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3328336118239599,
+      "learning_rate": 4.181251273165283e-06,
+      "loss": 0.5501,
+      "step": 8840
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.138457546966545,
+      "learning_rate": 4.1802287941767e-06,
+      "loss": 0.6561,
+      "step": 8841
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3715548532469324,
+      "learning_rate": 4.179206350415509e-06,
+      "loss": 0.6392,
+      "step": 8842
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2128162312914117,
+      "learning_rate": 4.17818394192565e-06,
+      "loss": 0.5289,
+      "step": 8843
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.384401050973362,
+      "learning_rate": 4.177161568751058e-06,
+      "loss": 0.6206,
+      "step": 8844
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.430337265148639,
+      "learning_rate": 4.176139230935666e-06,
+      "loss": 0.6504,
+      "step": 8845
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4477065876026014,
+      "learning_rate": 4.175116928523405e-06,
+      "loss": 0.6745,
+      "step": 8846
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3246645719311891,
+      "learning_rate": 4.174094661558209e-06,
+      "loss": 0.6612,
+      "step": 8847
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4236268448826337,
+      "learning_rate": 4.173072430084002e-06,
+      "loss": 0.6456,
+      "step": 8848
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.387491983404803,
+      "learning_rate": 4.172050234144716e-06,
+      "loss": 0.5415,
+      "step": 8849
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4473466419389314,
+      "learning_rate": 4.171028073784274e-06,
+      "loss": 0.7004,
+      "step": 8850
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3485396497035413,
+      "learning_rate": 4.170005949046604e-06,
+      "loss": 0.6122,
+      "step": 8851
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5587744129889831,
+      "learning_rate": 4.168983859975625e-06,
+      "loss": 0.6445,
+      "step": 8852
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3328240554978366,
+      "learning_rate": 4.16796180661526e-06,
+      "loss": 0.661,
+      "step": 8853
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3263662081544236,
+      "learning_rate": 4.16693978900943e-06,
+      "loss": 0.5576,
+      "step": 8854
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4941792089595896,
+      "learning_rate": 4.165917807202055e-06,
+      "loss": 0.6402,
+      "step": 8855
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2224525024732784,
+      "learning_rate": 4.164895861237046e-06,
+      "loss": 0.6941,
+      "step": 8856
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5650981552262306,
+      "learning_rate": 4.1638739511583224e-06,
+      "loss": 0.6117,
+      "step": 8857
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2322574761701113,
+      "learning_rate": 4.1628520770097994e-06,
+      "loss": 0.6626,
+      "step": 8858
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4021920294161896,
+      "learning_rate": 4.161830238835386e-06,
+      "loss": 0.6041,
+      "step": 8859
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5197621401509083,
+      "learning_rate": 4.160808436678992e-06,
+      "loss": 0.7175,
+      "step": 8860
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3054706180921003,
+      "learning_rate": 4.1597866705845295e-06,
+      "loss": 0.5814,
+      "step": 8861
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9088482336164438,
+      "learning_rate": 4.1587649405959065e-06,
+      "loss": 0.6142,
+      "step": 8862
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.117212869264639,
+      "learning_rate": 4.157743246757026e-06,
+      "loss": 0.7503,
+      "step": 8863
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4174472207567734,
+      "learning_rate": 4.156721589111794e-06,
+      "loss": 0.7279,
+      "step": 8864
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.064365874248402,
+      "learning_rate": 4.155699967704113e-06,
+      "loss": 0.5884,
+      "step": 8865
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4354110548922487,
+      "learning_rate": 4.154678382577887e-06,
+      "loss": 0.6187,
+      "step": 8866
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3156093759121343,
+      "learning_rate": 4.1536568337770114e-06,
+      "loss": 0.6709,
+      "step": 8867
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4356178127569528,
+      "learning_rate": 4.152635321345387e-06,
+      "loss": 0.5892,
+      "step": 8868
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3511035546693608,
+      "learning_rate": 4.151613845326912e-06,
+      "loss": 0.6618,
+      "step": 8869
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2848514129574926,
+      "learning_rate": 4.150592405765476e-06,
+      "loss": 0.5223,
+      "step": 8870
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3184699909676814,
+      "learning_rate": 4.1495710027049776e-06,
+      "loss": 0.6625,
+      "step": 8871
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5002108898575022,
+      "learning_rate": 4.148549636189308e-06,
+      "loss": 0.7642,
+      "step": 8872
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3328894130424231,
+      "learning_rate": 4.147528306262357e-06,
+      "loss": 0.7028,
+      "step": 8873
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3486194652138834,
+      "learning_rate": 4.146507012968013e-06,
+      "loss": 0.6613,
+      "step": 8874
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.525148498263408,
+      "learning_rate": 4.145485756350162e-06,
+      "loss": 0.6987,
+      "step": 8875
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.404528027251261,
+      "learning_rate": 4.144464536452693e-06,
+      "loss": 0.633,
+      "step": 8876
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6038625131679922,
+      "learning_rate": 4.1434433533194865e-06,
+      "loss": 0.6288,
+      "step": 8877
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6539952498993298,
+      "learning_rate": 4.142422206994428e-06,
+      "loss": 0.8693,
+      "step": 8878
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2563884708709365,
+      "learning_rate": 4.141401097521396e-06,
+      "loss": 0.7491,
+      "step": 8879
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.387993852364057,
+      "learning_rate": 4.140380024944273e-06,
+      "loss": 0.627,
+      "step": 8880
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3663689554820735,
+      "learning_rate": 4.139358989306933e-06,
+      "loss": 0.6193,
+      "step": 8881
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4319485087641977,
+      "learning_rate": 4.138337990653255e-06,
+      "loss": 0.7204,
+      "step": 8882
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.359365224595696,
+      "learning_rate": 4.137317029027111e-06,
+      "loss": 0.6316,
+      "step": 8883
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3027986005736176,
+      "learning_rate": 4.136296104472378e-06,
+      "loss": 0.6119,
+      "step": 8884
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.330657527876677,
+      "learning_rate": 4.1352752170329236e-06,
+      "loss": 0.6628,
+      "step": 8885
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4284239554200635,
+      "learning_rate": 4.1342543667526195e-06,
+      "loss": 0.6135,
+      "step": 8886
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1726182341598521,
+      "learning_rate": 4.133233553675334e-06,
+      "loss": 0.5469,
+      "step": 8887
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.54679491770651,
+      "learning_rate": 4.1322127778449315e-06,
+      "loss": 0.7164,
+      "step": 8888
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4331140979483161,
+      "learning_rate": 4.131192039305278e-06,
+      "loss": 0.6087,
+      "step": 8889
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3319545921209601,
+      "learning_rate": 4.1301713381002394e-06,
+      "loss": 0.7097,
+      "step": 8890
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5037867418396575,
+      "learning_rate": 4.129150674273675e-06,
+      "loss": 0.6177,
+      "step": 8891
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2455257590511901,
+      "learning_rate": 4.128130047869446e-06,
+      "loss": 0.612,
+      "step": 8892
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3307090998046032,
+      "learning_rate": 4.12710945893141e-06,
+      "loss": 0.5415,
+      "step": 8893
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3294353567338515,
+      "learning_rate": 4.1260889075034254e-06,
+      "loss": 0.5517,
+      "step": 8894
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.452766341203677,
+      "learning_rate": 4.125068393629346e-06,
+      "loss": 0.6808,
+      "step": 8895
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1601940376751714,
+      "learning_rate": 4.124047917353026e-06,
+      "loss": 0.5598,
+      "step": 8896
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2153629671209503,
+      "learning_rate": 4.123027478718318e-06,
+      "loss": 0.6575,
+      "step": 8897
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.183848591029906,
+      "learning_rate": 4.1220070777690736e-06,
+      "loss": 0.7788,
+      "step": 8898
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1303693028761759,
+      "learning_rate": 4.120986714549139e-06,
+      "loss": 0.6804,
+      "step": 8899
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.275494001331222,
+      "learning_rate": 4.119966389102363e-06,
+      "loss": 0.6255,
+      "step": 8900
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2157289823794037,
+      "learning_rate": 4.118946101472591e-06,
+      "loss": 0.6284,
+      "step": 8901
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3372738489928335,
+      "learning_rate": 4.117925851703669e-06,
+      "loss": 0.6682,
+      "step": 8902
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.391356686184438,
+      "learning_rate": 4.116905639839436e-06,
+      "loss": 0.593,
+      "step": 8903
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.247613389881829,
+      "learning_rate": 4.115885465923734e-06,
+      "loss": 0.6173,
+      "step": 8904
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5751425047301757,
+      "learning_rate": 4.114865330000405e-06,
+      "loss": 0.629,
+      "step": 8905
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1413469095882642,
+      "learning_rate": 4.113845232113282e-06,
+      "loss": 0.5666,
+      "step": 8906
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3729597509878482,
+      "learning_rate": 4.112825172306203e-06,
+      "loss": 0.6001,
+      "step": 8907
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.48592284615463,
+      "learning_rate": 4.111805150623002e-06,
+      "loss": 0.72,
+      "step": 8908
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2494724391603482,
+      "learning_rate": 4.110785167107514e-06,
+      "loss": 0.6445,
+      "step": 8909
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.269346837296491,
+      "learning_rate": 4.1097652218035666e-06,
+      "loss": 0.5931,
+      "step": 8910
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1353669543074576,
+      "learning_rate": 4.108745314754989e-06,
+      "loss": 0.5695,
+      "step": 8911
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3048780408163938,
+      "learning_rate": 4.107725446005612e-06,
+      "loss": 0.6269,
+      "step": 8912
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3135484403896984,
+      "learning_rate": 4.106705615599258e-06,
+      "loss": 0.6094,
+      "step": 8913
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.351716626992667,
+      "learning_rate": 4.1056858235797545e-06,
+      "loss": 0.6774,
+      "step": 8914
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5503923865797098,
+      "learning_rate": 4.104666069990921e-06,
+      "loss": 0.728,
+      "step": 8915
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7592286804406478,
+      "learning_rate": 4.103646354876584e-06,
+      "loss": 0.8137,
+      "step": 8916
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4629077004623914,
+      "learning_rate": 4.102626678280556e-06,
+      "loss": 0.6233,
+      "step": 8917
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3894586806434772,
+      "learning_rate": 4.101607040246659e-06,
+      "loss": 0.6474,
+      "step": 8918
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4234152308182386,
+      "learning_rate": 4.100587440818709e-06,
+      "loss": 0.771,
+      "step": 8919
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.580307282596117,
+      "learning_rate": 4.099567880040521e-06,
+      "loss": 0.5602,
+      "step": 8920
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4138805428589263,
+      "learning_rate": 4.098548357955904e-06,
+      "loss": 0.5473,
+      "step": 8921
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1944839567818442,
+      "learning_rate": 4.0975288746086724e-06,
+      "loss": 0.567,
+      "step": 8922
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2432894390830092,
+      "learning_rate": 4.0965094300426364e-06,
+      "loss": 0.6821,
+      "step": 8923
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2171833432997727,
+      "learning_rate": 4.0954900243016016e-06,
+      "loss": 0.6933,
+      "step": 8924
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6451910859248235,
+      "learning_rate": 4.094470657429374e-06,
+      "loss": 0.6023,
+      "step": 8925
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1572885232550305,
+      "learning_rate": 4.093451329469758e-06,
+      "loss": 0.6066,
+      "step": 8926
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3273361350211303,
+      "learning_rate": 4.092432040466561e-06,
+      "loss": 0.7087,
+      "step": 8927
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3878194585857027,
+      "learning_rate": 4.091412790463578e-06,
+      "loss": 0.6105,
+      "step": 8928
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4746971510269797,
+      "learning_rate": 4.090393579504612e-06,
+      "loss": 0.6831,
+      "step": 8929
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3165448900185022,
+      "learning_rate": 4.0893744076334606e-06,
+      "loss": 0.6527,
+      "step": 8930
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1348898430538972,
+      "learning_rate": 4.088355274893916e-06,
+      "loss": 0.608,
+      "step": 8931
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.237272017821911,
+      "learning_rate": 4.087336181329777e-06,
+      "loss": 0.5928,
+      "step": 8932
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.439084190206576,
+      "learning_rate": 4.086317126984835e-06,
+      "loss": 0.726,
+      "step": 8933
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1964609621881805,
+      "learning_rate": 4.085298111902882e-06,
+      "loss": 0.7064,
+      "step": 8934
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1402625147969883,
+      "learning_rate": 4.084279136127704e-06,
+      "loss": 0.637,
+      "step": 8935
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.366894703453319,
+      "learning_rate": 4.083260199703092e-06,
+      "loss": 0.5595,
+      "step": 8936
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.373972159143029,
+      "learning_rate": 4.082241302672831e-06,
+      "loss": 0.6819,
+      "step": 8937
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9727234170470123,
+      "learning_rate": 4.081222445080707e-06,
+      "loss": 0.573,
+      "step": 8938
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.390185546075743,
+      "learning_rate": 4.080203626970498e-06,
+      "loss": 0.6395,
+      "step": 8939
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4135722196885137,
+      "learning_rate": 4.079184848385988e-06,
+      "loss": 0.6078,
+      "step": 8940
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5312473368354456,
+      "learning_rate": 4.078166109370957e-06,
+      "loss": 0.7317,
+      "step": 8941
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4734067841350502,
+      "learning_rate": 4.077147409969181e-06,
+      "loss": 0.7148,
+      "step": 8942
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5328369082201097,
+      "learning_rate": 4.076128750224436e-06,
+      "loss": 0.5808,
+      "step": 8943
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.421937373312303,
+      "learning_rate": 4.0751101301804945e-06,
+      "loss": 0.5966,
+      "step": 8944
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4473052269846804,
+      "learning_rate": 4.074091549881134e-06,
+      "loss": 0.6991,
+      "step": 8945
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4497891817010407,
+      "learning_rate": 4.0730730093701185e-06,
+      "loss": 0.6522,
+      "step": 8946
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4120280691569387,
+      "learning_rate": 4.072054508691222e-06,
+      "loss": 0.6494,
+      "step": 8947
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4312613937614085,
+      "learning_rate": 4.07103604788821e-06,
+      "loss": 0.6205,
+      "step": 8948
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2794265817489756,
+      "learning_rate": 4.070017627004847e-06,
+      "loss": 0.6213,
+      "step": 8949
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.563625644779719,
+      "learning_rate": 4.068999246084898e-06,
+      "loss": 0.7155,
+      "step": 8950
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1992720949643747,
+      "learning_rate": 4.067980905172125e-06,
+      "loss": 0.7052,
+      "step": 8951
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.448991231924534,
+      "learning_rate": 4.066962604310288e-06,
+      "loss": 0.6677,
+      "step": 8952
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8448633378883772,
+      "learning_rate": 4.065944343543146e-06,
+      "loss": 0.7768,
+      "step": 8953
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2395498931182756,
+      "learning_rate": 4.0649261229144554e-06,
+      "loss": 0.5932,
+      "step": 8954
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3678397547214531,
+      "learning_rate": 4.063907942467971e-06,
+      "loss": 0.6444,
+      "step": 8955
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3704919766120292,
+      "learning_rate": 4.06288980224745e-06,
+      "loss": 0.68,
+      "step": 8956
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1659700905460433,
+      "learning_rate": 4.0618717022966395e-06,
+      "loss": 0.5599,
+      "step": 8957
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.328496691953845,
+      "learning_rate": 4.060853642659291e-06,
+      "loss": 0.61,
+      "step": 8958
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2803602604511068,
+      "learning_rate": 4.059835623379155e-06,
+      "loss": 0.6919,
+      "step": 8959
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7489973238684045,
+      "learning_rate": 4.058817644499973e-06,
+      "loss": 0.6755,
+      "step": 8960
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.331840802706706,
+      "learning_rate": 4.057799706065495e-06,
+      "loss": 0.5484,
+      "step": 8961
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6012182568644053,
+      "learning_rate": 4.056781808119461e-06,
+      "loss": 0.6489,
+      "step": 8962
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3634625350664902,
+      "learning_rate": 4.055763950705615e-06,
+      "loss": 0.732,
+      "step": 8963
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6299275251882985,
+      "learning_rate": 4.054746133867693e-06,
+      "loss": 0.5245,
+      "step": 8964
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5047392835649576,
+      "learning_rate": 4.053728357649436e-06,
+      "loss": 0.6872,
+      "step": 8965
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.314257765290702,
+      "learning_rate": 4.052710622094579e-06,
+      "loss": 0.6043,
+      "step": 8966
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7197234745175707,
+      "learning_rate": 4.051692927246857e-06,
+      "loss": 0.7516,
+      "step": 8967
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5573141767861025,
+      "learning_rate": 4.050675273149999e-06,
+      "loss": 0.7028,
+      "step": 8968
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4468408701684718,
+      "learning_rate": 4.0496576598477396e-06,
+      "loss": 0.6501,
+      "step": 8969
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4050886164695346,
+      "learning_rate": 4.048640087383809e-06,
+      "loss": 0.6471,
+      "step": 8970
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.309856175821882,
+      "learning_rate": 4.0476225558019324e-06,
+      "loss": 0.6406,
+      "step": 8971
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.381966871899665,
+      "learning_rate": 4.046605065145834e-06,
+      "loss": 0.6722,
+      "step": 8972
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3057806775649408,
+      "learning_rate": 4.04558761545924e-06,
+      "loss": 0.6264,
+      "step": 8973
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2452066852493653,
+      "learning_rate": 4.044570206785874e-06,
+      "loss": 0.6526,
+      "step": 8974
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3601413711132957,
+      "learning_rate": 4.043552839169451e-06,
+      "loss": 0.676,
+      "step": 8975
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5437960508929196,
+      "learning_rate": 4.042535512653693e-06,
+      "loss": 0.6227,
+      "step": 8976
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4682995492263187,
+      "learning_rate": 4.041518227282319e-06,
+      "loss": 0.5545,
+      "step": 8977
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8167718491280345,
+      "learning_rate": 4.040500983099038e-06,
+      "loss": 0.7565,
+      "step": 8978
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2983608731734668,
+      "learning_rate": 4.039483780147568e-06,
+      "loss": 0.6738,
+      "step": 8979
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1821467057166264,
+      "learning_rate": 4.038466618471619e-06,
+      "loss": 0.6506,
+      "step": 8980
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.523064156473597,
+      "learning_rate": 4.037449498114903e-06,
+      "loss": 0.7195,
+      "step": 8981
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4236782303847224,
+      "learning_rate": 4.036432419121123e-06,
+      "loss": 0.5966,
+      "step": 8982
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.243627223388809,
+      "learning_rate": 4.035415381533988e-06,
+      "loss": 0.7287,
+      "step": 8983
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3957830097724253,
+      "learning_rate": 4.0343983853972045e-06,
+      "loss": 0.7096,
+      "step": 8984
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0264564548401194,
+      "learning_rate": 4.0333814307544726e-06,
+      "loss": 0.6477,
+      "step": 8985
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5537933263122528,
+      "learning_rate": 4.032364517649491e-06,
+      "loss": 0.7132,
+      "step": 8986
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1719806483704023,
+      "learning_rate": 4.0313476461259615e-06,
+      "loss": 0.7302,
+      "step": 8987
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5906629620642112,
+      "learning_rate": 4.0303308162275835e-06,
+      "loss": 0.7217,
+      "step": 8988
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4652496330098737,
+      "learning_rate": 4.029314027998049e-06,
+      "loss": 0.6977,
+      "step": 8989
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1578724352827734,
+      "learning_rate": 4.028297281481051e-06,
+      "loss": 0.6714,
+      "step": 8990
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4470837362278235,
+      "learning_rate": 4.027280576720282e-06,
+      "loss": 0.6532,
+      "step": 8991
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3012222693828885,
+      "learning_rate": 4.026263913759437e-06,
+      "loss": 0.6475,
+      "step": 8992
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2932988700625863,
+      "learning_rate": 4.025247292642197e-06,
+      "loss": 0.6421,
+      "step": 8993
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0186265258790947,
+      "learning_rate": 4.024230713412253e-06,
+      "loss": 0.6417,
+      "step": 8994
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6534327796165174,
+      "learning_rate": 4.0232141761132894e-06,
+      "loss": 0.7402,
+      "step": 8995
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2405362249722045,
+      "learning_rate": 4.022197680788986e-06,
+      "loss": 0.7244,
+      "step": 8996
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2697713045583179,
+      "learning_rate": 4.021181227483026e-06,
+      "loss": 0.5244,
+      "step": 8997
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3681910547533644,
+      "learning_rate": 4.02016481623909e-06,
+      "loss": 0.6727,
+      "step": 8998
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2818773685517644,
+      "learning_rate": 4.019148447100855e-06,
+      "loss": 0.7257,
+      "step": 8999
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3606829704075938,
+      "learning_rate": 4.018132120111992e-06,
+      "loss": 0.6943,
+      "step": 9000
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.444092227704479,
+      "learning_rate": 4.0171158353161805e-06,
+      "loss": 0.6593,
+      "step": 9001
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2459582131320004,
+      "learning_rate": 4.016099592757091e-06,
+      "loss": 0.7614,
+      "step": 9002
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3019609535326975,
+      "learning_rate": 4.015083392478393e-06,
+      "loss": 0.6349,
+      "step": 9003
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0890604650426225,
+      "learning_rate": 4.014067234523756e-06,
+      "loss": 0.7312,
+      "step": 9004
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4353810827901847,
+      "learning_rate": 4.0130511189368435e-06,
+      "loss": 0.6951,
+      "step": 9005
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3195553127404447,
+      "learning_rate": 4.012035045761326e-06,
+      "loss": 0.468,
+      "step": 9006
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.311770847121976,
+      "learning_rate": 4.0110190150408594e-06,
+      "loss": 0.6428,
+      "step": 9007
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1783522081788782,
+      "learning_rate": 4.01000302681911e-06,
+      "loss": 0.6837,
+      "step": 9008
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.354270737231472,
+      "learning_rate": 4.008987081139734e-06,
+      "loss": 0.7433,
+      "step": 9009
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.181818702179306,
+      "learning_rate": 4.007971178046393e-06,
+      "loss": 0.5464,
+      "step": 9010
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2547118585186043,
+      "learning_rate": 4.006955317582737e-06,
+      "loss": 0.5697,
+      "step": 9011
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6858753912270301,
+      "learning_rate": 4.005939499792424e-06,
+      "loss": 0.736,
+      "step": 9012
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.410065915097856,
+      "learning_rate": 4.004923724719106e-06,
+      "loss": 0.6241,
+      "step": 9013
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6893131312750977,
+      "learning_rate": 4.0039079924064285e-06,
+      "loss": 0.6067,
+      "step": 9014
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2796632453019217,
+      "learning_rate": 4.0028923028980435e-06,
+      "loss": 0.6042,
+      "step": 9015
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2492616318064627,
+      "learning_rate": 4.0018766562375984e-06,
+      "loss": 0.6423,
+      "step": 9016
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5770154595861865,
+      "learning_rate": 4.000861052468736e-06,
+      "loss": 0.7252,
+      "step": 9017
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.786942965152114,
+      "learning_rate": 3.9998454916350996e-06,
+      "loss": 0.6984,
+      "step": 9018
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1070833265552986,
+      "learning_rate": 3.998829973780329e-06,
+      "loss": 0.6246,
+      "step": 9019
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.388838395996351,
+      "learning_rate": 3.997814498948064e-06,
+      "loss": 0.6615,
+      "step": 9020
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3974238705919322,
+      "learning_rate": 3.996799067181943e-06,
+      "loss": 0.7122,
+      "step": 9021
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2853318596985792,
+      "learning_rate": 3.995783678525601e-06,
+      "loss": 0.608,
+      "step": 9022
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.401997809104501,
+      "learning_rate": 3.994768333022669e-06,
+      "loss": 0.5841,
+      "step": 9023
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2553441336536233,
+      "learning_rate": 3.993753030716783e-06,
+      "loss": 0.5754,
+      "step": 9024
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2531515015360177,
+      "learning_rate": 3.992737771651567e-06,
+      "loss": 0.6099,
+      "step": 9025
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.6439506603761944,
+      "learning_rate": 3.991722555870655e-06,
+      "loss": 0.6204,
+      "step": 9026
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4718314884116392,
+      "learning_rate": 3.990707383417668e-06,
+      "loss": 0.6602,
+      "step": 9027
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3130887098470223,
+      "learning_rate": 3.989692254336235e-06,
+      "loss": 0.5877,
+      "step": 9028
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6570479675782614,
+      "learning_rate": 3.988677168669974e-06,
+      "loss": 0.645,
+      "step": 9029
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.007336490977157,
+      "learning_rate": 3.987662126462507e-06,
+      "loss": 0.6007,
+      "step": 9030
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3296022734716344,
+      "learning_rate": 3.9866471277574554e-06,
+      "loss": 0.6436,
+      "step": 9031
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.423231227240757,
+      "learning_rate": 3.985632172598433e-06,
+      "loss": 0.7283,
+      "step": 9032
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2261987253835898,
+      "learning_rate": 3.984617261029054e-06,
+      "loss": 0.621,
+      "step": 9033
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3735527774838805,
+      "learning_rate": 3.983602393092931e-06,
+      "loss": 0.6783,
+      "step": 9034
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3158220905278093,
+      "learning_rate": 3.982587568833681e-06,
+      "loss": 0.6085,
+      "step": 9035
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0942094675393106,
+      "learning_rate": 3.981572788294907e-06,
+      "loss": 0.6046,
+      "step": 9036
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3702946475444056,
+      "learning_rate": 3.980558051520218e-06,
+      "loss": 0.652,
+      "step": 9037
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.322099852173322,
+      "learning_rate": 3.97954335855322e-06,
+      "loss": 0.6005,
+      "step": 9038
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2211446343254149,
+      "learning_rate": 3.978528709437518e-06,
+      "loss": 0.687,
+      "step": 9039
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.320148293759886,
+      "learning_rate": 3.977514104216711e-06,
+      "loss": 0.6631,
+      "step": 9040
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3295634909766,
+      "learning_rate": 3.9764995429344e-06,
+      "loss": 0.6888,
+      "step": 9041
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3608413772643402,
+      "learning_rate": 3.975485025634185e-06,
+      "loss": 0.7345,
+      "step": 9042
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4514606589907637,
+      "learning_rate": 3.9744705523596575e-06,
+      "loss": 0.5509,
+      "step": 9043
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4994183736374656,
+      "learning_rate": 3.973456123154415e-06,
+      "loss": 0.7071,
+      "step": 9044
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2200410562746409,
+      "learning_rate": 3.972441738062049e-06,
+      "loss": 0.5124,
+      "step": 9045
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.017553338059595,
+      "learning_rate": 3.971427397126153e-06,
+      "loss": 0.6032,
+      "step": 9046
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4526607164863785,
+      "learning_rate": 3.970413100390309e-06,
+      "loss": 0.7124,
+      "step": 9047
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3890555292457298,
+      "learning_rate": 3.9693988478981076e-06,
+      "loss": 0.6096,
+      "step": 9048
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0519477827825228,
+      "learning_rate": 3.9683846396931345e-06,
+      "loss": 0.7261,
+      "step": 9049
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.521909492355007,
+      "learning_rate": 3.967370475818971e-06,
+      "loss": 0.7331,
+      "step": 9050
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5491281382448006,
+      "learning_rate": 3.966356356319196e-06,
+      "loss": 0.6547,
+      "step": 9051
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4104774748109263,
+      "learning_rate": 3.965342281237391e-06,
+      "loss": 0.5778,
+      "step": 9052
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3224499235387257,
+      "learning_rate": 3.964328250617134e-06,
+      "loss": 0.6487,
+      "step": 9053
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4069183916269228,
+      "learning_rate": 3.9633142645019965e-06,
+      "loss": 0.6489,
+      "step": 9054
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1849213717308837,
+      "learning_rate": 3.962300322935556e-06,
+      "loss": 0.5007,
+      "step": 9055
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8806569275235465,
+      "learning_rate": 3.961286425961379e-06,
+      "loss": 0.6761,
+      "step": 9056
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2368739380120317,
+      "learning_rate": 3.9602725736230405e-06,
+      "loss": 0.647,
+      "step": 9057
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5792652614735867,
+      "learning_rate": 3.959258765964104e-06,
+      "loss": 0.6124,
+      "step": 9058
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.229154295677476,
+      "learning_rate": 3.958245003028136e-06,
+      "loss": 0.6788,
+      "step": 9059
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4708229760650278,
+      "learning_rate": 3.957231284858701e-06,
+      "loss": 0.6721,
+      "step": 9060
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2731345412170567,
+      "learning_rate": 3.956217611499359e-06,
+      "loss": 0.534,
+      "step": 9061
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.375786872759154,
+      "learning_rate": 3.9552039829936705e-06,
+      "loss": 0.6247,
+      "step": 9062
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.999592751712679,
+      "learning_rate": 3.954190399385195e-06,
+      "loss": 0.5358,
+      "step": 9063
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3574261371891088,
+      "learning_rate": 3.953176860717488e-06,
+      "loss": 0.6329,
+      "step": 9064
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0689178081384334,
+      "learning_rate": 3.9521633670341005e-06,
+      "loss": 0.605,
+      "step": 9065
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.499238155507307,
+      "learning_rate": 3.9511499183785875e-06,
+      "loss": 0.7669,
+      "step": 9066
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7340449716216584,
+      "learning_rate": 3.950136514794499e-06,
+      "loss": 0.5591,
+      "step": 9067
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.603528959314479,
+      "learning_rate": 3.9491231563253815e-06,
+      "loss": 0.6585,
+      "step": 9068
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3543952829284531,
+      "learning_rate": 3.948109843014784e-06,
+      "loss": 0.7276,
+      "step": 9069
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.589438378856634,
+      "learning_rate": 3.947096574906248e-06,
+      "loss": 0.7457,
+      "step": 9070
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2379599100247844,
+      "learning_rate": 3.946083352043318e-06,
+      "loss": 0.659,
+      "step": 9071
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6858789211658265,
+      "learning_rate": 3.9450701744695325e-06,
+      "loss": 0.7182,
+      "step": 9072
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3191564448406576,
+      "learning_rate": 3.944057042228432e-06,
+      "loss": 0.6351,
+      "step": 9073
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.365392945171209,
+      "learning_rate": 3.9430439553635504e-06,
+      "loss": 0.7196,
+      "step": 9074
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.657487983003866,
+      "learning_rate": 3.942030913918427e-06,
+      "loss": 0.7883,
+      "step": 9075
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5895120523178838,
+      "learning_rate": 3.9410179179365895e-06,
+      "loss": 0.6618,
+      "step": 9076
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3998669232872252,
+      "learning_rate": 3.940004967461571e-06,
+      "loss": 0.5646,
+      "step": 9077
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3186497223988858,
+      "learning_rate": 3.9389920625368996e-06,
+      "loss": 0.6581,
+      "step": 9078
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1102375303263259,
+      "learning_rate": 3.937979203206103e-06,
+      "loss": 0.5204,
+      "step": 9079
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3595848442600214,
+      "learning_rate": 3.936966389512703e-06,
+      "loss": 0.6946,
+      "step": 9080
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3125358198156312,
+      "learning_rate": 3.935953621500226e-06,
+      "loss": 0.5221,
+      "step": 9081
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.558726572766821,
+      "learning_rate": 3.934940899212193e-06,
+      "loss": 0.6554,
+      "step": 9082
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2762145312242053,
+      "learning_rate": 3.9339282226921215e-06,
+      "loss": 0.6816,
+      "step": 9083
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4904319814389408,
+      "learning_rate": 3.932915591983526e-06,
+      "loss": 0.6008,
+      "step": 9084
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3519659334615322,
+      "learning_rate": 3.931903007129927e-06,
+      "loss": 0.6399,
+      "step": 9085
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0150631840490396,
+      "learning_rate": 3.930890468174833e-06,
+      "loss": 0.6458,
+      "step": 9086
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.752056653905289,
+      "learning_rate": 3.929877975161757e-06,
+      "loss": 0.6767,
+      "step": 9087
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4370650125914073,
+      "learning_rate": 3.928865528134206e-06,
+      "loss": 0.6646,
+      "step": 9088
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.32739740581813,
+      "learning_rate": 3.927853127135692e-06,
+      "loss": 0.6988,
+      "step": 9089
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2961918018820902,
+      "learning_rate": 3.926840772209714e-06,
+      "loss": 0.6262,
+      "step": 9090
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4588204971638412,
+      "learning_rate": 3.92582846339978e-06,
+      "loss": 0.6871,
+      "step": 9091
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4741248839516008,
+      "learning_rate": 3.924816200749387e-06,
+      "loss": 0.6272,
+      "step": 9092
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2882416423392562,
+      "learning_rate": 3.92380398430204e-06,
+      "loss": 0.6225,
+      "step": 9093
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3831657593605835,
+      "learning_rate": 3.92279181410123e-06,
+      "loss": 0.6633,
+      "step": 9094
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.355324064570219,
+      "learning_rate": 3.921779690190454e-06,
+      "loss": 0.6883,
+      "step": 9095
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.239988685395991,
+      "learning_rate": 3.920767612613209e-06,
+      "loss": 0.592,
+      "step": 9096
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.284919627189233,
+      "learning_rate": 3.919755581412982e-06,
+      "loss": 0.7756,
+      "step": 9097
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3713537692875974,
+      "learning_rate": 3.918743596633262e-06,
+      "loss": 0.6641,
+      "step": 9098
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.9544744320401275,
+      "learning_rate": 3.917731658317538e-06,
+      "loss": 0.5668,
+      "step": 9099
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2748817650703137,
+      "learning_rate": 3.916719766509297e-06,
+      "loss": 0.6382,
+      "step": 9100
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.213910686417965,
+      "learning_rate": 3.91570792125202e-06,
+      "loss": 0.6022,
+      "step": 9101
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2229715444574976,
+      "learning_rate": 3.914696122589187e-06,
+      "loss": 0.684,
+      "step": 9102
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.340665569185,
+      "learning_rate": 3.91368437056428e-06,
+      "loss": 0.6392,
+      "step": 9103
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3143252356503519,
+      "learning_rate": 3.912672665220773e-06,
+      "loss": 0.6298,
+      "step": 9104
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8736378992570673,
+      "learning_rate": 3.911661006602144e-06,
+      "loss": 0.6863,
+      "step": 9105
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9810313881246232,
+      "learning_rate": 3.910649394751867e-06,
+      "loss": 0.6192,
+      "step": 9106
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5241418336124002,
+      "learning_rate": 3.9096378297134115e-06,
+      "loss": 0.7112,
+      "step": 9107
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.314324644183073,
+      "learning_rate": 3.908626311530245e-06,
+      "loss": 0.6366,
+      "step": 9108
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3723074736323047,
+      "learning_rate": 3.907614840245836e-06,
+      "loss": 0.7031,
+      "step": 9109
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2289673062909992,
+      "learning_rate": 3.906603415903652e-06,
+      "loss": 0.5518,
+      "step": 9110
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3268115384594175,
+      "learning_rate": 3.905592038547155e-06,
+      "loss": 0.6482,
+      "step": 9111
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5277749111923806,
+      "learning_rate": 3.904580708219804e-06,
+      "loss": 0.6813,
+      "step": 9112
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3532830475605033,
+      "learning_rate": 3.903569424965059e-06,
+      "loss": 0.6759,
+      "step": 9113
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0129057164860946,
+      "learning_rate": 3.90255818882638e-06,
+      "loss": 0.6375,
+      "step": 9114
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2972587223572856,
+      "learning_rate": 3.901546999847217e-06,
+      "loss": 0.6045,
+      "step": 9115
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4250139501703407,
+      "learning_rate": 3.9005358580710275e-06,
+      "loss": 0.5924,
+      "step": 9116
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.431427927324773,
+      "learning_rate": 3.89952476354126e-06,
+      "loss": 0.7198,
+      "step": 9117
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4188775355362473,
+      "learning_rate": 3.898513716301366e-06,
+      "loss": 0.5624,
+      "step": 9118
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2906373156748983,
+      "learning_rate": 3.897502716394789e-06,
+      "loss": 0.5812,
+      "step": 9119
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2573604012970478,
+      "learning_rate": 3.896491763864976e-06,
+      "loss": 0.6055,
+      "step": 9120
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6841798649782416,
+      "learning_rate": 3.89548085875537e-06,
+      "loss": 0.6323,
+      "step": 9121
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3788690936459362,
+      "learning_rate": 3.894470001109412e-06,
+      "loss": 0.6288,
+      "step": 9122
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7807813626970326,
+      "learning_rate": 3.8934591909705385e-06,
+      "loss": 0.6588,
+      "step": 9123
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3073604143272552,
+      "learning_rate": 3.892448428382189e-06,
+      "loss": 0.633,
+      "step": 9124
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.427765276832121,
+      "learning_rate": 3.8914377133877976e-06,
+      "loss": 0.7174,
+      "step": 9125
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5196247614373404,
+      "learning_rate": 3.890427046030796e-06,
+      "loss": 0.6249,
+      "step": 9126
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2260264049855027,
+      "learning_rate": 3.889416426354614e-06,
+      "loss": 0.6286,
+      "step": 9127
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3654081531545645,
+      "learning_rate": 3.888405854402684e-06,
+      "loss": 0.6507,
+      "step": 9128
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5484346869928192,
+      "learning_rate": 3.887395330218429e-06,
+      "loss": 0.6121,
+      "step": 9129
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3477707922270008,
+      "learning_rate": 3.886384853845274e-06,
+      "loss": 0.6858,
+      "step": 9130
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3524345915341316,
+      "learning_rate": 3.885374425326642e-06,
+      "loss": 0.6329,
+      "step": 9131
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3667942258590748,
+      "learning_rate": 3.884364044705955e-06,
+      "loss": 0.6646,
+      "step": 9132
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4924902792687,
+      "learning_rate": 3.883353712026629e-06,
+      "loss": 0.7263,
+      "step": 9133
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3506847749481061,
+      "learning_rate": 3.8823434273320794e-06,
+      "loss": 0.5545,
+      "step": 9134
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4272378597469195,
+      "learning_rate": 3.881333190665723e-06,
+      "loss": 0.67,
+      "step": 9135
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.123749494536231,
+      "learning_rate": 3.880323002070971e-06,
+      "loss": 0.6748,
+      "step": 9136
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3471585429168538,
+      "learning_rate": 3.879312861591233e-06,
+      "loss": 0.6936,
+      "step": 9137
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.170309021914051,
+      "learning_rate": 3.878302769269917e-06,
+      "loss": 0.6673,
+      "step": 9138
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.293632868508903,
+      "learning_rate": 3.877292725150429e-06,
+      "loss": 0.585,
+      "step": 9139
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2580297982367716,
+      "learning_rate": 3.876282729276174e-06,
+      "loss": 0.564,
+      "step": 9140
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4677649401251363,
+      "learning_rate": 3.875272781690552e-06,
+      "loss": 0.5673,
+      "step": 9141
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7490487593407904,
+      "learning_rate": 3.8742628824369624e-06,
+      "loss": 0.5595,
+      "step": 9142
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2966295054099968,
+      "learning_rate": 3.8732530315588065e-06,
+      "loss": 0.7,
+      "step": 9143
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2750442617635527,
+      "learning_rate": 3.872243229099476e-06,
+      "loss": 0.5898,
+      "step": 9144
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3999545459909426,
+      "learning_rate": 3.871233475102365e-06,
+      "loss": 0.6626,
+      "step": 9145
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0845474147341398,
+      "learning_rate": 3.870223769610865e-06,
+      "loss": 0.567,
+      "step": 9146
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5273336837669274,
+      "learning_rate": 3.869214112668368e-06,
+      "loss": 0.6186,
+      "step": 9147
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3629368126104302,
+      "learning_rate": 3.868204504318258e-06,
+      "loss": 0.6693,
+      "step": 9148
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.439502436108344,
+      "learning_rate": 3.86719494460392e-06,
+      "loss": 0.6308,
+      "step": 9149
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5382086457857946,
+      "learning_rate": 3.866185433568741e-06,
+      "loss": 0.6279,
+      "step": 9150
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4593523439680887,
+      "learning_rate": 3.865175971256095e-06,
+      "loss": 0.5829,
+      "step": 9151
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1021726920686619,
+      "learning_rate": 3.864166557709368e-06,
+      "loss": 0.5509,
+      "step": 9152
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1386643727239667,
+      "learning_rate": 3.863157192971932e-06,
+      "loss": 0.6019,
+      "step": 9153
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.251118935556418,
+      "learning_rate": 3.8621478770871645e-06,
+      "loss": 0.5627,
+      "step": 9154
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5896536257432752,
+      "learning_rate": 3.861138610098435e-06,
+      "loss": 0.7502,
+      "step": 9155
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3638030157953553,
+      "learning_rate": 3.8601293920491165e-06,
+      "loss": 0.6421,
+      "step": 9156
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3478873358949466,
+      "learning_rate": 3.859120222982578e-06,
+      "loss": 0.6077,
+      "step": 9157
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1953422841210501,
+      "learning_rate": 3.858111102942184e-06,
+      "loss": 0.6061,
+      "step": 9158
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8346848802894546,
+      "learning_rate": 3.857102031971298e-06,
+      "loss": 0.75,
+      "step": 9159
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3726305752445662,
+      "learning_rate": 3.856093010113282e-06,
+      "loss": 0.6194,
+      "step": 9160
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4738595773089862,
+      "learning_rate": 3.855084037411501e-06,
+      "loss": 0.7184,
+      "step": 9161
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.183776089313033,
+      "learning_rate": 3.854075113909307e-06,
+      "loss": 0.5974,
+      "step": 9162
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.520301632743247,
+      "learning_rate": 3.853066239650055e-06,
+      "loss": 0.768,
+      "step": 9163
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.184921562739033,
+      "learning_rate": 3.852057414677102e-06,
+      "loss": 0.5921,
+      "step": 9164
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.243116959277835,
+      "learning_rate": 3.851048639033801e-06,
+      "loss": 0.668,
+      "step": 9165
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3602158376665405,
+      "learning_rate": 3.850039912763496e-06,
+      "loss": 0.6691,
+      "step": 9166
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2346125783119521,
+      "learning_rate": 3.849031235909539e-06,
+      "loss": 0.6257,
+      "step": 9167
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4295971695459766,
+      "learning_rate": 3.848022608515273e-06,
+      "loss": 0.6278,
+      "step": 9168
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2824350722366005,
+      "learning_rate": 3.84701403062404e-06,
+      "loss": 0.6413,
+      "step": 9169
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.356773395329441,
+      "learning_rate": 3.846005502279182e-06,
+      "loss": 0.602,
+      "step": 9170
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3029418119696303,
+      "learning_rate": 3.844997023524038e-06,
+      "loss": 0.6368,
+      "step": 9171
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3962542141903773,
+      "learning_rate": 3.843988594401946e-06,
+      "loss": 0.7089,
+      "step": 9172
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.149484142918413,
+      "learning_rate": 3.842980214956236e-06,
+      "loss": 0.6494,
+      "step": 9173
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0807764413073158,
+      "learning_rate": 3.841971885230243e-06,
+      "loss": 0.6998,
+      "step": 9174
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8552819992893226,
+      "learning_rate": 3.840963605267298e-06,
+      "loss": 0.6288,
+      "step": 9175
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1214730119769947,
+      "learning_rate": 3.83995537511073e-06,
+      "loss": 0.6909,
+      "step": 9176
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4438456946623788,
+      "learning_rate": 3.83894719480386e-06,
+      "loss": 0.6295,
+      "step": 9177
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.350159560700741,
+      "learning_rate": 3.837939064390016e-06,
+      "loss": 0.6065,
+      "step": 9178
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0514077174408196,
+      "learning_rate": 3.83693098391252e-06,
+      "loss": 0.6078,
+      "step": 9179
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4138905043430707,
+      "learning_rate": 3.835922953414688e-06,
+      "loss": 0.5746,
+      "step": 9180
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3372391045090213,
+      "learning_rate": 3.83491497293984e-06,
+      "loss": 0.6143,
+      "step": 9181
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5338374174820657,
+      "learning_rate": 3.833907042531289e-06,
+      "loss": 0.6424,
+      "step": 9182
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5883867702920418,
+      "learning_rate": 3.832899162232352e-06,
+      "loss": 0.6812,
+      "step": 9183
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3318031925753104,
+      "learning_rate": 3.8318913320863355e-06,
+      "loss": 0.644,
+      "step": 9184
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2340075979953355,
+      "learning_rate": 3.830883552136551e-06,
+      "loss": 0.6058,
+      "step": 9185
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6045945236330352,
+      "learning_rate": 3.829875822426304e-06,
+      "loss": 0.7548,
+      "step": 9186
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4770378853808561,
+      "learning_rate": 3.828868142998898e-06,
+      "loss": 0.5542,
+      "step": 9187
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1593392263673574,
+      "learning_rate": 3.827860513897635e-06,
+      "loss": 0.5349,
+      "step": 9188
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4125105028603149,
+      "learning_rate": 3.826852935165818e-06,
+      "loss": 0.7042,
+      "step": 9189
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4950449734873532,
+      "learning_rate": 3.825845406846743e-06,
+      "loss": 0.6645,
+      "step": 9190
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.115208390971038,
+      "learning_rate": 3.8248379289837065e-06,
+      "loss": 0.7748,
+      "step": 9191
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.251666909476321,
+      "learning_rate": 3.8238305016199995e-06,
+      "loss": 0.5854,
+      "step": 9192
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7970978321078195,
+      "learning_rate": 3.822823124798915e-06,
+      "loss": 0.6108,
+      "step": 9193
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4868469634942811,
+      "learning_rate": 3.8218157985637465e-06,
+      "loss": 0.7144,
+      "step": 9194
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6630009978759974,
+      "learning_rate": 3.820808522957774e-06,
+      "loss": 0.6328,
+      "step": 9195
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2953574768665432,
+      "learning_rate": 3.819801298024286e-06,
+      "loss": 0.5619,
+      "step": 9196
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3454882590298285,
+      "learning_rate": 3.818794123806565e-06,
+      "loss": 0.6536,
+      "step": 9197
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9159544212494335,
+      "learning_rate": 3.81778700034789e-06,
+      "loss": 0.6002,
+      "step": 9198
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1217901503549415,
+      "learning_rate": 3.816779927691542e-06,
+      "loss": 0.6617,
+      "step": 9199
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7912318938654945,
+      "learning_rate": 3.815772905880794e-06,
+      "loss": 0.7099,
+      "step": 9200
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1577730495636769,
+      "learning_rate": 3.814765934958924e-06,
+      "loss": 0.5951,
+      "step": 9201
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4203107739290197,
+      "learning_rate": 3.8137590149691992e-06,
+      "loss": 0.7024,
+      "step": 9202
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4274270044193298,
+      "learning_rate": 3.8127521459548934e-06,
+      "loss": 0.6104,
+      "step": 9203
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2730741382230468,
+      "learning_rate": 3.811745327959271e-06,
+      "loss": 0.5901,
+      "step": 9204
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4727704363022973,
+      "learning_rate": 3.810738561025599e-06,
+      "loss": 0.8091,
+      "step": 9205
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1296061664628936,
+      "learning_rate": 3.8097318451971375e-06,
+      "loss": 0.516,
+      "step": 9206
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2398118913897358,
+      "learning_rate": 3.80872518051715e-06,
+      "loss": 0.6469,
+      "step": 9207
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1422085578808752,
+      "learning_rate": 3.807718567028897e-06,
+      "loss": 0.6759,
+      "step": 9208
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.30286192469618,
+      "learning_rate": 3.8067120047756313e-06,
+      "loss": 0.6878,
+      "step": 9209
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5504432657417868,
+      "learning_rate": 3.8057054938006072e-06,
+      "loss": 0.6022,
+      "step": 9210
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7767334174358405,
+      "learning_rate": 3.804699034147078e-06,
+      "loss": 0.6213,
+      "step": 9211
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3116549622400657,
+      "learning_rate": 3.803692625858295e-06,
+      "loss": 0.5689,
+      "step": 9212
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4678377973190384,
+      "learning_rate": 3.8026862689775033e-06,
+      "loss": 0.6391,
+      "step": 9213
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9188033073676288,
+      "learning_rate": 3.801679963547949e-06,
+      "loss": 0.5856,
+      "step": 9214
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2586186526618415,
+      "learning_rate": 3.8006737096128763e-06,
+      "loss": 0.677,
+      "step": 9215
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.283230452906643,
+      "learning_rate": 3.7996675072155243e-06,
+      "loss": 0.5175,
+      "step": 9216
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4712578187416654,
+      "learning_rate": 3.7986613563991327e-06,
+      "loss": 0.7143,
+      "step": 9217
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2087637240671036,
+      "learning_rate": 3.7976552572069385e-06,
+      "loss": 0.5567,
+      "step": 9218
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5452250623304413,
+      "learning_rate": 3.7966492096821773e-06,
+      "loss": 0.677,
+      "step": 9219
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2263747878437157,
+      "learning_rate": 3.795643213868077e-06,
+      "loss": 0.6199,
+      "step": 9220
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3091866699263646,
+      "learning_rate": 3.7946372698078704e-06,
+      "loss": 0.6917,
+      "step": 9221
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2208920830908108,
+      "learning_rate": 3.793631377544787e-06,
+      "loss": 0.5382,
+      "step": 9222
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0913285642775243,
+      "learning_rate": 3.7926255371220487e-06,
+      "loss": 0.616,
+      "step": 9223
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1876406000713189,
+      "learning_rate": 3.7916197485828793e-06,
+      "loss": 0.5889,
+      "step": 9224
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2168506915405215,
+      "learning_rate": 3.7906140119705006e-06,
+      "loss": 0.6961,
+      "step": 9225
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1158488468080496,
+      "learning_rate": 3.7896083273281324e-06,
+      "loss": 0.5615,
+      "step": 9226
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2098346794074435,
+      "learning_rate": 3.7886026946989896e-06,
+      "loss": 0.6026,
+      "step": 9227
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.611694411170343,
+      "learning_rate": 3.787597114126286e-06,
+      "loss": 0.5942,
+      "step": 9228
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1418495042882155,
+      "learning_rate": 3.786591585653235e-06,
+      "loss": 0.576,
+      "step": 9229
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4910527351538385,
+      "learning_rate": 3.785586109323047e-06,
+      "loss": 0.7031,
+      "step": 9230
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.180950301375558,
+      "learning_rate": 3.7845806851789275e-06,
+      "loss": 0.6447,
+      "step": 9231
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4299067518410782,
+      "learning_rate": 3.783575313264083e-06,
+      "loss": 0.701,
+      "step": 9232
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2412088740198617,
+      "learning_rate": 3.7825699936217183e-06,
+      "loss": 0.6493,
+      "step": 9233
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5106451571026702,
+      "learning_rate": 3.7815647262950293e-06,
+      "loss": 0.6591,
+      "step": 9234
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3580174554396136,
+      "learning_rate": 3.7805595113272186e-06,
+      "loss": 0.5949,
+      "step": 9235
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3696644646135483,
+      "learning_rate": 3.7795543487614815e-06,
+      "loss": 0.6406,
+      "step": 9236
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0043518916052405,
+      "learning_rate": 3.7785492386410138e-06,
+      "loss": 0.6386,
+      "step": 9237
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4443361816526117,
+      "learning_rate": 3.7775441810090043e-06,
+      "loss": 0.6536,
+      "step": 9238
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.541774263391616,
+      "learning_rate": 3.7765391759086424e-06,
+      "loss": 0.6505,
+      "step": 9239
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0852688272434967,
+      "learning_rate": 3.7755342233831188e-06,
+      "loss": 0.6373,
+      "step": 9240
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2630136437266306,
+      "learning_rate": 3.7745293234756164e-06,
+      "loss": 0.6494,
+      "step": 9241
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4452235681529872,
+      "learning_rate": 3.7735244762293183e-06,
+      "loss": 0.6234,
+      "step": 9242
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3808028648310664,
+      "learning_rate": 3.7725196816874045e-06,
+      "loss": 0.7339,
+      "step": 9243
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4496014163185922,
+      "learning_rate": 3.771514939893055e-06,
+      "loss": 0.6167,
+      "step": 9244
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.161076036943806,
+      "learning_rate": 3.7705102508894425e-06,
+      "loss": 0.5866,
+      "step": 9245
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2976312832693258,
+      "learning_rate": 3.769505614719744e-06,
+      "loss": 0.6063,
+      "step": 9246
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.201866612024004,
+      "learning_rate": 3.7685010314271287e-06,
+      "loss": 0.6091,
+      "step": 9247
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5037734888007526,
+      "learning_rate": 3.7674965010547686e-06,
+      "loss": 0.6564,
+      "step": 9248
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5416820217687421,
+      "learning_rate": 3.766492023645827e-06,
+      "loss": 0.6602,
+      "step": 9249
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1786324882695098,
+      "learning_rate": 3.7654875992434714e-06,
+      "loss": 0.599,
+      "step": 9250
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3557584732205348,
+      "learning_rate": 3.764483227890864e-06,
+      "loss": 0.641,
+      "step": 9251
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4408819862898659,
+      "learning_rate": 3.763478909631162e-06,
+      "loss": 0.714,
+      "step": 9252
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4469821957801752,
+      "learning_rate": 3.7624746445075243e-06,
+      "loss": 0.6821,
+      "step": 9253
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5966862320792046,
+      "learning_rate": 3.761470432563109e-06,
+      "loss": 0.6432,
+      "step": 9254
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.503958108037097,
+      "learning_rate": 3.760466273841068e-06,
+      "loss": 0.6001,
+      "step": 9255
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2329627607076044,
+      "learning_rate": 3.759462168384551e-06,
+      "loss": 0.5658,
+      "step": 9256
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0372012817212253,
+      "learning_rate": 3.758458116236707e-06,
+      "loss": 0.5106,
+      "step": 9257
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3351645967453416,
+      "learning_rate": 3.757454117440682e-06,
+      "loss": 0.5946,
+      "step": 9258
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7354564100871144,
+      "learning_rate": 3.7564501720396242e-06,
+      "loss": 0.6058,
+      "step": 9259
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.204454905939865,
+      "learning_rate": 3.75544628007667e-06,
+      "loss": 0.7599,
+      "step": 9260
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4395468651553311,
+      "learning_rate": 3.75444244159496e-06,
+      "loss": 0.6422,
+      "step": 9261
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2267405350694114,
+      "learning_rate": 3.7534386566376345e-06,
+      "loss": 0.5322,
+      "step": 9262
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.098184949466445,
+      "learning_rate": 3.7524349252478243e-06,
+      "loss": 0.6795,
+      "step": 9263
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2876363862398708,
+      "learning_rate": 3.7514312474686643e-06,
+      "loss": 0.685,
+      "step": 9264
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9749148253972445,
+      "learning_rate": 3.7504276233432835e-06,
+      "loss": 0.6203,
+      "step": 9265
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4273027486622016,
+      "learning_rate": 3.749424052914813e-06,
+      "loss": 0.7868,
+      "step": 9266
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5167580098002234,
+      "learning_rate": 3.7484205362263737e-06,
+      "loss": 0.7786,
+      "step": 9267
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4651179646099155,
+      "learning_rate": 3.747417073321092e-06,
+      "loss": 0.7115,
+      "step": 9268
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3741820231769288,
+      "learning_rate": 3.74641366424209e-06,
+      "loss": 0.7522,
+      "step": 9269
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1872662787990653,
+      "learning_rate": 3.7454103090324835e-06,
+      "loss": 0.6279,
+      "step": 9270
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0576965630417217,
+      "learning_rate": 3.7444070077353896e-06,
+      "loss": 0.6218,
+      "step": 9271
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.361179160202802,
+      "learning_rate": 3.743403760393922e-06,
+      "loss": 0.648,
+      "step": 9272
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6633315133905706,
+      "learning_rate": 3.7424005670511964e-06,
+      "loss": 0.7227,
+      "step": 9273
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4675588486976046,
+      "learning_rate": 3.7413974277503183e-06,
+      "loss": 0.5417,
+      "step": 9274
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0412318522722457,
+      "learning_rate": 3.740394342534394e-06,
+      "loss": 0.4759,
+      "step": 9275
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4475778493979237,
+      "learning_rate": 3.739391311446531e-06,
+      "loss": 0.7622,
+      "step": 9276
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1762635350015236,
+      "learning_rate": 3.738388334529832e-06,
+      "loss": 0.6628,
+      "step": 9277
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3159133887036742,
+      "learning_rate": 3.7373854118273955e-06,
+      "loss": 0.6305,
+      "step": 9278
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4509198570454547,
+      "learning_rate": 3.7363825433823187e-06,
+      "loss": 0.718,
+      "step": 9279
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0692227092654993,
+      "learning_rate": 3.7353797292377002e-06,
+      "loss": 0.6792,
+      "step": 9280
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1509938649219138,
+      "learning_rate": 3.7343769694366294e-06,
+      "loss": 0.6956,
+      "step": 9281
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4642632126536468,
+      "learning_rate": 3.7333742640221994e-06,
+      "loss": 0.6544,
+      "step": 9282
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3487369927286164,
+      "learning_rate": 3.732371613037498e-06,
+      "loss": 0.6496,
+      "step": 9283
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5180300104434292,
+      "learning_rate": 3.7313690165256134e-06,
+      "loss": 0.6453,
+      "step": 9284
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3443783736104997,
+      "learning_rate": 3.730366474529625e-06,
+      "loss": 0.5943,
+      "step": 9285
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2984265380576663,
+      "learning_rate": 3.7293639870926173e-06,
+      "loss": 0.5567,
+      "step": 9286
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2036496062828559,
+      "learning_rate": 3.728361554257671e-06,
+      "loss": 0.6209,
+      "step": 9287
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3711093069081781,
+      "learning_rate": 3.7273591760678594e-06,
+      "loss": 0.5937,
+      "step": 9288
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6890864763893296,
+      "learning_rate": 3.7263568525662574e-06,
+      "loss": 0.6064,
+      "step": 9289
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2566656453533909,
+      "learning_rate": 3.7253545837959383e-06,
+      "loss": 0.6508,
+      "step": 9290
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.695007242682219,
+      "learning_rate": 3.7243523697999733e-06,
+      "loss": 0.8543,
+      "step": 9291
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2968095047411683,
+      "learning_rate": 3.723350210621426e-06,
+      "loss": 0.6305,
+      "step": 9292
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3337778420137851,
+      "learning_rate": 3.722348106303364e-06,
+      "loss": 0.9068,
+      "step": 9293
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4539606332041066,
+      "learning_rate": 3.7213460568888493e-06,
+      "loss": 0.8838,
+      "step": 9294
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.092697741549937,
+      "learning_rate": 3.7203440624209438e-06,
+      "loss": 0.6627,
+      "step": 9295
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2584416976781505,
+      "learning_rate": 3.7193421229427017e-06,
+      "loss": 0.6095,
+      "step": 9296
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4158646261592058,
+      "learning_rate": 3.7183402384971818e-06,
+      "loss": 0.6712,
+      "step": 9297
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5407508339934604,
+      "learning_rate": 3.7173384091274372e-06,
+      "loss": 0.6476,
+      "step": 9298
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.350041127840252,
+      "learning_rate": 3.716336634876516e-06,
+      "loss": 0.6678,
+      "step": 9299
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4387667379330462,
+      "learning_rate": 3.7153349157874687e-06,
+      "loss": 0.6675,
+      "step": 9300
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3985116097413124,
+      "learning_rate": 3.7143332519033413e-06,
+      "loss": 0.6095,
+      "step": 9301
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.13894418572736,
+      "learning_rate": 3.7133316432671794e-06,
+      "loss": 0.5191,
+      "step": 9302
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.292779113242487,
+      "learning_rate": 3.7123300899220193e-06,
+      "loss": 0.5523,
+      "step": 9303
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2771628642182071,
+      "learning_rate": 3.711328591910904e-06,
+      "loss": 0.6734,
+      "step": 9304
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4386965365480955,
+      "learning_rate": 3.710327149276871e-06,
+      "loss": 0.6085,
+      "step": 9305
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5036076099824387,
+      "learning_rate": 3.70932576206295e-06,
+      "loss": 0.64,
+      "step": 9306
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4953528676760655,
+      "learning_rate": 3.708324430312177e-06,
+      "loss": 0.567,
+      "step": 9307
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3247832265438737,
+      "learning_rate": 3.7073231540675792e-06,
+      "loss": 0.7545,
+      "step": 9308
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4928495873632595,
+      "learning_rate": 3.7063219333721857e-06,
+      "loss": 0.6584,
+      "step": 9309
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9867413702712164,
+      "learning_rate": 3.7053207682690184e-06,
+      "loss": 0.5433,
+      "step": 9310
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3578269786138621,
+      "learning_rate": 3.704319658801102e-06,
+      "loss": 0.6586,
+      "step": 9311
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.327431345512031,
+      "learning_rate": 3.7033186050114547e-06,
+      "loss": 0.7192,
+      "step": 9312
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5535324448243435,
+      "learning_rate": 3.7023176069430965e-06,
+      "loss": 0.7419,
+      "step": 9313
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4472284621909237,
+      "learning_rate": 3.7013166646390384e-06,
+      "loss": 0.7043,
+      "step": 9314
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2208371700779053,
+      "learning_rate": 3.700315778142297e-06,
+      "loss": 0.6053,
+      "step": 9315
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5977860918012143,
+      "learning_rate": 3.699314947495881e-06,
+      "loss": 0.6487,
+      "step": 9316
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.299494285712963,
+      "learning_rate": 3.698314172742799e-06,
+      "loss": 0.5488,
+      "step": 9317
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4754053545580765,
+      "learning_rate": 3.6973134539260545e-06,
+      "loss": 0.6451,
+      "step": 9318
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.873665571324839,
+      "learning_rate": 3.6963127910886526e-06,
+      "loss": 0.703,
+      "step": 9319
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.56893082590164,
+      "learning_rate": 3.695312184273595e-06,
+      "loss": 0.6679,
+      "step": 9320
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2951313456327986,
+      "learning_rate": 3.694311633523878e-06,
+      "loss": 0.6476,
+      "step": 9321
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.90357508672841,
+      "learning_rate": 3.693311138882497e-06,
+      "loss": 0.6124,
+      "step": 9322
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2866029882891943,
+      "learning_rate": 3.6923107003924486e-06,
+      "loss": 0.655,
+      "step": 9323
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.298685377013262,
+      "learning_rate": 3.691310318096719e-06,
+      "loss": 0.5617,
+      "step": 9324
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2470981300071802,
+      "learning_rate": 3.690309992038301e-06,
+      "loss": 0.5309,
+      "step": 9325
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1366111865652775,
+      "learning_rate": 3.689309722260179e-06,
+      "loss": 0.498,
+      "step": 9326
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0318348284625607,
+      "learning_rate": 3.6883095088053387e-06,
+      "loss": 0.6913,
+      "step": 9327
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3718434361466183,
+      "learning_rate": 3.6873093517167584e-06,
+      "loss": 0.6267,
+      "step": 9328
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2450242306911554,
+      "learning_rate": 3.6863092510374198e-06,
+      "loss": 0.5946,
+      "step": 9329
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3730577069238465,
+      "learning_rate": 3.685309206810297e-06,
+      "loss": 0.6698,
+      "step": 9330
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3236973274775945,
+      "learning_rate": 3.684309219078368e-06,
+      "loss": 0.6366,
+      "step": 9331
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2855545560711108,
+      "learning_rate": 3.6833092878846e-06,
+      "loss": 0.5952,
+      "step": 9332
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6381360291189182,
+      "learning_rate": 3.6823094132719645e-06,
+      "loss": 0.5944,
+      "step": 9333
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2195454395950773,
+      "learning_rate": 3.68130959528343e-06,
+      "loss": 0.6112,
+      "step": 9334
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4934909544827404,
+      "learning_rate": 3.680309833961958e-06,
+      "loss": 0.6409,
+      "step": 9335
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2308863539399706,
+      "learning_rate": 3.6793101293505106e-06,
+      "loss": 0.5853,
+      "step": 9336
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.202885768603601,
+      "learning_rate": 3.678310481492048e-06,
+      "loss": 0.7591,
+      "step": 9337
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.443125565783363,
+      "learning_rate": 3.6773108904295294e-06,
+      "loss": 0.6825,
+      "step": 9338
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6054341117022293,
+      "learning_rate": 3.6763113562059077e-06,
+      "loss": 0.7126,
+      "step": 9339
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9273874731064715,
+      "learning_rate": 3.6753118788641327e-06,
+      "loss": 0.7704,
+      "step": 9340
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3733750693149194,
+      "learning_rate": 3.67431245844716e-06,
+      "loss": 0.6625,
+      "step": 9341
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6950563156142744,
+      "learning_rate": 3.67331309499793e-06,
+      "loss": 0.7748,
+      "step": 9342
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9157076388165812,
+      "learning_rate": 3.672313788559392e-06,
+      "loss": 0.726,
+      "step": 9343
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3263581229186918,
+      "learning_rate": 3.6713145391744877e-06,
+      "loss": 0.6464,
+      "step": 9344
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1981479269494444,
+      "learning_rate": 3.6703153468861585e-06,
+      "loss": 0.6059,
+      "step": 9345
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2554346076066187,
+      "learning_rate": 3.6693162117373377e-06,
+      "loss": 0.5651,
+      "step": 9346
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3551424515947474,
+      "learning_rate": 3.668317133770963e-06,
+      "loss": 0.6765,
+      "step": 9347
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.370309622399288,
+      "learning_rate": 3.667318113029968e-06,
+      "loss": 0.6906,
+      "step": 9348
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3663047403814366,
+      "learning_rate": 3.6663191495572827e-06,
+      "loss": 0.6925,
+      "step": 9349
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2272956288749506,
+      "learning_rate": 3.6653202433958313e-06,
+      "loss": 0.5682,
+      "step": 9350
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4737932731964705,
+      "learning_rate": 3.664321394588542e-06,
+      "loss": 0.6556,
+      "step": 9351
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7024098324630872,
+      "learning_rate": 3.663322603178339e-06,
+      "loss": 0.7564,
+      "step": 9352
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2236800057503412,
+      "learning_rate": 3.66232386920814e-06,
+      "loss": 0.4953,
+      "step": 9353
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3698574826981487,
+      "learning_rate": 3.661325192720862e-06,
+      "loss": 0.6773,
+      "step": 9354
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2138362218163112,
+      "learning_rate": 3.660326573759422e-06,
+      "loss": 0.6098,
+      "step": 9355
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.635619694329298,
+      "learning_rate": 3.659328012366735e-06,
+      "loss": 0.675,
+      "step": 9356
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3207639993892053,
+      "learning_rate": 3.658329508585707e-06,
+      "loss": 0.6567,
+      "step": 9357
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4338106478979713,
+      "learning_rate": 3.657331062459249e-06,
+      "loss": 0.681,
+      "step": 9358
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.074898838367308,
+      "learning_rate": 3.6563326740302664e-06,
+      "loss": 0.4772,
+      "step": 9359
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.340040223496613,
+      "learning_rate": 3.6553343433416598e-06,
+      "loss": 0.6154,
+      "step": 9360
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4486098590847736,
+      "learning_rate": 3.6543360704363305e-06,
+      "loss": 0.7081,
+      "step": 9361
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0985160926826276,
+      "learning_rate": 3.6533378553571786e-06,
+      "loss": 0.8505,
+      "step": 9362
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.355893620640527,
+      "learning_rate": 3.6523396981470995e-06,
+      "loss": 0.9005,
+      "step": 9363
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2253538965161956,
+      "learning_rate": 3.6513415988489824e-06,
+      "loss": 0.6948,
+      "step": 9364
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.533650344987133,
+      "learning_rate": 3.6503435575057205e-06,
+      "loss": 0.6236,
+      "step": 9365
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4901207381426071,
+      "learning_rate": 3.6493455741602035e-06,
+      "loss": 0.6889,
+      "step": 9366
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4534038009234145,
+      "learning_rate": 3.6483476488553143e-06,
+      "loss": 0.696,
+      "step": 9367
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.241088367707276,
+      "learning_rate": 3.6473497816339375e-06,
+      "loss": 0.6042,
+      "step": 9368
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.199201381614752,
+      "learning_rate": 3.6463519725389516e-06,
+      "loss": 0.5898,
+      "step": 9369
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.679329401174984,
+      "learning_rate": 3.645354221613239e-06,
+      "loss": 0.6109,
+      "step": 9370
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2548401955803787,
+      "learning_rate": 3.6443565288996698e-06,
+      "loss": 0.6216,
+      "step": 9371
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3077204705321832,
+      "learning_rate": 3.6433588944411207e-06,
+      "loss": 0.6345,
+      "step": 9372
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1889009216141504,
+      "learning_rate": 3.642361318280461e-06,
+      "loss": 0.546,
+      "step": 9373
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0579497116110528,
+      "learning_rate": 3.64136380046056e-06,
+      "loss": 0.6125,
+      "step": 9374
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3559984584541485,
+      "learning_rate": 3.640366341024281e-06,
+      "loss": 0.6911,
+      "step": 9375
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3248363866960537,
+      "learning_rate": 3.639368940014489e-06,
+      "loss": 0.6288,
+      "step": 9376
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3128567741361759,
+      "learning_rate": 3.638371597474044e-06,
+      "loss": 0.6063,
+      "step": 9377
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1099483801079286,
+      "learning_rate": 3.637374313445805e-06,
+      "loss": 0.6542,
+      "step": 9378
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3609972790128002,
+      "learning_rate": 3.6363770879726247e-06,
+      "loss": 0.6332,
+      "step": 9379
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1182164072825869,
+      "learning_rate": 3.635379921097359e-06,
+      "loss": 0.5787,
+      "step": 9380
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2958320696998178,
+      "learning_rate": 3.634382812862857e-06,
+      "loss": 0.6587,
+      "step": 9381
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.298532637321419,
+      "learning_rate": 3.6333857633119673e-06,
+      "loss": 0.5419,
+      "step": 9382
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3296134258647605,
+      "learning_rate": 3.632388772487534e-06,
+      "loss": 0.5842,
+      "step": 9383
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1885627860704682,
+      "learning_rate": 3.6313918404324e-06,
+      "loss": 0.5823,
+      "step": 9384
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.592732565491914,
+      "learning_rate": 3.63039496718941e-06,
+      "loss": 0.6441,
+      "step": 9385
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4085262055152779,
+      "learning_rate": 3.6293981528013965e-06,
+      "loss": 0.6553,
+      "step": 9386
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2943055756602995,
+      "learning_rate": 3.6284013973111962e-06,
+      "loss": 0.6275,
+      "step": 9387
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1641695308449882,
+      "learning_rate": 3.627404700761645e-06,
+      "loss": 0.5768,
+      "step": 9388
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3539747012859051,
+      "learning_rate": 3.6264080631955683e-06,
+      "loss": 0.6204,
+      "step": 9389
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3267985459958107,
+      "learning_rate": 3.6254114846557977e-06,
+      "loss": 0.6342,
+      "step": 9390
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.247922952724555,
+      "learning_rate": 3.624414965185156e-06,
+      "loss": 0.5809,
+      "step": 9391
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3311303612990695,
+      "learning_rate": 3.623418504826468e-06,
+      "loss": 0.6335,
+      "step": 9392
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7194572851035088,
+      "learning_rate": 3.622422103622551e-06,
+      "loss": 0.7184,
+      "step": 9393
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.505516534840998,
+      "learning_rate": 3.621425761616224e-06,
+      "loss": 0.6671,
+      "step": 9394
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.420522677765055,
+      "learning_rate": 3.6204294788503034e-06,
+      "loss": 0.6299,
+      "step": 9395
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2609731820258956,
+      "learning_rate": 3.6194332553676026e-06,
+      "loss": 0.5515,
+      "step": 9396
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2514485025920639,
+      "learning_rate": 3.618437091210927e-06,
+      "loss": 0.5803,
+      "step": 9397
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2594194271836758,
+      "learning_rate": 3.617440986423086e-06,
+      "loss": 0.5887,
+      "step": 9398
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4508742253254079,
+      "learning_rate": 3.616444941046887e-06,
+      "loss": 0.6456,
+      "step": 9399
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1186566991550342,
+      "learning_rate": 3.615448955125129e-06,
+      "loss": 0.4813,
+      "step": 9400
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3559940832172748,
+      "learning_rate": 3.614453028700613e-06,
+      "loss": 0.7146,
+      "step": 9401
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.407501052049212,
+      "learning_rate": 3.613457161816135e-06,
+      "loss": 0.6662,
+      "step": 9402
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2784258967414488,
+      "learning_rate": 3.6124613545144937e-06,
+      "loss": 0.7353,
+      "step": 9403
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3478442122458592,
+      "learning_rate": 3.6114656068384767e-06,
+      "loss": 0.6705,
+      "step": 9404
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3176885548365842,
+      "learning_rate": 3.610469918830874e-06,
+      "loss": 0.584,
+      "step": 9405
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.436041920467215,
+      "learning_rate": 3.6094742905344753e-06,
+      "loss": 0.6291,
+      "step": 9406
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2517794106956974,
+      "learning_rate": 3.6084787219920616e-06,
+      "loss": 0.636,
+      "step": 9407
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.405712924983509,
+      "learning_rate": 3.6074832132464165e-06,
+      "loss": 0.6289,
+      "step": 9408
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7939398844219958,
+      "learning_rate": 3.6064877643403194e-06,
+      "loss": 0.736,
+      "step": 9409
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4377531056734036,
+      "learning_rate": 3.605492375316548e-06,
+      "loss": 0.6059,
+      "step": 9410
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4094913858625175,
+      "learning_rate": 3.604497046217873e-06,
+      "loss": 0.7582,
+      "step": 9411
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3670868365732252,
+      "learning_rate": 3.6035017770870672e-06,
+      "loss": 0.5941,
+      "step": 9412
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3089167253679825,
+      "learning_rate": 3.6025065679669025e-06,
+      "loss": 0.6813,
+      "step": 9413
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4668619810081935,
+      "learning_rate": 3.601511418900143e-06,
+      "loss": 0.6354,
+      "step": 9414
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2903472611251199,
+      "learning_rate": 3.600516329929551e-06,
+      "loss": 0.6881,
+      "step": 9415
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2049278683597802,
+      "learning_rate": 3.599521301097888e-06,
+      "loss": 0.7125,
+      "step": 9416
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4372271088618123,
+      "learning_rate": 3.598526332447917e-06,
+      "loss": 0.7311,
+      "step": 9417
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3466243747780995,
+      "learning_rate": 3.597531424022388e-06,
+      "loss": 0.6375,
+      "step": 9418
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1350169949211224,
+      "learning_rate": 3.5965365758640587e-06,
+      "loss": 0.4821,
+      "step": 9419
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1925686610062747,
+      "learning_rate": 3.5955417880156766e-06,
+      "loss": 0.5747,
+      "step": 9420
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3142610050714123,
+      "learning_rate": 3.594547060519995e-06,
+      "loss": 0.6198,
+      "step": 9421
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5176524569517074,
+      "learning_rate": 3.5935523934197537e-06,
+      "loss": 0.6546,
+      "step": 9422
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3339765442270723,
+      "learning_rate": 3.592557786757699e-06,
+      "loss": 0.6296,
+      "step": 9423
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4353261856480584,
+      "learning_rate": 3.591563240576572e-06,
+      "loss": 0.5552,
+      "step": 9424
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.089285200823832,
+      "learning_rate": 3.590568754919107e-06,
+      "loss": 0.6024,
+      "step": 9425
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2200834515365822,
+      "learning_rate": 3.5895743298280423e-06,
+      "loss": 0.5575,
+      "step": 9426
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5125273347846293,
+      "learning_rate": 3.5885799653461107e-06,
+      "loss": 0.6899,
+      "step": 9427
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1780433283759013,
+      "learning_rate": 3.587585661516042e-06,
+      "loss": 0.5425,
+      "step": 9428
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2966138945867947,
+      "learning_rate": 3.5865914183805606e-06,
+      "loss": 0.6072,
+      "step": 9429
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0076280352371216,
+      "learning_rate": 3.585597235982394e-06,
+      "loss": 0.6157,
+      "step": 9430
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1888842513678342,
+      "learning_rate": 3.584603114364266e-06,
+      "loss": 0.6632,
+      "step": 9431
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1927018165813743,
+      "learning_rate": 3.5836090535688944e-06,
+      "loss": 0.6159,
+      "step": 9432
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1945119941048838,
+      "learning_rate": 3.582615053638996e-06,
+      "loss": 0.6534,
+      "step": 9433
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5781157659066167,
+      "learning_rate": 3.581621114617284e-06,
+      "loss": 0.7564,
+      "step": 9434
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4204388370755636,
+      "learning_rate": 3.580627236546474e-06,
+      "loss": 0.5859,
+      "step": 9435
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5078027662101936,
+      "learning_rate": 3.5796334194692704e-06,
+      "loss": 0.7076,
+      "step": 9436
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2579418049372366,
+      "learning_rate": 3.578639663428384e-06,
+      "loss": 0.6339,
+      "step": 9437
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2758432880879975,
+      "learning_rate": 3.5776459684665146e-06,
+      "loss": 0.6405,
+      "step": 9438
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.396494203426516,
+      "learning_rate": 3.5766523346263682e-06,
+      "loss": 0.7568,
+      "step": 9439
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.383329991460469,
+      "learning_rate": 3.5756587619506388e-06,
+      "loss": 0.6465,
+      "step": 9440
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.547401262180491,
+      "learning_rate": 3.5746652504820263e-06,
+      "loss": 0.6515,
+      "step": 9441
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1111699681101699,
+      "learning_rate": 3.5736718002632222e-06,
+      "loss": 0.6042,
+      "step": 9442
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2231553451696726,
+      "learning_rate": 3.572678411336916e-06,
+      "loss": 0.6395,
+      "step": 9443
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1756909542950105,
+      "learning_rate": 3.571685083745798e-06,
+      "loss": 0.5052,
+      "step": 9444
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4144974406299542,
+      "learning_rate": 3.5706918175325523e-06,
+      "loss": 0.6099,
+      "step": 9445
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2690996656596116,
+      "learning_rate": 3.5696986127398646e-06,
+      "loss": 0.5858,
+      "step": 9446
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4965112815204915,
+      "learning_rate": 3.5687054694104117e-06,
+      "loss": 0.6238,
+      "step": 9447
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4528917612361203,
+      "learning_rate": 3.5677123875868713e-06,
+      "loss": 0.6453,
+      "step": 9448
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5651443971968948,
+      "learning_rate": 3.56671936731192e-06,
+      "loss": 0.6619,
+      "step": 9449
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5014982598117634,
+      "learning_rate": 3.5657264086282317e-06,
+      "loss": 0.6856,
+      "step": 9450
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.532319683373957,
+      "learning_rate": 3.564733511578472e-06,
+      "loss": 0.7781,
+      "step": 9451
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.379863799718221,
+      "learning_rate": 3.5637406762053094e-06,
+      "loss": 0.7573,
+      "step": 9452
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8804114112760117,
+      "learning_rate": 3.562747902551411e-06,
+      "loss": 0.6491,
+      "step": 9453
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1000851080338387,
+      "learning_rate": 3.561755190659434e-06,
+      "loss": 0.5561,
+      "step": 9454
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.328207837613791,
+      "learning_rate": 3.5607625405720402e-06,
+      "loss": 0.6209,
+      "step": 9455
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.50383478681722,
+      "learning_rate": 3.559769952331884e-06,
+      "loss": 0.6666,
+      "step": 9456
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0254842616558866,
+      "learning_rate": 3.5587774259816234e-06,
+      "loss": 0.6564,
+      "step": 9457
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2283443274643924,
+      "learning_rate": 3.557784961563904e-06,
+      "loss": 0.5728,
+      "step": 9458
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.081435249304261,
+      "learning_rate": 3.556792559121377e-06,
+      "loss": 0.7364,
+      "step": 9459
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4963555130607922,
+      "learning_rate": 3.5558002186966906e-06,
+      "loss": 0.7998,
+      "step": 9460
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5314002029861502,
+      "learning_rate": 3.554807940332483e-06,
+      "loss": 0.6021,
+      "step": 9461
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6502292284379165,
+      "learning_rate": 3.5538157240713964e-06,
+      "loss": 0.6381,
+      "step": 9462
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.577989011303428,
+      "learning_rate": 3.5528235699560697e-06,
+      "loss": 0.7908,
+      "step": 9463
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7050646392815274,
+      "learning_rate": 3.5518314780291384e-06,
+      "loss": 0.7384,
+      "step": 9464
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.541128426947966,
+      "learning_rate": 3.5508394483332333e-06,
+      "loss": 0.7023,
+      "step": 9465
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6360731817832594,
+      "learning_rate": 3.549847480910983e-06,
+      "loss": 0.7632,
+      "step": 9466
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5514229942939597,
+      "learning_rate": 3.548855575805016e-06,
+      "loss": 0.7035,
+      "step": 9467
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3474662857079283,
+      "learning_rate": 3.547863733057959e-06,
+      "loss": 0.6466,
+      "step": 9468
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2891263546801506,
+      "learning_rate": 3.5468719527124294e-06,
+      "loss": 0.6036,
+      "step": 9469
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1238066592815523,
+      "learning_rate": 3.5458802348110488e-06,
+      "loss": 0.7139,
+      "step": 9470
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3066267249375503,
+      "learning_rate": 3.544888579396435e-06,
+      "loss": 0.6083,
+      "step": 9471
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.315916840055942,
+      "learning_rate": 3.543896986511197e-06,
+      "loss": 0.6631,
+      "step": 9472
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.086562364518976,
+      "learning_rate": 3.5429054561979485e-06,
+      "loss": 0.5859,
+      "step": 9473
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2047882797663834,
+      "learning_rate": 3.541913988499299e-06,
+      "loss": 0.6289,
+      "step": 9474
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3208759709683757,
+      "learning_rate": 3.5409225834578536e-06,
+      "loss": 0.6846,
+      "step": 9475
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.341854492541095,
+      "learning_rate": 3.5399312411162124e-06,
+      "loss": 0.638,
+      "step": 9476
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2899637556931038,
+      "learning_rate": 3.5389399615169774e-06,
+      "loss": 0.7015,
+      "step": 9477
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.518842135393398,
+      "learning_rate": 3.5379487447027483e-06,
+      "loss": 0.711,
+      "step": 9478
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.808666822068244,
+      "learning_rate": 3.5369575907161167e-06,
+      "loss": 0.7883,
+      "step": 9479
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.580009605997032,
+      "learning_rate": 3.5359664995996746e-06,
+      "loss": 0.785,
+      "step": 9480
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4451805142769008,
+      "learning_rate": 3.5349754713960126e-06,
+      "loss": 0.5361,
+      "step": 9481
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2406868420663508,
+      "learning_rate": 3.5339845061477195e-06,
+      "loss": 0.5902,
+      "step": 9482
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7174949342831802,
+      "learning_rate": 3.5329936038973756e-06,
+      "loss": 0.6932,
+      "step": 9483
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.244726800897068,
+      "learning_rate": 3.5320027646875643e-06,
+      "loss": 0.6027,
+      "step": 9484
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5045964866143062,
+      "learning_rate": 3.5310119885608625e-06,
+      "loss": 0.6975,
+      "step": 9485
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5891837388237737,
+      "learning_rate": 3.5300212755598497e-06,
+      "loss": 0.641,
+      "step": 9486
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4217462531512874,
+      "learning_rate": 3.5290306257270935e-06,
+      "loss": 0.9112,
+      "step": 9487
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.356705352203957,
+      "learning_rate": 3.5280400391051693e-06,
+      "loss": 0.684,
+      "step": 9488
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3211472399545936,
+      "learning_rate": 3.5270495157366434e-06,
+      "loss": 0.664,
+      "step": 9489
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3678401726198497,
+      "learning_rate": 3.5260590556640783e-06,
+      "loss": 0.5701,
+      "step": 9490
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3778357228057942,
+      "learning_rate": 3.5250686589300387e-06,
+      "loss": 0.6285,
+      "step": 9491
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.68242038535669,
+      "learning_rate": 3.524078325577084e-06,
+      "loss": 0.6282,
+      "step": 9492
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5273156546917426,
+      "learning_rate": 3.5230880556477707e-06,
+      "loss": 0.7625,
+      "step": 9493
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3501303623066858,
+      "learning_rate": 3.5220978491846534e-06,
+      "loss": 0.6915,
+      "step": 9494
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9342676323764266,
+      "learning_rate": 3.521107706230282e-06,
+      "loss": 0.6263,
+      "step": 9495
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2119884259510734,
+      "learning_rate": 3.5201176268272058e-06,
+      "loss": 0.6146,
+      "step": 9496
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2841843056502331,
+      "learning_rate": 3.519127611017973e-06,
+      "loss": 0.4897,
+      "step": 9497
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.1020978291090304,
+      "learning_rate": 3.5181376588451244e-06,
+      "loss": 0.692,
+      "step": 9498
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2527806291849974,
+      "learning_rate": 3.517147770351199e-06,
+      "loss": 0.6662,
+      "step": 9499
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3951988169032745,
+      "learning_rate": 3.5161579455787388e-06,
+      "loss": 0.7272,
+      "step": 9500
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.839684803827955,
+      "learning_rate": 3.5151681845702747e-06,
+      "loss": 0.6314,
+      "step": 9501
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2539307388897636,
+      "learning_rate": 3.5141784873683418e-06,
+      "loss": 0.5758,
+      "step": 9502
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4613503642438093,
+      "learning_rate": 3.513188854015466e-06,
+      "loss": 0.653,
+      "step": 9503
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2038044362742384,
+      "learning_rate": 3.5121992845541797e-06,
+      "loss": 0.8175,
+      "step": 9504
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2024090975309953,
+      "learning_rate": 3.5112097790270017e-06,
+      "loss": 0.5387,
+      "step": 9505
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2823082670110346,
+      "learning_rate": 3.5102203374764555e-06,
+      "loss": 0.6539,
+      "step": 9506
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3604877768208319,
+      "learning_rate": 3.5092309599450596e-06,
+      "loss": 0.6265,
+      "step": 9507
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.166979716778442,
+      "learning_rate": 3.50824164647533e-06,
+      "loss": 0.5368,
+      "step": 9508
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5224247205974066,
+      "learning_rate": 3.507252397109777e-06,
+      "loss": 0.5412,
+      "step": 9509
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3494367370229554,
+      "learning_rate": 3.506263211890914e-06,
+      "loss": 0.6964,
+      "step": 9510
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1528162753784301,
+      "learning_rate": 3.505274090861249e-06,
+      "loss": 0.7396,
+      "step": 9511
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2561219807659023,
+      "learning_rate": 3.504285034063285e-06,
+      "loss": 0.6444,
+      "step": 9512
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2747463617275017,
+      "learning_rate": 3.503296041539522e-06,
+      "loss": 0.6425,
+      "step": 9513
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.164557867743521,
+      "learning_rate": 3.5023071133324627e-06,
+      "loss": 0.6886,
+      "step": 9514
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3581485675793175,
+      "learning_rate": 3.501318249484604e-06,
+      "loss": 0.5893,
+      "step": 9515
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3782422265065144,
+      "learning_rate": 3.5003294500384367e-06,
+      "loss": 0.6624,
+      "step": 9516
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3122272687593939,
+      "learning_rate": 3.4993407150364513e-06,
+      "loss": 0.6146,
+      "step": 9517
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2524293802961401,
+      "learning_rate": 3.4983520445211403e-06,
+      "loss": 0.614,
+      "step": 9518
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.202654247418089,
+      "learning_rate": 3.497363438534984e-06,
+      "loss": 0.6469,
+      "step": 9519
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3375076242949755,
+      "learning_rate": 3.496374897120467e-06,
+      "loss": 0.5946,
+      "step": 9520
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4339573936159478,
+      "learning_rate": 3.495386420320071e-06,
+      "loss": 0.646,
+      "step": 9521
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7359187148125461,
+      "learning_rate": 3.4943980081762717e-06,
+      "loss": 0.7905,
+      "step": 9522
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3011015515028364,
+      "learning_rate": 3.493409660731542e-06,
+      "loss": 0.6868,
+      "step": 9523
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.060146241544337,
+      "learning_rate": 3.4924213780283545e-06,
+      "loss": 0.6894,
+      "step": 9524
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7337716603628994,
+      "learning_rate": 3.4914331601091793e-06,
+      "loss": 0.6018,
+      "step": 9525
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5732901347363935,
+      "learning_rate": 3.4904450070164803e-06,
+      "loss": 0.7122,
+      "step": 9526
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.511222838432739,
+      "learning_rate": 3.4894569187927204e-06,
+      "loss": 0.6972,
+      "step": 9527
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2679699733455316,
+      "learning_rate": 3.48846889548036e-06,
+      "loss": 0.6025,
+      "step": 9528
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.385046318129598,
+      "learning_rate": 3.4874809371218608e-06,
+      "loss": 0.6076,
+      "step": 9529
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.438108414650423,
+      "learning_rate": 3.486493043759673e-06,
+      "loss": 0.5997,
+      "step": 9530
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2748782395986495,
+      "learning_rate": 3.4855052154362485e-06,
+      "loss": 0.6617,
+      "step": 9531
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.264663246661394,
+      "learning_rate": 3.4845174521940384e-06,
+      "loss": 0.6354,
+      "step": 9532
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0950765326404692,
+      "learning_rate": 3.4835297540754896e-06,
+      "loss": 0.6502,
+      "step": 9533
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4801998927365139,
+      "learning_rate": 3.4825421211230437e-06,
+      "loss": 0.6148,
+      "step": 9534
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2224577482997676,
+      "learning_rate": 3.481554553379143e-06,
+      "loss": 0.5968,
+      "step": 9535
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2646001382687118,
+      "learning_rate": 3.4805670508862255e-06,
+      "loss": 0.6132,
+      "step": 9536
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.457901628404729,
+      "learning_rate": 3.4795796136867237e-06,
+      "loss": 0.5417,
+      "step": 9537
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.155782492147106,
+      "learning_rate": 3.4785922418230722e-06,
+      "loss": 0.5118,
+      "step": 9538
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.372535292336444,
+      "learning_rate": 3.4776049353377016e-06,
+      "loss": 0.7046,
+      "step": 9539
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3832317082987826,
+      "learning_rate": 3.476617694273039e-06,
+      "loss": 0.5907,
+      "step": 9540
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3078129671316046,
+      "learning_rate": 3.4756305186715046e-06,
+      "loss": 0.6008,
+      "step": 9541
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4547802392237101,
+      "learning_rate": 3.474643408575521e-06,
+      "loss": 0.6635,
+      "step": 9542
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4869066837677232,
+      "learning_rate": 3.47365636402751e-06,
+      "loss": 0.6716,
+      "step": 9543
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2555534921307057,
+      "learning_rate": 3.4726693850698824e-06,
+      "loss": 0.6837,
+      "step": 9544
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2004480447795491,
+      "learning_rate": 3.4716824717450538e-06,
+      "loss": 0.6224,
+      "step": 9545
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3384193615050237,
+      "learning_rate": 3.4706956240954317e-06,
+      "loss": 0.6967,
+      "step": 9546
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3813793589288532,
+      "learning_rate": 3.4697088421634266e-06,
+      "loss": 0.6124,
+      "step": 9547
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.267661712011807,
+      "learning_rate": 3.4687221259914394e-06,
+      "loss": 0.5776,
+      "step": 9548
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4554965868842267,
+      "learning_rate": 3.467735475621873e-06,
+      "loss": 0.7471,
+      "step": 9549
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4061491721843424,
+      "learning_rate": 3.4667488910971252e-06,
+      "loss": 0.6501,
+      "step": 9550
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8460203717467616,
+      "learning_rate": 3.4657623724595944e-06,
+      "loss": 0.6233,
+      "step": 9551
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.422030775683194,
+      "learning_rate": 3.4647759197516687e-06,
+      "loss": 0.7185,
+      "step": 9552
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2953102439273096,
+      "learning_rate": 3.4637895330157423e-06,
+      "loss": 0.7372,
+      "step": 9553
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4932560701441773,
+      "learning_rate": 3.4628032122942024e-06,
+      "loss": 0.6705,
+      "step": 9554
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7577980669503397,
+      "learning_rate": 3.461816957629429e-06,
+      "loss": 0.6223,
+      "step": 9555
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1733225147002173,
+      "learning_rate": 3.4608307690638075e-06,
+      "loss": 0.6139,
+      "step": 9556
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.226142018146141,
+      "learning_rate": 3.459844646639717e-06,
+      "loss": 0.5478,
+      "step": 9557
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.270083163967127,
+      "learning_rate": 3.4588585903995314e-06,
+      "loss": 0.6235,
+      "step": 9558
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2274635506920843,
+      "learning_rate": 3.4578726003856245e-06,
+      "loss": 0.6151,
+      "step": 9559
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.225341251107578,
+      "learning_rate": 3.4568866766403665e-06,
+      "loss": 0.7073,
+      "step": 9560
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9974791483444387,
+      "learning_rate": 3.4559008192061254e-06,
+      "loss": 0.5436,
+      "step": 9561
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1892166268686275,
+      "learning_rate": 3.4549150281252635e-06,
+      "loss": 0.5771,
+      "step": 9562
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3377919847698923,
+      "learning_rate": 3.453929303440145e-06,
+      "loss": 0.6126,
+      "step": 9563
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3576851973259276,
+      "learning_rate": 3.4529436451931263e-06,
+      "loss": 0.6408,
+      "step": 9564
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1289029711055012,
+      "learning_rate": 3.4519580534265674e-06,
+      "loss": 0.6058,
+      "step": 9565
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.33997810602606,
+      "learning_rate": 3.4509725281828155e-06,
+      "loss": 0.6482,
+      "step": 9566
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5136830026987165,
+      "learning_rate": 3.449987069504226e-06,
+      "loss": 0.5878,
+      "step": 9567
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4657838842743778,
+      "learning_rate": 3.4490016774331426e-06,
+      "loss": 0.6546,
+      "step": 9568
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4339141745387756,
+      "learning_rate": 3.448016352011914e-06,
+      "loss": 0.6212,
+      "step": 9569
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.114836234159153,
+      "learning_rate": 3.4470310932828766e-06,
+      "loss": 0.5945,
+      "step": 9570
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3991737867765295,
+      "learning_rate": 3.4460459012883716e-06,
+      "loss": 0.5843,
+      "step": 9571
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6631674457816559,
+      "learning_rate": 3.4450607760707373e-06,
+      "loss": 0.684,
+      "step": 9572
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.538361013351043,
+      "learning_rate": 3.4440757176723037e-06,
+      "loss": 0.6176,
+      "step": 9573
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3654776461196216,
+      "learning_rate": 3.4430907261354e-06,
+      "loss": 0.6,
+      "step": 9574
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.34523645608255,
+      "learning_rate": 3.442105801502356e-06,
+      "loss": 0.6067,
+      "step": 9575
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4934113045694501,
+      "learning_rate": 3.441120943815497e-06,
+      "loss": 0.6541,
+      "step": 9576
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3989512610002834,
+      "learning_rate": 3.440136153117142e-06,
+      "loss": 0.6138,
+      "step": 9577
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.181388571506882,
+      "learning_rate": 3.4391514294496097e-06,
+      "loss": 0.6688,
+      "step": 9578
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0284286170265693,
+      "learning_rate": 3.438166772855218e-06,
+      "loss": 0.8187,
+      "step": 9579
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3998110954023018,
+      "learning_rate": 3.4371821833762776e-06,
+      "loss": 0.6896,
+      "step": 9580
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.571562756132462,
+      "learning_rate": 3.4361976610550996e-06,
+      "loss": 0.7777,
+      "step": 9581
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1012991902130134,
+      "learning_rate": 3.43521320593399e-06,
+      "loss": 0.7387,
+      "step": 9582
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1574789743914147,
+      "learning_rate": 3.4342288180552556e-06,
+      "loss": 0.5649,
+      "step": 9583
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2803292486697944,
+      "learning_rate": 3.4332444974611946e-06,
+      "loss": 0.6511,
+      "step": 9584
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2056134405627137,
+      "learning_rate": 3.4322602441941064e-06,
+      "loss": 0.6396,
+      "step": 9585
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4888159501554776,
+      "learning_rate": 3.4312760582962877e-06,
+      "loss": 0.8366,
+      "step": 9586
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2904458980981584,
+      "learning_rate": 3.4302919398100325e-06,
+      "loss": 0.6198,
+      "step": 9587
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3363441379242356,
+      "learning_rate": 3.4293078887776264e-06,
+      "loss": 0.7255,
+      "step": 9588
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.370341822654092,
+      "learning_rate": 3.428323905241358e-06,
+      "loss": 0.6044,
+      "step": 9589
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1161941351522071,
+      "learning_rate": 3.427339989243514e-06,
+      "loss": 0.6721,
+      "step": 9590
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.288653783359367,
+      "learning_rate": 3.426356140826372e-06,
+      "loss": 0.6405,
+      "step": 9591
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4536205738458632,
+      "learning_rate": 3.4253723600322105e-06,
+      "loss": 0.7007,
+      "step": 9592
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3750773801562215,
+      "learning_rate": 3.424388646903305e-06,
+      "loss": 0.7674,
+      "step": 9593
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.289992342553495,
+      "learning_rate": 3.4234050014819308e-06,
+      "loss": 0.7256,
+      "step": 9594
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3834425526829113,
+      "learning_rate": 3.422421423810354e-06,
+      "loss": 0.6743,
+      "step": 9595
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3437407138487851,
+      "learning_rate": 3.421437913930842e-06,
+      "loss": 0.6976,
+      "step": 9596
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5089964111439922,
+      "learning_rate": 3.420454471885659e-06,
+      "loss": 0.7344,
+      "step": 9597
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.426528565690189,
+      "learning_rate": 3.4194710977170635e-06,
+      "loss": 0.7645,
+      "step": 9598
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2837118024460634,
+      "learning_rate": 3.4184877914673155e-06,
+      "loss": 0.6429,
+      "step": 9599
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1502406578870883,
+      "learning_rate": 3.41750455317867e-06,
+      "loss": 0.7059,
+      "step": 9600
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3029729639935799,
+      "learning_rate": 3.4165213828933797e-06,
+      "loss": 0.6141,
+      "step": 9601
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3984953658349193,
+      "learning_rate": 3.4155382806536897e-06,
+      "loss": 0.6842,
+      "step": 9602
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3097833286019218,
+      "learning_rate": 3.4145552465018483e-06,
+      "loss": 0.7316,
+      "step": 9603
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.283476402167272,
+      "learning_rate": 3.4135722804801004e-06,
+      "loss": 0.6615,
+      "step": 9604
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.191198601999279,
+      "learning_rate": 3.4125893826306855e-06,
+      "loss": 0.732,
+      "step": 9605
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4129462059162852,
+      "learning_rate": 3.411606552995838e-06,
+      "loss": 0.6189,
+      "step": 9606
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0493066902722474,
+      "learning_rate": 3.4106237916177954e-06,
+      "loss": 0.7,
+      "step": 9607
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1521395584238936,
+      "learning_rate": 3.4096410985387898e-06,
+      "loss": 0.7928,
+      "step": 9608
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3503776655746773,
+      "learning_rate": 3.4086584738010455e-06,
+      "loss": 0.6583,
+      "step": 9609
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.346958045613319,
+      "learning_rate": 3.4076759174467923e-06,
+      "loss": 0.6705,
+      "step": 9610
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2978596113380358,
+      "learning_rate": 3.4066934295182496e-06,
+      "loss": 0.663,
+      "step": 9611
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1932515701870532,
+      "learning_rate": 3.4057110100576408e-06,
+      "loss": 0.61,
+      "step": 9612
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2362005192468093,
+      "learning_rate": 3.4047286591071794e-06,
+      "loss": 0.5923,
+      "step": 9613
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3902353027654786,
+      "learning_rate": 3.4037463767090807e-06,
+      "loss": 0.5985,
+      "step": 9614
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4430322490067,
+      "learning_rate": 3.402764162905554e-06,
+      "loss": 0.6055,
+      "step": 9615
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4234526281585564,
+      "learning_rate": 3.401782017738811e-06,
+      "loss": 0.6961,
+      "step": 9616
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4871234806849234,
+      "learning_rate": 3.400799941251052e-06,
+      "loss": 0.6367,
+      "step": 9617
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1654437078228468,
+      "learning_rate": 3.3998179334844823e-06,
+      "loss": 0.8814,
+      "step": 9618
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2237116798964722,
+      "learning_rate": 3.3988359944812997e-06,
+      "loss": 0.5769,
+      "step": 9619
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1449842836366648,
+      "learning_rate": 3.3978541242837003e-06,
+      "loss": 0.6117,
+      "step": 9620
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.246836537995142,
+      "learning_rate": 3.396872322933876e-06,
+      "loss": 0.6326,
+      "step": 9621
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9859696627004761,
+      "learning_rate": 3.395890590474019e-06,
+      "loss": 0.6292,
+      "step": 9622
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3299494238864282,
+      "learning_rate": 3.3949089269463178e-06,
+      "loss": 0.5591,
+      "step": 9623
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6017796991938158,
+      "learning_rate": 3.3939273323929533e-06,
+      "loss": 0.634,
+      "step": 9624
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2559898036985957,
+      "learning_rate": 3.3929458068561073e-06,
+      "loss": 0.6578,
+      "step": 9625
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4988018394298306,
+      "learning_rate": 3.391964350377961e-06,
+      "loss": 0.6354,
+      "step": 9626
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.521052150253596,
+      "learning_rate": 3.3909829630006865e-06,
+      "loss": 0.6624,
+      "step": 9627
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1791335468844923,
+      "learning_rate": 3.390001644766458e-06,
+      "loss": 0.5933,
+      "step": 9628
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6666421381685592,
+      "learning_rate": 3.3890203957174437e-06,
+      "loss": 0.5929,
+      "step": 9629
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5118399384075154,
+      "learning_rate": 3.388039215895813e-06,
+      "loss": 0.7105,
+      "step": 9630
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.377750051611533,
+      "learning_rate": 3.3870581053437244e-06,
+      "loss": 0.5515,
+      "step": 9631
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4218751690742535,
+      "learning_rate": 3.3860770641033417e-06,
+      "loss": 0.6377,
+      "step": 9632
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4402827625336394,
+      "learning_rate": 3.385096092216823e-06,
+      "loss": 0.5564,
+      "step": 9633
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3627098345352502,
+      "learning_rate": 3.3841151897263234e-06,
+      "loss": 0.7421,
+      "step": 9634
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.518683929654102,
+      "learning_rate": 3.3831343566739904e-06,
+      "loss": 0.6732,
+      "step": 9635
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6440390269147425,
+      "learning_rate": 3.3821535931019744e-06,
+      "loss": 0.6605,
+      "step": 9636
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2047470918876717,
+      "learning_rate": 3.3811728990524245e-06,
+      "loss": 0.4728,
+      "step": 9637
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4712410333437456,
+      "learning_rate": 3.380192274567479e-06,
+      "loss": 0.5581,
+      "step": 9638
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.321992594594157,
+      "learning_rate": 3.379211719689278e-06,
+      "loss": 0.6359,
+      "step": 9639
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.384547701032402,
+      "learning_rate": 3.3782312344599598e-06,
+      "loss": 0.6813,
+      "step": 9640
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3600274427131693,
+      "learning_rate": 3.3772508189216586e-06,
+      "loss": 0.668,
+      "step": 9641
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5075672897679102,
+      "learning_rate": 3.376270473116503e-06,
+      "loss": 0.6092,
+      "step": 9642
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3974166231146468,
+      "learning_rate": 3.37529019708662e-06,
+      "loss": 0.6788,
+      "step": 9643
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3695982990509679,
+      "learning_rate": 3.3743099908741385e-06,
+      "loss": 0.6224,
+      "step": 9644
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.28895799529403,
+      "learning_rate": 3.3733298545211748e-06,
+      "loss": 0.6237,
+      "step": 9645
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1558660590649878,
+      "learning_rate": 3.37234978806985e-06,
+      "loss": 0.5771,
+      "step": 9646
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.290197319197699,
+      "learning_rate": 3.3713697915622806e-06,
+      "loss": 0.5327,
+      "step": 9647
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5354566395109612,
+      "learning_rate": 3.3703898650405796e-06,
+      "loss": 0.7329,
+      "step": 9648
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4840798051348278,
+      "learning_rate": 3.3694100085468535e-06,
+      "loss": 0.7176,
+      "step": 9649
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3760027020615557,
+      "learning_rate": 3.3684302221232106e-06,
+      "loss": 0.6938,
+      "step": 9650
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6541041656069255,
+      "learning_rate": 3.3674505058117556e-06,
+      "loss": 0.663,
+      "step": 9651
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2160986346223934,
+      "learning_rate": 3.36647085965459e-06,
+      "loss": 0.638,
+      "step": 9652
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.350128998384387,
+      "learning_rate": 3.365491283693807e-06,
+      "loss": 0.5869,
+      "step": 9653
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4677891946652868,
+      "learning_rate": 3.364511777971504e-06,
+      "loss": 0.6522,
+      "step": 9654
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.240549538428606,
+      "learning_rate": 3.363532342529775e-06,
+      "loss": 0.6303,
+      "step": 9655
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3199564352893791,
+      "learning_rate": 3.3625529774107028e-06,
+      "loss": 0.6625,
+      "step": 9656
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4016535155359564,
+      "learning_rate": 3.361573682656377e-06,
+      "loss": 0.5986,
+      "step": 9657
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2011721627034628,
+      "learning_rate": 3.360594458308879e-06,
+      "loss": 0.5739,
+      "step": 9658
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.368609962907448,
+      "learning_rate": 3.3596153044102897e-06,
+      "loss": 0.7378,
+      "step": 9659
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.445133108109583,
+      "learning_rate": 3.358636221002682e-06,
+      "loss": 0.6592,
+      "step": 9660
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3166228643582123,
+      "learning_rate": 3.357657208128132e-06,
+      "loss": 0.5595,
+      "step": 9661
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.548078975327515,
+      "learning_rate": 3.356678265828711e-06,
+      "loss": 0.6005,
+      "step": 9662
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4237190164024474,
+      "learning_rate": 3.3556993941464834e-06,
+      "loss": 0.8029,
+      "step": 9663
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1834928406646252,
+      "learning_rate": 3.354720593123514e-06,
+      "loss": 0.6007,
+      "step": 9664
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0513652201854555,
+      "learning_rate": 3.353741862801866e-06,
+      "loss": 0.7039,
+      "step": 9665
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2497327255542334,
+      "learning_rate": 3.352763203223598e-06,
+      "loss": 0.7014,
+      "step": 9666
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3717259411696068,
+      "learning_rate": 3.351784614430761e-06,
+      "loss": 0.6555,
+      "step": 9667
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3546637881693544,
+      "learning_rate": 3.3508060964654106e-06,
+      "loss": 0.6192,
+      "step": 9668
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3708099716388062,
+      "learning_rate": 3.349827649369596e-06,
+      "loss": 0.6136,
+      "step": 9669
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.0254714435236636,
+      "learning_rate": 3.348849273185362e-06,
+      "loss": 0.6338,
+      "step": 9670
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6578125910047445,
+      "learning_rate": 3.3478709679547527e-06,
+      "loss": 0.6585,
+      "step": 9671
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.424283679638457,
+      "learning_rate": 3.346892733719806e-06,
+      "loss": 0.73,
+      "step": 9672
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5823901315023,
+      "learning_rate": 3.3459145705225627e-06,
+      "loss": 0.659,
+      "step": 9673
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2807047076697664,
+      "learning_rate": 3.3449364784050515e-06,
+      "loss": 0.5637,
+      "step": 9674
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4494534348729675,
+      "learning_rate": 3.343958457409307e-06,
+      "loss": 0.6866,
+      "step": 9675
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4182675801404447,
+      "learning_rate": 3.3429805075773547e-06,
+      "loss": 0.6605,
+      "step": 9676
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6169468284274304,
+      "learning_rate": 3.3420026289512232e-06,
+      "loss": 0.6314,
+      "step": 9677
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4749340344181432,
+      "learning_rate": 3.341024821572928e-06,
+      "loss": 0.6785,
+      "step": 9678
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2755642799959488,
+      "learning_rate": 3.3400470854844925e-06,
+      "loss": 0.5307,
+      "step": 9679
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4419864409347536,
+      "learning_rate": 3.3390694207279308e-06,
+      "loss": 0.5985,
+      "step": 9680
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3190017189324041,
+      "learning_rate": 3.3380918273452557e-06,
+      "loss": 0.6352,
+      "step": 9681
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2516916421144841,
+      "learning_rate": 3.3371143053784737e-06,
+      "loss": 0.6076,
+      "step": 9682
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6774766969403585,
+      "learning_rate": 3.336136854869595e-06,
+      "loss": 0.7323,
+      "step": 9683
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.01427376432299,
+      "learning_rate": 3.3351594758606222e-06,
+      "loss": 0.6359,
+      "step": 9684
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.482835307783579,
+      "learning_rate": 3.334182168393554e-06,
+      "loss": 0.6682,
+      "step": 9685
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2772337412927657,
+      "learning_rate": 3.3332049325103867e-06,
+      "loss": 0.5549,
+      "step": 9686
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4076826259158353,
+      "learning_rate": 3.332227768253116e-06,
+      "loss": 0.6933,
+      "step": 9687
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4159491900639491,
+      "learning_rate": 3.3312506756637343e-06,
+      "loss": 0.5827,
+      "step": 9688
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2436887243696202,
+      "learning_rate": 3.3302736547842263e-06,
+      "loss": 0.6156,
+      "step": 9689
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2958625132794783,
+      "learning_rate": 3.3292967056565766e-06,
+      "loss": 0.6149,
+      "step": 9690
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2963746908842586,
+      "learning_rate": 3.3283198283227715e-06,
+      "loss": 0.5814,
+      "step": 9691
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5569192589109924,
+      "learning_rate": 3.3273430228247834e-06,
+      "loss": 0.6928,
+      "step": 9692
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2461317166785024,
+      "learning_rate": 3.3263662892045924e-06,
+      "loss": 0.6866,
+      "step": 9693
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.306375681490193,
+      "learning_rate": 3.3253896275041677e-06,
+      "loss": 0.5615,
+      "step": 9694
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3339548678321984,
+      "learning_rate": 3.324413037765483e-06,
+      "loss": 0.6367,
+      "step": 9695
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4205949849131154,
+      "learning_rate": 3.3234365200304997e-06,
+      "loss": 0.6641,
+      "step": 9696
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.154723157842722,
+      "learning_rate": 3.322460074341183e-06,
+      "loss": 0.5699,
+      "step": 9697
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4328054198021425,
+      "learning_rate": 3.321483700739494e-06,
+      "loss": 0.6152,
+      "step": 9698
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3171533521776995,
+      "learning_rate": 3.3205073992673885e-06,
+      "loss": 0.6546,
+      "step": 9699
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2702149873805286,
+      "learning_rate": 3.31953116996682e-06,
+      "loss": 0.7176,
+      "step": 9700
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3872460462665799,
+      "learning_rate": 3.3185550128797387e-06,
+      "loss": 0.6293,
+      "step": 9701
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2861291926938814,
+      "learning_rate": 3.317578928048096e-06,
+      "loss": 0.6029,
+      "step": 9702
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2364127338987168,
+      "learning_rate": 3.3166029155138324e-06,
+      "loss": 0.6178,
+      "step": 9703
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2810476539861106,
+      "learning_rate": 3.3156269753188895e-06,
+      "loss": 0.7141,
+      "step": 9704
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7790974026904123,
+      "learning_rate": 3.314651107505207e-06,
+      "loss": 0.609,
+      "step": 9705
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.520361606776559,
+      "learning_rate": 3.313675312114722e-06,
+      "loss": 0.7689,
+      "step": 9706
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5438341309701147,
+      "learning_rate": 3.312699589189362e-06,
+      "loss": 0.6684,
+      "step": 9707
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3056989217564805,
+      "learning_rate": 3.31172393877106e-06,
+      "loss": 0.5293,
+      "step": 9708
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2824680896302925,
+      "learning_rate": 3.310748360901741e-06,
+      "loss": 0.5724,
+      "step": 9709
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.380685060991457,
+      "learning_rate": 3.3097728556233256e-06,
+      "loss": 0.716,
+      "step": 9710
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5117340944394686,
+      "learning_rate": 3.308797422977734e-06,
+      "loss": 0.6751,
+      "step": 9711
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.462014003763691,
+      "learning_rate": 3.307822063006886e-06,
+      "loss": 0.6668,
+      "step": 9712
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1764735993635564,
+      "learning_rate": 3.3068467757526927e-06,
+      "loss": 0.4426,
+      "step": 9713
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4950889823208502,
+      "learning_rate": 3.3058715612570623e-06,
+      "loss": 0.671,
+      "step": 9714
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3566838853442265,
+      "learning_rate": 3.3048964195619044e-06,
+      "loss": 0.8107,
+      "step": 9715
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6578940077395083,
+      "learning_rate": 3.303921350709124e-06,
+      "loss": 0.6486,
+      "step": 9716
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.318558708273808,
+      "learning_rate": 3.3029463547406198e-06,
+      "loss": 0.6816,
+      "step": 9717
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5741648305372926,
+      "learning_rate": 3.30197143169829e-06,
+      "loss": 0.6634,
+      "step": 9718
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3303772706553039,
+      "learning_rate": 3.300996581624028e-06,
+      "loss": 0.6054,
+      "step": 9719
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5435977715734304,
+      "learning_rate": 3.30002180455973e-06,
+      "loss": 0.627,
+      "step": 9720
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4308124801367814,
+      "learning_rate": 3.299047100547278e-06,
+      "loss": 0.7084,
+      "step": 9721
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1230671426183936,
+      "learning_rate": 3.2980724696285626e-06,
+      "loss": 0.6696,
+      "step": 9722
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.397594567987838,
+      "learning_rate": 3.2970979118454616e-06,
+      "loss": 0.6458,
+      "step": 9723
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.488434711833693,
+      "learning_rate": 3.2961234272398578e-06,
+      "loss": 0.5598,
+      "step": 9724
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4143800748137876,
+      "learning_rate": 3.2951490158536238e-06,
+      "loss": 0.5991,
+      "step": 9725
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3442827348964208,
+      "learning_rate": 3.2941746777286344e-06,
+      "loss": 0.5356,
+      "step": 9726
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3539973974524309,
+      "learning_rate": 3.2932004129067592e-06,
+      "loss": 0.6384,
+      "step": 9727
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.226198406725117,
+      "learning_rate": 3.2922262214298613e-06,
+      "loss": 0.6719,
+      "step": 9728
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.379868245306721,
+      "learning_rate": 3.291252103339806e-06,
+      "loss": 0.6352,
+      "step": 9729
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.615298199412592,
+      "learning_rate": 3.2902780586784542e-06,
+      "loss": 0.6625,
+      "step": 9730
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3543994289728158,
+      "learning_rate": 3.2893040874876624e-06,
+      "loss": 0.6143,
+      "step": 9731
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.50137201994436,
+      "learning_rate": 3.2883301898092833e-06,
+      "loss": 0.6546,
+      "step": 9732
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5111556593853481,
+      "learning_rate": 3.287356365685168e-06,
+      "loss": 0.6591,
+      "step": 9733
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5061388489995808,
+      "learning_rate": 3.2863826151571654e-06,
+      "loss": 0.7499,
+      "step": 9734
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5154734767608276,
+      "learning_rate": 3.285408938267116e-06,
+      "loss": 0.6072,
+      "step": 9735
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2719099531870306,
+      "learning_rate": 3.284435335056865e-06,
+      "loss": 0.6318,
+      "step": 9736
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3110553679227526,
+      "learning_rate": 3.283461805568246e-06,
+      "loss": 0.6587,
+      "step": 9737
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6459306818309047,
+      "learning_rate": 3.2824883498430992e-06,
+      "loss": 0.5887,
+      "step": 9738
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.396858503653752,
+      "learning_rate": 3.2815149679232507e-06,
+      "loss": 0.6171,
+      "step": 9739
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3245377205078497,
+      "learning_rate": 3.2805416598505323e-06,
+      "loss": 0.5761,
+      "step": 9740
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5478170705441927,
+      "learning_rate": 3.2795684256667674e-06,
+      "loss": 0.7384,
+      "step": 9741
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2654855181417162,
+      "learning_rate": 3.278595265413781e-06,
+      "loss": 0.6189,
+      "step": 9742
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4219258814962723,
+      "learning_rate": 3.277622179133387e-06,
+      "loss": 0.5515,
+      "step": 9743
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.189997486882871,
+      "learning_rate": 3.276649166867406e-06,
+      "loss": 0.6176,
+      "step": 9744
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3904332099901298,
+      "learning_rate": 3.2756762286576482e-06,
+      "loss": 0.6693,
+      "step": 9745
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.383420996935016,
+      "learning_rate": 3.2747033645459226e-06,
+      "loss": 0.6065,
+      "step": 9746
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3365990189255215,
+      "learning_rate": 3.2737305745740353e-06,
+      "loss": 0.584,
+      "step": 9747
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4064294924412528,
+      "learning_rate": 3.2727578587837893e-06,
+      "loss": 0.6603,
+      "step": 9748
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4932878399263507,
+      "learning_rate": 3.271785217216987e-06,
+      "loss": 0.7842,
+      "step": 9749
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.47177070835505,
+      "learning_rate": 3.2708126499154225e-06,
+      "loss": 0.6874,
+      "step": 9750
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4976133859221432,
+      "learning_rate": 3.2698401569208883e-06,
+      "loss": 0.5932,
+      "step": 9751
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1033352636267146,
+      "learning_rate": 3.2688677382751754e-06,
+      "loss": 0.4843,
+      "step": 9752
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.344342844903466,
+      "learning_rate": 3.2678953940200744e-06,
+      "loss": 0.6075,
+      "step": 9753
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5733123182783768,
+      "learning_rate": 3.266923124197363e-06,
+      "loss": 0.6517,
+      "step": 9754
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.719727037414836,
+      "learning_rate": 3.265950928848825e-06,
+      "loss": 0.6354,
+      "step": 9755
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3797566383918212,
+      "learning_rate": 3.264978808016239e-06,
+      "loss": 0.6264,
+      "step": 9756
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3203052181052939,
+      "learning_rate": 3.264006761741376e-06,
+      "loss": 0.7226,
+      "step": 9757
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.505184022294216,
+      "learning_rate": 3.2630347900660094e-06,
+      "loss": 0.611,
+      "step": 9758
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.398604758100057,
+      "learning_rate": 3.2620628930319065e-06,
+      "loss": 0.635,
+      "step": 9759
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1468070429297756,
+      "learning_rate": 3.261091070680833e-06,
+      "loss": 0.5982,
+      "step": 9760
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2322960243171521,
+      "learning_rate": 3.260119323054546e-06,
+      "loss": 0.6362,
+      "step": 9761
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4092231085389293,
+      "learning_rate": 3.2591476501948075e-06,
+      "loss": 0.6337,
+      "step": 9762
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3468542080421848,
+      "learning_rate": 3.258176052143374e-06,
+      "loss": 0.6976,
+      "step": 9763
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4103772347099914,
+      "learning_rate": 3.257204528941993e-06,
+      "loss": 0.6754,
+      "step": 9764
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2893334016445914,
+      "learning_rate": 3.256233080632414e-06,
+      "loss": 0.7291,
+      "step": 9765
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0856275269282514,
+      "learning_rate": 3.255261707256383e-06,
+      "loss": 0.6434,
+      "step": 9766
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1849558887110054,
+      "learning_rate": 3.2542904088556437e-06,
+      "loss": 0.5821,
+      "step": 9767
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1657861285639255,
+      "learning_rate": 3.2533191854719336e-06,
+      "loss": 0.5764,
+      "step": 9768
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3742127198339655,
+      "learning_rate": 3.2523480371469863e-06,
+      "loss": 0.6906,
+      "step": 9769
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0553960332949108,
+      "learning_rate": 3.251376963922537e-06,
+      "loss": 0.7222,
+      "step": 9770
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6109043266486125,
+      "learning_rate": 3.250405965840315e-06,
+      "loss": 0.6201,
+      "step": 9771
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2324178134958934,
+      "learning_rate": 3.249435042942043e-06,
+      "loss": 0.7053,
+      "step": 9772
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3057703251329584,
+      "learning_rate": 3.2484641952694473e-06,
+      "loss": 0.7008,
+      "step": 9773
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1802454260400492,
+      "learning_rate": 3.2474934228642475e-06,
+      "loss": 0.5497,
+      "step": 9774
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.198867872792256,
+      "learning_rate": 3.246522725768156e-06,
+      "loss": 0.612,
+      "step": 9775
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4981880179925393,
+      "learning_rate": 3.2455521040228883e-06,
+      "loss": 0.6406,
+      "step": 9776
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4565260103780864,
+      "learning_rate": 3.2445815576701544e-06,
+      "loss": 0.6448,
+      "step": 9777
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1576521545233431,
+      "learning_rate": 3.2436110867516625e-06,
+      "loss": 0.7058,
+      "step": 9778
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0382087889783036,
+      "learning_rate": 3.242640691309111e-06,
+      "loss": 0.7663,
+      "step": 9779
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.433550891219091,
+      "learning_rate": 3.2416703713842036e-06,
+      "loss": 0.6309,
+      "step": 9780
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3980546252754429,
+      "learning_rate": 3.240700127018638e-06,
+      "loss": 0.5943,
+      "step": 9781
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4038161331203567,
+      "learning_rate": 3.239729958254104e-06,
+      "loss": 0.6241,
+      "step": 9782
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1450775807543387,
+      "learning_rate": 3.238759865132295e-06,
+      "loss": 0.5563,
+      "step": 9783
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4190511596731554,
+      "learning_rate": 3.2377898476948964e-06,
+      "loss": 0.6733,
+      "step": 9784
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.287944911456874,
+      "learning_rate": 3.2368199059835947e-06,
+      "loss": 0.5738,
+      "step": 9785
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7747502658584011,
+      "learning_rate": 3.235850040040066e-06,
+      "loss": 0.7127,
+      "step": 9786
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.428028056188389,
+      "learning_rate": 3.2348802499059915e-06,
+      "loss": 0.701,
+      "step": 9787
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5465317038707833,
+      "learning_rate": 3.2339105356230426e-06,
+      "loss": 0.7327,
+      "step": 9788
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.255027249490706,
+      "learning_rate": 3.2329408972328934e-06,
+      "loss": 0.5353,
+      "step": 9789
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8701219581958803,
+      "learning_rate": 3.2319713347772075e-06,
+      "loss": 0.7783,
+      "step": 9790
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3308856552346877,
+      "learning_rate": 3.231001848297651e-06,
+      "loss": 0.6403,
+      "step": 9791
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3452365548195444,
+      "learning_rate": 3.230032437835887e-06,
+      "loss": 0.6033,
+      "step": 9792
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3820138143747593,
+      "learning_rate": 3.2290631034335684e-06,
+      "loss": 0.6198,
+      "step": 9793
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.621419340391697,
+      "learning_rate": 3.2280938451323524e-06,
+      "loss": 0.7189,
+      "step": 9794
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.448700202826855,
+      "learning_rate": 3.2271246629738917e-06,
+      "loss": 0.6189,
+      "step": 9795
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5439485908698143,
+      "learning_rate": 3.2261555569998316e-06,
+      "loss": 0.7389,
+      "step": 9796
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.541942866356869,
+      "learning_rate": 3.2251865272518187e-06,
+      "loss": 0.7261,
+      "step": 9797
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.319812514820858,
+      "learning_rate": 3.224217573771492e-06,
+      "loss": 0.5934,
+      "step": 9798
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6115297373140551,
+      "learning_rate": 3.223248696600493e-06,
+      "loss": 0.6038,
+      "step": 9799
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4204151863723082,
+      "learning_rate": 3.2222798957804524e-06,
+      "loss": 0.6885,
+      "step": 9800
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4323943137412907,
+      "learning_rate": 3.2213111713530056e-06,
+      "loss": 0.5636,
+      "step": 9801
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3691155433200146,
+      "learning_rate": 3.2203425233597773e-06,
+      "loss": 0.6361,
+      "step": 9802
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4740047523439705,
+      "learning_rate": 3.2193739518423972e-06,
+      "loss": 0.6208,
+      "step": 9803
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.359516205766793,
+      "learning_rate": 3.2184054568424817e-06,
+      "loss": 0.6796,
+      "step": 9804
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2256950055544884,
+      "learning_rate": 3.217437038401652e-06,
+      "loss": 0.5908,
+      "step": 9805
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3310189677557378,
+      "learning_rate": 3.2164686965615223e-06,
+      "loss": 0.5878,
+      "step": 9806
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3374017203021766,
+      "learning_rate": 3.215500431363706e-06,
+      "loss": 0.5889,
+      "step": 9807
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3981256386537178,
+      "learning_rate": 3.2145322428498093e-06,
+      "loss": 0.6352,
+      "step": 9808
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4822248635071473,
+      "learning_rate": 3.2135641310614383e-06,
+      "loss": 0.7156,
+      "step": 9809
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2618242523303411,
+      "learning_rate": 3.2125960960401974e-06,
+      "loss": 0.6675,
+      "step": 9810
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3584271850824268,
+      "learning_rate": 3.211628137827682e-06,
+      "loss": 0.5827,
+      "step": 9811
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1529412014914417,
+      "learning_rate": 3.2106602564654877e-06,
+      "loss": 0.5814,
+      "step": 9812
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.31914898060717,
+      "learning_rate": 3.2096924519952067e-06,
+      "loss": 0.6264,
+      "step": 9813
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3242419424690282,
+      "learning_rate": 3.20872472445843e-06,
+      "loss": 0.6551,
+      "step": 9814
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1756183997665512,
+      "learning_rate": 3.207757073896741e-06,
+      "loss": 0.6616,
+      "step": 9815
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3715248517076992,
+      "learning_rate": 3.2067895003517207e-06,
+      "loss": 0.7091,
+      "step": 9816
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6009914729718715,
+      "learning_rate": 3.2058220038649513e-06,
+      "loss": 0.7327,
+      "step": 9817
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3321085941379815,
+      "learning_rate": 3.2048545844780044e-06,
+      "loss": 0.5618,
+      "step": 9818
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.200330033444637,
+      "learning_rate": 3.203887242232455e-06,
+      "loss": 0.4832,
+      "step": 9819
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.580439586556014,
+      "learning_rate": 3.2029199771698693e-06,
+      "loss": 0.7286,
+      "step": 9820
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9560128647291033,
+      "learning_rate": 3.2019527893318177e-06,
+      "loss": 0.6661,
+      "step": 9821
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3783438439227265,
+      "learning_rate": 3.2009856787598558e-06,
+      "loss": 0.6618,
+      "step": 9822
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3430947398605817,
+      "learning_rate": 3.2000186454955474e-06,
+      "loss": 0.7047,
+      "step": 9823
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3168655100004405,
+      "learning_rate": 3.1990516895804467e-06,
+      "loss": 0.6395,
+      "step": 9824
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4115821465063176,
+      "learning_rate": 3.198084811056107e-06,
+      "loss": 0.7048,
+      "step": 9825
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3916912525155904,
+      "learning_rate": 3.197118009964074e-06,
+      "loss": 0.6062,
+      "step": 9826
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3826932213992689,
+      "learning_rate": 3.1961512863458953e-06,
+      "loss": 0.6202,
+      "step": 9827
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7767495381685372,
+      "learning_rate": 3.195184640243115e-06,
+      "loss": 0.6364,
+      "step": 9828
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.6378134743111725,
+      "learning_rate": 3.1942180716972698e-06,
+      "loss": 0.7808,
+      "step": 9829
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4386102253940398,
+      "learning_rate": 3.1932515807498945e-06,
+      "loss": 0.6209,
+      "step": 9830
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5232446810567035,
+      "learning_rate": 3.1922851674425226e-06,
+      "loss": 0.7021,
+      "step": 9831
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2456651848101328,
+      "learning_rate": 3.191318831816685e-06,
+      "loss": 0.6304,
+      "step": 9832
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.361368535574474,
+      "learning_rate": 3.190352573913903e-06,
+      "loss": 0.6575,
+      "step": 9833
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3686889119815044,
+      "learning_rate": 3.189386393775703e-06,
+      "loss": 0.6324,
+      "step": 9834
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3713243810146627,
+      "learning_rate": 3.1884202914436024e-06,
+      "loss": 0.6996,
+      "step": 9835
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.258529094912021,
+      "learning_rate": 3.1874542669591148e-06,
+      "loss": 0.6102,
+      "step": 9836
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2207560597663536,
+      "learning_rate": 3.1864883203637544e-06,
+      "loss": 0.5593,
+      "step": 9837
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.27303060286429,
+      "learning_rate": 3.18552245169903e-06,
+      "loss": 0.5917,
+      "step": 9838
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1605369315852674,
+      "learning_rate": 3.1845566610064487e-06,
+      "loss": 0.5866,
+      "step": 9839
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6166105172126524,
+      "learning_rate": 3.1835909483275083e-06,
+      "loss": 0.5612,
+      "step": 9840
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3279841775087946,
+      "learning_rate": 3.1826253137037112e-06,
+      "loss": 0.6282,
+      "step": 9841
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3643203967935338,
+      "learning_rate": 3.1816597571765517e-06,
+      "loss": 0.6457,
+      "step": 9842
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4020703000435475,
+      "learning_rate": 3.180694278787524e-06,
+      "loss": 0.6874,
+      "step": 9843
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.309948305759791,
+      "learning_rate": 3.179728878578112e-06,
+      "loss": 0.6722,
+      "step": 9844
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2742696560716,
+      "learning_rate": 3.1787635565898046e-06,
+      "loss": 0.5499,
+      "step": 9845
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3676836169163051,
+      "learning_rate": 3.1777983128640855e-06,
+      "loss": 0.6781,
+      "step": 9846
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2363680653542222,
+      "learning_rate": 3.1768331474424285e-06,
+      "loss": 0.5643,
+      "step": 9847
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.230932727256702,
+      "learning_rate": 3.1758680603663127e-06,
+      "loss": 0.4931,
+      "step": 9848
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3122377328561707,
+      "learning_rate": 3.1749030516772084e-06,
+      "loss": 0.5702,
+      "step": 9849
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.143919930513717,
+      "learning_rate": 3.1739381214165864e-06,
+      "loss": 0.5124,
+      "step": 9850
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5523807910345955,
+      "learning_rate": 3.1729732696259085e-06,
+      "loss": 0.5837,
+      "step": 9851
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6168293101872127,
+      "learning_rate": 3.172008496346639e-06,
+      "loss": 0.624,
+      "step": 9852
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2089208458447385,
+      "learning_rate": 3.1710438016202362e-06,
+      "loss": 0.599,
+      "step": 9853
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6747529829162866,
+      "learning_rate": 3.170079185488153e-06,
+      "loss": 0.6752,
+      "step": 9854
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3527224314965178,
+      "learning_rate": 3.1691146479918423e-06,
+      "loss": 0.6505,
+      "step": 9855
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2752923375454952,
+      "learning_rate": 3.168150189172754e-06,
+      "loss": 0.5499,
+      "step": 9856
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4468237168944065,
+      "learning_rate": 3.1671858090723315e-06,
+      "loss": 0.7431,
+      "step": 9857
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3909057763139407,
+      "learning_rate": 3.1662215077320174e-06,
+      "loss": 0.5875,
+      "step": 9858
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4386452624472135,
+      "learning_rate": 3.165257285193248e-06,
+      "loss": 0.5823,
+      "step": 9859
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.280939224123671,
+      "learning_rate": 3.1642931414974586e-06,
+      "loss": 0.6091,
+      "step": 9860
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3317309476516201,
+      "learning_rate": 3.1633290766860834e-06,
+      "loss": 0.6469,
+      "step": 9861
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6095486216026091,
+      "learning_rate": 3.1623650908005476e-06,
+      "loss": 0.5966,
+      "step": 9862
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.280665052337773,
+      "learning_rate": 3.1614011838822755e-06,
+      "loss": 0.6323,
+      "step": 9863
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1427405627209148,
+      "learning_rate": 3.1604373559726915e-06,
+      "loss": 0.6743,
+      "step": 9864
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3717653192099957,
+      "learning_rate": 3.1594736071132086e-06,
+      "loss": 0.6395,
+      "step": 9865
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3970521516571677,
+      "learning_rate": 3.1585099373452454e-06,
+      "loss": 0.6585,
+      "step": 9866
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3325615202930439,
+      "learning_rate": 3.15754634671021e-06,
+      "loss": 0.6452,
+      "step": 9867
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2323145260133432,
+      "learning_rate": 3.1565828352495144e-06,
+      "loss": 0.5544,
+      "step": 9868
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4678027107815144,
+      "learning_rate": 3.1556194030045563e-06,
+      "loss": 0.6385,
+      "step": 9869
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.497739153477262,
+      "learning_rate": 3.154656050016742e-06,
+      "loss": 0.6053,
+      "step": 9870
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.412212440846435,
+      "learning_rate": 3.153692776327466e-06,
+      "loss": 0.5492,
+      "step": 9871
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5634442332470182,
+      "learning_rate": 3.1527295819781256e-06,
+      "loss": 0.7517,
+      "step": 9872
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.102158819353865,
+      "learning_rate": 3.151766467010107e-06,
+      "loss": 0.6258,
+      "step": 9873
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3841188630963044,
+      "learning_rate": 3.1508034314647994e-06,
+      "loss": 0.6294,
+      "step": 9874
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.394429048724333,
+      "learning_rate": 3.149840475383589e-06,
+      "loss": 0.7432,
+      "step": 9875
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1089917335268578,
+      "learning_rate": 3.1488775988078525e-06,
+      "loss": 0.5655,
+      "step": 9876
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2982395259421275,
+      "learning_rate": 3.1479148017789673e-06,
+      "loss": 0.7452,
+      "step": 9877
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1030037087872036,
+      "learning_rate": 3.146952084338308e-06,
+      "loss": 0.5173,
+      "step": 9878
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3119184660174454,
+      "learning_rate": 3.1459894465272467e-06,
+      "loss": 0.5424,
+      "step": 9879
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2073723891260977,
+      "learning_rate": 3.1450268883871467e-06,
+      "loss": 0.7162,
+      "step": 9880
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2354564061744502,
+      "learning_rate": 3.1440644099593716e-06,
+      "loss": 0.5351,
+      "step": 9881
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5280944101703235,
+      "learning_rate": 3.143102011285284e-06,
+      "loss": 0.6985,
+      "step": 9882
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8216969268008012,
+      "learning_rate": 3.142139692406237e-06,
+      "loss": 0.631,
+      "step": 9883
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3065462392307157,
+      "learning_rate": 3.1411774533635854e-06,
+      "loss": 0.6619,
+      "step": 9884
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5388356315450304,
+      "learning_rate": 3.1402152941986786e-06,
+      "loss": 0.7007,
+      "step": 9885
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6144592972572904,
+      "learning_rate": 3.1392532149528643e-06,
+      "loss": 0.7134,
+      "step": 9886
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.335095956447197,
+      "learning_rate": 3.1382912156674807e-06,
+      "loss": 0.599,
+      "step": 9887
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4482574433911004,
+      "learning_rate": 3.13732929638387e-06,
+      "loss": 0.7925,
+      "step": 9888
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6212423755225602,
+      "learning_rate": 3.136367457143369e-06,
+      "loss": 0.6144,
+      "step": 9889
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9865186154918261,
+      "learning_rate": 3.1354056979873104e-06,
+      "loss": 0.5659,
+      "step": 9890
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5271366231073393,
+      "learning_rate": 3.134444018957019e-06,
+      "loss": 0.6979,
+      "step": 9891
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9791265070248244,
+      "learning_rate": 3.1334824200938228e-06,
+      "loss": 0.6223,
+      "step": 9892
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7477921506771603,
+      "learning_rate": 3.132520901439047e-06,
+      "loss": 0.6696,
+      "step": 9893
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.228080497363743,
+      "learning_rate": 3.1315594630340052e-06,
+      "loss": 0.6655,
+      "step": 9894
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5957010020998865,
+      "learning_rate": 3.1305981049200134e-06,
+      "loss": 0.8624,
+      "step": 9895
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6167794686166468,
+      "learning_rate": 3.129636827138385e-06,
+      "loss": 0.7415,
+      "step": 9896
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.126803963206699,
+      "learning_rate": 3.128675629730429e-06,
+      "loss": 0.5765,
+      "step": 9897
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9933474523947506,
+      "learning_rate": 3.1277145127374475e-06,
+      "loss": 0.583,
+      "step": 9898
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.399085779963844,
+      "learning_rate": 3.1267534762007435e-06,
+      "loss": 0.7437,
+      "step": 9899
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6316551915726643,
+      "learning_rate": 3.125792520161616e-06,
+      "loss": 0.7036,
+      "step": 9900
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.169192161814114,
+      "learning_rate": 3.1248316446613563e-06,
+      "loss": 0.6367,
+      "step": 9901
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2646640964021076,
+      "learning_rate": 3.1238708497412566e-06,
+      "loss": 0.5181,
+      "step": 9902
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4200891983162545,
+      "learning_rate": 3.1229101354426063e-06,
+      "loss": 0.6015,
+      "step": 9903
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1611162530937016,
+      "learning_rate": 3.1219495018066888e-06,
+      "loss": 0.7302,
+      "step": 9904
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.50710666107025,
+      "learning_rate": 3.1209889488747813e-06,
+      "loss": 0.7497,
+      "step": 9905
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4612004269065175,
+      "learning_rate": 3.120028476688164e-06,
+      "loss": 0.6363,
+      "step": 9906
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3087378990329397,
+      "learning_rate": 3.119068085288111e-06,
+      "loss": 0.6463,
+      "step": 9907
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.376490395598786,
+      "learning_rate": 3.118107774715892e-06,
+      "loss": 0.7056,
+      "step": 9908
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5591507336360897,
+      "learning_rate": 3.1171475450127717e-06,
+      "loss": 0.5208,
+      "step": 9909
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3464891815597853,
+      "learning_rate": 3.1161873962200146e-06,
+      "loss": 0.6548,
+      "step": 9910
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1216079209917023,
+      "learning_rate": 3.1152273283788824e-06,
+      "loss": 0.5571,
+      "step": 9911
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3890553651609552,
+      "learning_rate": 3.114267341530627e-06,
+      "loss": 0.6223,
+      "step": 9912
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5200309150457896,
+      "learning_rate": 3.1133074357165055e-06,
+      "loss": 0.6791,
+      "step": 9913
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2668617323930689,
+      "learning_rate": 3.112347610977764e-06,
+      "loss": 0.6616,
+      "step": 9914
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4699725625417714,
+      "learning_rate": 3.111387867355652e-06,
+      "loss": 0.7423,
+      "step": 9915
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.643842987108769,
+      "learning_rate": 3.110428204891407e-06,
+      "loss": 0.5926,
+      "step": 9916
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2620829490754863,
+      "learning_rate": 3.109468623626272e-06,
+      "loss": 0.6368,
+      "step": 9917
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2685606193142105,
+      "learning_rate": 3.1085091236014808e-06,
+      "loss": 0.5673,
+      "step": 9918
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2224029531506027,
+      "learning_rate": 3.1075497048582635e-06,
+      "loss": 0.6014,
+      "step": 9919
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4582492749150453,
+      "learning_rate": 3.1065903674378507e-06,
+      "loss": 0.6784,
+      "step": 9920
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3757792590098343,
+      "learning_rate": 3.1056311113814675e-06,
+      "loss": 0.6569,
+      "step": 9921
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6522649826013087,
+      "learning_rate": 3.1046719367303334e-06,
+      "loss": 0.753,
+      "step": 9922
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3944241720294344,
+      "learning_rate": 3.1037128435256688e-06,
+      "loss": 0.6062,
+      "step": 9923
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2642631427168967,
+      "learning_rate": 3.102753831808685e-06,
+      "loss": 0.6985,
+      "step": 9924
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.265422309940266,
+      "learning_rate": 3.101794901620594e-06,
+      "loss": 0.5917,
+      "step": 9925
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4477075719588135,
+      "learning_rate": 3.1008360530026053e-06,
+      "loss": 0.6603,
+      "step": 9926
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3040314103938928,
+      "learning_rate": 3.099877285995921e-06,
+      "loss": 0.704,
+      "step": 9927
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.282663441030181,
+      "learning_rate": 3.09891860064174e-06,
+      "loss": 0.5521,
+      "step": 9928
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.321469661625787,
+      "learning_rate": 3.097959996981263e-06,
+      "loss": 0.7284,
+      "step": 9929
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4364482002664998,
+      "learning_rate": 3.0970014750556786e-06,
+      "loss": 0.7386,
+      "step": 9930
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.401913859854057,
+      "learning_rate": 3.0960430349061797e-06,
+      "loss": 0.604,
+      "step": 9931
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4172515696341912,
+      "learning_rate": 3.0950846765739518e-06,
+      "loss": 0.5863,
+      "step": 9932
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3094216940946946,
+      "learning_rate": 3.0941264001001796e-06,
+      "loss": 0.5938,
+      "step": 9933
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3190344218886116,
+      "learning_rate": 3.093168205526038e-06,
+      "loss": 0.619,
+      "step": 9934
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2259704647249685,
+      "learning_rate": 3.0922100928927062e-06,
+      "loss": 0.5367,
+      "step": 9935
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2545614758280603,
+      "learning_rate": 3.0912520622413573e-06,
+      "loss": 0.6404,
+      "step": 9936
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3708896925645122,
+      "learning_rate": 3.090294113613157e-06,
+      "loss": 0.7314,
+      "step": 9937
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.503976808448571,
+      "learning_rate": 3.0893362470492716e-06,
+      "loss": 0.6567,
+      "step": 9938
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6561507588078959,
+      "learning_rate": 3.0883784625908618e-06,
+      "loss": 0.7133,
+      "step": 9939
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3941076366437077,
+      "learning_rate": 3.0874207602790895e-06,
+      "loss": 0.6161,
+      "step": 9940
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.7389491279284783,
+      "learning_rate": 3.086463140155106e-06,
+      "loss": 0.7151,
+      "step": 9941
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4377428234825609,
+      "learning_rate": 3.085505602260061e-06,
+      "loss": 0.6797,
+      "step": 9942
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4904155750717998,
+      "learning_rate": 3.0845481466351047e-06,
+      "loss": 0.6121,
+      "step": 9943
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2256867333148664,
+      "learning_rate": 3.083590773321383e-06,
+      "loss": 0.6802,
+      "step": 9944
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3674927936564094,
+      "learning_rate": 3.082633482360032e-06,
+      "loss": 0.6439,
+      "step": 9945
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5864484883960945,
+      "learning_rate": 3.0816762737921896e-06,
+      "loss": 0.7764,
+      "step": 9946
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.350027316313769,
+      "learning_rate": 3.0807191476589926e-06,
+      "loss": 0.681,
+      "step": 9947
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4358983759066841,
+      "learning_rate": 3.0797621040015656e-06,
+      "loss": 0.6707,
+      "step": 9948
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.581581852569877,
+      "learning_rate": 3.0788051428610377e-06,
+      "loss": 0.7066,
+      "step": 9949
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4092140403192,
+      "learning_rate": 3.077848264278533e-06,
+      "loss": 0.7619,
+      "step": 9950
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2868428465512045,
+      "learning_rate": 3.07689146829517e-06,
+      "loss": 0.6732,
+      "step": 9951
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5440019583325288,
+      "learning_rate": 3.0759347549520614e-06,
+      "loss": 0.736,
+      "step": 9952
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3449352805835142,
+      "learning_rate": 3.074978124290322e-06,
+      "loss": 0.6189,
+      "step": 9953
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1578276039333995,
+      "learning_rate": 3.0740215763510617e-06,
+      "loss": 0.5916,
+      "step": 9954
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4514261825421018,
+      "learning_rate": 3.0730651111753824e-06,
+      "loss": 0.6053,
+      "step": 9955
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2882094047731694,
+      "learning_rate": 3.072108728804386e-06,
+      "loss": 0.5734,
+      "step": 9956
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0448253817887736,
+      "learning_rate": 3.0711524292791707e-06,
+      "loss": 0.7367,
+      "step": 9957
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.02602594715051,
+      "learning_rate": 3.0701962126408334e-06,
+      "loss": 0.5812,
+      "step": 9958
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3188313966388543,
+      "learning_rate": 3.069240078930461e-06,
+      "loss": 0.5695,
+      "step": 9959
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5135610785585731,
+      "learning_rate": 3.0682840281891433e-06,
+      "loss": 0.6711,
+      "step": 9960
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.285802425372793,
+      "learning_rate": 3.0673280604579623e-06,
+      "loss": 0.5858,
+      "step": 9961
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1492164160486134,
+      "learning_rate": 3.0663721757780006e-06,
+      "loss": 0.5463,
+      "step": 9962
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4171494951376262,
+      "learning_rate": 3.0654163741903315e-06,
+      "loss": 0.6439,
+      "step": 9963
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4402077282976846,
+      "learning_rate": 3.0644606557360303e-06,
+      "loss": 0.7203,
+      "step": 9964
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.303225836988199,
+      "learning_rate": 3.0635050204561658e-06,
+      "loss": 0.6497,
+      "step": 9965
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4473473785147384,
+      "learning_rate": 3.0625494683918012e-06,
+      "loss": 0.6016,
+      "step": 9966
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0017335763302633,
+      "learning_rate": 3.0615939995840017e-06,
+      "loss": 0.6865,
+      "step": 9967
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1484552403241688,
+      "learning_rate": 3.0606386140738253e-06,
+      "loss": 0.6166,
+      "step": 9968
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8045698025908599,
+      "learning_rate": 3.0596833119023283e-06,
+      "loss": 0.695,
+      "step": 9969
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7556013515445703,
+      "learning_rate": 3.058728093110559e-06,
+      "loss": 0.7692,
+      "step": 9970
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2095740633670495,
+      "learning_rate": 3.0577729577395663e-06,
+      "loss": 0.6311,
+      "step": 9971
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.182682450796763,
+      "learning_rate": 3.056817905830397e-06,
+      "loss": 0.7076,
+      "step": 9972
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.403101219874816,
+      "learning_rate": 3.055862937424089e-06,
+      "loss": 0.5989,
+      "step": 9973
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3313184032739251,
+      "learning_rate": 3.054908052561681e-06,
+      "loss": 0.6467,
+      "step": 9974
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.537105971160288,
+      "learning_rate": 3.053953251284205e-06,
+      "loss": 0.7429,
+      "step": 9975
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3623606563856194,
+      "learning_rate": 3.0529985336326933e-06,
+      "loss": 0.623,
+      "step": 9976
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2474494088919243,
+      "learning_rate": 3.0520438996481694e-06,
+      "loss": 0.5787,
+      "step": 9977
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0697114221940132,
+      "learning_rate": 3.0510893493716574e-06,
+      "loss": 0.6339,
+      "step": 9978
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.274834215655324,
+      "learning_rate": 3.0501348828441767e-06,
+      "loss": 0.6452,
+      "step": 9979
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.81858845473369,
+      "learning_rate": 3.0491805001067443e-06,
+      "loss": 0.8429,
+      "step": 9980
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3730497225994125,
+      "learning_rate": 3.048226201200369e-06,
+      "loss": 0.6295,
+      "step": 9981
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6438965619973587,
+      "learning_rate": 3.047271986166061e-06,
+      "loss": 0.7201,
+      "step": 9982
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2036003795343995,
+      "learning_rate": 3.0463178550448246e-06,
+      "loss": 0.6062,
+      "step": 9983
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3750745952290113,
+      "learning_rate": 3.0453638078776614e-06,
+      "loss": 0.6876,
+      "step": 9984
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2191463971022494,
+      "learning_rate": 3.044409844705568e-06,
+      "loss": 0.5665,
+      "step": 9985
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.239457437567063,
+      "learning_rate": 3.043455965569538e-06,
+      "loss": 0.5574,
+      "step": 9986
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3879209906030874,
+      "learning_rate": 3.042502170510565e-06,
+      "loss": 0.6382,
+      "step": 9987
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3867726820394914,
+      "learning_rate": 3.041548459569633e-06,
+      "loss": 0.6356,
+      "step": 9988
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5765002757348585,
+      "learning_rate": 3.0405948327877233e-06,
+      "loss": 0.6996,
+      "step": 9989
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.509383038046926,
+      "learning_rate": 3.039641290205818e-06,
+      "loss": 0.691,
+      "step": 9990
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4206009757439286,
+      "learning_rate": 3.038687831864895e-06,
+      "loss": 0.6249,
+      "step": 9991
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0868823970266268,
+      "learning_rate": 3.037734457805922e-06,
+      "loss": 0.5329,
+      "step": 9992
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.329350479226106,
+      "learning_rate": 3.036781168069869e-06,
+      "loss": 0.5979,
+      "step": 9993
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4461220661738046,
+      "learning_rate": 3.0358279626977034e-06,
+      "loss": 0.6705,
+      "step": 9994
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.289986004051368,
+      "learning_rate": 3.0348748417303826e-06,
+      "loss": 0.6244,
+      "step": 9995
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.251047794571385,
+      "learning_rate": 3.033921805208867e-06,
+      "loss": 0.693,
+      "step": 9996
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3422628602104818,
+      "learning_rate": 3.0329688531741097e-06,
+      "loss": 0.5659,
+      "step": 9997
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2215219987699601,
+      "learning_rate": 3.032015985667064e-06,
+      "loss": 0.5655,
+      "step": 9998
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0481707447862525,
+      "learning_rate": 3.0310632027286717e-06,
+      "loss": 0.63,
+      "step": 9999
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.265973159059951,
+      "learning_rate": 3.030110504399879e-06,
+      "loss": 0.6146,
+      "step": 10000
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4234849261060833,
+      "learning_rate": 3.029157890721627e-06,
+      "loss": 0.884,
+      "step": 10001
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2788981391845875,
+      "learning_rate": 3.0282053617348485e-06,
+      "loss": 0.6253,
+      "step": 10002
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.452906791430369,
+      "learning_rate": 3.027252917480476e-06,
+      "loss": 0.5881,
+      "step": 10003
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4112720043209583,
+      "learning_rate": 3.026300557999439e-06,
+      "loss": 0.7085,
+      "step": 10004
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0692181462324588,
+      "learning_rate": 3.0253482833326654e-06,
+      "loss": 0.5422,
+      "step": 10005
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4707952551520829,
+      "learning_rate": 3.0243960935210726e-06,
+      "loss": 0.6311,
+      "step": 10006
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.469596338261718,
+      "learning_rate": 3.023443988605578e-06,
+      "loss": 0.5863,
+      "step": 10007
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2174960357561606,
+      "learning_rate": 3.0224919686270982e-06,
+      "loss": 0.6157,
+      "step": 10008
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4299975004558716,
+      "learning_rate": 3.021540033626544e-06,
+      "loss": 0.6836,
+      "step": 10009
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1212296627136613,
+      "learning_rate": 3.0205881836448186e-06,
+      "loss": 0.6644,
+      "step": 10010
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3985318664809145,
+      "learning_rate": 3.0196364187228287e-06,
+      "loss": 0.9506,
+      "step": 10011
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.411876641675788,
+      "learning_rate": 3.0186847389014738e-06,
+      "loss": 0.6566,
+      "step": 10012
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4331062024106327,
+      "learning_rate": 3.0177331442216455e-06,
+      "loss": 0.6261,
+      "step": 10013
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7895323340904525,
+      "learning_rate": 3.0167816347242396e-06,
+      "loss": 0.7543,
+      "step": 10014
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2185860795687193,
+      "learning_rate": 3.0158302104501435e-06,
+      "loss": 0.6785,
+      "step": 10015
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4249111395757357,
+      "learning_rate": 3.0148788714402444e-06,
+      "loss": 0.7339,
+      "step": 10016
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3907762056136719,
+      "learning_rate": 3.0139276177354188e-06,
+      "loss": 0.6197,
+      "step": 10017
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3702957455497784,
+      "learning_rate": 3.0129764493765466e-06,
+      "loss": 0.7077,
+      "step": 10018
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3200982268079786,
+      "learning_rate": 3.012025366404504e-06,
+      "loss": 0.6388,
+      "step": 10019
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.028817261509046,
+      "learning_rate": 3.0110743688601578e-06,
+      "loss": 0.5883,
+      "step": 10020
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4228542230577983,
+      "learning_rate": 3.0101234567843747e-06,
+      "loss": 0.6682,
+      "step": 10021
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2341275461064545,
+      "learning_rate": 3.009172630218018e-06,
+      "loss": 0.6266,
+      "step": 10022
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1286156243271495,
+      "learning_rate": 3.0082218892019487e-06,
+      "loss": 0.551,
+      "step": 10023
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3507547218664502,
+      "learning_rate": 3.00727123377702e-06,
+      "loss": 0.7129,
+      "step": 10024
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3779810226168554,
+      "learning_rate": 3.0063206639840846e-06,
+      "loss": 0.7043,
+      "step": 10025
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.360081301468589,
+      "learning_rate": 3.00537017986399e-06,
+      "loss": 0.6337,
+      "step": 10026
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5376780643867916,
+      "learning_rate": 3.0044197814575825e-06,
+      "loss": 0.6495,
+      "step": 10027
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.074767533655776,
+      "learning_rate": 3.0034694688056998e-06,
+      "loss": 0.6383,
+      "step": 10028
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2836227586101487,
+      "learning_rate": 3.002519241949181e-06,
+      "loss": 0.576,
+      "step": 10029
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8100449564244276,
+      "learning_rate": 3.0015691009288605e-06,
+      "loss": 0.5733,
+      "step": 10030
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3230102168862812,
+      "learning_rate": 3.0006190457855643e-06,
+      "loss": 0.5293,
+      "step": 10031
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0483353794477415,
+      "learning_rate": 2.99966907656012e-06,
+      "loss": 0.6552,
+      "step": 10032
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.473180234046499,
+      "learning_rate": 2.9987191932933523e-06,
+      "loss": 0.6508,
+      "step": 10033
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3215746996983215,
+      "learning_rate": 2.997769396026078e-06,
+      "loss": 0.6429,
+      "step": 10034
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2630650093825286,
+      "learning_rate": 2.9968196847991117e-06,
+      "loss": 0.7247,
+      "step": 10035
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4400706280658562,
+      "learning_rate": 2.995870059653263e-06,
+      "loss": 0.6989,
+      "step": 10036
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4480117276755764,
+      "learning_rate": 2.994920520629344e-06,
+      "loss": 0.787,
+      "step": 10037
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.081778860743733,
+      "learning_rate": 2.9939710677681545e-06,
+      "loss": 0.5735,
+      "step": 10038
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.42305184609668,
+      "learning_rate": 2.9930217011104957e-06,
+      "loss": 0.6703,
+      "step": 10039
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.126328373553431,
+      "learning_rate": 2.992072420697164e-06,
+      "loss": 0.6729,
+      "step": 10040
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4994510417457896,
+      "learning_rate": 2.991123226568954e-06,
+      "loss": 0.7186,
+      "step": 10041
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2366630324521346,
+      "learning_rate": 2.990174118766651e-06,
+      "loss": 0.664,
+      "step": 10042
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4576644046704288,
+      "learning_rate": 2.9892250973310433e-06,
+      "loss": 0.7718,
+      "step": 10043
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3556073904671462,
+      "learning_rate": 2.98827616230291e-06,
+      "loss": 0.674,
+      "step": 10044
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.372964797733191,
+      "learning_rate": 2.987327313723033e-06,
+      "loss": 0.6357,
+      "step": 10045
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3453614913948482,
+      "learning_rate": 2.9863785516321806e-06,
+      "loss": 0.7604,
+      "step": 10046
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3333926326010124,
+      "learning_rate": 2.9854298760711285e-06,
+      "loss": 0.6861,
+      "step": 10047
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3420165761291072,
+      "learning_rate": 2.9844812870806417e-06,
+      "loss": 0.6454,
+      "step": 10048
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4276232918790521,
+      "learning_rate": 2.9835327847014816e-06,
+      "loss": 0.7446,
+      "step": 10049
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.212731813234314,
+      "learning_rate": 2.982584368974408e-06,
+      "loss": 0.502,
+      "step": 10050
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4409295599764622,
+      "learning_rate": 2.9816360399401777e-06,
+      "loss": 0.7997,
+      "step": 10051
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4118041791964504,
+      "learning_rate": 2.980687797639543e-06,
+      "loss": 0.5864,
+      "step": 10052
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2318336356072581,
+      "learning_rate": 2.979739642113251e-06,
+      "loss": 0.5574,
+      "step": 10053
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4398380398960962,
+      "learning_rate": 2.9787915734020446e-06,
+      "loss": 0.5937,
+      "step": 10054
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2110723636691547,
+      "learning_rate": 2.9778435915466674e-06,
+      "loss": 0.6597,
+      "step": 10055
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4773169018958436,
+      "learning_rate": 2.9768956965878526e-06,
+      "loss": 0.6743,
+      "step": 10056
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4583309210369861,
+      "learning_rate": 2.9759478885663362e-06,
+      "loss": 0.6482,
+      "step": 10057
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2300213688010084,
+      "learning_rate": 2.975000167522846e-06,
+      "loss": 0.5813,
+      "step": 10058
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3951471713581098,
+      "learning_rate": 2.9740525334981105e-06,
+      "loss": 0.6985,
+      "step": 10059
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6988827923086485,
+      "learning_rate": 2.9731049865328486e-06,
+      "loss": 0.8184,
+      "step": 10060
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3227072707809095,
+      "learning_rate": 2.9721575266677783e-06,
+      "loss": 0.6182,
+      "step": 10061
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1202225397279304,
+      "learning_rate": 2.971210153943617e-06,
+      "loss": 0.6203,
+      "step": 10062
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.302823513522959,
+      "learning_rate": 2.970262868401075e-06,
+      "loss": 0.6296,
+      "step": 10063
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2437606298008526,
+      "learning_rate": 2.9693156700808556e-06,
+      "loss": 0.6069,
+      "step": 10064
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4706641584767415,
+      "learning_rate": 2.968368559023664e-06,
+      "loss": 0.7081,
+      "step": 10065
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7023679568303849,
+      "learning_rate": 2.967421535270203e-06,
+      "loss": 0.7207,
+      "step": 10066
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.083354610955705,
+      "learning_rate": 2.966474598861164e-06,
+      "loss": 0.5672,
+      "step": 10067
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.516604161103265,
+      "learning_rate": 2.965527749837239e-06,
+      "loss": 0.7111,
+      "step": 10068
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5120940007063388,
+      "learning_rate": 2.9645809882391187e-06,
+      "loss": 0.7028,
+      "step": 10069
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.561310556603642,
+      "learning_rate": 2.9636343141074873e-06,
+      "loss": 0.61,
+      "step": 10070
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2014152710701997,
+      "learning_rate": 2.962687727483024e-06,
+      "loss": 0.8611,
+      "step": 10071
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1319219099897424,
+      "learning_rate": 2.961741228406405e-06,
+      "loss": 0.6054,
+      "step": 10072
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.181010878697717,
+      "learning_rate": 2.9607948169183077e-06,
+      "loss": 0.6295,
+      "step": 10073
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4163913654406555,
+      "learning_rate": 2.959848493059396e-06,
+      "loss": 0.6532,
+      "step": 10074
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.370208722548275,
+      "learning_rate": 2.958902256870339e-06,
+      "loss": 0.6294,
+      "step": 10075
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5918176786165106,
+      "learning_rate": 2.9579561083917984e-06,
+      "loss": 0.6708,
+      "step": 10076
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4336751253344542,
+      "learning_rate": 2.9570100476644325e-06,
+      "loss": 0.6711,
+      "step": 10077
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5303547863839158,
+      "learning_rate": 2.9560640747288925e-06,
+      "loss": 0.8027,
+      "step": 10078
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2537146887449273,
+      "learning_rate": 2.9551181896258317e-06,
+      "loss": 0.6195,
+      "step": 10079
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4192117043209727,
+      "learning_rate": 2.9541723923958975e-06,
+      "loss": 0.6778,
+      "step": 10080
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5707393715448446,
+      "learning_rate": 2.9532266830797336e-06,
+      "loss": 0.7558,
+      "step": 10081
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5511325703118533,
+      "learning_rate": 2.9522810617179742e-06,
+      "loss": 0.9781,
+      "step": 10082
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9727276542450325,
+      "learning_rate": 2.951335528351259e-06,
+      "loss": 0.7863,
+      "step": 10083
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2366688084943043,
+      "learning_rate": 2.9503900830202202e-06,
+      "loss": 0.6961,
+      "step": 10084
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2638579894637412,
+      "learning_rate": 2.9494447257654836e-06,
+      "loss": 0.5458,
+      "step": 10085
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7959001948647868,
+      "learning_rate": 2.948499456627674e-06,
+      "loss": 0.622,
+      "step": 10086
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1484955495840516,
+      "learning_rate": 2.94755427564741e-06,
+      "loss": 0.6786,
+      "step": 10087
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9789737832141407,
+      "learning_rate": 2.9466091828653127e-06,
+      "loss": 0.5731,
+      "step": 10088
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.159709861290248,
+      "learning_rate": 2.9456641783219897e-06,
+      "loss": 0.6961,
+      "step": 10089
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.362112132515206,
+      "learning_rate": 2.9447192620580532e-06,
+      "loss": 0.5884,
+      "step": 10090
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3065622201958955,
+      "learning_rate": 2.9437744341141084e-06,
+      "loss": 0.5797,
+      "step": 10091
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0129399449442087,
+      "learning_rate": 2.9428296945307533e-06,
+      "loss": 0.5202,
+      "step": 10092
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2352840072522904,
+      "learning_rate": 2.9418850433485877e-06,
+      "loss": 0.51,
+      "step": 10093
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5701901886558238,
+      "learning_rate": 2.9409404806082077e-06,
+      "loss": 0.6875,
+      "step": 10094
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3242893727728298,
+      "learning_rate": 2.939996006350201e-06,
+      "loss": 0.6362,
+      "step": 10095
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.647281376803744,
+      "learning_rate": 2.939051620615152e-06,
+      "loss": 0.5266,
+      "step": 10096
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.283219377240908,
+      "learning_rate": 2.938107323443645e-06,
+      "loss": 0.6361,
+      "step": 10097
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4987657169452107,
+      "learning_rate": 2.9371631148762592e-06,
+      "loss": 0.672,
+      "step": 10098
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2719750336208844,
+      "learning_rate": 2.936218994953568e-06,
+      "loss": 0.6369,
+      "step": 10099
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4361479070488652,
+      "learning_rate": 2.935274963716144e-06,
+      "loss": 0.4957,
+      "step": 10100
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.395573369099966,
+      "learning_rate": 2.934331021204551e-06,
+      "loss": 0.6183,
+      "step": 10101
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6476635642305824,
+      "learning_rate": 2.9333871674593573e-06,
+      "loss": 0.6098,
+      "step": 10102
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.438167752504769,
+      "learning_rate": 2.9324434025211173e-06,
+      "loss": 0.6449,
+      "step": 10103
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1988979700408784,
+      "learning_rate": 2.93149972643039e-06,
+      "loss": 0.6191,
+      "step": 10104
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4189161914245516,
+      "learning_rate": 2.930556139227726e-06,
+      "loss": 0.6861,
+      "step": 10105
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3067787165547828,
+      "learning_rate": 2.929612640953675e-06,
+      "loss": 0.6349,
+      "step": 10106
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2895211032311051,
+      "learning_rate": 2.9286692316487776e-06,
+      "loss": 0.678,
+      "step": 10107
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.112530710393557,
+      "learning_rate": 2.9277259113535774e-06,
+      "loss": 0.5978,
+      "step": 10108
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5351512903481894,
+      "learning_rate": 2.9267826801086103e-06,
+      "loss": 0.6928,
+      "step": 10109
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2071825260062035,
+      "learning_rate": 2.9258395379544085e-06,
+      "loss": 0.603,
+      "step": 10110
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3692105712076725,
+      "learning_rate": 2.9248964849315004e-06,
+      "loss": 0.6648,
+      "step": 10111
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4814137606177398,
+      "learning_rate": 2.9239535210804115e-06,
+      "loss": 0.6086,
+      "step": 10112
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3744504556832602,
+      "learning_rate": 2.923010646441666e-06,
+      "loss": 0.7315,
+      "step": 10113
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.353069624734594,
+      "learning_rate": 2.9220678610557773e-06,
+      "loss": 0.6345,
+      "step": 10114
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2154648589813846,
+      "learning_rate": 2.9211251649632587e-06,
+      "loss": 0.5785,
+      "step": 10115
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6363461862776396,
+      "learning_rate": 2.920182558204623e-06,
+      "loss": 0.6961,
+      "step": 10116
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.151224887489365,
+      "learning_rate": 2.919240040820376e-06,
+      "loss": 0.7196,
+      "step": 10117
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4604402976944506,
+      "learning_rate": 2.9182976128510177e-06,
+      "loss": 0.7477,
+      "step": 10118
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0451735392651067,
+      "learning_rate": 2.9173552743370454e-06,
+      "loss": 0.6006,
+      "step": 10119
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.569768750897323,
+      "learning_rate": 2.9164130253189594e-06,
+      "loss": 0.6999,
+      "step": 10120
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6786259434633861,
+      "learning_rate": 2.9154708658372433e-06,
+      "loss": 0.6928,
+      "step": 10121
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.682106626641055,
+      "learning_rate": 2.9145287959323852e-06,
+      "loss": 0.7269,
+      "step": 10122
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3338497825684092,
+      "learning_rate": 2.913586815644871e-06,
+      "loss": 0.6757,
+      "step": 10123
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4623285314977432,
+      "learning_rate": 2.912644925015179e-06,
+      "loss": 0.6945,
+      "step": 10124
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2002444197301811,
+      "learning_rate": 2.9117031240837825e-06,
+      "loss": 0.5538,
+      "step": 10125
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3631333694967802,
+      "learning_rate": 2.910761412891153e-06,
+      "loss": 0.7207,
+      "step": 10126
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4258208111509814,
+      "learning_rate": 2.9098197914777575e-06,
+      "loss": 0.5704,
+      "step": 10127
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0639634412905667,
+      "learning_rate": 2.9088782598840646e-06,
+      "loss": 0.4686,
+      "step": 10128
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.221254526694405,
+      "learning_rate": 2.9079368181505263e-06,
+      "loss": 0.5845,
+      "step": 10129
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4615602311897633,
+      "learning_rate": 2.9069954663176036e-06,
+      "loss": 0.6696,
+      "step": 10130
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.761574963483635,
+      "learning_rate": 2.9060542044257487e-06,
+      "loss": 0.6929,
+      "step": 10131
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8647488748820589,
+      "learning_rate": 2.9051130325154053e-06,
+      "loss": 0.5991,
+      "step": 10132
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1946223863234389,
+      "learning_rate": 2.904171950627022e-06,
+      "loss": 0.52,
+      "step": 10133
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4830931306060287,
+      "learning_rate": 2.9032309588010372e-06,
+      "loss": 0.8024,
+      "step": 10134
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1626633023898045,
+      "learning_rate": 2.9022900570778876e-06,
+      "loss": 0.5423,
+      "step": 10135
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.356860189597135,
+      "learning_rate": 2.9013492454980074e-06,
+      "loss": 0.5599,
+      "step": 10136
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2455330193962337,
+      "learning_rate": 2.9004085241018208e-06,
+      "loss": 0.6559,
+      "step": 10137
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3115298786489695,
+      "learning_rate": 2.899467892929761e-06,
+      "loss": 0.5841,
+      "step": 10138
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4943892225564295,
+      "learning_rate": 2.8985273520222414e-06,
+      "loss": 0.6775,
+      "step": 10139
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2758192209601396,
+      "learning_rate": 2.8975869014196797e-06,
+      "loss": 0.6757,
+      "step": 10140
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1210596042263146,
+      "learning_rate": 2.896646541162493e-06,
+      "loss": 0.7599,
+      "step": 10141
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.277051945388501,
+      "learning_rate": 2.895706271291089e-06,
+      "loss": 0.6561,
+      "step": 10142
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5379161970797122,
+      "learning_rate": 2.894766091845873e-06,
+      "loss": 0.6123,
+      "step": 10143
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.134048015620747,
+      "learning_rate": 2.893826002867247e-06,
+      "loss": 0.6793,
+      "step": 10144
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3672304206837442,
+      "learning_rate": 2.8928860043956062e-06,
+      "loss": 0.5975,
+      "step": 10145
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4336871307495167,
+      "learning_rate": 2.8919460964713507e-06,
+      "loss": 0.6372,
+      "step": 10146
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.514976158438903,
+      "learning_rate": 2.891006279134862e-06,
+      "loss": 0.6331,
+      "step": 10147
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.628943002964793,
+      "learning_rate": 2.8900665524265326e-06,
+      "loss": 0.5925,
+      "step": 10148
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3761101223381231,
+      "learning_rate": 2.889126916386744e-06,
+      "loss": 0.6686,
+      "step": 10149
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.198941954160972,
+      "learning_rate": 2.88818737105587e-06,
+      "loss": 0.6532,
+      "step": 10150
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2097311178180639,
+      "learning_rate": 2.8872479164742896e-06,
+      "loss": 0.6176,
+      "step": 10151
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9627934130217459,
+      "learning_rate": 2.886308552682371e-06,
+      "loss": 0.6441,
+      "step": 10152
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3946884269708448,
+      "learning_rate": 2.885369279720482e-06,
+      "loss": 0.6576,
+      "step": 10153
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3043911348310195,
+      "learning_rate": 2.884430097628984e-06,
+      "loss": 0.5656,
+      "step": 10154
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3892538385550979,
+      "learning_rate": 2.883491006448235e-06,
+      "loss": 0.6538,
+      "step": 10155
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5136887810665633,
+      "learning_rate": 2.8825520062185945e-06,
+      "loss": 0.6528,
+      "step": 10156
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.397463839375441,
+      "learning_rate": 2.881613096980407e-06,
+      "loss": 0.6539,
+      "step": 10157
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.445154182426802,
+      "learning_rate": 2.8806742787740237e-06,
+      "loss": 0.7613,
+      "step": 10158
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5944333326688827,
+      "learning_rate": 2.879735551639787e-06,
+      "loss": 0.6343,
+      "step": 10159
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4396160793316852,
+      "learning_rate": 2.8787969156180353e-06,
+      "loss": 0.7251,
+      "step": 10160
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.794972884539253,
+      "learning_rate": 2.877858370749105e-06,
+      "loss": 0.5705,
+      "step": 10161
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3844537891586621,
+      "learning_rate": 2.8769199170733265e-06,
+      "loss": 0.6488,
+      "step": 10162
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.302411749293348,
+      "learning_rate": 2.8759815546310254e-06,
+      "loss": 0.7163,
+      "step": 10163
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1556947073220856,
+      "learning_rate": 2.8750432834625312e-06,
+      "loss": 0.581,
+      "step": 10164
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6645107764270164,
+      "learning_rate": 2.874105103608156e-06,
+      "loss": 0.5686,
+      "step": 10165
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5917699558616543,
+      "learning_rate": 2.8731670151082213e-06,
+      "loss": 0.7201,
+      "step": 10166
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4693299712046077,
+      "learning_rate": 2.872229018003036e-06,
+      "loss": 0.6295,
+      "step": 10167
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1006227162022784,
+      "learning_rate": 2.871291112332909e-06,
+      "loss": 0.6491,
+      "step": 10168
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2429323568119333,
+      "learning_rate": 2.8703532981381437e-06,
+      "loss": 0.5387,
+      "step": 10169
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4277409590943102,
+      "learning_rate": 2.869415575459038e-06,
+      "loss": 0.6749,
+      "step": 10170
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.649936339460412,
+      "learning_rate": 2.8684779443358945e-06,
+      "loss": 0.6881,
+      "step": 10171
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4740817019897914,
+      "learning_rate": 2.8675404048089987e-06,
+      "loss": 0.7052,
+      "step": 10172
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.406620395394068,
+      "learning_rate": 2.8666029569186393e-06,
+      "loss": 0.651,
+      "step": 10173
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3560796782341749,
+      "learning_rate": 2.8656656007051055e-06,
+      "loss": 0.5942,
+      "step": 10174
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5611503919494671,
+      "learning_rate": 2.864728336208671e-06,
+      "loss": 0.6005,
+      "step": 10175
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6013098267008197,
+      "learning_rate": 2.8637911634696165e-06,
+      "loss": 0.6628,
+      "step": 10176
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3007735063455084,
+      "learning_rate": 2.862854082528214e-06,
+      "loss": 0.6034,
+      "step": 10177
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3893393085379016,
+      "learning_rate": 2.861917093424731e-06,
+      "loss": 0.62,
+      "step": 10178
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4409980449651525,
+      "learning_rate": 2.860980196199432e-06,
+      "loss": 0.6696,
+      "step": 10179
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5082138581398423,
+      "learning_rate": 2.8600433908925764e-06,
+      "loss": 0.7026,
+      "step": 10180
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4163535087819383,
+      "learning_rate": 2.859106677544424e-06,
+      "loss": 0.7332,
+      "step": 10181
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2542618050791339,
+      "learning_rate": 2.8581700561952274e-06,
+      "loss": 0.5985,
+      "step": 10182
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2868747882438378,
+      "learning_rate": 2.8572335268852297e-06,
+      "loss": 0.6519,
+      "step": 10183
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1778999748460008,
+      "learning_rate": 2.8562970896546815e-06,
+      "loss": 0.6142,
+      "step": 10184
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.108919842993341,
+      "learning_rate": 2.855360744543822e-06,
+      "loss": 0.5822,
+      "step": 10185
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.260822772017922,
+      "learning_rate": 2.8544244915928876e-06,
+      "loss": 0.5518,
+      "step": 10186
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5757962139541353,
+      "learning_rate": 2.8534883308421114e-06,
+      "loss": 0.6576,
+      "step": 10187
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2073758466264632,
+      "learning_rate": 2.852552262331719e-06,
+      "loss": 0.7112,
+      "step": 10188
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2862939813048608,
+      "learning_rate": 2.8516162861019437e-06,
+      "loss": 0.6634,
+      "step": 10189
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3510305628030865,
+      "learning_rate": 2.850680402192999e-06,
+      "loss": 0.6227,
+      "step": 10190
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4664299866663923,
+      "learning_rate": 2.8497446106451022e-06,
+      "loss": 0.553,
+      "step": 10191
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2267548186296162,
+      "learning_rate": 2.8488089114984725e-06,
+      "loss": 0.5828,
+      "step": 10192
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5192572623348293,
+      "learning_rate": 2.8478733047933103e-06,
+      "loss": 0.8272,
+      "step": 10193
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.252052230360452,
+      "learning_rate": 2.846937790569828e-06,
+      "loss": 0.5874,
+      "step": 10194
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5551273648509474,
+      "learning_rate": 2.8460023688682237e-06,
+      "loss": 0.6254,
+      "step": 10195
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.505314930513273,
+      "learning_rate": 2.8450670397286945e-06,
+      "loss": 0.811,
+      "step": 10196
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5020757490325192,
+      "learning_rate": 2.8441318031914344e-06,
+      "loss": 0.7353,
+      "step": 10197
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.264882635364458,
+      "learning_rate": 2.8431966592966297e-06,
+      "loss": 0.6165,
+      "step": 10198
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1507971632018608,
+      "learning_rate": 2.84226160808447e-06,
+      "loss": 0.5883,
+      "step": 10199
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1383305281412301,
+      "learning_rate": 2.8413266495951366e-06,
+      "loss": 0.6629,
+      "step": 10200
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3911972559620205,
+      "learning_rate": 2.8403917838688007e-06,
+      "loss": 0.5905,
+      "step": 10201
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.329722638523531,
+      "learning_rate": 2.8394570109456413e-06,
+      "loss": 0.6431,
+      "step": 10202
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3244080975409713,
+      "learning_rate": 2.838522330865826e-06,
+      "loss": 0.5586,
+      "step": 10203
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0376704521383984,
+      "learning_rate": 2.837587743669521e-06,
+      "loss": 0.5663,
+      "step": 10204
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4148083665666542,
+      "learning_rate": 2.8366532493968858e-06,
+      "loss": 0.6619,
+      "step": 10205
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4755709042056422,
+      "learning_rate": 2.835718848088076e-06,
+      "loss": 0.6976,
+      "step": 10206
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2607500507671694,
+      "learning_rate": 2.8347845397832517e-06,
+      "loss": 0.5856,
+      "step": 10207
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3360645949086118,
+      "learning_rate": 2.8338503245225544e-06,
+      "loss": 0.5905,
+      "step": 10208
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.85013597089306,
+      "learning_rate": 2.8329162023461355e-06,
+      "loss": 0.7016,
+      "step": 10209
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5215326906682403,
+      "learning_rate": 2.8319821732941345e-06,
+      "loss": 0.7256,
+      "step": 10210
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.39821648808393,
+      "learning_rate": 2.8310482374066843e-06,
+      "loss": 0.6156,
+      "step": 10211
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.287087842181638,
+      "learning_rate": 2.830114394723924e-06,
+      "loss": 0.5533,
+      "step": 10212
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4610292353450127,
+      "learning_rate": 2.8291806452859803e-06,
+      "loss": 0.7483,
+      "step": 10213
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3675405252176067,
+      "learning_rate": 2.82824698913298e-06,
+      "loss": 0.5487,
+      "step": 10214
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.408355178587845,
+      "learning_rate": 2.827313426305042e-06,
+      "loss": 0.775,
+      "step": 10215
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3356038545196625,
+      "learning_rate": 2.826379956842284e-06,
+      "loss": 0.691,
+      "step": 10216
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5105619349178236,
+      "learning_rate": 2.825446580784822e-06,
+      "loss": 0.6373,
+      "step": 10217
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4873035034508257,
+      "learning_rate": 2.824513298172763e-06,
+      "loss": 0.7737,
+      "step": 10218
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5191923790544952,
+      "learning_rate": 2.823580109046212e-06,
+      "loss": 0.6644,
+      "step": 10219
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2695415543102477,
+      "learning_rate": 2.822647013445272e-06,
+      "loss": 0.5626,
+      "step": 10220
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.533419824186297,
+      "learning_rate": 2.821714011410039e-06,
+      "loss": 0.669,
+      "step": 10221
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.187381334176573,
+      "learning_rate": 2.820781102980606e-06,
+      "loss": 0.6308,
+      "step": 10222
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2229950336438349,
+      "learning_rate": 2.8198482881970626e-06,
+      "loss": 0.7153,
+      "step": 10223
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3983238504640623,
+      "learning_rate": 2.8189155670994913e-06,
+      "loss": 0.6771,
+      "step": 10224
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4260898304685294,
+      "learning_rate": 2.8179829397279795e-06,
+      "loss": 0.6745,
+      "step": 10225
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3571633668153542,
+      "learning_rate": 2.8170504061225966e-06,
+      "loss": 0.6008,
+      "step": 10226
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.107627706729804,
+      "learning_rate": 2.8161179663234215e-06,
+      "loss": 0.6285,
+      "step": 10227
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2623558535039943,
+      "learning_rate": 2.8151856203705215e-06,
+      "loss": 0.5606,
+      "step": 10228
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5649871758902163,
+      "learning_rate": 2.814253368303961e-06,
+      "loss": 0.6807,
+      "step": 10229
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.363354114238057,
+      "learning_rate": 2.813321210163801e-06,
+      "loss": 0.6176,
+      "step": 10230
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3945475122716486,
+      "learning_rate": 2.8123891459900967e-06,
+      "loss": 0.7297,
+      "step": 10231
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3620917413627578,
+      "learning_rate": 2.8114571758229064e-06,
+      "loss": 0.5228,
+      "step": 10232
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5486249416805575,
+      "learning_rate": 2.8105252997022737e-06,
+      "loss": 0.6594,
+      "step": 10233
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4885564342657893,
+      "learning_rate": 2.809593517668243e-06,
+      "loss": 0.6895,
+      "step": 10234
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.393697802313881,
+      "learning_rate": 2.80866182976086e-06,
+      "loss": 0.7058,
+      "step": 10235
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3474098055577215,
+      "learning_rate": 2.8077302360201575e-06,
+      "loss": 0.6127,
+      "step": 10236
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.616909519062497,
+      "learning_rate": 2.80679873648617e-06,
+      "loss": 0.6734,
+      "step": 10237
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2558089924459952,
+      "learning_rate": 2.8058673311989247e-06,
+      "loss": 0.6458,
+      "step": 10238
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.40236724713278,
+      "learning_rate": 2.804936020198447e-06,
+      "loss": 0.6674,
+      "step": 10239
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4224110840035196,
+      "learning_rate": 2.8040048035247576e-06,
+      "loss": 0.728,
+      "step": 10240
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2667484744496191,
+      "learning_rate": 2.8030736812178717e-06,
+      "loss": 0.5712,
+      "step": 10241
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.489988370742548,
+      "learning_rate": 2.8021426533178007e-06,
+      "loss": 0.6743,
+      "step": 10242
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4891691052443161,
+      "learning_rate": 2.8012117198645596e-06,
+      "loss": 0.7489,
+      "step": 10243
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.034813487064621,
+      "learning_rate": 2.800280880898143e-06,
+      "loss": 0.724,
+      "step": 10244
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3637075294751402,
+      "learning_rate": 2.7993501364585584e-06,
+      "loss": 0.5897,
+      "step": 10245
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5868616171064844,
+      "learning_rate": 2.7984194865857994e-06,
+      "loss": 0.7038,
+      "step": 10246
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2570016115417089,
+      "learning_rate": 2.7974889313198584e-06,
+      "loss": 0.5567,
+      "step": 10247
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.084741323301433,
+      "learning_rate": 2.796558470700723e-06,
+      "loss": 0.5612,
+      "step": 10248
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2451968606314618,
+      "learning_rate": 2.795628104768376e-06,
+      "loss": 0.6155,
+      "step": 10249
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2639047995203139,
+      "learning_rate": 2.7946978335628023e-06,
+      "loss": 0.6154,
+      "step": 10250
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.010084081208681,
+      "learning_rate": 2.793767657123972e-06,
+      "loss": 0.6438,
+      "step": 10251
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.137372282798929,
+      "learning_rate": 2.792837575491857e-06,
+      "loss": 0.5571,
+      "step": 10252
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4915796713803546,
+      "learning_rate": 2.7919075887064296e-06,
+      "loss": 0.6548,
+      "step": 10253
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2150628542432775,
+      "learning_rate": 2.79097769680765e-06,
+      "loss": 0.6232,
+      "step": 10254
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4375018146230136,
+      "learning_rate": 2.790047899835479e-06,
+      "loss": 0.6451,
+      "step": 10255
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4184999972755985,
+      "learning_rate": 2.789118197829872e-06,
+      "loss": 0.6009,
+      "step": 10256
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.412896420645838,
+      "learning_rate": 2.7881885908307794e-06,
+      "loss": 0.6692,
+      "step": 10257
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.43292374452512,
+      "learning_rate": 2.7872590788781483e-06,
+      "loss": 0.6915,
+      "step": 10258
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5202297833146232,
+      "learning_rate": 2.7863296620119217e-06,
+      "loss": 0.7285,
+      "step": 10259
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3859067527092543,
+      "learning_rate": 2.7854003402720416e-06,
+      "loss": 0.6865,
+      "step": 10260
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6118464523126437,
+      "learning_rate": 2.7844711136984425e-06,
+      "loss": 0.6454,
+      "step": 10261
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2825362666425832,
+      "learning_rate": 2.7835419823310507e-06,
+      "loss": 0.6498,
+      "step": 10262
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0961059170343987,
+      "learning_rate": 2.7826129462097974e-06,
+      "loss": 0.5005,
+      "step": 10263
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1900392930400625,
+      "learning_rate": 2.781684005374604e-06,
+      "loss": 0.6005,
+      "step": 10264
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3310680215843842,
+      "learning_rate": 2.7807551598653905e-06,
+      "loss": 0.6292,
+      "step": 10265
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4903702241093506,
+      "learning_rate": 2.7798264097220694e-06,
+      "loss": 0.6921,
+      "step": 10266
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4671198864073125,
+      "learning_rate": 2.7788977549845503e-06,
+      "loss": 0.5876,
+      "step": 10267
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.2033840981646575,
+      "learning_rate": 2.777969195692745e-06,
+      "loss": 0.7471,
+      "step": 10268
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4404908684146482,
+      "learning_rate": 2.777040731886549e-06,
+      "loss": 0.6658,
+      "step": 10269
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4461019411643434,
+      "learning_rate": 2.7761123636058645e-06,
+      "loss": 0.7454,
+      "step": 10270
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.351170980847609,
+      "learning_rate": 2.775184090890585e-06,
+      "loss": 0.5881,
+      "step": 10271
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4030933659677705,
+      "learning_rate": 2.7742559137806e-06,
+      "loss": 0.6326,
+      "step": 10272
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3309185929253318,
+      "learning_rate": 2.773327832315796e-06,
+      "loss": 0.5962,
+      "step": 10273
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2590050992193125,
+      "learning_rate": 2.7723998465360537e-06,
+      "loss": 0.5573,
+      "step": 10274
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3425682419301213,
+      "learning_rate": 2.771471956481251e-06,
+      "loss": 0.6259,
+      "step": 10275
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.43485649421561,
+      "learning_rate": 2.770544162191261e-06,
+      "loss": 0.691,
+      "step": 10276
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2541251120323451,
+      "learning_rate": 2.7696164637059527e-06,
+      "loss": 0.608,
+      "step": 10277
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4338506522762922,
+      "learning_rate": 2.7686888610651936e-06,
+      "loss": 0.6486,
+      "step": 10278
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4387656924793977,
+      "learning_rate": 2.7677613543088432e-06,
+      "loss": 0.6124,
+      "step": 10279
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1889772138728383,
+      "learning_rate": 2.76683394347676e-06,
+      "loss": 0.5664,
+      "step": 10280
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6095406618406038,
+      "learning_rate": 2.7659066286087947e-06,
+      "loss": 0.6305,
+      "step": 10281
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3816398187602086,
+      "learning_rate": 2.7649794097447956e-06,
+      "loss": 0.5026,
+      "step": 10282
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2937804992081223,
+      "learning_rate": 2.7640522869246134e-06,
+      "loss": 0.6044,
+      "step": 10283
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.366875043915891,
+      "learning_rate": 2.7631252601880816e-06,
+      "loss": 0.5414,
+      "step": 10284
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0735182498696592,
+      "learning_rate": 2.762198329575037e-06,
+      "loss": 0.7532,
+      "step": 10285
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4388473377474356,
+      "learning_rate": 2.7612714951253183e-06,
+      "loss": 0.6636,
+      "step": 10286
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7495765703470052,
+      "learning_rate": 2.7603447568787456e-06,
+      "loss": 0.7128,
+      "step": 10287
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1836611409175537,
+      "learning_rate": 2.759418114875148e-06,
+      "loss": 0.5746,
+      "step": 10288
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.740618892849135,
+      "learning_rate": 2.7584915691543444e-06,
+      "loss": 0.6691,
+      "step": 10289
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4132306258243037,
+      "learning_rate": 2.7575651197561504e-06,
+      "loss": 0.6704,
+      "step": 10290
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3919345550390703,
+      "learning_rate": 2.7566387667203776e-06,
+      "loss": 0.8076,
+      "step": 10291
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1095747580980975,
+      "learning_rate": 2.7557125100868327e-06,
+      "loss": 0.651,
+      "step": 10292
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0341976651447593,
+      "learning_rate": 2.75478634989532e-06,
+      "loss": 0.6383,
+      "step": 10293
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2516285245232424,
+      "learning_rate": 2.753860286185637e-06,
+      "loss": 0.6202,
+      "step": 10294
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4228227612926703,
+      "learning_rate": 2.7529343189975788e-06,
+      "loss": 0.6411,
+      "step": 10295
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4004221418074914,
+      "learning_rate": 2.7520084483709397e-06,
+      "loss": 0.9036,
+      "step": 10296
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3979339901735133,
+      "learning_rate": 2.7510826743455037e-06,
+      "loss": 0.6791,
+      "step": 10297
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.493240632800153,
+      "learning_rate": 2.7501569969610544e-06,
+      "loss": 0.6559,
+      "step": 10298
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0982196531156343,
+      "learning_rate": 2.7492314162573687e-06,
+      "loss": 0.5768,
+      "step": 10299
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2177001712439217,
+      "learning_rate": 2.74830593227422e-06,
+      "loss": 0.6183,
+      "step": 10300
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4395027715241062,
+      "learning_rate": 2.7473805450513846e-06,
+      "loss": 0.7349,
+      "step": 10301
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.577212291321699,
+      "learning_rate": 2.7464552546286217e-06,
+      "loss": 0.7587,
+      "step": 10302
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3695012635724364,
+      "learning_rate": 2.7455300610456935e-06,
+      "loss": 0.66,
+      "step": 10303
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2495187261478014,
+      "learning_rate": 2.744604964342364e-06,
+      "loss": 0.5756,
+      "step": 10304
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3889760576823142,
+      "learning_rate": 2.743679964558378e-06,
+      "loss": 0.5291,
+      "step": 10305
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5640618891148932,
+      "learning_rate": 2.7427550617334905e-06,
+      "loss": 0.6664,
+      "step": 10306
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2731109579129725,
+      "learning_rate": 2.741830255907445e-06,
+      "loss": 0.7552,
+      "step": 10307
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.280883047317916,
+      "learning_rate": 2.740905547119983e-06,
+      "loss": 0.6411,
+      "step": 10308
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3725307361316703,
+      "learning_rate": 2.7399809354108415e-06,
+      "loss": 0.634,
+      "step": 10309
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6274026551790406,
+      "learning_rate": 2.73905642081975e-06,
+      "loss": 0.6361,
+      "step": 10310
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5398526484731727,
+      "learning_rate": 2.7381320033864434e-06,
+      "loss": 0.6788,
+      "step": 10311
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.494575055694621,
+      "learning_rate": 2.73720768315064e-06,
+      "loss": 0.7646,
+      "step": 10312
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3283622369901613,
+      "learning_rate": 2.7362834601520603e-06,
+      "loss": 0.5978,
+      "step": 10313
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2557749043805477,
+      "learning_rate": 2.735359334430424e-06,
+      "loss": 0.5665,
+      "step": 10314
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.76575473173603,
+      "learning_rate": 2.73443530602544e-06,
+      "loss": 0.7551,
+      "step": 10315
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1917461048093714,
+      "learning_rate": 2.733511374976817e-06,
+      "loss": 0.6328,
+      "step": 10316
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1951523325239313,
+      "learning_rate": 2.7325875413242585e-06,
+      "loss": 0.5867,
+      "step": 10317
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2878666461969892,
+      "learning_rate": 2.7316638051074605e-06,
+      "loss": 0.6179,
+      "step": 10318
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.541080596054616,
+      "learning_rate": 2.7307401663661247e-06,
+      "loss": 0.5839,
+      "step": 10319
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.173934702682459,
+      "learning_rate": 2.729816625139934e-06,
+      "loss": 0.7206,
+      "step": 10320
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.141055555206971,
+      "learning_rate": 2.728893181468581e-06,
+      "loss": 0.7078,
+      "step": 10321
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7309621400261852,
+      "learning_rate": 2.7279698353917476e-06,
+      "loss": 0.6754,
+      "step": 10322
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3640709531169577,
+      "learning_rate": 2.727046586949107e-06,
+      "loss": 0.5982,
+      "step": 10323
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.291767735170827,
+      "learning_rate": 2.7261234361803383e-06,
+      "loss": 0.6281,
+      "step": 10324
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3920628166960844,
+      "learning_rate": 2.72520038312511e-06,
+      "loss": 0.6755,
+      "step": 10325
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.554557369349747,
+      "learning_rate": 2.7242774278230876e-06,
+      "loss": 0.7449,
+      "step": 10326
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2810028625103855,
+      "learning_rate": 2.723354570313932e-06,
+      "loss": 0.6322,
+      "step": 10327
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1823919475962297,
+      "learning_rate": 2.7224318106372995e-06,
+      "loss": 0.5906,
+      "step": 10328
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2466046847155667,
+      "learning_rate": 2.721509148832847e-06,
+      "loss": 0.7386,
+      "step": 10329
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.323154687244469,
+      "learning_rate": 2.7205865849402187e-06,
+      "loss": 0.7112,
+      "step": 10330
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4254270089199594,
+      "learning_rate": 2.7196641189990623e-06,
+      "loss": 0.6479,
+      "step": 10331
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3409995102352008,
+      "learning_rate": 2.7187417510490176e-06,
+      "loss": 0.7222,
+      "step": 10332
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.225546447003138,
+      "learning_rate": 2.7178194811297204e-06,
+      "loss": 0.5775,
+      "step": 10333
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.502449044934449,
+      "learning_rate": 2.7168973092808025e-06,
+      "loss": 0.6132,
+      "step": 10334
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.373226123128492,
+      "learning_rate": 2.7159752355418927e-06,
+      "loss": 0.7175,
+      "step": 10335
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.412809686756924,
+      "learning_rate": 2.7150532599526114e-06,
+      "loss": 0.6765,
+      "step": 10336
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8761252834779307,
+      "learning_rate": 2.7141313825525845e-06,
+      "loss": 0.6912,
+      "step": 10337
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2980782163147815,
+      "learning_rate": 2.7132096033814193e-06,
+      "loss": 0.6868,
+      "step": 10338
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3035917709691722,
+      "learning_rate": 2.7122879224787315e-06,
+      "loss": 0.5567,
+      "step": 10339
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2851965853403362,
+      "learning_rate": 2.7113663398841296e-06,
+      "loss": 0.6314,
+      "step": 10340
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5606710801367445,
+      "learning_rate": 2.7104448556372086e-06,
+      "loss": 0.7146,
+      "step": 10341
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7993134467739789,
+      "learning_rate": 2.709523469777573e-06,
+      "loss": 0.6315,
+      "step": 10342
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4425250566425079,
+      "learning_rate": 2.7086021823448155e-06,
+      "loss": 0.6397,
+      "step": 10343
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2445244439821141,
+      "learning_rate": 2.7076809933785254e-06,
+      "loss": 0.6715,
+      "step": 10344
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.333069313110237,
+      "learning_rate": 2.7067599029182884e-06,
+      "loss": 0.6679,
+      "step": 10345
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.373406307904956,
+      "learning_rate": 2.7058389110036835e-06,
+      "loss": 0.5898,
+      "step": 10346
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4443722016802236,
+      "learning_rate": 2.704918017674294e-06,
+      "loss": 0.7531,
+      "step": 10347
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4698085324566585,
+      "learning_rate": 2.7039972229696847e-06,
+      "loss": 0.6148,
+      "step": 10348
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4705370036760148,
+      "learning_rate": 2.70307652692943e-06,
+      "loss": 0.5935,
+      "step": 10349
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4581921692147726,
+      "learning_rate": 2.702155929593093e-06,
+      "loss": 0.7707,
+      "step": 10350
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4511928319102452,
+      "learning_rate": 2.7012354310002333e-06,
+      "loss": 0.6092,
+      "step": 10351
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1880011248025895,
+      "learning_rate": 2.700315031190406e-06,
+      "loss": 0.546,
+      "step": 10352
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1747951730943982,
+      "learning_rate": 2.6993947302031643e-06,
+      "loss": 0.5539,
+      "step": 10353
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3605498836285401,
+      "learning_rate": 2.6984745280780524e-06,
+      "loss": 0.7091,
+      "step": 10354
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1724114376420292,
+      "learning_rate": 2.6975544248546195e-06,
+      "loss": 0.6575,
+      "step": 10355
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.265833887840222,
+      "learning_rate": 2.696634420572397e-06,
+      "loss": 0.5215,
+      "step": 10356
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3819804216622888,
+      "learning_rate": 2.6957145152709253e-06,
+      "loss": 0.6868,
+      "step": 10357
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.076983515121069,
+      "learning_rate": 2.694794708989733e-06,
+      "loss": 0.7528,
+      "step": 10358
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4156570466216083,
+      "learning_rate": 2.6938750017683457e-06,
+      "loss": 0.766,
+      "step": 10359
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3259968987860298,
+      "learning_rate": 2.692955393646286e-06,
+      "loss": 0.626,
+      "step": 10360
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5073392373494046,
+      "learning_rate": 2.6920358846630688e-06,
+      "loss": 0.6268,
+      "step": 10361
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2439925746204417,
+      "learning_rate": 2.691116474858214e-06,
+      "loss": 0.5981,
+      "step": 10362
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2127882093908913,
+      "learning_rate": 2.690197164271224e-06,
+      "loss": 0.5927,
+      "step": 10363
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2858698849756787,
+      "learning_rate": 2.6892779529416045e-06,
+      "loss": 0.5996,
+      "step": 10364
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5441938868064626,
+      "learning_rate": 2.6883588409088597e-06,
+      "loss": 0.69,
+      "step": 10365
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.127785552280323,
+      "learning_rate": 2.6874398282124837e-06,
+      "loss": 0.4807,
+      "step": 10366
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5808252174043307,
+      "learning_rate": 2.686520914891968e-06,
+      "loss": 0.6143,
+      "step": 10367
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.41429735822334,
+      "learning_rate": 2.6856021009868016e-06,
+      "loss": 0.747,
+      "step": 10368
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.081704320537324,
+      "learning_rate": 2.6846833865364674e-06,
+      "loss": 0.5197,
+      "step": 10369
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4756689615031184,
+      "learning_rate": 2.6837647715804444e-06,
+      "loss": 0.716,
+      "step": 10370
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5001741866255893,
+      "learning_rate": 2.682846256158206e-06,
+      "loss": 0.6655,
+      "step": 10371
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3189110551969518,
+      "learning_rate": 2.6819278403092258e-06,
+      "loss": 0.6116,
+      "step": 10372
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3791904646021074,
+      "learning_rate": 2.6810095240729715e-06,
+      "loss": 0.7083,
+      "step": 10373
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3824509177596622,
+      "learning_rate": 2.6800913074888984e-06,
+      "loss": 0.5863,
+      "step": 10374
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3888321801817995,
+      "learning_rate": 2.679173190596471e-06,
+      "loss": 0.6572,
+      "step": 10375
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4492914030201212,
+      "learning_rate": 2.6782551734351396e-06,
+      "loss": 0.6563,
+      "step": 10376
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2567958755643798,
+      "learning_rate": 2.6773372560443543e-06,
+      "loss": 0.5707,
+      "step": 10377
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.284087703041164,
+      "learning_rate": 2.6764194384635607e-06,
+      "loss": 0.6287,
+      "step": 10378
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2740980979580945,
+      "learning_rate": 2.6755017207321964e-06,
+      "loss": 0.5972,
+      "step": 10379
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.228987305672955,
+      "learning_rate": 2.6745841028897034e-06,
+      "loss": 0.6021,
+      "step": 10380
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9389829942676535,
+      "learning_rate": 2.6736665849755073e-06,
+      "loss": 0.6366,
+      "step": 10381
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.41481781104072,
+      "learning_rate": 2.6727491670290405e-06,
+      "loss": 0.6334,
+      "step": 10382
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4352118956487117,
+      "learning_rate": 2.6718318490897256e-06,
+      "loss": 0.7241,
+      "step": 10383
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5069284667350864,
+      "learning_rate": 2.6709146311969813e-06,
+      "loss": 0.6025,
+      "step": 10384
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3090453362618293,
+      "learning_rate": 2.669997513390222e-06,
+      "loss": 0.6411,
+      "step": 10385
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2959639085657708,
+      "learning_rate": 2.669080495708859e-06,
+      "loss": 0.4847,
+      "step": 10386
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.426477561514024,
+      "learning_rate": 2.668163578192298e-06,
+      "loss": 0.6821,
+      "step": 10387
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.245891649236314,
+      "learning_rate": 2.6672467608799413e-06,
+      "loss": 0.5861,
+      "step": 10388
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2770630273462393,
+      "learning_rate": 2.666330043811185e-06,
+      "loss": 0.6197,
+      "step": 10389
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3171128728362553,
+      "learning_rate": 2.6654134270254263e-06,
+      "loss": 0.5895,
+      "step": 10390
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.260041037842768,
+      "learning_rate": 2.6644969105620535e-06,
+      "loss": 0.6767,
+      "step": 10391
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4605403769030665,
+      "learning_rate": 2.663580494460446e-06,
+      "loss": 0.5656,
+      "step": 10392
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6086437182179452,
+      "learning_rate": 2.6626641787599904e-06,
+      "loss": 0.6916,
+      "step": 10393
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4704224355909312,
+      "learning_rate": 2.66174796350006e-06,
+      "loss": 0.7559,
+      "step": 10394
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.442911180647317,
+      "learning_rate": 2.660831848720028e-06,
+      "loss": 0.6352,
+      "step": 10395
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3677298762335157,
+      "learning_rate": 2.6599158344592606e-06,
+      "loss": 0.5633,
+      "step": 10396
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.442616457065481,
+      "learning_rate": 2.658999920757119e-06,
+      "loss": 0.7234,
+      "step": 10397
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2754531120131427,
+      "learning_rate": 2.6580841076529697e-06,
+      "loss": 0.5296,
+      "step": 10398
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2497951277399033,
+      "learning_rate": 2.657168395186157e-06,
+      "loss": 0.6015,
+      "step": 10399
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3662073941171236,
+      "learning_rate": 2.656252783396038e-06,
+      "loss": 0.7028,
+      "step": 10400
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.446871450711779,
+      "learning_rate": 2.655337272321957e-06,
+      "loss": 0.6547,
+      "step": 10401
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.368940014629459,
+      "learning_rate": 2.654421862003256e-06,
+      "loss": 0.6059,
+      "step": 10402
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.540432351869988,
+      "learning_rate": 2.653506552479271e-06,
+      "loss": 0.7408,
+      "step": 10403
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1704303117249157,
+      "learning_rate": 2.6525913437893346e-06,
+      "loss": 0.5852,
+      "step": 10404
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4970118803904715,
+      "learning_rate": 2.651676235972777e-06,
+      "loss": 0.6006,
+      "step": 10405
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2727738095797023,
+      "learning_rate": 2.650761229068921e-06,
+      "loss": 0.7468,
+      "step": 10406
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5498970988913376,
+      "learning_rate": 2.649846323117086e-06,
+      "loss": 0.6086,
+      "step": 10407
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.036616415317378,
+      "learning_rate": 2.6489315181565896e-06,
+      "loss": 0.641,
+      "step": 10408
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6095112231243742,
+      "learning_rate": 2.648016814226742e-06,
+      "loss": 0.6983,
+      "step": 10409
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2177575656754063,
+      "learning_rate": 2.6471022113668505e-06,
+      "loss": 0.5822,
+      "step": 10410
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6289205636672852,
+      "learning_rate": 2.646187709616217e-06,
+      "loss": 0.6812,
+      "step": 10411
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4481445800568797,
+      "learning_rate": 2.645273309014139e-06,
+      "loss": 0.6871,
+      "step": 10412
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2691492164869964,
+      "learning_rate": 2.6443590095999127e-06,
+      "loss": 0.6098,
+      "step": 10413
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5218458139240303,
+      "learning_rate": 2.6434448114128252e-06,
+      "loss": 0.7054,
+      "step": 10414
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2053146497330673,
+      "learning_rate": 2.642530714492161e-06,
+      "loss": 0.5951,
+      "step": 10415
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3860563250137292,
+      "learning_rate": 2.6416167188772052e-06,
+      "loss": 0.6465,
+      "step": 10416
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2019583926930413,
+      "learning_rate": 2.6407028246072287e-06,
+      "loss": 0.6741,
+      "step": 10417
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7619209528142792,
+      "learning_rate": 2.6397890317215067e-06,
+      "loss": 0.6937,
+      "step": 10418
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4728588988189613,
+      "learning_rate": 2.6388753402593083e-06,
+      "loss": 0.6508,
+      "step": 10419
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4114942350834256,
+      "learning_rate": 2.637961750259894e-06,
+      "loss": 0.6675,
+      "step": 10420
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.303975784386149,
+      "learning_rate": 2.6370482617625236e-06,
+      "loss": 0.7399,
+      "step": 10421
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6155487905987578,
+      "learning_rate": 2.636134874806451e-06,
+      "loss": 0.7346,
+      "step": 10422
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4336706636802008,
+      "learning_rate": 2.6352215894309306e-06,
+      "loss": 0.6029,
+      "step": 10423
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6032700244884868,
+      "learning_rate": 2.6343084056752032e-06,
+      "loss": 0.6901,
+      "step": 10424
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.305805231567921,
+      "learning_rate": 2.6333953235785105e-06,
+      "loss": 0.5714,
+      "step": 10425
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3876579023656173,
+      "learning_rate": 2.632482343180094e-06,
+      "loss": 0.682,
+      "step": 10426
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1400183155363612,
+      "learning_rate": 2.631569464519183e-06,
+      "loss": 0.5405,
+      "step": 10427
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3048548057612703,
+      "learning_rate": 2.6306566876350072e-06,
+      "loss": 0.6757,
+      "step": 10428
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3307876700260417,
+      "learning_rate": 2.6297440125667904e-06,
+      "loss": 0.7926,
+      "step": 10429
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.624591310891051,
+      "learning_rate": 2.6288314393537522e-06,
+      "loss": 0.6474,
+      "step": 10430
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2427574080105626,
+      "learning_rate": 2.627918968035108e-06,
+      "loss": 0.6232,
+      "step": 10431
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3525793762922516,
+      "learning_rate": 2.627006598650067e-06,
+      "loss": 0.6341,
+      "step": 10432
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6375227801715244,
+      "learning_rate": 2.6260943312378384e-06,
+      "loss": 0.6703,
+      "step": 10433
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2646737719538277,
+      "learning_rate": 2.6251821658376265e-06,
+      "loss": 0.6073,
+      "step": 10434
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1731776418961641,
+      "learning_rate": 2.624270102488621e-06,
+      "loss": 0.6054,
+      "step": 10435
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5238945626350848,
+      "learning_rate": 2.6233581412300222e-06,
+      "loss": 0.6549,
+      "step": 10436
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4394447108096964,
+      "learning_rate": 2.6224462821010185e-06,
+      "loss": 0.6193,
+      "step": 10437
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5992174335633105,
+      "learning_rate": 2.621534525140792e-06,
+      "loss": 0.6063,
+      "step": 10438
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.388448014031598,
+      "learning_rate": 2.620622870388524e-06,
+      "loss": 0.684,
+      "step": 10439
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2229433737956086,
+      "learning_rate": 2.619711317883389e-06,
+      "loss": 0.6024,
+      "step": 10440
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4287716374729507,
+      "learning_rate": 2.6187998676645633e-06,
+      "loss": 0.7252,
+      "step": 10441
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2504464182595756,
+      "learning_rate": 2.617888519771207e-06,
+      "loss": 0.5388,
+      "step": 10442
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3287793237230587,
+      "learning_rate": 2.616977274242488e-06,
+      "loss": 0.6522,
+      "step": 10443
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.516830837714648,
+      "learning_rate": 2.616066131117563e-06,
+      "loss": 0.676,
+      "step": 10444
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6632800774796912,
+      "learning_rate": 2.6151550904355853e-06,
+      "loss": 0.7089,
+      "step": 10445
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5174329300761935,
+      "learning_rate": 2.6142441522357055e-06,
+      "loss": 0.7746,
+      "step": 10446
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2499855095356764,
+      "learning_rate": 2.6133333165570667e-06,
+      "loss": 0.6075,
+      "step": 10447
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2808179015562307,
+      "learning_rate": 2.612422583438811e-06,
+      "loss": 0.6938,
+      "step": 10448
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1676822287360586,
+      "learning_rate": 2.6115119529200748e-06,
+      "loss": 0.5043,
+      "step": 10449
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5250049576341604,
+      "learning_rate": 2.6106014250399868e-06,
+      "loss": 0.747,
+      "step": 10450
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.552985595774383,
+      "learning_rate": 2.6096909998376794e-06,
+      "loss": 0.6597,
+      "step": 10451
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8721571775685815,
+      "learning_rate": 2.6087806773522755e-06,
+      "loss": 0.5701,
+      "step": 10452
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5167345626765119,
+      "learning_rate": 2.6078704576228876e-06,
+      "loss": 0.6945,
+      "step": 10453
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5562776774275242,
+      "learning_rate": 2.6069603406886347e-06,
+      "loss": 0.6659,
+      "step": 10454
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.617334125538553,
+      "learning_rate": 2.606050326588626e-06,
+      "loss": 0.7317,
+      "step": 10455
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4357422779242883,
+      "learning_rate": 2.605140415361967e-06,
+      "loss": 0.7168,
+      "step": 10456
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7665162843351572,
+      "learning_rate": 2.6042306070477576e-06,
+      "loss": 0.7125,
+      "step": 10457
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3597051887489975,
+      "learning_rate": 2.6033209016850926e-06,
+      "loss": 0.6303,
+      "step": 10458
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.246570951454681,
+      "learning_rate": 2.60241129931307e-06,
+      "loss": 0.636,
+      "step": 10459
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.550280032078867,
+      "learning_rate": 2.601501799970769e-06,
+      "loss": 0.7389,
+      "step": 10460
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4366923732051886,
+      "learning_rate": 2.6005924036972807e-06,
+      "loss": 0.5913,
+      "step": 10461
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.826800294965755,
+      "learning_rate": 2.59968311053168e-06,
+      "loss": 0.7466,
+      "step": 10462
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6839402047154874,
+      "learning_rate": 2.5987739205130413e-06,
+      "loss": 0.7605,
+      "step": 10463
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3139496496224843,
+      "learning_rate": 2.597864833680436e-06,
+      "loss": 0.5977,
+      "step": 10464
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4380532171115783,
+      "learning_rate": 2.596955850072928e-06,
+      "loss": 0.6559,
+      "step": 10465
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4939073876076072,
+      "learning_rate": 2.5960469697295794e-06,
+      "loss": 0.8508,
+      "step": 10466
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2133439241997146,
+      "learning_rate": 2.595138192689446e-06,
+      "loss": 0.5172,
+      "step": 10467
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2798067825050263,
+      "learning_rate": 2.594229518991579e-06,
+      "loss": 0.6479,
+      "step": 10468
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3507452869788885,
+      "learning_rate": 2.593320948675029e-06,
+      "loss": 0.6286,
+      "step": 10469
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.354071669733383,
+      "learning_rate": 2.5924124817788377e-06,
+      "loss": 0.6443,
+      "step": 10470
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4243795629936264,
+      "learning_rate": 2.591504118342044e-06,
+      "loss": 0.5932,
+      "step": 10471
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4870112081317082,
+      "learning_rate": 2.5905958584036826e-06,
+      "loss": 0.6526,
+      "step": 10472
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4020737415452045,
+      "learning_rate": 2.5896877020027813e-06,
+      "loss": 0.581,
+      "step": 10473
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1492154164857005,
+      "learning_rate": 2.588779649178371e-06,
+      "loss": 0.4663,
+      "step": 10474
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4013291869758937,
+      "learning_rate": 2.5878716999694675e-06,
+      "loss": 0.6307,
+      "step": 10475
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7744097668838907,
+      "learning_rate": 2.5869638544150876e-06,
+      "loss": 0.7919,
+      "step": 10476
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7899190363553825,
+      "learning_rate": 2.586056112554248e-06,
+      "loss": 0.7691,
+      "step": 10477
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3177815670046917,
+      "learning_rate": 2.5851484744259503e-06,
+      "loss": 0.5716,
+      "step": 10478
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.545546661489331,
+      "learning_rate": 2.5842409400692026e-06,
+      "loss": 0.7571,
+      "step": 10479
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5060804001727202,
+      "learning_rate": 2.5833335095230016e-06,
+      "loss": 0.6957,
+      "step": 10480
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1286882531362021,
+      "learning_rate": 2.5824261828263424e-06,
+      "loss": 0.5978,
+      "step": 10481
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4750890254996047,
+      "learning_rate": 2.5815189600182145e-06,
+      "loss": 0.6132,
+      "step": 10482
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6332848354900316,
+      "learning_rate": 2.5806118411376e-06,
+      "loss": 0.6615,
+      "step": 10483
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.442869802469921,
+      "learning_rate": 2.579704826223488e-06,
+      "loss": 0.6458,
+      "step": 10484
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3601873812708578,
+      "learning_rate": 2.578797915314848e-06,
+      "loss": 0.6651,
+      "step": 10485
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2538250945723575,
+      "learning_rate": 2.577891108450651e-06,
+      "loss": 0.5639,
+      "step": 10486
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0403375213659336,
+      "learning_rate": 2.5769844056698702e-06,
+      "loss": 0.605,
+      "step": 10487
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.425227635293567,
+      "learning_rate": 2.576077807011465e-06,
+      "loss": 0.5835,
+      "step": 10488
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4827244974937004,
+      "learning_rate": 2.575171312514395e-06,
+      "loss": 0.6909,
+      "step": 10489
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4813743351019237,
+      "learning_rate": 2.5742649222176147e-06,
+      "loss": 0.5868,
+      "step": 10490
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6084087762064954,
+      "learning_rate": 2.573358636160071e-06,
+      "loss": 0.6985,
+      "step": 10491
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.230931884320263,
+      "learning_rate": 2.572452454380715e-06,
+      "loss": 0.6803,
+      "step": 10492
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1299426840913407,
+      "learning_rate": 2.571546376918479e-06,
+      "loss": 0.6015,
+      "step": 10493
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3594078122860795,
+      "learning_rate": 2.570640403812306e-06,
+      "loss": 0.5883,
+      "step": 10494
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2330894090679456,
+      "learning_rate": 2.569734535101127e-06,
+      "loss": 0.5946,
+      "step": 10495
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3245569521349014,
+      "learning_rate": 2.5688287708238647e-06,
+      "loss": 0.6697,
+      "step": 10496
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4231235209539148,
+      "learning_rate": 2.5679231110194454e-06,
+      "loss": 0.6956,
+      "step": 10497
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.407162992944031,
+      "learning_rate": 2.567017555726787e-06,
+      "loss": 0.6937,
+      "step": 10498
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.518519573794622,
+      "learning_rate": 2.5661121049848026e-06,
+      "loss": 0.6345,
+      "step": 10499
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2348634660470479,
+      "learning_rate": 2.5652067588324015e-06,
+      "loss": 0.6366,
+      "step": 10500
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7105057174612974,
+      "learning_rate": 2.564301517308487e-06,
+      "loss": 0.6436,
+      "step": 10501
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1392267828847586,
+      "learning_rate": 2.5633963804519623e-06,
+      "loss": 0.7167,
+      "step": 10502
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4505490715060987,
+      "learning_rate": 2.5624913483017246e-06,
+      "loss": 0.6622,
+      "step": 10503
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.183688297398832,
+      "learning_rate": 2.5615864208966573e-06,
+      "loss": 0.5704,
+      "step": 10504
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2929312107753892,
+      "learning_rate": 2.560681598275655e-06,
+      "loss": 0.5852,
+      "step": 10505
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.464583627153267,
+      "learning_rate": 2.559776880477596e-06,
+      "loss": 0.7358,
+      "step": 10506
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.387341927767103,
+      "learning_rate": 2.55887226754136e-06,
+      "loss": 0.6463,
+      "step": 10507
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4948640623561524,
+      "learning_rate": 2.5579677595058196e-06,
+      "loss": 0.6016,
+      "step": 10508
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.365075765297574,
+      "learning_rate": 2.55706335640984e-06,
+      "loss": 0.664,
+      "step": 10509
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4090374029867132,
+      "learning_rate": 2.5561590582922934e-06,
+      "loss": 0.6544,
+      "step": 10510
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0078834793904075,
+      "learning_rate": 2.5552548651920317e-06,
+      "loss": 0.6443,
+      "step": 10511
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1768972486820422,
+      "learning_rate": 2.5543507771479137e-06,
+      "loss": 0.557,
+      "step": 10512
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7176137341663436,
+      "learning_rate": 2.553446794198793e-06,
+      "loss": 0.6794,
+      "step": 10513
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.717596779084223,
+      "learning_rate": 2.552542916383507e-06,
+      "loss": 0.6803,
+      "step": 10514
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4025151909951679,
+      "learning_rate": 2.551639143740905e-06,
+      "loss": 0.6995,
+      "step": 10515
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1113733038496962,
+      "learning_rate": 2.5507354763098226e-06,
+      "loss": 0.6496,
+      "step": 10516
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4839903048467964,
+      "learning_rate": 2.5498319141290904e-06,
+      "loss": 0.6577,
+      "step": 10517
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.298216564801722,
+      "learning_rate": 2.548928457237538e-06,
+      "loss": 0.597,
+      "step": 10518
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0675049841655118,
+      "learning_rate": 2.5480251056739874e-06,
+      "loss": 0.6056,
+      "step": 10519
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2980179848627822,
+      "learning_rate": 2.5471218594772602e-06,
+      "loss": 0.6574,
+      "step": 10520
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4325613599417972,
+      "learning_rate": 2.5462187186861697e-06,
+      "loss": 0.6851,
+      "step": 10521
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3303688397912803,
+      "learning_rate": 2.5453156833395254e-06,
+      "loss": 0.7555,
+      "step": 10522
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.319326347167239,
+      "learning_rate": 2.544412753476134e-06,
+      "loss": 0.5962,
+      "step": 10523
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3372193596394981,
+      "learning_rate": 2.543509929134794e-06,
+      "loss": 0.693,
+      "step": 10524
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5143734759415968,
+      "learning_rate": 2.5426072103543043e-06,
+      "loss": 0.5551,
+      "step": 10525
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4676237577058084,
+      "learning_rate": 2.5417045971734556e-06,
+      "loss": 0.6906,
+      "step": 10526
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2717719780353933,
+      "learning_rate": 2.5408020896310333e-06,
+      "loss": 0.5977,
+      "step": 10527
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0395905885897585,
+      "learning_rate": 2.5398996877658256e-06,
+      "loss": 0.5987,
+      "step": 10528
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.442010694504482,
+      "learning_rate": 2.5389973916166037e-06,
+      "loss": 0.5912,
+      "step": 10529
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.626396379910216,
+      "learning_rate": 2.5380952012221473e-06,
+      "loss": 0.786,
+      "step": 10530
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.237802984076152,
+      "learning_rate": 2.5371931166212216e-06,
+      "loss": 0.6259,
+      "step": 10531
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3531280038717588,
+      "learning_rate": 2.536291137852594e-06,
+      "loss": 0.7377,
+      "step": 10532
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5297130294279588,
+      "learning_rate": 2.535389264955023e-06,
+      "loss": 0.6579,
+      "step": 10533
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.255070299081247,
+      "learning_rate": 2.534487497967262e-06,
+      "loss": 0.5945,
+      "step": 10534
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7968120856992356,
+      "learning_rate": 2.5335858369280674e-06,
+      "loss": 0.7027,
+      "step": 10535
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5004572518097314,
+      "learning_rate": 2.5326842818761814e-06,
+      "loss": 0.7533,
+      "step": 10536
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8084970122138888,
+      "learning_rate": 2.5317828328503445e-06,
+      "loss": 0.7474,
+      "step": 10537
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3765074149165393,
+      "learning_rate": 2.5308814898892985e-06,
+      "loss": 0.6001,
+      "step": 10538
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5865560723000385,
+      "learning_rate": 2.529980253031774e-06,
+      "loss": 0.7616,
+      "step": 10539
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3942176725249678,
+      "learning_rate": 2.529079122316499e-06,
+      "loss": 0.7402,
+      "step": 10540
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3913558066098477,
+      "learning_rate": 2.5281780977821967e-06,
+      "loss": 0.6171,
+      "step": 10541
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2619992053188875,
+      "learning_rate": 2.5272771794675866e-06,
+      "loss": 0.5923,
+      "step": 10542
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0238184565567936,
+      "learning_rate": 2.526376367411384e-06,
+      "loss": 0.5216,
+      "step": 10543
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.180411989307027,
+      "learning_rate": 2.5254756616522953e-06,
+      "loss": 0.6233,
+      "step": 10544
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2863549576243638,
+      "learning_rate": 2.52457506222903e-06,
+      "loss": 0.6585,
+      "step": 10545
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6852426204235276,
+      "learning_rate": 2.52367456918029e-06,
+      "loss": 0.6993,
+      "step": 10546
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4346616263772718,
+      "learning_rate": 2.522774182544765e-06,
+      "loss": 0.6269,
+      "step": 10547
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2271270596011135,
+      "learning_rate": 2.5218739023611523e-06,
+      "loss": 0.5931,
+      "step": 10548
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0084920712109078,
+      "learning_rate": 2.5209737286681367e-06,
+      "loss": 0.6112,
+      "step": 10549
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2305125150293437,
+      "learning_rate": 2.5200736615044012e-06,
+      "loss": 0.5927,
+      "step": 10550
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4189546350519266,
+      "learning_rate": 2.5191737009086235e-06,
+      "loss": 0.5991,
+      "step": 10551
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.017572652238589,
+      "learning_rate": 2.5182738469194757e-06,
+      "loss": 0.6602,
+      "step": 10552
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3795158361676065,
+      "learning_rate": 2.5173740995756313e-06,
+      "loss": 0.7066,
+      "step": 10553
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6005632216672507,
+      "learning_rate": 2.5164744589157488e-06,
+      "loss": 0.6069,
+      "step": 10554
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.246056267638046,
+      "learning_rate": 2.5155749249784876e-06,
+      "loss": 0.6066,
+      "step": 10555
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.32555761702842,
+      "learning_rate": 2.514675497802508e-06,
+      "loss": 0.6221,
+      "step": 10556
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.470161556296513,
+      "learning_rate": 2.513776177426457e-06,
+      "loss": 0.6664,
+      "step": 10557
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1625150095285801,
+      "learning_rate": 2.51287696388898e-06,
+      "loss": 0.6598,
+      "step": 10558
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4657494327344405,
+      "learning_rate": 2.5119778572287195e-06,
+      "loss": 0.6379,
+      "step": 10559
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5141304226709646,
+      "learning_rate": 2.5110788574843113e-06,
+      "loss": 0.7194,
+      "step": 10560
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.225737925095066,
+      "learning_rate": 2.510179964694388e-06,
+      "loss": 0.6201,
+      "step": 10561
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.416294995374682,
+      "learning_rate": 2.5092811788975754e-06,
+      "loss": 0.681,
+      "step": 10562
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2202522619407832,
+      "learning_rate": 2.508382500132499e-06,
+      "loss": 0.6541,
+      "step": 10563
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4090837070489348,
+      "learning_rate": 2.5074839284377774e-06,
+      "loss": 0.6832,
+      "step": 10564
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3937597114896705,
+      "learning_rate": 2.506585463852019e-06,
+      "loss": 0.6662,
+      "step": 10565
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4573179977891093,
+      "learning_rate": 2.5056871064138383e-06,
+      "loss": 0.6865,
+      "step": 10566
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3268892094348461,
+      "learning_rate": 2.504788856161838e-06,
+      "loss": 0.6233,
+      "step": 10567
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3103021938255204,
+      "learning_rate": 2.5038907131346168e-06,
+      "loss": 0.5582,
+      "step": 10568
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6063297512672878,
+      "learning_rate": 2.5029926773707713e-06,
+      "loss": 0.6743,
+      "step": 10569
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4511273319926494,
+      "learning_rate": 2.50209474890889e-06,
+      "loss": 0.7972,
+      "step": 10570
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4158530632462174,
+      "learning_rate": 2.5011969277875643e-06,
+      "loss": 0.6391,
+      "step": 10571
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2575887392069907,
+      "learning_rate": 2.5002992140453674e-06,
+      "loss": 0.6751,
+      "step": 10572
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.375420363898702,
+      "learning_rate": 2.499401607720883e-06,
+      "loss": 0.5351,
+      "step": 10573
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.366551393969518,
+      "learning_rate": 2.49850410885268e-06,
+      "loss": 0.6499,
+      "step": 10574
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5979363961697783,
+      "learning_rate": 2.4976067174793268e-06,
+      "loss": 0.6102,
+      "step": 10575
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3381692185919065,
+      "learning_rate": 2.496709433639386e-06,
+      "loss": 0.6903,
+      "step": 10576
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.777904683410583,
+      "learning_rate": 2.495812257371416e-06,
+      "loss": 0.6331,
+      "step": 10577
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3605130830560657,
+      "learning_rate": 2.4949151887139705e-06,
+      "loss": 0.7555,
+      "step": 10578
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2973824244337258,
+      "learning_rate": 2.4940182277055987e-06,
+      "loss": 0.6726,
+      "step": 10579
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5436081064894196,
+      "learning_rate": 2.4931213743848428e-06,
+      "loss": 0.6428,
+      "step": 10580
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.204412116842133,
+      "learning_rate": 2.4922246287902458e-06,
+      "loss": 0.5858,
+      "step": 10581
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3239979355926736,
+      "learning_rate": 2.491327990960342e-06,
+      "loss": 0.6243,
+      "step": 10582
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2293413083656182,
+      "learning_rate": 2.4904314609336616e-06,
+      "loss": 0.5763,
+      "step": 10583
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5655191509804365,
+      "learning_rate": 2.4895350387487304e-06,
+      "loss": 0.7606,
+      "step": 10584
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.426351916423831,
+      "learning_rate": 2.48863872444407e-06,
+      "loss": 0.6727,
+      "step": 10585
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5501605542804777,
+      "learning_rate": 2.4877425180581964e-06,
+      "loss": 0.6994,
+      "step": 10586
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4030390475830379,
+      "learning_rate": 2.486846419629622e-06,
+      "loss": 0.6959,
+      "step": 10587
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3450294325293493,
+      "learning_rate": 2.4859504291968522e-06,
+      "loss": 0.6464,
+      "step": 10588
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.510694627156924,
+      "learning_rate": 2.485054546798395e-06,
+      "loss": 0.682,
+      "step": 10589
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1288312833552034,
+      "learning_rate": 2.4841587724727416e-06,
+      "loss": 0.694,
+      "step": 10590
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.373586410557063,
+      "learning_rate": 2.4832631062583906e-06,
+      "loss": 0.664,
+      "step": 10591
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4978048639731507,
+      "learning_rate": 2.4823675481938293e-06,
+      "loss": 0.658,
+      "step": 10592
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5408795824462476,
+      "learning_rate": 2.4814720983175413e-06,
+      "loss": 0.6794,
+      "step": 10593
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.424321146133001,
+      "learning_rate": 2.4805767566680057e-06,
+      "loss": 0.6016,
+      "step": 10594
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9674967950758064,
+      "learning_rate": 2.4796815232836965e-06,
+      "loss": 0.5954,
+      "step": 10595
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.295500347401912,
+      "learning_rate": 2.478786398203089e-06,
+      "loss": 0.6756,
+      "step": 10596
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2498069358685024,
+      "learning_rate": 2.4778913814646433e-06,
+      "loss": 0.6659,
+      "step": 10597
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.400438196174517,
+      "learning_rate": 2.47699647310682e-06,
+      "loss": 0.5996,
+      "step": 10598
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3602187552324738,
+      "learning_rate": 2.4761016731680792e-06,
+      "loss": 0.6385,
+      "step": 10599
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.565184949390043,
+      "learning_rate": 2.4752069816868707e-06,
+      "loss": 0.674,
+      "step": 10600
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4738179788869152,
+      "learning_rate": 2.474312398701641e-06,
+      "loss": 0.8422,
+      "step": 10601
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3773074751044612,
+      "learning_rate": 2.473417924250833e-06,
+      "loss": 0.6304,
+      "step": 10602
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3387449903426212,
+      "learning_rate": 2.472523558372883e-06,
+      "loss": 0.4987,
+      "step": 10603
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6992232326896266,
+      "learning_rate": 2.4716293011062248e-06,
+      "loss": 0.6308,
+      "step": 10604
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2672732722757898,
+      "learning_rate": 2.470735152489287e-06,
+      "loss": 0.683,
+      "step": 10605
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.468842165366907,
+      "learning_rate": 2.4698411125604903e-06,
+      "loss": 0.6729,
+      "step": 10606
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5765110004449414,
+      "learning_rate": 2.46894718135826e-06,
+      "loss": 0.7379,
+      "step": 10607
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2954455823859328,
+      "learning_rate": 2.4680533589210028e-06,
+      "loss": 0.5986,
+      "step": 10608
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5305929915323948,
+      "learning_rate": 2.467159645287133e-06,
+      "loss": 0.6176,
+      "step": 10609
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2381134068880757,
+      "learning_rate": 2.4662660404950544e-06,
+      "loss": 0.6045,
+      "step": 10610
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5408609540623401,
+      "learning_rate": 2.465372544583167e-06,
+      "loss": 0.7013,
+      "step": 10611
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4050785142201583,
+      "learning_rate": 2.4644791575898665e-06,
+      "loss": 0.6607,
+      "step": 10612
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4600905878546526,
+      "learning_rate": 2.4635858795535413e-06,
+      "loss": 0.6162,
+      "step": 10613
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.507241410703118,
+      "learning_rate": 2.4626927105125834e-06,
+      "loss": 0.7189,
+      "step": 10614
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5493810670175447,
+      "learning_rate": 2.461799650505369e-06,
+      "loss": 0.6434,
+      "step": 10615
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.303084050122424,
+      "learning_rate": 2.4609066995702745e-06,
+      "loss": 0.6715,
+      "step": 10616
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2357374635184797,
+      "learning_rate": 2.460013857745676e-06,
+      "loss": 0.5833,
+      "step": 10617
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0464693068545916,
+      "learning_rate": 2.4591211250699393e-06,
+      "loss": 0.5794,
+      "step": 10618
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2548868514628524,
+      "learning_rate": 2.4582285015814263e-06,
+      "loss": 0.6309,
+      "step": 10619
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4074901478595827,
+      "learning_rate": 2.457335987318495e-06,
+      "loss": 0.7039,
+      "step": 10620
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3054933410048843,
+      "learning_rate": 2.4564435823194977e-06,
+      "loss": 0.6226,
+      "step": 10621
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.174782193041706,
+      "learning_rate": 2.455551286622788e-06,
+      "loss": 0.6221,
+      "step": 10622
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2824468313196267,
+      "learning_rate": 2.454659100266703e-06,
+      "loss": 0.5215,
+      "step": 10623
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2402500793035822,
+      "learning_rate": 2.4537670232895866e-06,
+      "loss": 0.6885,
+      "step": 10624
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2581977582964508,
+      "learning_rate": 2.4528750557297745e-06,
+      "loss": 0.6211,
+      "step": 10625
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5815569621166221,
+      "learning_rate": 2.4519831976255892e-06,
+      "loss": 0.6752,
+      "step": 10626
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2399315472119898,
+      "learning_rate": 2.451091449015364e-06,
+      "loss": 0.6523,
+      "step": 10627
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.354459114643782,
+      "learning_rate": 2.4501998099374147e-06,
+      "loss": 0.6251,
+      "step": 10628
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3346078408987996,
+      "learning_rate": 2.4493082804300585e-06,
+      "loss": 0.651,
+      "step": 10629
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5869353172581662,
+      "learning_rate": 2.4484168605316062e-06,
+      "loss": 0.7328,
+      "step": 10630
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5377498643223804,
+      "learning_rate": 2.4475255502803624e-06,
+      "loss": 0.7523,
+      "step": 10631
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.450025606822622,
+      "learning_rate": 2.4466343497146345e-06,
+      "loss": 0.6629,
+      "step": 10632
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.294583238390682,
+      "learning_rate": 2.445743258872711e-06,
+      "loss": 0.5494,
+      "step": 10633
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.388338316365262,
+      "learning_rate": 2.4448522777928903e-06,
+      "loss": 0.5115,
+      "step": 10634
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5988091122642234,
+      "learning_rate": 2.4439614065134576e-06,
+      "loss": 0.6399,
+      "step": 10635
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9216840273785536,
+      "learning_rate": 2.443070645072696e-06,
+      "loss": 0.6102,
+      "step": 10636
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3714560659032122,
+      "learning_rate": 2.4421799935088837e-06,
+      "loss": 0.6591,
+      "step": 10637
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3971129608479953,
+      "learning_rate": 2.441289451860293e-06,
+      "loss": 0.7206,
+      "step": 10638
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2369966979937843,
+      "learning_rate": 2.4403990201651915e-06,
+      "loss": 0.5822,
+      "step": 10639
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0369197952985685,
+      "learning_rate": 2.4395086984618486e-06,
+      "loss": 0.5714,
+      "step": 10640
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4157430564246838,
+      "learning_rate": 2.4386184867885154e-06,
+      "loss": 0.6187,
+      "step": 10641
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.421974884121493,
+      "learning_rate": 2.437728385183453e-06,
+      "loss": 0.6035,
+      "step": 10642
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2773834317232544,
+      "learning_rate": 2.436838393684907e-06,
+      "loss": 0.6918,
+      "step": 10643
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7498272046488577,
+      "learning_rate": 2.435948512331125e-06,
+      "loss": 0.8126,
+      "step": 10644
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4338768257252492,
+      "learning_rate": 2.4350587411603454e-06,
+      "loss": 0.6684,
+      "step": 10645
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.493354408860496,
+      "learning_rate": 2.434169080210802e-06,
+      "loss": 0.56,
+      "step": 10646
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2772154155637594,
+      "learning_rate": 2.433279529520732e-06,
+      "loss": 0.5859,
+      "step": 10647
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3120194342543756,
+      "learning_rate": 2.432390089128354e-06,
+      "loss": 0.6422,
+      "step": 10648
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5693750288935533,
+      "learning_rate": 2.4315007590718913e-06,
+      "loss": 0.6313,
+      "step": 10649
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2127223505945726,
+      "learning_rate": 2.4306115393895645e-06,
+      "loss": 0.5385,
+      "step": 10650
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1814712897181296,
+      "learning_rate": 2.4297224301195795e-06,
+      "loss": 0.5638,
+      "step": 10651
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7538997736843196,
+      "learning_rate": 2.4288334313001467e-06,
+      "loss": 0.7098,
+      "step": 10652
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3900564684338432,
+      "learning_rate": 2.4279445429694683e-06,
+      "loss": 0.7866,
+      "step": 10653
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2622558175759173,
+      "learning_rate": 2.427055765165741e-06,
+      "loss": 0.585,
+      "step": 10654
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.492732406107266,
+      "learning_rate": 2.4261670979271577e-06,
+      "loss": 0.584,
+      "step": 10655
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.533887465824394,
+      "learning_rate": 2.4252785412919065e-06,
+      "loss": 0.603,
+      "step": 10656
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2984924712732793,
+      "learning_rate": 2.4243900952981685e-06,
+      "loss": 0.604,
+      "step": 10657
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1031823022185427,
+      "learning_rate": 2.423501759984128e-06,
+      "loss": 0.6157,
+      "step": 10658
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.237095106458682,
+      "learning_rate": 2.4226135353879516e-06,
+      "loss": 0.5947,
+      "step": 10659
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.402620856192415,
+      "learning_rate": 2.4217254215478137e-06,
+      "loss": 0.6865,
+      "step": 10660
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5071466236656015,
+      "learning_rate": 2.420837418501876e-06,
+      "loss": 0.5718,
+      "step": 10661
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4652652036398743,
+      "learning_rate": 2.4199495262882993e-06,
+      "loss": 0.8124,
+      "step": 10662
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5296867792630708,
+      "learning_rate": 2.419061744945237e-06,
+      "loss": 0.6791,
+      "step": 10663
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2878565062710432,
+      "learning_rate": 2.4181740745108377e-06,
+      "loss": 0.5766,
+      "step": 10664
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3179120661363808,
+      "learning_rate": 2.4172865150232528e-06,
+      "loss": 0.6021,
+      "step": 10665
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4472080632205748,
+      "learning_rate": 2.4163990665206153e-06,
+      "loss": 0.5397,
+      "step": 10666
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2709745201948839,
+      "learning_rate": 2.415511729041063e-06,
+      "loss": 0.7067,
+      "step": 10667
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4550692530525875,
+      "learning_rate": 2.414624502622731e-06,
+      "loss": 0.7275,
+      "step": 10668
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2941405629272869,
+      "learning_rate": 2.413737387303739e-06,
+      "loss": 0.609,
+      "step": 10669
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2124962614734347,
+      "learning_rate": 2.412850383122213e-06,
+      "loss": 0.6537,
+      "step": 10670
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0846964649860475,
+      "learning_rate": 2.411963490116267e-06,
+      "loss": 0.6347,
+      "step": 10671
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4744238913757362,
+      "learning_rate": 2.4110767083240145e-06,
+      "loss": 0.7018,
+      "step": 10672
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.584573864348503,
+      "learning_rate": 2.410190037783562e-06,
+      "loss": 0.6529,
+      "step": 10673
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4131854773747468,
+      "learning_rate": 2.4093034785330087e-06,
+      "loss": 0.5673,
+      "step": 10674
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4417685708912111,
+      "learning_rate": 2.408417030610457e-06,
+      "loss": 0.6153,
+      "step": 10675
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3096111613099815,
+      "learning_rate": 2.4075306940539988e-06,
+      "loss": 0.6982,
+      "step": 10676
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.451346564540652,
+      "learning_rate": 2.4066444689017173e-06,
+      "loss": 0.6673,
+      "step": 10677
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3547477253926845,
+      "learning_rate": 2.4057583551917e-06,
+      "loss": 0.6309,
+      "step": 10678
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6262330222370887,
+      "learning_rate": 2.4048723529620246e-06,
+      "loss": 0.504,
+      "step": 10679
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1844546198915427,
+      "learning_rate": 2.403986462250763e-06,
+      "loss": 0.6323,
+      "step": 10680
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2617939209368307,
+      "learning_rate": 2.4031006830959853e-06,
+      "loss": 0.6341,
+      "step": 10681
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4855334682639678,
+      "learning_rate": 2.4022150155357526e-06,
+      "loss": 0.5209,
+      "step": 10682
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2359218109654464,
+      "learning_rate": 2.40132945960813e-06,
+      "loss": 0.6014,
+      "step": 10683
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.638993271686653,
+      "learning_rate": 2.4004440153511642e-06,
+      "loss": 0.6891,
+      "step": 10684
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.316923669036391,
+      "learning_rate": 2.399558682802911e-06,
+      "loss": 0.631,
+      "step": 10685
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4739063673971418,
+      "learning_rate": 2.3986734620014146e-06,
+      "loss": 0.6635,
+      "step": 10686
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4049832984045922,
+      "learning_rate": 2.3977883529847087e-06,
+      "loss": 0.6509,
+      "step": 10687
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5933785436416603,
+      "learning_rate": 2.396903355790835e-06,
+      "loss": 0.4958,
+      "step": 10688
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4250429820588455,
+      "learning_rate": 2.396018470457821e-06,
+      "loss": 0.653,
+      "step": 10689
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3044823006619148,
+      "learning_rate": 2.395133697023693e-06,
+      "loss": 0.5942,
+      "step": 10690
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2652008868161304,
+      "learning_rate": 2.394249035526472e-06,
+      "loss": 0.6809,
+      "step": 10691
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.30852233646923,
+      "learning_rate": 2.3933644860041703e-06,
+      "loss": 0.5547,
+      "step": 10692
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3835052088631852,
+      "learning_rate": 2.3924800484948045e-06,
+      "loss": 0.6301,
+      "step": 10693
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0015430219761359,
+      "learning_rate": 2.3915957230363783e-06,
+      "loss": 0.6572,
+      "step": 10694
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4126712655066005,
+      "learning_rate": 2.390711509666893e-06,
+      "loss": 0.6517,
+      "step": 10695
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4508365094401758,
+      "learning_rate": 2.389827408424345e-06,
+      "loss": 0.6942,
+      "step": 10696
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.332746770111835,
+      "learning_rate": 2.388943419346727e-06,
+      "loss": 0.5664,
+      "step": 10697
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3545949227845597,
+      "learning_rate": 2.388059542472025e-06,
+      "loss": 0.6201,
+      "step": 10698
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4426034592717762,
+      "learning_rate": 2.3871757778382216e-06,
+      "loss": 0.6417,
+      "step": 10699
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4328939908202507,
+      "learning_rate": 2.386292125483292e-06,
+      "loss": 0.6738,
+      "step": 10700
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6122396838449173,
+      "learning_rate": 2.385408585445215e-06,
+      "loss": 0.6605,
+      "step": 10701
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0487058537221357,
+      "learning_rate": 2.38452515776195e-06,
+      "loss": 0.6611,
+      "step": 10702
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2461180917899934,
+      "learning_rate": 2.3836418424714665e-06,
+      "loss": 0.6068,
+      "step": 10703
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4824517316684187,
+      "learning_rate": 2.3827586396117207e-06,
+      "loss": 0.705,
+      "step": 10704
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5163901160552378,
+      "learning_rate": 2.3818755492206624e-06,
+      "loss": 0.7807,
+      "step": 10705
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.28562166617956,
+      "learning_rate": 2.3809925713362437e-06,
+      "loss": 0.7457,
+      "step": 10706
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2866268425723857,
+      "learning_rate": 2.3801097059964075e-06,
+      "loss": 0.6832,
+      "step": 10707
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5617289114783393,
+      "learning_rate": 2.3792269532390923e-06,
+      "loss": 0.7088,
+      "step": 10708
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3633208665159193,
+      "learning_rate": 2.378344313102231e-06,
+      "loss": 0.664,
+      "step": 10709
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9012683266177377,
+      "learning_rate": 2.377461785623752e-06,
+      "loss": 0.5743,
+      "step": 10710
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3673769502223525,
+      "learning_rate": 2.3765793708415825e-06,
+      "loss": 0.6868,
+      "step": 10711
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4267799958139447,
+      "learning_rate": 2.3756970687936397e-06,
+      "loss": 0.6469,
+      "step": 10712
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2432500685076067,
+      "learning_rate": 2.3748148795178387e-06,
+      "loss": 0.5524,
+      "step": 10713
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.551247495961787,
+      "learning_rate": 2.373932803052089e-06,
+      "loss": 0.7686,
+      "step": 10714
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5703639019716746,
+      "learning_rate": 2.3730508394342957e-06,
+      "loss": 0.7341,
+      "step": 10715
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2591360604588477,
+      "learning_rate": 2.372168988702358e-06,
+      "loss": 0.535,
+      "step": 10716
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.412214218091047,
+      "learning_rate": 2.3712872508941714e-06,
+      "loss": 0.648,
+      "step": 10717
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2638377577209665,
+      "learning_rate": 2.3704056260476237e-06,
+      "loss": 0.5624,
+      "step": 10718
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5098530123107794,
+      "learning_rate": 2.369524114200607e-06,
+      "loss": 0.5654,
+      "step": 10719
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3964204786607395,
+      "learning_rate": 2.368642715390994e-06,
+      "loss": 0.6499,
+      "step": 10720
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.532895212056541,
+      "learning_rate": 2.3677614296566654e-06,
+      "loss": 0.77,
+      "step": 10721
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6222987660814066,
+      "learning_rate": 2.3668802570354903e-06,
+      "loss": 0.6944,
+      "step": 10722
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.346916389841755,
+      "learning_rate": 2.365999197565335e-06,
+      "loss": 0.6058,
+      "step": 10723
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2513206202705935,
+      "learning_rate": 2.3651182512840604e-06,
+      "loss": 0.5218,
+      "step": 10724
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3516588076544582,
+      "learning_rate": 2.3642374182295207e-06,
+      "loss": 0.5742,
+      "step": 10725
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2478097759497866,
+      "learning_rate": 2.363356698439573e-06,
+      "loss": 0.604,
+      "step": 10726
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2477692360725057,
+      "learning_rate": 2.362476091952059e-06,
+      "loss": 0.6688,
+      "step": 10727
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3436142991979432,
+      "learning_rate": 2.361595598804819e-06,
+      "loss": 0.5763,
+      "step": 10728
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8120655704911084,
+      "learning_rate": 2.360715219035694e-06,
+      "loss": 0.7309,
+      "step": 10729
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.376886023046629,
+      "learning_rate": 2.3598349526825137e-06,
+      "loss": 0.6945,
+      "step": 10730
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2909088612398851,
+      "learning_rate": 2.358954799783106e-06,
+      "loss": 0.6624,
+      "step": 10731
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1231850893828592,
+      "learning_rate": 2.358074760375292e-06,
+      "loss": 0.7102,
+      "step": 10732
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5876647748758548,
+      "learning_rate": 2.3571948344968896e-06,
+      "loss": 0.6571,
+      "step": 10733
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2150200172480161,
+      "learning_rate": 2.356315022185712e-06,
+      "loss": 0.5774,
+      "step": 10734
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5549427171133026,
+      "learning_rate": 2.3554353234795624e-06,
+      "loss": 0.6019,
+      "step": 10735
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2978995825557464,
+      "learning_rate": 2.3545557384162494e-06,
+      "loss": 0.839,
+      "step": 10736
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.262318505699832,
+      "learning_rate": 2.3536762670335696e-06,
+      "loss": 0.5438,
+      "step": 10737
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4397453687933166,
+      "learning_rate": 2.3527969093693105e-06,
+      "loss": 0.676,
+      "step": 10738
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1284088504622987,
+      "learning_rate": 2.3519176654612657e-06,
+      "loss": 0.5051,
+      "step": 10739
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4535095480816518,
+      "learning_rate": 2.3510385353472163e-06,
+      "loss": 0.5809,
+      "step": 10740
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.298994504643067,
+      "learning_rate": 2.3501595190649405e-06,
+      "loss": 0.6387,
+      "step": 10741
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3851675021264762,
+      "learning_rate": 2.3492806166522125e-06,
+      "loss": 0.5797,
+      "step": 10742
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.267501230278989,
+      "learning_rate": 2.3484018281467973e-06,
+      "loss": 0.5584,
+      "step": 10743
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2568212135571062,
+      "learning_rate": 2.3475231535864653e-06,
+      "loss": 0.5562,
+      "step": 10744
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.96558821349941,
+      "learning_rate": 2.346644593008966e-06,
+      "loss": 0.7142,
+      "step": 10745
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8038497539667147,
+      "learning_rate": 2.3457661464520605e-06,
+      "loss": 0.6192,
+      "step": 10746
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3440985793047642,
+      "learning_rate": 2.3448878139534948e-06,
+      "loss": 0.593,
+      "step": 10747
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4668326336629236,
+      "learning_rate": 2.344009595551013e-06,
+      "loss": 0.7487,
+      "step": 10748
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8202181955887742,
+      "learning_rate": 2.3431314912823543e-06,
+      "loss": 0.4866,
+      "step": 10749
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3816203423976192,
+      "learning_rate": 2.3422535011852522e-06,
+      "loss": 0.6597,
+      "step": 10750
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3313203147358585,
+      "learning_rate": 2.3413756252974366e-06,
+      "loss": 0.7524,
+      "step": 10751
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5529927509665744,
+      "learning_rate": 2.3404978636566312e-06,
+      "loss": 0.6311,
+      "step": 10752
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2143397673751568,
+      "learning_rate": 2.3396202163005536e-06,
+      "loss": 0.6417,
+      "step": 10753
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3141556364584484,
+      "learning_rate": 2.338742683266923e-06,
+      "loss": 0.5923,
+      "step": 10754
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3025233410422694,
+      "learning_rate": 2.3378652645934473e-06,
+      "loss": 0.6205,
+      "step": 10755
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0170449039713385,
+      "learning_rate": 2.3369879603178264e-06,
+      "loss": 0.495,
+      "step": 10756
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2208991054483478,
+      "learning_rate": 2.336110770477766e-06,
+      "loss": 0.6228,
+      "step": 10757
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4445641062575025,
+      "learning_rate": 2.3352336951109584e-06,
+      "loss": 0.6217,
+      "step": 10758
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3290181036823108,
+      "learning_rate": 2.3343567342550933e-06,
+      "loss": 0.6589,
+      "step": 10759
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2694664520848644,
+      "learning_rate": 2.3334798879478563e-06,
+      "loss": 0.5893,
+      "step": 10760
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3811267402803369,
+      "learning_rate": 2.332603156226926e-06,
+      "loss": 0.6139,
+      "step": 10761
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2446974398235942,
+      "learning_rate": 2.3317265391299813e-06,
+      "loss": 0.5722,
+      "step": 10762
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5546637289240142,
+      "learning_rate": 2.330850036694687e-06,
+      "loss": 0.5954,
+      "step": 10763
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.402312277503259,
+      "learning_rate": 2.329973648958712e-06,
+      "loss": 0.5989,
+      "step": 10764
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2455249884347328,
+      "learning_rate": 2.3290973759597165e-06,
+      "loss": 0.783,
+      "step": 10765
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4869337935649833,
+      "learning_rate": 2.328221217735355e-06,
+      "loss": 0.6446,
+      "step": 10766
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2331670294651178,
+      "learning_rate": 2.3273451743232777e-06,
+      "loss": 0.5883,
+      "step": 10767
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0228098837187523,
+      "learning_rate": 2.326469245761131e-06,
+      "loss": 0.6726,
+      "step": 10768
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3850438745426341,
+      "learning_rate": 2.3255934320865555e-06,
+      "loss": 0.6555,
+      "step": 10769
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4014569876466072,
+      "learning_rate": 2.324717733337186e-06,
+      "loss": 0.6113,
+      "step": 10770
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3262056531838333,
+      "learning_rate": 2.3238421495506515e-06,
+      "loss": 0.5931,
+      "step": 10771
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2046645455438625,
+      "learning_rate": 2.322966680764582e-06,
+      "loss": 0.6041,
+      "step": 10772
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1302475768526894,
+      "learning_rate": 2.322091327016597e-06,
+      "loss": 0.5258,
+      "step": 10773
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4074263449443265,
+      "learning_rate": 2.3212160883443107e-06,
+      "loss": 0.5971,
+      "step": 10774
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.250865697217463,
+      "learning_rate": 2.320340964785336e-06,
+      "loss": 0.7477,
+      "step": 10775
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4481793602279114,
+      "learning_rate": 2.3194659563772754e-06,
+      "loss": 0.7135,
+      "step": 10776
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.396711295128649,
+      "learning_rate": 2.318591063157736e-06,
+      "loss": 0.653,
+      "step": 10777
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4405356511923326,
+      "learning_rate": 2.317716285164309e-06,
+      "loss": 0.6384,
+      "step": 10778
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4983328756653274,
+      "learning_rate": 2.316841622434586e-06,
+      "loss": 0.7237,
+      "step": 10779
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2011411794966151,
+      "learning_rate": 2.3159670750061563e-06,
+      "loss": 0.5996,
+      "step": 10780
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.273408349789211,
+      "learning_rate": 2.3150926429165967e-06,
+      "loss": 0.5608,
+      "step": 10781
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.318590345585312,
+      "learning_rate": 2.3142183262034865e-06,
+      "loss": 0.6652,
+      "step": 10782
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5011979864634863,
+      "learning_rate": 2.3133441249043973e-06,
+      "loss": 0.7225,
+      "step": 10783
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0165872005366428,
+      "learning_rate": 2.3124700390568945e-06,
+      "loss": 0.6604,
+      "step": 10784
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4289334357689503,
+      "learning_rate": 2.311596068698539e-06,
+      "loss": 0.6874,
+      "step": 10785
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.471242290307068,
+      "learning_rate": 2.3107222138668863e-06,
+      "loss": 0.6576,
+      "step": 10786
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2706158158126584,
+      "learning_rate": 2.3098484745994933e-06,
+      "loss": 0.6392,
+      "step": 10787
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.758056172141532,
+      "learning_rate": 2.3089748509339007e-06,
+      "loss": 0.7337,
+      "step": 10788
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.707555029013058,
+      "learning_rate": 2.30810134290765e-06,
+      "loss": 0.5878,
+      "step": 10789
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.404673979544091,
+      "learning_rate": 2.307227950558281e-06,
+      "loss": 0.5336,
+      "step": 10790
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.7948062559055757,
+      "learning_rate": 2.306354673923325e-06,
+      "loss": 0.6859,
+      "step": 10791
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4864880206114184,
+      "learning_rate": 2.3054815130403073e-06,
+      "loss": 0.6296,
+      "step": 10792
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.332889337281238,
+      "learning_rate": 2.3046084679467506e-06,
+      "loss": 0.6621,
+      "step": 10793
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2254901175282291,
+      "learning_rate": 2.3037355386801683e-06,
+      "loss": 0.644,
+      "step": 10794
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.997015120717411,
+      "learning_rate": 2.3028627252780794e-06,
+      "loss": 0.5877,
+      "step": 10795
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3094926596876428,
+      "learning_rate": 2.3019900277779816e-06,
+      "loss": 0.5828,
+      "step": 10796
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2436928632317366,
+      "learning_rate": 2.3011174462173836e-06,
+      "loss": 0.5681,
+      "step": 10797
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4225208068961983,
+      "learning_rate": 2.3002449806337807e-06,
+      "loss": 0.7243,
+      "step": 10798
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2160533486230554,
+      "learning_rate": 2.2993726310646603e-06,
+      "loss": 0.7307,
+      "step": 10799
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1133309005047918,
+      "learning_rate": 2.2985003975475146e-06,
+      "loss": 0.5895,
+      "step": 10800
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4768994264323667,
+      "learning_rate": 2.2976282801198237e-06,
+      "loss": 0.741,
+      "step": 10801
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3073977065321876,
+      "learning_rate": 2.2967562788190633e-06,
+      "loss": 0.6236,
+      "step": 10802
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1428746466423583,
+      "learning_rate": 2.2958843936827064e-06,
+      "loss": 0.6069,
+      "step": 10803
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.0746763174994127,
+      "learning_rate": 2.2950126247482178e-06,
+      "loss": 0.762,
+      "step": 10804
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3277888084860872,
+      "learning_rate": 2.2941409720530642e-06,
+      "loss": 0.7052,
+      "step": 10805
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2787325862341792,
+      "learning_rate": 2.2932694356346986e-06,
+      "loss": 0.7281,
+      "step": 10806
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.432085161069723,
+      "learning_rate": 2.292398015530571e-06,
+      "loss": 0.7387,
+      "step": 10807
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2611216054370953,
+      "learning_rate": 2.2915267117781328e-06,
+      "loss": 0.5089,
+      "step": 10808
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.483626643079064,
+      "learning_rate": 2.2906555244148233e-06,
+      "loss": 0.674,
+      "step": 10809
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3008931851409455,
+      "learning_rate": 2.2897844534780805e-06,
+      "loss": 0.6595,
+      "step": 10810
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2319621579991304,
+      "learning_rate": 2.2889134990053356e-06,
+      "loss": 0.5847,
+      "step": 10811
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.165596873619203,
+      "learning_rate": 2.2880426610340135e-06,
+      "loss": 0.6993,
+      "step": 10812
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3612839594787214,
+      "learning_rate": 2.2871719396015423e-06,
+      "loss": 0.6006,
+      "step": 10813
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4228189902797403,
+      "learning_rate": 2.2863013347453305e-06,
+      "loss": 0.6878,
+      "step": 10814
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4662891459880032,
+      "learning_rate": 2.2854308465027963e-06,
+      "loss": 0.65,
+      "step": 10815
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2206712339107453,
+      "learning_rate": 2.2845604749113457e-06,
+      "loss": 0.6511,
+      "step": 10816
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6083311465651133,
+      "learning_rate": 2.2836902200083762e-06,
+      "loss": 0.6409,
+      "step": 10817
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4651934259772552,
+      "learning_rate": 2.2828200818312886e-06,
+      "loss": 0.6874,
+      "step": 10818
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.340562699584537,
+      "learning_rate": 2.2819500604174733e-06,
+      "loss": 0.6297,
+      "step": 10819
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.498127924335464,
+      "learning_rate": 2.2810801558043177e-06,
+      "loss": 0.5561,
+      "step": 10820
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2806119227289077,
+      "learning_rate": 2.2802103680292037e-06,
+      "loss": 0.5637,
+      "step": 10821
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4285543156003477,
+      "learning_rate": 2.279340697129505e-06,
+      "loss": 0.6008,
+      "step": 10822
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6252676022645616,
+      "learning_rate": 2.2784711431426e-06,
+      "loss": 0.7037,
+      "step": 10823
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4110987650575526,
+      "learning_rate": 2.277601706105847e-06,
+      "loss": 0.6187,
+      "step": 10824
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3677177039424648,
+      "learning_rate": 2.2767323860566137e-06,
+      "loss": 0.6106,
+      "step": 10825
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.199931853761699,
+      "learning_rate": 2.2758631830322546e-06,
+      "loss": 0.5853,
+      "step": 10826
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2175808855932309,
+      "learning_rate": 2.274994097070122e-06,
+      "loss": 0.5741,
+      "step": 10827
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4155355615603242,
+      "learning_rate": 2.2741251282075617e-06,
+      "loss": 0.5921,
+      "step": 10828
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0793729854992977,
+      "learning_rate": 2.2732562764819157e-06,
+      "loss": 0.6678,
+      "step": 10829
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4916733788208691,
+      "learning_rate": 2.272387541930518e-06,
+      "loss": 0.7218,
+      "step": 10830
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2619021432162894,
+      "learning_rate": 2.2715189245907066e-06,
+      "loss": 0.5549,
+      "step": 10831
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.575278705212369,
+      "learning_rate": 2.2706504244997997e-06,
+      "loss": 0.697,
+      "step": 10832
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6292189317372898,
+      "learning_rate": 2.269782041695124e-06,
+      "loss": 0.7345,
+      "step": 10833
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5608534343181182,
+      "learning_rate": 2.2689137762139952e-06,
+      "loss": 0.6687,
+      "step": 10834
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.993090411790578,
+      "learning_rate": 2.2680456280937234e-06,
+      "loss": 0.601,
+      "step": 10835
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.477282555546667,
+      "learning_rate": 2.267177597371616e-06,
+      "loss": 0.6989,
+      "step": 10836
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1897104617282346,
+      "learning_rate": 2.266309684084971e-06,
+      "loss": 0.5757,
+      "step": 10837
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6880023316652462,
+      "learning_rate": 2.265441888271091e-06,
+      "loss": 0.6337,
+      "step": 10838
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2624570284543915,
+      "learning_rate": 2.264574209967262e-06,
+      "loss": 0.5892,
+      "step": 10839
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.136552474853612,
+      "learning_rate": 2.263706649210769e-06,
+      "loss": 0.6952,
+      "step": 10840
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2584742625516434,
+      "learning_rate": 2.2628392060388986e-06,
+      "loss": 0.6926,
+      "step": 10841
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3717540721430639,
+      "learning_rate": 2.2619718804889203e-06,
+      "loss": 0.6499,
+      "step": 10842
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.170722941977824,
+      "learning_rate": 2.26110467259811e-06,
+      "loss": 0.5723,
+      "step": 10843
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3104641077941794,
+      "learning_rate": 2.260237582403732e-06,
+      "loss": 0.5867,
+      "step": 10844
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3018919581780055,
+      "learning_rate": 2.259370609943046e-06,
+      "loss": 0.6421,
+      "step": 10845
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3428397710783455,
+      "learning_rate": 2.2585037552533097e-06,
+      "loss": 0.6377,
+      "step": 10846
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2221622916891197,
+      "learning_rate": 2.2576370183717704e-06,
+      "loss": 0.6227,
+      "step": 10847
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4131141538857535,
+      "learning_rate": 2.2567703993356775e-06,
+      "loss": 0.6245,
+      "step": 10848
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2693838213384114,
+      "learning_rate": 2.2559038981822724e-06,
+      "loss": 0.5372,
+      "step": 10849
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3671298951569484,
+      "learning_rate": 2.255037514948785e-06,
+      "loss": 0.5778,
+      "step": 10850
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4623580588038565,
+      "learning_rate": 2.2541712496724495e-06,
+      "loss": 0.6599,
+      "step": 10851
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.621726962153251,
+      "learning_rate": 2.2533051023904913e-06,
+      "loss": 0.7165,
+      "step": 10852
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3700237224586893,
+      "learning_rate": 2.25243907314013e-06,
+      "loss": 0.606,
+      "step": 10853
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7201143775205356,
+      "learning_rate": 2.2515731619585814e-06,
+      "loss": 0.5555,
+      "step": 10854
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2481543767268137,
+      "learning_rate": 2.250707368883052e-06,
+      "loss": 0.6128,
+      "step": 10855
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.395758927486731,
+      "learning_rate": 2.249841693950755e-06,
+      "loss": 0.6589,
+      "step": 10856
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0498174527796613,
+      "learning_rate": 2.2489761371988826e-06,
+      "loss": 0.6344,
+      "step": 10857
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4495837700260417,
+      "learning_rate": 2.2481106986646314e-06,
+      "loss": 0.5725,
+      "step": 10858
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.236750617632983,
+      "learning_rate": 2.247245378385195e-06,
+      "loss": 0.6199,
+      "step": 10859
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.377272937078904,
+      "learning_rate": 2.246380176397752e-06,
+      "loss": 0.6311,
+      "step": 10860
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6465674379788557,
+      "learning_rate": 2.245515092739488e-06,
+      "loss": 0.6282,
+      "step": 10861
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4097866823771157,
+      "learning_rate": 2.2446501274475747e-06,
+      "loss": 0.6647,
+      "step": 10862
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.109788899120709,
+      "learning_rate": 2.2437852805591815e-06,
+      "loss": 0.6822,
+      "step": 10863
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2312594310370377,
+      "learning_rate": 2.242920552111473e-06,
+      "loss": 0.6412,
+      "step": 10864
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.517333963553632,
+      "learning_rate": 2.2420559421416075e-06,
+      "loss": 0.665,
+      "step": 10865
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4560520127961798,
+      "learning_rate": 2.241191450686743e-06,
+      "loss": 0.652,
+      "step": 10866
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.744977990537173,
+      "learning_rate": 2.2403270777840262e-06,
+      "loss": 0.7886,
+      "step": 10867
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2660083446094774,
+      "learning_rate": 2.239462823470598e-06,
+      "loss": 0.6525,
+      "step": 10868
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4648102910326448,
+      "learning_rate": 2.238598687783603e-06,
+      "loss": 0.542,
+      "step": 10869
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8132054217556643,
+      "learning_rate": 2.237734670760172e-06,
+      "loss": 0.6295,
+      "step": 10870
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5000020842725728,
+      "learning_rate": 2.236870772437433e-06,
+      "loss": 0.5466,
+      "step": 10871
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3018535164493725,
+      "learning_rate": 2.2360069928525118e-06,
+      "loss": 0.5939,
+      "step": 10872
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.369576293258269,
+      "learning_rate": 2.2351433320425237e-06,
+      "loss": 0.5962,
+      "step": 10873
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3541490094967454,
+      "learning_rate": 2.234279790044588e-06,
+      "loss": 0.6835,
+      "step": 10874
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4181765759676754,
+      "learning_rate": 2.2334163668958055e-06,
+      "loss": 0.6052,
+      "step": 10875
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2208553330240854,
+      "learning_rate": 2.2325530626332852e-06,
+      "loss": 0.53,
+      "step": 10876
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3339283155598225,
+      "learning_rate": 2.231689877294123e-06,
+      "loss": 0.7133,
+      "step": 10877
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0355800408567926,
+      "learning_rate": 2.2308268109154126e-06,
+      "loss": 0.6849,
+      "step": 10878
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3253866411234185,
+      "learning_rate": 2.229963863534241e-06,
+      "loss": 0.5999,
+      "step": 10879
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5367091992628474,
+      "learning_rate": 2.229101035187692e-06,
+      "loss": 0.7586,
+      "step": 10880
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5474667282590118,
+      "learning_rate": 2.2282383259128416e-06,
+      "loss": 0.6976,
+      "step": 10881
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7776265369491708,
+      "learning_rate": 2.227375735746764e-06,
+      "loss": 0.6133,
+      "step": 10882
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1189128368877705,
+      "learning_rate": 2.2265132647265243e-06,
+      "loss": 0.6955,
+      "step": 10883
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.310805734227794,
+      "learning_rate": 2.225650912889188e-06,
+      "loss": 0.5791,
+      "step": 10884
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3813686948733772,
+      "learning_rate": 2.224788680271811e-06,
+      "loss": 0.6342,
+      "step": 10885
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3169071128902656,
+      "learning_rate": 2.223926566911445e-06,
+      "loss": 0.5247,
+      "step": 10886
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3014911515904015,
+      "learning_rate": 2.223064572845137e-06,
+      "loss": 0.6425,
+      "step": 10887
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3253812526498046,
+      "learning_rate": 2.222202698109929e-06,
+      "loss": 0.6524,
+      "step": 10888
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.575439250201321,
+      "learning_rate": 2.221340942742858e-06,
+      "loss": 0.6356,
+      "step": 10889
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1150197572749525,
+      "learning_rate": 2.2204793067809545e-06,
+      "loss": 0.5854,
+      "step": 10890
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1928982517256983,
+      "learning_rate": 2.219617790261244e-06,
+      "loss": 0.5977,
+      "step": 10891
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.544456768790033,
+      "learning_rate": 2.218756393220753e-06,
+      "loss": 0.6599,
+      "step": 10892
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2371109512467047,
+      "learning_rate": 2.2178951156964895e-06,
+      "loss": 0.7201,
+      "step": 10893
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4302077674318001,
+      "learning_rate": 2.2170339577254714e-06,
+      "loss": 0.5799,
+      "step": 10894
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7110180511028357,
+      "learning_rate": 2.2161729193447024e-06,
+      "loss": 0.7764,
+      "step": 10895
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.370825457501849,
+      "learning_rate": 2.2153120005911825e-06,
+      "loss": 0.6718,
+      "step": 10896
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3560042022828749,
+      "learning_rate": 2.214451201501908e-06,
+      "loss": 0.6829,
+      "step": 10897
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5047413945467418,
+      "learning_rate": 2.213590522113867e-06,
+      "loss": 0.7854,
+      "step": 10898
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2723622986535519,
+      "learning_rate": 2.212729962464051e-06,
+      "loss": 0.5778,
+      "step": 10899
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9314305715379554,
+      "learning_rate": 2.2118695225894343e-06,
+      "loss": 0.5739,
+      "step": 10900
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3696702785684196,
+      "learning_rate": 2.2110092025269915e-06,
+      "loss": 0.6102,
+      "step": 10901
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.193123942090771,
+      "learning_rate": 2.210149002313697e-06,
+      "loss": 0.6318,
+      "step": 10902
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3275980665443134,
+      "learning_rate": 2.2092889219865127e-06,
+      "loss": 0.6151,
+      "step": 10903
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4985420325836718,
+      "learning_rate": 2.208428961582399e-06,
+      "loss": 0.6175,
+      "step": 10904
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.345780296369336,
+      "learning_rate": 2.2075691211383094e-06,
+      "loss": 0.6544,
+      "step": 10905
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5937712815225862,
+      "learning_rate": 2.2067094006911943e-06,
+      "loss": 0.6951,
+      "step": 10906
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.199498626752051,
+      "learning_rate": 2.205849800277997e-06,
+      "loss": 0.5373,
+      "step": 10907
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.546491448805278,
+      "learning_rate": 2.204990319935657e-06,
+      "loss": 0.6254,
+      "step": 10908
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3902832950738278,
+      "learning_rate": 2.2041309597011057e-06,
+      "loss": 0.6083,
+      "step": 10909
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0982017188549564,
+      "learning_rate": 2.2032717196112776e-06,
+      "loss": 0.7517,
+      "step": 10910
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4467004484049002,
+      "learning_rate": 2.2024125997030885e-06,
+      "loss": 0.6157,
+      "step": 10911
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2554065961839016,
+      "learning_rate": 2.2015536000134624e-06,
+      "loss": 0.6153,
+      "step": 10912
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6241570431383037,
+      "learning_rate": 2.2006947205793107e-06,
+      "loss": 0.8234,
+      "step": 10913
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4988628418346805,
+      "learning_rate": 2.1998359614375412e-06,
+      "loss": 0.6901,
+      "step": 10914
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.069729180197872,
+      "learning_rate": 2.198977322625057e-06,
+      "loss": 0.727,
+      "step": 10915
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.300769719449283,
+      "learning_rate": 2.1981188041787534e-06,
+      "loss": 0.6104,
+      "step": 10916
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.232721599873563,
+      "learning_rate": 2.1972604061355285e-06,
+      "loss": 0.6834,
+      "step": 10917
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3727939212309785,
+      "learning_rate": 2.196402128532264e-06,
+      "loss": 0.6303,
+      "step": 10918
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.629698241248217,
+      "learning_rate": 2.1955439714058422e-06,
+      "loss": 0.675,
+      "step": 10919
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3486124636619419,
+      "learning_rate": 2.1946859347931442e-06,
+      "loss": 0.6702,
+      "step": 10920
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2827366051053848,
+      "learning_rate": 2.193828018731039e-06,
+      "loss": 0.6447,
+      "step": 10921
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4759464773690196,
+      "learning_rate": 2.192970223256393e-06,
+      "loss": 0.6748,
+      "step": 10922
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.227124783436434,
+      "learning_rate": 2.1921125484060692e-06,
+      "loss": 0.6025,
+      "step": 10923
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1890830694360999,
+      "learning_rate": 2.191254994216922e-06,
+      "loss": 0.6094,
+      "step": 10924
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.420668349561729,
+      "learning_rate": 2.1903975607258025e-06,
+      "loss": 0.6626,
+      "step": 10925
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4106221900892248,
+      "learning_rate": 2.1895402479695566e-06,
+      "loss": 0.6632,
+      "step": 10926
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.26052072076845,
+      "learning_rate": 2.1886830559850264e-06,
+      "loss": 0.5799,
+      "step": 10927
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.942062716672347,
+      "learning_rate": 2.187825984809049e-06,
+      "loss": 0.6165,
+      "step": 10928
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3041021398775299,
+      "learning_rate": 2.186969034478448e-06,
+      "loss": 0.5418,
+      "step": 10929
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2339050893555896,
+      "learning_rate": 2.186112205030054e-06,
+      "loss": 0.5336,
+      "step": 10930
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1278932242998945,
+      "learning_rate": 2.1852554965006854e-06,
+      "loss": 0.5585,
+      "step": 10931
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4238188142194967,
+      "learning_rate": 2.1843989089271572e-06,
+      "loss": 0.6461,
+      "step": 10932
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2633697206523153,
+      "learning_rate": 2.183542442346279e-06,
+      "loss": 0.6019,
+      "step": 10933
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2760718676562848,
+      "learning_rate": 2.182686096794852e-06,
+      "loss": 0.563,
+      "step": 10934
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1846797954263768,
+      "learning_rate": 2.181829872309682e-06,
+      "loss": 0.737,
+      "step": 10935
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9749476908931551,
+      "learning_rate": 2.180973768927555e-06,
+      "loss": 0.6473,
+      "step": 10936
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5481397266185644,
+      "learning_rate": 2.1801177866852658e-06,
+      "loss": 0.6888,
+      "step": 10937
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5501793775990536,
+      "learning_rate": 2.179261925619595e-06,
+      "loss": 0.6689,
+      "step": 10938
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6265033274896397,
+      "learning_rate": 2.1784061857673217e-06,
+      "loss": 0.6878,
+      "step": 10939
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0714050572819784,
+      "learning_rate": 2.177550567165218e-06,
+      "loss": 0.7303,
+      "step": 10940
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.360007202946974,
+      "learning_rate": 2.176695069850053e-06,
+      "loss": 0.6281,
+      "step": 10941
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.326596301292963,
+      "learning_rate": 2.1758396938585885e-06,
+      "loss": 0.5627,
+      "step": 10942
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3272073696407323,
+      "learning_rate": 2.174984439227582e-06,
+      "loss": 0.6462,
+      "step": 10943
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.914453971128191,
+      "learning_rate": 2.174129305993784e-06,
+      "loss": 0.6938,
+      "step": 10944
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2837195768699132,
+      "learning_rate": 2.173274294193945e-06,
+      "loss": 0.6735,
+      "step": 10945
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1025442103940954,
+      "learning_rate": 2.1724194038648057e-06,
+      "loss": 0.5599,
+      "step": 10946
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5959794953115913,
+      "learning_rate": 2.1715646350431015e-06,
+      "loss": 0.6651,
+      "step": 10947
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2159967993842042,
+      "learning_rate": 2.1707099877655634e-06,
+      "loss": 0.5645,
+      "step": 10948
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1758786724217027,
+      "learning_rate": 2.1698554620689178e-06,
+      "loss": 0.5335,
+      "step": 10949
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2280189872163032,
+      "learning_rate": 2.169001057989889e-06,
+      "loss": 0.6376,
+      "step": 10950
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1458899136641603,
+      "learning_rate": 2.1681467755651882e-06,
+      "loss": 0.5952,
+      "step": 10951
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3242381790364044,
+      "learning_rate": 2.1672926148315247e-06,
+      "loss": 0.6306,
+      "step": 10952
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3684605710070927,
+      "learning_rate": 2.1664385758256106e-06,
+      "loss": 0.6156,
+      "step": 10953
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6105168775894225,
+      "learning_rate": 2.165584658584138e-06,
+      "loss": 0.6802,
+      "step": 10954
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3307087043962096,
+      "learning_rate": 2.1647308631438068e-06,
+      "loss": 0.614,
+      "step": 10955
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.287583217232786,
+      "learning_rate": 2.163877189541305e-06,
+      "loss": 0.6636,
+      "step": 10956
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3554228468574867,
+      "learning_rate": 2.1630236378133168e-06,
+      "loss": 0.6568,
+      "step": 10957
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5394509795740536,
+      "learning_rate": 2.1621702079965214e-06,
+      "loss": 0.6527,
+      "step": 10958
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5475955196631508,
+      "learning_rate": 2.16131690012759e-06,
+      "loss": 0.7549,
+      "step": 10959
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.415767075069152,
+      "learning_rate": 2.1604637142431978e-06,
+      "loss": 0.6522,
+      "step": 10960
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4694387318022804,
+      "learning_rate": 2.1596106503800013e-06,
+      "loss": 0.6837,
+      "step": 10961
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3944155088800954,
+      "learning_rate": 2.1587577085746596e-06,
+      "loss": 0.6806,
+      "step": 10962
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.142980612074145,
+      "learning_rate": 2.157904888863829e-06,
+      "loss": 0.6035,
+      "step": 10963
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2696696610790619,
+      "learning_rate": 2.157052191284154e-06,
+      "loss": 0.641,
+      "step": 10964
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.50314874063214,
+      "learning_rate": 2.1561996158722785e-06,
+      "loss": 0.717,
+      "step": 10965
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4357504466661073,
+      "learning_rate": 2.1553471626648392e-06,
+      "loss": 0.6349,
+      "step": 10966
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4207834683892677,
+      "learning_rate": 2.1544948316984658e-06,
+      "loss": 0.6312,
+      "step": 10967
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.44816142878197,
+      "learning_rate": 2.1536426230097906e-06,
+      "loss": 0.6843,
+      "step": 10968
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2681798182154789,
+      "learning_rate": 2.1527905366354292e-06,
+      "loss": 0.7559,
+      "step": 10969
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5324879013187946,
+      "learning_rate": 2.1519385726119972e-06,
+      "loss": 0.5919,
+      "step": 10970
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2296644397483267,
+      "learning_rate": 2.151086730976112e-06,
+      "loss": 0.67,
+      "step": 10971
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.527459954106091,
+      "learning_rate": 2.1502350117643704e-06,
+      "loss": 0.5831,
+      "step": 10972
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3775000773417732,
+      "learning_rate": 2.1493834150133795e-06,
+      "loss": 0.643,
+      "step": 10973
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5314154175449115,
+      "learning_rate": 2.1485319407597315e-06,
+      "loss": 0.6557,
+      "step": 10974
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2588094550389404,
+      "learning_rate": 2.1476805890400172e-06,
+      "loss": 0.6794,
+      "step": 10975
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5132453452884214,
+      "learning_rate": 2.14682935989082e-06,
+      "loss": 0.7933,
+      "step": 10976
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2784759825853071,
+      "learning_rate": 2.1459782533487176e-06,
+      "loss": 0.5948,
+      "step": 10977
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3462521695755951,
+      "learning_rate": 2.14512726945029e-06,
+      "loss": 0.6843,
+      "step": 10978
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6322827799290867,
+      "learning_rate": 2.1442764082321e-06,
+      "loss": 0.617,
+      "step": 10979
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3780381202876015,
+      "learning_rate": 2.1434256697307103e-06,
+      "loss": 0.5568,
+      "step": 10980
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5867947702023173,
+      "learning_rate": 2.142575053982684e-06,
+      "loss": 0.6596,
+      "step": 10981
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3318151028555487,
+      "learning_rate": 2.1417245610245714e-06,
+      "loss": 0.6552,
+      "step": 10982
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3871636970143424,
+      "learning_rate": 2.14087419089292e-06,
+      "loss": 0.6421,
+      "step": 10983
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4815016925943694,
+      "learning_rate": 2.140023943624272e-06,
+      "loss": 0.7613,
+      "step": 10984
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4711396643081145,
+      "learning_rate": 2.139173819255163e-06,
+      "loss": 0.613,
+      "step": 10985
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1389034409130405,
+      "learning_rate": 2.13832381782213e-06,
+      "loss": 0.5758,
+      "step": 10986
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5622717650086417,
+      "learning_rate": 2.1374739393616923e-06,
+      "loss": 0.6781,
+      "step": 10987
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3819043683677987,
+      "learning_rate": 2.1366241839103764e-06,
+      "loss": 0.6474,
+      "step": 10988
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5893310504108706,
+      "learning_rate": 2.135774551504698e-06,
+      "loss": 0.7036,
+      "step": 10989
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0401391390740182,
+      "learning_rate": 2.1349250421811622e-06,
+      "loss": 0.6402,
+      "step": 10990
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4463631039575005,
+      "learning_rate": 2.1340756559762808e-06,
+      "loss": 0.5877,
+      "step": 10991
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7719554973901035,
+      "learning_rate": 2.1332263929265505e-06,
+      "loss": 0.7282,
+      "step": 10992
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4889673412309097,
+      "learning_rate": 2.132377253068467e-06,
+      "loss": 0.6863,
+      "step": 10993
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9347946953026585,
+      "learning_rate": 2.1315282364385197e-06,
+      "loss": 0.6131,
+      "step": 10994
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.032205554229832,
+      "learning_rate": 2.1306793430731908e-06,
+      "loss": 0.5921,
+      "step": 10995
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4945533110034508,
+      "learning_rate": 2.1298305730089626e-06,
+      "loss": 0.6727,
+      "step": 10996
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.184377721271408,
+      "learning_rate": 2.1289819262823065e-06,
+      "loss": 0.6795,
+      "step": 10997
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.603628190300405,
+      "learning_rate": 2.1281334029296913e-06,
+      "loss": 0.9026,
+      "step": 10998
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4080496352659655,
+      "learning_rate": 2.1272850029875802e-06,
+      "loss": 0.6972,
+      "step": 10999
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3306410354362546,
+      "learning_rate": 2.1264367264924295e-06,
+      "loss": 0.6476,
+      "step": 11000
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5141156776277538,
+      "learning_rate": 2.125588573480693e-06,
+      "loss": 0.6803,
+      "step": 11001
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1554581926020493,
+      "learning_rate": 2.1247405439888173e-06,
+      "loss": 0.6421,
+      "step": 11002
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.669898930769512,
+      "learning_rate": 2.1238926380532423e-06,
+      "loss": 0.5658,
+      "step": 11003
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3576736726990695,
+      "learning_rate": 2.1230448557104087e-06,
+      "loss": 0.6382,
+      "step": 11004
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2475654278063784,
+      "learning_rate": 2.1221971969967424e-06,
+      "loss": 0.6019,
+      "step": 11005
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3233395282048694,
+      "learning_rate": 2.121349661948673e-06,
+      "loss": 0.6305,
+      "step": 11006
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2715821401318244,
+      "learning_rate": 2.12050225060262e-06,
+      "loss": 0.5953,
+      "step": 11007
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4734422497069983,
+      "learning_rate": 2.119654962994998e-06,
+      "loss": 0.7557,
+      "step": 11008
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0947674191609056,
+      "learning_rate": 2.1188077991622174e-06,
+      "loss": 0.722,
+      "step": 11009
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3062267508177319,
+      "learning_rate": 2.1179607591406807e-06,
+      "loss": 0.6069,
+      "step": 11010
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8584288444812218,
+      "learning_rate": 2.117113842966792e-06,
+      "loss": 0.7468,
+      "step": 11011
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9374444261923269,
+      "learning_rate": 2.1162670506769408e-06,
+      "loss": 0.5975,
+      "step": 11012
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2599302850838408,
+      "learning_rate": 2.1154203823075146e-06,
+      "loss": 0.7472,
+      "step": 11013
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2839879826870815,
+      "learning_rate": 2.1145738378949004e-06,
+      "loss": 0.6267,
+      "step": 11014
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3177753376237378,
+      "learning_rate": 2.113727417475475e-06,
+      "loss": 0.5599,
+      "step": 11015
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0701536045803597,
+      "learning_rate": 2.112881121085611e-06,
+      "loss": 0.6088,
+      "step": 11016
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3783230912895474,
+      "learning_rate": 2.112034948761674e-06,
+      "loss": 0.692,
+      "step": 11017
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.292749128606463,
+      "learning_rate": 2.111188900540028e-06,
+      "loss": 0.6354,
+      "step": 11018
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.21556408648744,
+      "learning_rate": 2.110342976457029e-06,
+      "loss": 0.5506,
+      "step": 11019
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4165727263510315,
+      "learning_rate": 2.1094971765490266e-06,
+      "loss": 0.717,
+      "step": 11020
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2396574068906772,
+      "learning_rate": 2.108651500852367e-06,
+      "loss": 0.614,
+      "step": 11021
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3052873288137734,
+      "learning_rate": 2.107805949403396e-06,
+      "loss": 0.6216,
+      "step": 11022
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.584404096336593,
+      "learning_rate": 2.10696052223844e-06,
+      "loss": 0.5248,
+      "step": 11023
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2724854123790816,
+      "learning_rate": 2.1061152193938355e-06,
+      "loss": 0.6032,
+      "step": 11024
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3820982132793647,
+      "learning_rate": 2.1052700409059057e-06,
+      "loss": 0.6807,
+      "step": 11025
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2002581646230546,
+      "learning_rate": 2.1044249868109683e-06,
+      "loss": 0.6131,
+      "step": 11026
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4907828307497293,
+      "learning_rate": 2.1035800571453386e-06,
+      "loss": 0.6278,
+      "step": 11027
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.255229961059989,
+      "learning_rate": 2.1027352519453225e-06,
+      "loss": 0.6459,
+      "step": 11028
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9400011089426938,
+      "learning_rate": 2.1018905712472285e-06,
+      "loss": 0.6458,
+      "step": 11029
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.962853895634733,
+      "learning_rate": 2.1010460150873495e-06,
+      "loss": 0.7356,
+      "step": 11030
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4823500827970495,
+      "learning_rate": 2.1002015835019777e-06,
+      "loss": 0.6944,
+      "step": 11031
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.601332971248508,
+      "learning_rate": 2.0993572765274044e-06,
+      "loss": 0.6421,
+      "step": 11032
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8523933413154459,
+      "learning_rate": 2.098513094199909e-06,
+      "loss": 0.6174,
+      "step": 11033
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3587488516322008,
+      "learning_rate": 2.0976690365557673e-06,
+      "loss": 0.6478,
+      "step": 11034
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3002168162386825,
+      "learning_rate": 2.096825103631251e-06,
+      "loss": 0.5772,
+      "step": 11035
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3634937669571403,
+      "learning_rate": 2.0959812954626256e-06,
+      "loss": 0.5315,
+      "step": 11036
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3498169301523923,
+      "learning_rate": 2.0951376120861523e-06,
+      "loss": 0.6445,
+      "step": 11037
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5267461868280616,
+      "learning_rate": 2.0942940535380823e-06,
+      "loss": 0.6939,
+      "step": 11038
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2979650631397752,
+      "learning_rate": 2.093450619854671e-06,
+      "loss": 0.6786,
+      "step": 11039
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2697508000361595,
+      "learning_rate": 2.0926073110721613e-06,
+      "loss": 0.6974,
+      "step": 11040
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0066167166849973,
+      "learning_rate": 2.0917641272267867e-06,
+      "loss": 0.6733,
+      "step": 11041
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.87585438327892,
+      "learning_rate": 2.0909210683547865e-06,
+      "loss": 0.6,
+      "step": 11042
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3740747641271287,
+      "learning_rate": 2.0900781344923872e-06,
+      "loss": 0.7676,
+      "step": 11043
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.32949189636448,
+      "learning_rate": 2.0892353256758107e-06,
+      "loss": 0.6058,
+      "step": 11044
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0340397219886333,
+      "learning_rate": 2.0883926419412748e-06,
+      "loss": 0.4487,
+      "step": 11045
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.274276079372929,
+      "learning_rate": 2.08755008332499e-06,
+      "loss": 0.5964,
+      "step": 11046
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0535567612814756,
+      "learning_rate": 2.0867076498631685e-06,
+      "loss": 0.6502,
+      "step": 11047
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.997065392455074,
+      "learning_rate": 2.0858653415920034e-06,
+      "loss": 0.6914,
+      "step": 11048
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1572368784650298,
+      "learning_rate": 2.0850231585476965e-06,
+      "loss": 0.5528,
+      "step": 11049
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.908274033994722,
+      "learning_rate": 2.084181100766437e-06,
+      "loss": 0.6813,
+      "step": 11050
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3020903003855921,
+      "learning_rate": 2.083339168284409e-06,
+      "loss": 0.6192,
+      "step": 11051
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6006824203042227,
+      "learning_rate": 2.0824973611377924e-06,
+      "loss": 0.6439,
+      "step": 11052
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3434740268001581,
+      "learning_rate": 2.0816556793627624e-06,
+      "loss": 0.6097,
+      "step": 11053
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4554880703253192,
+      "learning_rate": 2.0808141229954876e-06,
+      "loss": 0.6218,
+      "step": 11054
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3287535242937165,
+      "learning_rate": 2.079972692072131e-06,
+      "loss": 0.6121,
+      "step": 11055
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.206058229366394,
+      "learning_rate": 2.0791313866288485e-06,
+      "loss": 0.6374,
+      "step": 11056
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2736800280702953,
+      "learning_rate": 2.078290206701798e-06,
+      "loss": 0.5853,
+      "step": 11057
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.544660608200405,
+      "learning_rate": 2.0774491523271243e-06,
+      "loss": 0.6361,
+      "step": 11058
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4748122398185974,
+      "learning_rate": 2.0766082235409695e-06,
+      "loss": 0.6811,
+      "step": 11059
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.343644621396862,
+      "learning_rate": 2.0757674203794696e-06,
+      "loss": 0.637,
+      "step": 11060
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4503073020050388,
+      "learning_rate": 2.074926742878757e-06,
+      "loss": 0.7209,
+      "step": 11061
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3742704699364516,
+      "learning_rate": 2.074086191074956e-06,
+      "loss": 0.6689,
+      "step": 11062
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1029166261980783,
+      "learning_rate": 2.0732457650041884e-06,
+      "loss": 0.6387,
+      "step": 11063
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1229162400690054,
+      "learning_rate": 2.072405464702566e-06,
+      "loss": 0.5732,
+      "step": 11064
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4162945302070722,
+      "learning_rate": 2.071565290206205e-06,
+      "loss": 0.6536,
+      "step": 11065
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.244651380764473,
+      "learning_rate": 2.0707252415512015e-06,
+      "loss": 0.6325,
+      "step": 11066
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4915538468249092,
+      "learning_rate": 2.06988531877366e-06,
+      "loss": 0.6698,
+      "step": 11067
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4706006170475923,
+      "learning_rate": 2.069045521909672e-06,
+      "loss": 0.5989,
+      "step": 11068
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1911261385587668,
+      "learning_rate": 2.068205850995326e-06,
+      "loss": 0.5555,
+      "step": 11069
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7281531094746472,
+      "learning_rate": 2.067366306066704e-06,
+      "loss": 0.615,
+      "step": 11070
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9905913083599003,
+      "learning_rate": 2.0665268871598832e-06,
+      "loss": 0.6063,
+      "step": 11071
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3115179757024644,
+      "learning_rate": 2.065687594310935e-06,
+      "loss": 0.626,
+      "step": 11072
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5390743236631732,
+      "learning_rate": 2.064848427555926e-06,
+      "loss": 0.5672,
+      "step": 11073
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4604026054124066,
+      "learning_rate": 2.064009386930915e-06,
+      "loss": 0.5649,
+      "step": 11074
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6159118038921891,
+      "learning_rate": 2.0631704724719613e-06,
+      "loss": 0.6539,
+      "step": 11075
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7280877911799484,
+      "learning_rate": 2.0623316842151132e-06,
+      "loss": 0.5857,
+      "step": 11076
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3158252341135124,
+      "learning_rate": 2.061493022196415e-06,
+      "loss": 0.5597,
+      "step": 11077
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3200214092283906,
+      "learning_rate": 2.0606544864519064e-06,
+      "loss": 0.6372,
+      "step": 11078
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.468048505359808,
+      "learning_rate": 2.0598160770176208e-06,
+      "loss": 0.6094,
+      "step": 11079
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.532085687000873,
+      "learning_rate": 2.0589777939295855e-06,
+      "loss": 0.5787,
+      "step": 11080
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5334008997023691,
+      "learning_rate": 2.0581396372238254e-06,
+      "loss": 0.7226,
+      "step": 11081
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3552958963862363,
+      "learning_rate": 2.0573016069363554e-06,
+      "loss": 0.7006,
+      "step": 11082
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3886597899321098,
+      "learning_rate": 2.056463703103192e-06,
+      "loss": 0.6991,
+      "step": 11083
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0576514484057862,
+      "learning_rate": 2.0556259257603355e-06,
+      "loss": 0.5833,
+      "step": 11084
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7900678922922995,
+      "learning_rate": 2.0547882749437923e-06,
+      "loss": 0.7307,
+      "step": 11085
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3295720824540165,
+      "learning_rate": 2.053950750689556e-06,
+      "loss": 0.6077,
+      "step": 11086
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0740237958063696,
+      "learning_rate": 2.053113353033618e-06,
+      "loss": 0.6267,
+      "step": 11087
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2906149542586776,
+      "learning_rate": 2.0522760820119615e-06,
+      "loss": 0.5757,
+      "step": 11088
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7541809968703108,
+      "learning_rate": 2.0514389376605646e-06,
+      "loss": 0.5252,
+      "step": 11089
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3783865140821434,
+      "learning_rate": 2.0506019200154076e-06,
+      "loss": 0.7308,
+      "step": 11090
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3363922390843257,
+      "learning_rate": 2.049765029112453e-06,
+      "loss": 0.5693,
+      "step": 11091
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.179324616926461,
+      "learning_rate": 2.0489282649876633e-06,
+      "loss": 0.6418,
+      "step": 11092
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3452183176766073,
+      "learning_rate": 2.0480916276770005e-06,
+      "loss": 0.5961,
+      "step": 11093
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2626645417474625,
+      "learning_rate": 2.0472551172164152e-06,
+      "loss": 0.6229,
+      "step": 11094
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2418442151270708,
+      "learning_rate": 2.046418733641853e-06,
+      "loss": 0.6018,
+      "step": 11095
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9557588913114907,
+      "learning_rate": 2.0455824769892556e-06,
+      "loss": 0.6335,
+      "step": 11096
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.570490247303884,
+      "learning_rate": 2.0447463472945593e-06,
+      "loss": 0.6441,
+      "step": 11097
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3919667936054316,
+      "learning_rate": 2.0439103445936946e-06,
+      "loss": 0.8415,
+      "step": 11098
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4707544410910742,
+      "learning_rate": 2.0430744689225833e-06,
+      "loss": 0.6218,
+      "step": 11099
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3216782932618205,
+      "learning_rate": 2.0422387203171503e-06,
+      "loss": 0.6889,
+      "step": 11100
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1881984220667026,
+      "learning_rate": 2.0414030988133076e-06,
+      "loss": 0.6859,
+      "step": 11101
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3492318714335165,
+      "learning_rate": 2.04056760444696e-06,
+      "loss": 0.5789,
+      "step": 11102
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5003353649481828,
+      "learning_rate": 2.039732237254015e-06,
+      "loss": 0.7467,
+      "step": 11103
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4051498970686924,
+      "learning_rate": 2.0388969972703688e-06,
+      "loss": 0.6303,
+      "step": 11104
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2294880515047952,
+      "learning_rate": 2.0380618845319132e-06,
+      "loss": 0.6129,
+      "step": 11105
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.100712377583581,
+      "learning_rate": 2.0372268990745347e-06,
+      "loss": 0.6518,
+      "step": 11106
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2574180657573906,
+      "learning_rate": 2.036392040934114e-06,
+      "loss": 0.6047,
+      "step": 11107
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.47531123401695,
+      "learning_rate": 2.03555731014653e-06,
+      "loss": 0.6311,
+      "step": 11108
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.516690911808481,
+      "learning_rate": 2.0347227067476478e-06,
+      "loss": 0.714,
+      "step": 11109
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2958200620229912,
+      "learning_rate": 2.033888230773336e-06,
+      "loss": 0.6125,
+      "step": 11110
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1939805899927696,
+      "learning_rate": 2.0330538822594532e-06,
+      "loss": 0.7457,
+      "step": 11111
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2961671157756356,
+      "learning_rate": 2.0322196612418527e-06,
+      "loss": 0.6815,
+      "step": 11112
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2090893035835921,
+      "learning_rate": 2.031385567756383e-06,
+      "loss": 0.5113,
+      "step": 11113
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2801232540179646,
+      "learning_rate": 2.030551601838887e-06,
+      "loss": 0.5966,
+      "step": 11114
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5389852053031412,
+      "learning_rate": 2.029717763525202e-06,
+      "loss": 0.6267,
+      "step": 11115
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.176265644519001,
+      "learning_rate": 2.02888405285116e-06,
+      "loss": 0.6562,
+      "step": 11116
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4865287167680907,
+      "learning_rate": 2.0280504698525856e-06,
+      "loss": 0.5122,
+      "step": 11117
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3028567174392265,
+      "learning_rate": 2.0272170145653025e-06,
+      "loss": 0.641,
+      "step": 11118
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0820900028340315,
+      "learning_rate": 2.0263836870251277e-06,
+      "loss": 0.622,
+      "step": 11119
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.243385789744784,
+      "learning_rate": 2.025550487267864e-06,
+      "loss": 0.5239,
+      "step": 11120
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5398935415775266,
+      "learning_rate": 2.0247174153293226e-06,
+      "loss": 0.5927,
+      "step": 11121
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.305711782203539,
+      "learning_rate": 2.0238844712453e-06,
+      "loss": 0.6071,
+      "step": 11122
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7262566041410223,
+      "learning_rate": 2.02305165505159e-06,
+      "loss": 0.5964,
+      "step": 11123
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.235221636489542,
+      "learning_rate": 2.0222189667839805e-06,
+      "loss": 0.5723,
+      "step": 11124
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3946351199508158,
+      "learning_rate": 2.0213864064782516e-06,
+      "loss": 0.5447,
+      "step": 11125
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1684688435348018,
+      "learning_rate": 2.020553974170186e-06,
+      "loss": 0.6831,
+      "step": 11126
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5621835108936977,
+      "learning_rate": 2.019721669895549e-06,
+      "loss": 0.6071,
+      "step": 11127
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5846902665979288,
+      "learning_rate": 2.018889493690111e-06,
+      "loss": 0.5692,
+      "step": 11128
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4529914222296052,
+      "learning_rate": 2.01805744558963e-06,
+      "loss": 0.7283,
+      "step": 11129
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7495218541472197,
+      "learning_rate": 2.0172255256298623e-06,
+      "loss": 0.5326,
+      "step": 11130
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3938476660903794,
+      "learning_rate": 2.016393733846557e-06,
+      "loss": 0.5897,
+      "step": 11131
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3191429106639563,
+      "learning_rate": 2.015562070275457e-06,
+      "loss": 0.5998,
+      "step": 11132
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7571103533653347,
+      "learning_rate": 2.0147305349523e-06,
+      "loss": 0.6444,
+      "step": 11133
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7062579999768899,
+      "learning_rate": 2.013899127912824e-06,
+      "loss": 0.7098,
+      "step": 11134
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4184172100918986,
+      "learning_rate": 2.0130678491927495e-06,
+      "loss": 0.608,
+      "step": 11135
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4974047779343458,
+      "learning_rate": 2.0122366988278035e-06,
+      "loss": 0.6435,
+      "step": 11136
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.359791347740473,
+      "learning_rate": 2.0114056768537005e-06,
+      "loss": 0.6534,
+      "step": 11137
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.277119315016452,
+      "learning_rate": 2.0105747833061516e-06,
+      "loss": 0.6556,
+      "step": 11138
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4734099242989411,
+      "learning_rate": 2.009744018220863e-06,
+      "loss": 0.6524,
+      "step": 11139
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9956524716832025,
+      "learning_rate": 2.0089133816335303e-06,
+      "loss": 0.6804,
+      "step": 11140
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2822160078094964,
+      "learning_rate": 2.0080828735798554e-06,
+      "loss": 0.5857,
+      "step": 11141
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7529744401226597,
+      "learning_rate": 2.0072524940955213e-06,
+      "loss": 0.7003,
+      "step": 11142
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4631684963217455,
+      "learning_rate": 2.00642224321621e-06,
+      "loss": 0.7745,
+      "step": 11143
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.343271917730862,
+      "learning_rate": 2.005592120977606e-06,
+      "loss": 0.6256,
+      "step": 11144
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.389050684866307,
+      "learning_rate": 2.0047621274153746e-06,
+      "loss": 0.6676,
+      "step": 11145
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3885078313839971,
+      "learning_rate": 2.0039322625651865e-06,
+      "loss": 0.7392,
+      "step": 11146
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.054263854367133,
+      "learning_rate": 2.0031025264627015e-06,
+      "loss": 0.589,
+      "step": 11147
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5369114021365033,
+      "learning_rate": 2.002272919143576e-06,
+      "loss": 0.6625,
+      "step": 11148
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5751429963548118,
+      "learning_rate": 2.0014434406434584e-06,
+      "loss": 0.7608,
+      "step": 11149
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4196449220406657,
+      "learning_rate": 2.0006140909979937e-06,
+      "loss": 0.6753,
+      "step": 11150
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2237583381678763,
+      "learning_rate": 1.9997848702428226e-06,
+      "loss": 0.612,
+      "step": 11151
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3643237713508432,
+      "learning_rate": 1.9989557784135792e-06,
+      "loss": 0.5638,
+      "step": 11152
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.747084028156461,
+      "learning_rate": 1.9981268155458864e-06,
+      "loss": 0.6958,
+      "step": 11153
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6623059346822908,
+      "learning_rate": 1.9972979816753717e-06,
+      "loss": 0.7035,
+      "step": 11154
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.272175401409592,
+      "learning_rate": 1.9964692768376497e-06,
+      "loss": 0.5116,
+      "step": 11155
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3359853593895337,
+      "learning_rate": 1.9956407010683324e-06,
+      "loss": 0.4496,
+      "step": 11156
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.460767682986614,
+      "learning_rate": 1.994812254403025e-06,
+      "loss": 0.656,
+      "step": 11157
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4369669403094847,
+      "learning_rate": 1.9939839368773267e-06,
+      "loss": 0.5577,
+      "step": 11158
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4611418104922984,
+      "learning_rate": 1.9931557485268365e-06,
+      "loss": 0.6504,
+      "step": 11159
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.577281727332985,
+      "learning_rate": 1.992327689387137e-06,
+      "loss": 0.6719,
+      "step": 11160
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2162057472888068,
+      "learning_rate": 1.991499759493817e-06,
+      "loss": 0.5812,
+      "step": 11161
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1914453259866291,
+      "learning_rate": 1.990671958882454e-06,
+      "loss": 0.5895,
+      "step": 11162
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5581598890718753,
+      "learning_rate": 1.9898442875886157e-06,
+      "loss": 0.5719,
+      "step": 11163
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6242456953788769,
+      "learning_rate": 1.9890167456478748e-06,
+      "loss": 0.6118,
+      "step": 11164
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4009183258281184,
+      "learning_rate": 1.9881893330957893e-06,
+      "loss": 0.6622,
+      "step": 11165
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4665242517973296,
+      "learning_rate": 1.9873620499679157e-06,
+      "loss": 0.6033,
+      "step": 11166
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5297833070080928,
+      "learning_rate": 1.986534896299805e-06,
+      "loss": 0.6779,
+      "step": 11167
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2626305777584257,
+      "learning_rate": 1.985707872126999e-06,
+      "loss": 0.5665,
+      "step": 11168
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.274063491883704,
+      "learning_rate": 1.984880977485041e-06,
+      "loss": 0.5887,
+      "step": 11169
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2880746164651629,
+      "learning_rate": 1.9840542124094635e-06,
+      "loss": 0.6433,
+      "step": 11170
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3294358300042206,
+      "learning_rate": 1.9832275769357906e-06,
+      "loss": 0.6394,
+      "step": 11171
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5393433982293547,
+      "learning_rate": 1.982401071099549e-06,
+      "loss": 0.7286,
+      "step": 11172
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7341623530890142,
+      "learning_rate": 1.9815746949362536e-06,
+      "loss": 0.6877,
+      "step": 11173
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1539137638330885,
+      "learning_rate": 1.980748448481416e-06,
+      "loss": 0.6094,
+      "step": 11174
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3167222060180603,
+      "learning_rate": 1.9799223317705426e-06,
+      "loss": 0.5937,
+      "step": 11175
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.0505809497911667,
+      "learning_rate": 1.97909634483913e-06,
+      "loss": 0.6869,
+      "step": 11176
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4315403915833589,
+      "learning_rate": 1.9782704877226796e-06,
+      "loss": 0.6259,
+      "step": 11177
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2395437620349283,
+      "learning_rate": 1.9774447604566715e-06,
+      "loss": 0.6095,
+      "step": 11178
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1522747729885252,
+      "learning_rate": 1.9766191630765964e-06,
+      "loss": 0.5539,
+      "step": 11179
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2072801351268705,
+      "learning_rate": 1.9757936956179308e-06,
+      "loss": 0.5477,
+      "step": 11180
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2861970448779403,
+      "learning_rate": 1.9749683581161417e-06,
+      "loss": 0.6237,
+      "step": 11181
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1041309907429764,
+      "learning_rate": 1.9741431506067012e-06,
+      "loss": 0.5859,
+      "step": 11182
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8120953245683844,
+      "learning_rate": 1.9733180731250683e-06,
+      "loss": 0.6946,
+      "step": 11183
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3877062340221011,
+      "learning_rate": 1.9724931257066988e-06,
+      "loss": 0.6722,
+      "step": 11184
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3502632843765767,
+      "learning_rate": 1.971668308387042e-06,
+      "loss": 0.6163,
+      "step": 11185
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4574432979049818,
+      "learning_rate": 1.970843621201541e-06,
+      "loss": 0.6812,
+      "step": 11186
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.351626178917795,
+      "learning_rate": 1.9700190641856378e-06,
+      "loss": 0.6247,
+      "step": 11187
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3568039555245344,
+      "learning_rate": 1.9691946373747626e-06,
+      "loss": 0.6675,
+      "step": 11188
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2455168053776966,
+      "learning_rate": 1.9683703408043447e-06,
+      "loss": 0.6259,
+      "step": 11189
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3697480814683929,
+      "learning_rate": 1.967546174509805e-06,
+      "loss": 0.6989,
+      "step": 11190
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5793804044903246,
+      "learning_rate": 1.9667221385265595e-06,
+      "loss": 0.7255,
+      "step": 11191
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4123782396346527,
+      "learning_rate": 1.9658982328900195e-06,
+      "loss": 0.6489,
+      "step": 11192
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3914223920711204,
+      "learning_rate": 1.9650744576355894e-06,
+      "loss": 0.6477,
+      "step": 11193
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3156927855275518,
+      "learning_rate": 1.9642508127986676e-06,
+      "loss": 0.6377,
+      "step": 11194
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.474620045676943,
+      "learning_rate": 1.963427298414653e-06,
+      "loss": 0.5284,
+      "step": 11195
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.421126408747985,
+      "learning_rate": 1.9626039145189263e-06,
+      "loss": 0.5518,
+      "step": 11196
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.320471612405293,
+      "learning_rate": 1.9617806611468756e-06,
+      "loss": 0.6442,
+      "step": 11197
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3641561116861607,
+      "learning_rate": 1.960957538333877e-06,
+      "loss": 0.6851,
+      "step": 11198
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2979279720826036,
+      "learning_rate": 1.9601345461153005e-06,
+      "loss": 0.7227,
+      "step": 11199
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0233974973707016,
+      "learning_rate": 1.959311684526513e-06,
+      "loss": 0.5581,
+      "step": 11200
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2462091968846551,
+      "learning_rate": 1.9584889536028728e-06,
+      "loss": 0.5536,
+      "step": 11201
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4842917046073414,
+      "learning_rate": 1.95766635337974e-06,
+      "loss": 0.6386,
+      "step": 11202
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.750618764772498,
+      "learning_rate": 1.956843883892457e-06,
+      "loss": 0.6163,
+      "step": 11203
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4659300685832726,
+      "learning_rate": 1.9560215451763685e-06,
+      "loss": 0.7233,
+      "step": 11204
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.425778228834499,
+      "learning_rate": 1.9551993372668153e-06,
+      "loss": 0.5981,
+      "step": 11205
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2407187708565808,
+      "learning_rate": 1.9543772601991277e-06,
+      "loss": 0.5614,
+      "step": 11206
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.331770112181938,
+      "learning_rate": 1.9535553140086322e-06,
+      "loss": 0.6716,
+      "step": 11207
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3875541777681981,
+      "learning_rate": 1.9527334987306495e-06,
+      "loss": 0.6197,
+      "step": 11208
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6904107718404648,
+      "learning_rate": 1.951911814400495e-06,
+      "loss": 0.6748,
+      "step": 11209
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2351386510548006,
+      "learning_rate": 1.9510902610534782e-06,
+      "loss": 0.5829,
+      "step": 11210
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.187042620241949,
+      "learning_rate": 1.950268838724902e-06,
+      "loss": 0.583,
+      "step": 11211
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0996787190562543,
+      "learning_rate": 1.949447547450067e-06,
+      "loss": 0.6202,
+      "step": 11212
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5830532308711278,
+      "learning_rate": 1.948626387264267e-06,
+      "loss": 0.7113,
+      "step": 11213
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4211282546729602,
+      "learning_rate": 1.9478053582027826e-06,
+      "loss": 0.6627,
+      "step": 11214
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1528059674350202,
+      "learning_rate": 1.9469844603009026e-06,
+      "loss": 0.6123,
+      "step": 11215
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2737015532005254,
+      "learning_rate": 1.946163693593899e-06,
+      "loss": 0.7045,
+      "step": 11216
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.588502447744087,
+      "learning_rate": 1.945343058117043e-06,
+      "loss": 0.7149,
+      "step": 11217
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3091001564248905,
+      "learning_rate": 1.9445225539055982e-06,
+      "loss": 0.561,
+      "step": 11218
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5066894454481108,
+      "learning_rate": 1.9437021809948232e-06,
+      "loss": 0.5741,
+      "step": 11219
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.365488121331566,
+      "learning_rate": 1.9428819394199753e-06,
+      "loss": 0.5579,
+      "step": 11220
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.285369088059283,
+      "learning_rate": 1.9420618292162974e-06,
+      "loss": 0.5477,
+      "step": 11221
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4119809078288519,
+      "learning_rate": 1.9412418504190316e-06,
+      "loss": 0.7386,
+      "step": 11222
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4900689322368983,
+      "learning_rate": 1.9404220030634173e-06,
+      "loss": 0.5982,
+      "step": 11223
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3353243772235024,
+      "learning_rate": 1.9396022871846836e-06,
+      "loss": 0.7332,
+      "step": 11224
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3939886900066436,
+      "learning_rate": 1.9387827028180556e-06,
+      "loss": 0.6168,
+      "step": 11225
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3753228794781949,
+      "learning_rate": 1.937963249998752e-06,
+      "loss": 0.5625,
+      "step": 11226
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6668417777597881,
+      "learning_rate": 1.9371439287619876e-06,
+      "loss": 0.8548,
+      "step": 11227
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.290937044085895,
+      "learning_rate": 1.9363247391429695e-06,
+      "loss": 0.5616,
+      "step": 11228
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2267154557372526,
+      "learning_rate": 1.935505681176899e-06,
+      "loss": 0.5821,
+      "step": 11229
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3521574243821723,
+      "learning_rate": 1.9346867548989767e-06,
+      "loss": 0.5338,
+      "step": 11230
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4647467980132653,
+      "learning_rate": 1.9338679603443922e-06,
+      "loss": 0.6484,
+      "step": 11231
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8895484862920642,
+      "learning_rate": 1.933049297548327e-06,
+      "loss": 0.7415,
+      "step": 11232
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3827380578665198,
+      "learning_rate": 1.932230766545966e-06,
+      "loss": 0.6494,
+      "step": 11233
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2860848247917347,
+      "learning_rate": 1.9314123673724805e-06,
+      "loss": 0.6291,
+      "step": 11234
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2130990844843719,
+      "learning_rate": 1.93059410006304e-06,
+      "loss": 0.5519,
+      "step": 11235
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3281803264916097,
+      "learning_rate": 1.9297759646528076e-06,
+      "loss": 0.6401,
+      "step": 11236
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0787969528052193,
+      "learning_rate": 1.928957961176937e-06,
+      "loss": 0.5668,
+      "step": 11237
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4557422644625027,
+      "learning_rate": 1.9281400896705865e-06,
+      "loss": 0.7054,
+      "step": 11238
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3535101290819556,
+      "learning_rate": 1.9273223501688943e-06,
+      "loss": 0.6725,
+      "step": 11239
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2567804049571456,
+      "learning_rate": 1.9265047427070054e-06,
+      "loss": 0.6503,
+      "step": 11240
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1168502564065996,
+      "learning_rate": 1.925687267320053e-06,
+      "loss": 0.499,
+      "step": 11241
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0922489200434393,
+      "learning_rate": 1.924869924043165e-06,
+      "loss": 0.6272,
+      "step": 11242
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2746830756641343,
+      "learning_rate": 1.9240527129114654e-06,
+      "loss": 0.6169,
+      "step": 11243
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2684712238264542,
+      "learning_rate": 1.9232356339600717e-06,
+      "loss": 0.62,
+      "step": 11244
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1516980846908837,
+      "learning_rate": 1.9224186872240946e-06,
+      "loss": 0.6008,
+      "step": 11245
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0615166413260066,
+      "learning_rate": 1.921601872738641e-06,
+      "loss": 0.5982,
+      "step": 11246
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.240060513090045,
+      "learning_rate": 1.920785190538809e-06,
+      "loss": 0.5969,
+      "step": 11247
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2015746817290747,
+      "learning_rate": 1.9199686406596968e-06,
+      "loss": 0.5309,
+      "step": 11248
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3373743345321534,
+      "learning_rate": 1.919152223136391e-06,
+      "loss": 0.5733,
+      "step": 11249
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3499419580109597,
+      "learning_rate": 1.918335938003977e-06,
+      "loss": 0.6208,
+      "step": 11250
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6191603660789005,
+      "learning_rate": 1.9175197852975304e-06,
+      "loss": 0.7223,
+      "step": 11251
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7001428855757001,
+      "learning_rate": 1.9167037650521215e-06,
+      "loss": 0.6853,
+      "step": 11252
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4519635352230316,
+      "learning_rate": 1.915887877302823e-06,
+      "loss": 0.6674,
+      "step": 11253
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2717394435434728,
+      "learning_rate": 1.9150721220846884e-06,
+      "loss": 0.5834,
+      "step": 11254
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7795472590244976,
+      "learning_rate": 1.914256499432774e-06,
+      "loss": 0.6596,
+      "step": 11255
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.50388638364815,
+      "learning_rate": 1.913441009382133e-06,
+      "loss": 0.6547,
+      "step": 11256
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4181586860615791,
+      "learning_rate": 1.9126256519678034e-06,
+      "loss": 0.6564,
+      "step": 11257
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.775206248791814,
+      "learning_rate": 1.9118104272248266e-06,
+      "loss": 0.5963,
+      "step": 11258
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2766679364385207,
+      "learning_rate": 1.910995335188234e-06,
+      "loss": 0.7193,
+      "step": 11259
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3551436323585406,
+      "learning_rate": 1.9101803758930514e-06,
+      "loss": 0.6567,
+      "step": 11260
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3320587890846176,
+      "learning_rate": 1.909365549374299e-06,
+      "loss": 0.5423,
+      "step": 11261
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8185395098121648,
+      "learning_rate": 1.9085508556669905e-06,
+      "loss": 0.645,
+      "step": 11262
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.070280674115213,
+      "learning_rate": 1.9077362948061404e-06,
+      "loss": 0.5625,
+      "step": 11263
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2554208045067194,
+      "learning_rate": 1.906921866826747e-06,
+      "loss": 0.5966,
+      "step": 11264
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.262049938178544,
+      "learning_rate": 1.9061075717638078e-06,
+      "loss": 0.5824,
+      "step": 11265
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6150144173804384,
+      "learning_rate": 1.9052934096523184e-06,
+      "loss": 0.6926,
+      "step": 11266
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4094964938295167,
+      "learning_rate": 1.9044793805272633e-06,
+      "loss": 0.6712,
+      "step": 11267
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1440832542646557,
+      "learning_rate": 1.9036654844236234e-06,
+      "loss": 0.5461,
+      "step": 11268
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2697310307974494,
+      "learning_rate": 1.9028517213763737e-06,
+      "loss": 0.5988,
+      "step": 11269
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6437845972605367,
+      "learning_rate": 1.902038091420481e-06,
+      "loss": 0.5378,
+      "step": 11270
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7375128380017841,
+      "learning_rate": 1.9012245945909153e-06,
+      "loss": 0.6725,
+      "step": 11271
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5165496140851686,
+      "learning_rate": 1.900411230922628e-06,
+      "loss": 0.671,
+      "step": 11272
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.100520144476482,
+      "learning_rate": 1.8995980004505715e-06,
+      "loss": 0.5042,
+      "step": 11273
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3548205224796652,
+      "learning_rate": 1.8987849032096973e-06,
+      "loss": 0.4904,
+      "step": 11274
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.47029287290065,
+      "learning_rate": 1.8979719392349388e-06,
+      "loss": 0.6215,
+      "step": 11275
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1298982531313873,
+      "learning_rate": 1.8971591085612367e-06,
+      "loss": 0.5576,
+      "step": 11276
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.59713710922103,
+      "learning_rate": 1.8963464112235185e-06,
+      "loss": 0.7078,
+      "step": 11277
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3432194281852095,
+      "learning_rate": 1.895533847256707e-06,
+      "loss": 0.6746,
+      "step": 11278
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1601404239826896,
+      "learning_rate": 1.89472141669572e-06,
+      "loss": 0.578,
+      "step": 11279
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7851995546784338,
+      "learning_rate": 1.8939091195754678e-06,
+      "loss": 0.6496,
+      "step": 11280
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6147219121981546,
+      "learning_rate": 1.8930969559308615e-06,
+      "loss": 0.6525,
+      "step": 11281
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4939065316511901,
+      "learning_rate": 1.8922849257967972e-06,
+      "loss": 0.6408,
+      "step": 11282
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2789152284629675,
+      "learning_rate": 1.891473029208169e-06,
+      "loss": 0.6098,
+      "step": 11283
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5374158030962286,
+      "learning_rate": 1.8906612661998698e-06,
+      "loss": 0.6411,
+      "step": 11284
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4285466064673482,
+      "learning_rate": 1.8898496368067804e-06,
+      "loss": 0.5404,
+      "step": 11285
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.381623489783538,
+      "learning_rate": 1.8890381410637786e-06,
+      "loss": 0.6288,
+      "step": 11286
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.617663605537347,
+      "learning_rate": 1.8882267790057374e-06,
+      "loss": 0.6316,
+      "step": 11287
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5750818617442512,
+      "learning_rate": 1.887415550667519e-06,
+      "loss": 0.5948,
+      "step": 11288
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2518044038152103,
+      "learning_rate": 1.8866044560839902e-06,
+      "loss": 0.6453,
+      "step": 11289
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3585302481539314,
+      "learning_rate": 1.8857934952899987e-06,
+      "loss": 0.6833,
+      "step": 11290
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.583509658916941,
+      "learning_rate": 1.884982668320398e-06,
+      "loss": 0.6182,
+      "step": 11291
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.464841036200851,
+      "learning_rate": 1.8841719752100313e-06,
+      "loss": 0.5863,
+      "step": 11292
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2501309105935954,
+      "learning_rate": 1.8833614159937307e-06,
+      "loss": 0.633,
+      "step": 11293
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5143850154885738,
+      "learning_rate": 1.8825509907063328e-06,
+      "loss": 0.7243,
+      "step": 11294
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.579477995755442,
+      "learning_rate": 1.881740699382662e-06,
+      "loss": 0.5687,
+      "step": 11295
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.288135860722438,
+      "learning_rate": 1.880930542057538e-06,
+      "loss": 0.6141,
+      "step": 11296
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3683330473126898,
+      "learning_rate": 1.880120518765775e-06,
+      "loss": 0.5128,
+      "step": 11297
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4895420851858383,
+      "learning_rate": 1.8793106295421797e-06,
+      "loss": 0.7186,
+      "step": 11298
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.299049967304551,
+      "learning_rate": 1.8785008744215606e-06,
+      "loss": 0.6162,
+      "step": 11299
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6061703399808331,
+      "learning_rate": 1.8776912534387076e-06,
+      "loss": 0.6638,
+      "step": 11300
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3609284571451763,
+      "learning_rate": 1.8768817666284166e-06,
+      "loss": 0.6396,
+      "step": 11301
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5143691305243976,
+      "learning_rate": 1.8760724140254716e-06,
+      "loss": 0.5892,
+      "step": 11302
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1205748302670815,
+      "learning_rate": 1.8752631956646522e-06,
+      "loss": 0.4998,
+      "step": 11303
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3017237590248298,
+      "learning_rate": 1.874454111580733e-06,
+      "loss": 0.6539,
+      "step": 11304
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1942097839987813,
+      "learning_rate": 1.873645161808481e-06,
+      "loss": 0.6595,
+      "step": 11305
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.247698725576874,
+      "learning_rate": 1.8728363463826577e-06,
+      "loss": 0.6997,
+      "step": 11306
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3051170323683854,
+      "learning_rate": 1.872027665338025e-06,
+      "loss": 0.5505,
+      "step": 11307
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5385120928853868,
+      "learning_rate": 1.871219118709326e-06,
+      "loss": 0.5651,
+      "step": 11308
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4432876346409684,
+      "learning_rate": 1.8704107065313116e-06,
+      "loss": 0.6186,
+      "step": 11309
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5041040656370601,
+      "learning_rate": 1.8696024288387189e-06,
+      "loss": 0.6609,
+      "step": 11310
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.241276489412275,
+      "learning_rate": 1.8687942856662815e-06,
+      "loss": 0.7028,
+      "step": 11311
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1451590694457758,
+      "learning_rate": 1.8679862770487273e-06,
+      "loss": 0.6616,
+      "step": 11312
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3753985455134354,
+      "learning_rate": 1.8671784030207762e-06,
+      "loss": 0.569,
+      "step": 11313
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5157230646435542,
+      "learning_rate": 1.8663706636171503e-06,
+      "loss": 0.7183,
+      "step": 11314
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2783667600876327,
+      "learning_rate": 1.8655630588725536e-06,
+      "loss": 0.6158,
+      "step": 11315
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4129588427835673,
+      "learning_rate": 1.8647555888216918e-06,
+      "loss": 0.8297,
+      "step": 11316
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3503015954867097,
+      "learning_rate": 1.8639482534992682e-06,
+      "loss": 0.6574,
+      "step": 11317
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3037877027997304,
+      "learning_rate": 1.863141052939969e-06,
+      "loss": 0.5345,
+      "step": 11318
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3014075989163552,
+      "learning_rate": 1.8623339871784869e-06,
+      "loss": 0.6116,
+      "step": 11319
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2827102807040174,
+      "learning_rate": 1.8615270562495014e-06,
+      "loss": 0.6214,
+      "step": 11320
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5081777585653922,
+      "learning_rate": 1.8607202601876873e-06,
+      "loss": 0.681,
+      "step": 11321
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.857851114263609,
+      "learning_rate": 1.8599135990277156e-06,
+      "loss": 0.6306,
+      "step": 11322
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3736189452302099,
+      "learning_rate": 1.8591070728042503e-06,
+      "loss": 0.6738,
+      "step": 11323
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.338466700705405,
+      "learning_rate": 1.8583006815519473e-06,
+      "loss": 0.6684,
+      "step": 11324
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2323164839293033,
+      "learning_rate": 1.8574944253054638e-06,
+      "loss": 0.6583,
+      "step": 11325
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3616894146998286,
+      "learning_rate": 1.8566883040994411e-06,
+      "loss": 0.6018,
+      "step": 11326
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3109918569946535,
+      "learning_rate": 1.855882317968523e-06,
+      "loss": 0.6026,
+      "step": 11327
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3605497606310857,
+      "learning_rate": 1.8550764669473454e-06,
+      "loss": 0.6207,
+      "step": 11328
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3736193913705623,
+      "learning_rate": 1.8542707510705355e-06,
+      "loss": 0.6318,
+      "step": 11329
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5502294452513783,
+      "learning_rate": 1.853465170372717e-06,
+      "loss": 0.6463,
+      "step": 11330
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5471625831501947,
+      "learning_rate": 1.8526597248885065e-06,
+      "loss": 0.7081,
+      "step": 11331
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2228065601749416,
+      "learning_rate": 1.8518544146525203e-06,
+      "loss": 0.552,
+      "step": 11332
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4242710754852428,
+      "learning_rate": 1.8510492396993595e-06,
+      "loss": 0.5969,
+      "step": 11333
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3675536742200538,
+      "learning_rate": 1.8502442000636246e-06,
+      "loss": 0.7442,
+      "step": 11334
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2576370446478964,
+      "learning_rate": 1.8494392957799146e-06,
+      "loss": 0.6361,
+      "step": 11335
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2933506166841813,
+      "learning_rate": 1.8486345268828115e-06,
+      "loss": 0.6172,
+      "step": 11336
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.280588410401709,
+      "learning_rate": 1.8478298934069028e-06,
+      "loss": 0.6635,
+      "step": 11337
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6029500148606686,
+      "learning_rate": 1.8470253953867635e-06,
+      "loss": 0.6166,
+      "step": 11338
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1754889441825696,
+      "learning_rate": 1.846221032856965e-06,
+      "loss": 0.5733,
+      "step": 11339
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4055322659977925,
+      "learning_rate": 1.8454168058520732e-06,
+      "loss": 0.632,
+      "step": 11340
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.007182890471149,
+      "learning_rate": 1.8446127144066445e-06,
+      "loss": 0.6047,
+      "step": 11341
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9448699990687381,
+      "learning_rate": 1.8438087585552371e-06,
+      "loss": 0.6516,
+      "step": 11342
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.363412683322897,
+      "learning_rate": 1.8430049383323983e-06,
+      "loss": 0.6506,
+      "step": 11343
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6805625517134684,
+      "learning_rate": 1.8422012537726646e-06,
+      "loss": 0.6086,
+      "step": 11344
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.118089825962739,
+      "learning_rate": 1.8413977049105774e-06,
+      "loss": 0.6404,
+      "step": 11345
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1650307048349042,
+      "learning_rate": 1.8405942917806657e-06,
+      "loss": 0.6138,
+      "step": 11346
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.399848149735003,
+      "learning_rate": 1.8397910144174536e-06,
+      "loss": 0.6457,
+      "step": 11347
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3564393896211058,
+      "learning_rate": 1.8389878728554594e-06,
+      "loss": 0.887,
+      "step": 11348
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2103824487048578,
+      "learning_rate": 1.8381848671291953e-06,
+      "loss": 0.5554,
+      "step": 11349
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9805854566456715,
+      "learning_rate": 1.8373819972731721e-06,
+      "loss": 0.6038,
+      "step": 11350
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2734941540938258,
+      "learning_rate": 1.8365792633218844e-06,
+      "loss": 0.5931,
+      "step": 11351
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2184794696488888,
+      "learning_rate": 1.835776665309833e-06,
+      "loss": 0.7321,
+      "step": 11352
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5333560862865185,
+      "learning_rate": 1.8349742032715072e-06,
+      "loss": 0.6629,
+      "step": 11353
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4845067600690043,
+      "learning_rate": 1.8341718772413852e-06,
+      "loss": 0.6169,
+      "step": 11354
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4704335410628255,
+      "learning_rate": 1.8333696872539502e-06,
+      "loss": 0.6257,
+      "step": 11355
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.396629693634587,
+      "learning_rate": 1.8325676333436727e-06,
+      "loss": 0.6948,
+      "step": 11356
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1208818102639126,
+      "learning_rate": 1.8317657155450174e-06,
+      "loss": 0.5013,
+      "step": 11357
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3895184013206885,
+      "learning_rate": 1.8309639338924456e-06,
+      "loss": 0.6787,
+      "step": 11358
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2435281578903514,
+      "learning_rate": 1.8301622884204096e-06,
+      "loss": 0.6174,
+      "step": 11359
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3678238799061697,
+      "learning_rate": 1.8293607791633617e-06,
+      "loss": 0.6239,
+      "step": 11360
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6467102974787906,
+      "learning_rate": 1.8285594061557421e-06,
+      "loss": 0.6901,
+      "step": 11361
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0872946516493636,
+      "learning_rate": 1.8277581694319884e-06,
+      "loss": 0.6624,
+      "step": 11362
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2165048257010007,
+      "learning_rate": 1.8269570690265308e-06,
+      "loss": 0.7508,
+      "step": 11363
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.669540672006909,
+      "learning_rate": 1.8261561049737946e-06,
+      "loss": 0.6919,
+      "step": 11364
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2451060594689896,
+      "learning_rate": 1.8253552773081995e-06,
+      "loss": 0.6479,
+      "step": 11365
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4957593520725294,
+      "learning_rate": 1.8245545860641584e-06,
+      "loss": 0.701,
+      "step": 11366
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.256391350023398,
+      "learning_rate": 1.8237540312760764e-06,
+      "loss": 0.6087,
+      "step": 11367
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.457445534926416,
+      "learning_rate": 1.822953612978362e-06,
+      "loss": 0.6392,
+      "step": 11368
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5354884171421104,
+      "learning_rate": 1.8221533312054024e-06,
+      "loss": 0.6895,
+      "step": 11369
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5472792554059627,
+      "learning_rate": 1.8213531859915939e-06,
+      "loss": 0.719,
+      "step": 11370
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4062905469163318,
+      "learning_rate": 1.8205531773713175e-06,
+      "loss": 0.5785,
+      "step": 11371
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5234755712517214,
+      "learning_rate": 1.8197533053789524e-06,
+      "loss": 0.6241,
+      "step": 11372
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5206678745562896,
+      "learning_rate": 1.8189535700488708e-06,
+      "loss": 0.6133,
+      "step": 11373
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.30529390256528,
+      "learning_rate": 1.818153971415439e-06,
+      "loss": 0.6566,
+      "step": 11374
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7142328039798467,
+      "learning_rate": 1.817354509513017e-06,
+      "loss": 0.6888,
+      "step": 11375
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2794986543032896,
+      "learning_rate": 1.81655518437596e-06,
+      "loss": 0.632,
+      "step": 11376
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3345459471086225,
+      "learning_rate": 1.8157559960386146e-06,
+      "loss": 0.6248,
+      "step": 11377
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4438731546739847,
+      "learning_rate": 1.8149569445353277e-06,
+      "loss": 0.6104,
+      "step": 11378
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3700433319791325,
+      "learning_rate": 1.8141580299004342e-06,
+      "loss": 0.6264,
+      "step": 11379
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0014629766414858,
+      "learning_rate": 1.8133592521682657e-06,
+      "loss": 0.6518,
+      "step": 11380
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6725036200281986,
+      "learning_rate": 1.812560611373147e-06,
+      "loss": 0.7685,
+      "step": 11381
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9901007364053414,
+      "learning_rate": 1.8117621075493979e-06,
+      "loss": 0.567,
+      "step": 11382
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3635433440420712,
+      "learning_rate": 1.8109637407313313e-06,
+      "loss": 0.673,
+      "step": 11383
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3107354931348303,
+      "learning_rate": 1.8101655109532552e-06,
+      "loss": 0.5796,
+      "step": 11384
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4692554830629976,
+      "learning_rate": 1.8093674182494698e-06,
+      "loss": 0.5653,
+      "step": 11385
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.315328098309237,
+      "learning_rate": 1.8085694626542755e-06,
+      "loss": 0.6695,
+      "step": 11386
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3912054660265873,
+      "learning_rate": 1.8077716442019554e-06,
+      "loss": 0.5983,
+      "step": 11387
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5838248620172197,
+      "learning_rate": 1.8069739629267991e-06,
+      "loss": 0.6642,
+      "step": 11388
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1065640138818738,
+      "learning_rate": 1.8061764188630831e-06,
+      "loss": 0.676,
+      "step": 11389
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4636438853136575,
+      "learning_rate": 1.8053790120450793e-06,
+      "loss": 0.5845,
+      "step": 11390
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.470168927847468,
+      "learning_rate": 1.8045817425070544e-06,
+      "loss": 0.697,
+      "step": 11391
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4881762696735392,
+      "learning_rate": 1.8037846102832667e-06,
+      "loss": 0.6922,
+      "step": 11392
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5034826149475289,
+      "learning_rate": 1.8029876154079755e-06,
+      "loss": 0.7056,
+      "step": 11393
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5459762156136612,
+      "learning_rate": 1.8021907579154257e-06,
+      "loss": 0.6013,
+      "step": 11394
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6778097461387294,
+      "learning_rate": 1.8013940378398593e-06,
+      "loss": 0.63,
+      "step": 11395
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4109981327585506,
+      "learning_rate": 1.8005974552155158e-06,
+      "loss": 0.6784,
+      "step": 11396
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2314357301076706,
+      "learning_rate": 1.7998010100766255e-06,
+      "loss": 0.5487,
+      "step": 11397
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5216513289782305,
+      "learning_rate": 1.799004702457413e-06,
+      "loss": 0.6668,
+      "step": 11398
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3704861862094282,
+      "learning_rate": 1.7982085323920973e-06,
+      "loss": 0.592,
+      "step": 11399
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4523137973467972,
+      "learning_rate": 1.7974124999148918e-06,
+      "loss": 0.6222,
+      "step": 11400
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5276400592551604,
+      "learning_rate": 1.7966166050600037e-06,
+      "loss": 0.6617,
+      "step": 11401
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3757693585852413,
+      "learning_rate": 1.795820847861633e-06,
+      "loss": 0.6167,
+      "step": 11402
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.473204128848606,
+      "learning_rate": 1.7950252283539776e-06,
+      "loss": 0.6201,
+      "step": 11403
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2838987457037343,
+      "learning_rate": 1.7942297465712282e-06,
+      "loss": 0.5947,
+      "step": 11404
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3482206496981584,
+      "learning_rate": 1.7934344025475625e-06,
+      "loss": 0.5926,
+      "step": 11405
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4939744546888547,
+      "learning_rate": 1.7926391963171642e-06,
+      "loss": 0.5891,
+      "step": 11406
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.376466744324281,
+      "learning_rate": 1.7918441279142025e-06,
+      "loss": 0.5582,
+      "step": 11407
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9979592432371743,
+      "learning_rate": 1.7910491973728433e-06,
+      "loss": 0.6831,
+      "step": 11408
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1941014212213235,
+      "learning_rate": 1.7902544047272468e-06,
+      "loss": 0.5516,
+      "step": 11409
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4958218051474519,
+      "learning_rate": 1.7894597500115657e-06,
+      "loss": 0.5404,
+      "step": 11410
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3012276851752,
+      "learning_rate": 1.788665233259953e-06,
+      "loss": 0.616,
+      "step": 11411
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3930388090003443,
+      "learning_rate": 1.7878708545065433e-06,
+      "loss": 0.6146,
+      "step": 11412
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5542606251863442,
+      "learning_rate": 1.7870766137854784e-06,
+      "loss": 0.7005,
+      "step": 11413
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3085301117644808,
+      "learning_rate": 1.7862825111308873e-06,
+      "loss": 0.67,
+      "step": 11414
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2345204359058661,
+      "learning_rate": 1.7854885465768939e-06,
+      "loss": 0.5996,
+      "step": 11415
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.110958364937,
+      "learning_rate": 1.7846947201576171e-06,
+      "loss": 0.7121,
+      "step": 11416
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.386780269235366,
+      "learning_rate": 1.7839010319071687e-06,
+      "loss": 0.7255,
+      "step": 11417
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2220306875227802,
+      "learning_rate": 1.7831074818596555e-06,
+      "loss": 0.5704,
+      "step": 11418
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4663758942346858,
+      "learning_rate": 1.7823140700491786e-06,
+      "loss": 0.6564,
+      "step": 11419
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2791231671511107,
+      "learning_rate": 1.7815207965098303e-06,
+      "loss": 0.6246,
+      "step": 11420
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4284224383292001,
+      "learning_rate": 1.780727661275703e-06,
+      "loss": 0.6912,
+      "step": 11421
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3642140031064118,
+      "learning_rate": 1.7799346643808796e-06,
+      "loss": 0.6626,
+      "step": 11422
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3304221582766251,
+      "learning_rate": 1.7791418058594318e-06,
+      "loss": 0.6963,
+      "step": 11423
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.253670020216008,
+      "learning_rate": 1.7783490857454354e-06,
+      "loss": 0.6031,
+      "step": 11424
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.346376730244213,
+      "learning_rate": 1.7775565040729536e-06,
+      "loss": 0.6913,
+      "step": 11425
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1858080773086435,
+      "learning_rate": 1.776764060876046e-06,
+      "loss": 0.6388,
+      "step": 11426
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9394713391610645,
+      "learning_rate": 1.7759717561887652e-06,
+      "loss": 0.6168,
+      "step": 11427
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.500547089458758,
+      "learning_rate": 1.7751795900451562e-06,
+      "loss": 0.55,
+      "step": 11428
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.55054765562916,
+      "learning_rate": 1.7743875624792662e-06,
+      "loss": 0.6395,
+      "step": 11429
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3049031567252685,
+      "learning_rate": 1.7735956735251224e-06,
+      "loss": 0.5829,
+      "step": 11430
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.403123332825039,
+      "learning_rate": 1.7728039232167603e-06,
+      "loss": 0.6536,
+      "step": 11431
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6894529933896811,
+      "learning_rate": 1.7720123115882009e-06,
+      "loss": 0.6368,
+      "step": 11432
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.103514719120459,
+      "learning_rate": 1.771220838673462e-06,
+      "loss": 0.6627,
+      "step": 11433
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.683290721774536,
+      "learning_rate": 1.770429504506554e-06,
+      "loss": 0.6096,
+      "step": 11434
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1918812185148349,
+      "learning_rate": 1.769638309121483e-06,
+      "loss": 0.5995,
+      "step": 11435
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1961877527803433,
+      "learning_rate": 1.7688472525522481e-06,
+      "loss": 0.5989,
+      "step": 11436
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.181294389906578,
+      "learning_rate": 1.7680563348328433e-06,
+      "loss": 0.7124,
+      "step": 11437
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.483187829990839,
+      "learning_rate": 1.7672655559972535e-06,
+      "loss": 0.626,
+      "step": 11438
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4920403799624826,
+      "learning_rate": 1.7664749160794642e-06,
+      "loss": 0.6852,
+      "step": 11439
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.579503507068608,
+      "learning_rate": 1.7656844151134495e-06,
+      "loss": 0.6544,
+      "step": 11440
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1665809879993936,
+      "learning_rate": 1.7648940531331787e-06,
+      "loss": 0.5801,
+      "step": 11441
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.334923258964295,
+      "learning_rate": 1.7641038301726155e-06,
+      "loss": 0.6005,
+      "step": 11442
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3093933652038405,
+      "learning_rate": 1.763313746265715e-06,
+      "loss": 0.6741,
+      "step": 11443
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2608763766666673,
+      "learning_rate": 1.7625238014464358e-06,
+      "loss": 0.5862,
+      "step": 11444
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.251819053163194,
+      "learning_rate": 1.7617339957487167e-06,
+      "loss": 0.6286,
+      "step": 11445
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.435495368346976,
+      "learning_rate": 1.760944329206498e-06,
+      "loss": 0.704,
+      "step": 11446
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3934144744081358,
+      "learning_rate": 1.760154801853719e-06,
+      "loss": 0.5833,
+      "step": 11447
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.557497328653194,
+      "learning_rate": 1.7593654137243005e-06,
+      "loss": 0.6517,
+      "step": 11448
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.393940584085333,
+      "learning_rate": 1.7585761648521688e-06,
+      "loss": 0.6313,
+      "step": 11449
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6783434079904365,
+      "learning_rate": 1.7577870552712378e-06,
+      "loss": 0.6775,
+      "step": 11450
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4558618528410874,
+      "learning_rate": 1.756998085015419e-06,
+      "loss": 0.6321,
+      "step": 11451
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6724015424868783,
+      "learning_rate": 1.7562092541186144e-06,
+      "loss": 0.5949,
+      "step": 11452
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5911324754497151,
+      "learning_rate": 1.7554205626147202e-06,
+      "loss": 0.7075,
+      "step": 11453
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6501692407216308,
+      "learning_rate": 1.7546320105376346e-06,
+      "loss": 0.6659,
+      "step": 11454
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3975831485991181,
+      "learning_rate": 1.7538435979212376e-06,
+      "loss": 0.7687,
+      "step": 11455
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5542122365453384,
+      "learning_rate": 1.7530553247994093e-06,
+      "loss": 0.7182,
+      "step": 11456
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0033947667774505,
+      "learning_rate": 1.7522671912060262e-06,
+      "loss": 0.7781,
+      "step": 11457
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.490337139820416,
+      "learning_rate": 1.751479197174955e-06,
+      "loss": 0.5417,
+      "step": 11458
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0994439881771083,
+      "learning_rate": 1.750691342740058e-06,
+      "loss": 0.6793,
+      "step": 11459
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3244095267035756,
+      "learning_rate": 1.7499036279351905e-06,
+      "loss": 0.581,
+      "step": 11460
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3459358334281706,
+      "learning_rate": 1.7491160527942002e-06,
+      "loss": 0.6614,
+      "step": 11461
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4930955787202345,
+      "learning_rate": 1.7483286173509372e-06,
+      "loss": 0.5808,
+      "step": 11462
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2449013308144619,
+      "learning_rate": 1.7475413216392317e-06,
+      "loss": 0.5978,
+      "step": 11463
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.422676424635376,
+      "learning_rate": 1.746754165692921e-06,
+      "loss": 0.6448,
+      "step": 11464
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2166062379524836,
+      "learning_rate": 1.7459671495458308e-06,
+      "loss": 0.5951,
+      "step": 11465
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3815120696857457,
+      "learning_rate": 1.7451802732317763e-06,
+      "loss": 0.5685,
+      "step": 11466
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.215119174260771,
+      "learning_rate": 1.7443935367845755e-06,
+      "loss": 0.5869,
+      "step": 11467
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5744603847049004,
+      "learning_rate": 1.7436069402380352e-06,
+      "loss": 0.6778,
+      "step": 11468
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.279380270481646,
+      "learning_rate": 1.742820483625957e-06,
+      "loss": 0.5694,
+      "step": 11469
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9697162434126738,
+      "learning_rate": 1.7420341669821373e-06,
+      "loss": 0.5657,
+      "step": 11470
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3829615706064493,
+      "learning_rate": 1.7412479903403628e-06,
+      "loss": 0.679,
+      "step": 11471
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3047270305796033,
+      "learning_rate": 1.7404619537344236e-06,
+      "loss": 0.5547,
+      "step": 11472
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.428377259893303,
+      "learning_rate": 1.7396760571980902e-06,
+      "loss": 0.6814,
+      "step": 11473
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3798571626845784,
+      "learning_rate": 1.7388903007651398e-06,
+      "loss": 0.7077,
+      "step": 11474
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7445042528414116,
+      "learning_rate": 1.7381046844693356e-06,
+      "loss": 0.6522,
+      "step": 11475
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5395992344995342,
+      "learning_rate": 1.737319208344438e-06,
+      "loss": 0.6433,
+      "step": 11476
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4159047758474541,
+      "learning_rate": 1.7365338724242003e-06,
+      "loss": 0.7148,
+      "step": 11477
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8195270480967451,
+      "learning_rate": 1.7357486767423699e-06,
+      "loss": 0.5871,
+      "step": 11478
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3700155496169506,
+      "learning_rate": 1.7349636213326876e-06,
+      "loss": 0.6084,
+      "step": 11479
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.777282948830275,
+      "learning_rate": 1.7341787062288928e-06,
+      "loss": 0.6435,
+      "step": 11480
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2722574435468283,
+      "learning_rate": 1.7333939314647097e-06,
+      "loss": 0.722,
+      "step": 11481
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5107400544247314,
+      "learning_rate": 1.7326092970738656e-06,
+      "loss": 0.5701,
+      "step": 11482
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4618466804499575,
+      "learning_rate": 1.7318248030900792e-06,
+      "loss": 0.5629,
+      "step": 11483
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3188998011844633,
+      "learning_rate": 1.7310404495470557e-06,
+      "loss": 0.5898,
+      "step": 11484
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5128697106762676,
+      "learning_rate": 1.7302562364785063e-06,
+      "loss": 0.4995,
+      "step": 11485
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4871967060890008,
+      "learning_rate": 1.729472163918129e-06,
+      "loss": 0.6345,
+      "step": 11486
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3460868846743739,
+      "learning_rate": 1.7286882318996162e-06,
+      "loss": 0.7276,
+      "step": 11487
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2816050707668007,
+      "learning_rate": 1.7279044404566554e-06,
+      "loss": 0.6732,
+      "step": 11488
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7010997548087303,
+      "learning_rate": 1.727120789622927e-06,
+      "loss": 0.8253,
+      "step": 11489
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4000570291431986,
+      "learning_rate": 1.7263372794321104e-06,
+      "loss": 0.6528,
+      "step": 11490
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3835641191642736,
+      "learning_rate": 1.7255539099178687e-06,
+      "loss": 0.6598,
+      "step": 11491
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2720164011015314,
+      "learning_rate": 1.7247706811138692e-06,
+      "loss": 0.6042,
+      "step": 11492
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3775695668988286,
+      "learning_rate": 1.723987593053768e-06,
+      "loss": 0.7092,
+      "step": 11493
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3046265862000876,
+      "learning_rate": 1.7232046457712164e-06,
+      "loss": 0.714,
+      "step": 11494
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.308076787176584,
+      "learning_rate": 1.7224218392998587e-06,
+      "loss": 0.6143,
+      "step": 11495
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.551480139531086,
+      "learning_rate": 1.7216391736733345e-06,
+      "loss": 0.5745,
+      "step": 11496
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2280163039899874,
+      "learning_rate": 1.7208566489252742e-06,
+      "loss": 0.6664,
+      "step": 11497
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.355832220492823,
+      "learning_rate": 1.7200742650893105e-06,
+      "loss": 0.7132,
+      "step": 11498
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.176545217850387,
+      "learning_rate": 1.7192920221990566e-06,
+      "loss": 0.5564,
+      "step": 11499
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.469549206869493,
+      "learning_rate": 1.7185099202881328e-06,
+      "loss": 0.5766,
+      "step": 11500
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3335086093735238,
+      "learning_rate": 1.7177279593901463e-06,
+      "loss": 0.6581,
+      "step": 11501
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3376441676303303,
+      "learning_rate": 1.7169461395386988e-06,
+      "loss": 0.6077,
+      "step": 11502
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.294130129683221,
+      "learning_rate": 1.7161644607673884e-06,
+      "loss": 0.5945,
+      "step": 11503
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3891453460399061,
+      "learning_rate": 1.7153829231098018e-06,
+      "loss": 0.6423,
+      "step": 11504
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.238029144048595,
+      "learning_rate": 1.7146015265995302e-06,
+      "loss": 0.6036,
+      "step": 11505
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3340968452995086,
+      "learning_rate": 1.7138202712701463e-06,
+      "loss": 0.6697,
+      "step": 11506
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3399603124505637,
+      "learning_rate": 1.7130391571552224e-06,
+      "loss": 0.6861,
+      "step": 11507
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4517308683922951,
+      "learning_rate": 1.712258184288328e-06,
+      "loss": 0.6112,
+      "step": 11508
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5545561382850215,
+      "learning_rate": 1.7114773527030215e-06,
+      "loss": 0.5916,
+      "step": 11509
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.482477077182164,
+      "learning_rate": 1.7106966624328575e-06,
+      "loss": 0.6266,
+      "step": 11510
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5116746964793142,
+      "learning_rate": 1.7099161135113828e-06,
+      "loss": 0.7208,
+      "step": 11511
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.393572246831658,
+      "learning_rate": 1.7091357059721409e-06,
+      "loss": 0.7033,
+      "step": 11512
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2840455829166497,
+      "learning_rate": 1.7083554398486663e-06,
+      "loss": 0.5992,
+      "step": 11513
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3754458858286231,
+      "learning_rate": 1.7075753151744885e-06,
+      "loss": 0.6297,
+      "step": 11514
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3526106005288594,
+      "learning_rate": 1.7067953319831327e-06,
+      "loss": 0.5117,
+      "step": 11515
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5922505781157084,
+      "learning_rate": 1.7060154903081177e-06,
+      "loss": 0.7287,
+      "step": 11516
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0216872923524747,
+      "learning_rate": 1.7052357901829502e-06,
+      "loss": 0.6264,
+      "step": 11517
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.38250518762084,
+      "learning_rate": 1.70445623164114e-06,
+      "loss": 0.5876,
+      "step": 11518
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9555134401507392,
+      "learning_rate": 1.7036768147161853e-06,
+      "loss": 0.6948,
+      "step": 11519
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.450146084547331,
+      "learning_rate": 1.7028975394415791e-06,
+      "loss": 0.7221,
+      "step": 11520
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.563547593635892,
+      "learning_rate": 1.702118405850809e-06,
+      "loss": 0.6425,
+      "step": 11521
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5559633059999247,
+      "learning_rate": 1.7013394139773537e-06,
+      "loss": 0.615,
+      "step": 11522
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3122850791639558,
+      "learning_rate": 1.700560563854694e-06,
+      "loss": 0.6499,
+      "step": 11523
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.561820693427513,
+      "learning_rate": 1.6997818555162915e-06,
+      "loss": 0.752,
+      "step": 11524
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2496785912186263,
+      "learning_rate": 1.6990032889956142e-06,
+      "loss": 0.6584,
+      "step": 11525
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.318635338703173,
+      "learning_rate": 1.6982248643261173e-06,
+      "loss": 0.6611,
+      "step": 11526
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0562101837548286,
+      "learning_rate": 1.6974465815412517e-06,
+      "loss": 0.6079,
+      "step": 11527
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.110177088739829,
+      "learning_rate": 1.6966684406744615e-06,
+      "loss": 0.6754,
+      "step": 11528
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4894852463313715,
+      "learning_rate": 1.6958904417591853e-06,
+      "loss": 0.6558,
+      "step": 11529
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3094211171427335,
+      "learning_rate": 1.6951125848288558e-06,
+      "loss": 0.5698,
+      "step": 11530
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3250626984760305,
+      "learning_rate": 1.6943348699168981e-06,
+      "loss": 0.6939,
+      "step": 11531
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1127267713659605,
+      "learning_rate": 1.6935572970567321e-06,
+      "loss": 0.6427,
+      "step": 11532
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2474204704360032,
+      "learning_rate": 1.6927798662817745e-06,
+      "loss": 0.6415,
+      "step": 11533
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3555330466961577,
+      "learning_rate": 1.6920025776254334e-06,
+      "loss": 0.6393,
+      "step": 11534
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4739984250095877,
+      "learning_rate": 1.6912254311211052e-06,
+      "loss": 0.6736,
+      "step": 11535
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1561151346075595,
+      "learning_rate": 1.6904484268021915e-06,
+      "loss": 0.6005,
+      "step": 11536
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4116996107700959,
+      "learning_rate": 1.689671564702079e-06,
+      "loss": 0.6767,
+      "step": 11537
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3336314523904613,
+      "learning_rate": 1.688894844854152e-06,
+      "loss": 0.7006,
+      "step": 11538
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3001667540683617,
+      "learning_rate": 1.6881182672917879e-06,
+      "loss": 0.6735,
+      "step": 11539
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4882845197274774,
+      "learning_rate": 1.6873418320483565e-06,
+      "loss": 0.6295,
+      "step": 11540
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2337484012837585,
+      "learning_rate": 1.6865655391572266e-06,
+      "loss": 0.6863,
+      "step": 11541
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.273049224169668,
+      "learning_rate": 1.6857893886517523e-06,
+      "loss": 0.5955,
+      "step": 11542
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4018083567692277,
+      "learning_rate": 1.6850133805652907e-06,
+      "loss": 0.6637,
+      "step": 11543
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4602921148633654,
+      "learning_rate": 1.6842375149311868e-06,
+      "loss": 0.7135,
+      "step": 11544
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3924999466278507,
+      "learning_rate": 1.6834617917827817e-06,
+      "loss": 0.7642,
+      "step": 11545
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6241768338592357,
+      "learning_rate": 1.6826862111534098e-06,
+      "loss": 0.6929,
+      "step": 11546
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2389865620253908,
+      "learning_rate": 1.6819107730763996e-06,
+      "loss": 0.6696,
+      "step": 11547
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3050129469896221,
+      "learning_rate": 1.681135477585073e-06,
+      "loss": 0.6689,
+      "step": 11548
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5465060071027807,
+      "learning_rate": 1.680360324712746e-06,
+      "loss": 0.6599,
+      "step": 11549
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1170319972114973,
+      "learning_rate": 1.6795853144927282e-06,
+      "loss": 0.6476,
+      "step": 11550
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3666850163867195,
+      "learning_rate": 1.6788104469583256e-06,
+      "loss": 0.6852,
+      "step": 11551
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4192730599297656,
+      "learning_rate": 1.678035722142835e-06,
+      "loss": 0.6397,
+      "step": 11552
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5625373201291142,
+      "learning_rate": 1.6772611400795475e-06,
+      "loss": 0.5529,
+      "step": 11553
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3042347032056354,
+      "learning_rate": 1.6764867008017493e-06,
+      "loss": 0.657,
+      "step": 11554
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3756602304053713,
+      "learning_rate": 1.6757124043427193e-06,
+      "loss": 0.6101,
+      "step": 11555
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.980092541417549,
+      "learning_rate": 1.674938250735731e-06,
+      "loss": 0.6148,
+      "step": 11556
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1520565788933324,
+      "learning_rate": 1.6741642400140513e-06,
+      "loss": 0.527,
+      "step": 11557
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8778996360981435,
+      "learning_rate": 1.6733903722109391e-06,
+      "loss": 0.5422,
+      "step": 11558
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2036235557133914,
+      "learning_rate": 1.672616647359655e-06,
+      "loss": 0.5752,
+      "step": 11559
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6177847432014223,
+      "learning_rate": 1.6718430654934408e-06,
+      "loss": 0.6685,
+      "step": 11560
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5717224246600416,
+      "learning_rate": 1.671069626645544e-06,
+      "loss": 0.6838,
+      "step": 11561
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.026072285929197,
+      "learning_rate": 1.6702963308491994e-06,
+      "loss": 0.6116,
+      "step": 11562
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3843918300601028,
+      "learning_rate": 1.669523178137637e-06,
+      "loss": 0.6461,
+      "step": 11563
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.214916286331747,
+      "learning_rate": 1.668750168544081e-06,
+      "loss": 0.6372,
+      "step": 11564
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2600793639071226,
+      "learning_rate": 1.6679773021017475e-06,
+      "loss": 0.5468,
+      "step": 11565
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1841560588888227,
+      "learning_rate": 1.6672045788438534e-06,
+      "loss": 0.681,
+      "step": 11566
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6214897084346425,
+      "learning_rate": 1.6664319988036e-06,
+      "loss": 0.7209,
+      "step": 11567
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2056508374616526,
+      "learning_rate": 1.6656595620141857e-06,
+      "loss": 0.6719,
+      "step": 11568
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2396178696913922,
+      "learning_rate": 1.664887268508808e-06,
+      "loss": 0.6064,
+      "step": 11569
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.116813678104552,
+      "learning_rate": 1.6641151183206516e-06,
+      "loss": 0.6495,
+      "step": 11570
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2739437761922365,
+      "learning_rate": 1.663343111482898e-06,
+      "loss": 0.5507,
+      "step": 11571
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5699579548837184,
+      "learning_rate": 1.6625712480287225e-06,
+      "loss": 0.6937,
+      "step": 11572
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9661950274052731,
+      "learning_rate": 1.661799527991293e-06,
+      "loss": 0.6485,
+      "step": 11573
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.521674891989812,
+      "learning_rate": 1.6610279514037725e-06,
+      "loss": 0.6145,
+      "step": 11574
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1293595470898117,
+      "learning_rate": 1.6602565182993152e-06,
+      "loss": 0.5145,
+      "step": 11575
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0660201992538023,
+      "learning_rate": 1.6594852287110758e-06,
+      "loss": 0.702,
+      "step": 11576
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4639055218641803,
+      "learning_rate": 1.6587140826721964e-06,
+      "loss": 0.7004,
+      "step": 11577
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1916000494185004,
+      "learning_rate": 1.657943080215812e-06,
+      "loss": 0.5665,
+      "step": 11578
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.248943700226712,
+      "learning_rate": 1.657172221375058e-06,
+      "loss": 0.5813,
+      "step": 11579
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2607217007149023,
+      "learning_rate": 1.656401506183058e-06,
+      "loss": 0.6266,
+      "step": 11580
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.465683691925965,
+      "learning_rate": 1.6556309346729322e-06,
+      "loss": 0.7161,
+      "step": 11581
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7033910994333583,
+      "learning_rate": 1.6548605068777935e-06,
+      "loss": 0.6,
+      "step": 11582
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.61365422365509,
+      "learning_rate": 1.6540902228307471e-06,
+      "loss": 0.6585,
+      "step": 11583
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8156241663045836,
+      "learning_rate": 1.6533200825648993e-06,
+      "loss": 0.6226,
+      "step": 11584
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2052283004729687,
+      "learning_rate": 1.6525500861133386e-06,
+      "loss": 0.5891,
+      "step": 11585
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4382745361569038,
+      "learning_rate": 1.6517802335091549e-06,
+      "loss": 0.6707,
+      "step": 11586
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3050449594929634,
+      "learning_rate": 1.6510105247854325e-06,
+      "loss": 0.6172,
+      "step": 11587
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2744565295422163,
+      "learning_rate": 1.6502409599752468e-06,
+      "loss": 0.5528,
+      "step": 11588
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3663683426641158,
+      "learning_rate": 1.6494715391116671e-06,
+      "loss": 0.6602,
+      "step": 11589
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6809813551781498,
+      "learning_rate": 1.6487022622277582e-06,
+      "loss": 0.5061,
+      "step": 11590
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.179371309425075,
+      "learning_rate": 1.6479331293565758e-06,
+      "loss": 0.6763,
+      "step": 11591
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2806966319332167,
+      "learning_rate": 1.6471641405311727e-06,
+      "loss": 0.6048,
+      "step": 11592
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5449809760056705,
+      "learning_rate": 1.6463952957845924e-06,
+      "loss": 0.6605,
+      "step": 11593
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3401492118843992,
+      "learning_rate": 1.6456265951498763e-06,
+      "loss": 0.5717,
+      "step": 11594
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2931785867019214,
+      "learning_rate": 1.6448580386600576e-06,
+      "loss": 0.6337,
+      "step": 11595
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2462182010785916,
+      "learning_rate": 1.6440896263481582e-06,
+      "loss": 0.733,
+      "step": 11596
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.317307859026465,
+      "learning_rate": 1.6433213582472036e-06,
+      "loss": 0.6655,
+      "step": 11597
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.654651999749323,
+      "learning_rate": 1.6425532343902051e-06,
+      "loss": 0.5346,
+      "step": 11598
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9202666814673965,
+      "learning_rate": 1.641785254810172e-06,
+      "loss": 0.6318,
+      "step": 11599
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3432534236633729,
+      "learning_rate": 1.6410174195401057e-06,
+      "loss": 0.5809,
+      "step": 11600
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2952460803214845,
+      "learning_rate": 1.6402497286129998e-06,
+      "loss": 0.6337,
+      "step": 11601
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1248687315539307,
+      "learning_rate": 1.6394821820618496e-06,
+      "loss": 0.6487,
+      "step": 11602
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.49366986081846,
+      "learning_rate": 1.63871477991963e-06,
+      "loss": 0.6702,
+      "step": 11603
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5009240769998113,
+      "learning_rate": 1.6379475222193248e-06,
+      "loss": 0.6823,
+      "step": 11604
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2772367802057587,
+      "learning_rate": 1.637180408993902e-06,
+      "loss": 0.5943,
+      "step": 11605
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.360723106008216,
+      "learning_rate": 1.636413440276326e-06,
+      "loss": 0.6329,
+      "step": 11606
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6250860855512943,
+      "learning_rate": 1.635646616099556e-06,
+      "loss": 0.7296,
+      "step": 11607
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3232935439552385,
+      "learning_rate": 1.634879936496544e-06,
+      "loss": 0.6751,
+      "step": 11608
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3813581183738675,
+      "learning_rate": 1.6341134015002352e-06,
+      "loss": 0.6372,
+      "step": 11609
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3616941911724623,
+      "learning_rate": 1.63334701114357e-06,
+      "loss": 0.6512,
+      "step": 11610
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4739913030245548,
+      "learning_rate": 1.63258076545948e-06,
+      "loss": 0.7702,
+      "step": 11611
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3459109575671455,
+      "learning_rate": 1.6318146644808952e-06,
+      "loss": 0.6283,
+      "step": 11612
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7055876269316073,
+      "learning_rate": 1.631048708240736e-06,
+      "loss": 0.7128,
+      "step": 11613
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1966403710635267,
+      "learning_rate": 1.6302828967719175e-06,
+      "loss": 0.5949,
+      "step": 11614
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1554114846702814,
+      "learning_rate": 1.6295172301073469e-06,
+      "loss": 0.6157,
+      "step": 11615
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5402522902329907,
+      "learning_rate": 1.6287517082799265e-06,
+      "loss": 0.6299,
+      "step": 11616
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0782931404594351,
+      "learning_rate": 1.6279863313225559e-06,
+      "loss": 0.5014,
+      "step": 11617
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3211859179861114,
+      "learning_rate": 1.6272210992681213e-06,
+      "loss": 0.6186,
+      "step": 11618
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.532381134290127,
+      "learning_rate": 1.626456012149506e-06,
+      "loss": 0.58,
+      "step": 11619
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.225786456805721,
+      "learning_rate": 1.6256910699995921e-06,
+      "loss": 0.4931,
+      "step": 11620
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4087284562881888,
+      "learning_rate": 1.624926272851245e-06,
+      "loss": 0.6685,
+      "step": 11621
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4667357031335921,
+      "learning_rate": 1.6241616207373339e-06,
+      "loss": 0.6175,
+      "step": 11622
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.458019139406689,
+      "learning_rate": 1.6233971136907168e-06,
+      "loss": 0.6642,
+      "step": 11623
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.637297997517545,
+      "learning_rate": 1.6226327517442453e-06,
+      "loss": 0.6812,
+      "step": 11624
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4554953925854865,
+      "learning_rate": 1.6218685349307667e-06,
+      "loss": 0.6168,
+      "step": 11625
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3536939788028837,
+      "learning_rate": 1.6211044632831185e-06,
+      "loss": 0.5941,
+      "step": 11626
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2006505557518448,
+      "learning_rate": 1.620340536834139e-06,
+      "loss": 0.6437,
+      "step": 11627
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5177120237066943,
+      "learning_rate": 1.6195767556166548e-06,
+      "loss": 0.6534,
+      "step": 11628
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4246215752428373,
+      "learning_rate": 1.6188131196634827e-06,
+      "loss": 0.7082,
+      "step": 11629
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3319350996106616,
+      "learning_rate": 1.6180496290074432e-06,
+      "loss": 0.7116,
+      "step": 11630
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.319061892347928,
+      "learning_rate": 1.617286283681343e-06,
+      "loss": 0.5798,
+      "step": 11631
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3227634572200564,
+      "learning_rate": 1.6165230837179846e-06,
+      "loss": 0.6422,
+      "step": 11632
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2050435412030105,
+      "learning_rate": 1.6157600291501646e-06,
+      "loss": 0.5771,
+      "step": 11633
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1630651359410895,
+      "learning_rate": 1.6149971200106723e-06,
+      "loss": 0.7136,
+      "step": 11634
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3735716148545347,
+      "learning_rate": 1.6142343563322955e-06,
+      "loss": 0.5535,
+      "step": 11635
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7564484368997064,
+      "learning_rate": 1.613471738147807e-06,
+      "loss": 0.7242,
+      "step": 11636
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4611463666945337,
+      "learning_rate": 1.6127092654899784e-06,
+      "loss": 0.7541,
+      "step": 11637
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.597669575821586,
+      "learning_rate": 1.6119469383915804e-06,
+      "loss": 0.6229,
+      "step": 11638
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5856058536565691,
+      "learning_rate": 1.6111847568853645e-06,
+      "loss": 0.6808,
+      "step": 11639
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1751447387916052,
+      "learning_rate": 1.6104227210040884e-06,
+      "loss": 0.6601,
+      "step": 11640
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4905800701390024,
+      "learning_rate": 1.6096608307804973e-06,
+      "loss": 0.6559,
+      "step": 11641
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2477137814262913,
+      "learning_rate": 1.608899086247331e-06,
+      "loss": 0.6526,
+      "step": 11642
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3271790298835284,
+      "learning_rate": 1.608137487437323e-06,
+      "loss": 0.6425,
+      "step": 11643
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1089807289207876,
+      "learning_rate": 1.6073760343831996e-06,
+      "loss": 0.6694,
+      "step": 11644
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0812494409522906,
+      "learning_rate": 1.6066147271176857e-06,
+      "loss": 0.5963,
+      "step": 11645
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3661133595509893,
+      "learning_rate": 1.6058535656734958e-06,
+      "loss": 0.6592,
+      "step": 11646
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.325901858404692,
+      "learning_rate": 1.6050925500833342e-06,
+      "loss": 0.6595,
+      "step": 11647
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5184455349665276,
+      "learning_rate": 1.604331680379908e-06,
+      "loss": 0.7016,
+      "step": 11648
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9711448857567614,
+      "learning_rate": 1.603570956595913e-06,
+      "loss": 0.6375,
+      "step": 11649
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8329899167683383,
+      "learning_rate": 1.602810378764037e-06,
+      "loss": 0.7104,
+      "step": 11650
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2463621758495285,
+      "learning_rate": 1.6020499469169659e-06,
+      "loss": 0.563,
+      "step": 11651
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.247107139962002,
+      "learning_rate": 1.601289661087374e-06,
+      "loss": 0.5952,
+      "step": 11652
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.765617155028935,
+      "learning_rate": 1.600529521307938e-06,
+      "loss": 0.6468,
+      "step": 11653
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1417540484943025,
+      "learning_rate": 1.5997695276113168e-06,
+      "loss": 0.5097,
+      "step": 11654
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7165766154312077,
+      "learning_rate": 1.599009680030173e-06,
+      "loss": 0.6491,
+      "step": 11655
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4080867552435625,
+      "learning_rate": 1.598249978597159e-06,
+      "loss": 0.6311,
+      "step": 11656
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3064292424846566,
+      "learning_rate": 1.5974904233449167e-06,
+      "loss": 0.5821,
+      "step": 11657
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6291590304666455,
+      "learning_rate": 1.59673101430609e-06,
+      "loss": 0.645,
+      "step": 11658
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1089226588955325,
+      "learning_rate": 1.595971751513311e-06,
+      "loss": 0.642,
+      "step": 11659
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.602096057270865,
+      "learning_rate": 1.5952126349992076e-06,
+      "loss": 0.611,
+      "step": 11660
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5078564534424783,
+      "learning_rate": 1.5944536647963998e-06,
+      "loss": 0.7553,
+      "step": 11661
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2035354863141132,
+      "learning_rate": 1.5936948409375007e-06,
+      "loss": 0.5405,
+      "step": 11662
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6021315112439174,
+      "learning_rate": 1.5929361634551227e-06,
+      "loss": 0.7026,
+      "step": 11663
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1261073918615574,
+      "learning_rate": 1.5921776323818655e-06,
+      "loss": 0.6812,
+      "step": 11664
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3072350542979791,
+      "learning_rate": 1.5914192477503248e-06,
+      "loss": 0.5811,
+      "step": 11665
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3631819401656533,
+      "learning_rate": 1.5906610095930908e-06,
+      "loss": 0.7124,
+      "step": 11666
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5476167625980177,
+      "learning_rate": 1.589902917942746e-06,
+      "loss": 0.6517,
+      "step": 11667
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2933348407658556,
+      "learning_rate": 1.5891449728318675e-06,
+      "loss": 0.6991,
+      "step": 11668
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1303122231773155,
+      "learning_rate": 1.5883871742930257e-06,
+      "loss": 0.5492,
+      "step": 11669
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1337667531960212,
+      "learning_rate": 1.5876295223587835e-06,
+      "loss": 0.6983,
+      "step": 11670
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.460406406566665,
+      "learning_rate": 1.586872017061704e-06,
+      "loss": 0.6899,
+      "step": 11671
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6935275669852092,
+      "learning_rate": 1.5861146584343318e-06,
+      "loss": 0.588,
+      "step": 11672
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0797103060588087,
+      "learning_rate": 1.585357446509218e-06,
+      "loss": 0.6489,
+      "step": 11673
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3769639168892847,
+      "learning_rate": 1.5846003813188993e-06,
+      "loss": 0.6457,
+      "step": 11674
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3834258201920975,
+      "learning_rate": 1.5838434628959081e-06,
+      "loss": 0.6351,
+      "step": 11675
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2678758854967962,
+      "learning_rate": 1.5830866912727722e-06,
+      "loss": 0.656,
+      "step": 11676
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.336639026464236,
+      "learning_rate": 1.582330066482009e-06,
+      "loss": 0.6853,
+      "step": 11677
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1271477926425197,
+      "learning_rate": 1.5815735885561385e-06,
+      "loss": 0.7294,
+      "step": 11678
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3883297652251672,
+      "learning_rate": 1.5808172575276615e-06,
+      "loss": 0.6437,
+      "step": 11679
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2640926222917643,
+      "learning_rate": 1.5800610734290805e-06,
+      "loss": 0.5305,
+      "step": 11680
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.35117274559013,
+      "learning_rate": 1.5793050362928935e-06,
+      "loss": 0.5614,
+      "step": 11681
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3137388681906044,
+      "learning_rate": 1.5785491461515878e-06,
+      "loss": 0.5933,
+      "step": 11682
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1238557872977013,
+      "learning_rate": 1.5777934030376445e-06,
+      "loss": 0.6799,
+      "step": 11683
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.228557762452577,
+      "learning_rate": 1.5770378069835412e-06,
+      "loss": 0.5481,
+      "step": 11684
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1712851424522315,
+      "learning_rate": 1.5762823580217463e-06,
+      "loss": 0.5307,
+      "step": 11685
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5525882510123583,
+      "learning_rate": 1.5755270561847236e-06,
+      "loss": 0.6634,
+      "step": 11686
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4873817327563987,
+      "learning_rate": 1.5747719015049294e-06,
+      "loss": 0.7264,
+      "step": 11687
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.471043936319395,
+      "learning_rate": 1.5740168940148137e-06,
+      "loss": 0.6144,
+      "step": 11688
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6110718668863677,
+      "learning_rate": 1.5732620337468258e-06,
+      "loss": 0.5986,
+      "step": 11689
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9074992164975257,
+      "learning_rate": 1.5725073207333963e-06,
+      "loss": 0.5455,
+      "step": 11690
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0062359908703409,
+      "learning_rate": 1.5717527550069628e-06,
+      "loss": 0.6983,
+      "step": 11691
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3057208394734987,
+      "learning_rate": 1.5709983365999481e-06,
+      "loss": 0.5945,
+      "step": 11692
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5328598774482909,
+      "learning_rate": 1.570244065544772e-06,
+      "loss": 0.6147,
+      "step": 11693
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1377171859046042,
+      "learning_rate": 1.5694899418738462e-06,
+      "loss": 0.543,
+      "step": 11694
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4191688112169194,
+      "learning_rate": 1.568735965619577e-06,
+      "loss": 0.6579,
+      "step": 11695
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4164679633393689,
+      "learning_rate": 1.5679821368143678e-06,
+      "loss": 0.601,
+      "step": 11696
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4665319947970863,
+      "learning_rate": 1.5672284554906087e-06,
+      "loss": 0.7233,
+      "step": 11697
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2441662629586208,
+      "learning_rate": 1.5664749216806858e-06,
+      "loss": 0.6196,
+      "step": 11698
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5475752017585027,
+      "learning_rate": 1.5657215354169841e-06,
+      "loss": 0.6255,
+      "step": 11699
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.191734284089039,
+      "learning_rate": 1.564968296731877e-06,
+      "loss": 0.664,
+      "step": 11700
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3478210043479801,
+      "learning_rate": 1.564215205657732e-06,
+      "loss": 0.6575,
+      "step": 11701
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0641921083752368,
+      "learning_rate": 1.563462262226912e-06,
+      "loss": 0.6709,
+      "step": 11702
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3911889246941957,
+      "learning_rate": 1.5627094664717719e-06,
+      "loss": 0.6201,
+      "step": 11703
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4312295026784074,
+      "learning_rate": 1.561956818424661e-06,
+      "loss": 0.6852,
+      "step": 11704
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2571119192446476,
+      "learning_rate": 1.561204318117921e-06,
+      "loss": 0.5531,
+      "step": 11705
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5349561921090473,
+      "learning_rate": 1.5604519655838918e-06,
+      "loss": 0.8019,
+      "step": 11706
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4108264935415005,
+      "learning_rate": 1.5596997608549036e-06,
+      "loss": 0.6281,
+      "step": 11707
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6933161993099346,
+      "learning_rate": 1.5589477039632755e-06,
+      "loss": 0.6223,
+      "step": 11708
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8681186418882747,
+      "learning_rate": 1.5581957949413295e-06,
+      "loss": 0.9022,
+      "step": 11709
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3352372156302474,
+      "learning_rate": 1.557444033821376e-06,
+      "loss": 0.6005,
+      "step": 11710
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3246903742193554,
+      "learning_rate": 1.5566924206357187e-06,
+      "loss": 0.6668,
+      "step": 11711
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1971878074351425,
+      "learning_rate": 1.5559409554166572e-06,
+      "loss": 0.6315,
+      "step": 11712
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3691302944705372,
+      "learning_rate": 1.5551896381964814e-06,
+      "loss": 0.6454,
+      "step": 11713
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3310773619408938,
+      "learning_rate": 1.554438469007482e-06,
+      "loss": 0.486,
+      "step": 11714
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3306846317032512,
+      "learning_rate": 1.5536874478819325e-06,
+      "loss": 0.6456,
+      "step": 11715
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2987793339189047,
+      "learning_rate": 1.5529365748521097e-06,
+      "loss": 0.5943,
+      "step": 11716
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.53268001421928,
+      "learning_rate": 1.5521858499502794e-06,
+      "loss": 0.5791,
+      "step": 11717
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3337431341050552,
+      "learning_rate": 1.5514352732087024e-06,
+      "loss": 0.6682,
+      "step": 11718
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5166336187954348,
+      "learning_rate": 1.5506848446596317e-06,
+      "loss": 0.7011,
+      "step": 11719
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2907554746191328,
+      "learning_rate": 1.5499345643353154e-06,
+      "loss": 0.6562,
+      "step": 11720
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4315233092363928,
+      "learning_rate": 1.5491844322679944e-06,
+      "loss": 0.7255,
+      "step": 11721
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2611717754347516,
+      "learning_rate": 1.5484344484899038e-06,
+      "loss": 0.6161,
+      "step": 11722
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2993810312159737,
+      "learning_rate": 1.5476846130332702e-06,
+      "loss": 0.6099,
+      "step": 11723
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4511365276917183,
+      "learning_rate": 1.546934925930319e-06,
+      "loss": 0.6433,
+      "step": 11724
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.663183146494316,
+      "learning_rate": 1.5461853872132648e-06,
+      "loss": 0.7532,
+      "step": 11725
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4596942787748939,
+      "learning_rate": 1.5454359969143161e-06,
+      "loss": 0.6328,
+      "step": 11726
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3518255896409246,
+      "learning_rate": 1.544686755065677e-06,
+      "loss": 0.6636,
+      "step": 11727
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1095634090684412,
+      "learning_rate": 1.5439376616995432e-06,
+      "loss": 0.623,
+      "step": 11728
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4276042868104413,
+      "learning_rate": 1.5431887168481051e-06,
+      "loss": 0.5542,
+      "step": 11729
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3697571485447313,
+      "learning_rate": 1.5424399205435464e-06,
+      "loss": 0.6354,
+      "step": 11730
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3556363963280509,
+      "learning_rate": 1.5416912728180433e-06,
+      "loss": 0.6308,
+      "step": 11731
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0195371234540473,
+      "learning_rate": 1.5409427737037713e-06,
+      "loss": 0.612,
+      "step": 11732
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3377604442017648,
+      "learning_rate": 1.540194423232889e-06,
+      "loss": 0.6523,
+      "step": 11733
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4013139775757986,
+      "learning_rate": 1.5394462214375593e-06,
+      "loss": 0.5609,
+      "step": 11734
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.506861234866339,
+      "learning_rate": 1.5386981683499325e-06,
+      "loss": 0.5805,
+      "step": 11735
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5254262384615038,
+      "learning_rate": 1.5379502640021544e-06,
+      "loss": 0.7262,
+      "step": 11736
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5314004848199272,
+      "learning_rate": 1.537202508426363e-06,
+      "loss": 0.6656,
+      "step": 11737
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4005291487825602,
+      "learning_rate": 1.5364549016546925e-06,
+      "loss": 0.6739,
+      "step": 11738
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3275752255333368,
+      "learning_rate": 1.5357074437192688e-06,
+      "loss": 0.6116,
+      "step": 11739
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2475521035391943,
+      "learning_rate": 1.5349601346522115e-06,
+      "loss": 0.553,
+      "step": 11740
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6173580884624452,
+      "learning_rate": 1.5342129744856327e-06,
+      "loss": 0.6915,
+      "step": 11741
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.340282784525938,
+      "learning_rate": 1.5334659632516424e-06,
+      "loss": 0.5554,
+      "step": 11742
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.500732892534024,
+      "learning_rate": 1.5327191009823405e-06,
+      "loss": 0.7398,
+      "step": 11743
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3658868199231098,
+      "learning_rate": 1.5319723877098202e-06,
+      "loss": 0.4743,
+      "step": 11744
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1406156170445285,
+      "learning_rate": 1.531225823466171e-06,
+      "loss": 0.6289,
+      "step": 11745
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2850148722817512,
+      "learning_rate": 1.5304794082834713e-06,
+      "loss": 0.6149,
+      "step": 11746
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3464827561246122,
+      "learning_rate": 1.5297331421938017e-06,
+      "loss": 0.664,
+      "step": 11747
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3434378852240008,
+      "learning_rate": 1.5289870252292265e-06,
+      "loss": 0.6304,
+      "step": 11748
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3490983624337967,
+      "learning_rate": 1.5282410574218072e-06,
+      "loss": 0.5942,
+      "step": 11749
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2328785997483307,
+      "learning_rate": 1.5274952388036052e-06,
+      "loss": 0.6315,
+      "step": 11750
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.250887741560516,
+      "learning_rate": 1.5267495694066631e-06,
+      "loss": 0.6464,
+      "step": 11751
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.504962425450196,
+      "learning_rate": 1.5260040492630286e-06,
+      "loss": 0.7856,
+      "step": 11752
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.002156006763748,
+      "learning_rate": 1.5252586784047374e-06,
+      "loss": 0.694,
+      "step": 11753
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0058335034683443,
+      "learning_rate": 1.5245134568638197e-06,
+      "loss": 0.6148,
+      "step": 11754
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.377070195185125,
+      "learning_rate": 1.523768384672299e-06,
+      "loss": 0.6482,
+      "step": 11755
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7158986636916231,
+      "learning_rate": 1.5230234618621908e-06,
+      "loss": 0.7071,
+      "step": 11756
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.032010377343387,
+      "learning_rate": 1.5222786884655117e-06,
+      "loss": 0.7196,
+      "step": 11757
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3103977451701643,
+      "learning_rate": 1.5215340645142607e-06,
+      "loss": 0.6035,
+      "step": 11758
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4048902709351658,
+      "learning_rate": 1.5207895900404363e-06,
+      "loss": 0.5318,
+      "step": 11759
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4452077959593588,
+      "learning_rate": 1.520045265076034e-06,
+      "loss": 0.5987,
+      "step": 11760
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2830717863172725,
+      "learning_rate": 1.5193010896530364e-06,
+      "loss": 0.5927,
+      "step": 11761
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6483593582356137,
+      "learning_rate": 1.518557063803423e-06,
+      "loss": 0.7534,
+      "step": 11762
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3276722917825678,
+      "learning_rate": 1.5178131875591662e-06,
+      "loss": 0.5718,
+      "step": 11763
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4183043200814258,
+      "learning_rate": 1.5170694609522306e-06,
+      "loss": 0.7229,
+      "step": 11764
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0153751823658839,
+      "learning_rate": 1.51632588401458e-06,
+      "loss": 0.6072,
+      "step": 11765
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6993910966236137,
+      "learning_rate": 1.5155824567781618e-06,
+      "loss": 0.6957,
+      "step": 11766
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4887354479274528,
+      "learning_rate": 1.5148391792749272e-06,
+      "loss": 0.6665,
+      "step": 11767
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4466803785088098,
+      "learning_rate": 1.5140960515368165e-06,
+      "loss": 0.7451,
+      "step": 11768
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2799851216615774,
+      "learning_rate": 1.5133530735957586e-06,
+      "loss": 0.6368,
+      "step": 11769
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2549331753173323,
+      "learning_rate": 1.512610245483686e-06,
+      "loss": 0.6139,
+      "step": 11770
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.246074934141199,
+      "learning_rate": 1.5118675672325184e-06,
+      "loss": 0.7073,
+      "step": 11771
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2975548881105896,
+      "learning_rate": 1.5111250388741693e-06,
+      "loss": 0.65,
+      "step": 11772
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5238731535178407,
+      "learning_rate": 1.510382660440548e-06,
+      "loss": 0.5965,
+      "step": 11773
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4856312945899495,
+      "learning_rate": 1.5096404319635533e-06,
+      "loss": 0.6596,
+      "step": 11774
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7244420846523472,
+      "learning_rate": 1.5088983534750867e-06,
+      "loss": 0.6309,
+      "step": 11775
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0672472234259935,
+      "learning_rate": 1.5081564250070291e-06,
+      "loss": 0.5356,
+      "step": 11776
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.45737595681345,
+      "learning_rate": 1.5074146465912687e-06,
+      "loss": 0.7416,
+      "step": 11777
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.065005876151099,
+      "learning_rate": 1.506673018259679e-06,
+      "loss": 0.5432,
+      "step": 11778
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9625061998190009,
+      "learning_rate": 1.50593154004413e-06,
+      "loss": 0.5878,
+      "step": 11779
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.151749488658262,
+      "learning_rate": 1.505190211976484e-06,
+      "loss": 0.6325,
+      "step": 11780
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.563050466446178,
+      "learning_rate": 1.5044490340885987e-06,
+      "loss": 0.6582,
+      "step": 11781
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1559951655995746,
+      "learning_rate": 1.5037080064123212e-06,
+      "loss": 0.6282,
+      "step": 11782
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3456355265182216,
+      "learning_rate": 1.5029671289795007e-06,
+      "loss": 0.5719,
+      "step": 11783
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9927936935327318,
+      "learning_rate": 1.502226401821968e-06,
+      "loss": 0.5646,
+      "step": 11784
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3914040153961225,
+      "learning_rate": 1.501485824971558e-06,
+      "loss": 0.7059,
+      "step": 11785
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1337040020043818,
+      "learning_rate": 1.5007453984600956e-06,
+      "loss": 0.5233,
+      "step": 11786
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4268582839428081,
+      "learning_rate": 1.5000051223193928e-06,
+      "loss": 0.7111,
+      "step": 11787
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8867675733277474,
+      "learning_rate": 1.4992649965812673e-06,
+      "loss": 0.58,
+      "step": 11788
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5080493149724625,
+      "learning_rate": 1.498525021277521e-06,
+      "loss": 0.6679,
+      "step": 11789
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5016084720099943,
+      "learning_rate": 1.497785196439953e-06,
+      "loss": 0.6278,
+      "step": 11790
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1449778768108219,
+      "learning_rate": 1.497045522100355e-06,
+      "loss": 0.6919,
+      "step": 11791
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4692524755175225,
+      "learning_rate": 1.4963059982905104e-06,
+      "loss": 0.7188,
+      "step": 11792
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1702727779705036,
+      "learning_rate": 1.4955666250422035e-06,
+      "loss": 0.6664,
+      "step": 11793
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2988673618674884,
+      "learning_rate": 1.4948274023872005e-06,
+      "loss": 0.5375,
+      "step": 11794
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.429971887398272,
+      "learning_rate": 1.4940883303572724e-06,
+      "loss": 0.6488,
+      "step": 11795
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3836283484978027,
+      "learning_rate": 1.493349408984176e-06,
+      "loss": 0.6884,
+      "step": 11796
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3471866376196813,
+      "learning_rate": 1.4926106382996658e-06,
+      "loss": 0.6134,
+      "step": 11797
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4202708029341373,
+      "learning_rate": 1.491872018335488e-06,
+      "loss": 0.7807,
+      "step": 11798
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2871539218638959,
+      "learning_rate": 1.4911335491233818e-06,
+      "loss": 0.6363,
+      "step": 11799
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6145329936890989,
+      "learning_rate": 1.4903952306950808e-06,
+      "loss": 0.6273,
+      "step": 11800
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.084114290935755,
+      "learning_rate": 1.489657063082316e-06,
+      "loss": 0.5888,
+      "step": 11801
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4225063679333412,
+      "learning_rate": 1.4889190463168019e-06,
+      "loss": 0.5946,
+      "step": 11802
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.327332370257213,
+      "learning_rate": 1.4881811804302577e-06,
+      "loss": 0.6802,
+      "step": 11803
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6605398756218162,
+      "learning_rate": 1.487443465454389e-06,
+      "loss": 0.581,
+      "step": 11804
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4286163073371707,
+      "learning_rate": 1.4867059014208979e-06,
+      "loss": 0.6656,
+      "step": 11805
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2610456770403062,
+      "learning_rate": 1.4859684883614788e-06,
+      "loss": 0.5287,
+      "step": 11806
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4217100730015115,
+      "learning_rate": 1.4852312263078183e-06,
+      "loss": 0.6156,
+      "step": 11807
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1059379311870854,
+      "learning_rate": 1.4844941152916026e-06,
+      "loss": 0.6831,
+      "step": 11808
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3582392615417758,
+      "learning_rate": 1.483757155344503e-06,
+      "loss": 0.5787,
+      "step": 11809
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2582197047778862,
+      "learning_rate": 1.4830203464981879e-06,
+      "loss": 0.6217,
+      "step": 11810
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.436820001332946,
+      "learning_rate": 1.4822836887843244e-06,
+      "loss": 0.5962,
+      "step": 11811
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2696504043820112,
+      "learning_rate": 1.4815471822345618e-06,
+      "loss": 0.5836,
+      "step": 11812
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5740130173884141,
+      "learning_rate": 1.4808108268805549e-06,
+      "loss": 0.6154,
+      "step": 11813
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2829488554620658,
+      "learning_rate": 1.4800746227539437e-06,
+      "loss": 0.5649,
+      "step": 11814
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4373148340242416,
+      "learning_rate": 1.479338569886366e-06,
+      "loss": 0.6497,
+      "step": 11815
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1924930041598782,
+      "learning_rate": 1.47860266830945e-06,
+      "loss": 0.6545,
+      "step": 11816
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.191339221868033,
+      "learning_rate": 1.4778669180548195e-06,
+      "loss": 0.6236,
+      "step": 11817
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1656580016096136,
+      "learning_rate": 1.4771313191540927e-06,
+      "loss": 0.5194,
+      "step": 11818
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4991294759414038,
+      "learning_rate": 1.4763958716388798e-06,
+      "loss": 0.7319,
+      "step": 11819
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5131651307357061,
+      "learning_rate": 1.4756605755407816e-06,
+      "loss": 0.7111,
+      "step": 11820
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5141415700755378,
+      "learning_rate": 1.474925430891399e-06,
+      "loss": 0.7859,
+      "step": 11821
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4552421214347695,
+      "learning_rate": 1.4741904377223215e-06,
+      "loss": 0.6881,
+      "step": 11822
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3325112070940386,
+      "learning_rate": 1.473455596065133e-06,
+      "loss": 0.601,
+      "step": 11823
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.420514082241096,
+      "learning_rate": 1.4727209059514114e-06,
+      "loss": 0.655,
+      "step": 11824
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.351636333604007,
+      "learning_rate": 1.4719863674127272e-06,
+      "loss": 0.7123,
+      "step": 11825
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3298095657004625,
+      "learning_rate": 1.4712519804806485e-06,
+      "loss": 0.5895,
+      "step": 11826
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3710972303488504,
+      "learning_rate": 1.4705177451867287e-06,
+      "loss": 0.7129,
+      "step": 11827
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2836605374106893,
+      "learning_rate": 1.4697836615625233e-06,
+      "loss": 0.5953,
+      "step": 11828
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2328579527350987,
+      "learning_rate": 1.4690497296395773e-06,
+      "loss": 0.648,
+      "step": 11829
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4438603557474332,
+      "learning_rate": 1.4683159494494259e-06,
+      "loss": 0.6609,
+      "step": 11830
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3363037259816857,
+      "learning_rate": 1.4675823210236046e-06,
+      "loss": 0.6798,
+      "step": 11831
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3218015295819194,
+      "learning_rate": 1.4668488443936385e-06,
+      "loss": 0.6744,
+      "step": 11832
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3506761038009811,
+      "learning_rate": 1.466115519591046e-06,
+      "loss": 0.6954,
+      "step": 11833
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3118474847345902,
+      "learning_rate": 1.46538234664734e-06,
+      "loss": 0.4851,
+      "step": 11834
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4813295476683872,
+      "learning_rate": 1.4646493255940253e-06,
+      "loss": 0.6884,
+      "step": 11835
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.336652538350826,
+      "learning_rate": 1.4639164564626041e-06,
+      "loss": 0.5654,
+      "step": 11836
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.7878515612451613,
+      "learning_rate": 1.4631837392845694e-06,
+      "loss": 0.742,
+      "step": 11837
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5774733834225085,
+      "learning_rate": 1.4624511740914032e-06,
+      "loss": 0.6459,
+      "step": 11838
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.028792060711483,
+      "learning_rate": 1.4617187609145906e-06,
+      "loss": 0.624,
+      "step": 11839
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3605444114300862,
+      "learning_rate": 1.4609864997856026e-06,
+      "loss": 0.6599,
+      "step": 11840
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2401727789317256,
+      "learning_rate": 1.4602543907359068e-06,
+      "loss": 0.5696,
+      "step": 11841
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.466703791895573,
+      "learning_rate": 1.4595224337969633e-06,
+      "loss": 0.6655,
+      "step": 11842
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1458116853381424,
+      "learning_rate": 1.4587906290002242e-06,
+      "loss": 0.6668,
+      "step": 11843
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5141055191170385,
+      "learning_rate": 1.4580589763771413e-06,
+      "loss": 0.6406,
+      "step": 11844
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2909835562956553,
+      "learning_rate": 1.4573274759591499e-06,
+      "loss": 0.6133,
+      "step": 11845
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.227053557696498,
+      "learning_rate": 1.456596127777688e-06,
+      "loss": 0.6808,
+      "step": 11846
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4979014025551045,
+      "learning_rate": 1.4558649318641837e-06,
+      "loss": 0.6686,
+      "step": 11847
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2499460637677566,
+      "learning_rate": 1.4551338882500538e-06,
+      "loss": 0.6396,
+      "step": 11848
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6542335997542092,
+      "learning_rate": 1.4544029969667167e-06,
+      "loss": 0.8558,
+      "step": 11849
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.36830329569477,
+      "learning_rate": 1.4536722580455792e-06,
+      "loss": 0.6938,
+      "step": 11850
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.447673642157695,
+      "learning_rate": 1.4529416715180434e-06,
+      "loss": 0.6305,
+      "step": 11851
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3352059189187175,
+      "learning_rate": 1.452211237415503e-06,
+      "loss": 0.6121,
+      "step": 11852
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3946382619766742,
+      "learning_rate": 1.451480955769346e-06,
+      "loss": 0.7223,
+      "step": 11853
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6944142240005309,
+      "learning_rate": 1.4507508266109565e-06,
+      "loss": 0.644,
+      "step": 11854
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5032976064753032,
+      "learning_rate": 1.4500208499717088e-06,
+      "loss": 0.6575,
+      "step": 11855
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.233363469252691,
+      "learning_rate": 1.4492910258829717e-06,
+      "loss": 0.5851,
+      "step": 11856
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5338616593800316,
+      "learning_rate": 1.4485613543761068e-06,
+      "loss": 0.6651,
+      "step": 11857
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3615382565903689,
+      "learning_rate": 1.44783183548247e-06,
+      "loss": 0.7283,
+      "step": 11858
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2370035486091655,
+      "learning_rate": 1.4471024692334101e-06,
+      "loss": 0.5366,
+      "step": 11859
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2119599112987658,
+      "learning_rate": 1.44637325566027e-06,
+      "loss": 0.5991,
+      "step": 11860
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3205672129386636,
+      "learning_rate": 1.445644194794384e-06,
+      "loss": 0.6748,
+      "step": 11861
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4363367060930678,
+      "learning_rate": 1.4449152866670858e-06,
+      "loss": 0.7894,
+      "step": 11862
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.063154177441923,
+      "learning_rate": 1.4441865313096926e-06,
+      "loss": 0.7336,
+      "step": 11863
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2116481853808714,
+      "learning_rate": 1.4434579287535244e-06,
+      "loss": 0.6095,
+      "step": 11864
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4446977165806776,
+      "learning_rate": 1.4427294790298902e-06,
+      "loss": 0.625,
+      "step": 11865
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1880540369353698,
+      "learning_rate": 1.4420011821700924e-06,
+      "loss": 0.5899,
+      "step": 11866
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3205475219716016,
+      "learning_rate": 1.4412730382054285e-06,
+      "loss": 0.7066,
+      "step": 11867
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6936900170051736,
+      "learning_rate": 1.4405450471671861e-06,
+      "loss": 0.6608,
+      "step": 11868
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8173093639139886,
+      "learning_rate": 1.439817209086653e-06,
+      "loss": 0.7407,
+      "step": 11869
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0816361432302646,
+      "learning_rate": 1.4390895239951025e-06,
+      "loss": 0.7135,
+      "step": 11870
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.458345316939534,
+      "learning_rate": 1.4383619919238035e-06,
+      "loss": 0.6468,
+      "step": 11871
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5268826199008818,
+      "learning_rate": 1.4376346129040243e-06,
+      "loss": 0.6852,
+      "step": 11872
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1917345431795237,
+      "learning_rate": 1.4369073869670186e-06,
+      "loss": 0.5763,
+      "step": 11873
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.565577078635988,
+      "learning_rate": 1.4361803141440384e-06,
+      "loss": 0.679,
+      "step": 11874
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3106286864779149,
+      "learning_rate": 1.4354533944663274e-06,
+      "loss": 0.5323,
+      "step": 11875
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2516288663459139,
+      "learning_rate": 1.4347266279651224e-06,
+      "loss": 0.641,
+      "step": 11876
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.406007631048316,
+      "learning_rate": 1.4340000146716544e-06,
+      "loss": 0.657,
+      "step": 11877
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5135095258823914,
+      "learning_rate": 1.4332735546171466e-06,
+      "loss": 0.6392,
+      "step": 11878
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.61578698160189,
+      "learning_rate": 1.432547247832819e-06,
+      "loss": 0.74,
+      "step": 11879
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4798684925720194,
+      "learning_rate": 1.431821094349883e-06,
+      "loss": 0.6029,
+      "step": 11880
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.351035378194667,
+      "learning_rate": 1.4310950941995378e-06,
+      "loss": 0.6675,
+      "step": 11881
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2595693054130508,
+      "learning_rate": 1.4303692474129866e-06,
+      "loss": 0.6037,
+      "step": 11882
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4617601164547924,
+      "learning_rate": 1.4296435540214181e-06,
+      "loss": 0.7155,
+      "step": 11883
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7374793301573948,
+      "learning_rate": 1.4289180140560189e-06,
+      "loss": 0.6523,
+      "step": 11884
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5767089856066043,
+      "learning_rate": 1.428192627547965e-06,
+      "loss": 0.6541,
+      "step": 11885
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3042215173028582,
+      "learning_rate": 1.4274673945284278e-06,
+      "loss": 0.6386,
+      "step": 11886
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.644253421816978,
+      "learning_rate": 1.4267423150285758e-06,
+      "loss": 0.6055,
+      "step": 11887
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2136491009997221,
+      "learning_rate": 1.4260173890795631e-06,
+      "loss": 0.5318,
+      "step": 11888
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0778572244581663,
+      "learning_rate": 1.4252926167125413e-06,
+      "loss": 0.6631,
+      "step": 11889
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.061148772904463,
+      "learning_rate": 1.4245679979586591e-06,
+      "loss": 0.4626,
+      "step": 11890
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4143995196304702,
+      "learning_rate": 1.4238435328490525e-06,
+      "loss": 0.7627,
+      "step": 11891
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4961719849069632,
+      "learning_rate": 1.4231192214148538e-06,
+      "loss": 0.5951,
+      "step": 11892
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3818367271546352,
+      "learning_rate": 1.422395063687188e-06,
+      "loss": 0.6608,
+      "step": 11893
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3197724601771288,
+      "learning_rate": 1.421671059697175e-06,
+      "loss": 0.669,
+      "step": 11894
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0411223005408548,
+      "learning_rate": 1.4209472094759252e-06,
+      "loss": 0.5555,
+      "step": 11895
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4081446467845709,
+      "learning_rate": 1.4202235130545432e-06,
+      "loss": 0.5524,
+      "step": 11896
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3162996252526034,
+      "learning_rate": 1.4194999704641315e-06,
+      "loss": 0.6024,
+      "step": 11897
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2861120693949497,
+      "learning_rate": 1.4187765817357812e-06,
+      "loss": 0.671,
+      "step": 11898
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.321100188567007,
+      "learning_rate": 1.418053346900574e-06,
+      "loss": 0.6547,
+      "step": 11899
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5484053806349294,
+      "learning_rate": 1.4173302659895938e-06,
+      "loss": 0.7267,
+      "step": 11900
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3403473302454645,
+      "learning_rate": 1.4166073390339113e-06,
+      "loss": 0.6185,
+      "step": 11901
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3540920480526164,
+      "learning_rate": 1.4158845660645915e-06,
+      "loss": 0.642,
+      "step": 11902
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2949391373941364,
+      "learning_rate": 1.4151619471126943e-06,
+      "loss": 0.6707,
+      "step": 11903
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2853461641143724,
+      "learning_rate": 1.4144394822092712e-06,
+      "loss": 0.6039,
+      "step": 11904
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4757533335683637,
+      "learning_rate": 1.4137171713853714e-06,
+      "loss": 0.5872,
+      "step": 11905
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8484399586877873,
+      "learning_rate": 1.412995014672029e-06,
+      "loss": 0.656,
+      "step": 11906
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1333173307941513,
+      "learning_rate": 1.4122730121002808e-06,
+      "loss": 0.7149,
+      "step": 11907
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4244072384682054,
+      "learning_rate": 1.4115511637011514e-06,
+      "loss": 0.6809,
+      "step": 11908
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5533076639086099,
+      "learning_rate": 1.4108294695056606e-06,
+      "loss": 0.7307,
+      "step": 11909
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0874096728718177,
+      "learning_rate": 1.4101079295448217e-06,
+      "loss": 0.6619,
+      "step": 11910
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.349919335019567,
+      "learning_rate": 1.4093865438496395e-06,
+      "loss": 0.6563,
+      "step": 11911
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5742582748975067,
+      "learning_rate": 1.408665312451114e-06,
+      "loss": 0.6806,
+      "step": 11912
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4731514888281976,
+      "learning_rate": 1.4079442353802392e-06,
+      "loss": 0.6699,
+      "step": 11913
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3247415626111398,
+      "learning_rate": 1.4072233126679985e-06,
+      "loss": 0.6722,
+      "step": 11914
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4404427591553708,
+      "learning_rate": 1.406502544345375e-06,
+      "loss": 0.6839,
+      "step": 11915
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4558119431003154,
+      "learning_rate": 1.40578193044334e-06,
+      "loss": 0.7145,
+      "step": 11916
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.519215750345688,
+      "learning_rate": 1.4050614709928612e-06,
+      "loss": 0.7591,
+      "step": 11917
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4875487119545496,
+      "learning_rate": 1.4043411660248967e-06,
+      "loss": 0.6895,
+      "step": 11918
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.095424152564903,
+      "learning_rate": 1.4036210155703989e-06,
+      "loss": 0.6482,
+      "step": 11919
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5418495250131348,
+      "learning_rate": 1.4029010196603182e-06,
+      "loss": 0.6118,
+      "step": 11920
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4150869147845042,
+      "learning_rate": 1.4021811783255912e-06,
+      "loss": 0.5829,
+      "step": 11921
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3672915535646504,
+      "learning_rate": 1.4014614915971498e-06,
+      "loss": 0.6495,
+      "step": 11922
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1540863134206332,
+      "learning_rate": 1.4007419595059252e-06,
+      "loss": 0.6278,
+      "step": 11923
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4190841576103437,
+      "learning_rate": 1.4000225820828317e-06,
+      "loss": 0.5911,
+      "step": 11924
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3817983510655703,
+      "learning_rate": 1.3993033593587874e-06,
+      "loss": 0.6238,
+      "step": 11925
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.308329817951592,
+      "learning_rate": 1.3985842913646963e-06,
+      "loss": 0.6096,
+      "step": 11926
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.431231533279028,
+      "learning_rate": 1.3978653781314595e-06,
+      "loss": 0.6331,
+      "step": 11927
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4346936988972232,
+      "learning_rate": 1.3971466196899697e-06,
+      "loss": 0.5814,
+      "step": 11928
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4528930808954788,
+      "learning_rate": 1.3964280160711119e-06,
+      "loss": 0.5757,
+      "step": 11929
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4393527804188873,
+      "learning_rate": 1.3957095673057708e-06,
+      "loss": 0.567,
+      "step": 11930
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4475240858491545,
+      "learning_rate": 1.3949912734248156e-06,
+      "loss": 0.6752,
+      "step": 11931
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3279770765858352,
+      "learning_rate": 1.3942731344591127e-06,
+      "loss": 0.6298,
+      "step": 11932
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8743836019498168,
+      "learning_rate": 1.3935551504395246e-06,
+      "loss": 0.6416,
+      "step": 11933
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.209102669268369,
+      "learning_rate": 1.3928373213969038e-06,
+      "loss": 0.6162,
+      "step": 11934
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5051005989982464,
+      "learning_rate": 1.3921196473620975e-06,
+      "loss": 0.6837,
+      "step": 11935
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.38074948826881,
+      "learning_rate": 1.3914021283659445e-06,
+      "loss": 0.5932,
+      "step": 11936
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1581576371149989,
+      "learning_rate": 1.3906847644392773e-06,
+      "loss": 0.655,
+      "step": 11937
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2876177490062262,
+      "learning_rate": 1.3899675556129272e-06,
+      "loss": 0.5815,
+      "step": 11938
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5794832391099594,
+      "learning_rate": 1.38925050191771e-06,
+      "loss": 0.6687,
+      "step": 11939
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2808033325840433,
+      "learning_rate": 1.3885336033844388e-06,
+      "loss": 0.6037,
+      "step": 11940
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1963691558449312,
+      "learning_rate": 1.3878168600439246e-06,
+      "loss": 0.6388,
+      "step": 11941
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5083747487850225,
+      "learning_rate": 1.3871002719269616e-06,
+      "loss": 0.6409,
+      "step": 11942
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.471906070629513,
+      "learning_rate": 1.386383839064348e-06,
+      "loss": 0.6455,
+      "step": 11943
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.043415298921872,
+      "learning_rate": 1.3856675614868687e-06,
+      "loss": 0.7182,
+      "step": 11944
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.323549945788873,
+      "learning_rate": 1.3849514392253038e-06,
+      "loss": 0.575,
+      "step": 11945
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.035794901737774,
+      "learning_rate": 1.3842354723104267e-06,
+      "loss": 0.5383,
+      "step": 11946
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3509695083072613,
+      "learning_rate": 1.3835196607730028e-06,
+      "loss": 0.6462,
+      "step": 11947
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.143824090580593,
+      "learning_rate": 1.3828040046437963e-06,
+      "loss": 0.539,
+      "step": 11948
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9186971691090695,
+      "learning_rate": 1.3820885039535564e-06,
+      "loss": 0.7283,
+      "step": 11949
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.321782951905931,
+      "learning_rate": 1.3813731587330291e-06,
+      "loss": 0.738,
+      "step": 11950
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3948732431712607,
+      "learning_rate": 1.380657969012958e-06,
+      "loss": 0.608,
+      "step": 11951
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4508859627926103,
+      "learning_rate": 1.3799429348240746e-06,
+      "loss": 0.6429,
+      "step": 11952
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.389327273223278,
+      "learning_rate": 1.3792280561971056e-06,
+      "loss": 0.7083,
+      "step": 11953
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0720083996490348,
+      "learning_rate": 1.378513333162771e-06,
+      "loss": 0.5327,
+      "step": 11954
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3002164974808248,
+      "learning_rate": 1.3777987657517822e-06,
+      "loss": 0.5641,
+      "step": 11955
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2386578953964391,
+      "learning_rate": 1.3770843539948508e-06,
+      "loss": 0.6238,
+      "step": 11956
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2620261395467234,
+      "learning_rate": 1.3763700979226701e-06,
+      "loss": 0.5568,
+      "step": 11957
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1614347998737624,
+      "learning_rate": 1.3756559975659385e-06,
+      "loss": 0.6444,
+      "step": 11958
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4413327392994797,
+      "learning_rate": 1.3749420529553414e-06,
+      "loss": 0.6564,
+      "step": 11959
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4735917008534656,
+      "learning_rate": 1.3742282641215549e-06,
+      "loss": 0.7711,
+      "step": 11960
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2551847610781346,
+      "learning_rate": 1.3735146310952568e-06,
+      "loss": 0.6407,
+      "step": 11961
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4399574985067516,
+      "learning_rate": 1.3728011539071112e-06,
+      "loss": 0.702,
+      "step": 11962
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3620678061008475,
+      "learning_rate": 1.3720878325877785e-06,
+      "loss": 0.63,
+      "step": 11963
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3780182827136818,
+      "learning_rate": 1.3713746671679112e-06,
+      "loss": 0.6098,
+      "step": 11964
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3568501394389518,
+      "learning_rate": 1.3706616576781545e-06,
+      "loss": 0.577,
+      "step": 11965
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.310946001122419,
+      "learning_rate": 1.3699488041491526e-06,
+      "loss": 0.5812,
+      "step": 11966
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4211407324853764,
+      "learning_rate": 1.3692361066115318e-06,
+      "loss": 0.6553,
+      "step": 11967
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.000982104086666,
+      "learning_rate": 1.3685235650959228e-06,
+      "loss": 0.6013,
+      "step": 11968
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.0736533174636134,
+      "learning_rate": 1.3678111796329446e-06,
+      "loss": 0.7494,
+      "step": 11969
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.34179302793315,
+      "learning_rate": 1.3670989502532089e-06,
+      "loss": 0.5895,
+      "step": 11970
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.411193775825554,
+      "learning_rate": 1.3663868769873223e-06,
+      "loss": 0.6682,
+      "step": 11971
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0922226919828968,
+      "learning_rate": 1.365674959865884e-06,
+      "loss": 0.7331,
+      "step": 11972
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5487038542915335,
+      "learning_rate": 1.3649631989194856e-06,
+      "loss": 0.6719,
+      "step": 11973
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.337863258568273,
+      "learning_rate": 1.3642515941787171e-06,
+      "loss": 0.6055,
+      "step": 11974
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2562313101401246,
+      "learning_rate": 1.363540145674152e-06,
+      "loss": 0.5585,
+      "step": 11975
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3864482633742632,
+      "learning_rate": 1.3628288534363664e-06,
+      "loss": 0.7148,
+      "step": 11976
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4287259153446994,
+      "learning_rate": 1.362117717495926e-06,
+      "loss": 0.7222,
+      "step": 11977
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4644453544304528,
+      "learning_rate": 1.3614067378833895e-06,
+      "loss": 0.6685,
+      "step": 11978
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1566531643680225,
+      "learning_rate": 1.3606959146293086e-06,
+      "loss": 0.6522,
+      "step": 11979
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2446733021517211,
+      "learning_rate": 1.3599852477642283e-06,
+      "loss": 0.5674,
+      "step": 11980
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2727018103195937,
+      "learning_rate": 1.3592747373186915e-06,
+      "loss": 0.6953,
+      "step": 11981
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.119978273700866,
+      "learning_rate": 1.358564383323226e-06,
+      "loss": 0.6853,
+      "step": 11982
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2070663365055956,
+      "learning_rate": 1.357854185808357e-06,
+      "loss": 0.5582,
+      "step": 11983
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4205703409736024,
+      "learning_rate": 1.3571441448046086e-06,
+      "loss": 0.6043,
+      "step": 11984
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6264205734695718,
+      "learning_rate": 1.3564342603424864e-06,
+      "loss": 0.7045,
+      "step": 11985
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3073200725014311,
+      "learning_rate": 1.3557245324524999e-06,
+      "loss": 0.574,
+      "step": 11986
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3957457611953443,
+      "learning_rate": 1.3550149611651463e-06,
+      "loss": 0.6107,
+      "step": 11987
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2483048740869112,
+      "learning_rate": 1.3543055465109183e-06,
+      "loss": 0.552,
+      "step": 11988
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3525970195178598,
+      "learning_rate": 1.3535962885202997e-06,
+      "loss": 0.6483,
+      "step": 11989
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2618823098295184,
+      "learning_rate": 1.3528871872237687e-06,
+      "loss": 0.571,
+      "step": 11990
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6768887762670348,
+      "learning_rate": 1.3521782426517988e-06,
+      "loss": 0.7226,
+      "step": 11991
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.317801595585731,
+      "learning_rate": 1.3514694548348556e-06,
+      "loss": 0.6579,
+      "step": 11992
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5302615758380271,
+      "learning_rate": 1.3507608238033927e-06,
+      "loss": 0.7053,
+      "step": 11993
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1938781930720916,
+      "learning_rate": 1.350052349587866e-06,
+      "loss": 0.6627,
+      "step": 11994
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4522628833783935,
+      "learning_rate": 1.3493440322187185e-06,
+      "loss": 0.6095,
+      "step": 11995
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2797623507410432,
+      "learning_rate": 1.3486358717263887e-06,
+      "loss": 0.5684,
+      "step": 11996
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2286696162533473,
+      "learning_rate": 1.3479278681413072e-06,
+      "loss": 0.6583,
+      "step": 11997
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.477400601016369,
+      "learning_rate": 1.3472200214938974e-06,
+      "loss": 0.6152,
+      "step": 11998
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4608865705546468,
+      "learning_rate": 1.3465123318145817e-06,
+      "loss": 0.6487,
+      "step": 11999
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4470978410719642,
+      "learning_rate": 1.3458047991337664e-06,
+      "loss": 0.5914,
+      "step": 12000
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.2277032686699036,
+      "learning_rate": 1.3450974234818554e-06,
+      "loss": 0.7473,
+      "step": 12001
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4355940406681176,
+      "learning_rate": 1.34439020488925e-06,
+      "loss": 0.631,
+      "step": 12002
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.37407542523646,
+      "learning_rate": 1.3436831433863384e-06,
+      "loss": 0.6674,
+      "step": 12003
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.400069971327024,
+      "learning_rate": 1.342976239003505e-06,
+      "loss": 0.5591,
+      "step": 12004
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.414311391290379,
+      "learning_rate": 1.3422694917711276e-06,
+      "loss": 0.6031,
+      "step": 12005
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4214068256352372,
+      "learning_rate": 1.3415629017195763e-06,
+      "loss": 0.5866,
+      "step": 12006
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9711891267617678,
+      "learning_rate": 1.3408564688792153e-06,
+      "loss": 0.6459,
+      "step": 12007
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.469640774734688,
+      "learning_rate": 1.340150193280399e-06,
+      "loss": 0.6413,
+      "step": 12008
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.640328188218026,
+      "learning_rate": 1.339444074953482e-06,
+      "loss": 0.6984,
+      "step": 12009
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3986589360650312,
+      "learning_rate": 1.3387381139288064e-06,
+      "loss": 0.5463,
+      "step": 12010
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2843121716764194,
+      "learning_rate": 1.3380323102367055e-06,
+      "loss": 0.6541,
+      "step": 12011
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0956741181480558,
+      "learning_rate": 1.3373266639075134e-06,
+      "loss": 0.5102,
+      "step": 12012
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.54843621739104,
+      "learning_rate": 1.3366211749715519e-06,
+      "loss": 0.6933,
+      "step": 12013
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5513150103603932,
+      "learning_rate": 1.335915843459137e-06,
+      "loss": 0.6198,
+      "step": 12014
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.1480680316178526,
+      "learning_rate": 1.335210669400579e-06,
+      "loss": 0.6255,
+      "step": 12015
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1252665451731163,
+      "learning_rate": 1.3345056528261784e-06,
+      "loss": 0.5335,
+      "step": 12016
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5223357385665799,
+      "learning_rate": 1.3338007937662367e-06,
+      "loss": 0.6009,
+      "step": 12017
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9245410561113778,
+      "learning_rate": 1.3330960922510373e-06,
+      "loss": 0.6883,
+      "step": 12018
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0741381630535949,
+      "learning_rate": 1.3323915483108662e-06,
+      "loss": 0.5457,
+      "step": 12019
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3725169839855789,
+      "learning_rate": 1.3316871619759992e-06,
+      "loss": 0.6306,
+      "step": 12020
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.361676602285612,
+      "learning_rate": 1.3309829332767037e-06,
+      "loss": 0.5977,
+      "step": 12021
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3762335062549054,
+      "learning_rate": 1.330278862243244e-06,
+      "loss": 0.6391,
+      "step": 12022
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5208012514741749,
+      "learning_rate": 1.3295749489058741e-06,
+      "loss": 0.616,
+      "step": 12023
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3713957244068002,
+      "learning_rate": 1.3288711932948427e-06,
+      "loss": 0.6547,
+      "step": 12024
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5050264204093358,
+      "learning_rate": 1.3281675954403923e-06,
+      "loss": 0.7223,
+      "step": 12025
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5472608712099758,
+      "learning_rate": 1.3274641553727568e-06,
+      "loss": 0.5836,
+      "step": 12026
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3113167350958825,
+      "learning_rate": 1.3267608731221671e-06,
+      "loss": 0.5902,
+      "step": 12027
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4054278282394486,
+      "learning_rate": 1.3260577487188431e-06,
+      "loss": 0.6562,
+      "step": 12028
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.607029704164214,
+      "learning_rate": 1.3253547821930002e-06,
+      "loss": 0.6416,
+      "step": 12029
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.065122864973469,
+      "learning_rate": 1.3246519735748465e-06,
+      "loss": 0.6596,
+      "step": 12030
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3710778364450558,
+      "learning_rate": 1.3239493228945821e-06,
+      "loss": 0.6226,
+      "step": 12031
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4841368347139627,
+      "learning_rate": 1.323246830182403e-06,
+      "loss": 0.6706,
+      "step": 12032
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.300307680380195,
+      "learning_rate": 1.3225444954684962e-06,
+      "loss": 0.5522,
+      "step": 12033
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3117439085161837,
+      "learning_rate": 1.3218423187830409e-06,
+      "loss": 0.5311,
+      "step": 12034
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3227640746566782,
+      "learning_rate": 1.3211403001562155e-06,
+      "loss": 0.5258,
+      "step": 12035
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5759426855995153,
+      "learning_rate": 1.3204384396181819e-06,
+      "loss": 0.6304,
+      "step": 12036
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1761464716733825,
+      "learning_rate": 1.3197367371991048e-06,
+      "loss": 0.5163,
+      "step": 12037
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2712124021208884,
+      "learning_rate": 1.3190351929291367e-06,
+      "loss": 0.6041,
+      "step": 12038
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5238930736035816,
+      "learning_rate": 1.3183338068384243e-06,
+      "loss": 0.6462,
+      "step": 12039
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2811765582690113,
+      "learning_rate": 1.3176325789571075e-06,
+      "loss": 0.5357,
+      "step": 12040
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4679208938868895,
+      "learning_rate": 1.316931509315319e-06,
+      "loss": 0.562,
+      "step": 12041
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4379610015426427,
+      "learning_rate": 1.3162305979431883e-06,
+      "loss": 0.6277,
+      "step": 12042
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1326359952553096,
+      "learning_rate": 1.315529844870832e-06,
+      "loss": 0.5592,
+      "step": 12043
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6512530870426927,
+      "learning_rate": 1.3148292501283627e-06,
+      "loss": 0.7149,
+      "step": 12044
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.422014167353156,
+      "learning_rate": 1.314128813745889e-06,
+      "loss": 0.6589,
+      "step": 12045
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3255058977246192,
+      "learning_rate": 1.3134285357535093e-06,
+      "loss": 0.6756,
+      "step": 12046
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3206104448895664,
+      "learning_rate": 1.3127284161813153e-06,
+      "loss": 0.6678,
+      "step": 12047
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3771882616726225,
+      "learning_rate": 1.3120284550593936e-06,
+      "loss": 0.6093,
+      "step": 12048
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.345448202541536,
+      "learning_rate": 1.3113286524178232e-06,
+      "loss": 0.5993,
+      "step": 12049
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2051976014163641,
+      "learning_rate": 1.310629008286675e-06,
+      "loss": 0.6265,
+      "step": 12050
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6615105434605915,
+      "learning_rate": 1.3099295226960151e-06,
+      "loss": 0.667,
+      "step": 12051
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6643078540746978,
+      "learning_rate": 1.3092301956759008e-06,
+      "loss": 0.7296,
+      "step": 12052
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2637259930563036,
+      "learning_rate": 1.3085310272563872e-06,
+      "loss": 0.6647,
+      "step": 12053
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7062219443834703,
+      "learning_rate": 1.3078320174675141e-06,
+      "loss": 0.6378,
+      "step": 12054
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2491941363334933,
+      "learning_rate": 1.3071331663393233e-06,
+      "loss": 0.5282,
+      "step": 12055
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3922830279640184,
+      "learning_rate": 1.3064344739018452e-06,
+      "loss": 0.6636,
+      "step": 12056
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4759069920407206,
+      "learning_rate": 1.3057359401851034e-06,
+      "loss": 0.663,
+      "step": 12057
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2033306968927904,
+      "learning_rate": 1.305037565219116e-06,
+      "loss": 0.6136,
+      "step": 12058
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2979392086126358,
+      "learning_rate": 1.3043393490338918e-06,
+      "loss": 0.6256,
+      "step": 12059
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5759741286560343,
+      "learning_rate": 1.3036412916594398e-06,
+      "loss": 0.6629,
+      "step": 12060
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.533837990227598,
+      "learning_rate": 1.3029433931257524e-06,
+      "loss": 0.5708,
+      "step": 12061
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.425666112307212,
+      "learning_rate": 1.3022456534628191e-06,
+      "loss": 0.6928,
+      "step": 12062
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3054492678772782,
+      "learning_rate": 1.3015480727006274e-06,
+      "loss": 0.5939,
+      "step": 12063
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5925286373808643,
+      "learning_rate": 1.3008506508691516e-06,
+      "loss": 0.7111,
+      "step": 12064
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.264995115560874,
+      "learning_rate": 1.3001533879983618e-06,
+      "loss": 0.5606,
+      "step": 12065
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8620288899927793,
+      "learning_rate": 1.2994562841182213e-06,
+      "loss": 0.6514,
+      "step": 12066
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5476020269584791,
+      "learning_rate": 1.298759339258685e-06,
+      "loss": 0.6852,
+      "step": 12067
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2246194195917637,
+      "learning_rate": 1.2980625534497037e-06,
+      "loss": 0.5427,
+      "step": 12068
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2466031323218323,
+      "learning_rate": 1.2973659267212173e-06,
+      "loss": 0.6163,
+      "step": 12069
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3108164939587517,
+      "learning_rate": 1.2966694591031653e-06,
+      "loss": 0.6309,
+      "step": 12070
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0398131228288718,
+      "learning_rate": 1.2959731506254751e-06,
+      "loss": 0.587,
+      "step": 12071
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1940290333287142,
+      "learning_rate": 1.2952770013180655e-06,
+      "loss": 0.5895,
+      "step": 12072
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1196245473892457,
+      "learning_rate": 1.2945810112108554e-06,
+      "loss": 0.5482,
+      "step": 12073
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3099314074792578,
+      "learning_rate": 1.2938851803337516e-06,
+      "loss": 0.6588,
+      "step": 12074
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.330439270827365,
+      "learning_rate": 1.2931895087166551e-06,
+      "loss": 0.5302,
+      "step": 12075
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6903612924260791,
+      "learning_rate": 1.2924939963894617e-06,
+      "loss": 0.6952,
+      "step": 12076
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.664494372277042,
+      "learning_rate": 1.2917986433820567e-06,
+      "loss": 0.6417,
+      "step": 12077
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6416581044610086,
+      "learning_rate": 1.2911034497243251e-06,
+      "loss": 0.5874,
+      "step": 12078
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.402032047327968,
+      "learning_rate": 1.290408415446136e-06,
+      "loss": 0.757,
+      "step": 12079
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2316644675696422,
+      "learning_rate": 1.2897135405773602e-06,
+      "loss": 0.535,
+      "step": 12080
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3285171542937189,
+      "learning_rate": 1.2890188251478569e-06,
+      "loss": 0.6563,
+      "step": 12081
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5376100660411633,
+      "learning_rate": 1.2883242691874792e-06,
+      "loss": 0.6244,
+      "step": 12082
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2455082169006746,
+      "learning_rate": 1.2876298727260744e-06,
+      "loss": 0.6225,
+      "step": 12083
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1842371031760655,
+      "learning_rate": 1.2869356357934815e-06,
+      "loss": 0.6402,
+      "step": 12084
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4425206084190922,
+      "learning_rate": 1.2862415584195338e-06,
+      "loss": 0.5703,
+      "step": 12085
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2976568398006416,
+      "learning_rate": 1.2855476406340578e-06,
+      "loss": 0.6578,
+      "step": 12086
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9899255711574321,
+      "learning_rate": 1.28485388246687e-06,
+      "loss": 0.6177,
+      "step": 12087
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2745382622560306,
+      "learning_rate": 1.2841602839477868e-06,
+      "loss": 0.6978,
+      "step": 12088
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6993894739763384,
+      "learning_rate": 1.2834668451066118e-06,
+      "loss": 0.7569,
+      "step": 12089
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.266968228494469,
+      "learning_rate": 1.2827735659731438e-06,
+      "loss": 0.6282,
+      "step": 12090
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1550920001640665,
+      "learning_rate": 1.2820804465771742e-06,
+      "loss": 0.7562,
+      "step": 12091
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5162109225008447,
+      "learning_rate": 1.2813874869484866e-06,
+      "loss": 0.7409,
+      "step": 12092
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3414043465721832,
+      "learning_rate": 1.2806946871168634e-06,
+      "loss": 0.6074,
+      "step": 12093
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3652456710239338,
+      "learning_rate": 1.2800020471120717e-06,
+      "loss": 0.6361,
+      "step": 12094
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8265434197344455,
+      "learning_rate": 1.279309566963875e-06,
+      "loss": 0.6864,
+      "step": 12095
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2776429031346035,
+      "learning_rate": 1.2786172467020357e-06,
+      "loss": 0.6195,
+      "step": 12096
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4624462527846231,
+      "learning_rate": 1.2779250863562987e-06,
+      "loss": 0.7081,
+      "step": 12097
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2975510221979003,
+      "learning_rate": 1.2772330859564115e-06,
+      "loss": 0.6392,
+      "step": 12098
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4897693952759323,
+      "learning_rate": 1.276541245532109e-06,
+      "loss": 0.7162,
+      "step": 12099
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.28840315247448,
+      "learning_rate": 1.2758495651131226e-06,
+      "loss": 0.5638,
+      "step": 12100
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2123639357442129,
+      "learning_rate": 1.275158044729174e-06,
+      "loss": 0.5251,
+      "step": 12101
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3032710204931173,
+      "learning_rate": 1.2744666844099801e-06,
+      "loss": 0.5894,
+      "step": 12102
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3014313602200707,
+      "learning_rate": 1.2737754841852501e-06,
+      "loss": 0.542,
+      "step": 12103
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3579335773462742,
+      "learning_rate": 1.2730844440846862e-06,
+      "loss": 0.615,
+      "step": 12104
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3124015110006044,
+      "learning_rate": 1.2723935641379826e-06,
+      "loss": 0.6093,
+      "step": 12105
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0801540119993547,
+      "learning_rate": 1.2717028443748303e-06,
+      "loss": 0.6204,
+      "step": 12106
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1318250720812084,
+      "learning_rate": 1.2710122848249106e-06,
+      "loss": 0.6391,
+      "step": 12107
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7007350738394171,
+      "learning_rate": 1.2703218855178984e-06,
+      "loss": 0.6359,
+      "step": 12108
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2033522039505846,
+      "learning_rate": 1.2696316464834607e-06,
+      "loss": 0.6129,
+      "step": 12109
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.55638799097811,
+      "learning_rate": 1.2689415677512574e-06,
+      "loss": 0.5325,
+      "step": 12110
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5727603947632893,
+      "learning_rate": 1.2682516493509478e-06,
+      "loss": 0.5894,
+      "step": 12111
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2197369605795267,
+      "learning_rate": 1.267561891312174e-06,
+      "loss": 0.7225,
+      "step": 12112
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4287149243787578,
+      "learning_rate": 1.2668722936645773e-06,
+      "loss": 0.5957,
+      "step": 12113
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1040585555869364,
+      "learning_rate": 1.2661828564377948e-06,
+      "loss": 0.5539,
+      "step": 12114
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3091025107167098,
+      "learning_rate": 1.2654935796614481e-06,
+      "loss": 0.6773,
+      "step": 12115
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3767157962005951,
+      "learning_rate": 1.2648044633651607e-06,
+      "loss": 0.5895,
+      "step": 12116
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.394183510405307,
+      "learning_rate": 1.2641155075785444e-06,
+      "loss": 0.7619,
+      "step": 12117
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5678249531866828,
+      "learning_rate": 1.2634267123312049e-06,
+      "loss": 0.6386,
+      "step": 12118
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1612052990072628,
+      "learning_rate": 1.2627380776527415e-06,
+      "loss": 0.5999,
+      "step": 12119
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.196566613166791,
+      "learning_rate": 1.2620496035727448e-06,
+      "loss": 0.5636,
+      "step": 12120
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1358158879869125,
+      "learning_rate": 1.2613612901208022e-06,
+      "loss": 0.5453,
+      "step": 12121
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2175703684875545,
+      "learning_rate": 1.2606731373264936e-06,
+      "loss": 0.5245,
+      "step": 12122
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4507246707757817,
+      "learning_rate": 1.2599851452193851e-06,
+      "loss": 0.6949,
+      "step": 12123
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2709184202876787,
+      "learning_rate": 1.259297313829046e-06,
+      "loss": 0.5341,
+      "step": 12124
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.317596718453813,
+      "learning_rate": 1.258609643185032e-06,
+      "loss": 0.5453,
+      "step": 12125
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4671807555465808,
+      "learning_rate": 1.2579221333168945e-06,
+      "loss": 0.6186,
+      "step": 12126
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3936323563178403,
+      "learning_rate": 1.2572347842541766e-06,
+      "loss": 0.6088,
+      "step": 12127
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4216013984290823,
+      "learning_rate": 1.2565475960264145e-06,
+      "loss": 0.6372,
+      "step": 12128
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0759566913139773,
+      "learning_rate": 1.255860568663142e-06,
+      "loss": 0.7099,
+      "step": 12129
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3012737165510118,
+      "learning_rate": 1.2551737021938765e-06,
+      "loss": 0.628,
+      "step": 12130
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3990490924709753,
+      "learning_rate": 1.2544869966481389e-06,
+      "loss": 0.6983,
+      "step": 12131
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.450444002559874,
+      "learning_rate": 1.253800452055438e-06,
+      "loss": 0.6515,
+      "step": 12132
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.228371627187547,
+      "learning_rate": 1.2531140684452724e-06,
+      "loss": 0.6651,
+      "step": 12133
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.30578613198106,
+      "learning_rate": 1.2524278458471411e-06,
+      "loss": 0.6045,
+      "step": 12134
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7260575694375926,
+      "learning_rate": 1.2517417842905316e-06,
+      "loss": 0.6139,
+      "step": 12135
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.256821312456192,
+      "learning_rate": 1.2510558838049258e-06,
+      "loss": 0.5995,
+      "step": 12136
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1736037212192854,
+      "learning_rate": 1.2503701444197974e-06,
+      "loss": 0.6378,
+      "step": 12137
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3749746078541014,
+      "learning_rate": 1.249684566164614e-06,
+      "loss": 0.741,
+      "step": 12138
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3583094640799676,
+      "learning_rate": 1.248999149068838e-06,
+      "loss": 0.5863,
+      "step": 12139
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.368418856489613,
+      "learning_rate": 1.2483138931619227e-06,
+      "loss": 0.6327,
+      "step": 12140
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.387955410688036,
+      "learning_rate": 1.2476287984733154e-06,
+      "loss": 0.6958,
+      "step": 12141
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1406483061069417,
+      "learning_rate": 1.246943865032455e-06,
+      "loss": 0.608,
+      "step": 12142
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0601876430987245,
+      "learning_rate": 1.246259092868775e-06,
+      "loss": 0.6032,
+      "step": 12143
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3019146390501892,
+      "learning_rate": 1.2455744820117028e-06,
+      "loss": 0.5289,
+      "step": 12144
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5428552930374388,
+      "learning_rate": 1.2448900324906559e-06,
+      "loss": 0.6433,
+      "step": 12145
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0812898473519397,
+      "learning_rate": 1.2442057443350458e-06,
+      "loss": 0.6303,
+      "step": 12146
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.145765370228114,
+      "learning_rate": 1.2435216175742826e-06,
+      "loss": 0.5754,
+      "step": 12147
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2542021177369118,
+      "learning_rate": 1.242837652237759e-06,
+      "loss": 0.4756,
+      "step": 12148
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3671875576125267,
+      "learning_rate": 1.2421538483548706e-06,
+      "loss": 0.7099,
+      "step": 12149
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2616627224175296,
+      "learning_rate": 1.2414702059550016e-06,
+      "loss": 0.5536,
+      "step": 12150
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3494449917029019,
+      "learning_rate": 1.2407867250675253e-06,
+      "loss": 0.6491,
+      "step": 12151
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2854949555112656,
+      "learning_rate": 1.2401034057218181e-06,
+      "loss": 0.5703,
+      "step": 12152
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.46173389742557,
+      "learning_rate": 1.2394202479472406e-06,
+      "loss": 0.5372,
+      "step": 12153
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2344710077381311,
+      "learning_rate": 1.2387372517731505e-06,
+      "loss": 0.7265,
+      "step": 12154
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9124069541678249,
+      "learning_rate": 1.238054417228897e-06,
+      "loss": 0.591,
+      "step": 12155
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3513525375568103,
+      "learning_rate": 1.2373717443438227e-06,
+      "loss": 0.6777,
+      "step": 12156
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.390162520370915,
+      "learning_rate": 1.2366892331472657e-06,
+      "loss": 0.7298,
+      "step": 12157
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2267545713022707,
+      "learning_rate": 1.2360068836685536e-06,
+      "loss": 0.5116,
+      "step": 12158
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.903010878308355,
+      "learning_rate": 1.2353246959370086e-06,
+      "loss": 0.6792,
+      "step": 12159
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.288982483498578,
+      "learning_rate": 1.234642669981946e-06,
+      "loss": 0.7158,
+      "step": 12160
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1371258133739313,
+      "learning_rate": 1.2339608058326736e-06,
+      "loss": 0.8038,
+      "step": 12161
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.061899578487378,
+      "learning_rate": 1.2332791035184926e-06,
+      "loss": 0.6121,
+      "step": 12162
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4269288420979334,
+      "learning_rate": 1.2325975630686976e-06,
+      "loss": 0.583,
+      "step": 12163
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4718780029298313,
+      "learning_rate": 1.2319161845125744e-06,
+      "loss": 0.7148,
+      "step": 12164
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3619955538948028,
+      "learning_rate": 1.2312349678794072e-06,
+      "loss": 0.6297,
+      "step": 12165
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4413393425090126,
+      "learning_rate": 1.2305539131984646e-06,
+      "loss": 0.6668,
+      "step": 12166
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2920488823618506,
+      "learning_rate": 1.2298730204990162e-06,
+      "loss": 0.6263,
+      "step": 12167
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2101856770948858,
+      "learning_rate": 1.229192289810321e-06,
+      "loss": 0.5384,
+      "step": 12168
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.201565756479633,
+      "learning_rate": 1.228511721161631e-06,
+      "loss": 0.6084,
+      "step": 12169
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4698190018709785,
+      "learning_rate": 1.2278313145821924e-06,
+      "loss": 0.6405,
+      "step": 12170
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.438466388900323,
+      "learning_rate": 1.2271510701012412e-06,
+      "loss": 0.6572,
+      "step": 12171
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.447182346495059,
+      "learning_rate": 1.2264709877480147e-06,
+      "loss": 0.5896,
+      "step": 12172
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4412075676597806,
+      "learning_rate": 1.2257910675517315e-06,
+      "loss": 0.7667,
+      "step": 12173
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2379656692571261,
+      "learning_rate": 1.2251113095416113e-06,
+      "loss": 0.6379,
+      "step": 12174
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1879735091158325,
+      "learning_rate": 1.2244317137468658e-06,
+      "loss": 0.5689,
+      "step": 12175
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2262528581357823,
+      "learning_rate": 1.223752280196699e-06,
+      "loss": 0.6656,
+      "step": 12176
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1303799741774374,
+      "learning_rate": 1.2230730089203063e-06,
+      "loss": 0.6598,
+      "step": 12177
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.302874811951039,
+      "learning_rate": 1.2223938999468788e-06,
+      "loss": 0.6007,
+      "step": 12178
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4373492384513569,
+      "learning_rate": 1.2217149533055976e-06,
+      "loss": 0.5747,
+      "step": 12179
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1256764198694573,
+      "learning_rate": 1.22103616902564e-06,
+      "loss": 0.4796,
+      "step": 12180
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2285601946982936,
+      "learning_rate": 1.2203575471361733e-06,
+      "loss": 0.5909,
+      "step": 12181
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4159710264151937,
+      "learning_rate": 1.2196790876663617e-06,
+      "loss": 0.5613,
+      "step": 12182
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1992673108878091,
+      "learning_rate": 1.2190007906453605e-06,
+      "loss": 0.7141,
+      "step": 12183
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3664086931389368,
+      "learning_rate": 1.2183226561023132e-06,
+      "loss": 0.6765,
+      "step": 12184
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.669157975243307,
+      "learning_rate": 1.2176446840663647e-06,
+      "loss": 0.604,
+      "step": 12185
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2586036665308542,
+      "learning_rate": 1.216966874566648e-06,
+      "loss": 0.668,
+      "step": 12186
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9410960852256014,
+      "learning_rate": 1.21628922763229e-06,
+      "loss": 0.6513,
+      "step": 12187
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3790418895171974,
+      "learning_rate": 1.2156117432924108e-06,
+      "loss": 0.6952,
+      "step": 12188
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4308769023459258,
+      "learning_rate": 1.2149344215761216e-06,
+      "loss": 0.6414,
+      "step": 12189
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4597964294522257,
+      "learning_rate": 1.2142572625125332e-06,
+      "loss": 0.5875,
+      "step": 12190
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.379972802517962,
+      "learning_rate": 1.2135802661307388e-06,
+      "loss": 0.5751,
+      "step": 12191
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9993305170227996,
+      "learning_rate": 1.2129034324598343e-06,
+      "loss": 0.5903,
+      "step": 12192
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.51158020469434,
+      "learning_rate": 1.212226761528904e-06,
+      "loss": 0.6806,
+      "step": 12193
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0813300660703185,
+      "learning_rate": 1.2115502533670253e-06,
+      "loss": 0.5374,
+      "step": 12194
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5087140157808945,
+      "learning_rate": 1.2108739080032695e-06,
+      "loss": 0.648,
+      "step": 12195
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0937719136657216,
+      "learning_rate": 1.2101977254667007e-06,
+      "loss": 0.5702,
+      "step": 12196
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3533866342030958,
+      "learning_rate": 1.2095217057863762e-06,
+      "loss": 0.7898,
+      "step": 12197
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2633251666265055,
+      "learning_rate": 1.2088458489913458e-06,
+      "loss": 0.6434,
+      "step": 12198
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4062499660855632,
+      "learning_rate": 1.2081701551106506e-06,
+      "loss": 0.6969,
+      "step": 12199
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3158735351446857,
+      "learning_rate": 1.207494624173331e-06,
+      "loss": 0.5984,
+      "step": 12200
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4301146306074362,
+      "learning_rate": 1.2068192562084146e-06,
+      "loss": 0.5623,
+      "step": 12201
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.369723754208122,
+      "learning_rate": 1.2061440512449196e-06,
+      "loss": 0.6469,
+      "step": 12202
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4295445144283623,
+      "learning_rate": 1.205469009311866e-06,
+      "loss": 0.5645,
+      "step": 12203
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2716669033668255,
+      "learning_rate": 1.20479413043826e-06,
+      "loss": 0.6628,
+      "step": 12204
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.147964356951294,
+      "learning_rate": 1.2041194146531026e-06,
+      "loss": 0.5414,
+      "step": 12205
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.541965575603514,
+      "learning_rate": 1.2034448619853873e-06,
+      "loss": 0.6541,
+      "step": 12206
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.683465825110658,
+      "learning_rate": 1.202770472464101e-06,
+      "loss": 0.7581,
+      "step": 12207
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5757193956934954,
+      "learning_rate": 1.2020962461182268e-06,
+      "loss": 0.6171,
+      "step": 12208
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.340932125772874,
+      "learning_rate": 1.201422182976732e-06,
+      "loss": 0.8253,
+      "step": 12209
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.878361957515791,
+      "learning_rate": 1.200748283068588e-06,
+      "loss": 0.6984,
+      "step": 12210
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.325877770178706,
+      "learning_rate": 1.2000745464227515e-06,
+      "loss": 0.6269,
+      "step": 12211
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1156655255347785,
+      "learning_rate": 1.1994009730681744e-06,
+      "loss": 0.6687,
+      "step": 12212
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1286255420074038,
+      "learning_rate": 1.198727563033802e-06,
+      "loss": 0.6577,
+      "step": 12213
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.0333847090134585,
+      "learning_rate": 1.1980543163485726e-06,
+      "loss": 0.544,
+      "step": 12214
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3349250788518305,
+      "learning_rate": 1.1973812330414159e-06,
+      "loss": 0.5895,
+      "step": 12215
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.34863406504787,
+      "learning_rate": 1.1967083131412566e-06,
+      "loss": 0.5548,
+      "step": 12216
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2516768332767014,
+      "learning_rate": 1.1960355566770104e-06,
+      "loss": 0.5733,
+      "step": 12217
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.288936539675147,
+      "learning_rate": 1.1953629636775893e-06,
+      "loss": 0.5758,
+      "step": 12218
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.220733962086156,
+      "learning_rate": 1.1946905341718951e-06,
+      "loss": 0.5605,
+      "step": 12219
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7823980044466392,
+      "learning_rate": 1.1940182681888235e-06,
+      "loss": 0.6462,
+      "step": 12220
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3364992766017585,
+      "learning_rate": 1.1933461657572637e-06,
+      "loss": 0.6469,
+      "step": 12221
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3916903813477406,
+      "learning_rate": 1.1926742269060965e-06,
+      "loss": 0.6232,
+      "step": 12222
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.487965516869136,
+      "learning_rate": 1.1920024516641976e-06,
+      "loss": 0.6367,
+      "step": 12223
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.231889309015284,
+      "learning_rate": 1.1913308400604339e-06,
+      "loss": 0.5561,
+      "step": 12224
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2141789226423227,
+      "learning_rate": 1.1906593921236649e-06,
+      "loss": 0.5909,
+      "step": 12225
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.212616398017904,
+      "learning_rate": 1.1899881078827485e-06,
+      "loss": 0.5497,
+      "step": 12226
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0464330454096467,
+      "learning_rate": 1.189316987366526e-06,
+      "loss": 0.6158,
+      "step": 12227
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3513416488501024,
+      "learning_rate": 1.1886460306038405e-06,
+      "loss": 0.741,
+      "step": 12228
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5345362845823163,
+      "learning_rate": 1.1879752376235231e-06,
+      "loss": 0.6132,
+      "step": 12229
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3060205391363953,
+      "learning_rate": 1.1873046084544004e-06,
+      "loss": 0.5663,
+      "step": 12230
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5127794139817934,
+      "learning_rate": 1.1866341431252892e-06,
+      "loss": 0.7539,
+      "step": 12231
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5927638030296474,
+      "learning_rate": 1.185963841665001e-06,
+      "loss": 0.6782,
+      "step": 12232
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.364996502346154,
+      "learning_rate": 1.1852937041023433e-06,
+      "loss": 0.6231,
+      "step": 12233
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4777375248585431,
+      "learning_rate": 1.1846237304661095e-06,
+      "loss": 0.5641,
+      "step": 12234
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6657873536535757,
+      "learning_rate": 1.1839539207850898e-06,
+      "loss": 0.7086,
+      "step": 12235
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2946360295937336,
+      "learning_rate": 1.1832842750880702e-06,
+      "loss": 0.6408,
+      "step": 12236
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.332477707317221,
+      "learning_rate": 1.1826147934038262e-06,
+      "loss": 0.5735,
+      "step": 12237
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4568059280818784,
+      "learning_rate": 1.181945475761126e-06,
+      "loss": 0.6158,
+      "step": 12238
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5222856951626316,
+      "learning_rate": 1.181276322188732e-06,
+      "loss": 0.5866,
+      "step": 12239
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.339340736955866,
+      "learning_rate": 1.1806073327153988e-06,
+      "loss": 0.5225,
+      "step": 12240
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7819038479385425,
+      "learning_rate": 1.1799385073698749e-06,
+      "loss": 0.7538,
+      "step": 12241
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3119169923471903,
+      "learning_rate": 1.1792698461809e-06,
+      "loss": 0.5796,
+      "step": 12242
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3178350344523224,
+      "learning_rate": 1.1786013491772103e-06,
+      "loss": 0.6287,
+      "step": 12243
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6380217056416888,
+      "learning_rate": 1.1779330163875325e-06,
+      "loss": 0.6614,
+      "step": 12244
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9844810100095279,
+      "learning_rate": 1.1772648478405824e-06,
+      "loss": 0.6301,
+      "step": 12245
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3955614933603238,
+      "learning_rate": 1.176596843565077e-06,
+      "loss": 0.6174,
+      "step": 12246
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.273095675563449,
+      "learning_rate": 1.1759290035897208e-06,
+      "loss": 0.6925,
+      "step": 12247
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3115038256960487,
+      "learning_rate": 1.1752613279432112e-06,
+      "loss": 0.5508,
+      "step": 12248
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3052551316672845,
+      "learning_rate": 1.1745938166542414e-06,
+      "loss": 0.5611,
+      "step": 12249
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2606018394858174,
+      "learning_rate": 1.173926469751493e-06,
+      "loss": 0.5924,
+      "step": 12250
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2003273659261169,
+      "learning_rate": 1.1732592872636478e-06,
+      "loss": 0.6061,
+      "step": 12251
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4632385832947792,
+      "learning_rate": 1.172592269219373e-06,
+      "loss": 0.515,
+      "step": 12252
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1297195790605783,
+      "learning_rate": 1.1719254156473302e-06,
+      "loss": 0.6882,
+      "step": 12253
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.310651882694194,
+      "learning_rate": 1.1712587265761799e-06,
+      "loss": 0.6621,
+      "step": 12254
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3510707270469535,
+      "learning_rate": 1.1705922020345688e-06,
+      "loss": 0.9189,
+      "step": 12255
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3798601877228904,
+      "learning_rate": 1.16992584205114e-06,
+      "loss": 0.6257,
+      "step": 12256
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7215126073711042,
+      "learning_rate": 1.1692596466545275e-06,
+      "loss": 0.6813,
+      "step": 12257
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5436054775776773,
+      "learning_rate": 1.168593615873358e-06,
+      "loss": 0.6623,
+      "step": 12258
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.416144910766683,
+      "learning_rate": 1.1679277497362563e-06,
+      "loss": 0.5917,
+      "step": 12259
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4117185069803813,
+      "learning_rate": 1.167262048271831e-06,
+      "loss": 0.5155,
+      "step": 12260
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3150699690064709,
+      "learning_rate": 1.166596511508693e-06,
+      "loss": 0.5378,
+      "step": 12261
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.314279480238854,
+      "learning_rate": 1.1659311394754418e-06,
+      "loss": 0.621,
+      "step": 12262
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3933903685948268,
+      "learning_rate": 1.1652659322006659e-06,
+      "loss": 0.6604,
+      "step": 12263
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0546361719371156,
+      "learning_rate": 1.1646008897129546e-06,
+      "loss": 0.7024,
+      "step": 12264
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2818483319111322,
+      "learning_rate": 1.1639360120408843e-06,
+      "loss": 0.6296,
+      "step": 12265
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4019191335479309,
+      "learning_rate": 1.1632712992130278e-06,
+      "loss": 0.5999,
+      "step": 12266
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5888237229571727,
+      "learning_rate": 1.1626067512579481e-06,
+      "loss": 0.6454,
+      "step": 12267
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3972240162182887,
+      "learning_rate": 1.1619423682042008e-06,
+      "loss": 0.7375,
+      "step": 12268
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5611552322245024,
+      "learning_rate": 1.161278150080341e-06,
+      "loss": 0.692,
+      "step": 12269
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1345233653690492,
+      "learning_rate": 1.160614096914906e-06,
+      "loss": 0.5179,
+      "step": 12270
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9704444482749919,
+      "learning_rate": 1.1599502087364345e-06,
+      "loss": 0.5935,
+      "step": 12271
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4122684443834346,
+      "learning_rate": 1.1592864855734553e-06,
+      "loss": 0.6184,
+      "step": 12272
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3996608474370702,
+      "learning_rate": 1.1586229274544898e-06,
+      "loss": 0.5818,
+      "step": 12273
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6373807057889607,
+      "learning_rate": 1.157959534408052e-06,
+      "loss": 0.6459,
+      "step": 12274
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6185988095602535,
+      "learning_rate": 1.1572963064626497e-06,
+      "loss": 0.5924,
+      "step": 12275
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.043554060600506,
+      "learning_rate": 1.1566332436467815e-06,
+      "loss": 0.4701,
+      "step": 12276
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6512564127101965,
+      "learning_rate": 1.155970345988946e-06,
+      "loss": 0.5421,
+      "step": 12277
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1618734925162013,
+      "learning_rate": 1.1553076135176222e-06,
+      "loss": 0.5694,
+      "step": 12278
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2990313256493011,
+      "learning_rate": 1.1546450462612951e-06,
+      "loss": 0.6388,
+      "step": 12279
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4223325288741053,
+      "learning_rate": 1.1539826442484337e-06,
+      "loss": 0.6514,
+      "step": 12280
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2127364295860887,
+      "learning_rate": 1.1533204075075039e-06,
+      "loss": 0.4269,
+      "step": 12281
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4113173314865797,
+      "learning_rate": 1.1526583360669635e-06,
+      "loss": 0.585,
+      "step": 12282
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4263902939095128,
+      "learning_rate": 1.1519964299552611e-06,
+      "loss": 0.5527,
+      "step": 12283
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.619318779892476,
+      "learning_rate": 1.151334689200845e-06,
+      "loss": 0.716,
+      "step": 12284
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6699718262477954,
+      "learning_rate": 1.1506731138321474e-06,
+      "loss": 0.6817,
+      "step": 12285
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4319352867677717,
+      "learning_rate": 1.1500117038775977e-06,
+      "loss": 0.6469,
+      "step": 12286
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4204141843574138,
+      "learning_rate": 1.149350459365622e-06,
+      "loss": 0.6322,
+      "step": 12287
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.334356067213292,
+      "learning_rate": 1.1486893803246302e-06,
+      "loss": 0.4799,
+      "step": 12288
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5302648654173256,
+      "learning_rate": 1.1480284667830343e-06,
+      "loss": 0.6454,
+      "step": 12289
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4582820580842641,
+      "learning_rate": 1.1473677187692345e-06,
+      "loss": 0.6972,
+      "step": 12290
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8321778099181107,
+      "learning_rate": 1.1467071363116234e-06,
+      "loss": 0.7462,
+      "step": 12291
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5065831501708888,
+      "learning_rate": 1.1460467194385889e-06,
+      "loss": 0.6499,
+      "step": 12292
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4505262388212252,
+      "learning_rate": 1.1453864681785087e-06,
+      "loss": 0.7181,
+      "step": 12293
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4010312354849233,
+      "learning_rate": 1.1447263825597577e-06,
+      "loss": 0.6405,
+      "step": 12294
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3064657212465014,
+      "learning_rate": 1.1440664626107017e-06,
+      "loss": 0.6326,
+      "step": 12295
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3657567129726307,
+      "learning_rate": 1.143406708359695e-06,
+      "loss": 0.5701,
+      "step": 12296
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.150897359241321,
+      "learning_rate": 1.142747119835092e-06,
+      "loss": 0.5002,
+      "step": 12297
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.268933988742614,
+      "learning_rate": 1.1420876970652361e-06,
+      "loss": 0.6308,
+      "step": 12298
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1752841083464827,
+      "learning_rate": 1.1414284400784643e-06,
+      "loss": 0.6326,
+      "step": 12299
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8030078395005082,
+      "learning_rate": 1.1407693489031052e-06,
+      "loss": 0.6705,
+      "step": 12300
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1989749396052918,
+      "learning_rate": 1.1401104235674808e-06,
+      "loss": 0.4814,
+      "step": 12301
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2647586715343682,
+      "learning_rate": 1.1394516640999114e-06,
+      "loss": 0.6023,
+      "step": 12302
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.289042725290082,
+      "learning_rate": 1.1387930705286998e-06,
+      "loss": 0.5578,
+      "step": 12303
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.466144996728059,
+      "learning_rate": 1.1381346428821482e-06,
+      "loss": 0.6736,
+      "step": 12304
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1633350931787239,
+      "learning_rate": 1.1374763811885547e-06,
+      "loss": 0.5743,
+      "step": 12305
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3085647943931864,
+      "learning_rate": 1.1368182854762005e-06,
+      "loss": 0.5635,
+      "step": 12306
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0712478847134204,
+      "learning_rate": 1.1361603557733698e-06,
+      "loss": 0.6919,
+      "step": 12307
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7730073417001664,
+      "learning_rate": 1.135502592108334e-06,
+      "loss": 0.6456,
+      "step": 12308
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3835462427229641,
+      "learning_rate": 1.134844994509358e-06,
+      "loss": 0.6724,
+      "step": 12309
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3368612349926545,
+      "learning_rate": 1.1341875630047012e-06,
+      "loss": 0.6917,
+      "step": 12310
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3220893538138796,
+      "learning_rate": 1.1335302976226132e-06,
+      "loss": 0.6835,
+      "step": 12311
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5954559549428387,
+      "learning_rate": 1.1328731983913404e-06,
+      "loss": 0.7756,
+      "step": 12312
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3320685021593008,
+      "learning_rate": 1.13221626533912e-06,
+      "loss": 0.6794,
+      "step": 12313
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0806845893109822,
+      "learning_rate": 1.1315594984941786e-06,
+      "loss": 0.505,
+      "step": 12314
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0676800305287508,
+      "learning_rate": 1.1309028978847419e-06,
+      "loss": 0.638,
+      "step": 12315
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8274986904115782,
+      "learning_rate": 1.1302464635390247e-06,
+      "loss": 0.6267,
+      "step": 12316
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.725458369837837,
+      "learning_rate": 1.1295901954852356e-06,
+      "loss": 0.6106,
+      "step": 12317
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.233812300308324,
+      "learning_rate": 1.1289340937515753e-06,
+      "loss": 0.576,
+      "step": 12318
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4186784052513226,
+      "learning_rate": 1.1282781583662372e-06,
+      "loss": 0.6189,
+      "step": 12319
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.363542809528052,
+      "learning_rate": 1.1276223893574123e-06,
+      "loss": 0.6643,
+      "step": 12320
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5020947813787326,
+      "learning_rate": 1.1269667867532746e-06,
+      "loss": 0.7345,
+      "step": 12321
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3207577172068645,
+      "learning_rate": 1.1263113505820012e-06,
+      "loss": 0.638,
+      "step": 12322
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5195995249124434,
+      "learning_rate": 1.1256560808717576e-06,
+      "loss": 0.6138,
+      "step": 12323
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5221268104464178,
+      "learning_rate": 1.1250009776506982e-06,
+      "loss": 0.6931,
+      "step": 12324
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2829106928800893,
+      "learning_rate": 1.124346040946978e-06,
+      "loss": 0.601,
+      "step": 12325
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.623418396420793,
+      "learning_rate": 1.1236912707887404e-06,
+      "loss": 0.5595,
+      "step": 12326
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5010950671807541,
+      "learning_rate": 1.1230366672041216e-06,
+      "loss": 0.7661,
+      "step": 12327
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3511060257363883,
+      "learning_rate": 1.1223822302212517e-06,
+      "loss": 0.6314,
+      "step": 12328
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4566975159005038,
+      "learning_rate": 1.1217279598682518e-06,
+      "loss": 0.6716,
+      "step": 12329
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4546341723517486,
+      "learning_rate": 1.1210738561732405e-06,
+      "loss": 0.7658,
+      "step": 12330
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6607312241721526,
+      "learning_rate": 1.120419919164324e-06,
+      "loss": 0.6786,
+      "step": 12331
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3916965447404448,
+      "learning_rate": 1.1197661488696043e-06,
+      "loss": 0.6143,
+      "step": 12332
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7144684247343056,
+      "learning_rate": 1.1191125453171748e-06,
+      "loss": 0.7452,
+      "step": 12333
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2442175468181451,
+      "learning_rate": 1.118459108535122e-06,
+      "loss": 0.6253,
+      "step": 12334
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2905307511583617,
+      "learning_rate": 1.1178058385515268e-06,
+      "loss": 0.5779,
+      "step": 12335
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1821226554457274,
+      "learning_rate": 1.1171527353944601e-06,
+      "loss": 0.6652,
+      "step": 12336
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0946365749785074,
+      "learning_rate": 1.1164997990919868e-06,
+      "loss": 0.5242,
+      "step": 12337
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6240927111496828,
+      "learning_rate": 1.1158470296721691e-06,
+      "loss": 0.5245,
+      "step": 12338
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4574578312984803,
+      "learning_rate": 1.1151944271630517e-06,
+      "loss": 0.5285,
+      "step": 12339
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.305755956614687,
+      "learning_rate": 1.1145419915926836e-06,
+      "loss": 0.5903,
+      "step": 12340
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2342606620709746,
+      "learning_rate": 1.1138897229890995e-06,
+      "loss": 0.603,
+      "step": 12341
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5803911720441992,
+      "learning_rate": 1.1132376213803286e-06,
+      "loss": 0.69,
+      "step": 12342
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.420060700326728,
+      "learning_rate": 1.1125856867943936e-06,
+      "loss": 0.6601,
+      "step": 12343
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2736707108314904,
+      "learning_rate": 1.1119339192593077e-06,
+      "loss": 0.6787,
+      "step": 12344
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.450111521587073,
+      "learning_rate": 1.111282318803083e-06,
+      "loss": 0.6231,
+      "step": 12345
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.11403127387135,
+      "learning_rate": 1.1106308854537167e-06,
+      "loss": 0.6593,
+      "step": 12346
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4758959072581541,
+      "learning_rate": 1.1099796192392021e-06,
+      "loss": 0.6742,
+      "step": 12347
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3783938566077598,
+      "learning_rate": 1.109328520187528e-06,
+      "loss": 0.5823,
+      "step": 12348
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3675564296830662,
+      "learning_rate": 1.1086775883266725e-06,
+      "loss": 0.6886,
+      "step": 12349
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0057308694627545,
+      "learning_rate": 1.1080268236846076e-06,
+      "loss": 0.7401,
+      "step": 12350
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3932614546758875,
+      "learning_rate": 1.1073762262892979e-06,
+      "loss": 0.6711,
+      "step": 12351
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2913117000442849,
+      "learning_rate": 1.1067257961687005e-06,
+      "loss": 0.6246,
+      "step": 12352
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.329219396600821,
+      "learning_rate": 1.1060755333507673e-06,
+      "loss": 0.5345,
+      "step": 12353
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3636508479996896,
+      "learning_rate": 1.1054254378634399e-06,
+      "loss": 0.6377,
+      "step": 12354
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2229251255706026,
+      "learning_rate": 1.1047755097346541e-06,
+      "loss": 0.634,
+      "step": 12355
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.457951586126651,
+      "learning_rate": 1.1041257489923429e-06,
+      "loss": 0.711,
+      "step": 12356
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.31316096539369,
+      "learning_rate": 1.1034761556644214e-06,
+      "loss": 0.7263,
+      "step": 12357
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2982781676236705,
+      "learning_rate": 1.1028267297788092e-06,
+      "loss": 0.5712,
+      "step": 12358
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2617901587476195,
+      "learning_rate": 1.102177471363412e-06,
+      "loss": 0.5852,
+      "step": 12359
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4101764589995436,
+      "learning_rate": 1.1015283804461297e-06,
+      "loss": 0.6852,
+      "step": 12360
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2636926052527166,
+      "learning_rate": 1.1008794570548554e-06,
+      "loss": 0.6398,
+      "step": 12361
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5748805310396492,
+      "learning_rate": 1.100230701217473e-06,
+      "loss": 0.6622,
+      "step": 12362
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1635471009881202,
+      "learning_rate": 1.0995821129618656e-06,
+      "loss": 0.7075,
+      "step": 12363
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1543811761639162,
+      "learning_rate": 1.0989336923158999e-06,
+      "loss": 0.6664,
+      "step": 12364
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5009939649133106,
+      "learning_rate": 1.0982854393074399e-06,
+      "loss": 0.644,
+      "step": 12365
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4211189047062407,
+      "learning_rate": 1.0976373539643465e-06,
+      "loss": 0.7104,
+      "step": 12366
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3149586909083486,
+      "learning_rate": 1.0969894363144661e-06,
+      "loss": 0.662,
+      "step": 12367
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3424364653713254,
+      "learning_rate": 1.0963416863856425e-06,
+      "loss": 0.6808,
+      "step": 12368
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4260718310228864,
+      "learning_rate": 1.0956941042057106e-06,
+      "loss": 0.5794,
+      "step": 12369
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.490768255913994,
+      "learning_rate": 1.095046689802498e-06,
+      "loss": 0.7453,
+      "step": 12370
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1215501488609687,
+      "learning_rate": 1.0943994432038262e-06,
+      "loss": 0.5731,
+      "step": 12371
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4069961263753747,
+      "learning_rate": 1.0937523644375076e-06,
+      "loss": 0.5817,
+      "step": 12372
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5900180851619867,
+      "learning_rate": 1.0931054535313507e-06,
+      "loss": 0.6415,
+      "step": 12373
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3530068924986798,
+      "learning_rate": 1.0924587105131546e-06,
+      "loss": 0.5854,
+      "step": 12374
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2049982736366793,
+      "learning_rate": 1.0918121354107076e-06,
+      "loss": 0.5015,
+      "step": 12375
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2210061400233756,
+      "learning_rate": 1.091165728251799e-06,
+      "loss": 0.584,
+      "step": 12376
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.269360592691573,
+      "learning_rate": 1.090519489064204e-06,
+      "loss": 0.5868,
+      "step": 12377
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2741720996283181,
+      "learning_rate": 1.0898734178756937e-06,
+      "loss": 0.5828,
+      "step": 12378
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6241298135804718,
+      "learning_rate": 1.0892275147140307e-06,
+      "loss": 0.6517,
+      "step": 12379
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6682715017618517,
+      "learning_rate": 1.0885817796069697e-06,
+      "loss": 0.5288,
+      "step": 12380
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6026633019474905,
+      "learning_rate": 1.087936212582264e-06,
+      "loss": 0.6659,
+      "step": 12381
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2586195479782871,
+      "learning_rate": 1.0872908136676491e-06,
+      "loss": 0.761,
+      "step": 12382
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6346858308139232,
+      "learning_rate": 1.0866455828908634e-06,
+      "loss": 0.6793,
+      "step": 12383
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5800347714394238,
+      "learning_rate": 1.086000520279632e-06,
+      "loss": 0.6383,
+      "step": 12384
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.780041855652133,
+      "learning_rate": 1.0853556258616755e-06,
+      "loss": 0.7604,
+      "step": 12385
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4261073438983602,
+      "learning_rate": 1.0847108996647066e-06,
+      "loss": 0.6934,
+      "step": 12386
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2810184218256737,
+      "learning_rate": 1.0840663417164293e-06,
+      "loss": 0.5516,
+      "step": 12387
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3120478916340463,
+      "learning_rate": 1.0834219520445432e-06,
+      "loss": 0.6659,
+      "step": 12388
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.316399651909026,
+      "learning_rate": 1.0827777306767384e-06,
+      "loss": 0.5749,
+      "step": 12389
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1350850097140728,
+      "learning_rate": 1.082133677640697e-06,
+      "loss": 0.5944,
+      "step": 12390
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3063443194403435,
+      "learning_rate": 1.0814897929640984e-06,
+      "loss": 0.6359,
+      "step": 12391
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9893501733259071,
+      "learning_rate": 1.0808460766746104e-06,
+      "loss": 0.5818,
+      "step": 12392
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3781880420804458,
+      "learning_rate": 1.0802025287998953e-06,
+      "loss": 0.6644,
+      "step": 12393
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5446491090810117,
+      "learning_rate": 1.0795591493676072e-06,
+      "loss": 0.6142,
+      "step": 12394
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4301994892369019,
+      "learning_rate": 1.078915938405392e-06,
+      "loss": 0.6716,
+      "step": 12395
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3978401004550536,
+      "learning_rate": 1.078272895940895e-06,
+      "loss": 0.6858,
+      "step": 12396
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6675961636268188,
+      "learning_rate": 1.0776300220017437e-06,
+      "loss": 0.6797,
+      "step": 12397
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4938572864543158,
+      "learning_rate": 1.0769873166155653e-06,
+      "loss": 0.6191,
+      "step": 12398
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.613986014765033,
+      "learning_rate": 1.0763447798099813e-06,
+      "loss": 0.7452,
+      "step": 12399
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.24921102200595,
+      "learning_rate": 1.075702411612598e-06,
+      "loss": 0.7176,
+      "step": 12400
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.835443862374067,
+      "learning_rate": 1.0750602120510233e-06,
+      "loss": 0.6147,
+      "step": 12401
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5207534885587057,
+      "learning_rate": 1.0744181811528526e-06,
+      "loss": 0.6985,
+      "step": 12402
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4567564375628674,
+      "learning_rate": 1.073776318945675e-06,
+      "loss": 0.6078,
+      "step": 12403
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.262909165111246,
+      "learning_rate": 1.0731346254570735e-06,
+      "loss": 0.5797,
+      "step": 12404
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2589444689040596,
+      "learning_rate": 1.0724931007146227e-06,
+      "loss": 0.6013,
+      "step": 12405
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4589133752026653,
+      "learning_rate": 1.0718517447458904e-06,
+      "loss": 0.674,
+      "step": 12406
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2943412525315154,
+      "learning_rate": 1.071210557578437e-06,
+      "loss": 0.7049,
+      "step": 12407
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2591259518592768,
+      "learning_rate": 1.0705695392398146e-06,
+      "loss": 0.5584,
+      "step": 12408
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2203368043453033,
+      "learning_rate": 1.0699286897575718e-06,
+      "loss": 0.6149,
+      "step": 12409
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5734600433447528,
+      "learning_rate": 1.0692880091592457e-06,
+      "loss": 0.7446,
+      "step": 12410
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.274567334465294,
+      "learning_rate": 1.068647497472368e-06,
+      "loss": 0.563,
+      "step": 12411
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.428722516213334,
+      "learning_rate": 1.0680071547244636e-06,
+      "loss": 0.642,
+      "step": 12412
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3208269942964659,
+      "learning_rate": 1.0673669809430465e-06,
+      "loss": 0.6486,
+      "step": 12413
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3969099006535335,
+      "learning_rate": 1.066726976155632e-06,
+      "loss": 0.6023,
+      "step": 12414
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4985508628236868,
+      "learning_rate": 1.0660871403897177e-06,
+      "loss": 0.7054,
+      "step": 12415
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1039619104866647,
+      "learning_rate": 1.0654474736727988e-06,
+      "loss": 0.7003,
+      "step": 12416
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3203875273877836,
+      "learning_rate": 1.0648079760323675e-06,
+      "loss": 0.6364,
+      "step": 12417
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3619940925435043,
+      "learning_rate": 1.064168647495899e-06,
+      "loss": 0.6338,
+      "step": 12418
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4971516129796953,
+      "learning_rate": 1.0635294880908702e-06,
+      "loss": 0.6819,
+      "step": 12419
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.557756118504502,
+      "learning_rate": 1.0628904978447463e-06,
+      "loss": 0.7105,
+      "step": 12420
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.769106986919823,
+      "learning_rate": 1.0622516767849867e-06,
+      "loss": 0.7119,
+      "step": 12421
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.085228097929015,
+      "learning_rate": 1.0616130249390423e-06,
+      "loss": 0.6983,
+      "step": 12422
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5170481682241233,
+      "learning_rate": 1.0609745423343553e-06,
+      "loss": 0.6662,
+      "step": 12423
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1567997772731797,
+      "learning_rate": 1.0603362289983687e-06,
+      "loss": 0.5796,
+      "step": 12424
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2807673270710522,
+      "learning_rate": 1.0596980849585065e-06,
+      "loss": 0.5534,
+      "step": 12425
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0100984811062141,
+      "learning_rate": 1.0590601102421916e-06,
+      "loss": 0.5909,
+      "step": 12426
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1120065584468888,
+      "learning_rate": 1.0584223048768427e-06,
+      "loss": 0.6119,
+      "step": 12427
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4022413902496782,
+      "learning_rate": 1.0577846688898652e-06,
+      "loss": 0.7249,
+      "step": 12428
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0263463665954464,
+      "learning_rate": 1.0571472023086604e-06,
+      "loss": 0.612,
+      "step": 12429
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2904245995965467,
+      "learning_rate": 1.0565099051606214e-06,
+      "loss": 0.5833,
+      "step": 12430
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3919673426925865,
+      "learning_rate": 1.055872777473133e-06,
+      "loss": 0.7027,
+      "step": 12431
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4508508200500174,
+      "learning_rate": 1.0552358192735784e-06,
+      "loss": 0.6531,
+      "step": 12432
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4123781871854915,
+      "learning_rate": 1.0545990305893233e-06,
+      "loss": 0.7311,
+      "step": 12433
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6967125323850787,
+      "learning_rate": 1.053962411447736e-06,
+      "loss": 0.5995,
+      "step": 12434
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7928568798162139,
+      "learning_rate": 1.0533259618761738e-06,
+      "loss": 0.5712,
+      "step": 12435
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.36572525696927,
+      "learning_rate": 1.0526896819019817e-06,
+      "loss": 0.668,
+      "step": 12436
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2546869181917546,
+      "learning_rate": 1.0520535715525066e-06,
+      "loss": 0.6005,
+      "step": 12437
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7199926670132855,
+      "learning_rate": 1.0514176308550817e-06,
+      "loss": 0.7504,
+      "step": 12438
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3398653094161932,
+      "learning_rate": 1.0507818598370355e-06,
+      "loss": 0.6345,
+      "step": 12439
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2824342669431399,
+      "learning_rate": 1.0501462585256883e-06,
+      "loss": 0.6575,
+      "step": 12440
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4805398583381986,
+      "learning_rate": 1.0495108269483518e-06,
+      "loss": 0.6349,
+      "step": 12441
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4641134134221723,
+      "learning_rate": 1.0488755651323358e-06,
+      "loss": 0.6745,
+      "step": 12442
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.242375218068506,
+      "learning_rate": 1.048240473104934e-06,
+      "loss": 0.5308,
+      "step": 12443
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2862997735389967,
+      "learning_rate": 1.0476055508934408e-06,
+      "loss": 0.563,
+      "step": 12444
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.227206816385542,
+      "learning_rate": 1.04697079852514e-06,
+      "loss": 0.5771,
+      "step": 12445
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1327343967064516,
+      "learning_rate": 1.0463362160273076e-06,
+      "loss": 0.648,
+      "step": 12446
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1599943817771947,
+      "learning_rate": 1.0457018034272136e-06,
+      "loss": 0.6131,
+      "step": 12447
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3809182244013543,
+      "learning_rate": 1.04506756075212e-06,
+      "loss": 0.6734,
+      "step": 12448
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4040346497906762,
+      "learning_rate": 1.0444334880292794e-06,
+      "loss": 0.5796,
+      "step": 12449
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.1635683839467985,
+      "learning_rate": 1.0437995852859445e-06,
+      "loss": 0.5809,
+      "step": 12450
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5590844059687965,
+      "learning_rate": 1.0431658525493498e-06,
+      "loss": 0.6313,
+      "step": 12451
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7188546965687077,
+      "learning_rate": 1.0425322898467315e-06,
+      "loss": 0.7071,
+      "step": 12452
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3208087091381397,
+      "learning_rate": 1.0418988972053162e-06,
+      "loss": 0.5832,
+      "step": 12453
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4450788880961118,
+      "learning_rate": 1.0412656746523182e-06,
+      "loss": 0.7554,
+      "step": 12454
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3267575341332565,
+      "learning_rate": 1.0406326222149516e-06,
+      "loss": 0.5552,
+      "step": 12455
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2279480400017837,
+      "learning_rate": 1.0399997399204186e-06,
+      "loss": 0.5852,
+      "step": 12456
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4830632601973568,
+      "learning_rate": 1.0393670277959174e-06,
+      "loss": 0.6317,
+      "step": 12457
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3794782985774259,
+      "learning_rate": 1.038734485868635e-06,
+      "loss": 0.6401,
+      "step": 12458
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1433112024518117,
+      "learning_rate": 1.0381021141657526e-06,
+      "loss": 0.543,
+      "step": 12459
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.522895903085687,
+      "learning_rate": 1.037469912714449e-06,
+      "loss": 0.5205,
+      "step": 12460
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3969410111152942,
+      "learning_rate": 1.0368378815418856e-06,
+      "loss": 0.6723,
+      "step": 12461
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6131951803059696,
+      "learning_rate": 1.036206020675226e-06,
+      "loss": 0.6254,
+      "step": 12462
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3079163392070854,
+      "learning_rate": 1.0355743301416215e-06,
+      "loss": 0.6024,
+      "step": 12463
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3277069380281203,
+      "learning_rate": 1.0349428099682173e-06,
+      "loss": 0.6331,
+      "step": 12464
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1886779562314047,
+      "learning_rate": 1.0343114601821513e-06,
+      "loss": 0.6135,
+      "step": 12465
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5073219364100063,
+      "learning_rate": 1.033680280810554e-06,
+      "loss": 0.6674,
+      "step": 12466
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1004728013298117,
+      "learning_rate": 1.0330492718805469e-06,
+      "loss": 0.6866,
+      "step": 12467
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.37437516524938,
+      "learning_rate": 1.0324184334192505e-06,
+      "loss": 0.6196,
+      "step": 12468
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3139145313756462,
+      "learning_rate": 1.0317877654537672e-06,
+      "loss": 0.5646,
+      "step": 12469
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3989694896841902,
+      "learning_rate": 1.031157268011203e-06,
+      "loss": 0.5324,
+      "step": 12470
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2491092858344446,
+      "learning_rate": 1.0305269411186502e-06,
+      "loss": 0.5139,
+      "step": 12471
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3053459919483243,
+      "learning_rate": 1.0298967848031948e-06,
+      "loss": 0.716,
+      "step": 12472
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5949444934687318,
+      "learning_rate": 1.0292667990919164e-06,
+      "loss": 0.8614,
+      "step": 12473
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2706838150754018,
+      "learning_rate": 1.0286369840118859e-06,
+      "loss": 0.581,
+      "step": 12474
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3252790630338345,
+      "learning_rate": 1.0280073395901719e-06,
+      "loss": 0.6514,
+      "step": 12475
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3443714338282235,
+      "learning_rate": 1.0273778658538263e-06,
+      "loss": 0.6053,
+      "step": 12476
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.373604410707821,
+      "learning_rate": 1.0267485628299007e-06,
+      "loss": 0.5542,
+      "step": 12477
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.390969110797595,
+      "learning_rate": 1.026119430545441e-06,
+      "loss": 0.6305,
+      "step": 12478
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6867695377049923,
+      "learning_rate": 1.025490469027477e-06,
+      "loss": 0.6425,
+      "step": 12479
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1370168808407486,
+      "learning_rate": 1.0248616783030403e-06,
+      "loss": 0.5353,
+      "step": 12480
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3982299527198343,
+      "learning_rate": 1.0242330583991507e-06,
+      "loss": 0.5707,
+      "step": 12481
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3286096306707178,
+      "learning_rate": 1.0236046093428204e-06,
+      "loss": 0.602,
+      "step": 12482
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4348578349128291,
+      "learning_rate": 1.0229763311610563e-06,
+      "loss": 0.593,
+      "step": 12483
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4666203627256404,
+      "learning_rate": 1.0223482238808557e-06,
+      "loss": 0.5937,
+      "step": 12484
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3096762803767101,
+      "learning_rate": 1.0217202875292115e-06,
+      "loss": 0.6353,
+      "step": 12485
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2789505667306198,
+      "learning_rate": 1.021092522133108e-06,
+      "loss": 0.6355,
+      "step": 12486
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5836331329360542,
+      "learning_rate": 1.0204649277195178e-06,
+      "loss": 0.7104,
+      "step": 12487
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0895786641234948,
+      "learning_rate": 1.0198375043154142e-06,
+      "loss": 0.6881,
+      "step": 12488
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1422994392969303,
+      "learning_rate": 1.0192102519477565e-06,
+      "loss": 0.5655,
+      "step": 12489
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3823043955498069,
+      "learning_rate": 1.0185831706435007e-06,
+      "loss": 0.5918,
+      "step": 12490
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3391566047914016,
+      "learning_rate": 1.0179562604295929e-06,
+      "loss": 0.7071,
+      "step": 12491
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3111961489441541,
+      "learning_rate": 1.0173295213329714e-06,
+      "loss": 0.6101,
+      "step": 12492
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4788052567475731,
+      "learning_rate": 1.0167029533805733e-06,
+      "loss": 0.6142,
+      "step": 12493
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5290463487030732,
+      "learning_rate": 1.016076556599318e-06,
+      "loss": 0.7268,
+      "step": 12494
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2818124565199958,
+      "learning_rate": 1.0154503310161269e-06,
+      "loss": 0.6554,
+      "step": 12495
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5880486285180524,
+      "learning_rate": 1.014824276657909e-06,
+      "loss": 0.6861,
+      "step": 12496
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3657946299190453,
+      "learning_rate": 1.0141983935515675e-06,
+      "loss": 0.6445,
+      "step": 12497
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.375531282996147,
+      "learning_rate": 1.0135726817239983e-06,
+      "loss": 0.7624,
+      "step": 12498
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3773592879677727,
+      "learning_rate": 1.0129471412020886e-06,
+      "loss": 0.6796,
+      "step": 12499
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3684188828846535,
+      "learning_rate": 1.0123217720127203e-06,
+      "loss": 0.7009,
+      "step": 12500
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3780011950744895,
+      "learning_rate": 1.011696574182766e-06,
+      "loss": 0.6065,
+      "step": 12501
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6080522922074738,
+      "learning_rate": 1.0110715477390915e-06,
+      "loss": 0.5678,
+      "step": 12502
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.294124635544276,
+      "learning_rate": 1.0104466927085577e-06,
+      "loss": 0.6239,
+      "step": 12503
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3396295997415142,
+      "learning_rate": 1.0098220091180145e-06,
+      "loss": 0.6883,
+      "step": 12504
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3347230846440352,
+      "learning_rate": 1.0091974969943064e-06,
+      "loss": 0.5728,
+      "step": 12505
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4881840012857184,
+      "learning_rate": 1.0085731563642697e-06,
+      "loss": 0.7883,
+      "step": 12506
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3727756949283803,
+      "learning_rate": 1.0079489872547338e-06,
+      "loss": 0.5545,
+      "step": 12507
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4300020291628117,
+      "learning_rate": 1.0073249896925208e-06,
+      "loss": 0.7264,
+      "step": 12508
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6460518108369533,
+      "learning_rate": 1.006701163704445e-06,
+      "loss": 0.7575,
+      "step": 12509
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2053094034821221,
+      "learning_rate": 1.0060775093173126e-06,
+      "loss": 0.5573,
+      "step": 12510
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.297657069761888,
+      "learning_rate": 1.005454026557927e-06,
+      "loss": 0.5866,
+      "step": 12511
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.355194653753395,
+      "learning_rate": 1.0048307154530763e-06,
+      "loss": 0.5711,
+      "step": 12512
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2727961540912653,
+      "learning_rate": 1.0042075760295477e-06,
+      "loss": 0.6603,
+      "step": 12513
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3914019915220113,
+      "learning_rate": 1.0035846083141193e-06,
+      "loss": 0.6354,
+      "step": 12514
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.388801022026038,
+      "learning_rate": 1.002961812333561e-06,
+      "loss": 0.6507,
+      "step": 12515
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5225221557164232,
+      "learning_rate": 1.0023391881146349e-06,
+      "loss": 0.6139,
+      "step": 12516
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3289379282425255,
+      "learning_rate": 1.0017167356840974e-06,
+      "loss": 0.5557,
+      "step": 12517
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4006874841135255,
+      "learning_rate": 1.0010944550686968e-06,
+      "loss": 0.6478,
+      "step": 12518
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0825924317938505,
+      "learning_rate": 1.0004723462951732e-06,
+      "loss": 0.5726,
+      "step": 12519
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3252436234334521,
+      "learning_rate": 9.99850409390259e-07,
+      "loss": 0.6777,
+      "step": 12520
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.511912425603714,
+      "learning_rate": 9.992286443806825e-07,
+      "loss": 0.6581,
+      "step": 12521
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2499209989290445,
+      "learning_rate": 9.986070512931616e-07,
+      "loss": 0.6569,
+      "step": 12522
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.7477966287937083,
+      "learning_rate": 9.97985630154407e-07,
+      "loss": 0.6249,
+      "step": 12523
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2129963664485635,
+      "learning_rate": 9.973643809911238e-07,
+      "loss": 0.5753,
+      "step": 12524
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2649200655892703,
+      "learning_rate": 9.967433038300067e-07,
+      "loss": 0.7492,
+      "step": 12525
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7431790648199432,
+      "learning_rate": 9.96122398697746e-07,
+      "loss": 0.5904,
+      "step": 12526
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.263382960161445,
+      "learning_rate": 9.955016656210226e-07,
+      "loss": 0.6163,
+      "step": 12527
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3923500135551052,
+      "learning_rate": 9.948811046265095e-07,
+      "loss": 0.75,
+      "step": 12528
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2193730832982341,
+      "learning_rate": 9.942607157408784e-07,
+      "loss": 0.7,
+      "step": 12529
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.197762752479131,
+      "learning_rate": 9.936404989907828e-07,
+      "loss": 0.5519,
+      "step": 12530
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1714239453891517,
+      "learning_rate": 9.930204544028787e-07,
+      "loss": 0.657,
+      "step": 12531
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3961561452739855,
+      "learning_rate": 9.924005820038102e-07,
+      "loss": 0.6217,
+      "step": 12532
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1307486939516702,
+      "learning_rate": 9.917808818202135e-07,
+      "loss": 0.6085,
+      "step": 12533
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1640679848572026,
+      "learning_rate": 9.911613538787196e-07,
+      "loss": 0.5584,
+      "step": 12534
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.16392751058135,
+      "learning_rate": 9.905419982059488e-07,
+      "loss": 0.5425,
+      "step": 12535
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3144733837616673,
+      "learning_rate": 9.899228148285206e-07,
+      "loss": 0.6569,
+      "step": 12536
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4548476883041586,
+      "learning_rate": 9.89303803773039e-07,
+      "loss": 0.6334,
+      "step": 12537
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2391207242445095,
+      "learning_rate": 9.886849650661035e-07,
+      "loss": 0.5979,
+      "step": 12538
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4419240248118035,
+      "learning_rate": 9.880662987343103e-07,
+      "loss": 0.5874,
+      "step": 12539
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4067582704747692,
+      "learning_rate": 9.87447804804243e-07,
+      "loss": 0.691,
+      "step": 12540
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3809951926966344,
+      "learning_rate": 9.868294833024805e-07,
+      "loss": 0.5571,
+      "step": 12541
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2157349455752877,
+      "learning_rate": 9.862113342555924e-07,
+      "loss": 0.5688,
+      "step": 12542
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4135280625697786,
+      "learning_rate": 9.855933576901428e-07,
+      "loss": 0.896,
+      "step": 12543
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.31170647858904,
+      "learning_rate": 9.849755536326866e-07,
+      "loss": 0.7963,
+      "step": 12544
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4093430892845664,
+      "learning_rate": 9.843579221097722e-07,
+      "loss": 0.5418,
+      "step": 12545
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2621480010529251,
+      "learning_rate": 9.837404631479419e-07,
+      "loss": 0.5965,
+      "step": 12546
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4850970306773328,
+      "learning_rate": 9.831231767737305e-07,
+      "loss": 0.6628,
+      "step": 12547
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4947009328897878,
+      "learning_rate": 9.82506063013659e-07,
+      "loss": 0.6756,
+      "step": 12548
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3467563153451465,
+      "learning_rate": 9.818891218942511e-07,
+      "loss": 0.6078,
+      "step": 12549
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4519691416329399,
+      "learning_rate": 9.812723534420165e-07,
+      "loss": 0.6755,
+      "step": 12550
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4926691369745129,
+      "learning_rate": 9.806557576834591e-07,
+      "loss": 0.5598,
+      "step": 12551
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.209388516200725,
+      "learning_rate": 9.800393346450754e-07,
+      "loss": 0.6233,
+      "step": 12552
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4654754673754051,
+      "learning_rate": 9.794230843533525e-07,
+      "loss": 0.5583,
+      "step": 12553
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.231401572744767,
+      "learning_rate": 9.78807006834777e-07,
+      "loss": 0.6214,
+      "step": 12554
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4199848930282626,
+      "learning_rate": 9.781911021158181e-07,
+      "loss": 0.669,
+      "step": 12555
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6434694716649174,
+      "learning_rate": 9.775753702229452e-07,
+      "loss": 0.5344,
+      "step": 12556
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2703809680537348,
+      "learning_rate": 9.769598111826166e-07,
+      "loss": 0.6153,
+      "step": 12557
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3281526030223392,
+      "learning_rate": 9.763444250212855e-07,
+      "loss": 0.5476,
+      "step": 12558
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4168106011327843,
+      "learning_rate": 9.757292117653955e-07,
+      "loss": 0.7221,
+      "step": 12559
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2847658355987663,
+      "learning_rate": 9.751141714413836e-07,
+      "loss": 0.6604,
+      "step": 12560
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0896091537582746,
+      "learning_rate": 9.7449930407568e-07,
+      "loss": 0.6965,
+      "step": 12561
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2136393438930901,
+      "learning_rate": 9.738846096947064e-07,
+      "loss": 0.6753,
+      "step": 12562
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.424669889873464,
+      "learning_rate": 9.732700883248763e-07,
+      "loss": 0.6888,
+      "step": 12563
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3127756009832294,
+      "learning_rate": 9.726557399925995e-07,
+      "loss": 0.5664,
+      "step": 12564
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4938048905395973,
+      "learning_rate": 9.72041564724277e-07,
+      "loss": 0.7093,
+      "step": 12565
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5524367098565695,
+      "learning_rate": 9.71427562546296e-07,
+      "loss": 0.6284,
+      "step": 12566
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5396984327153167,
+      "learning_rate": 9.708137334850465e-07,
+      "loss": 0.6968,
+      "step": 12567
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4362461547854593,
+      "learning_rate": 9.702000775669041e-07,
+      "loss": 0.6013,
+      "step": 12568
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.39133350493856,
+      "learning_rate": 9.695865948182392e-07,
+      "loss": 0.6101,
+      "step": 12569
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.923071697282676,
+      "learning_rate": 9.689732852654143e-07,
+      "loss": 0.6174,
+      "step": 12570
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2216950817233165,
+      "learning_rate": 9.683601489347843e-07,
+      "loss": 0.6534,
+      "step": 12571
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4135050833751437,
+      "learning_rate": 9.677471858526998e-07,
+      "loss": 0.5913,
+      "step": 12572
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.330099533658213,
+      "learning_rate": 9.671343960454965e-07,
+      "loss": 0.7937,
+      "step": 12573
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2999918364840313,
+      "learning_rate": 9.66521779539511e-07,
+      "loss": 0.694,
+      "step": 12574
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4733581211494984,
+      "learning_rate": 9.659093363610678e-07,
+      "loss": 0.5621,
+      "step": 12575
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2327696288678498,
+      "learning_rate": 9.652970665364847e-07,
+      "loss": 0.5651,
+      "step": 12576
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9986358354660444,
+      "learning_rate": 9.646849700920729e-07,
+      "loss": 0.5963,
+      "step": 12577
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3489070781835806,
+      "learning_rate": 9.640730470541343e-07,
+      "loss": 0.6299,
+      "step": 12578
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4493832841796912,
+      "learning_rate": 9.63461297448966e-07,
+      "loss": 0.6394,
+      "step": 12579
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2767715179506438,
+      "learning_rate": 9.628497213028553e-07,
+      "loss": 0.6672,
+      "step": 12580
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3635472625853329,
+      "learning_rate": 9.622383186420818e-07,
+      "loss": 0.593,
+      "step": 12581
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4700546455402435,
+      "learning_rate": 9.616270894929219e-07,
+      "loss": 0.6713,
+      "step": 12582
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.45680172528312,
+      "learning_rate": 9.610160338816404e-07,
+      "loss": 0.687,
+      "step": 12583
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.280890546214596,
+      "learning_rate": 9.604051518344948e-07,
+      "loss": 0.6898,
+      "step": 12584
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.776635167244844,
+      "learning_rate": 9.597944433777362e-07,
+      "loss": 0.733,
+      "step": 12585
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1176151929799958,
+      "learning_rate": 9.59183908537607e-07,
+      "loss": 0.58,
+      "step": 12586
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5168229642259545,
+      "learning_rate": 9.585735473403479e-07,
+      "loss": 0.8493,
+      "step": 12587
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.196735022101495,
+      "learning_rate": 9.579633598121824e-07,
+      "loss": 0.6056,
+      "step": 12588
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3668488951023345,
+      "learning_rate": 9.57353345979332e-07,
+      "loss": 0.6464,
+      "step": 12589
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.433544105072845,
+      "learning_rate": 9.567435058680146e-07,
+      "loss": 0.6205,
+      "step": 12590
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.142776899917553,
+      "learning_rate": 9.56133839504431e-07,
+      "loss": 0.7329,
+      "step": 12591
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.114830771213819,
+      "learning_rate": 9.55524346914784e-07,
+      "loss": 0.6423,
+      "step": 12592
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3846856686880533,
+      "learning_rate": 9.549150281252633e-07,
+      "loss": 0.6336,
+      "step": 12593
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2949199567319805,
+      "learning_rate": 9.543058831620528e-07,
+      "loss": 0.6977,
+      "step": 12594
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2127865362091124,
+      "learning_rate": 9.536969120513284e-07,
+      "loss": 0.7133,
+      "step": 12595
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4843195469542574,
+      "learning_rate": 9.530881148192578e-07,
+      "loss": 0.6498,
+      "step": 12596
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3314109823335707,
+      "learning_rate": 9.524794914920072e-07,
+      "loss": 0.6105,
+      "step": 12597
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1114903641772567,
+      "learning_rate": 9.51871042095725e-07,
+      "loss": 0.6056,
+      "step": 12598
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1045342820851092,
+      "learning_rate": 9.512627666565588e-07,
+      "loss": 0.6819,
+      "step": 12599
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5786156983602149,
+      "learning_rate": 9.506546652006504e-07,
+      "loss": 0.6531,
+      "step": 12600
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1808510233322325,
+      "learning_rate": 9.500467377541289e-07,
+      "loss": 0.56,
+      "step": 12601
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6478015305905807,
+      "learning_rate": 9.49438984343119e-07,
+      "loss": 0.6508,
+      "step": 12602
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1960069187393922,
+      "learning_rate": 9.48831404993737e-07,
+      "loss": 0.4643,
+      "step": 12603
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.36906669738475,
+      "learning_rate": 9.482239997320903e-07,
+      "loss": 0.6249,
+      "step": 12604
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3754501423340493,
+      "learning_rate": 9.476167685842852e-07,
+      "loss": 0.6634,
+      "step": 12605
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.515391168920032,
+      "learning_rate": 9.470097115764099e-07,
+      "loss": 0.6461,
+      "step": 12606
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.351000290153561,
+      "learning_rate": 9.464028287345551e-07,
+      "loss": 0.6736,
+      "step": 12607
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2732565408264067,
+      "learning_rate": 9.457961200847998e-07,
+      "loss": 0.5475,
+      "step": 12608
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3876813127997247,
+      "learning_rate": 9.451895856532117e-07,
+      "loss": 0.6607,
+      "step": 12609
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3944631009143442,
+      "learning_rate": 9.445832254658594e-07,
+      "loss": 0.6252,
+      "step": 12610
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2888474491984983,
+      "learning_rate": 9.439770395487974e-07,
+      "loss": 0.4714,
+      "step": 12611
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3521745011318804,
+      "learning_rate": 9.43371027928075e-07,
+      "loss": 0.5873,
+      "step": 12612
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4133933203587772,
+      "learning_rate": 9.427651906297347e-07,
+      "loss": 0.5757,
+      "step": 12613
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1806952759341172,
+      "learning_rate": 9.421595276798084e-07,
+      "loss": 0.6802,
+      "step": 12614
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5580000059350605,
+      "learning_rate": 9.415540391043276e-07,
+      "loss": 0.7147,
+      "step": 12615
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4323562098950877,
+      "learning_rate": 9.40948724929307e-07,
+      "loss": 0.6097,
+      "step": 12616
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4215992506462753,
+      "learning_rate": 9.403435851807579e-07,
+      "loss": 0.7305,
+      "step": 12617
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4205449642519152,
+      "learning_rate": 9.397386198846881e-07,
+      "loss": 0.6037,
+      "step": 12618
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4024722076946734,
+      "learning_rate": 9.39133829067092e-07,
+      "loss": 0.6236,
+      "step": 12619
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5122912523116763,
+      "learning_rate": 9.385292127539597e-07,
+      "loss": 0.6197,
+      "step": 12620
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3512343516373613,
+      "learning_rate": 9.379247709712725e-07,
+      "loss": 0.6206,
+      "step": 12621
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9741765764270961,
+      "learning_rate": 9.373205037450028e-07,
+      "loss": 0.6,
+      "step": 12622
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.456467028738514,
+      "learning_rate": 9.367164111011223e-07,
+      "loss": 0.7843,
+      "step": 12623
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4612878120779589,
+      "learning_rate": 9.361124930655841e-07,
+      "loss": 0.6188,
+      "step": 12624
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2039593898399485,
+      "learning_rate": 9.355087496643444e-07,
+      "loss": 0.6067,
+      "step": 12625
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1666863634248943,
+      "learning_rate": 9.349051809233472e-07,
+      "loss": 0.6342,
+      "step": 12626
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.532095002739339,
+      "learning_rate": 9.34301786868525e-07,
+      "loss": 0.7162,
+      "step": 12627
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3443169577346323,
+      "learning_rate": 9.336985675258109e-07,
+      "loss": 0.5868,
+      "step": 12628
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3840755377227467,
+      "learning_rate": 9.330955229211259e-07,
+      "loss": 0.5933,
+      "step": 12629
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.447361795084019,
+      "learning_rate": 9.324926530803835e-07,
+      "loss": 0.8318,
+      "step": 12630
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2779588525774848,
+      "learning_rate": 9.31889958029491e-07,
+      "loss": 0.6126,
+      "step": 12631
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3793616220441978,
+      "learning_rate": 9.312874377943454e-07,
+      "loss": 0.5913,
+      "step": 12632
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1676905684639576,
+      "learning_rate": 9.306850924008415e-07,
+      "loss": 0.6189,
+      "step": 12633
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3227782788200932,
+      "learning_rate": 9.300829218748625e-07,
+      "loss": 0.7881,
+      "step": 12634
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4840609039810044,
+      "learning_rate": 9.294809262422838e-07,
+      "loss": 0.6671,
+      "step": 12635
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5783608028658893,
+      "learning_rate": 9.288791055289759e-07,
+      "loss": 0.6362,
+      "step": 12636
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0659805683756565,
+      "learning_rate": 9.282774597607991e-07,
+      "loss": 0.6205,
+      "step": 12637
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1652055294745385,
+      "learning_rate": 9.276759889636084e-07,
+      "loss": 0.6784,
+      "step": 12638
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3945113876711321,
+      "learning_rate": 9.270746931632501e-07,
+      "loss": 0.707,
+      "step": 12639
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2728860101861992,
+      "learning_rate": 9.264735723855617e-07,
+      "loss": 0.6292,
+      "step": 12640
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2345090778607906,
+      "learning_rate": 9.258726266563789e-07,
+      "loss": 0.6713,
+      "step": 12641
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2887705710849235,
+      "learning_rate": 9.2527185600152e-07,
+      "loss": 0.5894,
+      "step": 12642
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4006122003444512,
+      "learning_rate": 9.246712604468061e-07,
+      "loss": 0.6607,
+      "step": 12643
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.349764115131035,
+      "learning_rate": 9.240708400180437e-07,
+      "loss": 0.6771,
+      "step": 12644
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4609567493253677,
+      "learning_rate": 9.234705947410355e-07,
+      "loss": 0.6814,
+      "step": 12645
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5763060743573467,
+      "learning_rate": 9.228705246415742e-07,
+      "loss": 0.6174,
+      "step": 12646
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4414042905994549,
+      "learning_rate": 9.222706297454459e-07,
+      "loss": 0.6578,
+      "step": 12647
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4248747287265822,
+      "learning_rate": 9.216709100784326e-07,
+      "loss": 0.6807,
+      "step": 12648
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.879910098062002,
+      "learning_rate": 9.210713656663023e-07,
+      "loss": 0.6082,
+      "step": 12649
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3425416412044928,
+      "learning_rate": 9.20471996534818e-07,
+      "loss": 0.681,
+      "step": 12650
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2771167869812807,
+      "learning_rate": 9.198728027097386e-07,
+      "loss": 0.5439,
+      "step": 12651
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4126270309222877,
+      "learning_rate": 9.192737842168126e-07,
+      "loss": 0.5601,
+      "step": 12652
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.778236085675598,
+      "learning_rate": 9.186749410817797e-07,
+      "loss": 0.6579,
+      "step": 12653
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5605336609699587,
+      "learning_rate": 9.180762733303745e-07,
+      "loss": 0.6936,
+      "step": 12654
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.337566135046257,
+      "learning_rate": 9.174777809883229e-07,
+      "loss": 0.6206,
+      "step": 12655
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2694648116190022,
+      "learning_rate": 9.168794640813428e-07,
+      "loss": 0.5654,
+      "step": 12656
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2249769977133633,
+      "learning_rate": 9.162813226351447e-07,
+      "loss": 0.5465,
+      "step": 12657
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2475483130975242,
+      "learning_rate": 9.156833566754347e-07,
+      "loss": 0.6049,
+      "step": 12658
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3907633985361383,
+      "learning_rate": 9.150855662279079e-07,
+      "loss": 0.5873,
+      "step": 12659
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1784181509881235,
+      "learning_rate": 9.144879513182498e-07,
+      "loss": 0.5345,
+      "step": 12660
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3161884744836907,
+      "learning_rate": 9.138905119721442e-07,
+      "loss": 0.6583,
+      "step": 12661
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3507767521965173,
+      "learning_rate": 9.13293248215264e-07,
+      "loss": 0.6597,
+      "step": 12662
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0451778413032085,
+      "learning_rate": 9.126961600732742e-07,
+      "loss": 0.6872,
+      "step": 12663
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9711717695496987,
+      "learning_rate": 9.120992475718333e-07,
+      "loss": 0.6178,
+      "step": 12664
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3565966795061097,
+      "learning_rate": 9.115025107365904e-07,
+      "loss": 0.5766,
+      "step": 12665
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4452830256664406,
+      "learning_rate": 9.109059495931932e-07,
+      "loss": 0.6457,
+      "step": 12666
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3075861800217718,
+      "learning_rate": 9.103095641672732e-07,
+      "loss": 0.5825,
+      "step": 12667
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3656162562310998,
+      "learning_rate": 9.097133544844577e-07,
+      "loss": 0.6792,
+      "step": 12668
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3496332474138357,
+      "learning_rate": 9.091173205703708e-07,
+      "loss": 0.739,
+      "step": 12669
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.107043740333823,
+      "learning_rate": 9.085214624506228e-07,
+      "loss": 0.517,
+      "step": 12670
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.370669879087982,
+      "learning_rate": 9.079257801508201e-07,
+      "loss": 0.5344,
+      "step": 12671
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2708257604227555,
+      "learning_rate": 9.0733027369656e-07,
+      "loss": 0.5012,
+      "step": 12672
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.431680371103871,
+      "learning_rate": 9.067349431134331e-07,
+      "loss": 0.6495,
+      "step": 12673
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.208817468666855,
+      "learning_rate": 9.061397884270217e-07,
+      "loss": 0.6476,
+      "step": 12674
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.236369776454015,
+      "learning_rate": 9.055448096628999e-07,
+      "loss": 0.5863,
+      "step": 12675
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.515498273301324,
+      "learning_rate": 9.049500068466377e-07,
+      "loss": 0.5924,
+      "step": 12676
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2250306926535857,
+      "learning_rate": 9.043553800037952e-07,
+      "loss": 0.6134,
+      "step": 12677
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2667732856654657,
+      "learning_rate": 9.037609291599214e-07,
+      "loss": 0.512,
+      "step": 12678
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.105028835760959,
+      "learning_rate": 9.031666543405637e-07,
+      "loss": 0.6549,
+      "step": 12679
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3870108525489002,
+      "learning_rate": 9.025725555712595e-07,
+      "loss": 0.6199,
+      "step": 12680
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6184312658759328,
+      "learning_rate": 9.019786328775382e-07,
+      "loss": 0.7337,
+      "step": 12681
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3335064065340008,
+      "learning_rate": 9.013848862849217e-07,
+      "loss": 0.6985,
+      "step": 12682
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.38175472909559,
+      "learning_rate": 9.007913158189236e-07,
+      "loss": 0.5624,
+      "step": 12683
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4381600762137094,
+      "learning_rate": 9.001979215050544e-07,
+      "loss": 0.5965,
+      "step": 12684
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8691332069544686,
+      "learning_rate": 8.996047033688083e-07,
+      "loss": 0.7847,
+      "step": 12685
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.979059757710423,
+      "learning_rate": 8.990116614356819e-07,
+      "loss": 0.6202,
+      "step": 12686
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9798554816060111,
+      "learning_rate": 8.984187957311579e-07,
+      "loss": 0.6867,
+      "step": 12687
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1937489300246162,
+      "learning_rate": 8.978261062807131e-07,
+      "loss": 0.6359,
+      "step": 12688
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9970229638628568,
+      "learning_rate": 8.972335931098159e-07,
+      "loss": 0.5018,
+      "step": 12689
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2497631407830274,
+      "learning_rate": 8.966412562439291e-07,
+      "loss": 0.6253,
+      "step": 12690
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3880093808511929,
+      "learning_rate": 8.960490957085061e-07,
+      "loss": 0.6344,
+      "step": 12691
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5169675243588172,
+      "learning_rate": 8.954571115289934e-07,
+      "loss": 0.6878,
+      "step": 12692
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.458224958943358,
+      "learning_rate": 8.948653037308286e-07,
+      "loss": 0.5402,
+      "step": 12693
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.700533537813842,
+      "learning_rate": 8.942736723394458e-07,
+      "loss": 0.6756,
+      "step": 12694
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5947134053068015,
+      "learning_rate": 8.936822173802667e-07,
+      "loss": 0.6156,
+      "step": 12695
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.158097607660641,
+      "learning_rate": 8.930909388787084e-07,
+      "loss": 0.6933,
+      "step": 12696
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.004199992745808,
+      "learning_rate": 8.92499836860179e-07,
+      "loss": 0.6857,
+      "step": 12697
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5476978828904633,
+      "learning_rate": 8.919089113500795e-07,
+      "loss": 0.6745,
+      "step": 12698
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2049845842803424,
+      "learning_rate": 8.913181623738032e-07,
+      "loss": 0.6126,
+      "step": 12699
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2787630948339743,
+      "learning_rate": 8.907275899567363e-07,
+      "loss": 0.6276,
+      "step": 12700
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5282217791042627,
+      "learning_rate": 8.901371941242554e-07,
+      "loss": 0.718,
+      "step": 12701
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2903708381616403,
+      "learning_rate": 8.895469749017344e-07,
+      "loss": 0.6631,
+      "step": 12702
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3457238711212236,
+      "learning_rate": 8.889569323145325e-07,
+      "loss": 0.6527,
+      "step": 12703
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.702317730818421,
+      "learning_rate": 8.883670663880078e-07,
+      "loss": 0.6539,
+      "step": 12704
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.137636731952159,
+      "learning_rate": 8.877773771475074e-07,
+      "loss": 0.5295,
+      "step": 12705
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3986904144017689,
+      "learning_rate": 8.871878646183718e-07,
+      "loss": 0.7383,
+      "step": 12706
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5680243705789378,
+      "learning_rate": 8.865985288259332e-07,
+      "loss": 0.6006,
+      "step": 12707
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5301949689884413,
+      "learning_rate": 8.860093697955152e-07,
+      "loss": 0.66,
+      "step": 12708
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5441455514559468,
+      "learning_rate": 8.854203875524403e-07,
+      "loss": 0.7543,
+      "step": 12709
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3292444926697138,
+      "learning_rate": 8.848315821220133e-07,
+      "loss": 0.553,
+      "step": 12710
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3599343191736022,
+      "learning_rate": 8.842429535295366e-07,
+      "loss": 0.7249,
+      "step": 12711
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.272588429447854,
+      "learning_rate": 8.836545018003084e-07,
+      "loss": 0.6226,
+      "step": 12712
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7010560711039306,
+      "learning_rate": 8.830662269596135e-07,
+      "loss": 0.6632,
+      "step": 12713
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.310799086933922,
+      "learning_rate": 8.824781290327317e-07,
+      "loss": 0.5491,
+      "step": 12714
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3855780210197386,
+      "learning_rate": 8.818902080449348e-07,
+      "loss": 0.6984,
+      "step": 12715
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.582113187218003,
+      "learning_rate": 8.813024640214873e-07,
+      "loss": 0.6369,
+      "step": 12716
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5333459800784965,
+      "learning_rate": 8.807148969876455e-07,
+      "loss": 0.6941,
+      "step": 12717
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.164015888219261,
+      "learning_rate": 8.801275069686593e-07,
+      "loss": 0.6648,
+      "step": 12718
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7771520735072404,
+      "learning_rate": 8.795402939897679e-07,
+      "loss": 0.7252,
+      "step": 12719
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4643674420747272,
+      "learning_rate": 8.789532580762095e-07,
+      "loss": 0.5669,
+      "step": 12720
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4395903489615,
+      "learning_rate": 8.783663992532048e-07,
+      "loss": 0.5496,
+      "step": 12721
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1624895473950052,
+      "learning_rate": 8.777797175459773e-07,
+      "loss": 0.6644,
+      "step": 12722
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5909189086972688,
+      "learning_rate": 8.771932129797356e-07,
+      "loss": 0.6737,
+      "step": 12723
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1372834164234378,
+      "learning_rate": 8.766068855796833e-07,
+      "loss": 0.6161,
+      "step": 12724
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3814127418699287,
+      "learning_rate": 8.760207353710165e-07,
+      "loss": 0.7048,
+      "step": 12725
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3248460245811151,
+      "learning_rate": 8.754347623789222e-07,
+      "loss": 0.5795,
+      "step": 12726
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4451574496794404,
+      "learning_rate": 8.748489666285842e-07,
+      "loss": 0.7965,
+      "step": 12727
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3305077356136568,
+      "learning_rate": 8.742633481451728e-07,
+      "loss": 0.6288,
+      "step": 12728
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3686284778932678,
+      "learning_rate": 8.736779069538521e-07,
+      "loss": 0.7456,
+      "step": 12729
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2138469403443808,
+      "learning_rate": 8.730926430797826e-07,
+      "loss": 0.5262,
+      "step": 12730
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4217290109330174,
+      "learning_rate": 8.72507556548114e-07,
+      "loss": 0.7163,
+      "step": 12731
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4524271734942087,
+      "learning_rate": 8.719226473839876e-07,
+      "loss": 0.6696,
+      "step": 12732
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4088195166627924,
+      "learning_rate": 8.713379156125385e-07,
+      "loss": 0.5803,
+      "step": 12733
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.398711272754709,
+      "learning_rate": 8.707533612588948e-07,
+      "loss": 0.5884,
+      "step": 12734
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.369013107024179,
+      "learning_rate": 8.701689843481753e-07,
+      "loss": 0.6061,
+      "step": 12735
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5308813445089646,
+      "learning_rate": 8.695847849054906e-07,
+      "loss": 0.6788,
+      "step": 12736
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2143283738459052,
+      "learning_rate": 8.690007629559482e-07,
+      "loss": 0.726,
+      "step": 12737
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.303820844734063,
+      "learning_rate": 8.684169185246444e-07,
+      "loss": 0.611,
+      "step": 12738
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1100509881559624,
+      "learning_rate": 8.67833251636665e-07,
+      "loss": 0.7105,
+      "step": 12739
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5581889663455515,
+      "learning_rate": 8.672497623170944e-07,
+      "loss": 0.6389,
+      "step": 12740
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3893298770130185,
+      "learning_rate": 8.666664505910055e-07,
+      "loss": 0.6728,
+      "step": 12741
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5056528098531599,
+      "learning_rate": 8.660833164834653e-07,
+      "loss": 0.7205,
+      "step": 12742
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2994667115721221,
+      "learning_rate": 8.65500360019531e-07,
+      "loss": 0.7447,
+      "step": 12743
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6181479987623177,
+      "learning_rate": 8.649175812242532e-07,
+      "loss": 0.6497,
+      "step": 12744
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5791767887752683,
+      "learning_rate": 8.643349801226791e-07,
+      "loss": 0.6962,
+      "step": 12745
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3820131849550734,
+      "learning_rate": 8.637525567398392e-07,
+      "loss": 0.6144,
+      "step": 12746
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1499784003538827,
+      "learning_rate": 8.631703111007645e-07,
+      "loss": 0.6903,
+      "step": 12747
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3751238648857895,
+      "learning_rate": 8.625882432304749e-07,
+      "loss": 0.6039,
+      "step": 12748
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3354955222422906,
+      "learning_rate": 8.62006353153983e-07,
+      "loss": 0.5315,
+      "step": 12749
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.274870398874321,
+      "learning_rate": 8.61424640896294e-07,
+      "loss": 0.5261,
+      "step": 12750
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.364013658400622,
+      "learning_rate": 8.608431064824052e-07,
+      "loss": 0.6643,
+      "step": 12751
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4472318895317278,
+      "learning_rate": 8.602617499373056e-07,
+      "loss": 0.665,
+      "step": 12752
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2444768267672406,
+      "learning_rate": 8.596805712859807e-07,
+      "loss": 0.5812,
+      "step": 12753
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.33540070619519,
+      "learning_rate": 8.590995705533994e-07,
+      "loss": 0.6262,
+      "step": 12754
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9238025122862856,
+      "learning_rate": 8.58518747764534e-07,
+      "loss": 0.6977,
+      "step": 12755
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.41957786017432,
+      "learning_rate": 8.579381029443412e-07,
+      "loss": 0.6742,
+      "step": 12756
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3259314258538248,
+      "learning_rate": 8.573576361177733e-07,
+      "loss": 0.6684,
+      "step": 12757
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8339796324760589,
+      "learning_rate": 8.567773473097735e-07,
+      "loss": 0.7402,
+      "step": 12758
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5737335244742097,
+      "learning_rate": 8.561972365452775e-07,
+      "loss": 0.6921,
+      "step": 12759
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3153768191194248,
+      "learning_rate": 8.556173038492171e-07,
+      "loss": 0.6527,
+      "step": 12760
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2521017803954804,
+      "learning_rate": 8.550375492465102e-07,
+      "loss": 0.7513,
+      "step": 12761
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2201136730068185,
+      "learning_rate": 8.5445797276207e-07,
+      "loss": 0.5284,
+      "step": 12762
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.176559692287077,
+      "learning_rate": 8.538785744208062e-07,
+      "loss": 0.5008,
+      "step": 12763
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4211297560047205,
+      "learning_rate": 8.532993542476108e-07,
+      "loss": 0.6543,
+      "step": 12764
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5160062466856505,
+      "learning_rate": 8.527203122673789e-07,
+      "loss": 0.6345,
+      "step": 12765
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.428647083529527,
+      "learning_rate": 8.521414485049917e-07,
+      "loss": 0.6457,
+      "step": 12766
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6320931684853328,
+      "learning_rate": 8.51562762985324e-07,
+      "loss": 0.6689,
+      "step": 12767
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4193765616786749,
+      "learning_rate": 8.509842557332437e-07,
+      "loss": 0.602,
+      "step": 12768
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5642026616954892,
+      "learning_rate": 8.504059267736097e-07,
+      "loss": 0.715,
+      "step": 12769
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5084762511686403,
+      "learning_rate": 8.49827776131274e-07,
+      "loss": 0.594,
+      "step": 12770
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2712349016443025,
+      "learning_rate": 8.492498038310843e-07,
+      "loss": 0.5959,
+      "step": 12771
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1991284516039529,
+      "learning_rate": 8.486720098978718e-07,
+      "loss": 0.5657,
+      "step": 12772
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3043438348826035,
+      "learning_rate": 8.480943943564701e-07,
+      "loss": 0.5382,
+      "step": 12773
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2318157986317473,
+      "learning_rate": 8.475169572316988e-07,
+      "loss": 0.6172,
+      "step": 12774
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.294178033774777,
+      "learning_rate": 8.469396985483724e-07,
+      "loss": 0.7059,
+      "step": 12775
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3309651440256698,
+      "learning_rate": 8.463626183312962e-07,
+      "loss": 0.5436,
+      "step": 12776
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5190774659841209,
+      "learning_rate": 8.457857166052674e-07,
+      "loss": 0.6438,
+      "step": 12777
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4422394805677041,
+      "learning_rate": 8.452089933950813e-07,
+      "loss": 0.668,
+      "step": 12778
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2314158311921264,
+      "learning_rate": 8.446324487255164e-07,
+      "loss": 0.5928,
+      "step": 12779
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4361984080269823,
+      "learning_rate": 8.440560826213485e-07,
+      "loss": 0.6012,
+      "step": 12780
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3384743126469765,
+      "learning_rate": 8.434798951073492e-07,
+      "loss": 0.6532,
+      "step": 12781
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2608502732596836,
+      "learning_rate": 8.429038862082734e-07,
+      "loss": 0.5934,
+      "step": 12782
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.685279777027409,
+      "learning_rate": 8.423280559488767e-07,
+      "loss": 0.6158,
+      "step": 12783
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3017309586879502,
+      "learning_rate": 8.417524043539038e-07,
+      "loss": 0.6243,
+      "step": 12784
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.957318980154921,
+      "learning_rate": 8.411769314480905e-07,
+      "loss": 0.505,
+      "step": 12785
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1357253662444415,
+      "learning_rate": 8.406016372561665e-07,
+      "loss": 0.5806,
+      "step": 12786
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5476337617435392,
+      "learning_rate": 8.400265218028525e-07,
+      "loss": 0.5602,
+      "step": 12787
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.178559037172487,
+      "learning_rate": 8.394515851128654e-07,
+      "loss": 0.7322,
+      "step": 12788
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5213239860391918,
+      "learning_rate": 8.388768272109105e-07,
+      "loss": 0.6176,
+      "step": 12789
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.205535199859058,
+      "learning_rate": 8.383022481216829e-07,
+      "loss": 0.5738,
+      "step": 12790
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3189980284723613,
+      "learning_rate": 8.37727847869878e-07,
+      "loss": 0.6056,
+      "step": 12791
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5763122067020277,
+      "learning_rate": 8.371536264801772e-07,
+      "loss": 0.7005,
+      "step": 12792
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2869711615698707,
+      "learning_rate": 8.365795839772561e-07,
+      "loss": 0.6701,
+      "step": 12793
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2478602118887239,
+      "learning_rate": 8.36005720385783e-07,
+      "loss": 0.5517,
+      "step": 12794
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3022155916002263,
+      "learning_rate": 8.354320357304163e-07,
+      "loss": 0.6361,
+      "step": 12795
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2324161878431836,
+      "learning_rate": 8.34858530035813e-07,
+      "loss": 0.6358,
+      "step": 12796
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3357631105973808,
+      "learning_rate": 8.342852033266119e-07,
+      "loss": 0.6437,
+      "step": 12797
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2790015035103872,
+      "learning_rate": 8.337120556274548e-07,
+      "loss": 0.5706,
+      "step": 12798
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.395505575486007,
+      "learning_rate": 8.331390869629702e-07,
+      "loss": 0.7019,
+      "step": 12799
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4212225287004294,
+      "learning_rate": 8.32566297357777e-07,
+      "loss": 0.6793,
+      "step": 12800
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2171276571861316,
+      "learning_rate": 8.319936868364925e-07,
+      "loss": 0.5952,
+      "step": 12801
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.011730263650144,
+      "learning_rate": 8.314212554237222e-07,
+      "loss": 0.6746,
+      "step": 12802
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0618620624533444,
+      "learning_rate": 8.308490031440641e-07,
+      "loss": 0.6866,
+      "step": 12803
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.441195093758219,
+      "learning_rate": 8.302769300221098e-07,
+      "loss": 0.6374,
+      "step": 12804
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4854328802202197,
+      "learning_rate": 8.29705036082441e-07,
+      "loss": 0.6614,
+      "step": 12805
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4056896133604042,
+      "learning_rate": 8.291333213496355e-07,
+      "loss": 0.6151,
+      "step": 12806
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4852910580438474,
+      "learning_rate": 8.285617858482609e-07,
+      "loss": 0.6596,
+      "step": 12807
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4559981093910845,
+      "learning_rate": 8.279904296028757e-07,
+      "loss": 0.788,
+      "step": 12808
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4037311449618686,
+      "learning_rate": 8.274192526380337e-07,
+      "loss": 0.6601,
+      "step": 12809
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3802168222322766,
+      "learning_rate": 8.268482549782797e-07,
+      "loss": 0.6088,
+      "step": 12810
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3987384382421,
+      "learning_rate": 8.262774366481496e-07,
+      "loss": 0.6191,
+      "step": 12811
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0590271868319392,
+      "learning_rate": 8.257067976721739e-07,
+      "loss": 0.686,
+      "step": 12812
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.848237187311206,
+      "learning_rate": 8.251363380748723e-07,
+      "loss": 0.5321,
+      "step": 12813
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.833650512090858,
+      "learning_rate": 8.24566057880763e-07,
+      "loss": 0.6456,
+      "step": 12814
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4920645260673226,
+      "learning_rate": 8.239959571143463e-07,
+      "loss": 0.6771,
+      "step": 12815
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3617386698128862,
+      "learning_rate": 8.234260358001256e-07,
+      "loss": 0.5947,
+      "step": 12816
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1013263237964754,
+      "learning_rate": 8.228562939625906e-07,
+      "loss": 0.6847,
+      "step": 12817
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4473442278353832,
+      "learning_rate": 8.222867316262217e-07,
+      "loss": 0.6842,
+      "step": 12818
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8627872494065632,
+      "learning_rate": 8.217173488154972e-07,
+      "loss": 0.6726,
+      "step": 12819
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4958349331955165,
+      "learning_rate": 8.211481455548837e-07,
+      "loss": 0.6897,
+      "step": 12820
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.532823324718162,
+      "learning_rate": 8.205791218688414e-07,
+      "loss": 0.7099,
+      "step": 12821
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9241930353451845,
+      "learning_rate": 8.20010277781822e-07,
+      "loss": 0.7143,
+      "step": 12822
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2565265890788777,
+      "learning_rate": 8.194416133182693e-07,
+      "loss": 0.6041,
+      "step": 12823
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4393383370306545,
+      "learning_rate": 8.188731285026219e-07,
+      "loss": 0.7277,
+      "step": 12824
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2703945598323172,
+      "learning_rate": 8.183048233593083e-07,
+      "loss": 0.5585,
+      "step": 12825
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4272801125825227,
+      "learning_rate": 8.177366979127499e-07,
+      "loss": 0.6493,
+      "step": 12826
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3465355343683258,
+      "learning_rate": 8.171687521873595e-07,
+      "loss": 0.5635,
+      "step": 12827
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4543939519922608,
+      "learning_rate": 8.166009862075435e-07,
+      "loss": 0.7067,
+      "step": 12828
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0846480407369357,
+      "learning_rate": 8.160333999977004e-07,
+      "loss": 0.6835,
+      "step": 12829
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2878699020145974,
+      "learning_rate": 8.154659935822201e-07,
+      "loss": 0.6805,
+      "step": 12830
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3889885903027859,
+      "learning_rate": 8.148987669854846e-07,
+      "loss": 0.5903,
+      "step": 12831
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.155713931906065,
+      "learning_rate": 8.143317202318718e-07,
+      "loss": 0.6413,
+      "step": 12832
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4696122756131884,
+      "learning_rate": 8.137648533457448e-07,
+      "loss": 0.6512,
+      "step": 12833
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9414559286184296,
+      "learning_rate": 8.131981663514665e-07,
+      "loss": 0.7533,
+      "step": 12834
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.682221413392598,
+      "learning_rate": 8.126316592733874e-07,
+      "loss": 0.775,
+      "step": 12835
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.32451787986107,
+      "learning_rate": 8.120653321358519e-07,
+      "loss": 0.5664,
+      "step": 12836
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3369238824322873,
+      "learning_rate": 8.114991849631953e-07,
+      "loss": 0.6314,
+      "step": 12837
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.493136131733075,
+      "learning_rate": 8.109332177797469e-07,
+      "loss": 0.6817,
+      "step": 12838
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6216379821896567,
+      "learning_rate": 8.103674306098291e-07,
+      "loss": 0.6678,
+      "step": 12839
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3619808049730808,
+      "learning_rate": 8.098018234777533e-07,
+      "loss": 0.5697,
+      "step": 12840
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2701253572026199,
+      "learning_rate": 8.092363964078237e-07,
+      "loss": 0.687,
+      "step": 12841
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5042669170813217,
+      "learning_rate": 8.086711494243405e-07,
+      "loss": 0.5517,
+      "step": 12842
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.37421709985203,
+      "learning_rate": 8.081060825515924e-07,
+      "loss": 0.5316,
+      "step": 12843
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4837064488136285,
+      "learning_rate": 8.075411958138623e-07,
+      "loss": 0.7637,
+      "step": 12844
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2827752404703743,
+      "learning_rate": 8.069764892354237e-07,
+      "loss": 0.5949,
+      "step": 12845
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5821169579916319,
+      "learning_rate": 8.064119628405443e-07,
+      "loss": 0.709,
+      "step": 12846
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5559896491061238,
+      "learning_rate": 8.058476166534818e-07,
+      "loss": 0.6331,
+      "step": 12847
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3906354040480182,
+      "learning_rate": 8.052834506984874e-07,
+      "loss": 0.6764,
+      "step": 12848
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7540853956872016,
+      "learning_rate": 8.047194649998063e-07,
+      "loss": 0.6039,
+      "step": 12849
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9988128320467631,
+      "learning_rate": 8.041556595816741e-07,
+      "loss": 0.5355,
+      "step": 12850
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4005889783376173,
+      "learning_rate": 8.035920344683157e-07,
+      "loss": 0.5906,
+      "step": 12851
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5259172892773913,
+      "learning_rate": 8.030285896839546e-07,
+      "loss": 0.7334,
+      "step": 12852
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6137120246342882,
+      "learning_rate": 8.024653252528025e-07,
+      "loss": 0.7409,
+      "step": 12853
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4913885369102653,
+      "learning_rate": 8.019022411990634e-07,
+      "loss": 0.5723,
+      "step": 12854
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1636453199214374,
+      "learning_rate": 8.013393375469347e-07,
+      "loss": 0.6168,
+      "step": 12855
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2697370747692482,
+      "learning_rate": 8.007766143206047e-07,
+      "loss": 0.747,
+      "step": 12856
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2051528131360023,
+      "learning_rate": 8.00214071544258e-07,
+      "loss": 0.668,
+      "step": 12857
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3807131909992143,
+      "learning_rate": 7.996517092420636e-07,
+      "loss": 0.5754,
+      "step": 12858
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2912078757739374,
+      "learning_rate": 7.99089527438191e-07,
+      "loss": 0.5282,
+      "step": 12859
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3219250545746217,
+      "learning_rate": 7.985275261567971e-07,
+      "loss": 0.6106,
+      "step": 12860
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.215408616483263,
+      "learning_rate": 7.97965705422033e-07,
+      "loss": 0.6699,
+      "step": 12861
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3941576717095725,
+      "learning_rate": 7.974040652580412e-07,
+      "loss": 0.6112,
+      "step": 12862
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2386410068542844,
+      "learning_rate": 7.968426056889561e-07,
+      "loss": 0.5647,
+      "step": 12863
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4148070702183784,
+      "learning_rate": 7.962813267389052e-07,
+      "loss": 0.6488,
+      "step": 12864
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1612248909352851,
+      "learning_rate": 7.957202284320076e-07,
+      "loss": 0.6126,
+      "step": 12865
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8473157224094137,
+      "learning_rate": 7.951593107923744e-07,
+      "loss": 0.654,
+      "step": 12866
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4485712453434678,
+      "learning_rate": 7.945985738441114e-07,
+      "loss": 0.6152,
+      "step": 12867
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3107123441872632,
+      "learning_rate": 7.940380176113149e-07,
+      "loss": 0.5054,
+      "step": 12868
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3112516001883412,
+      "learning_rate": 7.93477642118069e-07,
+      "loss": 0.7117,
+      "step": 12869
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2854999055921135,
+      "learning_rate": 7.929174473884593e-07,
+      "loss": 0.6509,
+      "step": 12870
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2752811944666875,
+      "learning_rate": 7.923574334465562e-07,
+      "loss": 0.5884,
+      "step": 12871
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3610147226566944,
+      "learning_rate": 7.917976003164246e-07,
+      "loss": 0.6388,
+      "step": 12872
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5011242485229213,
+      "learning_rate": 7.912379480221228e-07,
+      "loss": 0.6555,
+      "step": 12873
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.296483973145632,
+      "learning_rate": 7.906784765876985e-07,
+      "loss": 0.6352,
+      "step": 12874
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3932930566003954,
+      "learning_rate": 7.90119186037197e-07,
+      "loss": 0.6044,
+      "step": 12875
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.211066459902788,
+      "learning_rate": 7.895600763946482e-07,
+      "loss": 0.5867,
+      "step": 12876
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2506662904249528,
+      "learning_rate": 7.890011476840809e-07,
+      "loss": 0.7258,
+      "step": 12877
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3279202314352065,
+      "learning_rate": 7.884423999295122e-07,
+      "loss": 0.6307,
+      "step": 12878
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.377435462178004,
+      "learning_rate": 7.878838331549538e-07,
+      "loss": 0.7909,
+      "step": 12879
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.396254463668209,
+      "learning_rate": 7.873254473844077e-07,
+      "loss": 0.5778,
+      "step": 12880
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2701175986010265,
+      "learning_rate": 7.867672426418693e-07,
+      "loss": 0.6352,
+      "step": 12881
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3353236717054295,
+      "learning_rate": 7.86209218951326e-07,
+      "loss": 0.6699,
+      "step": 12882
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1963671296801064,
+      "learning_rate": 7.856513763367569e-07,
+      "loss": 0.6083,
+      "step": 12883
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.508609910378613,
+      "learning_rate": 7.850937148221332e-07,
+      "loss": 0.6176,
+      "step": 12884
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5473932233430883,
+      "learning_rate": 7.845362344314206e-07,
+      "loss": 0.7251,
+      "step": 12885
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4383901144122533,
+      "learning_rate": 7.839789351885746e-07,
+      "loss": 0.6682,
+      "step": 12886
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5737495463741775,
+      "learning_rate": 7.834218171175428e-07,
+      "loss": 0.7148,
+      "step": 12887
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3223578864422485,
+      "learning_rate": 7.828648802422667e-07,
+      "loss": 0.5919,
+      "step": 12888
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.313951785185997,
+      "learning_rate": 7.823081245866776e-07,
+      "loss": 0.5249,
+      "step": 12889
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5560580929066627,
+      "learning_rate": 7.817515501747041e-07,
+      "loss": 0.669,
+      "step": 12890
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3588985259363346,
+      "learning_rate": 7.811951570302594e-07,
+      "loss": 0.6891,
+      "step": 12891
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.268122810325069,
+      "learning_rate": 7.806389451772539e-07,
+      "loss": 0.6851,
+      "step": 12892
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3842135354555558,
+      "learning_rate": 7.800829146395922e-07,
+      "loss": 0.6479,
+      "step": 12893
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3553558129717218,
+      "learning_rate": 7.795270654411635e-07,
+      "loss": 0.6063,
+      "step": 12894
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2584392548852465,
+      "learning_rate": 7.789713976058578e-07,
+      "loss": 0.6445,
+      "step": 12895
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5532754362540637,
+      "learning_rate": 7.784159111575512e-07,
+      "loss": 0.6434,
+      "step": 12896
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5205201463335802,
+      "learning_rate": 7.778606061201154e-07,
+      "loss": 0.6918,
+      "step": 12897
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3553711999297955,
+      "learning_rate": 7.773054825174126e-07,
+      "loss": 0.6521,
+      "step": 12898
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0684801317994466,
+      "learning_rate": 7.767505403732961e-07,
+      "loss": 0.5621,
+      "step": 12899
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6017031891650304,
+      "learning_rate": 7.761957797116171e-07,
+      "loss": 0.7638,
+      "step": 12900
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.814389796253832,
+      "learning_rate": 7.756412005562114e-07,
+      "loss": 0.5646,
+      "step": 12901
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.388449937424128,
+      "learning_rate": 7.750868029309099e-07,
+      "loss": 0.5784,
+      "step": 12902
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.379358932616933,
+      "learning_rate": 7.745325868595394e-07,
+      "loss": 0.5442,
+      "step": 12903
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2208097151803055,
+      "learning_rate": 7.739785523659144e-07,
+      "loss": 0.6095,
+      "step": 12904
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3933346637440231,
+      "learning_rate": 7.734246994738426e-07,
+      "loss": 0.6606,
+      "step": 12905
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5944459348347342,
+      "learning_rate": 7.728710282071244e-07,
+      "loss": 0.6842,
+      "step": 12906
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.232397682070249,
+      "learning_rate": 7.723175385895509e-07,
+      "loss": 0.5467,
+      "step": 12907
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5760503581912715,
+      "learning_rate": 7.717642306449113e-07,
+      "loss": 0.5791,
+      "step": 12908
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5311390347720397,
+      "learning_rate": 7.712111043969772e-07,
+      "loss": 0.7039,
+      "step": 12909
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7223893154952696,
+      "learning_rate": 7.706581598695207e-07,
+      "loss": 0.6578,
+      "step": 12910
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2326473515721768,
+      "learning_rate": 7.701053970863037e-07,
+      "loss": 0.5823,
+      "step": 12911
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3575200310833524,
+      "learning_rate": 7.695528160710764e-07,
+      "loss": 0.6211,
+      "step": 12912
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1796391394888144,
+      "learning_rate": 7.690004168475868e-07,
+      "loss": 0.6179,
+      "step": 12913
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3996329878020732,
+      "learning_rate": 7.684481994395726e-07,
+      "loss": 0.6544,
+      "step": 12914
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3218511615983723,
+      "learning_rate": 7.678961638707633e-07,
+      "loss": 0.6489,
+      "step": 12915
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.574854337351131,
+      "learning_rate": 7.673443101648808e-07,
+      "loss": 0.7333,
+      "step": 12916
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3534255465487444,
+      "learning_rate": 7.667926383456392e-07,
+      "loss": 0.6248,
+      "step": 12917
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2851708228253744,
+      "learning_rate": 7.662411484367482e-07,
+      "loss": 0.605,
+      "step": 12918
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.024867553784085,
+      "learning_rate": 7.656898404619029e-07,
+      "loss": 0.7036,
+      "step": 12919
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5282356758757063,
+      "learning_rate": 7.651387144447942e-07,
+      "loss": 0.6693,
+      "step": 12920
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1314453203705126,
+      "learning_rate": 7.645877704091082e-07,
+      "loss": 0.6237,
+      "step": 12921
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3816049255626341,
+      "learning_rate": 7.640370083785175e-07,
+      "loss": 0.6119,
+      "step": 12922
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3829538212052552,
+      "learning_rate": 7.634864283766913e-07,
+      "loss": 0.5567,
+      "step": 12923
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6068277950743486,
+      "learning_rate": 7.629360304272882e-07,
+      "loss": 0.6419,
+      "step": 12924
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.517436289417842,
+      "learning_rate": 7.623858145539592e-07,
+      "loss": 0.6186,
+      "step": 12925
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2801259133244265,
+      "learning_rate": 7.618357807803517e-07,
+      "loss": 0.6639,
+      "step": 12926
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2876152558271812,
+      "learning_rate": 7.612859291300973e-07,
+      "loss": 0.5356,
+      "step": 12927
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3598632466936882,
+      "learning_rate": 7.607362596268281e-07,
+      "loss": 0.5427,
+      "step": 12928
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2181354366223542,
+      "learning_rate": 7.601867722941642e-07,
+      "loss": 0.557,
+      "step": 12929
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2161335740614259,
+      "learning_rate": 7.596374671557144e-07,
+      "loss": 0.6587,
+      "step": 12930
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3556582981379317,
+      "learning_rate": 7.590883442350883e-07,
+      "loss": 0.5625,
+      "step": 12931
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.163129279978658,
+      "learning_rate": 7.585394035558807e-07,
+      "loss": 0.5172,
+      "step": 12932
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2397230673858375,
+      "learning_rate": 7.579906451416813e-07,
+      "loss": 0.7626,
+      "step": 12933
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3003676664667045,
+      "learning_rate": 7.57442069016071e-07,
+      "loss": 0.6539,
+      "step": 12934
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3850062754033596,
+      "learning_rate": 7.568936752026229e-07,
+      "loss": 0.6043,
+      "step": 12935
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2947801027031407,
+      "learning_rate": 7.563454637249056e-07,
+      "loss": 0.6538,
+      "step": 12936
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1769985841952177,
+      "learning_rate": 7.557974346064728e-07,
+      "loss": 0.6312,
+      "step": 12937
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4785303948524031,
+      "learning_rate": 7.552495878708776e-07,
+      "loss": 0.7342,
+      "step": 12938
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.349489436453907,
+      "learning_rate": 7.547019235416609e-07,
+      "loss": 0.6228,
+      "step": 12939
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5585112390359448,
+      "learning_rate": 7.541544416423574e-07,
+      "loss": 0.7295,
+      "step": 12940
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2612410383163735,
+      "learning_rate": 7.536071421964941e-07,
+      "loss": 0.5317,
+      "step": 12941
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8052014729837167,
+      "learning_rate": 7.530600252275888e-07,
+      "loss": 0.6271,
+      "step": 12942
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4147010989374307,
+      "learning_rate": 7.52513090759151e-07,
+      "loss": 0.6237,
+      "step": 12943
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5412611922818644,
+      "learning_rate": 7.519663388146886e-07,
+      "loss": 0.7247,
+      "step": 12944
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6039202376458386,
+      "learning_rate": 7.514197694176905e-07,
+      "loss": 0.6962,
+      "step": 12945
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3053300968145325,
+      "learning_rate": 7.50873382591648e-07,
+      "loss": 0.5755,
+      "step": 12946
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5748740650481383,
+      "learning_rate": 7.503271783600402e-07,
+      "loss": 0.7285,
+      "step": 12947
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4651522603272304,
+      "learning_rate": 7.497811567463381e-07,
+      "loss": 0.6621,
+      "step": 12948
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4608858043644504,
+      "learning_rate": 7.492353177740047e-07,
+      "loss": 0.519,
+      "step": 12949
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3132557318021432,
+      "learning_rate": 7.486896614664962e-07,
+      "loss": 0.6339,
+      "step": 12950
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5262170130302426,
+      "learning_rate": 7.481441878472629e-07,
+      "loss": 0.5971,
+      "step": 12951
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.513386858560018,
+      "learning_rate": 7.475988969397424e-07,
+      "loss": 0.6846,
+      "step": 12952
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.346231166130108,
+      "learning_rate": 7.470537887673667e-07,
+      "loss": 0.6736,
+      "step": 12953
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6413952650409618,
+      "learning_rate": 7.465088633535639e-07,
+      "loss": 0.6384,
+      "step": 12954
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3721532188907593,
+      "learning_rate": 7.459641207217461e-07,
+      "loss": 0.6345,
+      "step": 12955
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.235503560824842,
+      "learning_rate": 7.454195608953252e-07,
+      "loss": 0.7575,
+      "step": 12956
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4969026191893147,
+      "learning_rate": 7.448751838977014e-07,
+      "loss": 0.6049,
+      "step": 12957
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.554631409029764,
+      "learning_rate": 7.443309897522671e-07,
+      "loss": 0.7244,
+      "step": 12958
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2060780146909564,
+      "learning_rate": 7.437869784824086e-07,
+      "loss": 0.6845,
+      "step": 12959
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5086563201202579,
+      "learning_rate": 7.432431501115012e-07,
+      "loss": 0.6997,
+      "step": 12960
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0151380434831805,
+      "learning_rate": 7.426995046629176e-07,
+      "loss": 0.6195,
+      "step": 12961
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1569430052460603,
+      "learning_rate": 7.421560421600182e-07,
+      "loss": 0.5746,
+      "step": 12962
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0607611544625628,
+      "learning_rate": 7.416127626261549e-07,
+      "loss": 0.665,
+      "step": 12963
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.386037067877568,
+      "learning_rate": 7.410696660846761e-07,
+      "loss": 0.587,
+      "step": 12964
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2992822248853941,
+      "learning_rate": 7.405267525589183e-07,
+      "loss": 0.6057,
+      "step": 12965
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2008634677541523,
+      "learning_rate": 7.399840220722127e-07,
+      "loss": 0.5304,
+      "step": 12966
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3784120715004156,
+      "learning_rate": 7.394414746478817e-07,
+      "loss": 0.6297,
+      "step": 12967
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2293568306259617,
+      "learning_rate": 7.388991103092374e-07,
+      "loss": 0.5548,
+      "step": 12968
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4823546986745397,
+      "learning_rate": 7.383569290795911e-07,
+      "loss": 0.7464,
+      "step": 12969
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4377217972785157,
+      "learning_rate": 7.378149309822374e-07,
+      "loss": 0.7244,
+      "step": 12970
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2229395153281504,
+      "learning_rate": 7.372731160404672e-07,
+      "loss": 0.656,
+      "step": 12971
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.311375990571317,
+      "learning_rate": 7.367314842775674e-07,
+      "loss": 0.6086,
+      "step": 12972
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0657450374280701,
+      "learning_rate": 7.361900357168084e-07,
+      "loss": 0.6284,
+      "step": 12973
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4853289574344548,
+      "learning_rate": 7.356487703814602e-07,
+      "loss": 0.584,
+      "step": 12974
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4677335155668068,
+      "learning_rate": 7.351076882947817e-07,
+      "loss": 0.5554,
+      "step": 12975
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6505305886403145,
+      "learning_rate": 7.345667894800246e-07,
+      "loss": 0.5654,
+      "step": 12976
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.21334296720466,
+      "learning_rate": 7.340260739604316e-07,
+      "loss": 0.5691,
+      "step": 12977
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5459444502745583,
+      "learning_rate": 7.334855417592385e-07,
+      "loss": 0.7204,
+      "step": 12978
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.363176405908029,
+      "learning_rate": 7.329451928996745e-07,
+      "loss": 0.5473,
+      "step": 12979
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4134169747940586,
+      "learning_rate": 7.324050274049599e-07,
+      "loss": 0.5268,
+      "step": 12980
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4949191406445828,
+      "learning_rate": 7.318650452983039e-07,
+      "loss": 0.6322,
+      "step": 12981
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.210738636233558,
+      "learning_rate": 7.313252466029131e-07,
+      "loss": 0.5875,
+      "step": 12982
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.184213642936643,
+      "learning_rate": 7.307856313419837e-07,
+      "loss": 0.6055,
+      "step": 12983
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6966404468552787,
+      "learning_rate": 7.302461995387033e-07,
+      "loss": 0.7356,
+      "step": 12984
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.396427479852999,
+      "learning_rate": 7.297069512162535e-07,
+      "loss": 0.7156,
+      "step": 12985
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4024142420479424,
+      "learning_rate": 7.291678863978052e-07,
+      "loss": 0.6764,
+      "step": 12986
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.392273346531989,
+      "learning_rate": 7.286290051065264e-07,
+      "loss": 0.6394,
+      "step": 12987
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4719159725681672,
+      "learning_rate": 7.280903073655704e-07,
+      "loss": 0.5833,
+      "step": 12988
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5619351476762247,
+      "learning_rate": 7.275517931980886e-07,
+      "loss": 0.6706,
+      "step": 12989
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1312332402619072,
+      "learning_rate": 7.270134626272229e-07,
+      "loss": 0.621,
+      "step": 12990
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.192825422584066,
+      "learning_rate": 7.264753156761028e-07,
+      "loss": 0.5678,
+      "step": 12991
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4669464955615235,
+      "learning_rate": 7.25937352367857e-07,
+      "loss": 0.6215,
+      "step": 12992
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3728679677022626,
+      "learning_rate": 7.253995727256024e-07,
+      "loss": 0.6516,
+      "step": 12993
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.373707756081858,
+      "learning_rate": 7.24861976772448e-07,
+      "loss": 0.6645,
+      "step": 12994
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7182394817625306,
+      "learning_rate": 7.243245645314955e-07,
+      "loss": 0.5793,
+      "step": 12995
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7539920460603886,
+      "learning_rate": 7.23787336025838e-07,
+      "loss": 0.6822,
+      "step": 12996
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3478910928306682,
+      "learning_rate": 7.232502912785633e-07,
+      "loss": 0.5876,
+      "step": 12997
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2764570514059719,
+      "learning_rate": 7.227134303127487e-07,
+      "loss": 0.6115,
+      "step": 12998
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.510537157631811,
+      "learning_rate": 7.22176753151464e-07,
+      "loss": 0.6102,
+      "step": 12999
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3902745420471354,
+      "learning_rate": 7.216402598177713e-07,
+      "loss": 0.6446,
+      "step": 13000
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5218196975599028,
+      "learning_rate": 7.21103950334725e-07,
+      "loss": 0.6103,
+      "step": 13001
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.167469377701914,
+      "learning_rate": 7.205678247253717e-07,
+      "loss": 0.5952,
+      "step": 13002
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.222496943577557,
+      "learning_rate": 7.200318830127495e-07,
+      "loss": 0.6651,
+      "step": 13003
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3468797862539765,
+      "learning_rate": 7.194961252198885e-07,
+      "loss": 0.6701,
+      "step": 13004
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2325835244357861,
+      "learning_rate": 7.189605513698139e-07,
+      "loss": 0.612,
+      "step": 13005
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3464345975873595,
+      "learning_rate": 7.184251614855369e-07,
+      "loss": 0.5676,
+      "step": 13006
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6021022850335171,
+      "learning_rate": 7.178899555900675e-07,
+      "loss": 0.7504,
+      "step": 13007
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.282086596904974,
+      "learning_rate": 7.173549337064035e-07,
+      "loss": 0.6982,
+      "step": 13008
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.103393168101913,
+      "learning_rate": 7.168200958575361e-07,
+      "loss": 0.6463,
+      "step": 13009
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6989198804872923,
+      "learning_rate": 7.162854420664478e-07,
+      "loss": 0.7319,
+      "step": 13010
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3953188402109424,
+      "learning_rate": 7.157509723561134e-07,
+      "loss": 0.6466,
+      "step": 13011
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4541877183773266,
+      "learning_rate": 7.152166867495042e-07,
+      "loss": 0.6511,
+      "step": 13012
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1587344817690803,
+      "learning_rate": 7.146825852695749e-07,
+      "loss": 0.6587,
+      "step": 13013
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2125571742446872,
+      "learning_rate": 7.141486679392778e-07,
+      "loss": 0.6727,
+      "step": 13014
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4493793271811286,
+      "learning_rate": 7.136149347815591e-07,
+      "loss": 0.6821,
+      "step": 13015
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4497155012559895,
+      "learning_rate": 7.130813858193525e-07,
+      "loss": 0.5661,
+      "step": 13016
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4039187804161926,
+      "learning_rate": 7.125480210755858e-07,
+      "loss": 0.6355,
+      "step": 13017
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2652532668066132,
+      "learning_rate": 7.120148405731803e-07,
+      "loss": 0.6694,
+      "step": 13018
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.335361165186553,
+      "learning_rate": 7.114818443350463e-07,
+      "loss": 0.5799,
+      "step": 13019
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4946558990329883,
+      "learning_rate": 7.109490323840884e-07,
+      "loss": 0.6106,
+      "step": 13020
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2295385211103476,
+      "learning_rate": 7.104164047432022e-07,
+      "loss": 0.6672,
+      "step": 13021
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4295524112954845,
+      "learning_rate": 7.098839614352775e-07,
+      "loss": 0.6468,
+      "step": 13022
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2443400765340373,
+      "learning_rate": 7.093517024831941e-07,
+      "loss": 0.6814,
+      "step": 13023
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3384532168804502,
+      "learning_rate": 7.088196279098225e-07,
+      "loss": 0.5688,
+      "step": 13024
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5880672891050844,
+      "learning_rate": 7.082877377380293e-07,
+      "loss": 0.6827,
+      "step": 13025
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4373581798697381,
+      "learning_rate": 7.077560319906696e-07,
+      "loss": 0.6764,
+      "step": 13026
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2285588712409445,
+      "learning_rate": 7.072245106905928e-07,
+      "loss": 0.55,
+      "step": 13027
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0820336565280597,
+      "learning_rate": 7.066931738606397e-07,
+      "loss": 0.6576,
+      "step": 13028
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.298533771501467,
+      "learning_rate": 7.061620215236415e-07,
+      "loss": 0.6184,
+      "step": 13029
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2318514989434053,
+      "learning_rate": 7.056310537024257e-07,
+      "loss": 0.6333,
+      "step": 13030
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.340884087906346,
+      "learning_rate": 7.051002704198074e-07,
+      "loss": 0.6915,
+      "step": 13031
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5100755317827483,
+      "learning_rate": 7.045696716985939e-07,
+      "loss": 0.6336,
+      "step": 13032
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4185758288518695,
+      "learning_rate": 7.040392575615896e-07,
+      "loss": 0.6733,
+      "step": 13033
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4226156565566468,
+      "learning_rate": 7.035090280315854e-07,
+      "loss": 0.6232,
+      "step": 13034
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4333064730845197,
+      "learning_rate": 7.029789831313677e-07,
+      "loss": 0.5975,
+      "step": 13035
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2995347220408993,
+      "learning_rate": 7.024491228837127e-07,
+      "loss": 0.7408,
+      "step": 13036
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1944640089359777,
+      "learning_rate": 7.0191944731139e-07,
+      "loss": 0.7474,
+      "step": 13037
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.341694468821271,
+      "learning_rate": 7.013899564371613e-07,
+      "loss": 0.6971,
+      "step": 13038
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2930861870886796,
+      "learning_rate": 7.008606502837784e-07,
+      "loss": 0.562,
+      "step": 13039
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5013297598484496,
+      "learning_rate": 7.003315288739893e-07,
+      "loss": 0.5621,
+      "step": 13040
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.228353336490899,
+      "learning_rate": 6.998025922305313e-07,
+      "loss": 0.5973,
+      "step": 13041
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5493439118235455,
+      "learning_rate": 6.992738403761307e-07,
+      "loss": 0.5729,
+      "step": 13042
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4076225517245726,
+      "learning_rate": 6.987452733335126e-07,
+      "loss": 0.5838,
+      "step": 13043
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7608760331160496,
+      "learning_rate": 6.982168911253895e-07,
+      "loss": 0.6551,
+      "step": 13044
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3760054915193254,
+      "learning_rate": 6.97688693774467e-07,
+      "loss": 0.6841,
+      "step": 13045
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1096128261729579,
+      "learning_rate": 6.971606813034432e-07,
+      "loss": 0.6607,
+      "step": 13046
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4486406354283687,
+      "learning_rate": 6.96632853735007e-07,
+      "loss": 0.6442,
+      "step": 13047
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1469870450219386,
+      "learning_rate": 6.961052110918432e-07,
+      "loss": 0.5972,
+      "step": 13048
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5456487527742921,
+      "learning_rate": 6.955777533966212e-07,
+      "loss": 0.7071,
+      "step": 13049
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8941841272999578,
+      "learning_rate": 6.950504806720115e-07,
+      "loss": 0.6632,
+      "step": 13050
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.95660935524023,
+      "learning_rate": 6.945233929406697e-07,
+      "loss": 0.7032,
+      "step": 13051
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0452573597038164,
+      "learning_rate": 6.939964902252466e-07,
+      "loss": 0.6451,
+      "step": 13052
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3188985224455796,
+      "learning_rate": 6.934697725483846e-07,
+      "loss": 0.64,
+      "step": 13053
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6094702717096294,
+      "learning_rate": 6.929432399327174e-07,
+      "loss": 0.6627,
+      "step": 13054
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.7624650214565163,
+      "learning_rate": 6.924168924008712e-07,
+      "loss": 0.5328,
+      "step": 13055
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2992679312801014,
+      "learning_rate": 6.918907299754651e-07,
+      "loss": 0.576,
+      "step": 13056
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9158123628199467,
+      "learning_rate": 6.913647526791078e-07,
+      "loss": 0.6594,
+      "step": 13057
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2859060221884835,
+      "learning_rate": 6.90838960534404e-07,
+      "loss": 0.591,
+      "step": 13058
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1909278616277719,
+      "learning_rate": 6.903133535639467e-07,
+      "loss": 0.5498,
+      "step": 13059
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.621585623409378,
+      "learning_rate": 6.897879317903233e-07,
+      "loss": 0.7353,
+      "step": 13060
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2397257921249858,
+      "learning_rate": 6.89262695236112e-07,
+      "loss": 0.5955,
+      "step": 13061
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1911998675128608,
+      "learning_rate": 6.887376439238813e-07,
+      "loss": 0.502,
+      "step": 13062
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.447015778479088,
+      "learning_rate": 6.882127778761987e-07,
+      "loss": 0.6604,
+      "step": 13063
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4046894582984453,
+      "learning_rate": 6.876880971156147e-07,
+      "loss": 0.6389,
+      "step": 13064
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4649340325332734,
+      "learning_rate": 6.871636016646754e-07,
+      "loss": 0.7188,
+      "step": 13065
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5703755759617464,
+      "learning_rate": 6.866392915459241e-07,
+      "loss": 0.7685,
+      "step": 13066
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3219514032240112,
+      "learning_rate": 6.861151667818866e-07,
+      "loss": 0.5585,
+      "step": 13067
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8673419980013501,
+      "learning_rate": 6.855912273950893e-07,
+      "loss": 0.6388,
+      "step": 13068
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2289100470388472,
+      "learning_rate": 6.850674734080454e-07,
+      "loss": 0.7233,
+      "step": 13069
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3696996647032535,
+      "learning_rate": 6.845439048432617e-07,
+      "loss": 0.5964,
+      "step": 13070
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4255629260927636,
+      "learning_rate": 6.840205217232376e-07,
+      "loss": 0.6407,
+      "step": 13071
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2996804164598061,
+      "learning_rate": 6.834973240704629e-07,
+      "loss": 0.5576,
+      "step": 13072
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9513927798929631,
+      "learning_rate": 6.829743119074239e-07,
+      "loss": 0.6995,
+      "step": 13073
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8916321983809752,
+      "learning_rate": 6.824514852565922e-07,
+      "loss": 0.7122,
+      "step": 13074
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4886043140749834,
+      "learning_rate": 6.819288441404343e-07,
+      "loss": 0.6415,
+      "step": 13075
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3620663413077667,
+      "learning_rate": 6.814063885814127e-07,
+      "loss": 0.6248,
+      "step": 13076
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4359146209943088,
+      "learning_rate": 6.80884118601976e-07,
+      "loss": 0.698,
+      "step": 13077
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2475314027280957,
+      "learning_rate": 6.803620342245676e-07,
+      "loss": 0.5494,
+      "step": 13078
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2953215801355962,
+      "learning_rate": 6.798401354716233e-07,
+      "loss": 0.6483,
+      "step": 13079
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.189365855728883,
+      "learning_rate": 6.793184223655691e-07,
+      "loss": 0.5026,
+      "step": 13080
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5715406466134945,
+      "learning_rate": 6.78796894928827e-07,
+      "loss": 0.606,
+      "step": 13081
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2034432429158683,
+      "learning_rate": 6.782755531838047e-07,
+      "loss": 0.6194,
+      "step": 13082
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2667330091732165,
+      "learning_rate": 6.77754397152906e-07,
+      "loss": 0.5827,
+      "step": 13083
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4231225693586311,
+      "learning_rate": 6.772334268585296e-07,
+      "loss": 0.6297,
+      "step": 13084
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.155634830980524,
+      "learning_rate": 6.767126423230574e-07,
+      "loss": 0.6233,
+      "step": 13085
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.524891474778729,
+      "learning_rate": 6.76192043568873e-07,
+      "loss": 0.7193,
+      "step": 13086
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3793062804911609,
+      "learning_rate": 6.756716306183459e-07,
+      "loss": 0.6245,
+      "step": 13087
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4808753896630231,
+      "learning_rate": 6.751514034938395e-07,
+      "loss": 0.5848,
+      "step": 13088
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2268596839611492,
+      "learning_rate": 6.746313622177097e-07,
+      "loss": 0.6007,
+      "step": 13089
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0854483604732492,
+      "learning_rate": 6.741115068123017e-07,
+      "loss": 0.5009,
+      "step": 13090
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1761947632620091,
+      "learning_rate": 6.73591837299959e-07,
+      "loss": 0.5455,
+      "step": 13091
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.136854376573398,
+      "learning_rate": 6.730723537030092e-07,
+      "loss": 0.666,
+      "step": 13092
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5063948212150127,
+      "learning_rate": 6.725530560437759e-07,
+      "loss": 0.6735,
+      "step": 13093
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6698260604153004,
+      "learning_rate": 6.720339443445772e-07,
+      "loss": 0.5664,
+      "step": 13094
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2060709353997503,
+      "learning_rate": 6.71515018627718e-07,
+      "loss": 0.589,
+      "step": 13095
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6756647343200628,
+      "learning_rate": 6.709962789154989e-07,
+      "loss": 0.6363,
+      "step": 13096
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3179317975631166,
+      "learning_rate": 6.704777252302108e-07,
+      "loss": 0.6198,
+      "step": 13097
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3808035081646632,
+      "learning_rate": 6.699593575941365e-07,
+      "loss": 0.6388,
+      "step": 13098
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.539377655425996,
+      "learning_rate": 6.694411760295538e-07,
+      "loss": 0.6357,
+      "step": 13099
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2201992901191758,
+      "learning_rate": 6.689231805587265e-07,
+      "loss": 0.6251,
+      "step": 13100
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3485489493588019,
+      "learning_rate": 6.684053712039168e-07,
+      "loss": 0.6254,
+      "step": 13101
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.533770356687894,
+      "learning_rate": 6.678877479873769e-07,
+      "loss": 0.6849,
+      "step": 13102
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2128712211374237,
+      "learning_rate": 6.673703109313456e-07,
+      "loss": 0.5442,
+      "step": 13103
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.396771161584175,
+      "learning_rate": 6.66853060058063e-07,
+      "loss": 0.5908,
+      "step": 13104
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3510977125804053,
+      "learning_rate": 6.663359953897547e-07,
+      "loss": 0.6245,
+      "step": 13105
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2172801178056247,
+      "learning_rate": 6.658191169486405e-07,
+      "loss": 0.6171,
+      "step": 13106
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2667616446380439,
+      "learning_rate": 6.653024247569312e-07,
+      "loss": 0.6323,
+      "step": 13107
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2664247295060218,
+      "learning_rate": 6.647859188368294e-07,
+      "loss": 0.7307,
+      "step": 13108
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4448602010177702,
+      "learning_rate": 6.642695992105347e-07,
+      "loss": 0.6968,
+      "step": 13109
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3389604124076693,
+      "learning_rate": 6.637534659002287e-07,
+      "loss": 0.6479,
+      "step": 13110
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.314427762028079,
+      "learning_rate": 6.632375189280948e-07,
+      "loss": 0.6902,
+      "step": 13111
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.026840130376221,
+      "learning_rate": 6.627217583163031e-07,
+      "loss": 0.697,
+      "step": 13112
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.589504046244746,
+      "learning_rate": 6.622061840870169e-07,
+      "loss": 0.735,
+      "step": 13113
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4052750422370366,
+      "learning_rate": 6.61690796262392e-07,
+      "loss": 0.5981,
+      "step": 13114
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3036197626441672,
+      "learning_rate": 6.61175594864576e-07,
+      "loss": 0.7099,
+      "step": 13115
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.471610047535004,
+      "learning_rate": 6.606605799157062e-07,
+      "loss": 0.5647,
+      "step": 13116
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.467711726339915,
+      "learning_rate": 6.601457514379173e-07,
+      "loss": 0.5913,
+      "step": 13117
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3019647436234028,
+      "learning_rate": 6.596311094533292e-07,
+      "loss": 0.5863,
+      "step": 13118
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2987949742990887,
+      "learning_rate": 6.591166539840599e-07,
+      "loss": 0.5114,
+      "step": 13119
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3958020284551382,
+      "learning_rate": 6.586023850522156e-07,
+      "loss": 0.6615,
+      "step": 13120
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2052774421456933,
+      "learning_rate": 6.580883026798957e-07,
+      "loss": 0.6039,
+      "step": 13121
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2334071795720867,
+      "learning_rate": 6.575744068891909e-07,
+      "loss": 0.5504,
+      "step": 13122
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2809914616596985,
+      "learning_rate": 6.570606977021842e-07,
+      "loss": 0.589,
+      "step": 13123
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7142442582455368,
+      "learning_rate": 6.565471751409541e-07,
+      "loss": 0.616,
+      "step": 13124
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3899000028368065,
+      "learning_rate": 6.56033839227564e-07,
+      "loss": 0.562,
+      "step": 13125
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.568981528393383,
+      "learning_rate": 6.555206899840727e-07,
+      "loss": 0.628,
+      "step": 13126
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.487562691308831,
+      "learning_rate": 6.550077274325345e-07,
+      "loss": 0.6453,
+      "step": 13127
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3839229111452056,
+      "learning_rate": 6.544949515949911e-07,
+      "loss": 0.5808,
+      "step": 13128
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2732846894337138,
+      "learning_rate": 6.539823624934777e-07,
+      "loss": 0.6088,
+      "step": 13129
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2283644504956563,
+      "learning_rate": 6.53469960150021e-07,
+      "loss": 0.5434,
+      "step": 13130
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6247427733333684,
+      "learning_rate": 6.529577445866409e-07,
+      "loss": 0.6611,
+      "step": 13131
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1176271881273325,
+      "learning_rate": 6.524457158253472e-07,
+      "loss": 0.5991,
+      "step": 13132
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2865866394934513,
+      "learning_rate": 6.519338738881442e-07,
+      "loss": 0.5534,
+      "step": 13133
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3187051348697905,
+      "learning_rate": 6.514222187970248e-07,
+      "loss": 0.5986,
+      "step": 13134
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3418794868050088,
+      "learning_rate": 6.509107505739798e-07,
+      "loss": 0.686,
+      "step": 13135
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3020529631580768,
+      "learning_rate": 6.503994692409838e-07,
+      "loss": 0.5154,
+      "step": 13136
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4432417009303415,
+      "learning_rate": 6.498883748200102e-07,
+      "loss": 0.6785,
+      "step": 13137
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.678876344162079,
+      "learning_rate": 6.493774673330217e-07,
+      "loss": 0.68,
+      "step": 13138
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4078575036393068,
+      "learning_rate": 6.488667468019727e-07,
+      "loss": 0.646,
+      "step": 13139
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7152451786512874,
+      "learning_rate": 6.483562132488103e-07,
+      "loss": 0.6999,
+      "step": 13140
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3290234698940988,
+      "learning_rate": 6.478458666954718e-07,
+      "loss": 0.673,
+      "step": 13141
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3750142110199237,
+      "learning_rate": 6.473357071638914e-07,
+      "loss": 0.6914,
+      "step": 13142
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3161740113499962,
+      "learning_rate": 6.468257346759888e-07,
+      "loss": 0.6467,
+      "step": 13143
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3533174389214317,
+      "learning_rate": 6.46315949253678e-07,
+      "loss": 0.6105,
+      "step": 13144
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9574345217712181,
+      "learning_rate": 6.458063509188678e-07,
+      "loss": 0.5981,
+      "step": 13145
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5829121629064395,
+      "learning_rate": 6.452969396934567e-07,
+      "loss": 0.6743,
+      "step": 13146
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.417662097192436,
+      "learning_rate": 6.447877155993343e-07,
+      "loss": 0.6605,
+      "step": 13147
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.211777612873131,
+      "learning_rate": 6.442786786583832e-07,
+      "loss": 0.5461,
+      "step": 13148
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5214315126964986,
+      "learning_rate": 6.437698288924777e-07,
+      "loss": 0.5792,
+      "step": 13149
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2993659507836646,
+      "learning_rate": 6.432611663234851e-07,
+      "loss": 0.6701,
+      "step": 13150
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6230087681185443,
+      "learning_rate": 6.427526909732617e-07,
+      "loss": 0.5776,
+      "step": 13151
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.290836326139727,
+      "learning_rate": 6.422444028636604e-07,
+      "loss": 0.6198,
+      "step": 13152
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5687257500175467,
+      "learning_rate": 6.417363020165235e-07,
+      "loss": 0.5995,
+      "step": 13153
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3165137915358234,
+      "learning_rate": 6.412283884536818e-07,
+      "loss": 0.5857,
+      "step": 13154
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4161530290723987,
+      "learning_rate": 6.407206621969648e-07,
+      "loss": 0.6374,
+      "step": 13155
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3694491470599046,
+      "learning_rate": 6.402131232681897e-07,
+      "loss": 0.6468,
+      "step": 13156
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8027069768272819,
+      "learning_rate": 6.397057716891669e-07,
+      "loss": 0.6679,
+      "step": 13157
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4993366992549582,
+      "learning_rate": 6.391986074816975e-07,
+      "loss": 0.6559,
+      "step": 13158
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3496004689684178,
+      "learning_rate": 6.38691630667575e-07,
+      "loss": 0.7062,
+      "step": 13159
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2328780261380936,
+      "learning_rate": 6.381848412685882e-07,
+      "loss": 0.6198,
+      "step": 13160
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2490116265346427,
+      "learning_rate": 6.376782393065111e-07,
+      "loss": 0.6557,
+      "step": 13161
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2712939093762459,
+      "learning_rate": 6.371718248031167e-07,
+      "loss": 0.6226,
+      "step": 13162
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2797019586302132,
+      "learning_rate": 6.36665597780165e-07,
+      "loss": 0.6133,
+      "step": 13163
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3974149580125519,
+      "learning_rate": 6.36159558259411e-07,
+      "loss": 0.7634,
+      "step": 13164
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4076204902949299,
+      "learning_rate": 6.35653706262599e-07,
+      "loss": 0.7245,
+      "step": 13165
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0650779572454618,
+      "learning_rate": 6.351480418114675e-07,
+      "loss": 0.5426,
+      "step": 13166
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9968306779132784,
+      "learning_rate": 6.346425649277454e-07,
+      "loss": 0.6164,
+      "step": 13167
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4958486479108608,
+      "learning_rate": 6.341372756331543e-07,
+      "loss": 0.6965,
+      "step": 13168
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2703857854154879,
+      "learning_rate": 6.336321739494072e-07,
+      "loss": 0.6309,
+      "step": 13169
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4986286077375037,
+      "learning_rate": 6.331272598982108e-07,
+      "loss": 0.5904,
+      "step": 13170
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2692186831936954,
+      "learning_rate": 6.326225335012615e-07,
+      "loss": 0.5778,
+      "step": 13171
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6450481205726797,
+      "learning_rate": 6.321179947802492e-07,
+      "loss": 0.6593,
+      "step": 13172
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.386350385974485,
+      "learning_rate": 6.316136437568538e-07,
+      "loss": 0.6635,
+      "step": 13173
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7659112414834781,
+      "learning_rate": 6.31109480452749e-07,
+      "loss": 0.6135,
+      "step": 13174
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6036826684325232,
+      "learning_rate": 6.306055048896004e-07,
+      "loss": 0.6212,
+      "step": 13175
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.277992032412767,
+      "learning_rate": 6.301017170890639e-07,
+      "loss": 0.5947,
+      "step": 13176
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6778980527333214,
+      "learning_rate": 6.295981170727883e-07,
+      "loss": 0.7133,
+      "step": 13177
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.377297322737271,
+      "learning_rate": 6.290947048624169e-07,
+      "loss": 0.6202,
+      "step": 13178
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.383457633237227,
+      "learning_rate": 6.285914804795784e-07,
+      "loss": 0.5805,
+      "step": 13179
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2580020353899806,
+      "learning_rate": 6.280884439459006e-07,
+      "loss": 0.6298,
+      "step": 13180
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3561157903750038,
+      "learning_rate": 6.275855952829995e-07,
+      "loss": 0.5864,
+      "step": 13181
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6329983805347301,
+      "learning_rate": 6.270829345124829e-07,
+      "loss": 0.6402,
+      "step": 13182
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2017793539470014,
+      "learning_rate": 6.265804616559518e-07,
+      "loss": 0.6285,
+      "step": 13183
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5266790496535814,
+      "learning_rate": 6.260781767349983e-07,
+      "loss": 0.628,
+      "step": 13184
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6032790010446116,
+      "learning_rate": 6.255760797712068e-07,
+      "loss": 0.5984,
+      "step": 13185
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.648178742186351,
+      "learning_rate": 6.250741707861535e-07,
+      "loss": 0.6758,
+      "step": 13186
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4546529216467716,
+      "learning_rate": 6.245724498014055e-07,
+      "loss": 0.6551,
+      "step": 13187
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4151735020636889,
+      "learning_rate": 6.240709168385251e-07,
+      "loss": 0.5977,
+      "step": 13188
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3055856323164121,
+      "learning_rate": 6.235695719190632e-07,
+      "loss": 0.6241,
+      "step": 13189
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1724693877140782,
+      "learning_rate": 6.230684150645633e-07,
+      "loss": 0.6164,
+      "step": 13190
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8391195853878153,
+      "learning_rate": 6.22567446296562e-07,
+      "loss": 0.5936,
+      "step": 13191
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3827520223795315,
+      "learning_rate": 6.220666656365865e-07,
+      "loss": 0.6035,
+      "step": 13192
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1038301358465765,
+      "learning_rate": 6.215660731061563e-07,
+      "loss": 0.6466,
+      "step": 13193
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.553923563520024,
+      "learning_rate": 6.210656687267835e-07,
+      "loss": 0.6842,
+      "step": 13194
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2671452704275366,
+      "learning_rate": 6.2056545251997e-07,
+      "loss": 0.5753,
+      "step": 13195
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4867361219989046,
+      "learning_rate": 6.200654245072146e-07,
+      "loss": 0.6849,
+      "step": 13196
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.110235872264556,
+      "learning_rate": 6.195655847100008e-07,
+      "loss": 0.6413,
+      "step": 13197
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4472713426125083,
+      "learning_rate": 6.1906593314981e-07,
+      "loss": 0.6942,
+      "step": 13198
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3246359447327747,
+      "learning_rate": 6.185664698481137e-07,
+      "loss": 0.6364,
+      "step": 13199
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3759033598072432,
+      "learning_rate": 6.180671948263739e-07,
+      "loss": 0.669,
+      "step": 13200
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.321191042542672,
+      "learning_rate": 6.175681081060453e-07,
+      "loss": 0.6538,
+      "step": 13201
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2098552989155522,
+      "learning_rate": 6.170692097085751e-07,
+      "loss": 0.5451,
+      "step": 13202
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.412987340988643,
+      "learning_rate": 6.165704996554039e-07,
+      "loss": 0.6558,
+      "step": 13203
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3040009902638359,
+      "learning_rate": 6.160719779679597e-07,
+      "loss": 0.7601,
+      "step": 13204
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7649782585872447,
+      "learning_rate": 6.155736446676652e-07,
+      "loss": 0.6444,
+      "step": 13205
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6231939980615127,
+      "learning_rate": 6.150754997759367e-07,
+      "loss": 0.6967,
+      "step": 13206
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4794097274745426,
+      "learning_rate": 6.145775433141798e-07,
+      "loss": 0.6055,
+      "step": 13207
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2095081265263639,
+      "learning_rate": 6.140797753037925e-07,
+      "loss": 0.581,
+      "step": 13208
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6671089166891249,
+      "learning_rate": 6.135821957661658e-07,
+      "loss": 0.6798,
+      "step": 13209
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9749856145514704,
+      "learning_rate": 6.130848047226806e-07,
+      "loss": 0.6109,
+      "step": 13210
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6121505102750981,
+      "learning_rate": 6.125876021947119e-07,
+      "loss": 0.6328,
+      "step": 13211
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4586537492423108,
+      "learning_rate": 6.120905882036244e-07,
+      "loss": 0.7217,
+      "step": 13212
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2839937878314927,
+      "learning_rate": 6.115937627707775e-07,
+      "loss": 0.5996,
+      "step": 13213
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3151066389258106,
+      "learning_rate": 6.110971259175208e-07,
+      "loss": 0.6017,
+      "step": 13214
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4832273703513712,
+      "learning_rate": 6.106006776651934e-07,
+      "loss": 0.6108,
+      "step": 13215
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0430844431002315,
+      "learning_rate": 6.101044180351318e-07,
+      "loss": 0.6913,
+      "step": 13216
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.504992628961373,
+      "learning_rate": 6.096083470486602e-07,
+      "loss": 0.6872,
+      "step": 13217
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3418660783893515,
+      "learning_rate": 6.091124647270958e-07,
+      "loss": 0.543,
+      "step": 13218
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.8505451240556443,
+      "learning_rate": 6.086167710917479e-07,
+      "loss": 0.5925,
+      "step": 13219
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2901087904059576,
+      "learning_rate": 6.081212661639163e-07,
+      "loss": 0.5788,
+      "step": 13220
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5838747561470343,
+      "learning_rate": 6.076259499648968e-07,
+      "loss": 0.6509,
+      "step": 13221
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2559275833708403,
+      "learning_rate": 6.071308225159712e-07,
+      "loss": 0.5226,
+      "step": 13222
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3680531544627412,
+      "learning_rate": 6.066358838384184e-07,
+      "loss": 0.5609,
+      "step": 13223
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3444842717255168,
+      "learning_rate": 6.061411339535062e-07,
+      "loss": 0.6635,
+      "step": 13224
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3070543443862985,
+      "learning_rate": 6.056465728824951e-07,
+      "loss": 0.595,
+      "step": 13225
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4346745378039052,
+      "learning_rate": 6.051522006466376e-07,
+      "loss": 0.6122,
+      "step": 13226
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3524901635170732,
+      "learning_rate": 6.046580172671773e-07,
+      "loss": 0.6241,
+      "step": 13227
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3809184482993904,
+      "learning_rate": 6.041640227653517e-07,
+      "loss": 0.6259,
+      "step": 13228
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.280072137261918,
+      "learning_rate": 6.036702171623876e-07,
+      "loss": 0.6397,
+      "step": 13229
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2432097367774932,
+      "learning_rate": 6.031766004795047e-07,
+      "loss": 0.6475,
+      "step": 13230
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4813086238716853,
+      "learning_rate": 6.026831727379168e-07,
+      "loss": 0.6039,
+      "step": 13231
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2983535714916197,
+      "learning_rate": 6.021899339588266e-07,
+      "loss": 0.5471,
+      "step": 13232
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2656407680471473,
+      "learning_rate": 6.016968841634279e-07,
+      "loss": 0.6249,
+      "step": 13233
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1294442593009573,
+      "learning_rate": 6.012040233729105e-07,
+      "loss": 0.6179,
+      "step": 13234
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1680130771785915,
+      "learning_rate": 6.007113516084528e-07,
+      "loss": 0.6951,
+      "step": 13235
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.488818794324025,
+      "learning_rate": 6.002188688912264e-07,
+      "loss": 0.6319,
+      "step": 13236
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4482783367969545,
+      "learning_rate": 5.997265752423936e-07,
+      "loss": 0.4389,
+      "step": 13237
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5096963133224814,
+      "learning_rate": 5.992344706831088e-07,
+      "loss": 0.6416,
+      "step": 13238
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.41896315685679,
+      "learning_rate": 5.987425552345222e-07,
+      "loss": 0.6781,
+      "step": 13239
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1254739455463447,
+      "learning_rate": 5.98250828917768e-07,
+      "loss": 0.5626,
+      "step": 13240
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6342887244019062,
+      "learning_rate": 5.977592917539799e-07,
+      "loss": 0.5352,
+      "step": 13241
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4574181584810768,
+      "learning_rate": 5.972679437642793e-07,
+      "loss": 0.6737,
+      "step": 13242
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3782529689506262,
+      "learning_rate": 5.967767849697803e-07,
+      "loss": 0.576,
+      "step": 13243
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5732780735022056,
+      "learning_rate": 5.962858153915896e-07,
+      "loss": 0.6023,
+      "step": 13244
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.845321859127021,
+      "learning_rate": 5.957950350508052e-07,
+      "loss": 0.7285,
+      "step": 13245
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2457730351853582,
+      "learning_rate": 5.953044439685157e-07,
+      "loss": 0.4773,
+      "step": 13246
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3831721282065017,
+      "learning_rate": 5.948140421658061e-07,
+      "loss": 0.5786,
+      "step": 13247
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4230759758853617,
+      "learning_rate": 5.943238296637455e-07,
+      "loss": 0.5848,
+      "step": 13248
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5704852505196074,
+      "learning_rate": 5.938338064834037e-07,
+      "loss": 0.6628,
+      "step": 13249
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.358189835856215,
+      "learning_rate": 5.933439726458357e-07,
+      "loss": 0.5713,
+      "step": 13250
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3323889455778262,
+      "learning_rate": 5.928543281720917e-07,
+      "loss": 0.6135,
+      "step": 13251
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4737977743319552,
+      "learning_rate": 5.923648730832121e-07,
+      "loss": 0.5758,
+      "step": 13252
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3187893740077452,
+      "learning_rate": 5.918756074002291e-07,
+      "loss": 0.6568,
+      "step": 13253
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9687926076443358,
+      "learning_rate": 5.913865311441714e-07,
+      "loss": 0.5807,
+      "step": 13254
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2487609858179585,
+      "learning_rate": 5.908976443360515e-07,
+      "loss": 0.5306,
+      "step": 13255
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3941097292608198,
+      "learning_rate": 5.904089469968782e-07,
+      "loss": 0.6444,
+      "step": 13256
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.658434797674734,
+      "learning_rate": 5.899204391476554e-07,
+      "loss": 0.6957,
+      "step": 13257
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.44309800337657,
+      "learning_rate": 5.894321208093712e-07,
+      "loss": 0.8054,
+      "step": 13258
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2615126800496275,
+      "learning_rate": 5.889439920030127e-07,
+      "loss": 0.6606,
+      "step": 13259
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2090029243323324,
+      "learning_rate": 5.884560527495542e-07,
+      "loss": 0.6112,
+      "step": 13260
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4003490880795486,
+      "learning_rate": 5.879683030699645e-07,
+      "loss": 0.7035,
+      "step": 13261
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3255091350810417,
+      "learning_rate": 5.874807429852031e-07,
+      "loss": 0.6222,
+      "step": 13262
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4538555630159975,
+      "learning_rate": 5.869933725162203e-07,
+      "loss": 0.5863,
+      "step": 13263
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.037756401290896,
+      "learning_rate": 5.865061916839615e-07,
+      "loss": 0.6001,
+      "step": 13264
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7025996113572208,
+      "learning_rate": 5.860192005093624e-07,
+      "loss": 0.6645,
+      "step": 13265
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.9050898699981405,
+      "learning_rate": 5.855323990133465e-07,
+      "loss": 0.6012,
+      "step": 13266
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2790710509518701,
+      "learning_rate": 5.850457872168358e-07,
+      "loss": 0.5997,
+      "step": 13267
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4015803166268836,
+      "learning_rate": 5.845593651407405e-07,
+      "loss": 0.6891,
+      "step": 13268
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1277314194330754,
+      "learning_rate": 5.840731328059629e-07,
+      "loss": 0.5434,
+      "step": 13269
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1156193524967988,
+      "learning_rate": 5.835870902333974e-07,
+      "loss": 0.6202,
+      "step": 13270
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.288211964248736,
+      "learning_rate": 5.831012374439293e-07,
+      "loss": 0.7017,
+      "step": 13271
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.541545891578065,
+      "learning_rate": 5.826155744584405e-07,
+      "loss": 0.6308,
+      "step": 13272
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.528643519411154,
+      "learning_rate": 5.821301012977965e-07,
+      "loss": 0.5951,
+      "step": 13273
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4261104888355498,
+      "learning_rate": 5.816448179828616e-07,
+      "loss": 0.5907,
+      "step": 13274
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3924203295284576,
+      "learning_rate": 5.811597245344908e-07,
+      "loss": 0.6272,
+      "step": 13275
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4112540002266158,
+      "learning_rate": 5.806748209735252e-07,
+      "loss": 0.6117,
+      "step": 13276
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4331073326073616,
+      "learning_rate": 5.801901073208066e-07,
+      "loss": 0.5873,
+      "step": 13277
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4323453021777643,
+      "learning_rate": 5.797055835971621e-07,
+      "loss": 0.6226,
+      "step": 13278
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2226466351030483,
+      "learning_rate": 5.792212498234134e-07,
+      "loss": 0.5608,
+      "step": 13279
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.638326602155504,
+      "learning_rate": 5.787371060203728e-07,
+      "loss": 0.738,
+      "step": 13280
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1390687658037895,
+      "learning_rate": 5.782531522088452e-07,
+      "loss": 0.5431,
+      "step": 13281
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.439977011729855,
+      "learning_rate": 5.777693884096275e-07,
+      "loss": 0.6099,
+      "step": 13282
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5661533855224978,
+      "learning_rate": 5.772858146435095e-07,
+      "loss": 0.6195,
+      "step": 13283
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2644110436170566,
+      "learning_rate": 5.768024309312681e-07,
+      "loss": 0.5602,
+      "step": 13284
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.064564169995072,
+      "learning_rate": 5.763192372936782e-07,
+      "loss": 0.5292,
+      "step": 13285
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4784601223647325,
+      "learning_rate": 5.758362337515028e-07,
+      "loss": 0.7045,
+      "step": 13286
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5109500082179266,
+      "learning_rate": 5.753534203254973e-07,
+      "loss": 0.5984,
+      "step": 13287
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1811864689878422,
+      "learning_rate": 5.748707970364092e-07,
+      "loss": 0.5841,
+      "step": 13288
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8709872967304204,
+      "learning_rate": 5.74388363904978e-07,
+      "loss": 0.5925,
+      "step": 13289
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4642158176616946,
+      "learning_rate": 5.73906120951937e-07,
+      "loss": 0.658,
+      "step": 13290
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2289710912862748,
+      "learning_rate": 5.734240681980052e-07,
+      "loss": 0.5416,
+      "step": 13291
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4824699256651286,
+      "learning_rate": 5.729422056639006e-07,
+      "loss": 0.5537,
+      "step": 13292
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3207757402089308,
+      "learning_rate": 5.724605333703303e-07,
+      "loss": 0.7164,
+      "step": 13293
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1640684437177449,
+      "learning_rate": 5.719790513379891e-07,
+      "loss": 0.5687,
+      "step": 13294
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2806597030494404,
+      "learning_rate": 5.714977595875715e-07,
+      "loss": 0.6548,
+      "step": 13295
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.38964750776423,
+      "learning_rate": 5.710166581397575e-07,
+      "loss": 0.7623,
+      "step": 13296
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.386448713215043,
+      "learning_rate": 5.705357470152218e-07,
+      "loss": 0.5732,
+      "step": 13297
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1729479879307776,
+      "learning_rate": 5.700550262346294e-07,
+      "loss": 0.6655,
+      "step": 13298
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6662496054477214,
+      "learning_rate": 5.695744958186383e-07,
+      "loss": 0.6665,
+      "step": 13299
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.241911879252673,
+      "learning_rate": 5.690941557878988e-07,
+      "loss": 0.599,
+      "step": 13300
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1528290259637528,
+      "learning_rate": 5.686140061630519e-07,
+      "loss": 0.5422,
+      "step": 13301
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.691728552990506,
+      "learning_rate": 5.6813404696473e-07,
+      "loss": 0.6963,
+      "step": 13302
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0964413967738804,
+      "learning_rate": 5.676542782135585e-07,
+      "loss": 0.6719,
+      "step": 13303
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4513747093691791,
+      "learning_rate": 5.671746999301542e-07,
+      "loss": 0.699,
+      "step": 13304
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.417818210991784,
+      "learning_rate": 5.666953121351254e-07,
+      "loss": 0.6509,
+      "step": 13305
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7812009810927656,
+      "learning_rate": 5.662161148490725e-07,
+      "loss": 0.5479,
+      "step": 13306
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5483821420756505,
+      "learning_rate": 5.657371080925866e-07,
+      "loss": 0.6703,
+      "step": 13307
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.437956073356041,
+      "learning_rate": 5.652582918862548e-07,
+      "loss": 0.7011,
+      "step": 13308
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2313104275344102,
+      "learning_rate": 5.647796662506493e-07,
+      "loss": 0.5886,
+      "step": 13309
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.263081622801871,
+      "learning_rate": 5.6430123120634e-07,
+      "loss": 0.6018,
+      "step": 13310
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.680296268935506,
+      "learning_rate": 5.638229867738854e-07,
+      "loss": 0.6482,
+      "step": 13311
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4897814971019936,
+      "learning_rate": 5.633449329738372e-07,
+      "loss": 0.6406,
+      "step": 13312
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.39423936932078,
+      "learning_rate": 5.628670698267386e-07,
+      "loss": 0.6265,
+      "step": 13313
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.461466011204462,
+      "learning_rate": 5.623893973531225e-07,
+      "loss": 0.6748,
+      "step": 13314
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4506784006176148,
+      "learning_rate": 5.619119155735192e-07,
+      "loss": 0.6615,
+      "step": 13315
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0801033467667143,
+      "learning_rate": 5.61434624508444e-07,
+      "loss": 0.571,
+      "step": 13316
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.16545489037679,
+      "learning_rate": 5.609575241784071e-07,
+      "loss": 0.6553,
+      "step": 13317
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4124488457831903,
+      "learning_rate": 5.604806146039126e-07,
+      "loss": 0.6406,
+      "step": 13318
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.278916361363598,
+      "learning_rate": 5.600038958054538e-07,
+      "loss": 0.7001,
+      "step": 13319
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9775149819240518,
+      "learning_rate": 5.595273678035156e-07,
+      "loss": 0.6934,
+      "step": 13320
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2081871221938718,
+      "learning_rate": 5.590510306185765e-07,
+      "loss": 0.6039,
+      "step": 13321
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.31384842543826,
+      "learning_rate": 5.585748842711047e-07,
+      "loss": 0.6204,
+      "step": 13322
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4299911064802073,
+      "learning_rate": 5.580989287815619e-07,
+      "loss": 0.6854,
+      "step": 13323
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7322103483597395,
+      "learning_rate": 5.576231641703994e-07,
+      "loss": 0.6444,
+      "step": 13324
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3448179015133837,
+      "learning_rate": 5.571475904580648e-07,
+      "loss": 0.6672,
+      "step": 13325
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.440063239583501,
+      "learning_rate": 5.566722076649933e-07,
+      "loss": 0.8037,
+      "step": 13326
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3181231316393502,
+      "learning_rate": 5.561970158116114e-07,
+      "loss": 0.5586,
+      "step": 13327
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1893824037075775,
+      "learning_rate": 5.557220149183412e-07,
+      "loss": 0.5097,
+      "step": 13328
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7588550894722832,
+      "learning_rate": 5.552472050055946e-07,
+      "loss": 0.6257,
+      "step": 13329
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3818986303246976,
+      "learning_rate": 5.547725860937741e-07,
+      "loss": 0.6637,
+      "step": 13330
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3513592064077635,
+      "learning_rate": 5.542981582032759e-07,
+      "loss": 0.6278,
+      "step": 13331
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.122203609858971,
+      "learning_rate": 5.538239213544855e-07,
+      "loss": 0.6774,
+      "step": 13332
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4586076133172117,
+      "learning_rate": 5.533498755677857e-07,
+      "loss": 0.5941,
+      "step": 13333
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.302075926514356,
+      "learning_rate": 5.528760208635436e-07,
+      "loss": 0.8572,
+      "step": 13334
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3975761017195787,
+      "learning_rate": 5.524023572621229e-07,
+      "loss": 0.6255,
+      "step": 13335
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8133390966252845,
+      "learning_rate": 5.519288847838783e-07,
+      "loss": 0.6594,
+      "step": 13336
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4604654918749174,
+      "learning_rate": 5.514556034491569e-07,
+      "loss": 0.5823,
+      "step": 13337
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.405425742988293,
+      "learning_rate": 5.509825132782948e-07,
+      "loss": 0.55,
+      "step": 13338
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0899405952873216,
+      "learning_rate": 5.505096142916233e-07,
+      "loss": 0.6305,
+      "step": 13339
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5162061403496516,
+      "learning_rate": 5.50036906509463e-07,
+      "loss": 0.6387,
+      "step": 13340
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3724792839321938,
+      "learning_rate": 5.495643899521269e-07,
+      "loss": 0.7301,
+      "step": 13341
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5848268007959427,
+      "learning_rate": 5.4909206463992e-07,
+      "loss": 0.5625,
+      "step": 13342
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0567067593978492,
+      "learning_rate": 5.486199305931411e-07,
+      "loss": 0.6983,
+      "step": 13343
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0832927170996136,
+      "learning_rate": 5.481479878320784e-07,
+      "loss": 0.6252,
+      "step": 13344
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2119029100037202,
+      "learning_rate": 5.476762363770088e-07,
+      "loss": 0.6587,
+      "step": 13345
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2201782581926792,
+      "learning_rate": 5.472046762482092e-07,
+      "loss": 0.5116,
+      "step": 13346
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3579347684036482,
+      "learning_rate": 5.467333074659409e-07,
+      "loss": 0.5832,
+      "step": 13347
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6256899673296852,
+      "learning_rate": 5.462621300504606e-07,
+      "loss": 0.7594,
+      "step": 13348
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1944217094774878,
+      "learning_rate": 5.457911440220154e-07,
+      "loss": 0.6348,
+      "step": 13349
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6130941666493575,
+      "learning_rate": 5.453203494008435e-07,
+      "loss": 0.673,
+      "step": 13350
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5229511574803567,
+      "learning_rate": 5.4484974620718e-07,
+      "loss": 0.7114,
+      "step": 13351
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5423492142173776,
+      "learning_rate": 5.443793344612425e-07,
+      "loss": 0.7119,
+      "step": 13352
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2535883226566045,
+      "learning_rate": 5.439091141832498e-07,
+      "loss": 0.5761,
+      "step": 13353
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.274374612920526,
+      "learning_rate": 5.434390853934063e-07,
+      "loss": 0.6713,
+      "step": 13354
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.616346939914211,
+      "learning_rate": 5.429692481119103e-07,
+      "loss": 0.8033,
+      "step": 13355
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2966942284341374,
+      "learning_rate": 5.424996023589524e-07,
+      "loss": 0.6157,
+      "step": 13356
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5469748102792436,
+      "learning_rate": 5.420301481547141e-07,
+      "loss": 0.7855,
+      "step": 13357
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4998969863020666,
+      "learning_rate": 5.415608855193688e-07,
+      "loss": 0.4838,
+      "step": 13358
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3773556625072831,
+      "learning_rate": 5.410918144730815e-07,
+      "loss": 0.6029,
+      "step": 13359
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.452311403953539,
+      "learning_rate": 5.406229350360087e-07,
+      "loss": 0.651,
+      "step": 13360
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.305969706592551,
+      "learning_rate": 5.401542472283006e-07,
+      "loss": 0.59,
+      "step": 13361
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9248349868450738,
+      "learning_rate": 5.39685751070097e-07,
+      "loss": 0.6206,
+      "step": 13362
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.470403807955159,
+      "learning_rate": 5.392174465815308e-07,
+      "loss": 0.6138,
+      "step": 13363
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2739073428357879,
+      "learning_rate": 5.387493337827254e-07,
+      "loss": 0.6518,
+      "step": 13364
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.642635878795365,
+      "learning_rate": 5.382814126937969e-07,
+      "loss": 0.599,
+      "step": 13365
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3091550820404887,
+      "learning_rate": 5.378136833348529e-07,
+      "loss": 0.5714,
+      "step": 13366
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3454146160651337,
+      "learning_rate": 5.373461457259926e-07,
+      "loss": 0.6042,
+      "step": 13367
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7430518691904926,
+      "learning_rate": 5.368787998873065e-07,
+      "loss": 0.5539,
+      "step": 13368
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4133517672337357,
+      "learning_rate": 5.364116458388802e-07,
+      "loss": 0.6619,
+      "step": 13369
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2710033780084968,
+      "learning_rate": 5.359446836007842e-07,
+      "loss": 0.5764,
+      "step": 13370
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7183483420763284,
+      "learning_rate": 5.354779131930882e-07,
+      "loss": 0.7522,
+      "step": 13371
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6466093119574245,
+      "learning_rate": 5.350113346358493e-07,
+      "loss": 0.677,
+      "step": 13372
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2568948800625337,
+      "learning_rate": 5.345449479491177e-07,
+      "loss": 0.6829,
+      "step": 13373
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.64579937084712,
+      "learning_rate": 5.340787531529346e-07,
+      "loss": 0.6923,
+      "step": 13374
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.596130860858748,
+      "learning_rate": 5.336127502673328e-07,
+      "loss": 0.7014,
+      "step": 13375
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2099778958742546,
+      "learning_rate": 5.331469393123401e-07,
+      "loss": 0.6278,
+      "step": 13376
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3873627233101944,
+      "learning_rate": 5.326813203079706e-07,
+      "loss": 0.6587,
+      "step": 13377
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6070628160293274,
+      "learning_rate": 5.322158932742328e-07,
+      "loss": 0.6237,
+      "step": 13378
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3546787703855394,
+      "learning_rate": 5.3175065823113e-07,
+      "loss": 0.6262,
+      "step": 13379
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.046906940713927,
+      "learning_rate": 5.312856151986523e-07,
+      "loss": 0.5371,
+      "step": 13380
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3931538217497177,
+      "learning_rate": 5.308207641967839e-07,
+      "loss": 0.68,
+      "step": 13381
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3122683784333369,
+      "learning_rate": 5.303561052455009e-07,
+      "loss": 0.6046,
+      "step": 13382
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2041389727883829,
+      "learning_rate": 5.298916383647695e-07,
+      "loss": 0.6955,
+      "step": 13383
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4775841268354664,
+      "learning_rate": 5.294273635745517e-07,
+      "loss": 0.5237,
+      "step": 13384
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5614838756094986,
+      "learning_rate": 5.289632808947959e-07,
+      "loss": 0.5529,
+      "step": 13385
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3390359025755187,
+      "learning_rate": 5.284993903454444e-07,
+      "loss": 0.5925,
+      "step": 13386
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4865340537277167,
+      "learning_rate": 5.280356919464341e-07,
+      "loss": 0.6857,
+      "step": 13387
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2395511282153429,
+      "learning_rate": 5.275721857176885e-07,
+      "loss": 0.6913,
+      "step": 13388
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4065900541151328,
+      "learning_rate": 5.271088716791273e-07,
+      "loss": 0.6879,
+      "step": 13389
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3366377203394646,
+      "learning_rate": 5.266457498506594e-07,
+      "loss": 0.7047,
+      "step": 13390
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7994205388237727,
+      "learning_rate": 5.261828202521868e-07,
+      "loss": 0.5834,
+      "step": 13391
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.070002845008023,
+      "learning_rate": 5.257200829036018e-07,
+      "loss": 0.7737,
+      "step": 13392
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3670449188361962,
+      "learning_rate": 5.252575378247887e-07,
+      "loss": 0.6437,
+      "step": 13393
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4465631936578809,
+      "learning_rate": 5.24795185035627e-07,
+      "loss": 0.598,
+      "step": 13394
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4509573168767003,
+      "learning_rate": 5.243330245559819e-07,
+      "loss": 0.6235,
+      "step": 13395
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.416504206908053,
+      "learning_rate": 5.238710564057137e-07,
+      "loss": 0.6527,
+      "step": 13396
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3294934148598525,
+      "learning_rate": 5.234092806046759e-07,
+      "loss": 0.6449,
+      "step": 13397
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7226131288346238,
+      "learning_rate": 5.229476971727115e-07,
+      "loss": 0.7233,
+      "step": 13398
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.428436050095497,
+      "learning_rate": 5.224863061296553e-07,
+      "loss": 0.6417,
+      "step": 13399
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1114109250851514,
+      "learning_rate": 5.220251074953348e-07,
+      "loss": 0.5752,
+      "step": 13400
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5373833339111491,
+      "learning_rate": 5.215641012895667e-07,
+      "loss": 0.6827,
+      "step": 13401
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3527605009293875,
+      "learning_rate": 5.211032875321659e-07,
+      "loss": 0.4687,
+      "step": 13402
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7216636651365607,
+      "learning_rate": 5.206426662429298e-07,
+      "loss": 0.638,
+      "step": 13403
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.271907603214871,
+      "learning_rate": 5.201822374416549e-07,
+      "loss": 0.5474,
+      "step": 13404
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.167753231503788,
+      "learning_rate": 5.197220011481274e-07,
+      "loss": 0.7294,
+      "step": 13405
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2062491656650762,
+      "learning_rate": 5.192619573821223e-07,
+      "loss": 0.5675,
+      "step": 13406
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.523351793348239,
+      "learning_rate": 5.188021061634102e-07,
+      "loss": 0.5565,
+      "step": 13407
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.528959850645937,
+      "learning_rate": 5.183424475117521e-07,
+      "loss": 0.6469,
+      "step": 13408
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6468856356525885,
+      "learning_rate": 5.178829814469006e-07,
+      "loss": 0.7204,
+      "step": 13409
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9613944352735948,
+      "learning_rate": 5.174237079885991e-07,
+      "loss": 0.536,
+      "step": 13410
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6981124816234074,
+      "learning_rate": 5.169646271565831e-07,
+      "loss": 0.6256,
+      "step": 13411
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6108415087461663,
+      "learning_rate": 5.165057389705835e-07,
+      "loss": 0.736,
+      "step": 13412
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4007116194107814,
+      "learning_rate": 5.160470434503156e-07,
+      "loss": 0.5568,
+      "step": 13413
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3498906010398835,
+      "learning_rate": 5.155885406154937e-07,
+      "loss": 0.6462,
+      "step": 13414
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3912135979834142,
+      "learning_rate": 5.151302304858192e-07,
+      "loss": 0.6074,
+      "step": 13415
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4049926267385737,
+      "learning_rate": 5.146721130809873e-07,
+      "loss": 0.6429,
+      "step": 13416
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1236036926881743,
+      "learning_rate": 5.142141884206842e-07,
+      "loss": 0.5236,
+      "step": 13417
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4473946142918936,
+      "learning_rate": 5.137564565245873e-07,
+      "loss": 0.6771,
+      "step": 13418
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3342831830519823,
+      "learning_rate": 5.132989174123659e-07,
+      "loss": 0.6112,
+      "step": 13419
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.434840706937629,
+      "learning_rate": 5.128415711036849e-07,
+      "loss": 0.602,
+      "step": 13420
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2329613438185365,
+      "learning_rate": 5.12384417618193e-07,
+      "loss": 0.5266,
+      "step": 13421
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.707389088559098,
+      "learning_rate": 5.119274569755378e-07,
+      "loss": 0.7227,
+      "step": 13422
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3550940696139417,
+      "learning_rate": 5.114706891953552e-07,
+      "loss": 0.6486,
+      "step": 13423
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5444843025271255,
+      "learning_rate": 5.110141142972735e-07,
+      "loss": 0.5865,
+      "step": 13424
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3647365719612705,
+      "learning_rate": 5.105577323009131e-07,
+      "loss": 0.5753,
+      "step": 13425
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4515693435135795,
+      "learning_rate": 5.101015432258843e-07,
+      "loss": 0.6054,
+      "step": 13426
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2481093407365997,
+      "learning_rate": 5.096455470917943e-07,
+      "loss": 0.6181,
+      "step": 13427
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2738675242916644,
+      "learning_rate": 5.091897439182342e-07,
+      "loss": 0.5925,
+      "step": 13428
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5693343514959421,
+      "learning_rate": 5.087341337247914e-07,
+      "loss": 0.755,
+      "step": 13429
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4401171662027623,
+      "learning_rate": 5.082787165310471e-07,
+      "loss": 0.6609,
+      "step": 13430
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.331512822476221,
+      "learning_rate": 5.078234923565684e-07,
+      "loss": 0.5782,
+      "step": 13431
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.578714405650059,
+      "learning_rate": 5.073684612209201e-07,
+      "loss": 0.7164,
+      "step": 13432
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4135970459898215,
+      "learning_rate": 5.069136231436539e-07,
+      "loss": 0.5973,
+      "step": 13433
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2602905616668147,
+      "learning_rate": 5.064589781443163e-07,
+      "loss": 0.6479,
+      "step": 13434
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.239450589125777,
+      "learning_rate": 5.060045262424446e-07,
+      "loss": 0.4811,
+      "step": 13435
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2141106231282228,
+      "learning_rate": 5.055502674575663e-07,
+      "loss": 0.6912,
+      "step": 13436
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2931121461835113,
+      "learning_rate": 5.05096201809202e-07,
+      "loss": 0.5743,
+      "step": 13437
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3361224876321776,
+      "learning_rate": 5.046423293168667e-07,
+      "loss": 0.5876,
+      "step": 13438
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3573777886832774,
+      "learning_rate": 5.041886500000603e-07,
+      "loss": 0.6373,
+      "step": 13439
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3539903594001872,
+      "learning_rate": 5.037351638782812e-07,
+      "loss": 0.6646,
+      "step": 13440
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7407213349852961,
+      "learning_rate": 5.032818709710152e-07,
+      "loss": 0.6716,
+      "step": 13441
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.544971601317596,
+      "learning_rate": 5.028287712977425e-07,
+      "loss": 0.7149,
+      "step": 13442
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4692951086497348,
+      "learning_rate": 5.023758648779336e-07,
+      "loss": 0.6779,
+      "step": 13443
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5680450071043457,
+      "learning_rate": 5.019231517310491e-07,
+      "loss": 0.7383,
+      "step": 13444
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0355326858551501,
+      "learning_rate": 5.014706318765466e-07,
+      "loss": 0.5774,
+      "step": 13445
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3435819463599163,
+      "learning_rate": 5.010183053338685e-07,
+      "loss": 0.6311,
+      "step": 13446
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2764288984753922,
+      "learning_rate": 5.00566172122453e-07,
+      "loss": 0.597,
+      "step": 13447
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.327521401172449,
+      "learning_rate": 5.001142322617314e-07,
+      "loss": 0.6959,
+      "step": 13448
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4335012228753126,
+      "learning_rate": 4.996624857711219e-07,
+      "loss": 0.6717,
+      "step": 13449
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2703194632164285,
+      "learning_rate": 4.992109326700384e-07,
+      "loss": 0.5622,
+      "step": 13450
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1764164015980163,
+      "learning_rate": 4.987595729778849e-07,
+      "loss": 0.708,
+      "step": 13451
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6626266802800653,
+      "learning_rate": 4.983084067140575e-07,
+      "loss": 0.5977,
+      "step": 13452
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3476054133556332,
+      "learning_rate": 4.978574338979442e-07,
+      "loss": 0.5646,
+      "step": 13453
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2378919278466993,
+      "learning_rate": 4.97406654548922e-07,
+      "loss": 0.5995,
+      "step": 13454
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2912943466445579,
+      "learning_rate": 4.96956068686365e-07,
+      "loss": 0.6808,
+      "step": 13455
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.70351511265427,
+      "learning_rate": 4.965056763296356e-07,
+      "loss": 0.6171,
+      "step": 13456
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3246474602728304,
+      "learning_rate": 4.960554774980853e-07,
+      "loss": 0.6232,
+      "step": 13457
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0548946781594644,
+      "learning_rate": 4.956054722110626e-07,
+      "loss": 0.5964,
+      "step": 13458
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2444298702203775,
+      "learning_rate": 4.951556604879049e-07,
+      "loss": 0.6248,
+      "step": 13459
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2141657511252302,
+      "learning_rate": 4.94706042347941e-07,
+      "loss": 0.6089,
+      "step": 13460
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2010339775243097,
+      "learning_rate": 4.942566178104924e-07,
+      "loss": 0.5996,
+      "step": 13461
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3051402564232646,
+      "learning_rate": 4.938073868948707e-07,
+      "loss": 0.7157,
+      "step": 13462
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4379347534413507,
+      "learning_rate": 4.933583496203831e-07,
+      "loss": 0.6034,
+      "step": 13463
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.160571967177185,
+      "learning_rate": 4.929095060063227e-07,
+      "loss": 0.6638,
+      "step": 13464
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2536138971860764,
+      "learning_rate": 4.92460856071979e-07,
+      "loss": 0.5504,
+      "step": 13465
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4000850644865306,
+      "learning_rate": 4.920123998366327e-07,
+      "loss": 0.5687,
+      "step": 13466
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2326738173582428,
+      "learning_rate": 4.915641373195512e-07,
+      "loss": 0.5517,
+      "step": 13467
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3429064200412186,
+      "learning_rate": 4.911160685400008e-07,
+      "loss": 0.6428,
+      "step": 13468
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2788229259760737,
+      "learning_rate": 4.906681935172342e-07,
+      "loss": 0.6456,
+      "step": 13469
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1791944432104207,
+      "learning_rate": 4.902205122704983e-07,
+      "loss": 0.5712,
+      "step": 13470
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3035410513190766,
+      "learning_rate": 4.897730248190313e-07,
+      "loss": 0.6388,
+      "step": 13471
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2594984707859025,
+      "learning_rate": 4.893257311820609e-07,
+      "loss": 0.7427,
+      "step": 13472
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1990736937694906,
+      "learning_rate": 4.888786313788108e-07,
+      "loss": 0.554,
+      "step": 13473
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4187222980302912,
+      "learning_rate": 4.88431725428492e-07,
+      "loss": 0.6938,
+      "step": 13474
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2510000078286594,
+      "learning_rate": 4.879850133503106e-07,
+      "loss": 0.5704,
+      "step": 13475
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.443326587319765,
+      "learning_rate": 4.875384951634616e-07,
+      "loss": 0.6875,
+      "step": 13476
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.680454131610627,
+      "learning_rate": 4.870921708871334e-07,
+      "loss": 0.554,
+      "step": 13477
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3621240660755078,
+      "learning_rate": 4.866460405405054e-07,
+      "loss": 0.6326,
+      "step": 13478
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0388352946095298,
+      "learning_rate": 4.862001041427488e-07,
+      "loss": 0.655,
+      "step": 13479
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4435463662769505,
+      "learning_rate": 4.857543617130256e-07,
+      "loss": 0.7159,
+      "step": 13480
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4974705474896384,
+      "learning_rate": 4.85308813270493e-07,
+      "loss": 0.6844,
+      "step": 13481
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3515345763877544,
+      "learning_rate": 4.848634588342932e-07,
+      "loss": 0.6564,
+      "step": 13482
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.346205205750705,
+      "learning_rate": 4.844182984235679e-07,
+      "loss": 0.5998,
+      "step": 13483
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.452750007839633,
+      "learning_rate": 4.839733320574457e-07,
+      "loss": 0.6843,
+      "step": 13484
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3811169810421096,
+      "learning_rate": 4.835285597550454e-07,
+      "loss": 0.6167,
+      "step": 13485
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2886076515940463,
+      "learning_rate": 4.830839815354827e-07,
+      "loss": 0.5891,
+      "step": 13486
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0094955967392458,
+      "learning_rate": 4.826395974178605e-07,
+      "loss": 0.5525,
+      "step": 13487
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2368090004572243,
+      "learning_rate": 4.821954074212759e-07,
+      "loss": 0.6227,
+      "step": 13488
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5691194694364168,
+      "learning_rate": 4.817514115648164e-07,
+      "loss": 0.6875,
+      "step": 13489
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3286171643137727,
+      "learning_rate": 4.813076098675601e-07,
+      "loss": 0.5822,
+      "step": 13490
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3585055767863823,
+      "learning_rate": 4.808640023485805e-07,
+      "loss": 0.6589,
+      "step": 13491
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4137276054642958,
+      "learning_rate": 4.804205890269398e-07,
+      "loss": 0.6811,
+      "step": 13492
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.897050048761827,
+      "learning_rate": 4.799773699216914e-07,
+      "loss": 0.6618,
+      "step": 13493
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3949108713342757,
+      "learning_rate": 4.795343450518825e-07,
+      "loss": 0.6908,
+      "step": 13494
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7383241569122154,
+      "learning_rate": 4.790915144365504e-07,
+      "loss": 0.7216,
+      "step": 13495
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2360416995364794,
+      "learning_rate": 4.786488780947246e-07,
+      "loss": 0.6068,
+      "step": 13496
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.445864891087476,
+      "learning_rate": 4.782064360454258e-07,
+      "loss": 0.654,
+      "step": 13497
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5682847026556699,
+      "learning_rate": 4.777641883076661e-07,
+      "loss": 0.6016,
+      "step": 13498
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5006776883888293,
+      "learning_rate": 4.773221349004531e-07,
+      "loss": 0.6242,
+      "step": 13499
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1684357053546783,
+      "learning_rate": 4.768802758427782e-07,
+      "loss": 0.608,
+      "step": 13500
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2539874096069605,
+      "learning_rate": 4.7643861115363234e-07,
+      "loss": 0.5827,
+      "step": 13501
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4137343216358218,
+      "learning_rate": 4.759971408519937e-07,
+      "loss": 0.6652,
+      "step": 13502
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3332149474959796,
+      "learning_rate": 4.755558649568337e-07,
+      "loss": 0.74,
+      "step": 13503
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7328384191602706,
+      "learning_rate": 4.7511478348711447e-07,
+      "loss": 0.6044,
+      "step": 13504
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3694201963472235,
+      "learning_rate": 4.7467389646178976e-07,
+      "loss": 0.7012,
+      "step": 13505
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5624287346977623,
+      "learning_rate": 4.742332038998082e-07,
+      "loss": 0.6205,
+      "step": 13506
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3044293369081974,
+      "learning_rate": 4.7379270582010416e-07,
+      "loss": 0.5762,
+      "step": 13507
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3283367711910314,
+      "learning_rate": 4.7335240224160695e-07,
+      "loss": 0.5412,
+      "step": 13508
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4966473503929665,
+      "learning_rate": 4.729122931832392e-07,
+      "loss": 0.6605,
+      "step": 13509
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3047701913868055,
+      "learning_rate": 4.7247237866391236e-07,
+      "loss": 0.5801,
+      "step": 13510
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7778498442492354,
+      "learning_rate": 4.7203265870253144e-07,
+      "loss": 0.6171,
+      "step": 13511
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2878191608231149,
+      "learning_rate": 4.715931333179913e-07,
+      "loss": 0.5711,
+      "step": 13512
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8504238057616191,
+      "learning_rate": 4.7115380252917897e-07,
+      "loss": 0.7589,
+      "step": 13513
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4754584886301665,
+      "learning_rate": 4.707146663549744e-07,
+      "loss": 0.7165,
+      "step": 13514
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2952927247476977,
+      "learning_rate": 4.7027572481424753e-07,
+      "loss": 0.6603,
+      "step": 13515
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3547850559441807,
+      "learning_rate": 4.6983697792586103e-07,
+      "loss": 0.6193,
+      "step": 13516
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1984738952761431,
+      "learning_rate": 4.6939842570867034e-07,
+      "loss": 0.6653,
+      "step": 13517
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1266930367927699,
+      "learning_rate": 4.6896006818151706e-07,
+      "loss": 0.6027,
+      "step": 13518
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3520950527324107,
+      "learning_rate": 4.685219053632423e-07,
+      "loss": 0.7287,
+      "step": 13519
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3759269629544837,
+      "learning_rate": 4.6808393727267307e-07,
+      "loss": 0.5084,
+      "step": 13520
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3530057756561196,
+      "learning_rate": 4.676461639286306e-07,
+      "loss": 0.5895,
+      "step": 13521
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9845412596786457,
+      "learning_rate": 4.6720858534992584e-07,
+      "loss": 0.6069,
+      "step": 13522
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4519724792197077,
+      "learning_rate": 4.667712015553627e-07,
+      "loss": 0.6794,
+      "step": 13523
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1640566385350761,
+      "learning_rate": 4.663340125637389e-07,
+      "loss": 0.6824,
+      "step": 13524
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4150358616243741,
+      "learning_rate": 4.658970183938383e-07,
+      "loss": 0.6578,
+      "step": 13525
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2544414273278464,
+      "learning_rate": 4.654602190644408e-07,
+      "loss": 0.6264,
+      "step": 13526
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7990475562688468,
+      "learning_rate": 4.650236145943171e-07,
+      "loss": 0.653,
+      "step": 13527
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2828773746813265,
+      "learning_rate": 4.6458720500222873e-07,
+      "loss": 0.5969,
+      "step": 13528
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2279983225741253,
+      "learning_rate": 4.6415099030692914e-07,
+      "loss": 0.59,
+      "step": 13529
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.509006342011548,
+      "learning_rate": 4.6371497052716265e-07,
+      "loss": 0.6427,
+      "step": 13530
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.191569442115888,
+      "learning_rate": 4.6327914568166763e-07,
+      "loss": 0.5699,
+      "step": 13531
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5720206587853596,
+      "learning_rate": 4.628435157891709e-07,
+      "loss": 0.9071,
+      "step": 13532
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4717855437278566,
+      "learning_rate": 4.624080808683923e-07,
+      "loss": 0.6721,
+      "step": 13533
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7173310839205977,
+      "learning_rate": 4.619728409380453e-07,
+      "loss": 0.544,
+      "step": 13534
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1267891705391826,
+      "learning_rate": 4.615377960168321e-07,
+      "loss": 0.7251,
+      "step": 13535
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1864969651402486,
+      "learning_rate": 4.611029461234473e-07,
+      "loss": 0.5486,
+      "step": 13536
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1086623252171288,
+      "learning_rate": 4.6066829127657807e-07,
+      "loss": 0.6527,
+      "step": 13537
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3897622421901044,
+      "learning_rate": 4.6023383149490066e-07,
+      "loss": 0.8034,
+      "step": 13538
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5099133410563086,
+      "learning_rate": 4.597995667970878e-07,
+      "loss": 0.725,
+      "step": 13539
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0469318367837877,
+      "learning_rate": 4.5936549720179855e-07,
+      "loss": 0.5423,
+      "step": 13540
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6148488793853046,
+      "learning_rate": 4.589316227276852e-07,
+      "loss": 0.5215,
+      "step": 13541
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8659468869354507,
+      "learning_rate": 4.584979433933956e-07,
+      "loss": 0.6289,
+      "step": 13542
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.087655639893616,
+      "learning_rate": 4.5806445921756146e-07,
+      "loss": 0.6245,
+      "step": 13543
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.529017355831857,
+      "learning_rate": 4.5763117021881467e-07,
+      "loss": 0.6671,
+      "step": 13544
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5053837396664398,
+      "learning_rate": 4.571980764157724e-07,
+      "loss": 0.5922,
+      "step": 13545
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.36130739181168,
+      "learning_rate": 4.567651778270465e-07,
+      "loss": 0.6181,
+      "step": 13546
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4016309906948936,
+      "learning_rate": 4.563324744712389e-07,
+      "loss": 0.5004,
+      "step": 13547
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.350258982884306,
+      "learning_rate": 4.5589996636694387e-07,
+      "loss": 0.6345,
+      "step": 13548
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4415285628451415,
+      "learning_rate": 4.5546765353274846e-07,
+      "loss": 0.5694,
+      "step": 13549
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2316321121043898,
+      "learning_rate": 4.5503553598722826e-07,
+      "loss": 0.5656,
+      "step": 13550
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3851659607114912,
+      "learning_rate": 4.546036137489529e-07,
+      "loss": 0.5994,
+      "step": 13551
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2050195095485425,
+      "learning_rate": 4.5417188683648417e-07,
+      "loss": 0.7017,
+      "step": 13552
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3956854039239828,
+      "learning_rate": 4.5374035526837334e-07,
+      "loss": 0.7615,
+      "step": 13553
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3532199778786387,
+      "learning_rate": 4.5330901906316506e-07,
+      "loss": 0.6409,
+      "step": 13554
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.515101932346298,
+      "learning_rate": 4.528778782393939e-07,
+      "loss": 0.5775,
+      "step": 13555
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4393367109686865,
+      "learning_rate": 4.5244693281558615e-07,
+      "loss": 0.6694,
+      "step": 13556
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.498429477753677,
+      "learning_rate": 4.5201618281026374e-07,
+      "loss": 0.5997,
+      "step": 13557
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.441490582336463,
+      "learning_rate": 4.5158562824193395e-07,
+      "loss": 0.5875,
+      "step": 13558
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6298892167062882,
+      "learning_rate": 4.511552691290988e-07,
+      "loss": 0.7361,
+      "step": 13559
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.25704742943567,
+      "learning_rate": 4.507251054902539e-07,
+      "loss": 0.6771,
+      "step": 13560
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5630853716679782,
+      "learning_rate": 4.5029513734388187e-07,
+      "loss": 0.6434,
+      "step": 13561
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4362130950647396,
+      "learning_rate": 4.498653647084611e-07,
+      "loss": 0.6476,
+      "step": 13562
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5809518910395877,
+      "learning_rate": 4.494357876024591e-07,
+      "loss": 0.6483,
+      "step": 13563
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5117986450883938,
+      "learning_rate": 4.490064060443361e-07,
+      "loss": 0.5972,
+      "step": 13564
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4140859847250233,
+      "learning_rate": 4.4857722005254345e-07,
+      "loss": 0.6737,
+      "step": 13565
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.069712841707881,
+      "learning_rate": 4.4814822964552363e-07,
+      "loss": 0.5806,
+      "step": 13566
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2187460303241884,
+      "learning_rate": 4.47719434841713e-07,
+      "loss": 0.5546,
+      "step": 13567
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3447995114417741,
+      "learning_rate": 4.4729083565953626e-07,
+      "loss": 0.6842,
+      "step": 13568
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4842649782423771,
+      "learning_rate": 4.468624321174109e-07,
+      "loss": 0.5917,
+      "step": 13569
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2020118422985155,
+      "learning_rate": 4.4643422423374825e-07,
+      "loss": 0.6322,
+      "step": 13570
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4051713188601984,
+      "learning_rate": 4.460062120269476e-07,
+      "loss": 0.7469,
+      "step": 13571
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.212621833027524,
+      "learning_rate": 4.4557839551540303e-07,
+      "loss": 0.5583,
+      "step": 13572
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.274536571582735,
+      "learning_rate": 4.4515077471749767e-07,
+      "loss": 0.6792,
+      "step": 13573
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.37781271681496,
+      "learning_rate": 4.4472334965160736e-07,
+      "loss": 0.6141,
+      "step": 13574
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4576105822861605,
+      "learning_rate": 4.442961203361013e-07,
+      "loss": 0.8641,
+      "step": 13575
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2813003041013011,
+      "learning_rate": 4.438690867893347e-07,
+      "loss": 0.6384,
+      "step": 13576
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.459515348376656,
+      "learning_rate": 4.4344224902966194e-07,
+      "loss": 0.7671,
+      "step": 13577
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3275882315660774,
+      "learning_rate": 4.430156070754249e-07,
+      "loss": 0.5926,
+      "step": 13578
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.265634246955758,
+      "learning_rate": 4.4258916094495394e-07,
+      "loss": 0.561,
+      "step": 13579
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.762335858958038,
+      "learning_rate": 4.421629106565778e-07,
+      "loss": 0.7482,
+      "step": 13580
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1936195645246301,
+      "learning_rate": 4.417368562286117e-07,
+      "loss": 0.4765,
+      "step": 13581
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0531956044461426,
+      "learning_rate": 4.4131099767936504e-07,
+      "loss": 0.6107,
+      "step": 13582
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5098877200390814,
+      "learning_rate": 4.4088533502713803e-07,
+      "loss": 0.6607,
+      "step": 13583
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4312628663290485,
+      "learning_rate": 4.4045986829022e-07,
+      "loss": 0.645,
+      "step": 13584
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6446766426763695,
+      "learning_rate": 4.400345974868986e-07,
+      "loss": 0.819,
+      "step": 13585
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5527628357380059,
+      "learning_rate": 4.396095226354441e-07,
+      "loss": 0.6117,
+      "step": 13586
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.473839167753101,
+      "learning_rate": 4.391846437541258e-07,
+      "loss": 0.7471,
+      "step": 13587
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5844260500645062,
+      "learning_rate": 4.3875996086120085e-07,
+      "loss": 0.6536,
+      "step": 13588
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.202192295659098,
+      "learning_rate": 4.38335473974919e-07,
+      "loss": 0.5559,
+      "step": 13589
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4841863513396443,
+      "learning_rate": 4.379111831135213e-07,
+      "loss": 0.7221,
+      "step": 13590
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.476678636659773,
+      "learning_rate": 4.374870882952398e-07,
+      "loss": 0.6336,
+      "step": 13591
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.406417806448294,
+      "learning_rate": 4.370631895382993e-07,
+      "loss": 0.7064,
+      "step": 13592
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4074669543669485,
+      "learning_rate": 4.3663948686091694e-07,
+      "loss": 0.5858,
+      "step": 13593
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5809256293904816,
+      "learning_rate": 4.362159802812971e-07,
+      "loss": 0.6561,
+      "step": 13594
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.474675031959856,
+      "learning_rate": 4.357926698176418e-07,
+      "loss": 0.7378,
+      "step": 13595
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1497852099340091,
+      "learning_rate": 4.3536955548814153e-07,
+      "loss": 0.5367,
+      "step": 13596
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4711890852902305,
+      "learning_rate": 4.3494663731097564e-07,
+      "loss": 0.7574,
+      "step": 13597
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3186444986105814,
+      "learning_rate": 4.3452391530432016e-07,
+      "loss": 0.5894,
+      "step": 13598
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4427381765119294,
+      "learning_rate": 4.341013894863405e-07,
+      "loss": 0.6546,
+      "step": 13599
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4250442792430205,
+      "learning_rate": 4.3367905987519275e-07,
+      "loss": 0.732,
+      "step": 13600
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2581039490915191,
+      "learning_rate": 4.332569264890252e-07,
+      "loss": 0.5136,
+      "step": 13601
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.464533909369156,
+      "learning_rate": 4.328349893459777e-07,
+      "loss": 0.6377,
+      "step": 13602
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.29331832418239,
+      "learning_rate": 4.3241324846418355e-07,
+      "loss": 0.6844,
+      "step": 13603
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.438692823500704,
+      "learning_rate": 4.3199170386176325e-07,
+      "loss": 0.6017,
+      "step": 13604
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.313199049713413,
+      "learning_rate": 4.31570355556834e-07,
+      "loss": 0.5906,
+      "step": 13605
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9035322262973555,
+      "learning_rate": 4.3114920356750076e-07,
+      "loss": 0.729,
+      "step": 13606
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3200830803667232,
+      "learning_rate": 4.3072824791186187e-07,
+      "loss": 0.6688,
+      "step": 13607
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1587375815929561,
+      "learning_rate": 4.3030748860800606e-07,
+      "loss": 0.763,
+      "step": 13608
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2296561868038267,
+      "learning_rate": 4.2988692567401515e-07,
+      "loss": 0.5389,
+      "step": 13609
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2962977563421627,
+      "learning_rate": 4.294665591279601e-07,
+      "loss": 0.6435,
+      "step": 13610
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.900205188734129,
+      "learning_rate": 4.290463889879082e-07,
+      "loss": 0.6384,
+      "step": 13611
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.520120016463019,
+      "learning_rate": 4.286264152719111e-07,
+      "loss": 0.609,
+      "step": 13612
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3789142663316503,
+      "learning_rate": 4.2820663799801876e-07,
+      "loss": 0.6171,
+      "step": 13613
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4869316437624605,
+      "learning_rate": 4.2778705718426907e-07,
+      "loss": 0.6777,
+      "step": 13614
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3602184140842701,
+      "learning_rate": 4.273676728486925e-07,
+      "loss": 0.6575,
+      "step": 13615
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7779826991847174,
+      "learning_rate": 4.2694848500931086e-07,
+      "loss": 0.6521,
+      "step": 13616
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2280381567055885,
+      "learning_rate": 4.2652949368413634e-07,
+      "loss": 0.6041,
+      "step": 13617
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.838429230552251,
+      "learning_rate": 4.2611069889117727e-07,
+      "loss": 0.7774,
+      "step": 13618
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6126585309178376,
+      "learning_rate": 4.2569210064842716e-07,
+      "loss": 0.6737,
+      "step": 13619
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.412608250516205,
+      "learning_rate": 4.252736989738737e-07,
+      "loss": 0.5636,
+      "step": 13620
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1427249349192425,
+      "learning_rate": 4.2485549388549927e-07,
+      "loss": 0.5648,
+      "step": 13621
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5180892091215463,
+      "learning_rate": 4.244374854012734e-07,
+      "loss": 0.8049,
+      "step": 13622
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2249727362742804,
+      "learning_rate": 4.2401967353915885e-07,
+      "loss": 0.6328,
+      "step": 13623
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0333727302845306,
+      "learning_rate": 4.236020583171108e-07,
+      "loss": 0.683,
+      "step": 13624
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4570308113400336,
+      "learning_rate": 4.2318463975307423e-07,
+      "loss": 0.6297,
+      "step": 13625
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4310029089462535,
+      "learning_rate": 4.227674178649865e-07,
+      "loss": 0.671,
+      "step": 13626
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3873677081493796,
+      "learning_rate": 4.2235039267077615e-07,
+      "loss": 0.609,
+      "step": 13627
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.284043431353784,
+      "learning_rate": 4.219335641883654e-07,
+      "loss": 0.4688,
+      "step": 13628
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3956427858931517,
+      "learning_rate": 4.215169324356666e-07,
+      "loss": 0.5139,
+      "step": 13629
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.767249626238597,
+      "learning_rate": 4.2110049743057933e-07,
+      "loss": 0.6569,
+      "step": 13630
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3610816735591633,
+      "learning_rate": 4.206842591910032e-07,
+      "loss": 0.5954,
+      "step": 13631
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4378124143338293,
+      "learning_rate": 4.202682177348227e-07,
+      "loss": 0.5892,
+      "step": 13632
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.149067064422466,
+      "learning_rate": 4.198523730799164e-07,
+      "loss": 0.6627,
+      "step": 13633
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3034804770413069,
+      "learning_rate": 4.194367252441545e-07,
+      "loss": 0.6708,
+      "step": 13634
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6110323863029181,
+      "learning_rate": 4.1902127424539695e-07,
+      "loss": 0.6259,
+      "step": 13635
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4521564145128014,
+      "learning_rate": 4.186060201014991e-07,
+      "loss": 0.6874,
+      "step": 13636
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5857115355799867,
+      "learning_rate": 4.181909628303027e-07,
+      "loss": 0.6484,
+      "step": 13637
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2110120412298098,
+      "learning_rate": 4.177761024496452e-07,
+      "loss": 0.5504,
+      "step": 13638
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4240965596696502,
+      "learning_rate": 4.1736143897735394e-07,
+      "loss": 0.7195,
+      "step": 13639
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4627097079171265,
+      "learning_rate": 4.1694697243124804e-07,
+      "loss": 0.6659,
+      "step": 13640
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3105862351252933,
+      "learning_rate": 4.1653270282913715e-07,
+      "loss": 0.671,
+      "step": 13641
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9822165195441167,
+      "learning_rate": 4.1611863018882424e-07,
+      "loss": 0.6904,
+      "step": 13642
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3240645600380019,
+      "learning_rate": 4.157047545281029e-07,
+      "loss": 0.7066,
+      "step": 13643
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4990584844702344,
+      "learning_rate": 4.152910758647577e-07,
+      "loss": 0.6404,
+      "step": 13644
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4510841840168718,
+      "learning_rate": 4.14877594216565e-07,
+      "loss": 0.6464,
+      "step": 13645
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4753927029220286,
+      "learning_rate": 4.14464309601294e-07,
+      "loss": 0.5532,
+      "step": 13646
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6363173718837818,
+      "learning_rate": 4.14051222036706e-07,
+      "loss": 0.6737,
+      "step": 13647
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4598569251885891,
+      "learning_rate": 4.1363833154054786e-07,
+      "loss": 0.6655,
+      "step": 13648
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4641798755567943,
+      "learning_rate": 4.1322563813056606e-07,
+      "loss": 0.5988,
+      "step": 13649
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3514911861367047,
+      "learning_rate": 4.1281314182449405e-07,
+      "loss": 0.6655,
+      "step": 13650
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4302458539952376,
+      "learning_rate": 4.1240084264005774e-07,
+      "loss": 0.6545,
+      "step": 13651
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.421656728082079,
+      "learning_rate": 4.1198874059497405e-07,
+      "loss": 0.6023,
+      "step": 13652
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0422095459952734,
+      "learning_rate": 4.1157683570695105e-07,
+      "loss": 0.5498,
+      "step": 13653
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.259896343535826,
+      "learning_rate": 4.111651279936929e-07,
+      "loss": 0.6272,
+      "step": 13654
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8419698356897387,
+      "learning_rate": 4.1075361747288657e-07,
+      "loss": 0.5551,
+      "step": 13655
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.293157482515941,
+      "learning_rate": 4.103423041622195e-07,
+      "loss": 0.6145,
+      "step": 13656
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3514239617255737,
+      "learning_rate": 4.099311880793655e-07,
+      "loss": 0.5943,
+      "step": 13657
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2138935571774545,
+      "learning_rate": 4.0952026924199027e-07,
+      "loss": 0.6881,
+      "step": 13658
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3657624026423956,
+      "learning_rate": 4.091095476677531e-07,
+      "loss": 0.6113,
+      "step": 13659
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7424634938154973,
+      "learning_rate": 4.0869902337430323e-07,
+      "loss": 0.6902,
+      "step": 13660
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2723032332848296,
+      "learning_rate": 4.082886963792815e-07,
+      "loss": 0.669,
+      "step": 13661
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.436697028402177,
+      "learning_rate": 4.0787856670032044e-07,
+      "loss": 0.6592,
+      "step": 13662
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3590764481129047,
+      "learning_rate": 4.0746863435504434e-07,
+      "loss": 0.6858,
+      "step": 13663
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0925141893416055,
+      "learning_rate": 4.070588993610697e-07,
+      "loss": 0.7006,
+      "step": 13664
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.361501846448554,
+      "learning_rate": 4.066493617360029e-07,
+      "loss": 0.5717,
+      "step": 13665
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5291985200031033,
+      "learning_rate": 4.062400214974432e-07,
+      "loss": 0.673,
+      "step": 13666
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3907006543508376,
+      "learning_rate": 4.0583087866298056e-07,
+      "loss": 0.5191,
+      "step": 13667
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.126044199985108,
+      "learning_rate": 4.054219332501974e-07,
+      "loss": 0.6461,
+      "step": 13668
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2016744322108601,
+      "learning_rate": 4.050131852766659e-07,
+      "loss": 0.5863,
+      "step": 13669
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4117966140122273,
+      "learning_rate": 4.046046347599514e-07,
+      "loss": 0.6177,
+      "step": 13670
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.393609965373271,
+      "learning_rate": 4.0419628171760927e-07,
+      "loss": 0.5483,
+      "step": 13671
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4712078054452686,
+      "learning_rate": 4.0378812616719056e-07,
+      "loss": 0.6708,
+      "step": 13672
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2962014591012763,
+      "learning_rate": 4.0338016812623007e-07,
+      "loss": 0.5676,
+      "step": 13673
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.399716416855015,
+      "learning_rate": 4.029724076122621e-07,
+      "loss": 0.6397,
+      "step": 13674
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3746016665237433,
+      "learning_rate": 4.025648446428082e-07,
+      "loss": 0.6209,
+      "step": 13675
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6445152580326536,
+      "learning_rate": 4.021574792353816e-07,
+      "loss": 0.6164,
+      "step": 13676
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.404733919893755,
+      "learning_rate": 4.017503114074883e-07,
+      "loss": 0.6458,
+      "step": 13677
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0712475199321152,
+      "learning_rate": 4.0134334117662375e-07,
+      "loss": 0.6272,
+      "step": 13678
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4711195235721775,
+      "learning_rate": 4.009365685602795e-07,
+      "loss": 0.5897,
+      "step": 13679
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4765522994707805,
+      "learning_rate": 4.005299935759327e-07,
+      "loss": 0.5502,
+      "step": 13680
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.175686562324145,
+      "learning_rate": 4.001236162410549e-07,
+      "loss": 0.7314,
+      "step": 13681
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3537083391051663,
+      "learning_rate": 3.997174365731105e-07,
+      "loss": 0.6396,
+      "step": 13682
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1967693259624856,
+      "learning_rate": 3.993114545895532e-07,
+      "loss": 0.5608,
+      "step": 13683
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4217060939333748,
+      "learning_rate": 3.989056703078292e-07,
+      "loss": 0.6297,
+      "step": 13684
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5089202872740828,
+      "learning_rate": 3.985000837453756e-07,
+      "loss": 0.5947,
+      "step": 13685
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3437148817587674,
+      "learning_rate": 3.9809469491962126e-07,
+      "loss": 0.5648,
+      "step": 13686
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.279301428108961,
+      "learning_rate": 3.976895038479872e-07,
+      "loss": 0.6167,
+      "step": 13687
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.436438421695576,
+      "learning_rate": 3.9728451054788396e-07,
+      "loss": 0.6494,
+      "step": 13688
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6546197931307416,
+      "learning_rate": 3.968797150367171e-07,
+      "loss": 0.6926,
+      "step": 13689
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5353363449285014,
+      "learning_rate": 3.9647511733188147e-07,
+      "loss": 0.5975,
+      "step": 13690
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4177213098311858,
+      "learning_rate": 3.960707174507611e-07,
+      "loss": 0.6211,
+      "step": 13691
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3480906804172414,
+      "learning_rate": 3.9566651541073586e-07,
+      "loss": 0.5524,
+      "step": 13692
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.028919506112487,
+      "learning_rate": 3.9526251122917526e-07,
+      "loss": 0.6246,
+      "step": 13693
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0978506260069414,
+      "learning_rate": 3.948587049234398e-07,
+      "loss": 0.6016,
+      "step": 13694
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5252411181592784,
+      "learning_rate": 3.944550965108818e-07,
+      "loss": 0.6044,
+      "step": 13695
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4151264740738017,
+      "learning_rate": 3.940516860088445e-07,
+      "loss": 0.5828,
+      "step": 13696
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6438490514186606,
+      "learning_rate": 3.936484734346663e-07,
+      "loss": 0.7727,
+      "step": 13697
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.463469610017703,
+      "learning_rate": 3.932454588056711e-07,
+      "loss": 0.7083,
+      "step": 13698
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.458278838034964,
+      "learning_rate": 3.928426421391773e-07,
+      "loss": 0.6013,
+      "step": 13699
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2387247802554406,
+      "learning_rate": 3.924400234524972e-07,
+      "loss": 0.5767,
+      "step": 13700
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3489583013848074,
+      "learning_rate": 3.9203760276293024e-07,
+      "loss": 0.5636,
+      "step": 13701
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2973448657782196,
+      "learning_rate": 3.9163538008777035e-07,
+      "loss": 0.5752,
+      "step": 13702
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.480970544842117,
+      "learning_rate": 3.9123335544430153e-07,
+      "loss": 0.6807,
+      "step": 13703
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2372304653884612,
+      "learning_rate": 3.9083152884979935e-07,
+      "loss": 0.6915,
+      "step": 13704
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3430833390867025,
+      "learning_rate": 3.9042990032153227e-07,
+      "loss": 0.544,
+      "step": 13705
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5837114004350337,
+      "learning_rate": 3.9002846987675704e-07,
+      "loss": 0.6885,
+      "step": 13706
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.768061568652613,
+      "learning_rate": 3.8962723753272645e-07,
+      "loss": 0.5473,
+      "step": 13707
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3858113415479685,
+      "learning_rate": 3.892262033066818e-07,
+      "loss": 0.5802,
+      "step": 13708
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3747812047098742,
+      "learning_rate": 3.8882536721585486e-07,
+      "loss": 0.6106,
+      "step": 13709
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5993154935651357,
+      "learning_rate": 3.884247292774718e-07,
+      "loss": 0.6867,
+      "step": 13710
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5876455166087833,
+      "learning_rate": 3.8802428950874884e-07,
+      "loss": 0.7012,
+      "step": 13711
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4095491218980736,
+      "learning_rate": 3.876240479268939e-07,
+      "loss": 0.6537,
+      "step": 13712
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.303969567787124,
+      "learning_rate": 3.872240045491055e-07,
+      "loss": 0.6606,
+      "step": 13713
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3282535805824918,
+      "learning_rate": 3.868241593925742e-07,
+      "loss": 0.568,
+      "step": 13714
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4493747792132607,
+      "learning_rate": 3.8642451247448477e-07,
+      "loss": 0.5946,
+      "step": 13715
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.375024690022403,
+      "learning_rate": 3.8602506381200666e-07,
+      "loss": 0.6699,
+      "step": 13716
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.481203547870678,
+      "learning_rate": 3.8562581342230897e-07,
+      "loss": 0.66,
+      "step": 13717
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5406793639878604,
+      "learning_rate": 3.8522676132254635e-07,
+      "loss": 0.5996,
+      "step": 13718
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5550577032433905,
+      "learning_rate": 3.848279075298678e-07,
+      "loss": 0.6459,
+      "step": 13719
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5122358525813615,
+      "learning_rate": 3.8442925206141237e-07,
+      "loss": 0.7432,
+      "step": 13720
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4978315348337259,
+      "learning_rate": 3.840307949343114e-07,
+      "loss": 0.6189,
+      "step": 13721
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2030115420780578,
+      "learning_rate": 3.8363253616568784e-07,
+      "loss": 0.6505,
+      "step": 13722
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.575962984608746,
+      "learning_rate": 3.8323447577265525e-07,
+      "loss": 0.6637,
+      "step": 13723
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.521966252304376,
+      "learning_rate": 3.828366137723183e-07,
+      "loss": 0.7891,
+      "step": 13724
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4978027591935532,
+      "learning_rate": 3.82438950181776e-07,
+      "loss": 0.7023,
+      "step": 13725
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.331351364928933,
+      "learning_rate": 3.820414850181153e-07,
+      "loss": 0.6231,
+      "step": 13726
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2570754482168307,
+      "learning_rate": 3.8164421829841756e-07,
+      "loss": 0.48,
+      "step": 13727
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3014214640322441,
+      "learning_rate": 3.81247150039753e-07,
+      "loss": 0.5845,
+      "step": 13728
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1874209867268395,
+      "learning_rate": 3.80850280259184e-07,
+      "loss": 0.6342,
+      "step": 13729
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.379852370760171,
+      "learning_rate": 3.8045360897376814e-07,
+      "loss": 0.569,
+      "step": 13730
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4866860654843455,
+      "learning_rate": 3.800571362005473e-07,
+      "loss": 0.675,
+      "step": 13731
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.176945522251395,
+      "learning_rate": 3.7966086195656e-07,
+      "loss": 0.6619,
+      "step": 13732
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3773516517617646,
+      "learning_rate": 3.7926478625883725e-07,
+      "loss": 0.6212,
+      "step": 13733
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6642884440514638,
+      "learning_rate": 3.7886890912439633e-07,
+      "loss": 0.6858,
+      "step": 13734
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4067318927193804,
+      "learning_rate": 3.784732305702504e-07,
+      "loss": 0.6873,
+      "step": 13735
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7349054441361356,
+      "learning_rate": 3.7807775061340257e-07,
+      "loss": 0.635,
+      "step": 13736
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4736473447507115,
+      "learning_rate": 3.776824692708475e-07,
+      "loss": 0.6682,
+      "step": 13737
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.625691624043796,
+      "learning_rate": 3.772873865595711e-07,
+      "loss": 0.7702,
+      "step": 13738
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5412383560608212,
+      "learning_rate": 3.768925024965503e-07,
+      "loss": 0.7593,
+      "step": 13739
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7801681985387239,
+      "learning_rate": 3.76497817098756e-07,
+      "loss": 0.6735,
+      "step": 13740
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4972688233429836,
+      "learning_rate": 3.761033303831474e-07,
+      "loss": 0.6895,
+      "step": 13741
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3654943260274472,
+      "learning_rate": 3.7570904236667536e-07,
+      "loss": 0.521,
+      "step": 13742
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3893332382916033,
+      "learning_rate": 3.7531495306628584e-07,
+      "loss": 0.6461,
+      "step": 13743
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0063948444314752,
+      "learning_rate": 3.749210624989125e-07,
+      "loss": 0.6081,
+      "step": 13744
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.420222666335132,
+      "learning_rate": 3.7452737068148173e-07,
+      "loss": 0.6299,
+      "step": 13745
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8923887636376913,
+      "learning_rate": 3.741338776309111e-07,
+      "loss": 0.5309,
+      "step": 13746
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3544802554309388,
+      "learning_rate": 3.7374058336410945e-07,
+      "loss": 0.5956,
+      "step": 13747
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5686236552219825,
+      "learning_rate": 3.733474878979798e-07,
+      "loss": 0.6588,
+      "step": 13748
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1746997408938893,
+      "learning_rate": 3.729545912494115e-07,
+      "loss": 0.554,
+      "step": 13749
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6886851211987264,
+      "learning_rate": 3.7256189343528926e-07,
+      "loss": 0.6504,
+      "step": 13750
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1624597890116803,
+      "learning_rate": 3.721693944724897e-07,
+      "loss": 0.5437,
+      "step": 13751
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4488732753660003,
+      "learning_rate": 3.717770943778759e-07,
+      "loss": 0.7883,
+      "step": 13752
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6321494254422577,
+      "learning_rate": 3.7138499316830946e-07,
+      "loss": 0.6593,
+      "step": 13753
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4627147213789606,
+      "learning_rate": 3.7099309086063794e-07,
+      "loss": 0.7035,
+      "step": 13754
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4574163223386665,
+      "learning_rate": 3.706013874717024e-07,
+      "loss": 0.6508,
+      "step": 13755
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4053132927119445,
+      "learning_rate": 3.702098830183359e-07,
+      "loss": 0.6312,
+      "step": 13756
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4671996619620113,
+      "learning_rate": 3.698185775173607e-07,
+      "loss": 0.6097,
+      "step": 13757
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.054651806021479,
+      "learning_rate": 3.694274709855944e-07,
+      "loss": 0.6093,
+      "step": 13758
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5070118151434817,
+      "learning_rate": 3.6903656343984293e-07,
+      "loss": 0.7919,
+      "step": 13759
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.371218911077962,
+      "learning_rate": 3.6864585489690297e-07,
+      "loss": 0.5535,
+      "step": 13760
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2998314303266962,
+      "learning_rate": 3.6825534537356545e-07,
+      "loss": 0.6272,
+      "step": 13761
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1263233652750588,
+      "learning_rate": 3.678650348866114e-07,
+      "loss": 0.6548,
+      "step": 13762
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2572114379435688,
+      "learning_rate": 3.6747492345281297e-07,
+      "loss": 0.5449,
+      "step": 13763
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6932295268284807,
+      "learning_rate": 3.670850110889346e-07,
+      "loss": 0.7167,
+      "step": 13764
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3441224351785122,
+      "learning_rate": 3.6669529781173053e-07,
+      "loss": 0.5883,
+      "step": 13765
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0641943130611766,
+      "learning_rate": 3.6630578363794965e-07,
+      "loss": 0.5303,
+      "step": 13766
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3189001639332918,
+      "learning_rate": 3.659164685843275e-07,
+      "loss": 0.6083,
+      "step": 13767
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4313588413723293,
+      "learning_rate": 3.655273526675962e-07,
+      "loss": 0.6518,
+      "step": 13768
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5779822523235996,
+      "learning_rate": 3.651384359044774e-07,
+      "loss": 0.5996,
+      "step": 13769
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3605085287324628,
+      "learning_rate": 3.647497183116799e-07,
+      "loss": 0.5713,
+      "step": 13770
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4382437015256149,
+      "learning_rate": 3.6436119990591214e-07,
+      "loss": 0.6813,
+      "step": 13771
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3960926563358365,
+      "learning_rate": 3.639728807038667e-07,
+      "loss": 0.6259,
+      "step": 13772
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1011092255269088,
+      "learning_rate": 3.6358476072223205e-07,
+      "loss": 0.6067,
+      "step": 13773
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1719378233567355,
+      "learning_rate": 3.631968399776864e-07,
+      "loss": 0.6496,
+      "step": 13774
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.455888944282111,
+      "learning_rate": 3.628091184868976e-07,
+      "loss": 0.6399,
+      "step": 13775
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.412792562225148,
+      "learning_rate": 3.6242159626653004e-07,
+      "loss": 0.6558,
+      "step": 13776
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4483800751646025,
+      "learning_rate": 3.6203427333323495e-07,
+      "loss": 0.6846,
+      "step": 13777
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.537104306304058,
+      "learning_rate": 3.616471497036561e-07,
+      "loss": 0.7728,
+      "step": 13778
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2000514317982787,
+      "learning_rate": 3.6126022539442975e-07,
+      "loss": 0.5977,
+      "step": 13779
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2602529583208286,
+      "learning_rate": 3.6087350042218194e-07,
+      "loss": 0.6098,
+      "step": 13780
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4274108128246945,
+      "learning_rate": 3.6048697480353225e-07,
+      "loss": 0.6737,
+      "step": 13781
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.402310890245512,
+      "learning_rate": 3.601006485550895e-07,
+      "loss": 0.6028,
+      "step": 13782
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.413174913324366,
+      "learning_rate": 3.597145216934556e-07,
+      "loss": 0.6601,
+      "step": 13783
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5385165046491363,
+      "learning_rate": 3.593285942352237e-07,
+      "loss": 0.6369,
+      "step": 13784
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.455344963341594,
+      "learning_rate": 3.5894286619697627e-07,
+      "loss": 0.6555,
+      "step": 13785
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4582092065027625,
+      "learning_rate": 3.585573375952911e-07,
+      "loss": 0.6343,
+      "step": 13786
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.396127118806889,
+      "learning_rate": 3.581720084467344e-07,
+      "loss": 0.7255,
+      "step": 13787
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.394953087845852,
+      "learning_rate": 3.57786878767864e-07,
+      "loss": 0.6008,
+      "step": 13788
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1644669627912274,
+      "learning_rate": 3.5740194857523e-07,
+      "loss": 0.5483,
+      "step": 13789
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.740705944027811,
+      "learning_rate": 3.570172178853731e-07,
+      "loss": 0.6188,
+      "step": 13790
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5998233681660585,
+      "learning_rate": 3.56632686714829e-07,
+      "loss": 0.6105,
+      "step": 13791
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2958547113957557,
+      "learning_rate": 3.5624835508011824e-07,
+      "loss": 0.5305,
+      "step": 13792
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0636651128224313,
+      "learning_rate": 3.5586422299775714e-07,
+      "loss": 0.7401,
+      "step": 13793
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.32457003211523,
+      "learning_rate": 3.554802904842547e-07,
+      "loss": 0.6187,
+      "step": 13794
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3144759439939766,
+      "learning_rate": 3.5509655755610773e-07,
+      "loss": 0.6136,
+      "step": 13795
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5906536937500197,
+      "learning_rate": 3.547130242298064e-07,
+      "loss": 0.6649,
+      "step": 13796
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3058291392868713,
+      "learning_rate": 3.5432969052183186e-07,
+      "loss": 0.567,
+      "step": 13797
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3986231244865113,
+      "learning_rate": 3.539465564486577e-07,
+      "loss": 0.6776,
+      "step": 13798
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0538698533892785,
+      "learning_rate": 3.5356362202674687e-07,
+      "loss": 0.6254,
+      "step": 13799
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4809329179690782,
+      "learning_rate": 3.5318088727255506e-07,
+      "loss": 0.6975,
+      "step": 13800
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2376587960151728,
+      "learning_rate": 3.527983522025291e-07,
+      "loss": 0.5566,
+      "step": 13801
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2148077598108802,
+      "learning_rate": 3.5241601683310924e-07,
+      "loss": 0.5571,
+      "step": 13802
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1984988743410936,
+      "learning_rate": 3.520338811807222e-07,
+      "loss": 0.5562,
+      "step": 13803
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3479291249381966,
+      "learning_rate": 3.516519452617922e-07,
+      "loss": 0.6524,
+      "step": 13804
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4466661811718935,
+      "learning_rate": 3.5127020909273e-07,
+      "loss": 0.7548,
+      "step": 13805
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2847370202808555,
+      "learning_rate": 3.508886726899402e-07,
+      "loss": 0.5872,
+      "step": 13806
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.528681245806517,
+      "learning_rate": 3.5050733606981855e-07,
+      "loss": 0.6523,
+      "step": 13807
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.271503097040181,
+      "learning_rate": 3.501261992487509e-07,
+      "loss": 0.6497,
+      "step": 13808
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6468575645742063,
+      "learning_rate": 3.4974526224311744e-07,
+      "loss": 0.656,
+      "step": 13809
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1537544448910317,
+      "learning_rate": 3.493645250692862e-07,
+      "loss": 0.6305,
+      "step": 13810
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.52794473116279,
+      "learning_rate": 3.4898398774361854e-07,
+      "loss": 0.6989,
+      "step": 13811
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3073835698090228,
+      "learning_rate": 3.486036502824675e-07,
+      "loss": 0.6106,
+      "step": 13812
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2686003641485681,
+      "learning_rate": 3.4822351270217656e-07,
+      "loss": 0.6057,
+      "step": 13813
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.515575034953158,
+      "learning_rate": 3.478435750190817e-07,
+      "loss": 0.7776,
+      "step": 13814
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6165094281934091,
+      "learning_rate": 3.474638372495098e-07,
+      "loss": 0.636,
+      "step": 13815
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.446797336670893,
+      "learning_rate": 3.4708429940977785e-07,
+      "loss": 0.6543,
+      "step": 13816
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1302667127135253,
+      "learning_rate": 3.467049615161966e-07,
+      "loss": 0.7399,
+      "step": 13817
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0166405517223231,
+      "learning_rate": 3.463258235850653e-07,
+      "loss": 0.7491,
+      "step": 13818
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3851146406392554,
+      "learning_rate": 3.459468856326792e-07,
+      "loss": 0.5991,
+      "step": 13819
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1231773223632504,
+      "learning_rate": 3.455681476753209e-07,
+      "loss": 0.5025,
+      "step": 13820
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5225624006210148,
+      "learning_rate": 3.451896097292634e-07,
+      "loss": 0.637,
+      "step": 13821
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1088736070774574,
+      "learning_rate": 3.448112718107766e-07,
+      "loss": 0.604,
+      "step": 13822
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3851446922423343,
+      "learning_rate": 3.444331339361168e-07,
+      "loss": 0.6141,
+      "step": 13823
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.486923067401751,
+      "learning_rate": 3.4405519612153326e-07,
+      "loss": 0.5855,
+      "step": 13824
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5350457860935425,
+      "learning_rate": 3.4367745838326807e-07,
+      "loss": 0.5365,
+      "step": 13825
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0280702887009086,
+      "learning_rate": 3.432999207375515e-07,
+      "loss": 0.638,
+      "step": 13826
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5441898475733298,
+      "learning_rate": 3.4292258320061e-07,
+      "loss": 0.6714,
+      "step": 13827
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.224194123008277,
+      "learning_rate": 3.425454457886551e-07,
+      "loss": 0.5484,
+      "step": 13828
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5672268620696401,
+      "learning_rate": 3.4216850851789663e-07,
+      "loss": 0.6761,
+      "step": 13829
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.461560050712595,
+      "learning_rate": 3.4179177140453045e-07,
+      "loss": 0.767,
+      "step": 13830
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4180039766402195,
+      "learning_rate": 3.414152344647459e-07,
+      "loss": 0.5785,
+      "step": 13831
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4549365943833406,
+      "learning_rate": 3.410388977147244e-07,
+      "loss": 0.6547,
+      "step": 13832
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.509571079446023,
+      "learning_rate": 3.406627611706376e-07,
+      "loss": 0.6231,
+      "step": 13833
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4355597566883018,
+      "learning_rate": 3.402868248486485e-07,
+      "loss": 0.504,
+      "step": 13834
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3147212496511749,
+      "learning_rate": 3.3991108876491263e-07,
+      "loss": 0.5054,
+      "step": 13835
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1428794448311785,
+      "learning_rate": 3.3953555293557483e-07,
+      "loss": 0.5395,
+      "step": 13836
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4687082834121714,
+      "learning_rate": 3.391602173767744e-07,
+      "loss": 0.6485,
+      "step": 13837
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6512149088279116,
+      "learning_rate": 3.387850821046401e-07,
+      "loss": 0.6664,
+      "step": 13838
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2036215563989543,
+      "learning_rate": 3.3841014713529184e-07,
+      "loss": 0.603,
+      "step": 13839
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0510258575869567,
+      "learning_rate": 3.380354124848412e-07,
+      "loss": 0.6458,
+      "step": 13840
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1041561450157036,
+      "learning_rate": 3.376608781693913e-07,
+      "loss": 0.483,
+      "step": 13841
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4128731711776026,
+      "learning_rate": 3.372865442050377e-07,
+      "loss": 0.6333,
+      "step": 13842
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2349089006790057,
+      "learning_rate": 3.3691241060786586e-07,
+      "loss": 0.5459,
+      "step": 13843
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.282822393325891,
+      "learning_rate": 3.3653847739395174e-07,
+      "loss": 0.5786,
+      "step": 13844
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5367499876802828,
+      "learning_rate": 3.36164744579367e-07,
+      "loss": 0.6207,
+      "step": 13845
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.198010784869128,
+      "learning_rate": 3.357912121801682e-07,
+      "loss": 0.5214,
+      "step": 13846
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3975482682143852,
+      "learning_rate": 3.354178802124103e-07,
+      "loss": 0.6059,
+      "step": 13847
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3578330007127584,
+      "learning_rate": 3.350447486921338e-07,
+      "loss": 0.5952,
+      "step": 13848
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4168730714230904,
+      "learning_rate": 3.346718176353747e-07,
+      "loss": 0.8002,
+      "step": 13849
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5525805396573553,
+      "learning_rate": 3.342990870581575e-07,
+      "loss": 0.6505,
+      "step": 13850
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.252628399941244,
+      "learning_rate": 3.339265569764993e-07,
+      "loss": 0.6234,
+      "step": 13851
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.273817814904429,
+      "learning_rate": 3.3355422740640905e-07,
+      "loss": 0.6083,
+      "step": 13852
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5682367239976138,
+      "learning_rate": 3.331820983638867e-07,
+      "loss": 0.6383,
+      "step": 13853
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0506318006495696,
+      "learning_rate": 3.3281016986492165e-07,
+      "loss": 0.5657,
+      "step": 13854
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3195778095504411,
+      "learning_rate": 3.3243844192549947e-07,
+      "loss": 0.6918,
+      "step": 13855
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1740199061771095,
+      "learning_rate": 3.3206691456159233e-07,
+      "loss": 0.52,
+      "step": 13856
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6307514166313204,
+      "learning_rate": 3.316955877891659e-07,
+      "loss": 0.7589,
+      "step": 13857
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4231790795804296,
+      "learning_rate": 3.313244616241773e-07,
+      "loss": 0.5947,
+      "step": 13858
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4957902359647184,
+      "learning_rate": 3.3095353608257385e-07,
+      "loss": 0.5935,
+      "step": 13859
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.960645423313288,
+      "learning_rate": 3.3058281118029553e-07,
+      "loss": 0.6892,
+      "step": 13860
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3719982428542996,
+      "learning_rate": 3.302122869332736e-07,
+      "loss": 0.6137,
+      "step": 13861
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.00189374509949,
+      "learning_rate": 3.2984196335742844e-07,
+      "loss": 0.6833,
+      "step": 13862
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3224119359535629,
+      "learning_rate": 3.294718404686775e-07,
+      "loss": 0.5809,
+      "step": 13863
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.612072575384235,
+      "learning_rate": 3.2910191828292083e-07,
+      "loss": 0.5922,
+      "step": 13864
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7527189949507604,
+      "learning_rate": 3.2873219681605894e-07,
+      "loss": 0.7237,
+      "step": 13865
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2175569470523837,
+      "learning_rate": 3.283626760839775e-07,
+      "loss": 0.673,
+      "step": 13866
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2280424014964586,
+      "learning_rate": 3.279933561025567e-07,
+      "loss": 0.679,
+      "step": 13867
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0389652383469992,
+      "learning_rate": 3.276242368876664e-07,
+      "loss": 0.5163,
+      "step": 13868
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.572270669850466,
+      "learning_rate": 3.2725531845516744e-07,
+      "loss": 0.6629,
+      "step": 13869
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4138505806134896,
+      "learning_rate": 3.2688660082091594e-07,
+      "loss": 0.5793,
+      "step": 13870
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3320793460273352,
+      "learning_rate": 3.2651808400075426e-07,
+      "loss": 0.5963,
+      "step": 13871
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4531372234860593,
+      "learning_rate": 3.2614976801051745e-07,
+      "loss": 0.6462,
+      "step": 13872
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1744472863371733,
+      "learning_rate": 3.257816528660357e-07,
+      "loss": 0.5971,
+      "step": 13873
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5577445265397432,
+      "learning_rate": 3.254137385831263e-07,
+      "loss": 0.6997,
+      "step": 13874
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5112680226045738,
+      "learning_rate": 3.250460251775994e-07,
+      "loss": 0.6715,
+      "step": 13875
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2904055045000475,
+      "learning_rate": 3.246785126652563e-07,
+      "loss": 0.5796,
+      "step": 13876
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2185644996901193,
+      "learning_rate": 3.243112010618893e-07,
+      "loss": 0.5764,
+      "step": 13877
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.101584275951094,
+      "learning_rate": 3.239440903832852e-07,
+      "loss": 0.5911,
+      "step": 13878
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.377621015355443,
+      "learning_rate": 3.2357718064521594e-07,
+      "loss": 0.5663,
+      "step": 13879
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.411288392312297,
+      "learning_rate": 3.2321047186345047e-07,
+      "loss": 0.5212,
+      "step": 13880
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8252426217840985,
+      "learning_rate": 3.2284396405374787e-07,
+      "loss": 0.7102,
+      "step": 13881
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1741920173403624,
+      "learning_rate": 3.2247765723185556e-07,
+      "loss": 0.5664,
+      "step": 13882
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5160276795664958,
+      "learning_rate": 3.221115514135159e-07,
+      "loss": 0.6331,
+      "step": 13883
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5720914999283198,
+      "learning_rate": 3.217456466144614e-07,
+      "loss": 0.6069,
+      "step": 13884
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.431977845127588,
+      "learning_rate": 3.213799428504155e-07,
+      "loss": 0.6502,
+      "step": 13885
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2148317506105635,
+      "learning_rate": 3.210144401370935e-07,
+      "loss": 0.5703,
+      "step": 13886
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5507791955534467,
+      "learning_rate": 3.206491384902005e-07,
+      "loss": 0.5716,
+      "step": 13887
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.450602270433892,
+      "learning_rate": 3.202840379254374e-07,
+      "loss": 0.619,
+      "step": 13888
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3684959081470334,
+      "learning_rate": 3.199191384584893e-07,
+      "loss": 0.6977,
+      "step": 13889
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0703742515517771,
+      "learning_rate": 3.1955444010504044e-07,
+      "loss": 0.7369,
+      "step": 13890
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5344718022741914,
+      "learning_rate": 3.19189942880761e-07,
+      "loss": 0.7079,
+      "step": 13891
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7143281032882973,
+      "learning_rate": 3.18825646801314e-07,
+      "loss": 0.6141,
+      "step": 13892
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7196489674747206,
+      "learning_rate": 3.184615518823547e-07,
+      "loss": 0.5926,
+      "step": 13893
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2022275036096228,
+      "learning_rate": 3.180976581395295e-07,
+      "loss": 0.541,
+      "step": 13894
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4871173605235193,
+      "learning_rate": 3.177339655884737e-07,
+      "loss": 0.6797,
+      "step": 13895
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1194459371375267,
+      "learning_rate": 3.173704742448186e-07,
+      "loss": 0.6505,
+      "step": 13896
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5135181208297823,
+      "learning_rate": 3.170071841241823e-07,
+      "loss": 0.6123,
+      "step": 13897
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3617864079744404,
+      "learning_rate": 3.1664409524217677e-07,
+      "loss": 0.6237,
+      "step": 13898
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3684629663962224,
+      "learning_rate": 3.1628120761440616e-07,
+      "loss": 0.6057,
+      "step": 13899
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7747420813094237,
+      "learning_rate": 3.1591852125646196e-07,
+      "loss": 0.7267,
+      "step": 13900
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4885009214033893,
+      "learning_rate": 3.1555603618393106e-07,
+      "loss": 0.6245,
+      "step": 13901
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3508923208414194,
+      "learning_rate": 3.151937524123905e-07,
+      "loss": 0.6369,
+      "step": 13902
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.417142046031898,
+      "learning_rate": 3.148316699574078e-07,
+      "loss": 0.575,
+      "step": 13903
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.567886120521376,
+      "learning_rate": 3.144697888345427e-07,
+      "loss": 0.6127,
+      "step": 13904
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6195732581716,
+      "learning_rate": 3.1410810905934564e-07,
+      "loss": 0.6248,
+      "step": 13905
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3974431817282753,
+      "learning_rate": 3.137466306473602e-07,
+      "loss": 0.5868,
+      "step": 13906
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2036553448700689,
+      "learning_rate": 3.1338535361411737e-07,
+      "loss": 0.6146,
+      "step": 13907
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2442073524107702,
+      "learning_rate": 3.1302427797514465e-07,
+      "loss": 0.5617,
+      "step": 13908
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.309851574134321,
+      "learning_rate": 3.1266340374595693e-07,
+      "loss": 0.6193,
+      "step": 13909
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7018079086517035,
+      "learning_rate": 3.1230273094206233e-07,
+      "loss": 0.6021,
+      "step": 13910
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2550883526789327,
+      "learning_rate": 3.1194225957895895e-07,
+      "loss": 0.6358,
+      "step": 13911
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1774486168796456,
+      "learning_rate": 3.115819896721378e-07,
+      "loss": 0.4748,
+      "step": 13912
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.30025966071799,
+      "learning_rate": 3.1122192123707985e-07,
+      "loss": 0.5361,
+      "step": 13913
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.252708198908199,
+      "learning_rate": 3.108620542892593e-07,
+      "loss": 0.6866,
+      "step": 13914
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3614817436497,
+      "learning_rate": 3.105023888441383e-07,
+      "loss": 0.6789,
+      "step": 13915
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.275739454699555,
+      "learning_rate": 3.1014292491717444e-07,
+      "loss": 0.5486,
+      "step": 13916
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.549287959626017,
+      "learning_rate": 3.0978366252381376e-07,
+      "loss": 0.6864,
+      "step": 13917
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3761273458013914,
+      "learning_rate": 3.0942460167949495e-07,
+      "loss": 0.5338,
+      "step": 13918
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.282775518672368,
+      "learning_rate": 3.0906574239964795e-07,
+      "loss": 0.6571,
+      "step": 13919
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3696844468989595,
+      "learning_rate": 3.08707084699692e-07,
+      "loss": 0.6284,
+      "step": 13920
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7414728865548663,
+      "learning_rate": 3.083486285950421e-07,
+      "loss": 0.5917,
+      "step": 13921
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5508935217207003,
+      "learning_rate": 3.0799037410109976e-07,
+      "loss": 0.6969,
+      "step": 13922
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.142177860040557,
+      "learning_rate": 3.076323212332605e-07,
+      "loss": 0.5793,
+      "step": 13923
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2806282047749646,
+      "learning_rate": 3.072744700069119e-07,
+      "loss": 0.6685,
+      "step": 13924
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3035918699062956,
+      "learning_rate": 3.06916820437429e-07,
+      "loss": 0.5352,
+      "step": 13925
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5607033775487629,
+      "learning_rate": 3.065593725401833e-07,
+      "loss": 0.563,
+      "step": 13926
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.562559302017641,
+      "learning_rate": 3.0620212633053424e-07,
+      "loss": 0.6563,
+      "step": 13927
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2623849304448396,
+      "learning_rate": 3.0584508182383346e-07,
+      "loss": 0.6513,
+      "step": 13928
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2079143107220922,
+      "learning_rate": 3.054882390354241e-07,
+      "loss": 0.7724,
+      "step": 13929
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.063516961748299,
+      "learning_rate": 3.0513159798063906e-07,
+      "loss": 0.7027,
+      "step": 13930
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4219365874192829,
+      "learning_rate": 3.0477515867480655e-07,
+      "loss": 0.6257,
+      "step": 13931
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2889066225754169,
+      "learning_rate": 3.0441892113324265e-07,
+      "loss": 0.6004,
+      "step": 13932
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6181672521561765,
+      "learning_rate": 3.0406288537125403e-07,
+      "loss": 0.6908,
+      "step": 13933
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.278552299151766,
+      "learning_rate": 3.0370705140414293e-07,
+      "loss": 0.6052,
+      "step": 13934
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.302641767735573,
+      "learning_rate": 3.0335141924719813e-07,
+      "loss": 0.7457,
+      "step": 13935
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6222730387861142,
+      "learning_rate": 3.029959889157036e-07,
+      "loss": 0.7273,
+      "step": 13936
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4568149004799504,
+      "learning_rate": 3.026407604249315e-07,
+      "loss": 0.7075,
+      "step": 13937
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.234207966712268,
+      "learning_rate": 3.0228573379014745e-07,
+      "loss": 0.5731,
+      "step": 13938
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4964309466174979,
+      "learning_rate": 3.019309090266087e-07,
+      "loss": 0.7056,
+      "step": 13939
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4425881241080192,
+      "learning_rate": 3.015762861495608e-07,
+      "loss": 0.6842,
+      "step": 13940
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.427914287460727,
+      "learning_rate": 3.0122186517424435e-07,
+      "loss": 0.6451,
+      "step": 13941
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3933217150423753,
+      "learning_rate": 3.0086764611589047e-07,
+      "loss": 0.6036,
+      "step": 13942
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0636016436738291,
+      "learning_rate": 3.0051362898971704e-07,
+      "loss": 0.6649,
+      "step": 13943
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1515619256377134,
+      "learning_rate": 3.0015981381094073e-07,
+      "loss": 0.5793,
+      "step": 13944
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6653377146122748,
+      "learning_rate": 2.9980620059476384e-07,
+      "loss": 0.5849,
+      "step": 13945
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.500211879992651,
+      "learning_rate": 2.9945278935638255e-07,
+      "loss": 0.597,
+      "step": 13946
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5347133907332755,
+      "learning_rate": 2.990995801109836e-07,
+      "loss": 0.6273,
+      "step": 13947
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2307778372198759,
+      "learning_rate": 2.987465728737443e-07,
+      "loss": 0.4749,
+      "step": 13948
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4508139002940164,
+      "learning_rate": 2.9839376765983583e-07,
+      "loss": 0.6228,
+      "step": 13949
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3894501871310587,
+      "learning_rate": 2.980411644844189e-07,
+      "loss": 0.6389,
+      "step": 13950
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.076281638688788,
+      "learning_rate": 2.976887633626435e-07,
+      "loss": 0.655,
+      "step": 13951
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3637308474237058,
+      "learning_rate": 2.973365643096554e-07,
+      "loss": 0.6392,
+      "step": 13952
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.405154830499082,
+      "learning_rate": 2.9698456734058856e-07,
+      "loss": 0.5685,
+      "step": 13953
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4858482779768594,
+      "learning_rate": 2.9663277247056923e-07,
+      "loss": 0.6165,
+      "step": 13954
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2978433067798485,
+      "learning_rate": 2.9628117971471416e-07,
+      "loss": 0.6365,
+      "step": 13955
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5177654358513324,
+      "learning_rate": 2.959297890881319e-07,
+      "loss": 0.6566,
+      "step": 13956
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3840345135018066,
+      "learning_rate": 2.955786006059247e-07,
+      "loss": 0.6097,
+      "step": 13957
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5566213362129302,
+      "learning_rate": 2.952276142831806e-07,
+      "loss": 0.5895,
+      "step": 13958
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.543654262035238,
+      "learning_rate": 2.9487683013498523e-07,
+      "loss": 0.6806,
+      "step": 13959
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3490352915891068,
+      "learning_rate": 2.9452624817641163e-07,
+      "loss": 0.6953,
+      "step": 13960
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3886929141309803,
+      "learning_rate": 2.9417586842252375e-07,
+      "loss": 0.5367,
+      "step": 13961
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4222299993098286,
+      "learning_rate": 2.9382569088837966e-07,
+      "loss": 0.6928,
+      "step": 13962
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2559160085616337,
+      "learning_rate": 2.934757155890272e-07,
+      "loss": 0.718,
+      "step": 13963
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.542306652712392,
+      "learning_rate": 2.93125942539505e-07,
+      "loss": 0.7192,
+      "step": 13964
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5714725590770158,
+      "learning_rate": 2.9277637175484376e-07,
+      "loss": 0.6011,
+      "step": 13965
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3822477764433196,
+      "learning_rate": 2.924270032500648e-07,
+      "loss": 0.5977,
+      "step": 13966
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.404186978603986,
+      "learning_rate": 2.9207783704018223e-07,
+      "loss": 0.6673,
+      "step": 13967
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.263067676669863,
+      "learning_rate": 2.9172887314020017e-07,
+      "loss": 0.6231,
+      "step": 13968
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2704558701242197,
+      "learning_rate": 2.913801115651144e-07,
+      "loss": 0.5823,
+      "step": 13969
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3183329747357875,
+      "learning_rate": 2.910315523299123e-07,
+      "loss": 0.701,
+      "step": 13970
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2321814908429525,
+      "learning_rate": 2.906831954495715e-07,
+      "loss": 0.5685,
+      "step": 13971
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.58677235869384,
+      "learning_rate": 2.9033504093906207e-07,
+      "loss": 0.6898,
+      "step": 13972
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1702483192159727,
+      "learning_rate": 2.8998708881334437e-07,
+      "loss": 0.713,
+      "step": 13973
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7290462428901947,
+      "learning_rate": 2.896393390873714e-07,
+      "loss": 0.5864,
+      "step": 13974
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3961429322063568,
+      "learning_rate": 2.892917917760873e-07,
+      "loss": 0.6083,
+      "step": 13975
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3792568584334426,
+      "learning_rate": 2.8894444689442526e-07,
+      "loss": 0.6048,
+      "step": 13976
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2495319782773078,
+      "learning_rate": 2.885973044573126e-07,
+      "loss": 0.6571,
+      "step": 13977
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.422900637179702,
+      "learning_rate": 2.8825036447966745e-07,
+      "loss": 0.6094,
+      "step": 13978
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5088165376294798,
+      "learning_rate": 2.8790362697639685e-07,
+      "loss": 0.7475,
+      "step": 13979
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3538803578787775,
+      "learning_rate": 2.8755709196240264e-07,
+      "loss": 0.648,
+      "step": 13980
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5189365340061691,
+      "learning_rate": 2.8721075945257414e-07,
+      "loss": 0.5993,
+      "step": 13981
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3881687816524193,
+      "learning_rate": 2.868646294617966e-07,
+      "loss": 0.687,
+      "step": 13982
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.08594707655408,
+      "learning_rate": 2.86518702004942e-07,
+      "loss": 0.6567,
+      "step": 13983
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5828376340518395,
+      "learning_rate": 2.8617297709687577e-07,
+      "loss": 0.7594,
+      "step": 13984
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2525370002595289,
+      "learning_rate": 2.858274547524559e-07,
+      "loss": 0.7888,
+      "step": 13985
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.348410609671379,
+      "learning_rate": 2.854821349865289e-07,
+      "loss": 0.6213,
+      "step": 13986
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7815628561350507,
+      "learning_rate": 2.8513701781393456e-07,
+      "loss": 0.5499,
+      "step": 13987
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1910063389376786,
+      "learning_rate": 2.8479210324950266e-07,
+      "loss": 0.7222,
+      "step": 13988
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.113903419543021,
+      "learning_rate": 2.8444739130805587e-07,
+      "loss": 0.7073,
+      "step": 13989
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.579877032555118,
+      "learning_rate": 2.841028820044067e-07,
+      "loss": 0.6429,
+      "step": 13990
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2789706278228743,
+      "learning_rate": 2.837585753533589e-07,
+      "loss": 0.5547,
+      "step": 13991
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3355190554994991,
+      "learning_rate": 2.8341447136970954e-07,
+      "loss": 0.5672,
+      "step": 13992
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4818907998090982,
+      "learning_rate": 2.8307057006824514e-07,
+      "loss": 0.6718,
+      "step": 13993
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.361333079917159,
+      "learning_rate": 2.827268714637421e-07,
+      "loss": 0.6132,
+      "step": 13994
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1553340725656502,
+      "learning_rate": 2.823833755709721e-07,
+      "loss": 0.5149,
+      "step": 13995
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3451927354625786,
+      "learning_rate": 2.820400824046954e-07,
+      "loss": 0.5308,
+      "step": 13996
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4724803249308989,
+      "learning_rate": 2.816969919796636e-07,
+      "loss": 0.6761,
+      "step": 13997
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5290454627161623,
+      "learning_rate": 2.813541043106205e-07,
+      "loss": 0.6631,
+      "step": 13998
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4184274363435994,
+      "learning_rate": 2.810114194122998e-07,
+      "loss": 0.6158,
+      "step": 13999
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.354131728275325,
+      "learning_rate": 2.806689372994292e-07,
+      "loss": 0.6587,
+      "step": 14000
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2995906778466644,
+      "learning_rate": 2.803266579867242e-07,
+      "loss": 0.6363,
+      "step": 14001
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3265450240527639,
+      "learning_rate": 2.7998458148889354e-07,
+      "loss": 0.6438,
+      "step": 14002
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3757777453588793,
+      "learning_rate": 2.796427078206382e-07,
+      "loss": 0.5567,
+      "step": 14003
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3966506444279898,
+      "learning_rate": 2.793010369966487e-07,
+      "loss": 0.6055,
+      "step": 14004
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2938958212114418,
+      "learning_rate": 2.789595690316066e-07,
+      "loss": 0.6982,
+      "step": 14005
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5704745791923267,
+      "learning_rate": 2.7861830394018627e-07,
+      "loss": 0.6382,
+      "step": 14006
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3295554795135818,
+      "learning_rate": 2.7827724173705273e-07,
+      "loss": 0.7218,
+      "step": 14007
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4284836273708987,
+      "learning_rate": 2.779363824368619e-07,
+      "loss": 0.5867,
+      "step": 14008
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3332605929108134,
+      "learning_rate": 2.7759572605426057e-07,
+      "loss": 0.656,
+      "step": 14009
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2486491548951772,
+      "learning_rate": 2.7725527260388905e-07,
+      "loss": 0.587,
+      "step": 14010
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5276286659705518,
+      "learning_rate": 2.7691502210037744e-07,
+      "loss": 0.6936,
+      "step": 14011
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.468147536174512,
+      "learning_rate": 2.7657497455834457e-07,
+      "loss": 0.6764,
+      "step": 14012
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6532800545944937,
+      "learning_rate": 2.762351299924054e-07,
+      "loss": 0.622,
+      "step": 14013
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6418347829295543,
+      "learning_rate": 2.7589548841716274e-07,
+      "loss": 0.7074,
+      "step": 14014
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3285779124427815,
+      "learning_rate": 2.755560498472126e-07,
+      "loss": 0.6037,
+      "step": 14015
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2443532688758157,
+      "learning_rate": 2.752168142971406e-07,
+      "loss": 0.6539,
+      "step": 14016
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4680811216119771,
+      "learning_rate": 2.7487778178152457e-07,
+      "loss": 0.6772,
+      "step": 14017
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3683217072117697,
+      "learning_rate": 2.7453895231493445e-07,
+      "loss": 0.5924,
+      "step": 14018
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5457908402005334,
+      "learning_rate": 2.7420032591192856e-07,
+      "loss": 0.661,
+      "step": 14019
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3251769796164006,
+      "learning_rate": 2.7386190258706033e-07,
+      "loss": 0.5878,
+      "step": 14020
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.366679716857148,
+      "learning_rate": 2.735236823548715e-07,
+      "loss": 0.6671,
+      "step": 14021
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.718734382679096,
+      "learning_rate": 2.73185665229897e-07,
+      "loss": 0.5751,
+      "step": 14022
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1417293803798418,
+      "learning_rate": 2.7284785122666136e-07,
+      "loss": 0.7055,
+      "step": 14023
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2416515387372282,
+      "learning_rate": 2.7251024035968134e-07,
+      "loss": 0.6384,
+      "step": 14024
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.48855429688634,
+      "learning_rate": 2.721728326434653e-07,
+      "loss": 0.6649,
+      "step": 14025
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5043516206742353,
+      "learning_rate": 2.718356280925116e-07,
+      "loss": 0.7241,
+      "step": 14026
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5106594799876423,
+      "learning_rate": 2.71498626721311e-07,
+      "loss": 0.6887,
+      "step": 14027
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4739224039587728,
+      "learning_rate": 2.711618285443457e-07,
+      "loss": 0.7051,
+      "step": 14028
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2786436015428126,
+      "learning_rate": 2.7082523357608856e-07,
+      "loss": 0.7007,
+      "step": 14029
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.425670018711619,
+      "learning_rate": 2.7048884183100313e-07,
+      "loss": 0.7103,
+      "step": 14030
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1335298000090857,
+      "learning_rate": 2.70152653323546e-07,
+      "loss": 0.691,
+      "step": 14031
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3491288818996838,
+      "learning_rate": 2.6981666806816185e-07,
+      "loss": 0.6192,
+      "step": 14032
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2978289775500635,
+      "learning_rate": 2.694808860792919e-07,
+      "loss": 0.6893,
+      "step": 14033
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4624547888708337,
+      "learning_rate": 2.6914530737136346e-07,
+      "loss": 0.68,
+      "step": 14034
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2206817943462662,
+      "learning_rate": 2.6880993195879614e-07,
+      "loss": 0.6251,
+      "step": 14035
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3109407357172953,
+      "learning_rate": 2.684747598560045e-07,
+      "loss": 0.7257,
+      "step": 14036
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2199584511419646,
+      "learning_rate": 2.6813979107738864e-07,
+      "loss": 0.5441,
+      "step": 14037
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4466388091387163,
+      "learning_rate": 2.678050256373449e-07,
+      "loss": 0.6129,
+      "step": 14038
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.731675482204236,
+      "learning_rate": 2.674704635502584e-07,
+      "loss": 0.612,
+      "step": 14039
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2119310197938824,
+      "learning_rate": 2.6713610483050654e-07,
+      "loss": 0.5619,
+      "step": 14040
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0180833304658337,
+      "learning_rate": 2.668019494924562e-07,
+      "loss": 0.6354,
+      "step": 14041
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.77100903869668,
+      "learning_rate": 2.6646799755046746e-07,
+      "loss": 0.6412,
+      "step": 14042
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4970329331579497,
+      "learning_rate": 2.661342490188917e-07,
+      "loss": 0.6723,
+      "step": 14043
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4223013299306906,
+      "learning_rate": 2.658007039120697e-07,
+      "loss": 0.7301,
+      "step": 14044
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1900988802941563,
+      "learning_rate": 2.654673622443349e-07,
+      "loss": 0.6462,
+      "step": 14045
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4481137606136758,
+      "learning_rate": 2.65134224030012e-07,
+      "loss": 0.5409,
+      "step": 14046
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3941566748195626,
+      "learning_rate": 2.6480128928341675e-07,
+      "loss": 0.6948,
+      "step": 14047
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3142789882352068,
+      "learning_rate": 2.6446855801885606e-07,
+      "loss": 0.6569,
+      "step": 14048
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.119911397544226,
+      "learning_rate": 2.64136030250628e-07,
+      "loss": 0.6114,
+      "step": 14049
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0694483170706126,
+      "learning_rate": 2.638037059930215e-07,
+      "loss": 0.7658,
+      "step": 14050
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.384412804569858,
+      "learning_rate": 2.634715852603187e-07,
+      "loss": 0.648,
+      "step": 14051
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4194607872146914,
+      "learning_rate": 2.631396680667908e-07,
+      "loss": 0.6964,
+      "step": 14052
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7172609187848318,
+      "learning_rate": 2.6280795442669925e-07,
+      "loss": 0.6178,
+      "step": 14053
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3608348371343577,
+      "learning_rate": 2.6247644435430263e-07,
+      "loss": 0.6305,
+      "step": 14054
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.810157636795424,
+      "learning_rate": 2.621451378638418e-07,
+      "loss": 0.7627,
+      "step": 14055
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2539595911537844,
+      "learning_rate": 2.618140349695575e-07,
+      "loss": 0.6259,
+      "step": 14056
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6706296775465443,
+      "learning_rate": 2.6148313568567674e-07,
+      "loss": 0.5803,
+      "step": 14057
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2843324797746831,
+      "learning_rate": 2.611524400264187e-07,
+      "loss": 0.5986,
+      "step": 14058
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4575477330269095,
+      "learning_rate": 2.6082194800599424e-07,
+      "loss": 0.5395,
+      "step": 14059
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.116877146570642,
+      "learning_rate": 2.6049165963860425e-07,
+      "loss": 0.7766,
+      "step": 14060
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0973954476130285,
+      "learning_rate": 2.601615749384451e-07,
+      "loss": 0.6304,
+      "step": 14061
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.467644558795657,
+      "learning_rate": 2.5983169391969823e-07,
+      "loss": 0.6658,
+      "step": 14062
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2459836171070493,
+      "learning_rate": 2.595020165965401e-07,
+      "loss": 0.5733,
+      "step": 14063
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2368669192355322,
+      "learning_rate": 2.591725429831382e-07,
+      "loss": 0.5564,
+      "step": 14064
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4445928572332112,
+      "learning_rate": 2.5884327309365074e-07,
+      "loss": 0.5888,
+      "step": 14065
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0934982218214306,
+      "learning_rate": 2.5851420694222697e-07,
+      "loss": 0.4814,
+      "step": 14066
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.354184037786973,
+      "learning_rate": 2.581853445430077e-07,
+      "loss": 0.6089,
+      "step": 14067
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.516342580953839,
+      "learning_rate": 2.5785668591012393e-07,
+      "loss": 0.7964,
+      "step": 14068
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6789194144899082,
+      "learning_rate": 2.57528231057701e-07,
+      "loss": 0.6053,
+      "step": 14069
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3990228219031162,
+      "learning_rate": 2.571999799998509e-07,
+      "loss": 0.6421,
+      "step": 14070
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5021699647661506,
+      "learning_rate": 2.568719327506808e-07,
+      "loss": 0.612,
+      "step": 14071
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.368479572912971,
+      "learning_rate": 2.565440893242882e-07,
+      "loss": 0.621,
+      "step": 14072
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5142686411762463,
+      "learning_rate": 2.5621644973475914e-07,
+      "loss": 0.7849,
+      "step": 14073
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.425958371594337,
+      "learning_rate": 2.558890139961745e-07,
+      "loss": 0.6532,
+      "step": 14074
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2607917969923583,
+      "learning_rate": 2.5556178212260474e-07,
+      "loss": 0.6635,
+      "step": 14075
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.126398751124451,
+      "learning_rate": 2.552347541281114e-07,
+      "loss": 0.6847,
+      "step": 14076
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0678209036131472,
+      "learning_rate": 2.549079300267482e-07,
+      "loss": 0.5816,
+      "step": 14077
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4892522664287735,
+      "learning_rate": 2.5458130983255835e-07,
+      "loss": 0.5368,
+      "step": 14078
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.239266491084396,
+      "learning_rate": 2.5425489355957956e-07,
+      "loss": 0.6853,
+      "step": 14079
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5694698494560668,
+      "learning_rate": 2.539286812218361e-07,
+      "loss": 0.5188,
+      "step": 14080
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9497707244119943,
+      "learning_rate": 2.5360267283334796e-07,
+      "loss": 0.555,
+      "step": 14081
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4378330943028688,
+      "learning_rate": 2.5327686840812326e-07,
+      "loss": 0.7011,
+      "step": 14082
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6650140858317237,
+      "learning_rate": 2.5295126796016366e-07,
+      "loss": 0.6959,
+      "step": 14083
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5132094136877088,
+      "learning_rate": 2.526258715034602e-07,
+      "loss": 0.7117,
+      "step": 14084
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4374415750872656,
+      "learning_rate": 2.523006790519955e-07,
+      "loss": 0.6471,
+      "step": 14085
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2063443230411524,
+      "learning_rate": 2.5197569061974404e-07,
+      "loss": 0.6039,
+      "step": 14086
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.790713743023204,
+      "learning_rate": 2.5165090622067343e-07,
+      "loss": 0.5715,
+      "step": 14087
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3617602627562087,
+      "learning_rate": 2.513263258687365e-07,
+      "loss": 0.5809,
+      "step": 14088
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4002475677912645,
+      "learning_rate": 2.510019495778837e-07,
+      "loss": 0.6215,
+      "step": 14089
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2999301712941955,
+      "learning_rate": 2.5067777736205444e-07,
+      "loss": 0.6051,
+      "step": 14090
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0779327237743013,
+      "learning_rate": 2.503538092351782e-07,
+      "loss": 0.5346,
+      "step": 14091
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.466575786415882,
+      "learning_rate": 2.5003004521117646e-07,
+      "loss": 0.7927,
+      "step": 14092
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2289311104401883,
+      "learning_rate": 2.4970648530396157e-07,
+      "loss": 0.4781,
+      "step": 14093
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2882110399415194,
+      "learning_rate": 2.4938312952744016e-07,
+      "loss": 0.5613,
+      "step": 14094
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3085157630634285,
+      "learning_rate": 2.490599778955055e-07,
+      "loss": 0.6877,
+      "step": 14095
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.137793989762916,
+      "learning_rate": 2.4873703042204323e-07,
+      "loss": 0.643,
+      "step": 14096
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.478283360961364,
+      "learning_rate": 2.484142871209344e-07,
+      "loss": 0.6827,
+      "step": 14097
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0998473963630861,
+      "learning_rate": 2.480917480060441e-07,
+      "loss": 0.6785,
+      "step": 14098
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.307658547684547,
+      "learning_rate": 2.477694130912356e-07,
+      "loss": 0.5484,
+      "step": 14099
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3429285741612234,
+      "learning_rate": 2.474472823903584e-07,
+      "loss": 0.6548,
+      "step": 14100
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.815829906229226,
+      "learning_rate": 2.4712535591725706e-07,
+      "loss": 0.6958,
+      "step": 14101
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4188050116218516,
+      "learning_rate": 2.4680363368576376e-07,
+      "loss": 0.583,
+      "step": 14102
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2747600861610582,
+      "learning_rate": 2.464821157097036e-07,
+      "loss": 0.5942,
+      "step": 14103
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.499765281158626,
+      "learning_rate": 2.461608020028944e-07,
+      "loss": 0.6687,
+      "step": 14104
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.243600265112361,
+      "learning_rate": 2.458396925791434e-07,
+      "loss": 0.6562,
+      "step": 14105
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.499995495436273,
+      "learning_rate": 2.4551878745224745e-07,
+      "loss": 0.6077,
+      "step": 14106
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.429528771638644,
+      "learning_rate": 2.4519808663599876e-07,
+      "loss": 0.6288,
+      "step": 14107
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3657549061932954,
+      "learning_rate": 2.44877590144178e-07,
+      "loss": 0.6222,
+      "step": 14108
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4482939943408273,
+      "learning_rate": 2.445572979905575e-07,
+      "loss": 0.6944,
+      "step": 14109
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3348505685040994,
+      "learning_rate": 2.442372101889007e-07,
+      "loss": 0.7208,
+      "step": 14110
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.493756519070979,
+      "learning_rate": 2.4391732675296207e-07,
+      "loss": 0.6618,
+      "step": 14111
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2023118235891577,
+      "learning_rate": 2.4359764769648907e-07,
+      "loss": 0.6695,
+      "step": 14112
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2502113184859402,
+      "learning_rate": 2.4327817303321786e-07,
+      "loss": 0.6882,
+      "step": 14113
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.364237582689243,
+      "learning_rate": 2.4295890277687695e-07,
+      "loss": 0.6458,
+      "step": 14114
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7436760978284773,
+      "learning_rate": 2.4263983694118754e-07,
+      "loss": 0.7294,
+      "step": 14115
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3071603221563264,
+      "learning_rate": 2.423209755398587e-07,
+      "loss": 0.5543,
+      "step": 14116
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.328905411679625,
+      "learning_rate": 2.4200231858659384e-07,
+      "loss": 0.5288,
+      "step": 14117
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5991598134644789,
+      "learning_rate": 2.416838660950854e-07,
+      "loss": 0.5834,
+      "step": 14118
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3419914527206616,
+      "learning_rate": 2.4136561807901916e-07,
+      "loss": 0.6876,
+      "step": 14119
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.61792116119808,
+      "learning_rate": 2.410475745520702e-07,
+      "loss": 0.6502,
+      "step": 14120
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3516546379432002,
+      "learning_rate": 2.4072973552790535e-07,
+      "loss": 0.5774,
+      "step": 14121
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3674809298147,
+      "learning_rate": 2.4041210102018375e-07,
+      "loss": 0.6961,
+      "step": 14122
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4119340534082276,
+      "learning_rate": 2.40094671042555e-07,
+      "loss": 0.5728,
+      "step": 14123
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.216623919517199,
+      "learning_rate": 2.397774456086577e-07,
+      "loss": 0.6501,
+      "step": 14124
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1973196857569386,
+      "learning_rate": 2.3946042473212526e-07,
+      "loss": 0.5693,
+      "step": 14125
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2357851346304363,
+      "learning_rate": 2.391436084265814e-07,
+      "loss": 0.6104,
+      "step": 14126
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4367502760070978,
+      "learning_rate": 2.3882699670563947e-07,
+      "loss": 0.5729,
+      "step": 14127
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3367738285849045,
+      "learning_rate": 2.385105895829054e-07,
+      "loss": 0.544,
+      "step": 14128
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2739757312770787,
+      "learning_rate": 2.3819438707197495e-07,
+      "loss": 0.6733,
+      "step": 14129
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.9252092168545847,
+      "learning_rate": 2.378783891864378e-07,
+      "loss": 0.6401,
+      "step": 14130
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.596677849203751,
+      "learning_rate": 2.3756259593987086e-07,
+      "loss": 0.6862,
+      "step": 14131
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2618772125143796,
+      "learning_rate": 2.372470073458466e-07,
+      "loss": 0.5132,
+      "step": 14132
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.400309159906369,
+      "learning_rate": 2.3693162341792532e-07,
+      "loss": 0.648,
+      "step": 14133
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6845091214292567,
+      "learning_rate": 2.3661644416966057e-07,
+      "loss": 0.6926,
+      "step": 14134
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.344818048469335,
+      "learning_rate": 2.3630146961459543e-07,
+      "loss": 0.6027,
+      "step": 14135
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4167827085204114,
+      "learning_rate": 2.3598669976626576e-07,
+      "loss": 0.6919,
+      "step": 14136
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4606900898919648,
+      "learning_rate": 2.3567213463819738e-07,
+      "loss": 0.6379,
+      "step": 14137
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4085161454048878,
+      "learning_rate": 2.353577742439084e-07,
+      "loss": 0.7208,
+      "step": 14138
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.554268588520786,
+      "learning_rate": 2.3504361859690628e-07,
+      "loss": 0.6133,
+      "step": 14139
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4072170361488534,
+      "learning_rate": 2.347296677106925e-07,
+      "loss": 0.8424,
+      "step": 14140
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3342217858031868,
+      "learning_rate": 2.344159215987585e-07,
+      "loss": 0.646,
+      "step": 14141
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6411093048413103,
+      "learning_rate": 2.341023802745851e-07,
+      "loss": 0.623,
+      "step": 14142
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4897031641893035,
+      "learning_rate": 2.3378904375164714e-07,
+      "loss": 0.7059,
+      "step": 14143
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1348990306652276,
+      "learning_rate": 2.3347591204340881e-07,
+      "loss": 0.644,
+      "step": 14144
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1416240973402672,
+      "learning_rate": 2.3316298516332604e-07,
+      "loss": 0.4987,
+      "step": 14145
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.606552434384712,
+      "learning_rate": 2.3285026312484583e-07,
+      "loss": 0.6527,
+      "step": 14146
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4876915486633164,
+      "learning_rate": 2.3253774594140633e-07,
+      "loss": 0.7064,
+      "step": 14147
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8194182970006765,
+      "learning_rate": 2.32225433626439e-07,
+      "loss": 0.6139,
+      "step": 14148
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.394510368084764,
+      "learning_rate": 2.3191332619336204e-07,
+      "loss": 0.6701,
+      "step": 14149
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2657249045322017,
+      "learning_rate": 2.3160142365558914e-07,
+      "loss": 0.5846,
+      "step": 14150
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.252954647177413,
+      "learning_rate": 2.312897260265229e-07,
+      "loss": 0.6463,
+      "step": 14151
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4779544457562137,
+      "learning_rate": 2.309782333195576e-07,
+      "loss": 0.6867,
+      "step": 14152
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.52201563315751,
+      "learning_rate": 2.3066694554807867e-07,
+      "loss": 0.679,
+      "step": 14153
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4811453395010807,
+      "learning_rate": 2.3035586272546207e-07,
+      "loss": 0.709,
+      "step": 14154
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.468407821468986,
+      "learning_rate": 2.300449848650782e-07,
+      "loss": 0.6348,
+      "step": 14155
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3081176770868639,
+      "learning_rate": 2.2973431198028417e-07,
+      "loss": 0.6279,
+      "step": 14156
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0200649830369741,
+      "learning_rate": 2.2942384408442986e-07,
+      "loss": 0.6022,
+      "step": 14157
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0570286181630986,
+      "learning_rate": 2.2911358119085792e-07,
+      "loss": 0.5776,
+      "step": 14158
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.487593410142504,
+      "learning_rate": 2.2880352331290102e-07,
+      "loss": 0.6688,
+      "step": 14159
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3397167038725217,
+      "learning_rate": 2.2849367046388294e-07,
+      "loss": 0.6213,
+      "step": 14160
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3519052379703,
+      "learning_rate": 2.2818402265711858e-07,
+      "loss": 0.5927,
+      "step": 14161
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.455358661431255,
+      "learning_rate": 2.27874579905914e-07,
+      "loss": 0.6935,
+      "step": 14162
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.415704607604364,
+      "learning_rate": 2.2756534222356687e-07,
+      "loss": 0.7052,
+      "step": 14163
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4028944520071085,
+      "learning_rate": 2.2725630962336542e-07,
+      "loss": 0.6471,
+      "step": 14164
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3476227185798266,
+      "learning_rate": 2.2694748211858964e-07,
+      "loss": 0.5903,
+      "step": 14165
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.316983234899001,
+      "learning_rate": 2.2663885972251165e-07,
+      "loss": 0.6241,
+      "step": 14166
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4373821856848648,
+      "learning_rate": 2.2633044244839143e-07,
+      "loss": 0.5769,
+      "step": 14167
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3564774346587292,
+      "learning_rate": 2.2602223030948445e-07,
+      "loss": 0.5658,
+      "step": 14168
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1792463357724605,
+      "learning_rate": 2.2571422331903458e-07,
+      "loss": 0.5521,
+      "step": 14169
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2559569821490784,
+      "learning_rate": 2.2540642149027735e-07,
+      "loss": 0.5936,
+      "step": 14170
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3392873582899567,
+      "learning_rate": 2.2509882483643942e-07,
+      "loss": 0.5737,
+      "step": 14171
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2801750979989095,
+      "learning_rate": 2.2479143337073904e-07,
+      "loss": 0.6007,
+      "step": 14172
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4108154719305817,
+      "learning_rate": 2.2448424710638684e-07,
+      "loss": 0.7586,
+      "step": 14173
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3454866169230066,
+      "learning_rate": 2.2417726605658164e-07,
+      "loss": 0.5911,
+      "step": 14174
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0845909140019914,
+      "learning_rate": 2.2387049023451458e-07,
+      "loss": 0.5195,
+      "step": 14175
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4790478630270627,
+      "learning_rate": 2.2356391965337064e-07,
+      "loss": 0.5392,
+      "step": 14176
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.66637918431735,
+      "learning_rate": 2.2325755432632267e-07,
+      "loss": 0.7823,
+      "step": 14177
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.471171419999833,
+      "learning_rate": 2.2295139426653622e-07,
+      "loss": 0.6334,
+      "step": 14178
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4891160900340215,
+      "learning_rate": 2.226454394871669e-07,
+      "loss": 0.6495,
+      "step": 14179
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3137160045385732,
+      "learning_rate": 2.2233969000136357e-07,
+      "loss": 0.5723,
+      "step": 14180
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6284635926826494,
+      "learning_rate": 2.2203414582226357e-07,
+      "loss": 0.5861,
+      "step": 14181
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4587699711378954,
+      "learning_rate": 2.2172880696299692e-07,
+      "loss": 0.6398,
+      "step": 14182
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3369760382760258,
+      "learning_rate": 2.2142367343668592e-07,
+      "loss": 0.6411,
+      "step": 14183
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2795280616847455,
+      "learning_rate": 2.2111874525644228e-07,
+      "loss": 0.686,
+      "step": 14184
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.33358701365021,
+      "learning_rate": 2.2081402243536833e-07,
+      "loss": 0.6076,
+      "step": 14185
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3753143738771545,
+      "learning_rate": 2.2050950498656078e-07,
+      "loss": 0.6044,
+      "step": 14186
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.605348952700997,
+      "learning_rate": 2.2020519292310361e-07,
+      "loss": 0.6048,
+      "step": 14187
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2783679202660891,
+      "learning_rate": 2.199010862580747e-07,
+      "loss": 0.7253,
+      "step": 14188
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2176590709374098,
+      "learning_rate": 2.1959718500454196e-07,
+      "loss": 0.6885,
+      "step": 14189
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.511145743334303,
+      "learning_rate": 2.1929348917556436e-07,
+      "loss": 0.6453,
+      "step": 14190
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.333807489373787,
+      "learning_rate": 2.189899987841937e-07,
+      "loss": 0.6175,
+      "step": 14191
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4430163978644883,
+      "learning_rate": 2.186867138434695e-07,
+      "loss": 0.6232,
+      "step": 14192
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6534637126370473,
+      "learning_rate": 2.183836343664264e-07,
+      "loss": 0.6585,
+      "step": 14193
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5537792615713115,
+      "learning_rate": 2.1808076036608783e-07,
+      "loss": 0.6526,
+      "step": 14194
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2573309552455298,
+      "learning_rate": 2.1777809185546894e-07,
+      "loss": 0.6191,
+      "step": 14195
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4651972037036716,
+      "learning_rate": 2.17475628847576e-07,
+      "loss": 0.5454,
+      "step": 14196
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5077098644320344,
+      "learning_rate": 2.171733713554064e-07,
+      "loss": 0.6596,
+      "step": 14197
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2678276471781929,
+      "learning_rate": 2.1687131939194862e-07,
+      "loss": 0.6038,
+      "step": 14198
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3785187335037543,
+      "learning_rate": 2.165694729701834e-07,
+      "loss": 0.6269,
+      "step": 14199
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6850235426306854,
+      "learning_rate": 2.1626783210308032e-07,
+      "loss": 0.6518,
+      "step": 14200
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5245409561440888,
+      "learning_rate": 2.15966396803603e-07,
+      "loss": 0.6886,
+      "step": 14201
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4241837498903762,
+      "learning_rate": 2.1566516708470487e-07,
+      "loss": 0.5573,
+      "step": 14202
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4529591248565703,
+      "learning_rate": 2.1536414295932896e-07,
+      "loss": 0.7591,
+      "step": 14203
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0916898993313477,
+      "learning_rate": 2.1506332444041212e-07,
+      "loss": 0.6061,
+      "step": 14204
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5163675960413128,
+      "learning_rate": 2.1476271154088013e-07,
+      "loss": 0.6461,
+      "step": 14205
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3339426600238906,
+      "learning_rate": 2.1446230427365323e-07,
+      "loss": 0.6288,
+      "step": 14206
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5338739786056246,
+      "learning_rate": 2.1416210265163828e-07,
+      "loss": 0.5801,
+      "step": 14207
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3499578587904422,
+      "learning_rate": 2.1386210668773554e-07,
+      "loss": 0.5687,
+      "step": 14208
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9736117815447303,
+      "learning_rate": 2.1356231639483917e-07,
+      "loss": 0.5929,
+      "step": 14209
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.26572079994536,
+      "learning_rate": 2.1326273178582822e-07,
+      "loss": 0.5758,
+      "step": 14210
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6002255027595649,
+      "learning_rate": 2.1296335287357918e-07,
+      "loss": 0.6273,
+      "step": 14211
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2723209053952782,
+      "learning_rate": 2.1266417967095665e-07,
+      "loss": 0.5596,
+      "step": 14212
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4525728269742264,
+      "learning_rate": 2.123652121908154e-07,
+      "loss": 0.6254,
+      "step": 14213
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.519049712219173,
+      "learning_rate": 2.1206645044600404e-07,
+      "loss": 0.5568,
+      "step": 14214
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6851651387085038,
+      "learning_rate": 2.1176789444936065e-07,
+      "loss": 0.6984,
+      "step": 14215
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0862230644689819,
+      "learning_rate": 2.114695442137149e-07,
+      "loss": 0.7011,
+      "step": 14216
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4780186816289247,
+      "learning_rate": 2.1117139975188716e-07,
+      "loss": 0.6075,
+      "step": 14217
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7877570782703898,
+      "learning_rate": 2.1087346107668937e-07,
+      "loss": 0.5011,
+      "step": 14218
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1451623336960106,
+      "learning_rate": 2.1057572820092576e-07,
+      "loss": 0.6544,
+      "step": 14219
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4161979220994936,
+      "learning_rate": 2.1027820113738884e-07,
+      "loss": 0.5619,
+      "step": 14220
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4913898081479546,
+      "learning_rate": 2.099808798988656e-07,
+      "loss": 0.6192,
+      "step": 14221
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2115181168631124,
+      "learning_rate": 2.0968376449813198e-07,
+      "loss": 0.6078,
+      "step": 14222
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.10223950128014,
+      "learning_rate": 2.093868549479544e-07,
+      "loss": 0.5385,
+      "step": 14223
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.363647624027204,
+      "learning_rate": 2.0909015126109488e-07,
+      "loss": 0.6576,
+      "step": 14224
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3083357350870477,
+      "learning_rate": 2.0879365345030044e-07,
+      "loss": 0.5842,
+      "step": 14225
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1364473271713016,
+      "learning_rate": 2.0849736152831256e-07,
+      "loss": 0.7317,
+      "step": 14226
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1732158568871809,
+      "learning_rate": 2.0820127550786551e-07,
+      "loss": 0.5988,
+      "step": 14227
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3420253522234955,
+      "learning_rate": 2.0790539540168074e-07,
+      "loss": 0.7349,
+      "step": 14228
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5180172972895103,
+      "learning_rate": 2.0760972122247425e-07,
+      "loss": 0.6117,
+      "step": 14229
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.299283920815701,
+      "learning_rate": 2.0731425298295083e-07,
+      "loss": 0.5814,
+      "step": 14230
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2649324710960417,
+      "learning_rate": 2.070189906958081e-07,
+      "loss": 0.6159,
+      "step": 14231
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4925804749887608,
+      "learning_rate": 2.0672393437373372e-07,
+      "loss": 0.586,
+      "step": 14232
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.51928590771092,
+      "learning_rate": 2.06429084029407e-07,
+      "loss": 0.6543,
+      "step": 14233
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4286910124228749,
+      "learning_rate": 2.061344396754994e-07,
+      "loss": 0.712,
+      "step": 14234
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3443756564827813,
+      "learning_rate": 2.058400013246703e-07,
+      "loss": 0.6475,
+      "step": 14235
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4743157347551359,
+      "learning_rate": 2.0554576898957346e-07,
+      "loss": 0.5601,
+      "step": 14236
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1075251963074906,
+      "learning_rate": 2.052517426828532e-07,
+      "loss": 0.6914,
+      "step": 14237
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2143658305282279,
+      "learning_rate": 2.0495792241714386e-07,
+      "loss": 0.6625,
+      "step": 14238
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.70402964475404,
+      "learning_rate": 2.04664308205072e-07,
+      "loss": 0.5948,
+      "step": 14239
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2669493175436493,
+      "learning_rate": 2.0437090005925475e-07,
+      "loss": 0.6357,
+      "step": 14240
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3580852758339574,
+      "learning_rate": 2.0407769799229925e-07,
+      "loss": 0.6862,
+      "step": 14241
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3423697133794303,
+      "learning_rate": 2.0378470201680767e-07,
+      "loss": 0.5716,
+      "step": 14242
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.726153019615091,
+      "learning_rate": 2.0349191214536767e-07,
+      "loss": 0.6903,
+      "step": 14243
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2107057913465964,
+      "learning_rate": 2.0319932839056365e-07,
+      "loss": 0.703,
+      "step": 14244
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.398204728393341,
+      "learning_rate": 2.029069507649678e-07,
+      "loss": 0.6314,
+      "step": 14245
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7214687868492347,
+      "learning_rate": 2.0261477928114282e-07,
+      "loss": 0.6856,
+      "step": 14246
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.236920217825028,
+      "learning_rate": 2.0232281395164587e-07,
+      "loss": 0.5852,
+      "step": 14247
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2626073402120614,
+      "learning_rate": 2.0203105478902252e-07,
+      "loss": 0.657,
+      "step": 14248
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2387604117140327,
+      "learning_rate": 2.0173950180581047e-07,
+      "loss": 0.692,
+      "step": 14249
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1879624458411453,
+      "learning_rate": 2.0144815501453863e-07,
+      "loss": 0.6827,
+      "step": 14250
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0489528396031544,
+      "learning_rate": 2.0115701442772528e-07,
+      "loss": 0.6298,
+      "step": 14251
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0526184827331284,
+      "learning_rate": 2.0086608005788376e-07,
+      "loss": 0.6779,
+      "step": 14252
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.870617112106903,
+      "learning_rate": 2.0057535191751465e-07,
+      "loss": 0.6128,
+      "step": 14253
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4122446068114003,
+      "learning_rate": 2.002848300191118e-07,
+      "loss": 0.5494,
+      "step": 14254
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.228531662698104,
+      "learning_rate": 1.9999451437515915e-07,
+      "loss": 0.7827,
+      "step": 14255
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4457377754843195,
+      "learning_rate": 1.997044049981328e-07,
+      "loss": 0.6229,
+      "step": 14256
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3080842465640754,
+      "learning_rate": 1.994145019004984e-07,
+      "loss": 0.5966,
+      "step": 14257
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2795247086674597,
+      "learning_rate": 1.9912480509471476e-07,
+      "loss": 0.6125,
+      "step": 14258
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1160781776926996,
+      "learning_rate": 1.988353145932298e-07,
+      "loss": 0.558,
+      "step": 14259
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1609324669028587,
+      "learning_rate": 1.985460304084852e-07,
+      "loss": 0.6834,
+      "step": 14260
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3704650774313107,
+      "learning_rate": 1.9825695255290934e-07,
+      "loss": 0.6572,
+      "step": 14261
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3968444299360268,
+      "learning_rate": 1.9796808103892783e-07,
+      "loss": 0.6646,
+      "step": 14262
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5883691515138696,
+      "learning_rate": 1.9767941587895246e-07,
+      "loss": 0.576,
+      "step": 14263
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3543318544415548,
+      "learning_rate": 1.9739095708538714e-07,
+      "loss": 0.6081,
+      "step": 14264
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4921018058207491,
+      "learning_rate": 1.9710270467062865e-07,
+      "loss": 0.66,
+      "step": 14265
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.592031399393966,
+      "learning_rate": 1.9681465864706372e-07,
+      "loss": 0.4678,
+      "step": 14266
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6141810265028982,
+      "learning_rate": 1.9652681902707028e-07,
+      "loss": 0.633,
+      "step": 14267
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1329065751417176,
+      "learning_rate": 1.9623918582301726e-07,
+      "loss": 0.6752,
+      "step": 14268
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2416923372621005,
+      "learning_rate": 1.9595175904726481e-07,
+      "loss": 0.5816,
+      "step": 14269
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2218361622878076,
+      "learning_rate": 1.9566453871216417e-07,
+      "loss": 0.5976,
+      "step": 14270
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3720565724206475,
+      "learning_rate": 1.953775248300588e-07,
+      "loss": 0.63,
+      "step": 14271
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.52004347373098,
+      "learning_rate": 1.9509071741328155e-07,
+      "loss": 0.7378,
+      "step": 14272
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.355216206014599,
+      "learning_rate": 1.9480411647415708e-07,
+      "loss": 0.6319,
+      "step": 14273
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3955187202936992,
+      "learning_rate": 1.9451772202500163e-07,
+      "loss": 0.6264,
+      "step": 14274
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3054960673187996,
+      "learning_rate": 1.94231534078122e-07,
+      "loss": 0.6682,
+      "step": 14275
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9934957028777471,
+      "learning_rate": 1.9394555264581672e-07,
+      "loss": 0.6643,
+      "step": 14276
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3978556045925308,
+      "learning_rate": 1.936597777403737e-07,
+      "loss": 0.6478,
+      "step": 14277
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5259078448317525,
+      "learning_rate": 1.9337420937407536e-07,
+      "loss": 0.7787,
+      "step": 14278
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2717604266081497,
+      "learning_rate": 1.9308884755919132e-07,
+      "loss": 0.6479,
+      "step": 14279
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1580951244665842,
+      "learning_rate": 1.9280369230798568e-07,
+      "loss": 0.5576,
+      "step": 14280
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4107409205120758,
+      "learning_rate": 1.925187436327114e-07,
+      "loss": 0.6664,
+      "step": 14281
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7931608429419208,
+      "learning_rate": 1.9223400154561313e-07,
+      "loss": 0.6484,
+      "step": 14282
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3882321008202174,
+      "learning_rate": 1.9194946605892773e-07,
+      "loss": 0.5902,
+      "step": 14283
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3577581406878803,
+      "learning_rate": 1.9166513718488155e-07,
+      "loss": 0.5605,
+      "step": 14284
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1899583551602968,
+      "learning_rate": 1.9138101493569428e-07,
+      "loss": 0.6298,
+      "step": 14285
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3601238310779613,
+      "learning_rate": 1.9109709932357333e-07,
+      "loss": 0.5124,
+      "step": 14286
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3591666492839942,
+      "learning_rate": 1.9081339036071956e-07,
+      "loss": 0.5445,
+      "step": 14287
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3341348815238612,
+      "learning_rate": 1.9052988805932538e-07,
+      "loss": 0.5802,
+      "step": 14288
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.7575574719487634,
+      "learning_rate": 1.902465924315733e-07,
+      "loss": 0.5557,
+      "step": 14289
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.521549966104881,
+      "learning_rate": 1.899635034896369e-07,
+      "loss": 0.7431,
+      "step": 14290
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.486025240491682,
+      "learning_rate": 1.8968062124568142e-07,
+      "loss": 0.5888,
+      "step": 14291
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1643701704162046,
+      "learning_rate": 1.8939794571186332e-07,
+      "loss": 0.6684,
+      "step": 14292
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5987465704825001,
+      "learning_rate": 1.8911547690032838e-07,
+      "loss": 0.6857,
+      "step": 14293
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6738571673942566,
+      "learning_rate": 1.8883321482321583e-07,
+      "loss": 0.6648,
+      "step": 14294
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3565184344699572,
+      "learning_rate": 1.8855115949265535e-07,
+      "loss": 0.8796,
+      "step": 14295
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2703014857572623,
+      "learning_rate": 1.882693109207684e-07,
+      "loss": 0.6512,
+      "step": 14296
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0355313373629704,
+      "learning_rate": 1.879876691196636e-07,
+      "loss": 0.5106,
+      "step": 14297
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2441779583149777,
+      "learning_rate": 1.8770623410144628e-07,
+      "loss": 0.5804,
+      "step": 14298
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4875767915745333,
+      "learning_rate": 1.8742500587820955e-07,
+      "loss": 0.6026,
+      "step": 14299
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3420011652723256,
+      "learning_rate": 1.8714398446203873e-07,
+      "loss": 0.6084,
+      "step": 14300
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1619812022403346,
+      "learning_rate": 1.8686316986500974e-07,
+      "loss": 0.6911,
+      "step": 14301
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3802525653784312,
+      "learning_rate": 1.86582562099189e-07,
+      "loss": 0.636,
+      "step": 14302
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6473626514558053,
+      "learning_rate": 1.8630216117663747e-07,
+      "loss": 0.8048,
+      "step": 14303
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5572479022754482,
+      "learning_rate": 1.86021967109401e-07,
+      "loss": 0.8043,
+      "step": 14304
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.423328145714096,
+      "learning_rate": 1.8574197990952225e-07,
+      "loss": 0.7766,
+      "step": 14305
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6367108633869876,
+      "learning_rate": 1.8546219958903267e-07,
+      "loss": 0.6043,
+      "step": 14306
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3378131081395384,
+      "learning_rate": 1.8518262615995541e-07,
+      "loss": 0.7261,
+      "step": 14307
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.302316457841628,
+      "learning_rate": 1.8490325963430368e-07,
+      "loss": 0.6183,
+      "step": 14308
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6456959419734192,
+      "learning_rate": 1.8462410002408228e-07,
+      "loss": 0.6479,
+      "step": 14309
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2138090714648653,
+      "learning_rate": 1.843451473412877e-07,
+      "loss": 0.5648,
+      "step": 14310
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4104162935053646,
+      "learning_rate": 1.8406640159790767e-07,
+      "loss": 0.6339,
+      "step": 14311
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6588277985373718,
+      "learning_rate": 1.837878628059192e-07,
+      "loss": 0.6047,
+      "step": 14312
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2914784026944053,
+      "learning_rate": 1.835095309772933e-07,
+      "loss": 0.649,
+      "step": 14313
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4364726899793505,
+      "learning_rate": 1.8323140612399038e-07,
+      "loss": 0.6954,
+      "step": 14314
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4870530833419613,
+      "learning_rate": 1.829534882579598e-07,
+      "loss": 0.6615,
+      "step": 14315
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2213668503110673,
+      "learning_rate": 1.8267577739114695e-07,
+      "loss": 0.5806,
+      "step": 14316
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5417339141127766,
+      "learning_rate": 1.8239827353548457e-07,
+      "loss": 0.6911,
+      "step": 14317
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4038712319837285,
+      "learning_rate": 1.8212097670289807e-07,
+      "loss": 0.616,
+      "step": 14318
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.478027642880037,
+      "learning_rate": 1.8184388690530242e-07,
+      "loss": 0.6361,
+      "step": 14319
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3760793262266429,
+      "learning_rate": 1.8156700415460582e-07,
+      "loss": 0.6805,
+      "step": 14320
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9247229634349368,
+      "learning_rate": 1.812903284627071e-07,
+      "loss": 0.7494,
+      "step": 14321
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.564839299350958,
+      "learning_rate": 1.8101385984149343e-07,
+      "loss": 0.556,
+      "step": 14322
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3510023136456022,
+      "learning_rate": 1.8073759830284754e-07,
+      "loss": 0.5993,
+      "step": 14323
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7969512013185833,
+      "learning_rate": 1.804615438586399e-07,
+      "loss": 0.6293,
+      "step": 14324
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3029719467466658,
+      "learning_rate": 1.801856965207338e-07,
+      "loss": 0.8273,
+      "step": 14325
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0900275950757579,
+      "learning_rate": 1.79910056300982e-07,
+      "loss": 0.5465,
+      "step": 14326
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6991126193703616,
+      "learning_rate": 1.796346232112306e-07,
+      "loss": 0.7003,
+      "step": 14327
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.328223230717909,
+      "learning_rate": 1.7935939726331453e-07,
+      "loss": 0.694,
+      "step": 14328
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5419258419134225,
+      "learning_rate": 1.7908437846906158e-07,
+      "loss": 0.6938,
+      "step": 14329
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1078601149170533,
+      "learning_rate": 1.7880956684028894e-07,
+      "loss": 0.6002,
+      "step": 14330
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0710987092061623,
+      "learning_rate": 1.7853496238880718e-07,
+      "loss": 0.6016,
+      "step": 14331
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.475809845505849,
+      "learning_rate": 1.7826056512641633e-07,
+      "loss": 0.6694,
+      "step": 14332
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.291555191647312,
+      "learning_rate": 1.7798637506490745e-07,
+      "loss": 0.7058,
+      "step": 14333
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.444428456147284,
+      "learning_rate": 1.7771239221606285e-07,
+      "loss": 0.6954,
+      "step": 14334
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.51151286710312,
+      "learning_rate": 1.7743861659165695e-07,
+      "loss": 0.7011,
+      "step": 14335
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2851201894407778,
+      "learning_rate": 1.7716504820345427e-07,
+      "loss": 0.5774,
+      "step": 14336
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4991367646842668,
+      "learning_rate": 1.768916870632098e-07,
+      "loss": 0.7597,
+      "step": 14337
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1407283179233139,
+      "learning_rate": 1.7661853318267142e-07,
+      "loss": 0.678,
+      "step": 14338
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6021221057015982,
+      "learning_rate": 1.7634558657357748e-07,
+      "loss": 0.5706,
+      "step": 14339
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.8986092292573518,
+      "learning_rate": 1.7607284724765583e-07,
+      "loss": 0.7591,
+      "step": 14340
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3886259523955036,
+      "learning_rate": 1.7580031521662765e-07,
+      "loss": 0.6251,
+      "step": 14341
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3835812300214483,
+      "learning_rate": 1.7552799049220415e-07,
+      "loss": 0.6233,
+      "step": 14342
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.237477683974181,
+      "learning_rate": 1.752558730860876e-07,
+      "loss": 0.5878,
+      "step": 14343
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4391387805376923,
+      "learning_rate": 1.7498396300997146e-07,
+      "loss": 0.6684,
+      "step": 14344
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4766025932834785,
+      "learning_rate": 1.747122602755391e-07,
+      "loss": 0.6953,
+      "step": 14345
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2895176673799589,
+      "learning_rate": 1.7444076489446958e-07,
+      "loss": 0.6297,
+      "step": 14346
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4091747089068138,
+      "learning_rate": 1.7416947687842633e-07,
+      "loss": 0.6325,
+      "step": 14347
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3439100202460472,
+      "learning_rate": 1.738983962390678e-07,
+      "loss": 0.5868,
+      "step": 14348
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3777041283313747,
+      "learning_rate": 1.736275229880441e-07,
+      "loss": 0.669,
+      "step": 14349
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6594536848073693,
+      "learning_rate": 1.733568571369948e-07,
+      "loss": 0.6397,
+      "step": 14350
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5403087655376952,
+      "learning_rate": 1.7308639869755062e-07,
+      "loss": 0.6435,
+      "step": 14351
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8760260588540532,
+      "learning_rate": 1.728161476813345e-07,
+      "loss": 0.6928,
+      "step": 14352
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4664763301935695,
+      "learning_rate": 1.7254610409995875e-07,
+      "loss": 0.5952,
+      "step": 14353
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4766297409970863,
+      "learning_rate": 1.7227626796502807e-07,
+      "loss": 0.6372,
+      "step": 14354
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0541159264686797,
+      "learning_rate": 1.720066392881381e-07,
+      "loss": 0.5486,
+      "step": 14355
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.26337600100342,
+      "learning_rate": 1.7173721808087573e-07,
+      "loss": 0.5819,
+      "step": 14356
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4621139575606308,
+      "learning_rate": 1.7146800435481837e-07,
+      "loss": 0.5882,
+      "step": 14357
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4418818608424362,
+      "learning_rate": 1.7119899812153394e-07,
+      "loss": 0.5689,
+      "step": 14358
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6154477058068566,
+      "learning_rate": 1.7093019939258327e-07,
+      "loss": 0.698,
+      "step": 14359
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.649656991562924,
+      "learning_rate": 1.706616081795165e-07,
+      "loss": 0.7026,
+      "step": 14360
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2596128187782538,
+      "learning_rate": 1.7039322449387664e-07,
+      "loss": 0.4964,
+      "step": 14361
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3107263699640224,
+      "learning_rate": 1.7012504834719556e-07,
+      "loss": 0.6677,
+      "step": 14362
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4211299790473289,
+      "learning_rate": 1.6985707975099742e-07,
+      "loss": 0.6638,
+      "step": 14363
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6486904226257157,
+      "learning_rate": 1.6958931871679908e-07,
+      "loss": 0.5834,
+      "step": 14364
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5075090157318876,
+      "learning_rate": 1.6932176525610465e-07,
+      "loss": 0.6664,
+      "step": 14365
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.23524521941306,
+      "learning_rate": 1.6905441938041222e-07,
+      "loss": 0.7397,
+      "step": 14366
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5033549407115334,
+      "learning_rate": 1.687872811012109e-07,
+      "loss": 0.6323,
+      "step": 14367
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.523820698076294,
+      "learning_rate": 1.685203504299804e-07,
+      "loss": 0.6373,
+      "step": 14368
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4930944691531454,
+      "learning_rate": 1.6825362737818985e-07,
+      "loss": 0.6405,
+      "step": 14369
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.187740412068091,
+      "learning_rate": 1.6798711195730233e-07,
+      "loss": 0.5401,
+      "step": 14370
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2958663419953016,
+      "learning_rate": 1.677208041787698e-07,
+      "loss": 0.6698,
+      "step": 14371
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2547414249740825,
+      "learning_rate": 1.6745470405403697e-07,
+      "loss": 0.5402,
+      "step": 14372
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.382614818518251,
+      "learning_rate": 1.6718881159453693e-07,
+      "loss": 0.611,
+      "step": 14373
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6501582073257446,
+      "learning_rate": 1.6692312681169775e-07,
+      "loss": 0.6834,
+      "step": 14374
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.301213516224009,
+      "learning_rate": 1.6665764971693587e-07,
+      "loss": 0.567,
+      "step": 14375
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.155625281978456,
+      "learning_rate": 1.6639238032165883e-07,
+      "loss": 0.6321,
+      "step": 14376
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.341059305405279,
+      "learning_rate": 1.6612731863726638e-07,
+      "loss": 0.8239,
+      "step": 14377
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4951159914946002,
+      "learning_rate": 1.6586246467514833e-07,
+      "loss": 0.741,
+      "step": 14378
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.219378101746413,
+      "learning_rate": 1.6559781844668666e-07,
+      "loss": 0.4872,
+      "step": 14379
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6743321920751963,
+      "learning_rate": 1.6533337996325337e-07,
+      "loss": 0.6351,
+      "step": 14380
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.236303078467631,
+      "learning_rate": 1.650691492362122e-07,
+      "loss": 0.6243,
+      "step": 14381
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2964831316170182,
+      "learning_rate": 1.6480512627691792e-07,
+      "loss": 0.6086,
+      "step": 14382
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5005754035680974,
+      "learning_rate": 1.6454131109671534e-07,
+      "loss": 0.5692,
+      "step": 14383
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4776444585345403,
+      "learning_rate": 1.6427770370694208e-07,
+      "loss": 0.6338,
+      "step": 14384
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1489137798558835,
+      "learning_rate": 1.6401430411892572e-07,
+      "loss": 0.6983,
+      "step": 14385
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4216688167951603,
+      "learning_rate": 1.63751112343985e-07,
+      "loss": 0.6976,
+      "step": 14386
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4313162499967493,
+      "learning_rate": 1.6348812839342976e-07,
+      "loss": 0.6522,
+      "step": 14387
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1169033120340368,
+      "learning_rate": 1.6322535227856152e-07,
+      "loss": 0.5582,
+      "step": 14388
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5052484711788159,
+      "learning_rate": 1.6296278401067122e-07,
+      "loss": 0.5219,
+      "step": 14389
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4581085021268991,
+      "learning_rate": 1.6270042360104377e-07,
+      "loss": 0.7277,
+      "step": 14390
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4456301204035484,
+      "learning_rate": 1.624382710609512e-07,
+      "loss": 0.6412,
+      "step": 14391
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3468111245622691,
+      "learning_rate": 1.621763264016607e-07,
+      "loss": 0.6819,
+      "step": 14392
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3956027730789773,
+      "learning_rate": 1.6191458963442762e-07,
+      "loss": 0.6503,
+      "step": 14393
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4260521892235074,
+      "learning_rate": 1.6165306077049969e-07,
+      "loss": 0.6361,
+      "step": 14394
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.305783592917671,
+      "learning_rate": 1.6139173982111567e-07,
+      "loss": 0.6068,
+      "step": 14395
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3934050557549706,
+      "learning_rate": 1.611306267975038e-07,
+      "loss": 0.6711,
+      "step": 14396
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3405629511992097,
+      "learning_rate": 1.6086972171088678e-07,
+      "loss": 0.5915,
+      "step": 14397
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4104840229074458,
+      "learning_rate": 1.606090245724745e-07,
+      "loss": 0.6421,
+      "step": 14398
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4428078271030724,
+      "learning_rate": 1.603485353934703e-07,
+      "loss": 0.8108,
+      "step": 14399
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2373205573285668,
+      "learning_rate": 1.6008825418506846e-07,
+      "loss": 0.5797,
+      "step": 14400
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1506183314995526,
+      "learning_rate": 1.5982818095845286e-07,
+      "loss": 0.7144,
+      "step": 14401
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3010095528499006,
+      "learning_rate": 1.5956831572480013e-07,
+      "loss": 0.6525,
+      "step": 14402
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4707680659709184,
+      "learning_rate": 1.5930865849527744e-07,
+      "loss": 0.6205,
+      "step": 14403
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3624185957190205,
+      "learning_rate": 1.5904920928104196e-07,
+      "loss": 0.6272,
+      "step": 14404
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3000620008957675,
+      "learning_rate": 1.5878996809324366e-07,
+      "loss": 0.651,
+      "step": 14405
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4808142443320111,
+      "learning_rate": 1.5853093494302195e-07,
+      "loss": 0.6375,
+      "step": 14406
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0691617674554337,
+      "learning_rate": 1.582721098415091e-07,
+      "loss": 0.7105,
+      "step": 14407
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.024975457187923,
+      "learning_rate": 1.5801349279982726e-07,
+      "loss": 0.6107,
+      "step": 14408
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3803979302531157,
+      "learning_rate": 1.577550838290881e-07,
+      "loss": 0.6161,
+      "step": 14409
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.593434449281508,
+      "learning_rate": 1.5749688294039778e-07,
+      "loss": 0.6541,
+      "step": 14410
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4138611914799952,
+      "learning_rate": 1.5723889014485128e-07,
+      "loss": 0.6445,
+      "step": 14411
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.349712776205667,
+      "learning_rate": 1.5698110545353474e-07,
+      "loss": 0.6357,
+      "step": 14412
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3020600286659505,
+      "learning_rate": 1.567235288775265e-07,
+      "loss": 0.5636,
+      "step": 14413
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2347039488552012,
+      "learning_rate": 1.564661604278944e-07,
+      "loss": 0.5973,
+      "step": 14414
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6118395367756424,
+      "learning_rate": 1.5620900011569906e-07,
+      "loss": 0.7178,
+      "step": 14415
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.235891824446211,
+      "learning_rate": 1.5595204795199047e-07,
+      "loss": 0.5497,
+      "step": 14416
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3990037697271724,
+      "learning_rate": 1.5569530394780986e-07,
+      "loss": 0.6495,
+      "step": 14417
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.044986923974736,
+      "learning_rate": 1.5543876811419222e-07,
+      "loss": 0.6674,
+      "step": 14418
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4452725999292961,
+      "learning_rate": 1.5518244046215936e-07,
+      "loss": 0.6036,
+      "step": 14419
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2968843720366048,
+      "learning_rate": 1.5492632100272686e-07,
+      "loss": 0.5663,
+      "step": 14420
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.27885122346733,
+      "learning_rate": 1.546704097469015e-07,
+      "loss": 0.6541,
+      "step": 14421
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3468992381797182,
+      "learning_rate": 1.544147067056795e-07,
+      "loss": 0.5587,
+      "step": 14422
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4109336594301374,
+      "learning_rate": 1.5415921189004922e-07,
+      "loss": 0.5711,
+      "step": 14423
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5708930807085784,
+      "learning_rate": 1.539039253109892e-07,
+      "loss": 0.6254,
+      "step": 14424
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9979985298313061,
+      "learning_rate": 1.5364884697947113e-07,
+      "loss": 0.6601,
+      "step": 14425
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4805839938455114,
+      "learning_rate": 1.533939769064563e-07,
+      "loss": 0.6833,
+      "step": 14426
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5454344290577595,
+      "learning_rate": 1.5313931510289482e-07,
+      "loss": 0.6844,
+      "step": 14427
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3436518711406649,
+      "learning_rate": 1.528848615797318e-07,
+      "loss": 0.6542,
+      "step": 14428
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7503273143020202,
+      "learning_rate": 1.526306163479019e-07,
+      "loss": 0.7215,
+      "step": 14429
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4112420804590915,
+      "learning_rate": 1.523765794183296e-07,
+      "loss": 0.5847,
+      "step": 14430
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4408885947778822,
+      "learning_rate": 1.521227508019324e-07,
+      "loss": 0.6303,
+      "step": 14431
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2691529946108167,
+      "learning_rate": 1.518691305096165e-07,
+      "loss": 0.5385,
+      "step": 14432
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.121853766889347,
+      "learning_rate": 1.5161571855228264e-07,
+      "loss": 0.5117,
+      "step": 14433
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3928814863882226,
+      "learning_rate": 1.5136251494081822e-07,
+      "loss": 0.5783,
+      "step": 14434
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.255858264041917,
+      "learning_rate": 1.5110951968610564e-07,
+      "loss": 0.5316,
+      "step": 14435
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.313289879812934,
+      "learning_rate": 1.5085673279901626e-07,
+      "loss": 0.5272,
+      "step": 14436
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4591854690936936,
+      "learning_rate": 1.5060415429041186e-07,
+      "loss": 0.6868,
+      "step": 14437
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2743056701856132,
+      "learning_rate": 1.503517841711477e-07,
+      "loss": 0.6179,
+      "step": 14438
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4496942975772762,
+      "learning_rate": 1.5009962245206845e-07,
+      "loss": 0.7496,
+      "step": 14439
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.411201266372185,
+      "learning_rate": 1.4984766914400928e-07,
+      "loss": 0.5464,
+      "step": 14440
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3665228943034804,
+      "learning_rate": 1.4959592425779768e-07,
+      "loss": 0.6455,
+      "step": 14441
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5211446102168429,
+      "learning_rate": 1.4934438780425108e-07,
+      "loss": 0.5936,
+      "step": 14442
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3714068625740021,
+      "learning_rate": 1.4909305979417976e-07,
+      "loss": 0.7048,
+      "step": 14443
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3757935053261172,
+      "learning_rate": 1.488419402383834e-07,
+      "loss": 0.6026,
+      "step": 14444
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2482007625910712,
+      "learning_rate": 1.4859102914765278e-07,
+      "loss": 0.604,
+      "step": 14445
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7108995708752384,
+      "learning_rate": 1.4834032653276986e-07,
+      "loss": 0.666,
+      "step": 14446
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3209053584864638,
+      "learning_rate": 1.4808983240450881e-07,
+      "loss": 0.6113,
+      "step": 14447
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5594720300503295,
+      "learning_rate": 1.4783954677363376e-07,
+      "loss": 0.6235,
+      "step": 14448
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3005894110829166,
+      "learning_rate": 1.4758946965089894e-07,
+      "loss": 0.5433,
+      "step": 14449
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2793760662393348,
+      "learning_rate": 1.473396010470518e-07,
+      "loss": 0.571,
+      "step": 14450
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5413406212969119,
+      "learning_rate": 1.470899409728299e-07,
+      "loss": 0.5902,
+      "step": 14451
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0022633675198567,
+      "learning_rate": 1.4684048943896022e-07,
+      "loss": 0.6894,
+      "step": 14452
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2412061273199229,
+      "learning_rate": 1.4659124645616418e-07,
+      "loss": 0.6488,
+      "step": 14453
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.320123680499981,
+      "learning_rate": 1.4634221203515097e-07,
+      "loss": 0.6237,
+      "step": 14454
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4337198144436234,
+      "learning_rate": 1.4609338618662318e-07,
+      "loss": 0.577,
+      "step": 14455
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.42033507191154,
+      "learning_rate": 1.458447689212722e-07,
+      "loss": 0.6378,
+      "step": 14456
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4681208046741776,
+      "learning_rate": 1.455963602497823e-07,
+      "loss": 0.7231,
+      "step": 14457
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4123241247425236,
+      "learning_rate": 1.4534816018282882e-07,
+      "loss": 0.6761,
+      "step": 14458
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3299250554667443,
+      "learning_rate": 1.4510016873107657e-07,
+      "loss": 0.8916,
+      "step": 14459
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.398240498981903,
+      "learning_rate": 1.4485238590518147e-07,
+      "loss": 0.5862,
+      "step": 14460
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3231762445468183,
+      "learning_rate": 1.4460481171579334e-07,
+      "loss": 0.6207,
+      "step": 14461
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2624111257381485,
+      "learning_rate": 1.4435744617354975e-07,
+      "loss": 0.5855,
+      "step": 14462
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4176577305059983,
+      "learning_rate": 1.4411028928908054e-07,
+      "loss": 0.837,
+      "step": 14463
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.454282253748915,
+      "learning_rate": 1.4386334107300727e-07,
+      "loss": 0.6411,
+      "step": 14464
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3853501756948146,
+      "learning_rate": 1.4361660153594137e-07,
+      "loss": 0.7601,
+      "step": 14465
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3223676737026313,
+      "learning_rate": 1.4337007068848607e-07,
+      "loss": 0.5882,
+      "step": 14466
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4443553361077819,
+      "learning_rate": 1.4312374854123456e-07,
+      "loss": 0.6302,
+      "step": 14467
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5131222613244617,
+      "learning_rate": 1.4287763510477225e-07,
+      "loss": 0.6414,
+      "step": 14468
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5402143702871502,
+      "learning_rate": 1.4263173038967627e-07,
+      "loss": 0.6477,
+      "step": 14469
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.684472535387732,
+      "learning_rate": 1.42386034406512e-07,
+      "loss": 0.7309,
+      "step": 14470
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2169758671350939,
+      "learning_rate": 1.421405471658388e-07,
+      "loss": 0.5191,
+      "step": 14471
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.612077079521614,
+      "learning_rate": 1.4189526867820491e-07,
+      "loss": 0.6514,
+      "step": 14472
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4283306261027837,
+      "learning_rate": 1.4165019895415133e-07,
+      "loss": 0.6248,
+      "step": 14473
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3577488402954057,
+      "learning_rate": 1.4140533800420853e-07,
+      "loss": 0.603,
+      "step": 14474
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4785043388688668,
+      "learning_rate": 1.4116068583889865e-07,
+      "loss": 0.6648,
+      "step": 14475
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0398179883220229,
+      "learning_rate": 1.409162424687366e-07,
+      "loss": 0.5976,
+      "step": 14476
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.241270075649878,
+      "learning_rate": 1.4067200790422454e-07,
+      "loss": 0.5271,
+      "step": 14477
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5176985224413253,
+      "learning_rate": 1.4042798215585796e-07,
+      "loss": 0.7032,
+      "step": 14478
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3128138197920136,
+      "learning_rate": 1.401841652341246e-07,
+      "loss": 0.5087,
+      "step": 14479
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.081425821099929,
+      "learning_rate": 1.3994055714950105e-07,
+      "loss": 0.7516,
+      "step": 14480
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7021570346230674,
+      "learning_rate": 1.3969715791245565e-07,
+      "loss": 0.692,
+      "step": 14481
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6260918059797644,
+      "learning_rate": 1.3945396753344776e-07,
+      "loss": 0.6596,
+      "step": 14482
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2843739495414848,
+      "learning_rate": 1.3921098602292793e-07,
+      "loss": 0.6861,
+      "step": 14483
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.545403943035126,
+      "learning_rate": 1.389682133913378e-07,
+      "loss": 0.7157,
+      "step": 14484
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3123865497276626,
+      "learning_rate": 1.3872564964910905e-07,
+      "loss": 0.6176,
+      "step": 14485
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6646655440819083,
+      "learning_rate": 1.384832948066661e-07,
+      "loss": 0.5892,
+      "step": 14486
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3695565531996745,
+      "learning_rate": 1.382411488744234e-07,
+      "loss": 0.5991,
+      "step": 14487
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2899939751937253,
+      "learning_rate": 1.3799921186278597e-07,
+      "loss": 0.6346,
+      "step": 14488
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3130190689667427,
+      "learning_rate": 1.3775748378215047e-07,
+      "loss": 0.6065,
+      "step": 14489
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0670802833509032,
+      "learning_rate": 1.3751596464290529e-07,
+      "loss": 0.6811,
+      "step": 14490
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3369713895819142,
+      "learning_rate": 1.372746544554282e-07,
+      "loss": 0.6418,
+      "step": 14491
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.398978185498732,
+      "learning_rate": 1.3703355323008872e-07,
+      "loss": 0.6699,
+      "step": 14492
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5865993369274547,
+      "learning_rate": 1.3679266097724797e-07,
+      "loss": 0.6586,
+      "step": 14493
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3845345944855576,
+      "learning_rate": 1.3655197770725826e-07,
+      "loss": 0.6051,
+      "step": 14494
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5764502075189288,
+      "learning_rate": 1.363115034304602e-07,
+      "loss": 0.6917,
+      "step": 14495
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.463053647139865,
+      "learning_rate": 1.3607123815718993e-07,
+      "loss": 0.5984,
+      "step": 14496
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5597694324059632,
+      "learning_rate": 1.358311818977709e-07,
+      "loss": 0.6392,
+      "step": 14497
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5843553474799885,
+      "learning_rate": 1.3559133466251874e-07,
+      "loss": 0.7241,
+      "step": 14498
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6215990652874124,
+      "learning_rate": 1.3535169646174073e-07,
+      "loss": 0.7066,
+      "step": 14499
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3764123185148138,
+      "learning_rate": 1.3511226730573478e-07,
+      "loss": 0.6342,
+      "step": 14500
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4601337186821428,
+      "learning_rate": 1.3487304720478933e-07,
+      "loss": 0.6958,
+      "step": 14501
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.397250507515905,
+      "learning_rate": 1.346340361691839e-07,
+      "loss": 0.6344,
+      "step": 14502
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.77772072754753,
+      "learning_rate": 1.343952342091892e-07,
+      "loss": 0.7802,
+      "step": 14503
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9352901253254768,
+      "learning_rate": 1.3415664133506812e-07,
+      "loss": 0.6387,
+      "step": 14504
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2224989677810014,
+      "learning_rate": 1.3391825755707356e-07,
+      "loss": 0.5903,
+      "step": 14505
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4621795970758882,
+      "learning_rate": 1.3368008288544843e-07,
+      "loss": 0.5696,
+      "step": 14506
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2417805896629892,
+      "learning_rate": 1.3344211733042788e-07,
+      "loss": 0.5984,
+      "step": 14507
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4715583830822427,
+      "learning_rate": 1.3320436090223765e-07,
+      "loss": 0.63,
+      "step": 14508
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2646266293609678,
+      "learning_rate": 1.3296681361109564e-07,
+      "loss": 0.7782,
+      "step": 14509
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1149887824615212,
+      "learning_rate": 1.327294754672087e-07,
+      "loss": 0.5204,
+      "step": 14510
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3984784821955982,
+      "learning_rate": 1.324923464807759e-07,
+      "loss": 0.6263,
+      "step": 14511
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6938149692017035,
+      "learning_rate": 1.3225542666198854e-07,
+      "loss": 0.6623,
+      "step": 14512
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0348241946026326,
+      "learning_rate": 1.3201871602102512e-07,
+      "loss": 0.7223,
+      "step": 14513
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2575318574982606,
+      "learning_rate": 1.3178221456806028e-07,
+      "loss": 0.6969,
+      "step": 14514
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4614704931627114,
+      "learning_rate": 1.3154592231325535e-07,
+      "loss": 0.6948,
+      "step": 14515
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.458317101462919,
+      "learning_rate": 1.3130983926676445e-07,
+      "loss": 0.5854,
+      "step": 14516
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3210772749154607,
+      "learning_rate": 1.310739654387333e-07,
+      "loss": 0.7032,
+      "step": 14517
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5211978943316578,
+      "learning_rate": 1.308383008392977e-07,
+      "loss": 0.5827,
+      "step": 14518
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4207251636703169,
+      "learning_rate": 1.3060284547858403e-07,
+      "loss": 0.5927,
+      "step": 14519
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3499110903611689,
+      "learning_rate": 1.3036759936671084e-07,
+      "loss": 0.7235,
+      "step": 14520
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.675799450464721,
+      "learning_rate": 1.3013256251378725e-07,
+      "loss": 0.6159,
+      "step": 14521
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3771868673847478,
+      "learning_rate": 1.2989773492991298e-07,
+      "loss": 0.5658,
+      "step": 14522
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5391967064884393,
+      "learning_rate": 1.2966311662517994e-07,
+      "loss": 0.6544,
+      "step": 14523
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3452610893959476,
+      "learning_rate": 1.2942870760966952e-07,
+      "loss": 0.6001,
+      "step": 14524
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3863288870474173,
+      "learning_rate": 1.2919450789345477e-07,
+      "loss": 0.6848,
+      "step": 14525
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3791979403237797,
+      "learning_rate": 1.289605174865993e-07,
+      "loss": 0.8748,
+      "step": 14526
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1018899654973553,
+      "learning_rate": 1.2872673639915955e-07,
+      "loss": 0.6981,
+      "step": 14527
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3840953358512522,
+      "learning_rate": 1.2849316464118078e-07,
+      "loss": 0.6268,
+      "step": 14528
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4618117256042447,
+      "learning_rate": 1.282598022226994e-07,
+      "loss": 0.6318,
+      "step": 14529
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7115909365570827,
+      "learning_rate": 1.2802664915374575e-07,
+      "loss": 0.7329,
+      "step": 14530
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.473941504489849,
+      "learning_rate": 1.2779370544433567e-07,
+      "loss": 0.6383,
+      "step": 14531
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2198910347566083,
+      "learning_rate": 1.275609711044823e-07,
+      "loss": 0.6254,
+      "step": 14532
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1980337843437538,
+      "learning_rate": 1.2732844614418482e-07,
+      "loss": 0.5659,
+      "step": 14533
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1883174720226046,
+      "learning_rate": 1.270961305734364e-07,
+      "loss": 0.7061,
+      "step": 14534
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.293198551270385,
+      "learning_rate": 1.2686402440222013e-07,
+      "loss": 0.5536,
+      "step": 14535
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5122814013700308,
+      "learning_rate": 1.2663212764050858e-07,
+      "loss": 0.6299,
+      "step": 14536
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2496053404644794,
+      "learning_rate": 1.2640044029826937e-07,
+      "loss": 0.5398,
+      "step": 14537
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4415106787216476,
+      "learning_rate": 1.261689623854573e-07,
+      "loss": 0.6107,
+      "step": 14538
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6367660973829108,
+      "learning_rate": 1.2593769391201827e-07,
+      "loss": 0.6496,
+      "step": 14539
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.155209417967403,
+      "learning_rate": 1.257066348878927e-07,
+      "loss": 0.6349,
+      "step": 14540
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.002954461364405,
+      "learning_rate": 1.254757853230082e-07,
+      "loss": 0.7829,
+      "step": 14541
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1119499363632517,
+      "learning_rate": 1.252451452272857e-07,
+      "loss": 0.5975,
+      "step": 14542
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.388250146035823,
+      "learning_rate": 1.2501471461063564e-07,
+      "loss": 0.6313,
+      "step": 14543
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.410712711724459,
+      "learning_rate": 1.247844934829606e-07,
+      "loss": 0.6343,
+      "step": 14544
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.855611519300243,
+      "learning_rate": 1.2455448185415386e-07,
+      "loss": 0.5597,
+      "step": 14545
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5521809595778402,
+      "learning_rate": 1.2432467973409857e-07,
+      "loss": 0.6157,
+      "step": 14546
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.16876524587885,
+      "learning_rate": 1.2409508713267072e-07,
+      "loss": 0.5604,
+      "step": 14547
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4744149609047958,
+      "learning_rate": 1.238657040597374e-07,
+      "loss": 0.5948,
+      "step": 14548
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2810138180408333,
+      "learning_rate": 1.2363653052515302e-07,
+      "loss": 0.6595,
+      "step": 14549
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.406450164519609,
+      "learning_rate": 1.234075665387674e-07,
+      "loss": 0.8265,
+      "step": 14550
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8645196478784571,
+      "learning_rate": 1.2317881211041993e-07,
+      "loss": 0.706,
+      "step": 14551
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4205234448869284,
+      "learning_rate": 1.2295026724993998e-07,
+      "loss": 0.6103,
+      "step": 14552
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4392238529200228,
+      "learning_rate": 1.2272193196714854e-07,
+      "loss": 0.6209,
+      "step": 14553
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.268558397911429,
+      "learning_rate": 1.2249380627185781e-07,
+      "loss": 0.4926,
+      "step": 14554
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.313686525607636,
+      "learning_rate": 1.2226589017387158e-07,
+      "loss": 0.6157,
+      "step": 14555
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2744335763225358,
+      "learning_rate": 1.2203818368298315e-07,
+      "loss": 0.6471,
+      "step": 14556
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4407797672841818,
+      "learning_rate": 1.2181068680897745e-07,
+      "loss": 0.6376,
+      "step": 14557
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2276174237309698,
+      "learning_rate": 1.2158339956163113e-07,
+      "loss": 0.5795,
+      "step": 14558
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3567121511795461,
+      "learning_rate": 1.2135632195071133e-07,
+      "loss": 0.6771,
+      "step": 14559
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.489076536572611,
+      "learning_rate": 1.211294539859753e-07,
+      "loss": 0.5703,
+      "step": 14560
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7521063120602836,
+      "learning_rate": 1.2090279567717246e-07,
+      "loss": 0.6509,
+      "step": 14561
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5576233840726008,
+      "learning_rate": 1.2067634703404218e-07,
+      "loss": 0.7198,
+      "step": 14562
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3480466358620267,
+      "learning_rate": 1.204501080663173e-07,
+      "loss": 0.7471,
+      "step": 14563
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6373309609912963,
+      "learning_rate": 1.202240787837178e-07,
+      "loss": 0.5927,
+      "step": 14564
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4648812344679658,
+      "learning_rate": 1.1999825919595808e-07,
+      "loss": 0.6467,
+      "step": 14565
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6661026378267825,
+      "learning_rate": 1.1977264931274158e-07,
+      "loss": 0.7422,
+      "step": 14566
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.417398144091873,
+      "learning_rate": 1.1954724914376215e-07,
+      "loss": 0.7057,
+      "step": 14567
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4593760737744326,
+      "learning_rate": 1.1932205869870762e-07,
+      "loss": 0.6588,
+      "step": 14568
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3841827102938629,
+      "learning_rate": 1.1909707798725412e-07,
+      "loss": 0.6031,
+      "step": 14569
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5725223439251121,
+      "learning_rate": 1.1887230701906893e-07,
+      "loss": 0.7204,
+      "step": 14570
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8051948088937315,
+      "learning_rate": 1.186477458038121e-07,
+      "loss": 0.628,
+      "step": 14571
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.561402720636023,
+      "learning_rate": 1.1842339435113259e-07,
+      "loss": 0.6295,
+      "step": 14572
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2401163638527488,
+      "learning_rate": 1.1819925267067211e-07,
+      "loss": 0.5484,
+      "step": 14573
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.672949556492963,
+      "learning_rate": 1.1797532077206187e-07,
+      "loss": 0.7452,
+      "step": 14574
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.350074754641563,
+      "learning_rate": 1.177515986649247e-07,
+      "loss": 0.67,
+      "step": 14575
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5026903229561501,
+      "learning_rate": 1.1752808635887514e-07,
+      "loss": 0.6393,
+      "step": 14576
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2061033804094243,
+      "learning_rate": 1.1730478386351719e-07,
+      "loss": 0.5706,
+      "step": 14577
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2481566729631326,
+      "learning_rate": 1.1708169118844759e-07,
+      "loss": 0.6431,
+      "step": 14578
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2725261730071478,
+      "learning_rate": 1.1685880834325203e-07,
+      "loss": 0.6308,
+      "step": 14579
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8216435715765107,
+      "learning_rate": 1.1663613533750839e-07,
+      "loss": 0.7031,
+      "step": 14580
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3012783941592472,
+      "learning_rate": 1.1641367218078736e-07,
+      "loss": 0.5861,
+      "step": 14581
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3147382823331877,
+      "learning_rate": 1.1619141888264574e-07,
+      "loss": 0.6843,
+      "step": 14582
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.410289755562629,
+      "learning_rate": 1.1596937545263587e-07,
+      "loss": 0.5443,
+      "step": 14583
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2706783094709204,
+      "learning_rate": 1.1574754190030014e-07,
+      "loss": 0.6022,
+      "step": 14584
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.185205754521489,
+      "learning_rate": 1.1552591823516978e-07,
+      "loss": 0.5018,
+      "step": 14585
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2087101484357465,
+      "learning_rate": 1.1530450446676889e-07,
+      "loss": 0.5527,
+      "step": 14586
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3424446654051108,
+      "learning_rate": 1.1508330060461204e-07,
+      "loss": 0.6477,
+      "step": 14587
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3454693879385053,
+      "learning_rate": 1.1486230665820552e-07,
+      "loss": 0.6067,
+      "step": 14588
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2887031880189814,
+      "learning_rate": 1.1464152263704565e-07,
+      "loss": 0.5871,
+      "step": 14589
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.386656407214221,
+      "learning_rate": 1.1442094855061925e-07,
+      "loss": 0.593,
+      "step": 14590
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2038764376451943,
+      "learning_rate": 1.1420058440840598e-07,
+      "loss": 0.535,
+      "step": 14591
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4796972890313282,
+      "learning_rate": 1.1398043021987438e-07,
+      "loss": 0.5981,
+      "step": 14592
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.278696910815545,
+      "learning_rate": 1.1376048599448519e-07,
+      "loss": 0.6789,
+      "step": 14593
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.370044479486013,
+      "learning_rate": 1.1354075174169088e-07,
+      "loss": 0.6301,
+      "step": 14594
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2712193641749203,
+      "learning_rate": 1.1332122747093277e-07,
+      "loss": 0.5898,
+      "step": 14595
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.813635364247161,
+      "learning_rate": 1.1310191319164498e-07,
+      "loss": 0.6829,
+      "step": 14596
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.177400107580191,
+      "learning_rate": 1.1288280891325165e-07,
+      "loss": 0.5347,
+      "step": 14597
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.413824349172784,
+      "learning_rate": 1.1266391464516802e-07,
+      "loss": 0.6841,
+      "step": 14598
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.222193029225326,
+      "learning_rate": 1.12445230396801e-07,
+      "loss": 0.5713,
+      "step": 14599
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.121584668805231,
+      "learning_rate": 1.1222675617754752e-07,
+      "loss": 0.5619,
+      "step": 14600
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4463564524416168,
+      "learning_rate": 1.1200849199679564e-07,
+      "loss": 0.5921,
+      "step": 14601
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3843271153127625,
+      "learning_rate": 1.1179043786392507e-07,
+      "loss": 0.577,
+      "step": 14602
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.612010539844321,
+      "learning_rate": 1.1157259378830665e-07,
+      "loss": 0.705,
+      "step": 14603
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1932929074522272,
+      "learning_rate": 1.1135495977930011e-07,
+      "loss": 0.6783,
+      "step": 14604
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.660161053895346,
+      "learning_rate": 1.1113753584625853e-07,
+      "loss": 0.5779,
+      "step": 14605
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2346450832222506,
+      "learning_rate": 1.1092032199852609e-07,
+      "loss": 0.6668,
+      "step": 14606
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4237104083081324,
+      "learning_rate": 1.1070331824543534e-07,
+      "loss": 0.6592,
+      "step": 14607
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.478280494482155,
+      "learning_rate": 1.1048652459631216e-07,
+      "loss": 0.7155,
+      "step": 14608
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2485659061180212,
+      "learning_rate": 1.1026994106047296e-07,
+      "loss": 0.7682,
+      "step": 14609
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5223297413042636,
+      "learning_rate": 1.1005356764722364e-07,
+      "loss": 0.6392,
+      "step": 14610
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.526188569024392,
+      "learning_rate": 1.0983740436586343e-07,
+      "loss": 0.6567,
+      "step": 14611
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2367881443817126,
+      "learning_rate": 1.0962145122568046e-07,
+      "loss": 0.6326,
+      "step": 14612
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3790285424280346,
+      "learning_rate": 1.0940570823595565e-07,
+      "loss": 0.6407,
+      "step": 14613
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.295471624292166,
+      "learning_rate": 1.0919017540595933e-07,
+      "loss": 0.7218,
+      "step": 14614
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3374320393823227,
+      "learning_rate": 1.0897485274495357e-07,
+      "loss": 0.6026,
+      "step": 14615
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.025799922408355,
+      "learning_rate": 1.0875974026219149e-07,
+      "loss": 0.5375,
+      "step": 14616
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0997518017574448,
+      "learning_rate": 1.0854483796691629e-07,
+      "loss": 0.6446,
+      "step": 14617
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.170170117124883,
+      "learning_rate": 1.0833014586836388e-07,
+      "loss": 0.5989,
+      "step": 14618
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4299130925568468,
+      "learning_rate": 1.0811566397575912e-07,
+      "loss": 0.6157,
+      "step": 14619
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3839244275864235,
+      "learning_rate": 1.0790139229831909e-07,
+      "loss": 0.6169,
+      "step": 14620
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6154495782287757,
+      "learning_rate": 1.0768733084525196e-07,
+      "loss": 0.7096,
+      "step": 14621
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2416713894924671,
+      "learning_rate": 1.0747347962575539e-07,
+      "loss": 0.6602,
+      "step": 14622
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4613980923956669,
+      "learning_rate": 1.0725983864901978e-07,
+      "loss": 0.6108,
+      "step": 14623
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3721199868432428,
+      "learning_rate": 1.0704640792422616e-07,
+      "loss": 0.656,
+      "step": 14624
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8453827799244547,
+      "learning_rate": 1.0683318746054439e-07,
+      "loss": 0.6986,
+      "step": 14625
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8888592871320788,
+      "learning_rate": 1.066201772671388e-07,
+      "loss": 0.6447,
+      "step": 14626
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.168963057160967,
+      "learning_rate": 1.0640737735316264e-07,
+      "loss": 0.6124,
+      "step": 14627
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.628574335754763,
+      "learning_rate": 1.0619478772776027e-07,
+      "loss": 0.6264,
+      "step": 14628
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3514578335752319,
+      "learning_rate": 1.0598240840006658e-07,
+      "loss": 0.6232,
+      "step": 14629
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9510884237049264,
+      "learning_rate": 1.0577023937920816e-07,
+      "loss": 0.6749,
+      "step": 14630
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3731004421912507,
+      "learning_rate": 1.055582806743033e-07,
+      "loss": 0.6448,
+      "step": 14631
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2588328268598143,
+      "learning_rate": 1.0534653229445913e-07,
+      "loss": 0.6938,
+      "step": 14632
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.246789250704401,
+      "learning_rate": 1.0513499424877505e-07,
+      "loss": 0.6576,
+      "step": 14633
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4395274951548822,
+      "learning_rate": 1.0492366654634211e-07,
+      "loss": 0.5898,
+      "step": 14634
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5551839609875027,
+      "learning_rate": 1.0471254919624141e-07,
+      "loss": 0.6725,
+      "step": 14635
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2484312560747697,
+      "learning_rate": 1.0450164220754456e-07,
+      "loss": 0.5238,
+      "step": 14636
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6292871693545143,
+      "learning_rate": 1.0429094558931485e-07,
+      "loss": 0.7031,
+      "step": 14637
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3513479373277768,
+      "learning_rate": 1.0408045935060674e-07,
+      "loss": 0.6435,
+      "step": 14638
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3760518448991372,
+      "learning_rate": 1.0387018350046519e-07,
+      "loss": 0.6696,
+      "step": 14639
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1933061360276738,
+      "learning_rate": 1.0366011804792631e-07,
+      "loss": 0.6107,
+      "step": 14640
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2694674643513268,
+      "learning_rate": 1.0345026300201622e-07,
+      "loss": 0.5389,
+      "step": 14641
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1908675139863885,
+      "learning_rate": 1.0324061837175436e-07,
+      "loss": 0.6837,
+      "step": 14642
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3587871666810558,
+      "learning_rate": 1.0303118416614798e-07,
+      "loss": 0.5789,
+      "step": 14643
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4463217139577955,
+      "learning_rate": 1.0282196039419823e-07,
+      "loss": 0.6423,
+      "step": 14644
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.288377004825599,
+      "learning_rate": 1.0261294706489511e-07,
+      "loss": 0.6239,
+      "step": 14645
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4168977288258213,
+      "learning_rate": 1.0240414418722088e-07,
+      "loss": 0.6014,
+      "step": 14646
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5019924001922447,
+      "learning_rate": 1.021955517701484e-07,
+      "loss": 0.6938,
+      "step": 14647
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5288686586859361,
+      "learning_rate": 1.0198716982264045e-07,
+      "loss": 0.8473,
+      "step": 14648
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.525877108312384,
+      "learning_rate": 1.0177899835365323e-07,
+      "loss": 0.6642,
+      "step": 14649
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2222591609831204,
+      "learning_rate": 1.015710373721307e-07,
+      "loss": 0.5816,
+      "step": 14650
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1220146697258364,
+      "learning_rate": 1.0136328688700958e-07,
+      "loss": 0.6491,
+      "step": 14651
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.362749970154329,
+      "learning_rate": 1.0115574690721886e-07,
+      "loss": 0.6584,
+      "step": 14652
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1397388524980407,
+      "learning_rate": 1.0094841744167583e-07,
+      "loss": 0.5739,
+      "step": 14653
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3704780417256046,
+      "learning_rate": 1.0074129849928948e-07,
+      "loss": 0.575,
+      "step": 14654
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3685590615421432,
+      "learning_rate": 1.005343900889616e-07,
+      "loss": 0.5696,
+      "step": 14655
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.100385880982696,
+      "learning_rate": 1.0032769221958227e-07,
+      "loss": 0.7209,
+      "step": 14656
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5144456295141435,
+      "learning_rate": 1.0012120490003385e-07,
+      "loss": 0.6118,
+      "step": 14657
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.424032648668969,
+      "learning_rate": 9.99149281391898e-08,
+      "loss": 0.6466,
+      "step": 14658
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4365403638878318,
+      "learning_rate": 9.970886194591467e-08,
+      "loss": 0.6519,
+      "step": 14659
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4313777578549503,
+      "learning_rate": 9.950300632906363e-08,
+      "loss": 0.7211,
+      "step": 14660
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2574804176079433,
+      "learning_rate": 9.92973612974818e-08,
+      "loss": 0.6011,
+      "step": 14661
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2711504219322551,
+      "learning_rate": 9.909192686000713e-08,
+      "loss": 0.4741,
+      "step": 14662
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3468230216772281,
+      "learning_rate": 9.888670302546754e-08,
+      "loss": 0.6702,
+      "step": 14663
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4515839557919092,
+      "learning_rate": 9.8681689802681e-08,
+      "loss": 0.6409,
+      "step": 14664
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.287963297479683,
+      "learning_rate": 9.847688720045878e-08,
+      "loss": 0.5978,
+      "step": 14665
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1958219362280407,
+      "learning_rate": 9.827229522759996e-08,
+      "loss": 0.6262,
+      "step": 14666
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1708537910208445,
+      "learning_rate": 9.806791389289861e-08,
+      "loss": 0.5021,
+      "step": 14667
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0899194492567046,
+      "learning_rate": 9.786374320513547e-08,
+      "loss": 0.6784,
+      "step": 14668
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0122143654126288,
+      "learning_rate": 9.765978317308522e-08,
+      "loss": 0.5464,
+      "step": 14669
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.393513809318925,
+      "learning_rate": 9.745603380551194e-08,
+      "loss": 0.6487,
+      "step": 14670
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7500907304951996,
+      "learning_rate": 9.72524951111714e-08,
+      "loss": 0.6754,
+      "step": 14671
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.30330492433707,
+      "learning_rate": 9.704916709881052e-08,
+      "loss": 0.6209,
+      "step": 14672
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5509349621516577,
+      "learning_rate": 9.684604977716616e-08,
+      "loss": 0.7616,
+      "step": 14673
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9322444649217287,
+      "learning_rate": 9.664314315496692e-08,
+      "loss": 0.7013,
+      "step": 14674
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2147627435238606,
+      "learning_rate": 9.644044724093249e-08,
+      "loss": 0.6862,
+      "step": 14675
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4954198538193388,
+      "learning_rate": 9.6237962043772e-08,
+      "loss": 0.5631,
+      "step": 14676
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5106494018713983,
+      "learning_rate": 9.603568757218795e-08,
+      "loss": 0.737,
+      "step": 14677
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2605696057033868,
+      "learning_rate": 9.583362383487283e-08,
+      "loss": 0.583,
+      "step": 14678
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4739033889318411,
+      "learning_rate": 9.5631770840508e-08,
+      "loss": 0.6225,
+      "step": 14679
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3087438017288562,
+      "learning_rate": 9.543012859776934e-08,
+      "loss": 0.7236,
+      "step": 14680
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2759124032754587,
+      "learning_rate": 9.5228697115321e-08,
+      "loss": 0.6356,
+      "step": 14681
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2609955776461919,
+      "learning_rate": 9.502747640181887e-08,
+      "loss": 0.5775,
+      "step": 14682
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6694932188597476,
+      "learning_rate": 9.482646646591042e-08,
+      "loss": 0.5544,
+      "step": 14683
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.389014236361807,
+      "learning_rate": 9.462566731623213e-08,
+      "loss": 0.6956,
+      "step": 14684
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4517901331850114,
+      "learning_rate": 9.442507896141484e-08,
+      "loss": 0.6027,
+      "step": 14685
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2559318699529654,
+      "learning_rate": 9.422470141007667e-08,
+      "loss": 0.6319,
+      "step": 14686
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3029489270762136,
+      "learning_rate": 9.402453467082906e-08,
+      "loss": 0.667,
+      "step": 14687
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4468677375698793,
+      "learning_rate": 9.382457875227346e-08,
+      "loss": 0.6419,
+      "step": 14688
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9623683318824844,
+      "learning_rate": 9.36248336630019e-08,
+      "loss": 0.6124,
+      "step": 14689
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0375508726131197,
+      "learning_rate": 9.342529941159862e-08,
+      "loss": 0.5732,
+      "step": 14690
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0759543660410624,
+      "learning_rate": 9.322597600663785e-08,
+      "loss": 0.5998,
+      "step": 14691
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5874952087124155,
+      "learning_rate": 9.3026863456685e-08,
+      "loss": 0.6614,
+      "step": 14692
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4965499145020909,
+      "learning_rate": 9.282796177029596e-08,
+      "loss": 0.6405,
+      "step": 14693
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.435080496435972,
+      "learning_rate": 9.262927095601782e-08,
+      "loss": 0.6563,
+      "step": 14694
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0618538058761757,
+      "learning_rate": 9.243079102238983e-08,
+      "loss": 0.592,
+      "step": 14695
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6281637917466705,
+      "learning_rate": 9.22325219779402e-08,
+      "loss": 0.725,
+      "step": 14696
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.278351786071308,
+      "learning_rate": 9.20344638311893e-08,
+      "loss": 0.5963,
+      "step": 14697
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2631126146421596,
+      "learning_rate": 9.183661659064869e-08,
+      "loss": 0.6362,
+      "step": 14698
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1235868422585462,
+      "learning_rate": 9.163898026481876e-08,
+      "loss": 0.503,
+      "step": 14699
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4318903527082194,
+      "learning_rate": 9.144155486219442e-08,
+      "loss": 0.6037,
+      "step": 14700
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3837168981404027,
+      "learning_rate": 9.124434039125774e-08,
+      "loss": 0.6478,
+      "step": 14701
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.531115770841403,
+      "learning_rate": 9.104733686048417e-08,
+      "loss": 0.7232,
+      "step": 14702
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4690165988261625,
+      "learning_rate": 9.085054427833972e-08,
+      "loss": 0.6162,
+      "step": 14703
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.36188410868196,
+      "learning_rate": 9.065396265327986e-08,
+      "loss": 0.5801,
+      "step": 14704
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4798618113509776,
+      "learning_rate": 9.045759199375393e-08,
+      "loss": 0.7548,
+      "step": 14705
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.181965088455338,
+      "learning_rate": 9.02614323081985e-08,
+      "loss": 0.6144,
+      "step": 14706
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4165246808133651,
+      "learning_rate": 9.006548360504463e-08,
+      "loss": 0.6812,
+      "step": 14707
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4359085080500422,
+      "learning_rate": 8.986974589271169e-08,
+      "loss": 0.5715,
+      "step": 14708
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5089682784825555,
+      "learning_rate": 8.967421917961072e-08,
+      "loss": 0.6384,
+      "step": 14709
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4400665234680625,
+      "learning_rate": 8.947890347414556e-08,
+      "loss": 0.7138,
+      "step": 14710
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4214107912710787,
+      "learning_rate": 8.928379878470727e-08,
+      "loss": 0.5516,
+      "step": 14711
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.255511645877489,
+      "learning_rate": 8.908890511968083e-08,
+      "loss": 0.602,
+      "step": 14712
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3472533855073903,
+      "learning_rate": 8.889422248744173e-08,
+      "loss": 0.6203,
+      "step": 14713
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2740163632869606,
+      "learning_rate": 8.869975089635552e-08,
+      "loss": 0.6422,
+      "step": 14714
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4253762868295046,
+      "learning_rate": 8.85054903547794e-08,
+      "loss": 0.5639,
+      "step": 14715
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.544108518411187,
+      "learning_rate": 8.831144087106003e-08,
+      "loss": 0.828,
+      "step": 14716
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3598072328550206,
+      "learning_rate": 8.811760245353739e-08,
+      "loss": 0.6313,
+      "step": 14717
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4728652432819078,
+      "learning_rate": 8.792397511054097e-08,
+      "loss": 0.5988,
+      "step": 14718
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4353857196209479,
+      "learning_rate": 8.773055885039072e-08,
+      "loss": 0.6045,
+      "step": 14719
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3647400411942672,
+      "learning_rate": 8.753735368139838e-08,
+      "loss": 0.6923,
+      "step": 14720
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4397778715139624,
+      "learning_rate": 8.734435961186782e-08,
+      "loss": 0.6406,
+      "step": 14721
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2428360963606013,
+      "learning_rate": 8.715157665008967e-08,
+      "loss": 0.5444,
+      "step": 14722
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2600022506379585,
+      "learning_rate": 8.69590048043506e-08,
+      "loss": 0.5467,
+      "step": 14723
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.414557781459269,
+      "learning_rate": 8.676664408292457e-08,
+      "loss": 0.5707,
+      "step": 14724
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1362516890336594,
+      "learning_rate": 8.657449449407884e-08,
+      "loss": 0.6777,
+      "step": 14725
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.138874056664755,
+      "learning_rate": 8.63825560460696e-08,
+      "loss": 0.6594,
+      "step": 14726
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2937515580851573,
+      "learning_rate": 8.619082874714413e-08,
+      "loss": 0.6233,
+      "step": 14727
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.305784105896576,
+      "learning_rate": 8.599931260554417e-08,
+      "loss": 0.6249,
+      "step": 14728
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.324199383027967,
+      "learning_rate": 8.580800762949704e-08,
+      "loss": 0.6253,
+      "step": 14729
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3018417037360586,
+      "learning_rate": 8.561691382722337e-08,
+      "loss": 0.5013,
+      "step": 14730
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2554699690560638,
+      "learning_rate": 8.542603120693716e-08,
+      "loss": 0.6312,
+      "step": 14731
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.276092336439173,
+      "learning_rate": 8.523535977683905e-08,
+      "loss": 0.5968,
+      "step": 14732
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2745860016926338,
+      "learning_rate": 8.504489954512363e-08,
+      "loss": 0.595,
+      "step": 14733
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0641932785375587,
+      "learning_rate": 8.485465051997488e-08,
+      "loss": 0.5882,
+      "step": 14734
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6634935114359914,
+      "learning_rate": 8.466461270956794e-08,
+      "loss": 0.6562,
+      "step": 14735
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4727850698813258,
+      "learning_rate": 8.447478612207072e-08,
+      "loss": 0.5917,
+      "step": 14736
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3415320749938757,
+      "learning_rate": 8.428517076563836e-08,
+      "loss": 0.5943,
+      "step": 14737
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.415228600939866,
+      "learning_rate": 8.409576664842045e-08,
+      "loss": 0.7291,
+      "step": 14738
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5994672221111974,
+      "learning_rate": 8.39065737785566e-08,
+      "loss": 0.6805,
+      "step": 14739
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5587811496558783,
+      "learning_rate": 8.371759216417474e-08,
+      "loss": 0.7283,
+      "step": 14740
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3668760337874857,
+      "learning_rate": 8.352882181339783e-08,
+      "loss": 0.6211,
+      "step": 14741
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1650724832167965,
+      "learning_rate": 8.334026273433659e-08,
+      "loss": 0.5901,
+      "step": 14742
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0877427804263218,
+      "learning_rate": 8.315191493509456e-08,
+      "loss": 0.6217,
+      "step": 14743
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3454409954853783,
+      "learning_rate": 8.296377842376524e-08,
+      "loss": 0.6869,
+      "step": 14744
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.411831364921299,
+      "learning_rate": 8.277585320843273e-08,
+      "loss": 0.6959,
+      "step": 14745
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0320100557106433,
+      "learning_rate": 8.258813929717391e-08,
+      "loss": 0.7753,
+      "step": 14746
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.641240243040886,
+      "learning_rate": 8.240063669805343e-08,
+      "loss": 0.7735,
+      "step": 14747
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3061414708758663,
+      "learning_rate": 8.221334541912984e-08,
+      "loss": 0.5946,
+      "step": 14748
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3703828652291785,
+      "learning_rate": 8.202626546845172e-08,
+      "loss": 0.6602,
+      "step": 14749
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9671827150042208,
+      "learning_rate": 8.183939685405762e-08,
+      "loss": 0.5392,
+      "step": 14750
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3552911576182805,
+      "learning_rate": 8.165273958397835e-08,
+      "loss": 0.5808,
+      "step": 14751
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.5872109778872736,
+      "learning_rate": 8.146629366623415e-08,
+      "loss": 0.6668,
+      "step": 14752
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9351241656830083,
+      "learning_rate": 8.128005910883751e-08,
+      "loss": 0.7135,
+      "step": 14753
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.246635501271347,
+      "learning_rate": 8.109403591979148e-08,
+      "loss": 0.534,
+      "step": 14754
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9552701640275698,
+      "learning_rate": 8.090822410708965e-08,
+      "loss": 0.5886,
+      "step": 14755
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5050496643210483,
+      "learning_rate": 8.072262367871675e-08,
+      "loss": 0.7669,
+      "step": 14756
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.306550909659642,
+      "learning_rate": 8.053723464264863e-08,
+      "loss": 0.5533,
+      "step": 14757
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.254961324665419,
+      "learning_rate": 8.035205700685167e-08,
+      "loss": 0.5647,
+      "step": 14758
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2983411450479927,
+      "learning_rate": 8.016709077928397e-08,
+      "loss": 0.708,
+      "step": 14759
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.216288901160909,
+      "learning_rate": 7.998233596789251e-08,
+      "loss": 0.6464,
+      "step": 14760
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0822906706332116,
+      "learning_rate": 7.97977925806187e-08,
+      "loss": 0.6844,
+      "step": 14761
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3223141767719553,
+      "learning_rate": 7.961346062539066e-08,
+      "loss": 0.6528,
+      "step": 14762
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2614317234047443,
+      "learning_rate": 7.942934011013037e-08,
+      "loss": 0.6232,
+      "step": 14763
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5723877212004527,
+      "learning_rate": 7.924543104275095e-08,
+      "loss": 0.5797,
+      "step": 14764
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3395581082981651,
+      "learning_rate": 7.906173343115386e-08,
+      "loss": 0.5343,
+      "step": 14765
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7696030126479843,
+      "learning_rate": 7.88782472832339e-08,
+      "loss": 0.6768,
+      "step": 14766
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5763527842236031,
+      "learning_rate": 7.86949726068753e-08,
+      "loss": 0.7068,
+      "step": 14767
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0779532608134161,
+      "learning_rate": 7.851190940995402e-08,
+      "loss": 0.6059,
+      "step": 14768
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.099676523090801,
+      "learning_rate": 7.832905770033705e-08,
+      "loss": 0.5043,
+      "step": 14769
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.371953236234462,
+      "learning_rate": 7.814641748588148e-08,
+      "loss": 0.608,
+      "step": 14770
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.310592453575445,
+      "learning_rate": 7.796398877443545e-08,
+      "loss": 0.5963,
+      "step": 14771
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4610920960874303,
+      "learning_rate": 7.778177157383993e-08,
+      "loss": 0.6141,
+      "step": 14772
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9660835267181923,
+      "learning_rate": 7.759976589192309e-08,
+      "loss": 0.5889,
+      "step": 14773
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2901270039647637,
+      "learning_rate": 7.7417971736507e-08,
+      "loss": 0.5906,
+      "step": 14774
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8908282523804592,
+      "learning_rate": 7.723638911540432e-08,
+      "loss": 0.6566,
+      "step": 14775
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9651518685209153,
+      "learning_rate": 7.705501803641769e-08,
+      "loss": 0.6832,
+      "step": 14776
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5634923071210602,
+      "learning_rate": 7.687385850734086e-08,
+      "loss": 0.6389,
+      "step": 14777
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5536862595690786,
+      "learning_rate": 7.669291053595818e-08,
+      "loss": 0.5621,
+      "step": 14778
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1303139020956943,
+      "learning_rate": 7.651217413004674e-08,
+      "loss": 0.6613,
+      "step": 14779
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5871894320299254,
+      "learning_rate": 7.633164929737202e-08,
+      "loss": 0.5949,
+      "step": 14780
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4310797804203916,
+      "learning_rate": 7.615133604569113e-08,
+      "loss": 0.6178,
+      "step": 14781
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.945438658154778,
+      "learning_rate": 7.597123438275399e-08,
+      "loss": 0.6363,
+      "step": 14782
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3170146465835373,
+      "learning_rate": 7.579134431629942e-08,
+      "loss": 0.6589,
+      "step": 14783
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2334235244180993,
+      "learning_rate": 7.561166585405789e-08,
+      "loss": 0.5672,
+      "step": 14784
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2446963333096173,
+      "learning_rate": 7.543219900374987e-08,
+      "loss": 0.6246,
+      "step": 14785
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.160632204359799,
+      "learning_rate": 7.525294377308812e-08,
+      "loss": 0.6127,
+      "step": 14786
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4177341557406513,
+      "learning_rate": 7.507390016977478e-08,
+      "loss": 0.6325,
+      "step": 14787
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4469791527498848,
+      "learning_rate": 7.489506820150483e-08,
+      "loss": 0.6835,
+      "step": 14788
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.421689573372018,
+      "learning_rate": 7.47164478759621e-08,
+      "loss": 0.7077,
+      "step": 14789
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5344775050044515,
+      "learning_rate": 7.453803920082381e-08,
+      "loss": 0.6575,
+      "step": 14790
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5557787230950326,
+      "learning_rate": 7.435984218375436e-08,
+      "loss": 0.503,
+      "step": 14791
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3719285125727962,
+      "learning_rate": 7.418185683241319e-08,
+      "loss": 0.6605,
+      "step": 14792
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4468139868972703,
+      "learning_rate": 7.400408315444807e-08,
+      "loss": 0.659,
+      "step": 14793
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2822698612053542,
+      "learning_rate": 7.382652115749789e-08,
+      "loss": 0.6269,
+      "step": 14794
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.614449313443492,
+      "learning_rate": 7.364917084919376e-08,
+      "loss": 0.607,
+      "step": 14795
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3478835790816452,
+      "learning_rate": 7.347203223715571e-08,
+      "loss": 0.6323,
+      "step": 14796
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2331374661376613,
+      "learning_rate": 7.329510532899708e-08,
+      "loss": 0.609,
+      "step": 14797
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6003687729818756,
+      "learning_rate": 7.311839013231959e-08,
+      "loss": 0.6389,
+      "step": 14798
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0847323652184138,
+      "learning_rate": 7.294188665471769e-08,
+      "loss": 0.5956,
+      "step": 14799
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4497484408694723,
+      "learning_rate": 7.276559490377644e-08,
+      "loss": 0.6161,
+      "step": 14800
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8040338727240899,
+      "learning_rate": 7.258951488707089e-08,
+      "loss": 0.6453,
+      "step": 14801
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3898543090252458,
+      "learning_rate": 7.241364661216776e-08,
+      "loss": 0.6356,
+      "step": 14802
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5161360525469378,
+      "learning_rate": 7.223799008662491e-08,
+      "loss": 0.6461,
+      "step": 14803
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3947606647596351,
+      "learning_rate": 7.206254531799018e-08,
+      "loss": 0.6737,
+      "step": 14804
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3634570422581898,
+      "learning_rate": 7.188731231380253e-08,
+      "loss": 0.6185,
+      "step": 14805
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.280973282049882,
+      "learning_rate": 7.171229108159262e-08,
+      "loss": 0.5593,
+      "step": 14806
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3848898365492106,
+      "learning_rate": 7.153748162888164e-08,
+      "loss": 0.6477,
+      "step": 14807
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.228361508762804,
+      "learning_rate": 7.136288396318192e-08,
+      "loss": 0.6221,
+      "step": 14808
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2234914273855486,
+      "learning_rate": 7.118849809199524e-08,
+      "loss": 0.6306,
+      "step": 14809
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4604709279403496,
+      "learning_rate": 7.10143240228156e-08,
+      "loss": 0.6926,
+      "step": 14810
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1757761070003003,
+      "learning_rate": 7.084036176312814e-08,
+      "loss": 0.5817,
+      "step": 14811
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3512819943189074,
+      "learning_rate": 7.066661132040853e-08,
+      "loss": 0.6061,
+      "step": 14812
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3487073503375386,
+      "learning_rate": 7.049307270212246e-08,
+      "loss": 0.6802,
+      "step": 14813
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2471803745549868,
+      "learning_rate": 7.031974591572732e-08,
+      "loss": 0.652,
+      "step": 14814
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.373040775486465,
+      "learning_rate": 7.014663096867214e-08,
+      "loss": 0.7033,
+      "step": 14815
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4769327380330448,
+      "learning_rate": 6.99737278683954e-08,
+      "loss": 0.7115,
+      "step": 14816
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4358831342360996,
+      "learning_rate": 6.980103662232728e-08,
+      "loss": 0.6588,
+      "step": 14817
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7212256765464744,
+      "learning_rate": 6.962855723788909e-08,
+      "loss": 0.7499,
+      "step": 14818
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.075382798521717,
+      "learning_rate": 6.945628972249208e-08,
+      "loss": 0.6694,
+      "step": 14819
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.290322441030323,
+      "learning_rate": 6.928423408353979e-08,
+      "loss": 0.6666,
+      "step": 14820
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0880424982148025,
+      "learning_rate": 6.91123903284241e-08,
+      "loss": 0.671,
+      "step": 14821
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0881818998149706,
+      "learning_rate": 6.894075846453185e-08,
+      "loss": 0.6952,
+      "step": 14822
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2730932302982534,
+      "learning_rate": 6.876933849923662e-08,
+      "loss": 0.6427,
+      "step": 14823
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.288253484095809,
+      "learning_rate": 6.859813043990526e-08,
+      "loss": 0.6885,
+      "step": 14824
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7703768903255173,
+      "learning_rate": 6.842713429389525e-08,
+      "loss": 0.7333,
+      "step": 14825
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.914008888431789,
+      "learning_rate": 6.825635006855458e-08,
+      "loss": 0.6456,
+      "step": 14826
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2986168148443007,
+      "learning_rate": 6.808577777122183e-08,
+      "loss": 0.6376,
+      "step": 14827
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1293274024403817,
+      "learning_rate": 6.79154174092278e-08,
+      "loss": 0.6387,
+      "step": 14828
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4482350268053596,
+      "learning_rate": 6.77452689898922e-08,
+      "loss": 0.6303,
+      "step": 14829
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1808909361851454,
+      "learning_rate": 6.757533252052695e-08,
+      "loss": 0.612,
+      "step": 14830
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.431486618418371,
+      "learning_rate": 6.74056080084351e-08,
+      "loss": 0.6155,
+      "step": 14831
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.046935809073082,
+      "learning_rate": 6.723609546090915e-08,
+      "loss": 0.6333,
+      "step": 14832
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.156619870310285,
+      "learning_rate": 6.706679488523494e-08,
+      "loss": 0.7025,
+      "step": 14833
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.273998497472841,
+      "learning_rate": 6.689770628868609e-08,
+      "loss": 0.6105,
+      "step": 14834
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6028453473906472,
+      "learning_rate": 6.672882967852956e-08,
+      "loss": 0.8064,
+      "step": 14835
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.736612270519681,
+      "learning_rate": 6.656016506202234e-08,
+      "loss": 0.7206,
+      "step": 14836
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.349562171707979,
+      "learning_rate": 6.639171244641252e-08,
+      "loss": 0.6201,
+      "step": 14837
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1923320742622758,
+      "learning_rate": 6.622347183893818e-08,
+      "loss": 0.5743,
+      "step": 14838
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3433662809514648,
+      "learning_rate": 6.605544324682855e-08,
+      "loss": 0.5521,
+      "step": 14839
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1251235214151667,
+      "learning_rate": 6.58876266773062e-08,
+      "loss": 0.5997,
+      "step": 14840
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2273678350188753,
+      "learning_rate": 6.572002213758144e-08,
+      "loss": 0.5852,
+      "step": 14841
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4447562880294957,
+      "learning_rate": 6.555262963485576e-08,
+      "loss": 0.5786,
+      "step": 14842
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4036736430311143,
+      "learning_rate": 6.53854491763234e-08,
+      "loss": 0.7092,
+      "step": 14843
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3395244829858775,
+      "learning_rate": 6.521848076916859e-08,
+      "loss": 0.6549,
+      "step": 14844
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.52648824210296,
+      "learning_rate": 6.505172442056562e-08,
+      "loss": 0.7383,
+      "step": 14845
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3003718743788946,
+      "learning_rate": 6.488518013768096e-08,
+      "loss": 0.6772,
+      "step": 14846
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4817608018785136,
+      "learning_rate": 6.471884792767169e-08,
+      "loss": 0.6599,
+      "step": 14847
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1885369654021567,
+      "learning_rate": 6.455272779768429e-08,
+      "loss": 0.5787,
+      "step": 14848
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5974194994284059,
+      "learning_rate": 6.438681975485805e-08,
+      "loss": 0.6167,
+      "step": 14849
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4180070084433638,
+      "learning_rate": 6.422112380632228e-08,
+      "loss": 0.6806,
+      "step": 14850
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.164904681582355,
+      "learning_rate": 6.40556399591985e-08,
+      "loss": 0.5827,
+      "step": 14851
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2360347789829014,
+      "learning_rate": 6.389036822059546e-08,
+      "loss": 0.6527,
+      "step": 14852
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1796596611805028,
+      "learning_rate": 6.372530859761694e-08,
+      "loss": 0.6128,
+      "step": 14853
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.348533235803264,
+      "learning_rate": 6.356046109735614e-08,
+      "loss": 0.6789,
+      "step": 14854
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7014251980477757,
+      "learning_rate": 6.339582572689573e-08,
+      "loss": 0.6001,
+      "step": 14855
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9505695725931067,
+      "learning_rate": 6.323140249331116e-08,
+      "loss": 0.8489,
+      "step": 14856
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6683786202429194,
+      "learning_rate": 6.306719140366791e-08,
+      "loss": 0.643,
+      "step": 14857
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3349379975409266,
+      "learning_rate": 6.290319246502363e-08,
+      "loss": 0.612,
+      "step": 14858
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3722827493143333,
+      "learning_rate": 6.273940568442327e-08,
+      "loss": 0.4911,
+      "step": 14859
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3929264675644684,
+      "learning_rate": 6.25758310689073e-08,
+      "loss": 0.7158,
+      "step": 14860
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.8159861641966755,
+      "learning_rate": 6.241246862550398e-08,
+      "loss": 0.6009,
+      "step": 14861
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8193314614790286,
+      "learning_rate": 6.224931836123327e-08,
+      "loss": 0.6405,
+      "step": 14862
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5885274069753788,
+      "learning_rate": 6.208638028310677e-08,
+      "loss": 0.6356,
+      "step": 14863
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.459708806812008,
+      "learning_rate": 6.192365439812553e-08,
+      "loss": 0.6699,
+      "step": 14864
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1571776485912115,
+      "learning_rate": 6.176114071328287e-08,
+      "loss": 0.4847,
+      "step": 14865
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1760215366376003,
+      "learning_rate": 6.159883923556154e-08,
+      "loss": 0.5374,
+      "step": 14866
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4577135424634506,
+      "learning_rate": 6.143674997193649e-08,
+      "loss": 0.672,
+      "step": 14867
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3605847286169155,
+      "learning_rate": 6.127487292937328e-08,
+      "loss": 0.6266,
+      "step": 14868
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4896544126565034,
+      "learning_rate": 6.111320811482802e-08,
+      "loss": 0.6202,
+      "step": 14869
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5796099498617173,
+      "learning_rate": 6.095175553524735e-08,
+      "loss": 0.608,
+      "step": 14870
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4490392776146428,
+      "learning_rate": 6.079051519757018e-08,
+      "loss": 0.8506,
+      "step": 14871
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.451400975315284,
+      "learning_rate": 6.062948710872373e-08,
+      "loss": 0.7637,
+      "step": 14872
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3767329135426696,
+      "learning_rate": 6.046867127563027e-08,
+      "loss": 0.6046,
+      "step": 14873
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4601151689847593,
+      "learning_rate": 6.030806770519815e-08,
+      "loss": 0.6251,
+      "step": 14874
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5358311089097954,
+      "learning_rate": 6.014767640432905e-08,
+      "loss": 0.5804,
+      "step": 14875
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5930609303142187,
+      "learning_rate": 5.998749737991693e-08,
+      "loss": 0.6674,
+      "step": 14876
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4647426489296707,
+      "learning_rate": 5.98275306388435e-08,
+      "loss": 0.6585,
+      "step": 14877
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5520673889588035,
+      "learning_rate": 5.96677761879838e-08,
+      "loss": 0.6608,
+      "step": 14878
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1007170709098617,
+      "learning_rate": 5.9508234034202364e-08,
+      "loss": 0.5414,
+      "step": 14879
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5770195617041958,
+      "learning_rate": 5.9348904184354795e-08,
+      "loss": 0.6032,
+      "step": 14880
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.46834791343997,
+      "learning_rate": 5.918978664528841e-08,
+      "loss": 0.6849,
+      "step": 14881
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5113216698628473,
+      "learning_rate": 5.903088142384106e-08,
+      "loss": 0.6385,
+      "step": 14882
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3992089984422178,
+      "learning_rate": 5.887218852684007e-08,
+      "loss": 0.6357,
+      "step": 14883
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4120043841105556,
+      "learning_rate": 5.871370796110665e-08,
+      "loss": 0.749,
+      "step": 14884
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3786987610106751,
+      "learning_rate": 5.855543973344924e-08,
+      "loss": 0.583,
+      "step": 14885
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4794015878556388,
+      "learning_rate": 5.8397383850670176e-08,
+      "loss": 0.6869,
+      "step": 14886
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3628429236073871,
+      "learning_rate": 5.823954031956069e-08,
+      "loss": 0.6484,
+      "step": 14887
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.617931241771711,
+      "learning_rate": 5.808190914690426e-08,
+      "loss": 0.7106,
+      "step": 14888
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4247730221400798,
+      "learning_rate": 5.7924490339474335e-08,
+      "loss": 0.599,
+      "step": 14889
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4087689314170262,
+      "learning_rate": 5.776728390403552e-08,
+      "loss": 0.5921,
+      "step": 14890
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6025749683581156,
+      "learning_rate": 5.761028984734407e-08,
+      "loss": 0.7897,
+      "step": 14891
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1574076694076951,
+      "learning_rate": 5.74535081761457e-08,
+      "loss": 0.6586,
+      "step": 14892
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5244947156483106,
+      "learning_rate": 5.7296938897177245e-08,
+      "loss": 0.6299,
+      "step": 14893
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.521725904724582,
+      "learning_rate": 5.7140582017167764e-08,
+      "loss": 0.57,
+      "step": 14894
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.092164783068618,
+      "learning_rate": 5.6984437542835225e-08,
+      "loss": 0.5364,
+      "step": 14895
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4432882940390774,
+      "learning_rate": 5.682850548089036e-08,
+      "loss": 0.6698,
+      "step": 14896
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.364610538717114,
+      "learning_rate": 5.6672785838033925e-08,
+      "loss": 0.6054,
+      "step": 14897
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7117523628187687,
+      "learning_rate": 5.651727862095668e-08,
+      "loss": 0.6713,
+      "step": 14898
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5346345303559163,
+      "learning_rate": 5.636198383634217e-08,
+      "loss": 0.5917,
+      "step": 14899
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4058291869447217,
+      "learning_rate": 5.620690149086283e-08,
+      "loss": 0.643,
+      "step": 14900
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3211722532056789,
+      "learning_rate": 5.6052031591183884e-08,
+      "loss": 0.6277,
+      "step": 14901
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1810982867770576,
+      "learning_rate": 5.5897374143960013e-08,
+      "loss": 0.5225,
+      "step": 14902
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.522492294513865,
+      "learning_rate": 5.574292915583646e-08,
+      "loss": 0.7025,
+      "step": 14903
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.400772410559154,
+      "learning_rate": 5.558869663345123e-08,
+      "loss": 0.711,
+      "step": 14904
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2630013382438057,
+      "learning_rate": 5.543467658343127e-08,
+      "loss": 0.4879,
+      "step": 14905
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3641782009397614,
+      "learning_rate": 5.5280869012395154e-08,
+      "loss": 0.5599,
+      "step": 14906
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3332037092520999,
+      "learning_rate": 5.5127273926953165e-08,
+      "loss": 0.6244,
+      "step": 14907
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.322151575695447,
+      "learning_rate": 5.4973891333704453e-08,
+      "loss": 0.6765,
+      "step": 14908
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3977481737917996,
+      "learning_rate": 5.482072123924098e-08,
+      "loss": 0.607,
+      "step": 14909
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.159938800546897,
+      "learning_rate": 5.46677636501447e-08,
+      "loss": 0.5336,
+      "step": 14910
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.251269484028178,
+      "learning_rate": 5.4515018572988687e-08,
+      "loss": 0.6011,
+      "step": 14911
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.053762331494429,
+      "learning_rate": 5.436248601433658e-08,
+      "loss": 0.8262,
+      "step": 14912
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0926187907704576,
+      "learning_rate": 5.4210165980742025e-08,
+      "loss": 0.5833,
+      "step": 14913
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.414126206356319,
+      "learning_rate": 5.405805847875256e-08,
+      "loss": 0.605,
+      "step": 14914
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4773314285703165,
+      "learning_rate": 5.3906163514902966e-08,
+      "loss": 0.6446,
+      "step": 14915
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3360673904896028,
+      "learning_rate": 5.3754481095721344e-08,
+      "loss": 0.6504,
+      "step": 14916
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4917678656344353,
+      "learning_rate": 5.3603011227725265e-08,
+      "loss": 0.6241,
+      "step": 14917
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.353767761613045,
+      "learning_rate": 5.3451753917423964e-08,
+      "loss": 0.5985,
+      "step": 14918
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5011231251358088,
+      "learning_rate": 5.330070917131724e-08,
+      "loss": 0.6343,
+      "step": 14919
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1219383853884175,
+      "learning_rate": 5.3149876995896575e-08,
+      "loss": 0.5684,
+      "step": 14920
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6318168910212,
+      "learning_rate": 5.299925739764233e-08,
+      "loss": 0.7176,
+      "step": 14921
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2602021598056048,
+      "learning_rate": 5.284885038302823e-08,
+      "loss": 0.6398,
+      "step": 14922
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3318650045066223,
+      "learning_rate": 5.269865595851631e-08,
+      "loss": 0.6031,
+      "step": 14923
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5573987267963625,
+      "learning_rate": 5.2548674130561974e-08,
+      "loss": 0.6386,
+      "step": 14924
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5968879554535444,
+      "learning_rate": 5.239890490560895e-08,
+      "loss": 0.6869,
+      "step": 14925
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.737113798288398,
+      "learning_rate": 5.224934829009376e-08,
+      "loss": 0.6171,
+      "step": 14926
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8101885568680651,
+      "learning_rate": 5.21000042904446e-08,
+      "loss": 0.5897,
+      "step": 14927
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3023628649470165,
+      "learning_rate": 5.195087291307632e-08,
+      "loss": 0.5926,
+      "step": 14928
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5720158562357234,
+      "learning_rate": 5.1801954164399925e-08,
+      "loss": 0.5341,
+      "step": 14929
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4427388547472217,
+      "learning_rate": 5.165324805081362e-08,
+      "loss": 0.7768,
+      "step": 14930
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2392971143643068,
+      "learning_rate": 5.1504754578707294e-08,
+      "loss": 0.5976,
+      "step": 14931
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4006372807896634,
+      "learning_rate": 5.135647375446307e-08,
+      "loss": 0.5893,
+      "step": 14932
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.431242138833272,
+      "learning_rate": 5.120840558445195e-08,
+      "loss": 0.6193,
+      "step": 14933
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2740779830972004,
+      "learning_rate": 5.106055007503774e-08,
+      "loss": 0.5944,
+      "step": 14934
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3842418145218374,
+      "learning_rate": 5.091290723257314e-08,
+      "loss": 0.6311,
+      "step": 14935
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4940600194545453,
+      "learning_rate": 5.076547706340307e-08,
+      "loss": 0.6583,
+      "step": 14936
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3677116585607125,
+      "learning_rate": 5.061825957386302e-08,
+      "loss": 0.6217,
+      "step": 14937
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6338141499900474,
+      "learning_rate": 5.047125477027959e-08,
+      "loss": 0.6501,
+      "step": 14938
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2309842976306526,
+      "learning_rate": 5.0324462658969395e-08,
+      "loss": 0.6328,
+      "step": 14939
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1083060444849577,
+      "learning_rate": 5.017788324624073e-08,
+      "loss": 0.6994,
+      "step": 14940
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6543769181672587,
+      "learning_rate": 5.003151653839189e-08,
+      "loss": 0.7979,
+      "step": 14941
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.617350936332202,
+      "learning_rate": 4.988536254171284e-08,
+      "loss": 0.6664,
+      "step": 14942
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0837527059847982,
+      "learning_rate": 4.9739421262484656e-08,
+      "loss": 0.6317,
+      "step": 14943
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.044713823626411,
+      "learning_rate": 4.959369270697789e-08,
+      "loss": 0.6645,
+      "step": 14944
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3832106600989502,
+      "learning_rate": 4.944817688145642e-08,
+      "loss": 0.6502,
+      "step": 14945
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.729220305943598,
+      "learning_rate": 4.93028737921708e-08,
+      "loss": 0.56,
+      "step": 14946
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6025758924306666,
+      "learning_rate": 4.915778344536715e-08,
+      "loss": 0.57,
+      "step": 14947
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3182180376700214,
+      "learning_rate": 4.9012905847279913e-08,
+      "loss": 0.6219,
+      "step": 14948
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0544053943054115,
+      "learning_rate": 4.886824100413412e-08,
+      "loss": 0.6034,
+      "step": 14949
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4125904388689738,
+      "learning_rate": 4.872378892214702e-08,
+      "loss": 0.4987,
+      "step": 14950
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3983294048446022,
+      "learning_rate": 4.8579549607525866e-08,
+      "loss": 0.6217,
+      "step": 14951
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.484786180906282,
+      "learning_rate": 4.843552306646904e-08,
+      "loss": 0.703,
+      "step": 14952
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7015885568155291,
+      "learning_rate": 4.829170930516491e-08,
+      "loss": 0.6759,
+      "step": 14953
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0577180606347973,
+      "learning_rate": 4.814810832979411e-08,
+      "loss": 0.6156,
+      "step": 14954
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3459482676963792,
+      "learning_rate": 4.8004720146527795e-08,
+      "loss": 0.6529,
+      "step": 14955
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.29563342519362,
+      "learning_rate": 4.786154476152716e-08,
+      "loss": 0.6454,
+      "step": 14956
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1456172644429858,
+      "learning_rate": 4.771858218094505e-08,
+      "loss": 0.5411,
+      "step": 14957
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5606154060292217,
+      "learning_rate": 4.7575832410924893e-08,
+      "loss": 0.7007,
+      "step": 14958
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4426350575081166,
+      "learning_rate": 4.743329545760122e-08,
+      "loss": 0.6494,
+      "step": 14959
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5008989037130973,
+      "learning_rate": 4.7290971327098035e-08,
+      "loss": 0.7126,
+      "step": 14960
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3206444673490547,
+      "learning_rate": 4.71488600255321e-08,
+      "loss": 0.6254,
+      "step": 14961
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4389485440987129,
+      "learning_rate": 4.7006961559010745e-08,
+      "loss": 0.6572,
+      "step": 14962
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3254856814669516,
+      "learning_rate": 4.686527593363133e-08,
+      "loss": 0.6494,
+      "step": 14963
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3528266229934631,
+      "learning_rate": 4.67238031554812e-08,
+      "loss": 0.6507,
+      "step": 14964
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4432564528237255,
+      "learning_rate": 4.6582543230641596e-08,
+      "loss": 0.6382,
+      "step": 14965
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3117364353102008,
+      "learning_rate": 4.644149616518212e-08,
+      "loss": 0.7288,
+      "step": 14966
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3348753795679213,
+      "learning_rate": 4.630066196516347e-08,
+      "loss": 0.6112,
+      "step": 14967
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2157190096977502,
+      "learning_rate": 4.6160040636637484e-08,
+      "loss": 0.62,
+      "step": 14968
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2937718349395049,
+      "learning_rate": 4.6019632185647645e-08,
+      "loss": 0.6672,
+      "step": 14969
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2274866133089972,
+      "learning_rate": 4.587943661822747e-08,
+      "loss": 0.5666,
+      "step": 14970
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4248523061170046,
+      "learning_rate": 4.5739453940401024e-08,
+      "loss": 0.5863,
+      "step": 14971
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4829940615712565,
+      "learning_rate": 4.5599684158184055e-08,
+      "loss": 0.6422,
+      "step": 14972
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2904390678755668,
+      "learning_rate": 4.5460127277582863e-08,
+      "loss": 0.5698,
+      "step": 14973
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3807368826431812,
+      "learning_rate": 4.532078330459433e-08,
+      "loss": 0.6373,
+      "step": 14974
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.314450116867437,
+      "learning_rate": 4.518165224520643e-08,
+      "loss": 0.645,
+      "step": 14975
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3971046921521726,
+      "learning_rate": 4.504273410539772e-08,
+      "loss": 0.6681,
+      "step": 14976
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4009477577582812,
+      "learning_rate": 4.490402889113787e-08,
+      "loss": 0.6661,
+      "step": 14977
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.426958927915053,
+      "learning_rate": 4.4765536608388227e-08,
+      "loss": 0.6317,
+      "step": 14978
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5890024598619212,
+      "learning_rate": 4.4627257263098465e-08,
+      "loss": 0.8,
+      "step": 14979
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.427068267761344,
+      "learning_rate": 4.448919086121217e-08,
+      "loss": 0.5673,
+      "step": 14980
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0631879032288352,
+      "learning_rate": 4.435133740866182e-08,
+      "loss": 0.5846,
+      "step": 14981
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4800554695042591,
+      "learning_rate": 4.4213696911371006e-08,
+      "loss": 0.652,
+      "step": 14982
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3394656077263447,
+      "learning_rate": 4.4076269375255e-08,
+      "loss": 0.7196,
+      "step": 14983
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0972833368013013,
+      "learning_rate": 4.393905480621907e-08,
+      "loss": 0.5872,
+      "step": 14984
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4583038369380046,
+      "learning_rate": 4.380205321016018e-08,
+      "loss": 0.6236,
+      "step": 14985
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.430094694148463,
+      "learning_rate": 4.366526459296472e-08,
+      "loss": 0.6945,
+      "step": 14986
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0078437891394576,
+      "learning_rate": 4.352868896051077e-08,
+      "loss": 0.621,
+      "step": 14987
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2113067411859888,
+      "learning_rate": 4.3392326318668076e-08,
+      "loss": 0.6833,
+      "step": 14988
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2918060419848865,
+      "learning_rate": 4.3256176673295846e-08,
+      "loss": 0.6434,
+      "step": 14989
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7140057940567048,
+      "learning_rate": 4.312024003024551e-08,
+      "loss": 0.4628,
+      "step": 14990
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3420915729741194,
+      "learning_rate": 4.2984516395357944e-08,
+      "loss": 0.7093,
+      "step": 14991
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5766098675830136,
+      "learning_rate": 4.284900577446516e-08,
+      "loss": 0.6599,
+      "step": 14992
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.450455268541348,
+      "learning_rate": 4.271370817339082e-08,
+      "loss": 0.7297,
+      "step": 14993
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6516610270587784,
+      "learning_rate": 4.2578623597949174e-08,
+      "loss": 0.6582,
+      "step": 14994
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3327697383825943,
+      "learning_rate": 4.244375205394502e-08,
+      "loss": 0.6074,
+      "step": 14995
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.255461592149346,
+      "learning_rate": 4.230909354717372e-08,
+      "loss": 0.7703,
+      "step": 14996
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.409853486950479,
+      "learning_rate": 4.217464808342175e-08,
+      "loss": 0.6085,
+      "step": 14997
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.545636017021854,
+      "learning_rate": 4.204041566846673e-08,
+      "loss": 0.606,
+      "step": 14998
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.707871172448941,
+      "learning_rate": 4.1906396308077356e-08,
+      "loss": 0.5981,
+      "step": 14999
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4228620436565989,
+      "learning_rate": 4.177259000801237e-08,
+      "loss": 0.5747,
+      "step": 15000
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.391551433762031,
+      "learning_rate": 4.163899677402161e-08,
+      "loss": 0.6399,
+      "step": 15001
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5320760642868354,
+      "learning_rate": 4.150561661184604e-08,
+      "loss": 0.6337,
+      "step": 15002
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4008445529339617,
+      "learning_rate": 4.1372449527217754e-08,
+      "loss": 0.6983,
+      "step": 15003
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2705428795734803,
+      "learning_rate": 4.123949552585826e-08,
+      "loss": 0.5985,
+      "step": 15004
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4370058838996278,
+      "learning_rate": 4.1106754613481347e-08,
+      "loss": 0.5647,
+      "step": 15005
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3205780356043022,
+      "learning_rate": 4.097422679579188e-08,
+      "loss": 0.5838,
+      "step": 15006
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5358206971259498,
+      "learning_rate": 4.0841912078483645e-08,
+      "loss": 0.5978,
+      "step": 15007
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.374948156698008,
+      "learning_rate": 4.0709810467243204e-08,
+      "loss": 0.624,
+      "step": 15008
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3211016413200056,
+      "learning_rate": 4.0577921967747126e-08,
+      "loss": 0.6038,
+      "step": 15009
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2589821169381314,
+      "learning_rate": 4.0446246585663093e-08,
+      "loss": 0.6035,
+      "step": 15010
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.481970947771804,
+      "learning_rate": 4.031478432664937e-08,
+      "loss": 0.7115,
+      "step": 15011
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3459650325102688,
+      "learning_rate": 4.018353519635476e-08,
+      "loss": 0.6398,
+      "step": 15012
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4643712119604468,
+      "learning_rate": 4.0052499200420314e-08,
+      "loss": 0.7521,
+      "step": 15013
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5313241079897972,
+      "learning_rate": 3.9921676344475966e-08,
+      "loss": 0.734,
+      "step": 15014
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4209210106440793,
+      "learning_rate": 3.979106663414389e-08,
+      "loss": 0.5973,
+      "step": 15015
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4163838991602287,
+      "learning_rate": 3.9660670075036823e-08,
+      "loss": 0.6837,
+      "step": 15016
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1814292973976728,
+      "learning_rate": 3.953048667275805e-08,
+      "loss": 0.5124,
+      "step": 15017
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2411164983360117,
+      "learning_rate": 3.9400516432901434e-08,
+      "loss": 0.5323,
+      "step": 15018
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5738872331243576,
+      "learning_rate": 3.927075936105307e-08,
+      "loss": 0.5768,
+      "step": 15019
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4172365941164071,
+      "learning_rate": 3.914121546278793e-08,
+      "loss": 0.7492,
+      "step": 15020
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5461771554969286,
+      "learning_rate": 3.901188474367323e-08,
+      "loss": 0.6554,
+      "step": 15021
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2964695463381208,
+      "learning_rate": 3.8882767209266756e-08,
+      "loss": 0.5565,
+      "step": 15022
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2693256106768607,
+      "learning_rate": 3.875386286511684e-08,
+      "loss": 0.5466,
+      "step": 15023
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1873855409327119,
+      "learning_rate": 3.8625171716762385e-08,
+      "loss": 0.7579,
+      "step": 15024
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9248488011304613,
+      "learning_rate": 3.849669376973397e-08,
+      "loss": 0.6081,
+      "step": 15025
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4471066336267888,
+      "learning_rate": 3.836842902955273e-08,
+      "loss": 0.6144,
+      "step": 15026
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4470974306728384,
+      "learning_rate": 3.8240377501729817e-08,
+      "loss": 0.7109,
+      "step": 15027
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5445136908502084,
+      "learning_rate": 3.8112539191768604e-08,
+      "loss": 0.6732,
+      "step": 15028
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2226541688779273,
+      "learning_rate": 3.7984914105162474e-08,
+      "loss": 0.655,
+      "step": 15029
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3158949783683982,
+      "learning_rate": 3.785750224739482e-08,
+      "loss": 0.5359,
+      "step": 15030
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.481074328681093,
+      "learning_rate": 3.7730303623942364e-08,
+      "loss": 0.6358,
+      "step": 15031
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.424005413869755,
+      "learning_rate": 3.760331824027019e-08,
+      "loss": 0.6075,
+      "step": 15032
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.148992447343872,
+      "learning_rate": 3.7476546101835045e-08,
+      "loss": 0.7191,
+      "step": 15033
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1569481814385079,
+      "learning_rate": 3.7349987214084784e-08,
+      "loss": 0.7518,
+      "step": 15034
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5856458724572164,
+      "learning_rate": 3.722364158245839e-08,
+      "loss": 0.6335,
+      "step": 15035
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.399257907778065,
+      "learning_rate": 3.709750921238486e-08,
+      "loss": 0.6393,
+      "step": 15036
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1010179750487545,
+      "learning_rate": 3.697159010928375e-08,
+      "loss": 0.6382,
+      "step": 15037
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3610038033425824,
+      "learning_rate": 3.6845884278566836e-08,
+      "loss": 0.659,
+      "step": 15038
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1545057901636322,
+      "learning_rate": 3.672039172563646e-08,
+      "loss": 0.5006,
+      "step": 15039
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3200512035062055,
+      "learning_rate": 3.659511245588387e-08,
+      "loss": 0.5579,
+      "step": 15040
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2916781863141344,
+      "learning_rate": 3.647004647469365e-08,
+      "loss": 0.6556,
+      "step": 15041
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3442616789516304,
+      "learning_rate": 3.634519378744039e-08,
+      "loss": 0.5848,
+      "step": 15042
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1079440578075577,
+      "learning_rate": 3.622055439948813e-08,
+      "loss": 0.6547,
+      "step": 15043
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3252538457973129,
+      "learning_rate": 3.609612831619369e-08,
+      "loss": 0.6557,
+      "step": 15044
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2549400900945817,
+      "learning_rate": 3.597191554290391e-08,
+      "loss": 0.6265,
+      "step": 15045
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.374192572115594,
+      "learning_rate": 3.584791608495675e-08,
+      "loss": 0.6444,
+      "step": 15046
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0240697203845475,
+      "learning_rate": 3.5724129947680156e-08,
+      "loss": 0.6549,
+      "step": 15047
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3350091102873174,
+      "learning_rate": 3.560055713639321e-08,
+      "loss": 0.6278,
+      "step": 15048
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2731650584156367,
+      "learning_rate": 3.547719765640778e-08,
+      "loss": 0.5649,
+      "step": 15049
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1989097255184031,
+      "learning_rate": 3.5354051513022405e-08,
+      "loss": 0.6589,
+      "step": 15050
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4283572962692204,
+      "learning_rate": 3.523111871153118e-08,
+      "loss": 0.6098,
+      "step": 15051
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5159842215976378,
+      "learning_rate": 3.510839925721543e-08,
+      "loss": 0.67,
+      "step": 15052
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8241255007384902,
+      "learning_rate": 3.4985893155349835e-08,
+      "loss": 0.6561,
+      "step": 15053
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3457021214771792,
+      "learning_rate": 3.4863600411197404e-08,
+      "loss": 0.6315,
+      "step": 15054
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4508349892134085,
+      "learning_rate": 3.474152103001449e-08,
+      "loss": 0.6465,
+      "step": 15055
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0218081386848499,
+      "learning_rate": 3.461965501704634e-08,
+      "loss": 0.6397,
+      "step": 15056
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3428960384085191,
+      "learning_rate": 3.449800237753043e-08,
+      "loss": 0.6255,
+      "step": 15057
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3527520398098596,
+      "learning_rate": 3.4376563116693685e-08,
+      "loss": 0.6243,
+      "step": 15058
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2992201979161815,
+      "learning_rate": 3.425533723975527e-08,
+      "loss": 0.5671,
+      "step": 15059
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3802020864145956,
+      "learning_rate": 3.413432475192435e-08,
+      "loss": 0.6883,
+      "step": 15060
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2897133461809849,
+      "learning_rate": 3.401352565840121e-08,
+      "loss": 0.629,
+      "step": 15061
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.527689799655525,
+      "learning_rate": 3.3892939964376704e-08,
+      "loss": 0.5861,
+      "step": 15062
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3429853727052017,
+      "learning_rate": 3.377256767503223e-08,
+      "loss": 0.6323,
+      "step": 15063
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5026715135567084,
+      "learning_rate": 3.365240879554144e-08,
+      "loss": 0.668,
+      "step": 15064
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2177397837427708,
+      "learning_rate": 3.353246333106741e-08,
+      "loss": 0.5096,
+      "step": 15065
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5400190308792263,
+      "learning_rate": 3.3412731286763256e-08,
+      "loss": 0.6303,
+      "step": 15066
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2527871307200038,
+      "learning_rate": 3.329321266777652e-08,
+      "loss": 0.6035,
+      "step": 15067
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3911307353145372,
+      "learning_rate": 3.3173907479240875e-08,
+      "loss": 0.629,
+      "step": 15068
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4453134638597738,
+      "learning_rate": 3.3054815726285e-08,
+      "loss": 0.7407,
+      "step": 15069
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3629764675019285,
+      "learning_rate": 3.2935937414025345e-08,
+      "loss": 0.5845,
+      "step": 15070
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4926178273459931,
+      "learning_rate": 3.281727254757061e-08,
+      "loss": 0.5859,
+      "step": 15071
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2573938361662627,
+      "learning_rate": 3.269882113202005e-08,
+      "loss": 0.5694,
+      "step": 15072
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3934510096088772,
+      "learning_rate": 3.258058317246404e-08,
+      "loss": 0.6084,
+      "step": 15073
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5458689909610845,
+      "learning_rate": 3.2462558673983516e-08,
+      "loss": 0.7067,
+      "step": 15074
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3755974639628594,
+      "learning_rate": 3.234474764165052e-08,
+      "loss": 0.7325,
+      "step": 15075
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2967376464217522,
+      "learning_rate": 3.222715008052657e-08,
+      "loss": 0.5836,
+      "step": 15076
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0339856294589032,
+      "learning_rate": 3.2109765995665957e-08,
+      "loss": 0.6194,
+      "step": 15077
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2885099790715946,
+      "learning_rate": 3.1992595392112966e-08,
+      "loss": 0.6551,
+      "step": 15078
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9953111645035786,
+      "learning_rate": 3.1875638274902476e-08,
+      "loss": 0.6636,
+      "step": 15079
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2321559281807346,
+      "learning_rate": 3.1758894649060456e-08,
+      "loss": 0.6589,
+      "step": 15080
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3549918377794796,
+      "learning_rate": 3.164236451960345e-08,
+      "loss": 0.6461,
+      "step": 15081
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4266009320429116,
+      "learning_rate": 3.152604789153912e-08,
+      "loss": 0.4842,
+      "step": 15082
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.368895660530362,
+      "learning_rate": 3.140994476986625e-08,
+      "loss": 0.5699,
+      "step": 15083
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2699304194227579,
+      "learning_rate": 3.129405515957307e-08,
+      "loss": 0.5681,
+      "step": 15084
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.186140683642399,
+      "learning_rate": 3.117837906564114e-08,
+      "loss": 0.6175,
+      "step": 15085
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5487323029733808,
+      "learning_rate": 3.106291649303983e-08,
+      "loss": 0.656,
+      "step": 15086
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.531208725658464,
+      "learning_rate": 3.094766744673128e-08,
+      "loss": 0.5804,
+      "step": 15087
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3473910010971504,
+      "learning_rate": 3.083263193166819e-08,
+      "loss": 0.8097,
+      "step": 15088
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5483359394660834,
+      "learning_rate": 3.071780995279439e-08,
+      "loss": 0.6826,
+      "step": 15089
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5039815014786764,
+      "learning_rate": 3.060320151504315e-08,
+      "loss": 0.6352,
+      "step": 15090
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5320966720623026,
+      "learning_rate": 3.0488806623339414e-08,
+      "loss": 0.6307,
+      "step": 15091
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0263531707612095,
+      "learning_rate": 3.0374625282599826e-08,
+      "loss": 0.6266,
+      "step": 15092
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3508653439316651,
+      "learning_rate": 3.026065749773044e-08,
+      "loss": 0.6399,
+      "step": 15093
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4870100486098665,
+      "learning_rate": 3.014690327362846e-08,
+      "loss": 0.5825,
+      "step": 15094
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1621676586272898,
+      "learning_rate": 3.003336261518275e-08,
+      "loss": 0.5501,
+      "step": 15095
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2941441433410428,
+      "learning_rate": 2.992003552727219e-08,
+      "loss": 0.6102,
+      "step": 15096
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5064594604793176,
+      "learning_rate": 2.980692201476676e-08,
+      "loss": 0.6565,
+      "step": 15097
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5050713585353686,
+      "learning_rate": 2.969402208252703e-08,
+      "loss": 0.524,
+      "step": 15098
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.454416131552231,
+      "learning_rate": 2.9581335735404672e-08,
+      "loss": 0.774,
+      "step": 15099
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4576624114177876,
+      "learning_rate": 2.9468862978241917e-08,
+      "loss": 0.6016,
+      "step": 15100
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4421088450399489,
+      "learning_rate": 2.9356603815872132e-08,
+      "loss": 0.6091,
+      "step": 15101
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5977930594828598,
+      "learning_rate": 2.9244558253119226e-08,
+      "loss": 0.611,
+      "step": 15102
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.324903525180418,
+      "learning_rate": 2.9132726294798242e-08,
+      "loss": 0.6184,
+      "step": 15103
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.404386650449847,
+      "learning_rate": 2.9021107945714777e-08,
+      "loss": 0.6346,
+      "step": 15104
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3363341696771225,
+      "learning_rate": 2.8909703210664995e-08,
+      "loss": 0.6556,
+      "step": 15105
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.228929790185743,
+      "learning_rate": 2.8798512094436738e-08,
+      "loss": 0.5614,
+      "step": 15106
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6354999305954057,
+      "learning_rate": 2.8687534601807842e-08,
+      "loss": 0.5269,
+      "step": 15107
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.378646186672206,
+      "learning_rate": 2.8576770737547276e-08,
+      "loss": 0.5829,
+      "step": 15108
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.336519304162388,
+      "learning_rate": 2.8466220506414565e-08,
+      "loss": 0.6201,
+      "step": 15109
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.609959108086178,
+      "learning_rate": 2.8355883913160907e-08,
+      "loss": 0.6146,
+      "step": 15110
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7846429528772865,
+      "learning_rate": 2.8245760962527514e-08,
+      "loss": 0.5788,
+      "step": 15111
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2100416043711208,
+      "learning_rate": 2.813585165924615e-08,
+      "loss": 0.6089,
+      "step": 15112
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4867015262091814,
+      "learning_rate": 2.802615600804026e-08,
+      "loss": 0.6649,
+      "step": 15113
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4879100691343992,
+      "learning_rate": 2.79166740136233e-08,
+      "loss": 0.6082,
+      "step": 15114
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.707559964361194,
+      "learning_rate": 2.780740568070095e-08,
+      "loss": 0.6334,
+      "step": 15115
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6275265191836232,
+      "learning_rate": 2.769835101396723e-08,
+      "loss": 0.5758,
+      "step": 15116
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.488992074102693,
+      "learning_rate": 2.75895100181095e-08,
+      "loss": 0.6029,
+      "step": 15117
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6134996034625992,
+      "learning_rate": 2.7480882697805133e-08,
+      "loss": 0.6257,
+      "step": 15118
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3995604848473076,
+      "learning_rate": 2.7372469057721506e-08,
+      "loss": 0.6304,
+      "step": 15119
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6379759626608907,
+      "learning_rate": 2.7264269102517117e-08,
+      "loss": 0.6113,
+      "step": 15120
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1838460242103026,
+      "learning_rate": 2.7156282836842683e-08,
+      "loss": 0.5957,
+      "step": 15121
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.605005947429566,
+      "learning_rate": 2.704851026533728e-08,
+      "loss": 0.6061,
+      "step": 15122
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1903459571730484,
+      "learning_rate": 2.6940951392632753e-08,
+      "loss": 0.6078,
+      "step": 15123
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4419683704227177,
+      "learning_rate": 2.6833606223351515e-08,
+      "loss": 0.8086,
+      "step": 15124
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0504559761972045,
+      "learning_rate": 2.6726474762105992e-08,
+      "loss": 0.684,
+      "step": 15125
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4505486986211746,
+      "learning_rate": 2.661955701349972e-08,
+      "loss": 0.6366,
+      "step": 15126
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0940504972448424,
+      "learning_rate": 2.6512852982127357e-08,
+      "loss": 0.6412,
+      "step": 15127
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.197872996265428,
+      "learning_rate": 2.6406362672574125e-08,
+      "loss": 0.6393,
+      "step": 15128
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3058067106092879,
+      "learning_rate": 2.6300086089416366e-08,
+      "loss": 0.675,
+      "step": 15129
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.424626377443051,
+      "learning_rate": 2.6194023237221533e-08,
+      "loss": 0.6295,
+      "step": 15130
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3369569966307597,
+      "learning_rate": 2.6088174120545985e-08,
+      "loss": 0.5998,
+      "step": 15131
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.524501976073389,
+      "learning_rate": 2.5982538743939413e-08,
+      "loss": 0.6496,
+      "step": 15132
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.811005735501903,
+      "learning_rate": 2.5877117111940964e-08,
+      "loss": 0.6635,
+      "step": 15133
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2684860012186512,
+      "learning_rate": 2.577190922908035e-08,
+      "loss": 0.5856,
+      "step": 15134
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2843417141454325,
+      "learning_rate": 2.5666915099879508e-08,
+      "loss": 0.6525,
+      "step": 15135
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1810578327912808,
+      "learning_rate": 2.556213472884983e-08,
+      "loss": 0.6553,
+      "step": 15136
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2755865918078557,
+      "learning_rate": 2.5457568120493825e-08,
+      "loss": 0.6925,
+      "step": 15137
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4973204955009873,
+      "learning_rate": 2.5353215279304566e-08,
+      "loss": 0.8091,
+      "step": 15138
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5633648378484621,
+      "learning_rate": 2.5249076209767353e-08,
+      "loss": 0.7109,
+      "step": 15139
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5072034025572525,
+      "learning_rate": 2.5145150916356386e-08,
+      "loss": 0.5663,
+      "step": 15140
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3471505932105543,
+      "learning_rate": 2.5041439403537537e-08,
+      "loss": 0.6154,
+      "step": 15141
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4014048179966072,
+      "learning_rate": 2.493794167576835e-08,
+      "loss": 0.6849,
+      "step": 15142
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3352795670795388,
+      "learning_rate": 2.4834657737495828e-08,
+      "loss": 0.6043,
+      "step": 15143
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9762585614303028,
+      "learning_rate": 2.473158759315808e-08,
+      "loss": 0.6639,
+      "step": 15144
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3949704974541215,
+      "learning_rate": 2.4628731247184902e-08,
+      "loss": 0.5677,
+      "step": 15145
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.40201609416387,
+      "learning_rate": 2.452608870399553e-08,
+      "loss": 0.6976,
+      "step": 15146
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4081782092007122,
+      "learning_rate": 2.4423659968001446e-08,
+      "loss": 0.601,
+      "step": 15147
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0876899119932193,
+      "learning_rate": 2.4321445043603565e-08,
+      "loss": 0.8407,
+      "step": 15148
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2157999170968259,
+      "learning_rate": 2.421944393519504e-08,
+      "loss": 0.5875,
+      "step": 15149
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.439590328952095,
+      "learning_rate": 2.411765664715793e-08,
+      "loss": 0.6589,
+      "step": 15150
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1542916783122155,
+      "learning_rate": 2.401608318386761e-08,
+      "loss": 0.6723,
+      "step": 15151
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4301732004699126,
+      "learning_rate": 2.3914723549687825e-08,
+      "loss": 0.5851,
+      "step": 15152
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5735745590840113,
+      "learning_rate": 2.381357774897508e-08,
+      "loss": 0.5932,
+      "step": 15153
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5479734497351127,
+      "learning_rate": 2.3712645786075905e-08,
+      "loss": 0.6058,
+      "step": 15154
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2283477157719203,
+      "learning_rate": 2.3611927665326275e-08,
+      "loss": 0.5695,
+      "step": 15155
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.646860458819075,
+      "learning_rate": 2.3511423391055498e-08,
+      "loss": 0.6286,
+      "step": 15156
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5701695111286427,
+      "learning_rate": 2.341113296758235e-08,
+      "loss": 0.7482,
+      "step": 15157
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3360433598953583,
+      "learning_rate": 2.3311056399216158e-08,
+      "loss": 0.6819,
+      "step": 15158
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.266185703325197,
+      "learning_rate": 2.3211193690257373e-08,
+      "loss": 0.5838,
+      "step": 15159
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2745869950020972,
+      "learning_rate": 2.3111544844997003e-08,
+      "loss": 0.5857,
+      "step": 15160
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4038418162574726,
+      "learning_rate": 2.3012109867718845e-08,
+      "loss": 0.6036,
+      "step": 15161
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.215732739124014,
+      "learning_rate": 2.291288876269393e-08,
+      "loss": 0.498,
+      "step": 15162
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2306661787981636,
+      "learning_rate": 2.2813881534186622e-08,
+      "loss": 0.562,
+      "step": 15163
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.59901591735826,
+      "learning_rate": 2.271508818645185e-08,
+      "loss": 0.6247,
+      "step": 15164
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.231562056612065,
+      "learning_rate": 2.2616508723734555e-08,
+      "loss": 0.612,
+      "step": 15165
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.412388110019993,
+      "learning_rate": 2.2518143150271344e-08,
+      "loss": 0.6965,
+      "step": 15166
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2024948610198105,
+      "learning_rate": 2.2419991470289393e-08,
+      "loss": 0.7389,
+      "step": 15167
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5691932449401627,
+      "learning_rate": 2.232205368800533e-08,
+      "loss": 0.6568,
+      "step": 15168
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4072529287713922,
+      "learning_rate": 2.222432980762912e-08,
+      "loss": 0.6722,
+      "step": 15169
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8615481033062722,
+      "learning_rate": 2.2126819833358516e-08,
+      "loss": 0.6514,
+      "step": 15170
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5540838565818351,
+      "learning_rate": 2.2029523769385718e-08,
+      "loss": 0.5899,
+      "step": 15171
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5094136789398918,
+      "learning_rate": 2.1932441619890722e-08,
+      "loss": 0.6183,
+      "step": 15172
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2116748906859334,
+      "learning_rate": 2.1835573389045185e-08,
+      "loss": 0.4924,
+      "step": 15173
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.501467413187285,
+      "learning_rate": 2.1738919081012446e-08,
+      "loss": 0.54,
+      "step": 15174
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2635783687780513,
+      "learning_rate": 2.1642478699944736e-08,
+      "loss": 0.6177,
+      "step": 15175
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.435156963950307,
+      "learning_rate": 2.1546252249988186e-08,
+      "loss": 0.6455,
+      "step": 15176
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.151442488250847,
+      "learning_rate": 2.1450239735276158e-08,
+      "loss": 0.5504,
+      "step": 15177
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.47670132298952,
+      "learning_rate": 2.135444115993479e-08,
+      "loss": 0.6614,
+      "step": 15178
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3301488367457697,
+      "learning_rate": 2.1258856528081906e-08,
+      "loss": 0.64,
+      "step": 15179
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4787493209012124,
+      "learning_rate": 2.116348584382366e-08,
+      "loss": 0.598,
+      "step": 15180
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3662637247234797,
+      "learning_rate": 2.1068329111259e-08,
+      "loss": 0.67,
+      "step": 15181
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4771911418835266,
+      "learning_rate": 2.097338633447743e-08,
+      "loss": 0.6178,
+      "step": 15182
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3456581674840558,
+      "learning_rate": 2.087865751755791e-08,
+      "loss": 0.6673,
+      "step": 15183
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4165421421370097,
+      "learning_rate": 2.0784142664571626e-08,
+      "loss": 0.6446,
+      "step": 15184
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4322014488964743,
+      "learning_rate": 2.0689841779579777e-08,
+      "loss": 0.652,
+      "step": 15185
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6520694215714624,
+      "learning_rate": 2.059575486663523e-08,
+      "loss": 0.6512,
+      "step": 15186
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.313942157628635,
+      "learning_rate": 2.0501881929780865e-08,
+      "loss": 0.6965,
+      "step": 15187
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2641997203057918,
+      "learning_rate": 2.040822297305012e-08,
+      "loss": 0.6445,
+      "step": 15188
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3567541739221538,
+      "learning_rate": 2.031477800046866e-08,
+      "loss": 0.6631,
+      "step": 15189
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6387913881130214,
+      "learning_rate": 2.0221547016051614e-08,
+      "loss": 0.6629,
+      "step": 15190
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4445973967835932,
+      "learning_rate": 2.012853002380466e-08,
+      "loss": 0.6604,
+      "step": 15191
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4940732403423893,
+      "learning_rate": 2.003572702772627e-08,
+      "loss": 0.7014,
+      "step": 15192
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4193105465366675,
+      "learning_rate": 1.9943138031803255e-08,
+      "loss": 0.6461,
+      "step": 15193
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4100020216934672,
+      "learning_rate": 1.9850763040014654e-08,
+      "loss": 0.7108,
+      "step": 15194
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3786439347929418,
+      "learning_rate": 1.9758602056330623e-08,
+      "loss": 0.691,
+      "step": 15195
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5984722859613054,
+      "learning_rate": 1.9666655084710772e-08,
+      "loss": 0.7927,
+      "step": 15196
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2533130765591423,
+      "learning_rate": 1.957492212910639e-08,
+      "loss": 0.5675,
+      "step": 15197
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3504481825844472,
+      "learning_rate": 1.9483403193459873e-08,
+      "loss": 0.6304,
+      "step": 15198
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2730550768108815,
+      "learning_rate": 1.939209828170363e-08,
+      "loss": 0.6061,
+      "step": 15199
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5538943140657662,
+      "learning_rate": 1.9301007397761197e-08,
+      "loss": 0.665,
+      "step": 15200
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3063987341246521,
+      "learning_rate": 1.9210130545547213e-08,
+      "loss": 0.5933,
+      "step": 15201
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2625894575292143,
+      "learning_rate": 1.9119467728966888e-08,
+      "loss": 0.6688,
+      "step": 15202
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4228152450192617,
+      "learning_rate": 1.9029018951915446e-08,
+      "loss": 0.6453,
+      "step": 15203
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8505443310331295,
+      "learning_rate": 1.8938784218281435e-08,
+      "loss": 0.6571,
+      "step": 15204
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3837343176315353,
+      "learning_rate": 1.884876353194065e-08,
+      "loss": 0.6741,
+      "step": 15205
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5434159562310885,
+      "learning_rate": 1.875895689676166e-08,
+      "loss": 0.6197,
+      "step": 15206
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1994572987459744,
+      "learning_rate": 1.866936431660471e-08,
+      "loss": 0.6055,
+      "step": 15207
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.153340063253554,
+      "learning_rate": 1.8579985795318946e-08,
+      "loss": 0.6016,
+      "step": 15208
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4374959769023485,
+      "learning_rate": 1.849082133674518e-08,
+      "loss": 0.6231,
+      "step": 15209
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2860236981162638,
+      "learning_rate": 1.8401870944715903e-08,
+      "loss": 0.6034,
+      "step": 15210
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2462616369372466,
+      "learning_rate": 1.8313134623051955e-08,
+      "loss": 0.68,
+      "step": 15211
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3698852799522359,
+      "learning_rate": 1.822461237556805e-08,
+      "loss": 0.521,
+      "step": 15212
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8946265784747545,
+      "learning_rate": 1.813630420606727e-08,
+      "loss": 0.607,
+      "step": 15213
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2321429901748675,
+      "learning_rate": 1.80482101183449e-08,
+      "loss": 0.6399,
+      "step": 15214
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.71884060835746,
+      "learning_rate": 1.7960330116186807e-08,
+      "loss": 0.6432,
+      "step": 15215
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.618207439917371,
+      "learning_rate": 1.7872664203368305e-08,
+      "loss": 0.5867,
+      "step": 15216
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3454834003464602,
+      "learning_rate": 1.7785212383656934e-08,
+      "loss": 0.5404,
+      "step": 15217
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3315389402709397,
+      "learning_rate": 1.7697974660811357e-08,
+      "loss": 0.697,
+      "step": 15218
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3481762080822248,
+      "learning_rate": 1.761095103858024e-08,
+      "loss": 0.6023,
+      "step": 15219
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0568659370513875,
+      "learning_rate": 1.7524141520702808e-08,
+      "loss": 0.6165,
+      "step": 15220
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3915656766143079,
+      "learning_rate": 1.743754611090942e-08,
+      "loss": 0.5386,
+      "step": 15221
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2586673362880842,
+      "learning_rate": 1.7351164812921538e-08,
+      "loss": 0.6059,
+      "step": 15222
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3848087932677646,
+      "learning_rate": 1.726499763045064e-08,
+      "loss": 0.6202,
+      "step": 15223
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5288159555859289,
+      "learning_rate": 1.717904456720043e-08,
+      "loss": 0.62,
+      "step": 15224
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.460339087134942,
+      "learning_rate": 1.7093305626864065e-08,
+      "loss": 0.7168,
+      "step": 15225
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2063706156820138,
+      "learning_rate": 1.7007780813125817e-08,
+      "loss": 0.5329,
+      "step": 15226
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2855723268378938,
+      "learning_rate": 1.692247012966053e-08,
+      "loss": 0.5429,
+      "step": 15227
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.486448947801812,
+      "learning_rate": 1.6837373580135265e-08,
+      "loss": 0.6614,
+      "step": 15228
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1150323889978715,
+      "learning_rate": 1.675249116820543e-08,
+      "loss": 0.6195,
+      "step": 15229
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.521027209585193,
+      "learning_rate": 1.666782289751978e-08,
+      "loss": 0.617,
+      "step": 15230
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2850028576878891,
+      "learning_rate": 1.6583368771716513e-08,
+      "loss": 0.7027,
+      "step": 15231
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4267055651220437,
+      "learning_rate": 1.6499128794423836e-08,
+      "loss": 0.7859,
+      "step": 15232
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2897507979310543,
+      "learning_rate": 1.6415102969263297e-08,
+      "loss": 0.5295,
+      "step": 15233
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5953600752743218,
+      "learning_rate": 1.6331291299844233e-08,
+      "loss": 0.6719,
+      "step": 15234
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1399546952160435,
+      "learning_rate": 1.6247693789768758e-08,
+      "loss": 0.5586,
+      "step": 15235
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3845994357464522,
+      "learning_rate": 1.616431044262956e-08,
+      "loss": 0.6446,
+      "step": 15236
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2616720740752456,
+      "learning_rate": 1.6081141262009326e-08,
+      "loss": 0.5501,
+      "step": 15237
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9907236289950381,
+      "learning_rate": 1.5998186251482416e-08,
+      "loss": 0.6617,
+      "step": 15238
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2362281200620742,
+      "learning_rate": 1.5915445414613208e-08,
+      "loss": 0.5109,
+      "step": 15239
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4078482994545638,
+      "learning_rate": 1.5832918754958293e-08,
+      "loss": 0.6281,
+      "step": 15240
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.438397807206498,
+      "learning_rate": 1.5750606276062063e-08,
+      "loss": 0.7071,
+      "step": 15241
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4805550048896485,
+      "learning_rate": 1.5668507981463354e-08,
+      "loss": 0.8305,
+      "step": 15242
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2463647168483358,
+      "learning_rate": 1.5586623874689345e-08,
+      "loss": 0.5303,
+      "step": 15243
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.221767193200098,
+      "learning_rate": 1.550495395925944e-08,
+      "loss": 0.6631,
+      "step": 15244
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5115714912707765,
+      "learning_rate": 1.54234982386825e-08,
+      "loss": 0.6689,
+      "step": 15245
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2090121789218493,
+      "learning_rate": 1.5342256716459058e-08,
+      "loss": 0.556,
+      "step": 15246
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3846478031333902,
+      "learning_rate": 1.5261229396079657e-08,
+      "loss": 0.5654,
+      "step": 15247
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9215855971695954,
+      "learning_rate": 1.518041628102762e-08,
+      "loss": 0.5426,
+      "step": 15248
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.277403010313088,
+      "learning_rate": 1.5099817374774615e-08,
+      "loss": 0.6076,
+      "step": 15249
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4711812812542144,
+      "learning_rate": 1.501943268078454e-08,
+      "loss": 0.6441,
+      "step": 15250
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4329076083850636,
+      "learning_rate": 1.4939262202511297e-08,
+      "loss": 0.5845,
+      "step": 15251
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4089764498324595,
+      "learning_rate": 1.4859305943400459e-08,
+      "loss": 0.6338,
+      "step": 15252
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3315932849023244,
+      "learning_rate": 1.4779563906888172e-08,
+      "loss": 0.5598,
+      "step": 15253
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4353053059640337,
+      "learning_rate": 1.4700036096400028e-08,
+      "loss": 0.6529,
+      "step": 15254
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3103627565594116,
+      "learning_rate": 1.4620722515354957e-08,
+      "loss": 0.5262,
+      "step": 15255
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.448573827718736,
+      "learning_rate": 1.454162316715968e-08,
+      "loss": 0.7144,
+      "step": 15256
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3038642141279464,
+      "learning_rate": 1.4462738055214254e-08,
+      "loss": 0.6057,
+      "step": 15257
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5717111455476551,
+      "learning_rate": 1.4384067182908746e-08,
+      "loss": 0.7073,
+      "step": 15258
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5008073427432411,
+      "learning_rate": 1.4305610553623228e-08,
+      "loss": 0.6619,
+      "step": 15259
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.738143903742245,
+      "learning_rate": 1.4227368170728894e-08,
+      "loss": 0.5732,
+      "step": 15260
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4431934662304737,
+      "learning_rate": 1.4149340037589166e-08,
+      "loss": 0.6214,
+      "step": 15261
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4727235730029145,
+      "learning_rate": 1.4071526157556358e-08,
+      "loss": 0.6915,
+      "step": 15262
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2960206557012652,
+      "learning_rate": 1.3993926533973912e-08,
+      "loss": 0.542,
+      "step": 15263
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0658667545341727,
+      "learning_rate": 1.3916541170176934e-08,
+      "loss": 0.5605,
+      "step": 15264
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5028786679457455,
+      "learning_rate": 1.3839370069490543e-08,
+      "loss": 0.6821,
+      "step": 15265
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3611511844514022,
+      "learning_rate": 1.3762413235232086e-08,
+      "loss": 0.5515,
+      "step": 15266
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4766782946050525,
+      "learning_rate": 1.3685670670706697e-08,
+      "loss": 0.6991,
+      "step": 15267
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.440893937516582,
+      "learning_rate": 1.3609142379213957e-08,
+      "loss": 0.7045,
+      "step": 15268
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.491154562211241,
+      "learning_rate": 1.3532828364041239e-08,
+      "loss": 0.69,
+      "step": 15269
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.411131407389213,
+      "learning_rate": 1.3456728628468141e-08,
+      "loss": 0.6406,
+      "step": 15270
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3196385795844878,
+      "learning_rate": 1.3380843175765379e-08,
+      "loss": 0.6138,
+      "step": 15271
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2422664590159607,
+      "learning_rate": 1.3305172009193678e-08,
+      "loss": 0.6282,
+      "step": 15272
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3114971084238602,
+      "learning_rate": 1.3229715132004328e-08,
+      "loss": 0.6105,
+      "step": 15273
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4772805830092166,
+      "learning_rate": 1.3154472547440289e-08,
+      "loss": 0.6822,
+      "step": 15274
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.952186050327038,
+      "learning_rate": 1.3079444258735086e-08,
+      "loss": 0.6553,
+      "step": 15275
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4938686308774614,
+      "learning_rate": 1.3004630269112806e-08,
+      "loss": 0.6922,
+      "step": 15276
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5385752107757262,
+      "learning_rate": 1.29300305817881e-08,
+      "loss": 0.7841,
+      "step": 15277
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4752707816572954,
+      "learning_rate": 1.2855645199966738e-08,
+      "loss": 0.6452,
+      "step": 15278
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5571173640810212,
+      "learning_rate": 1.2781474126845051e-08,
+      "loss": 0.6415,
+      "step": 15279
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.459239085502038,
+      "learning_rate": 1.270751736561049e-08,
+      "loss": 0.6631,
+      "step": 15280
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4712004405316188,
+      "learning_rate": 1.2633774919441622e-08,
+      "loss": 0.7318,
+      "step": 15281
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.437266568624729,
+      "learning_rate": 1.256024679150647e-08,
+      "loss": 0.5884,
+      "step": 15282
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3952256660498326,
+      "learning_rate": 1.2486932984965282e-08,
+      "loss": 0.6145,
+      "step": 15283
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3297110913028032,
+      "learning_rate": 1.241383350296832e-08,
+      "loss": 0.5298,
+      "step": 15284
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.495321391638634,
+      "learning_rate": 1.2340948348656955e-08,
+      "loss": 0.7889,
+      "step": 15285
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0714416942515885,
+      "learning_rate": 1.2268277525163685e-08,
+      "loss": 0.6263,
+      "step": 15286
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2443510747180584,
+      "learning_rate": 1.21958210356099e-08,
+      "loss": 0.5917,
+      "step": 15287
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4882564824861158,
+      "learning_rate": 1.2123578883110887e-08,
+      "loss": 0.6769,
+      "step": 15288
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1749615003671003,
+      "learning_rate": 1.2051551070769719e-08,
+      "loss": 0.5924,
+      "step": 15289
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2525597128856927,
+      "learning_rate": 1.1979737601682252e-08,
+      "loss": 0.6082,
+      "step": 15290
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2673938597442873,
+      "learning_rate": 1.1908138478934905e-08,
+      "loss": 0.6032,
+      "step": 15291
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.298318041137735,
+      "learning_rate": 1.1836753705602998e-08,
+      "loss": 0.7341,
+      "step": 15292
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3291050674421392,
+      "learning_rate": 1.1765583284755743e-08,
+      "loss": 0.6728,
+      "step": 15293
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2746278191775402,
+      "learning_rate": 1.1694627219450694e-08,
+      "loss": 0.596,
+      "step": 15294
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3668952060466373,
+      "learning_rate": 1.1623885512737076e-08,
+      "loss": 0.5758,
+      "step": 15295
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7154710574575347,
+      "learning_rate": 1.1553358167654683e-08,
+      "loss": 0.6374,
+      "step": 15296
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.426348620456702,
+      "learning_rate": 1.1483045187233866e-08,
+      "loss": 0.7011,
+      "step": 15297
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2894252517935714,
+      "learning_rate": 1.1412946574497208e-08,
+      "loss": 0.6046,
+      "step": 15298
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4676892418086833,
+      "learning_rate": 1.134306233245619e-08,
+      "loss": 0.6307,
+      "step": 15299
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3929955717617237,
+      "learning_rate": 1.1273392464114518e-08,
+      "loss": 0.5907,
+      "step": 15300
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4246924028004198,
+      "learning_rate": 1.1203936972465356e-08,
+      "loss": 0.5993,
+      "step": 15301
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6122542973652871,
+      "learning_rate": 1.1134695860493539e-08,
+      "loss": 0.6297,
+      "step": 15302
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.459338902548803,
+      "learning_rate": 1.1065669131175016e-08,
+      "loss": 0.588,
+      "step": 15303
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2717052758463747,
+      "learning_rate": 1.0996856787475197e-08,
+      "loss": 0.5878,
+      "step": 15304
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6168439809795685,
+      "learning_rate": 1.0928258832351712e-08,
+      "loss": 0.6622,
+      "step": 15305
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3589824061544815,
+      "learning_rate": 1.0859875268752206e-08,
+      "loss": 0.6779,
+      "step": 15306
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.426037007134735,
+      "learning_rate": 1.079170609961544e-08,
+      "loss": 0.681,
+      "step": 15307
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4152885409079081,
+      "learning_rate": 1.072375132787018e-08,
+      "loss": 0.5528,
+      "step": 15308
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3933606863671646,
+      "learning_rate": 1.0656010956437979e-08,
+      "loss": 0.5433,
+      "step": 15309
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8990854277627456,
+      "learning_rate": 1.0588484988228175e-08,
+      "loss": 0.68,
+      "step": 15310
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3885358462900712,
+      "learning_rate": 1.0521173426142894e-08,
+      "loss": 0.5022,
+      "step": 15311
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.845452131810857,
+      "learning_rate": 1.0454076273075375e-08,
+      "loss": 0.6441,
+      "step": 15312
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5555738399811196,
+      "learning_rate": 1.0387193531908313e-08,
+      "loss": 0.6389,
+      "step": 15313
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.593829391020196,
+      "learning_rate": 1.0320525205516629e-08,
+      "loss": 0.7067,
+      "step": 15314
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2564310100111882,
+      "learning_rate": 1.025407129676359e-08,
+      "loss": 0.6457,
+      "step": 15315
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3247959923483397,
+      "learning_rate": 1.018783180850691e-08,
+      "loss": 0.5834,
+      "step": 15316
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4061950369479648,
+      "learning_rate": 1.0121806743591534e-08,
+      "loss": 0.5542,
+      "step": 15317
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3448896008369862,
+      "learning_rate": 1.0055996104854637e-08,
+      "loss": 0.5589,
+      "step": 15318
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8241630453169562,
+      "learning_rate": 9.990399895125624e-09,
+      "loss": 0.7188,
+      "step": 15319
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3611481786088837,
+      "learning_rate": 9.925018117221686e-09,
+      "loss": 0.6207,
+      "step": 15320
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.698357801508171,
+      "learning_rate": 9.859850773953905e-09,
+      "loss": 0.7146,
+      "step": 15321
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4419903884354546,
+      "learning_rate": 9.794897868121711e-09,
+      "loss": 0.6173,
+      "step": 15322
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.265096134100969,
+      "learning_rate": 9.73015940251676e-09,
+      "loss": 0.7545,
+      "step": 15323
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.42082128554563,
+      "learning_rate": 9.665635379920157e-09,
+      "loss": 0.6755,
+      "step": 15324
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5466423327495238,
+      "learning_rate": 9.6013258031058e-09,
+      "loss": 0.5376,
+      "step": 15325
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2535451847523953,
+      "learning_rate": 9.537230674836473e-09,
+      "loss": 0.6233,
+      "step": 15326
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.325298587154244,
+      "learning_rate": 9.473349997867198e-09,
+      "loss": 0.5779,
+      "step": 15327
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2175656507066823,
+      "learning_rate": 9.409683774941891e-09,
+      "loss": 0.5836,
+      "step": 15328
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4490589093233763,
+      "learning_rate": 9.346232008797252e-09,
+      "loss": 0.6069,
+      "step": 15329
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7179557236544565,
+      "learning_rate": 9.282994702159986e-09,
+      "loss": 0.5771,
+      "step": 15330
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5389343703614486,
+      "learning_rate": 9.219971857747368e-09,
+      "loss": 0.5424,
+      "step": 15331
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6461593679578308,
+      "learning_rate": 9.157163478267783e-09,
+      "loss": 0.6778,
+      "step": 15332
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9210560778612651,
+      "learning_rate": 9.094569566419632e-09,
+      "loss": 0.5878,
+      "step": 15333
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8437582018266545,
+      "learning_rate": 9.032190124893536e-09,
+      "loss": 0.5743,
+      "step": 15334
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3184176340416505,
+      "learning_rate": 8.970025156369578e-09,
+      "loss": 0.6082,
+      "step": 15335
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.745185593263662,
+      "learning_rate": 8.908074663519505e-09,
+      "loss": 0.745,
+      "step": 15336
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5610731285046586,
+      "learning_rate": 8.846338649005082e-09,
+      "loss": 0.6263,
+      "step": 15337
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3367085841093937,
+      "learning_rate": 8.784817115479183e-09,
+      "loss": 0.5926,
+      "step": 15338
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2603344977315003,
+      "learning_rate": 8.723510065585806e-09,
+      "loss": 0.5716,
+      "step": 15339
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3931310715139982,
+      "learning_rate": 8.662417501959507e-09,
+      "loss": 0.6281,
+      "step": 15340
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3495346918784377,
+      "learning_rate": 8.601539427225414e-09,
+      "loss": 0.5845,
+      "step": 15341
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2338902119605717,
+      "learning_rate": 8.540875843999208e-09,
+      "loss": 0.5772,
+      "step": 15342
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6453438418938604,
+      "learning_rate": 8.480426754888805e-09,
+      "loss": 0.5753,
+      "step": 15343
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4272502114328447,
+      "learning_rate": 8.42019216249046e-09,
+      "loss": 0.6198,
+      "step": 15344
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4625092172229166,
+      "learning_rate": 8.360172069393768e-09,
+      "loss": 0.5572,
+      "step": 15345
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0621461311652125,
+      "learning_rate": 8.300366478177779e-09,
+      "loss": 0.684,
+      "step": 15346
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0259273282010726,
+      "learning_rate": 8.240775391411548e-09,
+      "loss": 0.559,
+      "step": 15347
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.264942471409948,
+      "learning_rate": 8.18139881165636e-09,
+      "loss": 0.6464,
+      "step": 15348
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2515458172017921,
+      "learning_rate": 8.122236741464618e-09,
+      "loss": 0.6215,
+      "step": 15349
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0439695216591858,
+      "learning_rate": 8.063289183377065e-09,
+      "loss": 0.6882,
+      "step": 15350
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4913170084163059,
+      "learning_rate": 8.004556139927788e-09,
+      "loss": 0.6112,
+      "step": 15351
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4399060888497404,
+      "learning_rate": 7.946037613640878e-09,
+      "loss": 0.615,
+      "step": 15352
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.651470143045369,
+      "learning_rate": 7.887733607030435e-09,
+      "loss": 0.6345,
+      "step": 15353
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6515206311933999,
+      "learning_rate": 7.82964412260223e-09,
+      "loss": 0.7065,
+      "step": 15354
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4474641975674682,
+      "learning_rate": 7.77176916285205e-09,
+      "loss": 0.7051,
+      "step": 15355
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.281253663689237,
+      "learning_rate": 7.714108730267345e-09,
+      "loss": 0.7114,
+      "step": 15356
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.273090192802488,
+      "learning_rate": 7.656662827326134e-09,
+      "loss": 0.5691,
+      "step": 15357
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4406288247215207,
+      "learning_rate": 7.599431456495888e-09,
+      "loss": 0.7493,
+      "step": 15358
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3096851998238832,
+      "learning_rate": 7.542414620237414e-09,
+      "loss": 0.544,
+      "step": 15359
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2832035258743706,
+      "learning_rate": 7.485612321000424e-09,
+      "loss": 0.6035,
+      "step": 15360
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4138045194190092,
+      "learning_rate": 7.4290245612246294e-09,
+      "loss": 0.7483,
+      "step": 15361
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5128149493002263,
+      "learning_rate": 7.372651343343639e-09,
+      "loss": 0.6756,
+      "step": 15362
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.265639365863903,
+      "learning_rate": 7.316492669778297e-09,
+      "loss": 0.5888,
+      "step": 15363
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6982093937920688,
+      "learning_rate": 7.260548542943335e-09,
+      "loss": 0.6152,
+      "step": 15364
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.175516339146047,
+      "learning_rate": 7.2048189652412784e-09,
+      "loss": 0.5497,
+      "step": 15365
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.217949036580795,
+      "learning_rate": 7.149303939067986e-09,
+      "loss": 0.5795,
+      "step": 15366
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3983991867807901,
+      "learning_rate": 7.094003466808774e-09,
+      "loss": 0.6994,
+      "step": 15367
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6092554070231468,
+      "learning_rate": 7.038917550840074e-09,
+      "loss": 0.6865,
+      "step": 15368
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6665641376592666,
+      "learning_rate": 6.984046193528881e-09,
+      "loss": 0.5931,
+      "step": 15369
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7588787529075878,
+      "learning_rate": 6.9293893972338654e-09,
+      "loss": 0.6633,
+      "step": 15370
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0418463139899914,
+      "learning_rate": 6.874947164302592e-09,
+      "loss": 0.6432,
+      "step": 15371
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4886132969273695,
+      "learning_rate": 6.820719497074857e-09,
+      "loss": 0.6492,
+      "step": 15372
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7016675207782173,
+      "learning_rate": 6.7667063978815724e-09,
+      "loss": 0.5995,
+      "step": 15373
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0832792507972644,
+      "learning_rate": 6.712907869043661e-09,
+      "loss": 0.5792,
+      "step": 15374
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4421158230545923,
+      "learning_rate": 6.659323912872606e-09,
+      "loss": 0.6519,
+      "step": 15375
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1850432617427737,
+      "learning_rate": 6.60595453167101e-09,
+      "loss": 0.5485,
+      "step": 15376
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3550746316847975,
+      "learning_rate": 6.552799727732595e-09,
+      "loss": 0.6795,
+      "step": 15377
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4344295879613569,
+      "learning_rate": 6.4998595033410886e-09,
+      "loss": 0.6617,
+      "step": 15378
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5733490634654768,
+      "learning_rate": 6.447133860771893e-09,
+      "loss": 0.6418,
+      "step": 15379
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8034981099602585,
+      "learning_rate": 6.394622802290973e-09,
+      "loss": 0.6391,
+      "step": 15380
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4161025871305482,
+      "learning_rate": 6.342326330153747e-09,
+      "loss": 0.5592,
+      "step": 15381
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1077416908186675,
+      "learning_rate": 6.290244446608973e-09,
+      "loss": 0.5774,
+      "step": 15382
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0074874846882365,
+      "learning_rate": 6.238377153893749e-09,
+      "loss": 0.5674,
+      "step": 15383
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.426867058427212,
+      "learning_rate": 6.186724454236847e-09,
+      "loss": 0.677,
+      "step": 15384
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3364190928375435,
+      "learning_rate": 6.13528634985816e-09,
+      "loss": 0.6058,
+      "step": 15385
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3083174030136382,
+      "learning_rate": 6.084062842968696e-09,
+      "loss": 0.5922,
+      "step": 15386
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2201677711978982,
+      "learning_rate": 6.0330539357689175e-09,
+      "loss": 0.5954,
+      "step": 15387
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2585389633225823,
+      "learning_rate": 5.98225963045096e-09,
+      "loss": 0.6259,
+      "step": 15388
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4362792735779497,
+      "learning_rate": 5.9316799291969654e-09,
+      "loss": 0.7294,
+      "step": 15389
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1481390127160411,
+      "learning_rate": 5.881314834181862e-09,
+      "loss": 0.6504,
+      "step": 15390
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4739111619209828,
+      "learning_rate": 5.831164347568918e-09,
+      "loss": 0.6316,
+      "step": 15391
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5957596434907066,
+      "learning_rate": 5.781228471513633e-09,
+      "loss": 0.7103,
+      "step": 15392
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3460690363033068,
+      "learning_rate": 5.731507208160958e-09,
+      "loss": 0.5675,
+      "step": 15393
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8325281127937385,
+      "learning_rate": 5.682000559649181e-09,
+      "loss": 0.6817,
+      "step": 15394
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1859297365314276,
+      "learning_rate": 5.63270852810438e-09,
+      "loss": 0.6685,
+      "step": 15395
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2900414225314583,
+      "learning_rate": 5.583631115644861e-09,
+      "loss": 0.5947,
+      "step": 15396
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2647935289272643,
+      "learning_rate": 5.534768324380046e-09,
+      "loss": 0.6458,
+      "step": 15397
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5564638267134163,
+      "learning_rate": 5.486120156409924e-09,
+      "loss": 0.5884,
+      "step": 15398
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5785841688197513,
+      "learning_rate": 5.437686613823934e-09,
+      "loss": 0.6148,
+      "step": 15399
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.634018939932586,
+      "learning_rate": 5.389467698704298e-09,
+      "loss": 0.6309,
+      "step": 15400
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.682260308473768,
+      "learning_rate": 5.341463413123249e-09,
+      "loss": 0.5285,
+      "step": 15401
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.589082859251898,
+      "learning_rate": 5.29367375914247e-09,
+      "loss": 0.7438,
+      "step": 15402
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5262839803428934,
+      "learning_rate": 5.246098738816985e-09,
+      "loss": 0.5958,
+      "step": 15403
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.786134301429008,
+      "learning_rate": 5.198738354190158e-09,
+      "loss": 0.6912,
+      "step": 15404
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.044691546039755,
+      "learning_rate": 5.151592607298139e-09,
+      "loss": 0.6341,
+      "step": 15405
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5830304227569727,
+      "learning_rate": 5.104661500165975e-09,
+      "loss": 0.6961,
+      "step": 15406
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1366555986789406,
+      "learning_rate": 5.057945034810385e-09,
+      "loss": 0.5723,
+      "step": 15407
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2123046269254552,
+      "learning_rate": 5.0114432132397636e-09,
+      "loss": 0.5945,
+      "step": 15408
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1597009176697197,
+      "learning_rate": 4.9651560374514015e-09,
+      "loss": 0.6503,
+      "step": 15409
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3892355416214381,
+      "learning_rate": 4.919083509434819e-09,
+      "loss": 0.6202,
+      "step": 15410
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2759229078140943,
+      "learning_rate": 4.873225631170098e-09,
+      "loss": 0.5797,
+      "step": 15411
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.459408262329021,
+      "learning_rate": 4.827582404627884e-09,
+      "loss": 0.6792,
+      "step": 15412
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.320323696039065,
+      "learning_rate": 4.782153831768832e-09,
+      "loss": 0.6792,
+      "step": 15413
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5578628517059878,
+      "learning_rate": 4.736939914545824e-09,
+      "loss": 0.6322,
+      "step": 15414
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.468108614744506,
+      "learning_rate": 4.691940654901195e-09,
+      "loss": 0.5773,
+      "step": 15415
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4143607156198632,
+      "learning_rate": 4.647156054769508e-09,
+      "loss": 0.6094,
+      "step": 15416
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1862364309253517,
+      "learning_rate": 4.602586116074781e-09,
+      "loss": 0.5976,
+      "step": 15417
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5619239166724999,
+      "learning_rate": 4.558230840731592e-09,
+      "loss": 0.719,
+      "step": 15418
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6806588717386872,
+      "learning_rate": 4.514090230647305e-09,
+      "loss": 0.5853,
+      "step": 15419
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4952416457733706,
+      "learning_rate": 4.470164287717627e-09,
+      "loss": 0.6997,
+      "step": 15420
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2502604793460557,
+      "learning_rate": 4.4264530138310445e-09,
+      "loss": 0.6405,
+      "step": 15421
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.346095048538475,
+      "learning_rate": 4.382956410865502e-09,
+      "loss": 0.5752,
+      "step": 15422
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 64.257075641149,
+      "learning_rate": 4.339674480690059e-09,
+      "loss": 0.615,
+      "step": 15423
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3920802903104463,
+      "learning_rate": 4.296607225164895e-09,
+      "loss": 0.5808,
+      "step": 15424
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3547082889939077,
+      "learning_rate": 4.253754646140196e-09,
+      "loss": 0.6402,
+      "step": 15425
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3318623142404868,
+      "learning_rate": 4.211116745458377e-09,
+      "loss": 0.5725,
+      "step": 15426
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4067871362346283,
+      "learning_rate": 4.168693524950196e-09,
+      "loss": 0.7135,
+      "step": 15427
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6032877665284415,
+      "learning_rate": 4.1264849864403044e-09,
+      "loss": 0.6007,
+      "step": 15428
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.540830639228882,
+      "learning_rate": 4.084491131741697e-09,
+      "loss": 0.627,
+      "step": 15429
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6112444189480521,
+      "learning_rate": 4.042711962658486e-09,
+      "loss": 0.6823,
+      "step": 15430
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5638337437744565,
+      "learning_rate": 4.001147480987011e-09,
+      "loss": 0.6064,
+      "step": 15431
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.535063395305323,
+      "learning_rate": 3.959797688513067e-09,
+      "loss": 0.5558,
+      "step": 15432
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2880026855998208,
+      "learning_rate": 3.9186625870130115e-09,
+      "loss": 0.5852,
+      "step": 15433
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5398197963252935,
+      "learning_rate": 3.877742178254873e-09,
+      "loss": 0.7221,
+      "step": 15434
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6258245755819964,
+      "learning_rate": 3.837036463997246e-09,
+      "loss": 0.7721,
+      "step": 15435
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2766268439653323,
+      "learning_rate": 3.796545445989286e-09,
+      "loss": 0.6699,
+      "step": 15436
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5202174158365984,
+      "learning_rate": 3.756269125970713e-09,
+      "loss": 0.6973,
+      "step": 15437
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2466041465307525,
+      "learning_rate": 3.7162075056729196e-09,
+      "loss": 0.6454,
+      "step": 15438
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3910038256632706,
+      "learning_rate": 3.6763605868167516e-09,
+      "loss": 0.5853,
+      "step": 15439
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4188656756588534,
+      "learning_rate": 3.636728371114173e-09,
+      "loss": 0.6574,
+      "step": 15440
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3657148317964087,
+      "learning_rate": 3.5973108602699313e-09,
+      "loss": 0.5611,
+      "step": 15441
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4109682801710888,
+      "learning_rate": 3.558108055976006e-09,
+      "loss": 0.5636,
+      "step": 15442
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1014381893661418,
+      "learning_rate": 3.5191199599182712e-09,
+      "loss": 0.5757,
+      "step": 15443
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6033078968754095,
+      "learning_rate": 3.4803465737714983e-09,
+      "loss": 0.711,
+      "step": 15444
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3840727380458344,
+      "learning_rate": 3.4417878992021315e-09,
+      "loss": 0.5742,
+      "step": 15445
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6829880029001125,
+      "learning_rate": 3.403443937867179e-09,
+      "loss": 0.748,
+      "step": 15446
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.271886670493069,
+      "learning_rate": 3.365314691414212e-09,
+      "loss": 0.7866,
+      "step": 15447
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1490517168290517,
+      "learning_rate": 3.3274001614819195e-09,
+      "loss": 0.6519,
+      "step": 15448
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4119826782869835,
+      "learning_rate": 3.289700349698999e-09,
+      "loss": 0.623,
+      "step": 15449
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4452728682549512,
+      "learning_rate": 3.252215257686375e-09,
+      "loss": 0.6429,
+      "step": 15450
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5244835404438397,
+      "learning_rate": 3.2149448870544273e-09,
+      "loss": 0.7824,
+      "step": 15451
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2215574694744835,
+      "learning_rate": 3.1778892394046525e-09,
+      "loss": 0.6115,
+      "step": 15452
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.043786412212102,
+      "learning_rate": 3.1410483163291093e-09,
+      "loss": 0.6273,
+      "step": 15453
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.148212339519425,
+      "learning_rate": 3.104422119411532e-09,
+      "loss": 0.5859,
+      "step": 15454
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4526966879448286,
+      "learning_rate": 3.0680106502256613e-09,
+      "loss": 0.7013,
+      "step": 15455
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1514147921884983,
+      "learning_rate": 3.0318139103363564e-09,
+      "loss": 0.5351,
+      "step": 15456
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6423236091493285,
+      "learning_rate": 2.9958319012984847e-09,
+      "loss": 0.7046,
+      "step": 15457
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2672506257124678,
+      "learning_rate": 2.9600646246591425e-09,
+      "loss": 0.563,
+      "step": 15458
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4777976830728348,
+      "learning_rate": 2.9245120819543226e-09,
+      "loss": 0.5808,
+      "step": 15459
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2460662973106458,
+      "learning_rate": 2.8891742747128025e-09,
+      "loss": 0.6295,
+      "step": 15460
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.372765466814767,
+      "learning_rate": 2.854051204452257e-09,
+      "loss": 0.4888,
+      "step": 15461
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6306324462960082,
+      "learning_rate": 2.8191428726820347e-09,
+      "loss": 0.5266,
+      "step": 15462
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.298475315998152,
+      "learning_rate": 2.7844492809031567e-09,
+      "loss": 0.678,
+      "step": 15463
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9748860947624656,
+      "learning_rate": 2.749970430605542e-09,
+      "loss": 0.6304,
+      "step": 15464
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.001955343562789,
+      "learning_rate": 2.715706323271339e-09,
+      "loss": 0.6279,
+      "step": 15465
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.190629852253006,
+      "learning_rate": 2.681656960372703e-09,
+      "loss": 0.5755,
+      "step": 15466
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3348690930271412,
+      "learning_rate": 2.6478223433734627e-09,
+      "loss": 0.6226,
+      "step": 15467
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4032313064725124,
+      "learning_rate": 2.614202473726346e-09,
+      "loss": 0.6262,
+      "step": 15468
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.423756062666398,
+      "learning_rate": 2.5807973528768626e-09,
+      "loss": 0.7193,
+      "step": 15469
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4428886222388264,
+      "learning_rate": 2.547606982260531e-09,
+      "loss": 0.6702,
+      "step": 15470
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2888874085955113,
+      "learning_rate": 2.5146313633028775e-09,
+      "loss": 0.6492,
+      "step": 15471
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.553865729423203,
+      "learning_rate": 2.481870497422212e-09,
+      "loss": 0.7051,
+      "step": 15472
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4839904549736855,
+      "learning_rate": 2.449324386025187e-09,
+      "loss": 0.7277,
+      "step": 15473
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6496680781922086,
+      "learning_rate": 2.416993030511239e-09,
+      "loss": 0.707,
+      "step": 15474
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2877567212874537,
+      "learning_rate": 2.3848764322687014e-09,
+      "loss": 0.5975,
+      "step": 15475
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.32321314440367,
+      "learning_rate": 2.3529745926786916e-09,
+      "loss": 0.6772,
+      "step": 15476
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.386647962641244,
+      "learning_rate": 2.3212875131117805e-09,
+      "loss": 0.6368,
+      "step": 15477
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.345403438486055,
+      "learning_rate": 2.2898151949296566e-09,
+      "loss": 0.5918,
+      "step": 15478
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4466630180213516,
+      "learning_rate": 2.258557639484571e-09,
+      "loss": 0.6604,
+      "step": 15479
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3127541295212268,
+      "learning_rate": 2.227514848119894e-09,
+      "loss": 0.6648,
+      "step": 15480
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9456885384653635,
+      "learning_rate": 2.196686822169003e-09,
+      "loss": 0.6866,
+      "step": 15481
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4117724439230415,
+      "learning_rate": 2.1660735629580598e-09,
+      "loss": 0.6909,
+      "step": 15482
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1576262305948073,
+      "learning_rate": 2.1356750718010133e-09,
+      "loss": 0.6774,
+      "step": 15483
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2686852000710964,
+      "learning_rate": 2.1054913500051512e-09,
+      "loss": 0.6258,
+      "step": 15484
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1818745448996766,
+      "learning_rate": 2.0755223988672136e-09,
+      "loss": 0.6767,
+      "step": 15485
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8719068748833614,
+      "learning_rate": 2.0457682196750595e-09,
+      "loss": 0.7209,
+      "step": 15486
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8426979326223234,
+      "learning_rate": 2.0162288137071105e-09,
+      "loss": 0.6461,
+      "step": 15487
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3119176283303162,
+      "learning_rate": 1.9869041822329073e-09,
+      "loss": 0.5573,
+      "step": 15488
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4245533980211578,
+      "learning_rate": 1.957794326513107e-09,
+      "loss": 0.6757,
+      "step": 15489
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2916110362053417,
+      "learning_rate": 1.928899247797822e-09,
+      "loss": 0.5956,
+      "step": 15490
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3654104421992859,
+      "learning_rate": 1.9002189473288356e-09,
+      "loss": 0.5953,
+      "step": 15491
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4682960140376777,
+      "learning_rate": 1.8717534263390513e-09,
+      "loss": 0.5445,
+      "step": 15492
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3667174618519853,
+      "learning_rate": 1.8435026860513794e-09,
+      "loss": 0.7032,
+      "step": 15493
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.349397021989839,
+      "learning_rate": 1.8154667276798488e-09,
+      "loss": 0.6748,
+      "step": 15494
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4310174148337154,
+      "learning_rate": 1.7876455524290514e-09,
+      "loss": 0.6654,
+      "step": 15495
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.289736569647144,
+      "learning_rate": 1.7600391614952529e-09,
+      "loss": 0.6276,
+      "step": 15496
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2530571931488002,
+      "learning_rate": 1.7326475560636157e-09,
+      "loss": 0.5689,
+      "step": 15497
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6170719929926938,
+      "learning_rate": 1.7054707373126423e-09,
+      "loss": 0.6259,
+      "step": 15498
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.995762709219735,
+      "learning_rate": 1.6785087064086213e-09,
+      "loss": 0.6345,
+      "step": 15499
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0869412598115455,
+      "learning_rate": 1.6517614645111812e-09,
+      "loss": 0.7053,
+      "step": 15500
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4596338276985976,
+      "learning_rate": 1.6252290127699577e-09,
+      "loss": 0.5301,
+      "step": 15501
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5638849195378592,
+      "learning_rate": 1.5989113523240396e-09,
+      "loss": 0.6661,
+      "step": 15502
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2558929597492754,
+      "learning_rate": 1.5728084843052992e-09,
+      "loss": 0.536,
+      "step": 15503
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4274666592301963,
+      "learning_rate": 1.546920409834507e-09,
+      "loss": 0.5825,
+      "step": 15504
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3507211279490308,
+      "learning_rate": 1.5212471300252163e-09,
+      "loss": 0.5901,
+      "step": 15505
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.700245881061265,
+      "learning_rate": 1.4957886459798787e-09,
+      "loss": 0.6857,
+      "step": 15506
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.230912490548262,
+      "learning_rate": 1.4705449587931742e-09,
+      "loss": 0.5282,
+      "step": 15507
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6495737275692546,
+      "learning_rate": 1.4455160695492354e-09,
+      "loss": 0.589,
+      "step": 15508
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4291307944655687,
+      "learning_rate": 1.4207019793238686e-09,
+      "loss": 0.5666,
+      "step": 15509
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3301369264662832,
+      "learning_rate": 1.3961026891828878e-09,
+      "loss": 0.7062,
+      "step": 15510
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3381685374463839,
+      "learning_rate": 1.3717182001843355e-09,
+      "loss": 0.7006,
+      "step": 15511
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2950751795787012,
+      "learning_rate": 1.347548513375707e-09,
+      "loss": 0.615,
+      "step": 15512
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1667325209247046,
+      "learning_rate": 1.3235936297956165e-09,
+      "loss": 0.705,
+      "step": 15513
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4473520869570637,
+      "learning_rate": 1.299853550472685e-09,
+      "loss": 0.7724,
+      "step": 15514
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3838194168159323,
+      "learning_rate": 1.276328276428318e-09,
+      "loss": 0.5824,
+      "step": 15515
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5105620191286517,
+      "learning_rate": 1.2530178086728183e-09,
+      "loss": 0.6145,
+      "step": 15516
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3226668710141465,
+      "learning_rate": 1.2299221482076074e-09,
+      "loss": 0.6722,
+      "step": 15517
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3754109075264167,
+      "learning_rate": 1.2070412960257793e-09,
+      "loss": 0.6263,
+      "step": 15518
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.622134551309241,
+      "learning_rate": 1.1843752531104368e-09,
+      "loss": 0.678,
+      "step": 15519
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2333804305658724,
+      "learning_rate": 1.1619240204352455e-09,
+      "loss": 0.6412,
+      "step": 15520
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4537784305854013,
+      "learning_rate": 1.1396875989655443e-09,
+      "loss": 0.5888,
+      "step": 15521
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1447854562357092,
+      "learning_rate": 1.1176659896561247e-09,
+      "loss": 0.5684,
+      "step": 15522
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2829046372693305,
+      "learning_rate": 1.0958591934534525e-09,
+      "loss": 0.562,
+      "step": 15523
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3243363702432505,
+      "learning_rate": 1.0742672112951103e-09,
+      "loss": 0.5049,
+      "step": 15524
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4349676077155407,
+      "learning_rate": 1.052890044109245e-09,
+      "loss": 0.6393,
+      "step": 15525
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3851469933835296,
+      "learning_rate": 1.0317276928134557e-09,
+      "loss": 0.6541,
+      "step": 15526
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3787730790420154,
+      "learning_rate": 1.0107801583175702e-09,
+      "loss": 0.6264,
+      "step": 15527
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2296323451223936,
+      "learning_rate": 9.900474415219795e-10,
+      "loss": 0.5697,
+      "step": 15528
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2800605398423783,
+      "learning_rate": 9.695295433170826e-10,
+      "loss": 0.6574,
+      "step": 15529
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4258005941985337,
+      "learning_rate": 9.492264645855065e-10,
+      "loss": 0.6273,
+      "step": 15530
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5456722548873811,
+      "learning_rate": 9.291382061987763e-10,
+      "loss": 0.5772,
+      "step": 15531
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.212651448150818,
+      "learning_rate": 9.092647690206457e-10,
+      "loss": 0.6239,
+      "step": 15532
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3437977658100568,
+      "learning_rate": 8.896061539048762e-10,
+      "loss": 0.6762,
+      "step": 15533
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4320304431933648,
+      "learning_rate": 8.701623616963472e-10,
+      "loss": 0.6779,
+      "step": 15534
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3099053470783506,
+      "learning_rate": 8.509333932310571e-10,
+      "loss": 0.6141,
+      "step": 15535
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.380838811310172,
+      "learning_rate": 8.319192493344563e-10,
+      "loss": 0.6373,
+      "step": 15536
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.495683813671648,
+      "learning_rate": 8.131199308236692e-10,
+      "loss": 0.7086,
+      "step": 15537
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7728181810621289,
+      "learning_rate": 7.945354385074933e-10,
+      "loss": 0.5498,
+      "step": 15538
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.372219691852769,
+      "learning_rate": 7.761657731836236e-10,
+      "loss": 0.564,
+      "step": 15539
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1441436462890742,
+      "learning_rate": 7.580109356419841e-10,
+      "loss": 0.5434,
+      "step": 15540
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3696098491635436,
+      "learning_rate": 7.400709266625062e-10,
+      "loss": 0.6623,
+      "step": 15541
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3111541039309211,
+      "learning_rate": 7.223457470156847e-10,
+      "loss": 0.5572,
+      "step": 15542
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3082004948933048,
+      "learning_rate": 7.04835397464243e-10,
+      "loss": 0.6028,
+      "step": 15543
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.285551036542428,
+      "learning_rate": 6.87539878759802e-10,
+      "loss": 0.638,
+      "step": 15544
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.205062649002289,
+      "learning_rate": 6.704591916462111e-10,
+      "loss": 0.5061,
+      "step": 15545
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.216930144694498,
+      "learning_rate": 6.535933368567726e-10,
+      "loss": 0.4792,
+      "step": 15546
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9624318083870836,
+      "learning_rate": 6.369423151164622e-10,
+      "loss": 0.5795,
+      "step": 15547
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2057774080364536,
+      "learning_rate": 6.205061271408186e-10,
+      "loss": 0.5525,
+      "step": 15548
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.244505792844563,
+      "learning_rate": 6.042847736364987e-10,
+      "loss": 0.5653,
+      "step": 15549
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.320898319139121,
+      "learning_rate": 5.882782553001676e-10,
+      "loss": 0.6114,
+      "step": 15550
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4971952537073752,
+      "learning_rate": 5.724865728201634e-10,
+      "loss": 0.5948,
+      "step": 15551
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4053391478129467,
+      "learning_rate": 5.569097268742773e-10,
+      "loss": 0.6119,
+      "step": 15552
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.451376369474033,
+      "learning_rate": 5.415477181325291e-10,
+      "loss": 0.6328,
+      "step": 15553
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5324106069627226,
+      "learning_rate": 5.264005472549461e-10,
+      "loss": 0.6051,
+      "step": 15554
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3502507643687012,
+      "learning_rate": 5.114682148921191e-10,
+      "loss": 0.7467,
+      "step": 15555
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5185967390879758,
+      "learning_rate": 4.967507216863121e-10,
+      "loss": 0.619,
+      "step": 15556
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0783915614482362,
+      "learning_rate": 4.822480682692421e-10,
+      "loss": 0.661,
+      "step": 15557
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2518009128744059,
+      "learning_rate": 4.679602552648543e-10,
+      "loss": 0.5884,
+      "step": 15558
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4932447516057774,
+      "learning_rate": 4.538872832865471e-10,
+      "loss": 0.6525,
+      "step": 15559
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5916538465918373,
+      "learning_rate": 4.4002915293939186e-10,
+      "loss": 0.6822,
+      "step": 15560
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3106627065560392,
+      "learning_rate": 4.2638586481846823e-10,
+      "loss": 0.6635,
+      "step": 15561
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2722800563629562,
+      "learning_rate": 4.1295741951108417e-10,
+      "loss": 0.6108,
+      "step": 15562
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.456585027536061,
+      "learning_rate": 3.9974381759289027e-10,
+      "loss": 0.6229,
+      "step": 15563
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3786475898868948,
+      "learning_rate": 3.867450596328759e-10,
+      "loss": 0.6741,
+      "step": 15564
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7386370951716614,
+      "learning_rate": 3.739611461889281e-10,
+      "loss": 0.6712,
+      "step": 15565
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3459826184432473,
+      "learning_rate": 3.6139207781060723e-10,
+      "loss": 0.5684,
+      "step": 15566
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.634145963506164,
+      "learning_rate": 3.490378550380369e-10,
+      "loss": 0.6649,
+      "step": 15567
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.897636001087922,
+      "learning_rate": 3.368984784024587e-10,
+      "loss": 0.7262,
+      "step": 15568
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.357811520218764,
+      "learning_rate": 3.2497394842512244e-10,
+      "loss": 0.6564,
+      "step": 15569
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4680398896264144,
+      "learning_rate": 3.1326426561839597e-10,
+      "loss": 0.666,
+      "step": 15570
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.490321646475494,
+      "learning_rate": 3.017694304852103e-10,
+      "loss": 0.5809,
+      "step": 15571
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.069646186887372,
+      "learning_rate": 2.9048944352072505e-10,
+      "loss": 0.6249,
+      "step": 15572
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2670903619488851,
+      "learning_rate": 2.794243052084422e-10,
+      "loss": 0.6164,
+      "step": 15573
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4044881535214677,
+      "learning_rate": 2.685740160240924e-10,
+      "loss": 0.6322,
+      "step": 15574
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.542908285876712,
+      "learning_rate": 2.5793857643396924e-10,
+      "loss": 0.5857,
+      "step": 15575
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.153896982572405,
+      "learning_rate": 2.475179868960398e-10,
+      "loss": 0.7302,
+      "step": 15576
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0561139210234807,
+      "learning_rate": 2.3731224785661365e-10,
+      "loss": 0.665,
+      "step": 15577
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2895370812983267,
+      "learning_rate": 2.273213597553392e-10,
+      "loss": 0.6453,
+      "step": 15578
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.407052796262983,
+      "learning_rate": 2.1754532302076247e-10,
+      "loss": 0.6376,
+      "step": 15579
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4947422138588473,
+      "learning_rate": 2.0798413807365802e-10,
+      "loss": 0.6865,
+      "step": 15580
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.328962059992125,
+      "learning_rate": 1.9863780532480836e-10,
+      "loss": 0.5426,
+      "step": 15581
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1771466349480377,
+      "learning_rate": 1.89506325175004e-10,
+      "loss": 0.5684,
+      "step": 15582
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.263484701501509,
+      "learning_rate": 1.80589698017819e-10,
+      "loss": 0.5942,
+      "step": 15583
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2742262574577274,
+      "learning_rate": 1.718879242357252e-10,
+      "loss": 0.5884,
+      "step": 15584
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5579254410709003,
+      "learning_rate": 1.6340100420286776e-10,
+      "loss": 0.6601,
+      "step": 15585
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5358877724368114,
+      "learning_rate": 1.5512893828395493e-10,
+      "loss": 0.5879,
+      "step": 15586
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.070015386579493,
+      "learning_rate": 1.470717268342581e-10,
+      "loss": 0.6359,
+      "step": 15587
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2935740563830411,
+      "learning_rate": 1.392293702001668e-10,
+      "loss": 0.5493,
+      "step": 15588
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3477732508094546,
+      "learning_rate": 1.316018687191889e-10,
+      "loss": 0.5601,
+      "step": 15589
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6828276045025978,
+      "learning_rate": 1.2418922271772993e-10,
+      "loss": 0.7142,
+      "step": 15590
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1429643878783875,
+      "learning_rate": 1.1699143251553414e-10,
+      "loss": 0.6436,
+      "step": 15591
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0563004538053447,
+      "learning_rate": 1.1000849842179862e-10,
+      "loss": 0.7222,
+      "step": 15592
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3673260540328136,
+      "learning_rate": 1.032404207362836e-10,
+      "loss": 0.6118,
+      "step": 15593
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4939552631340536,
+      "learning_rate": 9.668719974986751e-11,
+      "loss": 0.7169,
+      "step": 15594
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4769977789520943,
+      "learning_rate": 9.03488357439919e-11,
+      "loss": 0.6109,
+      "step": 15595
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3307599432133919,
+      "learning_rate": 8.422532899121649e-11,
+      "loss": 0.6396,
+      "step": 15596
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5567647250514738,
+      "learning_rate": 7.831667975466417e-11,
+      "loss": 0.6624,
+      "step": 15597
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3659699483224692,
+      "learning_rate": 7.2622888288576e-11,
+      "loss": 0.6502,
+      "step": 15598
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4366301376366217,
+      "learning_rate": 6.714395483720105e-11,
+      "loss": 0.72,
+      "step": 15599
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.173461017759572,
+      "learning_rate": 6.187987963590658e-11,
+      "loss": 0.685,
+      "step": 15600
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.54044016112656,
+      "learning_rate": 5.6830662911178115e-11,
+      "loss": 0.6259,
+      "step": 15601
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6351172420280626,
+      "learning_rate": 5.199630488006424e-11,
+      "loss": 0.5907,
+      "step": 15602
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3512082798771217,
+      "learning_rate": 4.737680575017667e-11,
+      "loss": 0.6409,
+      "step": 15603
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3452450767874837,
+      "learning_rate": 4.297216571969021e-11,
+      "loss": 0.6286,
+      "step": 15604
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2655867086635184,
+      "learning_rate": 3.878238497845299e-11,
+      "loss": 0.6241,
+      "step": 15605
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2038211222937976,
+      "learning_rate": 3.4807463705766044e-11,
+      "loss": 0.6413,
+      "step": 15606
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3824777829985602,
+      "learning_rate": 3.104740207371393e-11,
+      "loss": 0.6307,
+      "step": 15607
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3394123560318547,
+      "learning_rate": 2.7502200242723874e-11,
+      "loss": 0.9087,
+      "step": 15608
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2837961186567455,
+      "learning_rate": 2.417185836545155e-11,
+      "loss": 0.7496,
+      "step": 15609
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5306220737819043,
+      "learning_rate": 2.1056376585115723e-11,
+      "loss": 0.6736,
+      "step": 15610
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.360666470717302,
+      "learning_rate": 1.815575503549827e-11,
+      "loss": 0.6226,
+      "step": 15611
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1566414149559052,
+      "learning_rate": 1.5469993840944163e-11,
+      "loss": 0.4574,
+      "step": 15612
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2379786096641863,
+      "learning_rate": 1.2999093117471717e-11,
+      "loss": 0.5833,
+      "step": 15613
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6767233265032588,
+      "learning_rate": 1.0743052971107225e-11,
+      "loss": 0.6602,
+      "step": 15614
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9982571623196719,
+      "learning_rate": 8.701873498995205e-12,
+      "loss": 0.582,
+      "step": 15615
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4573687080042876,
+      "learning_rate": 6.875554787733052e-12,
+      "loss": 0.6872,
+      "step": 15616
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3943155130514362,
+      "learning_rate": 5.2640969172568225e-12,
+      "loss": 0.5013,
+      "step": 15617
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.286574151866362,
+      "learning_rate": 3.867499955845233e-12,
+      "loss": 0.6569,
+      "step": 15618
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4363450720567288,
+      "learning_rate": 2.6857639640054387e-12,
+      "loss": 0.7109,
+      "step": 15619
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5992108912053928,
+      "learning_rate": 1.7188889922525876e-12,
+      "loss": 0.5969,
+      "step": 15620
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2026211439387164,
+      "learning_rate": 9.668750822200423e-13,
+      "loss": 0.649,
+      "step": 15621
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.342717656085749,
+      "learning_rate": 4.2972226665938255e-13,
+      "loss": 0.5576,
+      "step": 15622
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3973943238883246,
+      "learning_rate": 1.0743056777506866e-13,
+      "loss": 0.5862,
+      "step": 15623
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3177786711792432,
+      "learning_rate": 0.0,
+      "loss": 0.6158,
+      "step": 15624
+    },
+    {
+      "epoch": 1.0,
+      "step": 15624,
+      "total_flos": 7456711478362112.0,
+      "train_loss": 0.6766228766485294,
+      "train_runtime": 70767.1188,
+      "train_samples_per_second": 14.131,
+      "train_steps_per_second": 0.221
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 15624,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 2000,
+  "total_flos": 7456711478362112.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}