diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,46440 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 6630,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 219.80169607321315,
+      "learning_rate": 5.025125628140704e-08,
+      "loss": 1.4166,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 68.93088300407445,
+      "learning_rate": 1.0050251256281409e-07,
+      "loss": 1.3768,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 54.936783109940805,
+      "learning_rate": 1.5075376884422112e-07,
+      "loss": 1.35,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.147084785627328,
+      "learning_rate": 2.0100502512562817e-07,
+      "loss": 1.1854,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 56.07248625611662,
+      "learning_rate": 2.512562814070352e-07,
+      "loss": 1.3999,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 37.91675848222748,
+      "learning_rate": 3.0150753768844224e-07,
+      "loss": 1.2615,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 57.208281128557715,
+      "learning_rate": 3.5175879396984927e-07,
+      "loss": 1.2629,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 348.3970313714995,
+      "learning_rate": 4.0201005025125634e-07,
+      "loss": 1.2685,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 52.50230647732762,
+      "learning_rate": 4.5226130653266337e-07,
+      "loss": 1.3959,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 45.244460400126854,
+      "learning_rate": 5.025125628140704e-07,
+      "loss": 1.3011,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 46.12149197339764,
+      "learning_rate": 5.527638190954775e-07,
+      "loss": 1.1406,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 41.97396330351521,
+      "learning_rate": 6.030150753768845e-07,
+      "loss": 1.0228,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 45.40047132020229,
+      "learning_rate": 6.532663316582916e-07,
+      "loss": 1.0994,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 47.357890256244474,
+      "learning_rate": 7.035175879396985e-07,
+      "loss": 1.0124,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 45.92016237398398,
+      "learning_rate": 7.537688442211055e-07,
+      "loss": 0.958,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 36.41115963246688,
+      "learning_rate": 8.040201005025127e-07,
+      "loss": 0.9093,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 19.859992493272486,
+      "learning_rate": 8.542713567839197e-07,
+      "loss": 0.923,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 18.60349269113141,
+      "learning_rate": 9.045226130653267e-07,
+      "loss": 0.8999,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 57.136740770546616,
+      "learning_rate": 9.547738693467337e-07,
+      "loss": 0.8842,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 20.535094238126142,
+      "learning_rate": 1.0050251256281409e-06,
+      "loss": 0.9231,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 24.3905464446835,
+      "learning_rate": 1.0552763819095479e-06,
+      "loss": 0.8877,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 24.573896367893912,
+      "learning_rate": 1.105527638190955e-06,
+      "loss": 0.7757,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 132.19813272395143,
+      "learning_rate": 1.155778894472362e-06,
+      "loss": 0.7601,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 22.63715442823188,
+      "learning_rate": 1.206030150753769e-06,
+      "loss": 0.7715,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 21.10455964482046,
+      "learning_rate": 1.256281407035176e-06,
+      "loss": 0.8811,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 19.53217334448488,
+      "learning_rate": 1.3065326633165831e-06,
+      "loss": 0.7143,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 18.515790932608283,
+      "learning_rate": 1.35678391959799e-06,
+      "loss": 0.7327,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 109.51013540871278,
+      "learning_rate": 1.407035175879397e-06,
+      "loss": 0.6849,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 33.71827134253702,
+      "learning_rate": 1.457286432160804e-06,
+      "loss": 0.6873,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 29.189223156503477,
+      "learning_rate": 1.507537688442211e-06,
+      "loss": 0.8473,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 35.796275909570355,
+      "learning_rate": 1.5577889447236184e-06,
+      "loss": 0.6991,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 23.283695261880045,
+      "learning_rate": 1.6080402010050254e-06,
+      "loss": 0.7264,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 29.31591340941401,
+      "learning_rate": 1.6582914572864323e-06,
+      "loss": 0.7455,
+      "step": 33
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 26.707554037725632,
+      "learning_rate": 1.7085427135678393e-06,
+      "loss": 0.7368,
+      "step": 34
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 37.09136708374297,
+      "learning_rate": 1.7587939698492465e-06,
+      "loss": 0.6374,
+      "step": 35
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 40.854322655846346,
+      "learning_rate": 1.8090452261306535e-06,
+      "loss": 0.7454,
+      "step": 36
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 26.064397979559658,
+      "learning_rate": 1.8592964824120604e-06,
+      "loss": 0.6418,
+      "step": 37
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 26.509150666716433,
+      "learning_rate": 1.9095477386934674e-06,
+      "loss": 0.6377,
+      "step": 38
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 125.62839699992506,
+      "learning_rate": 1.9597989949748746e-06,
+      "loss": 0.7271,
+      "step": 39
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 48.21997249089312,
+      "learning_rate": 2.0100502512562818e-06,
+      "loss": 0.7225,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 88.40507848284909,
+      "learning_rate": 2.0603015075376885e-06,
+      "loss": 0.8219,
+      "step": 41
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 27.714556764927426,
+      "learning_rate": 2.1105527638190957e-06,
+      "loss": 0.7235,
+      "step": 42
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 25.23501982065819,
+      "learning_rate": 2.1608040201005025e-06,
+      "loss": 0.6816,
+      "step": 43
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 29.398637269021645,
+      "learning_rate": 2.21105527638191e-06,
+      "loss": 0.7052,
+      "step": 44
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 45.86690931113663,
+      "learning_rate": 2.261306532663317e-06,
+      "loss": 0.7121,
+      "step": 45
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 54.91948007458609,
+      "learning_rate": 2.311557788944724e-06,
+      "loss": 0.7021,
+      "step": 46
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 31.58202896907573,
+      "learning_rate": 2.3618090452261308e-06,
+      "loss": 0.6895,
+      "step": 47
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 32.674805702378215,
+      "learning_rate": 2.412060301507538e-06,
+      "loss": 0.6169,
+      "step": 48
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 46.388720790134435,
+      "learning_rate": 2.462311557788945e-06,
+      "loss": 0.6477,
+      "step": 49
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 79.95580591349393,
+      "learning_rate": 2.512562814070352e-06,
+      "loss": 0.6572,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 35.90023142018094,
+      "learning_rate": 2.562814070351759e-06,
+      "loss": 0.6494,
+      "step": 51
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 53.69914568614483,
+      "learning_rate": 2.6130653266331663e-06,
+      "loss": 0.6284,
+      "step": 52
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 28.33405194150982,
+      "learning_rate": 2.663316582914573e-06,
+      "loss": 0.6885,
+      "step": 53
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 27.997754632046586,
+      "learning_rate": 2.71356783919598e-06,
+      "loss": 0.5837,
+      "step": 54
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 45.2371368322122,
+      "learning_rate": 2.763819095477387e-06,
+      "loss": 0.7025,
+      "step": 55
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.389732964600146,
+      "learning_rate": 2.814070351758794e-06,
+      "loss": 0.7481,
+      "step": 56
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 84.75442800691671,
+      "learning_rate": 2.8643216080402013e-06,
+      "loss": 0.6395,
+      "step": 57
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 54.724916319146274,
+      "learning_rate": 2.914572864321608e-06,
+      "loss": 0.6724,
+      "step": 58
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 36.79325989363227,
+      "learning_rate": 2.9648241206030153e-06,
+      "loss": 0.6405,
+      "step": 59
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 35.726624372962526,
+      "learning_rate": 3.015075376884422e-06,
+      "loss": 0.6721,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 41.99486819155758,
+      "learning_rate": 3.065326633165829e-06,
+      "loss": 0.6044,
+      "step": 61
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 36.66938168093264,
+      "learning_rate": 3.115577889447237e-06,
+      "loss": 0.6386,
+      "step": 62
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 82.94525180209143,
+      "learning_rate": 3.165829145728643e-06,
+      "loss": 0.6665,
+      "step": 63
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 46.23617089605336,
+      "learning_rate": 3.2160804020100507e-06,
+      "loss": 0.6014,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 62.928348783059285,
+      "learning_rate": 3.266331658291458e-06,
+      "loss": 0.6793,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 40.400025749615885,
+      "learning_rate": 3.3165829145728647e-06,
+      "loss": 0.5949,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 43.274373268410514,
+      "learning_rate": 3.366834170854272e-06,
+      "loss": 0.6838,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 41.17740099944803,
+      "learning_rate": 3.4170854271356786e-06,
+      "loss": 0.6863,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 36.223954122007875,
+      "learning_rate": 3.467336683417086e-06,
+      "loss": 0.7148,
+      "step": 69
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 42.239556001754615,
+      "learning_rate": 3.517587939698493e-06,
+      "loss": 0.5889,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 72.26281145058014,
+      "learning_rate": 3.5678391959798997e-06,
+      "loss": 0.6382,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 51.064421074962404,
+      "learning_rate": 3.618090452261307e-06,
+      "loss": 0.6399,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 46.13746945750372,
+      "learning_rate": 3.6683417085427137e-06,
+      "loss": 0.5724,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 61.34816004780313,
+      "learning_rate": 3.718592964824121e-06,
+      "loss": 0.5273,
+      "step": 74
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 43.20589193064767,
+      "learning_rate": 3.768844221105528e-06,
+      "loss": 0.5118,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 30.10497917287366,
+      "learning_rate": 3.819095477386935e-06,
+      "loss": 0.561,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 31.968533737951724,
+      "learning_rate": 3.869346733668342e-06,
+      "loss": 0.6049,
+      "step": 77
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 44.26432209974167,
+      "learning_rate": 3.919597989949749e-06,
+      "loss": 0.6455,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 39.68079476845012,
+      "learning_rate": 3.969849246231156e-06,
+      "loss": 0.542,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 30.715115917294153,
+      "learning_rate": 4.0201005025125635e-06,
+      "loss": 0.5455,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 45.51910069229239,
+      "learning_rate": 4.07035175879397e-06,
+      "loss": 0.5784,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 59.678584465472404,
+      "learning_rate": 4.120603015075377e-06,
+      "loss": 0.6253,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 38.605599627687425,
+      "learning_rate": 4.170854271356784e-06,
+      "loss": 0.5777,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 29.301353927752366,
+      "learning_rate": 4.221105527638191e-06,
+      "loss": 0.5665,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6523416831756592,
+      "learning_rate": 4.271356783919598e-06,
+      "loss": 0.4574,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 28.83187389327923,
+      "learning_rate": 4.321608040201005e-06,
+      "loss": 0.6209,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 21.872822130085265,
+      "learning_rate": 4.3718592964824125e-06,
+      "loss": 0.5853,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 56.360847037427874,
+      "learning_rate": 4.42211055276382e-06,
+      "loss": 0.5734,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 48.68148087515708,
+      "learning_rate": 4.472361809045226e-06,
+      "loss": 0.5491,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 30.65555739079307,
+      "learning_rate": 4.522613065326634e-06,
+      "loss": 0.5757,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 29.256091037601923,
+      "learning_rate": 4.57286432160804e-06,
+      "loss": 0.5914,
+      "step": 91
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 27.8410115332911,
+      "learning_rate": 4.623115577889448e-06,
+      "loss": 0.6077,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 49.21321807169021,
+      "learning_rate": 4.673366834170855e-06,
+      "loss": 0.5944,
+      "step": 93
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 40.55362164051215,
+      "learning_rate": 4.7236180904522615e-06,
+      "loss": 0.674,
+      "step": 94
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 51.61931029001503,
+      "learning_rate": 4.773869346733669e-06,
+      "loss": 0.6142,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 27.6470664763442,
+      "learning_rate": 4.824120603015076e-06,
+      "loss": 0.6173,
+      "step": 96
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 34.293180346262666,
+      "learning_rate": 4.874371859296483e-06,
+      "loss": 0.5773,
+      "step": 97
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 38.65818714603588,
+      "learning_rate": 4.92462311557789e-06,
+      "loss": 0.657,
+      "step": 98
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 21.35105429407032,
+      "learning_rate": 4.974874371859297e-06,
+      "loss": 0.5465,
+      "step": 99
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 30.012254126098302,
+      "learning_rate": 5.025125628140704e-06,
+      "loss": 0.6993,
+      "step": 100
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 28.289640193861494,
+      "learning_rate": 5.0753768844221105e-06,
+      "loss": 0.5553,
+      "step": 101
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 102.01334548254385,
+      "learning_rate": 5.125628140703518e-06,
+      "loss": 0.6439,
+      "step": 102
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 34.36520299347059,
+      "learning_rate": 5.175879396984925e-06,
+      "loss": 0.5853,
+      "step": 103
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 25.425230295497396,
+      "learning_rate": 5.2261306532663325e-06,
+      "loss": 0.6361,
+      "step": 104
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 34.76338183402994,
+      "learning_rate": 5.2763819095477384e-06,
+      "loss": 0.5541,
+      "step": 105
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 46.525530607258666,
+      "learning_rate": 5.326633165829146e-06,
+      "loss": 0.5394,
+      "step": 106
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 27.541364641367817,
+      "learning_rate": 5.376884422110553e-06,
+      "loss": 0.6212,
+      "step": 107
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 25.842033740196463,
+      "learning_rate": 5.42713567839196e-06,
+      "loss": 0.6432,
+      "step": 108
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 35.818253585216006,
+      "learning_rate": 5.477386934673368e-06,
+      "loss": 0.5838,
+      "step": 109
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 36.4882165209498,
+      "learning_rate": 5.527638190954774e-06,
+      "loss": 0.5933,
+      "step": 110
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 37.5355979339956,
+      "learning_rate": 5.577889447236181e-06,
+      "loss": 0.6544,
+      "step": 111
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 41.39378007342803,
+      "learning_rate": 5.628140703517588e-06,
+      "loss": 0.5863,
+      "step": 112
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 111.1860664374623,
+      "learning_rate": 5.678391959798996e-06,
+      "loss": 0.4986,
+      "step": 113
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 33.26851464218771,
+      "learning_rate": 5.728643216080403e-06,
+      "loss": 0.5632,
+      "step": 114
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 21.775705044101347,
+      "learning_rate": 5.778894472361809e-06,
+      "loss": 0.6209,
+      "step": 115
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 27.742539018435583,
+      "learning_rate": 5.829145728643216e-06,
+      "loss": 0.5511,
+      "step": 116
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 22.962555775332277,
+      "learning_rate": 5.879396984924624e-06,
+      "loss": 0.7013,
+      "step": 117
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 22.126038778253104,
+      "learning_rate": 5.9296482412060305e-06,
+      "loss": 0.5704,
+      "step": 118
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 34.81754750465778,
+      "learning_rate": 5.979899497487438e-06,
+      "loss": 0.6602,
+      "step": 119
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 20.722454958381157,
+      "learning_rate": 6.030150753768844e-06,
+      "loss": 0.6041,
+      "step": 120
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.851637449637002,
+      "learning_rate": 6.080402010050252e-06,
+      "loss": 0.493,
+      "step": 121
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 21.763183463124886,
+      "learning_rate": 6.130653266331658e-06,
+      "loss": 0.598,
+      "step": 122
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 40.89684016180527,
+      "learning_rate": 6.180904522613066e-06,
+      "loss": 0.5659,
+      "step": 123
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 21.181290876327587,
+      "learning_rate": 6.231155778894474e-06,
+      "loss": 0.6299,
+      "step": 124
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 20.56180954584001,
+      "learning_rate": 6.28140703517588e-06,
+      "loss": 0.5623,
+      "step": 125
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 75.32726603198796,
+      "learning_rate": 6.331658291457286e-06,
+      "loss": 0.5382,
+      "step": 126
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 20.981367869000625,
+      "learning_rate": 6.381909547738694e-06,
+      "loss": 0.6155,
+      "step": 127
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 37.67101203691849,
+      "learning_rate": 6.4321608040201015e-06,
+      "loss": 0.6474,
+      "step": 128
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 24.158030044125677,
+      "learning_rate": 6.482412060301508e-06,
+      "loss": 0.5876,
+      "step": 129
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 20.280132606259215,
+      "learning_rate": 6.532663316582916e-06,
+      "loss": 0.5876,
+      "step": 130
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 17.6756733665952,
+      "learning_rate": 6.582914572864322e-06,
+      "loss": 0.5944,
+      "step": 131
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 16.17184035828278,
+      "learning_rate": 6.633165829145729e-06,
+      "loss": 0.6571,
+      "step": 132
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 14.491604437792551,
+      "learning_rate": 6.683417085427136e-06,
+      "loss": 0.5428,
+      "step": 133
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 107.3445942141567,
+      "learning_rate": 6.733668341708544e-06,
+      "loss": 0.597,
+      "step": 134
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 12.41658507054546,
+      "learning_rate": 6.7839195979899505e-06,
+      "loss": 0.5548,
+      "step": 135
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 14.919543575370547,
+      "learning_rate": 6.834170854271357e-06,
+      "loss": 0.6018,
+      "step": 136
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.76074869020652,
+      "learning_rate": 6.884422110552764e-06,
+      "loss": 0.4914,
+      "step": 137
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 17.592539228768032,
+      "learning_rate": 6.934673366834172e-06,
+      "loss": 0.5318,
+      "step": 138
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.846351562760573,
+      "learning_rate": 6.984924623115578e-06,
+      "loss": 0.5191,
+      "step": 139
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 12.298590056914627,
+      "learning_rate": 7.035175879396986e-06,
+      "loss": 0.5853,
+      "step": 140
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.331932940756646,
+      "learning_rate": 7.085427135678392e-06,
+      "loss": 0.592,
+      "step": 141
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 12.640841634849922,
+      "learning_rate": 7.1356783919597995e-06,
+      "loss": 0.577,
+      "step": 142
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 14.627934085382863,
+      "learning_rate": 7.185929648241206e-06,
+      "loss": 0.6412,
+      "step": 143
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.041339890524648,
+      "learning_rate": 7.236180904522614e-06,
+      "loss": 0.6223,
+      "step": 144
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 12.080668821939677,
+      "learning_rate": 7.2864321608040215e-06,
+      "loss": 0.5403,
+      "step": 145
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 28.59118964978665,
+      "learning_rate": 7.336683417085427e-06,
+      "loss": 0.5672,
+      "step": 146
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 22.512062985869857,
+      "learning_rate": 7.386934673366835e-06,
+      "loss": 0.572,
+      "step": 147
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 14.611894722368069,
+      "learning_rate": 7.437185929648242e-06,
+      "loss": 0.5913,
+      "step": 148
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 12.908800658783578,
+      "learning_rate": 7.487437185929649e-06,
+      "loss": 0.5541,
+      "step": 149
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 12.572046015264789,
+      "learning_rate": 7.537688442211056e-06,
+      "loss": 0.5756,
+      "step": 150
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.076290379152406,
+      "learning_rate": 7.587939698492463e-06,
+      "loss": 0.7968,
+      "step": 151
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.142987672905392,
+      "learning_rate": 7.63819095477387e-06,
+      "loss": 0.5895,
+      "step": 152
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.219279862892433,
+      "learning_rate": 7.688442211055276e-06,
+      "loss": 0.6431,
+      "step": 153
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 82.51032088105623,
+      "learning_rate": 7.738693467336685e-06,
+      "loss": 0.6039,
+      "step": 154
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 25.521460456044846,
+      "learning_rate": 7.788944723618092e-06,
+      "loss": 0.5741,
+      "step": 155
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 15.66496154723627,
+      "learning_rate": 7.839195979899498e-06,
+      "loss": 0.5678,
+      "step": 156
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 14.600602524240477,
+      "learning_rate": 7.889447236180905e-06,
+      "loss": 0.6139,
+      "step": 157
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.507798186601256,
+      "learning_rate": 7.939698492462312e-06,
+      "loss": 0.569,
+      "step": 158
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.306639039986711,
+      "learning_rate": 7.989949748743719e-06,
+      "loss": 0.541,
+      "step": 159
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.342281037687586,
+      "learning_rate": 8.040201005025127e-06,
+      "loss": 0.53,
+      "step": 160
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.30352598042552,
+      "learning_rate": 8.090452261306532e-06,
+      "loss": 0.6573,
+      "step": 161
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.30032789353799,
+      "learning_rate": 8.14070351758794e-06,
+      "loss": 0.5715,
+      "step": 162
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.5080432563917014,
+      "learning_rate": 8.190954773869347e-06,
+      "loss": 0.6203,
+      "step": 163
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.212160451857507,
+      "learning_rate": 8.241206030150754e-06,
+      "loss": 0.5741,
+      "step": 164
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8364686825594059,
+      "learning_rate": 8.291457286432163e-06,
+      "loss": 0.6082,
+      "step": 165
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.04524440603669,
+      "learning_rate": 8.341708542713568e-06,
+      "loss": 0.5691,
+      "step": 166
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 27.457121977496122,
+      "learning_rate": 8.391959798994976e-06,
+      "loss": 0.5014,
+      "step": 167
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 29.12754737938816,
+      "learning_rate": 8.442211055276383e-06,
+      "loss": 0.6045,
+      "step": 168
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2743047049679155,
+      "learning_rate": 8.49246231155779e-06,
+      "loss": 0.5052,
+      "step": 169
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.939720455650788,
+      "learning_rate": 8.542713567839196e-06,
+      "loss": 0.5941,
+      "step": 170
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 13.847402650316074,
+      "learning_rate": 8.592964824120603e-06,
+      "loss": 0.6353,
+      "step": 171
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.69994930362305,
+      "learning_rate": 8.64321608040201e-06,
+      "loss": 0.5709,
+      "step": 172
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 14.852017450906747,
+      "learning_rate": 8.693467336683418e-06,
+      "loss": 0.5974,
+      "step": 173
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.282759603287758,
+      "learning_rate": 8.743718592964825e-06,
+      "loss": 0.6649,
+      "step": 174
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.705648615815905,
+      "learning_rate": 8.793969849246232e-06,
+      "loss": 0.6479,
+      "step": 175
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 17.378780226562494,
+      "learning_rate": 8.84422110552764e-06,
+      "loss": 0.6284,
+      "step": 176
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.485875767852814,
+      "learning_rate": 8.894472361809045e-06,
+      "loss": 0.6334,
+      "step": 177
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 28.493604154824265,
+      "learning_rate": 8.944723618090452e-06,
+      "loss": 0.57,
+      "step": 178
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.02569682051414,
+      "learning_rate": 8.99497487437186e-06,
+      "loss": 0.6414,
+      "step": 179
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 23.94964353367036,
+      "learning_rate": 9.045226130653267e-06,
+      "loss": 0.5752,
+      "step": 180
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.201892832647694,
+      "learning_rate": 9.095477386934674e-06,
+      "loss": 0.6497,
+      "step": 181
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 34.56725707129777,
+      "learning_rate": 9.14572864321608e-06,
+      "loss": 0.6286,
+      "step": 182
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.365885747574012,
+      "learning_rate": 9.195979899497488e-06,
+      "loss": 0.5054,
+      "step": 183
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.315640561207363,
+      "learning_rate": 9.246231155778896e-06,
+      "loss": 0.4883,
+      "step": 184
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.234484347412982,
+      "learning_rate": 9.296482412060303e-06,
+      "loss": 0.5716,
+      "step": 185
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.013635084470593,
+      "learning_rate": 9.34673366834171e-06,
+      "loss": 0.5924,
+      "step": 186
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 14.630225264034479,
+      "learning_rate": 9.396984924623116e-06,
+      "loss": 0.4601,
+      "step": 187
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.696098457183469,
+      "learning_rate": 9.447236180904523e-06,
+      "loss": 0.577,
+      "step": 188
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.884969530672969,
+      "learning_rate": 9.49748743718593e-06,
+      "loss": 0.5676,
+      "step": 189
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.670258798400683,
+      "learning_rate": 9.547738693467338e-06,
+      "loss": 0.5952,
+      "step": 190
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.71830253774252,
+      "learning_rate": 9.597989949748745e-06,
+      "loss": 0.6696,
+      "step": 191
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.352870366648284,
+      "learning_rate": 9.648241206030152e-06,
+      "loss": 0.5667,
+      "step": 192
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 13.676226977398779,
+      "learning_rate": 9.698492462311559e-06,
+      "loss": 0.6238,
+      "step": 193
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.293976772997915,
+      "learning_rate": 9.748743718592965e-06,
+      "loss": 0.6028,
+      "step": 194
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 22.435553917528523,
+      "learning_rate": 9.798994974874372e-06,
+      "loss": 0.5388,
+      "step": 195
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.757970695912968,
+      "learning_rate": 9.84924623115578e-06,
+      "loss": 0.6014,
+      "step": 196
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7090923697651277,
+      "learning_rate": 9.899497487437186e-06,
+      "loss": 0.5893,
+      "step": 197
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.88306428787245,
+      "learning_rate": 9.949748743718594e-06,
+      "loss": 0.5427,
+      "step": 198
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 20.61088049278884,
+      "learning_rate": 1e-05,
+      "loss": 0.5998,
+      "step": 199
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.016234517570497,
+      "learning_rate": 9.99999940340072e-06,
+      "loss": 0.706,
+      "step": 200
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.800441976899956,
+      "learning_rate": 9.999997613603016e-06,
+      "loss": 0.6313,
+      "step": 201
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.668359423329318,
+      "learning_rate": 9.999994630607318e-06,
+      "loss": 0.5998,
+      "step": 202
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 29.442454035777118,
+      "learning_rate": 9.999990454414338e-06,
+      "loss": 0.547,
+      "step": 203
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.067155539653964,
+      "learning_rate": 9.999985085025072e-06,
+      "loss": 0.6207,
+      "step": 204
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 32.68446451394774,
+      "learning_rate": 9.999978522440803e-06,
+      "loss": 0.5438,
+      "step": 205
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 13.281531163040034,
+      "learning_rate": 9.999970766663094e-06,
+      "loss": 0.5807,
+      "step": 206
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.477145410296334,
+      "learning_rate": 9.999961817693799e-06,
+      "loss": 0.5582,
+      "step": 207
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.35353289465063,
+      "learning_rate": 9.99995167553505e-06,
+      "loss": 0.6438,
+      "step": 208
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.9351982301113475,
+      "learning_rate": 9.999940340189272e-06,
+      "loss": 0.5542,
+      "step": 209
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.046577237746433,
+      "learning_rate": 9.999927811659165e-06,
+      "loss": 0.5875,
+      "step": 210
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.878968149846823,
+      "learning_rate": 9.999914089947723e-06,
+      "loss": 0.6,
+      "step": 211
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.169102821627069,
+      "learning_rate": 9.999899175058218e-06,
+      "loss": 0.5599,
+      "step": 212
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.50112104914739,
+      "learning_rate": 9.999883066994212e-06,
+      "loss": 0.5484,
+      "step": 213
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 138.92279398438325,
+      "learning_rate": 9.999865765759545e-06,
+      "loss": 0.6151,
+      "step": 214
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.162819961893684,
+      "learning_rate": 9.999847271358347e-06,
+      "loss": 0.5758,
+      "step": 215
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.284957897146134,
+      "learning_rate": 9.999827583795034e-06,
+      "loss": 0.596,
+      "step": 216
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 27.69215261996493,
+      "learning_rate": 9.999806703074302e-06,
+      "loss": 0.5723,
+      "step": 217
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 26.909067557024127,
+      "learning_rate": 9.999784629201135e-06,
+      "loss": 0.5338,
+      "step": 218
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.64861562521606,
+      "learning_rate": 9.9997613621808e-06,
+      "loss": 0.6206,
+      "step": 219
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.976521658151706,
+      "learning_rate": 9.99973690201885e-06,
+      "loss": 0.5941,
+      "step": 220
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.777655350160853,
+      "learning_rate": 9.999711248721123e-06,
+      "loss": 0.5976,
+      "step": 221
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.745974814050962,
+      "learning_rate": 9.99968440229374e-06,
+      "loss": 0.6298,
+      "step": 222
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.291761308239028,
+      "learning_rate": 9.999656362743105e-06,
+      "loss": 0.6064,
+      "step": 223
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.79596931623397,
+      "learning_rate": 9.999627130075913e-06,
+      "loss": 0.6407,
+      "step": 224
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.600729461315963,
+      "learning_rate": 9.99959670429914e-06,
+      "loss": 0.5889,
+      "step": 225
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.51993555567592,
+      "learning_rate": 9.999565085420043e-06,
+      "loss": 0.5285,
+      "step": 226
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.439883313988274,
+      "learning_rate": 9.999532273446171e-06,
+      "loss": 0.609,
+      "step": 227
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.302480184408054,
+      "learning_rate": 9.999498268385355e-06,
+      "loss": 0.6047,
+      "step": 228
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 29.287480372108416,
+      "learning_rate": 9.999463070245708e-06,
+      "loss": 0.6137,
+      "step": 229
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 13.267635474103606,
+      "learning_rate": 9.999426679035628e-06,
+      "loss": 0.5282,
+      "step": 230
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 16.826599376523802,
+      "learning_rate": 9.999389094763803e-06,
+      "loss": 0.5646,
+      "step": 231
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.873440477203618,
+      "learning_rate": 9.9993503174392e-06,
+      "loss": 0.5036,
+      "step": 232
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 19.65972465759151,
+      "learning_rate": 9.999310347071074e-06,
+      "loss": 0.5581,
+      "step": 233
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 26.12962130035789,
+      "learning_rate": 9.999269183668962e-06,
+      "loss": 0.5532,
+      "step": 234
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.372674692441636,
+      "learning_rate": 9.99922682724269e-06,
+      "loss": 0.5737,
+      "step": 235
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.42407999298219,
+      "learning_rate": 9.999183277802362e-06,
+      "loss": 0.6267,
+      "step": 236
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 36.611472240219435,
+      "learning_rate": 9.999138535358373e-06,
+      "loss": 0.5672,
+      "step": 237
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.825855345960257,
+      "learning_rate": 9.9990925999214e-06,
+      "loss": 0.5957,
+      "step": 238
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 19.323100350984195,
+      "learning_rate": 9.999045471502406e-06,
+      "loss": 0.5767,
+      "step": 239
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.610436189144876,
+      "learning_rate": 9.998997150112636e-06,
+      "loss": 0.5068,
+      "step": 240
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.675084247743643,
+      "learning_rate": 9.998947635763622e-06,
+      "loss": 0.4517,
+      "step": 241
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 19.84468761589128,
+      "learning_rate": 9.99889692846718e-06,
+      "loss": 0.5475,
+      "step": 242
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.197872980064998,
+      "learning_rate": 9.998845028235411e-06,
+      "loss": 0.5209,
+      "step": 243
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.83477683089179,
+      "learning_rate": 9.998791935080701e-06,
+      "loss": 0.6186,
+      "step": 244
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.4993681294156,
+      "learning_rate": 9.998737649015719e-06,
+      "loss": 0.6267,
+      "step": 245
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 31.252219660148725,
+      "learning_rate": 9.99868217005342e-06,
+      "loss": 0.5104,
+      "step": 246
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.15122959085679,
+      "learning_rate": 9.998625498207044e-06,
+      "loss": 0.5399,
+      "step": 247
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 23.82904805404931,
+      "learning_rate": 9.998567633490115e-06,
+      "loss": 0.5158,
+      "step": 248
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.669607222407733,
+      "learning_rate": 9.998508575916442e-06,
+      "loss": 0.5174,
+      "step": 249
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.15240988221442,
+      "learning_rate": 9.998448325500118e-06,
+      "loss": 0.5424,
+      "step": 250
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.809479933372809,
+      "learning_rate": 9.998386882255522e-06,
+      "loss": 0.4983,
+      "step": 251
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.343704750718917,
+      "learning_rate": 9.998324246197315e-06,
+      "loss": 0.5393,
+      "step": 252
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.267931091924353,
+      "learning_rate": 9.998260417340447e-06,
+      "loss": 0.5441,
+      "step": 253
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.68703472201826,
+      "learning_rate": 9.998195395700149e-06,
+      "loss": 0.5243,
+      "step": 254
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.463114207456857,
+      "learning_rate": 9.998129181291937e-06,
+      "loss": 0.5313,
+      "step": 255
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.69128090124307,
+      "learning_rate": 9.998061774131613e-06,
+      "loss": 0.5809,
+      "step": 256
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.123272230025881,
+      "learning_rate": 9.997993174235262e-06,
+      "loss": 0.5827,
+      "step": 257
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.612474846339437,
+      "learning_rate": 9.997923381619257e-06,
+      "loss": 0.5972,
+      "step": 258
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.091467988522647,
+      "learning_rate": 9.997852396300251e-06,
+      "loss": 0.5681,
+      "step": 259
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.9099777000855305,
+      "learning_rate": 9.997780218295186e-06,
+      "loss": 0.5507,
+      "step": 260
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.08624697672159,
+      "learning_rate": 9.997706847621283e-06,
+      "loss": 0.5729,
+      "step": 261
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 14.13594760982386,
+      "learning_rate": 9.997632284296055e-06,
+      "loss": 0.5495,
+      "step": 262
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.334469223874015,
+      "learning_rate": 9.997556528337294e-06,
+      "loss": 0.5871,
+      "step": 263
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.107314307562153,
+      "learning_rate": 9.99747957976308e-06,
+      "loss": 0.5658,
+      "step": 264
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 29.624877151160735,
+      "learning_rate": 9.997401438591772e-06,
+      "loss": 0.6171,
+      "step": 265
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.840800982792306,
+      "learning_rate": 9.997322104842022e-06,
+      "loss": 0.5909,
+      "step": 266
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7693448468099302,
+      "learning_rate": 9.99724157853276e-06,
+      "loss": 0.599,
+      "step": 267
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.148578111689773,
+      "learning_rate": 9.997159859683202e-06,
+      "loss": 0.5699,
+      "step": 268
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.720413406550486,
+      "learning_rate": 9.997076948312854e-06,
+      "loss": 0.5882,
+      "step": 269
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.479917639919372,
+      "learning_rate": 9.996992844441495e-06,
+      "loss": 0.5104,
+      "step": 270
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.40208619326653,
+      "learning_rate": 9.9969075480892e-06,
+      "loss": 0.5415,
+      "step": 271
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.261932920368522,
+      "learning_rate": 9.996821059276324e-06,
+      "loss": 0.6097,
+      "step": 272
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.995186867538704,
+      "learning_rate": 9.996733378023507e-06,
+      "loss": 0.5487,
+      "step": 273
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.766011602556588,
+      "learning_rate": 9.99664450435167e-06,
+      "loss": 0.6235,
+      "step": 274
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.245219985860349,
+      "learning_rate": 9.996554438282022e-06,
+      "loss": 0.5595,
+      "step": 275
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.232355023753634,
+      "learning_rate": 9.99646317983606e-06,
+      "loss": 0.5108,
+      "step": 276
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.324443517614879,
+      "learning_rate": 9.996370729035562e-06,
+      "loss": 0.562,
+      "step": 277
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.675708072644669,
+      "learning_rate": 9.996277085902587e-06,
+      "loss": 0.5871,
+      "step": 278
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.202930918073277,
+      "learning_rate": 9.996182250459482e-06,
+      "loss": 0.6279,
+      "step": 279
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.543966258135516,
+      "learning_rate": 9.99608622272888e-06,
+      "loss": 0.5181,
+      "step": 280
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.0054898237369505,
+      "learning_rate": 9.9959890027337e-06,
+      "loss": 0.4936,
+      "step": 281
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.641326956964122,
+      "learning_rate": 9.995890590497137e-06,
+      "loss": 0.5829,
+      "step": 282
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.288768190807301,
+      "learning_rate": 9.995790986042678e-06,
+      "loss": 0.5789,
+      "step": 283
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.836510171047746,
+      "learning_rate": 9.995690189394095e-06,
+      "loss": 0.6051,
+      "step": 284
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.9540427198366,
+      "learning_rate": 9.995588200575439e-06,
+      "loss": 0.5723,
+      "step": 285
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.4835679648994535,
+      "learning_rate": 9.99548501961105e-06,
+      "loss": 0.6063,
+      "step": 286
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.4688368532425877,
+      "learning_rate": 9.995380646525552e-06,
+      "loss": 0.6053,
+      "step": 287
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.2781985653694194,
+      "learning_rate": 9.99527508134385e-06,
+      "loss": 0.5661,
+      "step": 288
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.057395199344807,
+      "learning_rate": 9.995168324091138e-06,
+      "loss": 0.6176,
+      "step": 289
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.287009804062839,
+      "learning_rate": 9.995060374792891e-06,
+      "loss": 0.5701,
+      "step": 290
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.0065877122942695,
+      "learning_rate": 9.994951233474873e-06,
+      "loss": 0.6082,
+      "step": 291
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.649890075906451,
+      "learning_rate": 9.994840900163127e-06,
+      "loss": 0.5821,
+      "step": 292
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8046918025983905,
+      "learning_rate": 9.994729374883982e-06,
+      "loss": 0.6069,
+      "step": 293
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.293678993348817,
+      "learning_rate": 9.994616657664056e-06,
+      "loss": 0.5658,
+      "step": 294
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.534239074937565,
+      "learning_rate": 9.994502748530243e-06,
+      "loss": 0.6102,
+      "step": 295
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.4955175456292853,
+      "learning_rate": 9.994387647509732e-06,
+      "loss": 0.5602,
+      "step": 296
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.674216498033882,
+      "learning_rate": 9.994271354629985e-06,
+      "loss": 0.5333,
+      "step": 297
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.799211361048443,
+      "learning_rate": 9.994153869918757e-06,
+      "loss": 0.5951,
+      "step": 298
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.317241033094687,
+      "learning_rate": 9.994035193404083e-06,
+      "loss": 0.5362,
+      "step": 299
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.277217311250674,
+      "learning_rate": 9.993915325114286e-06,
+      "loss": 0.5365,
+      "step": 300
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.71239699421881,
+      "learning_rate": 9.993794265077973e-06,
+      "loss": 0.4997,
+      "step": 301
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.2808155530701795,
+      "learning_rate": 9.993672013324027e-06,
+      "loss": 0.5376,
+      "step": 302
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.5451084889740905,
+      "learning_rate": 9.993548569881629e-06,
+      "loss": 0.5484,
+      "step": 303
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.315428158656323,
+      "learning_rate": 9.993423934780234e-06,
+      "loss": 0.4811,
+      "step": 304
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.035551424142229,
+      "learning_rate": 9.993298108049582e-06,
+      "loss": 0.5156,
+      "step": 305
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.18865103806143,
+      "learning_rate": 9.993171089719709e-06,
+      "loss": 0.6081,
+      "step": 306
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.7861730299504783,
+      "learning_rate": 9.993042879820919e-06,
+      "loss": 0.5759,
+      "step": 307
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.2473187991770525,
+      "learning_rate": 9.99291347838381e-06,
+      "loss": 0.6178,
+      "step": 308
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.460983528798518,
+      "learning_rate": 9.992782885439264e-06,
+      "loss": 0.6129,
+      "step": 309
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.2504658221365204,
+      "learning_rate": 9.992651101018445e-06,
+      "loss": 0.6251,
+      "step": 310
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.5471115468633485,
+      "learning_rate": 9.9925181251528e-06,
+      "loss": 0.6206,
+      "step": 311
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.559711856128409,
+      "learning_rate": 9.992383957874064e-06,
+      "loss": 0.4806,
+      "step": 312
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.3691960256501114,
+      "learning_rate": 9.992248599214253e-06,
+      "loss": 0.5963,
+      "step": 313
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.968602774473347,
+      "learning_rate": 9.992112049205674e-06,
+      "loss": 0.558,
+      "step": 314
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.0786187339854365,
+      "learning_rate": 9.991974307880907e-06,
+      "loss": 0.5641,
+      "step": 315
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.138642343520191,
+      "learning_rate": 9.991835375272825e-06,
+      "loss": 0.6231,
+      "step": 316
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.483354275951371,
+      "learning_rate": 9.991695251414584e-06,
+      "loss": 0.5677,
+      "step": 317
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4553910964484875,
+      "learning_rate": 9.991553936339621e-06,
+      "loss": 0.5458,
+      "step": 318
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.541641123752431,
+      "learning_rate": 9.99141143008166e-06,
+      "loss": 0.5739,
+      "step": 319
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.3400238903826867,
+      "learning_rate": 9.991267732674712e-06,
+      "loss": 0.6116,
+      "step": 320
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.142675903895313,
+      "learning_rate": 9.991122844153064e-06,
+      "loss": 0.5393,
+      "step": 321
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.787845243793807,
+      "learning_rate": 9.990976764551295e-06,
+      "loss": 0.5931,
+      "step": 322
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.145015952912078,
+      "learning_rate": 9.990829493904263e-06,
+      "loss": 0.4538,
+      "step": 323
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.363975419405047,
+      "learning_rate": 9.990681032247116e-06,
+      "loss": 0.5752,
+      "step": 324
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.896382169356612,
+      "learning_rate": 9.99053137961528e-06,
+      "loss": 0.5102,
+      "step": 325
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.220431137645542,
+      "learning_rate": 9.99038053604447e-06,
+      "loss": 0.6411,
+      "step": 326
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.368212409738737,
+      "learning_rate": 9.990228501570683e-06,
+      "loss": 0.5237,
+      "step": 327
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.545368920692188,
+      "learning_rate": 9.990075276230198e-06,
+      "loss": 0.5525,
+      "step": 328
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.526793870549361,
+      "learning_rate": 9.989920860059583e-06,
+      "loss": 0.6496,
+      "step": 329
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.39358924922123,
+      "learning_rate": 9.989765253095687e-06,
+      "loss": 0.5744,
+      "step": 330
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.892514219804371,
+      "learning_rate": 9.989608455375646e-06,
+      "loss": 0.5263,
+      "step": 331
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.899152726944619,
+      "learning_rate": 9.989450466936875e-06,
+      "loss": 0.5814,
+      "step": 332
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.51124859640849,
+      "learning_rate": 9.989291287817077e-06,
+      "loss": 0.55,
+      "step": 333
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.6007962146419947,
+      "learning_rate": 9.989130918054242e-06,
+      "loss": 0.5491,
+      "step": 334
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.625736061192414,
+      "learning_rate": 9.988969357686636e-06,
+      "loss": 0.6003,
+      "step": 335
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.681743846070172,
+      "learning_rate": 9.988806606752817e-06,
+      "loss": 0.6247,
+      "step": 336
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.738862838490249,
+      "learning_rate": 9.98864266529162e-06,
+      "loss": 0.5547,
+      "step": 337
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.794891102562757,
+      "learning_rate": 9.98847753334217e-06,
+      "loss": 0.5886,
+      "step": 338
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.345375679451182,
+      "learning_rate": 9.988311210943878e-06,
+      "loss": 0.6024,
+      "step": 339
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.346876133934082,
+      "learning_rate": 9.98814369813643e-06,
+      "loss": 0.5577,
+      "step": 340
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.821343446390748,
+      "learning_rate": 9.987974994959803e-06,
+      "loss": 0.6299,
+      "step": 341
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.68804329700119,
+      "learning_rate": 9.987805101454255e-06,
+      "loss": 0.6043,
+      "step": 342
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.888239691622705,
+      "learning_rate": 9.98763401766033e-06,
+      "loss": 0.6072,
+      "step": 343
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.775970261307909,
+      "learning_rate": 9.987461743618858e-06,
+      "loss": 0.5334,
+      "step": 344
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.066907007927615,
+      "learning_rate": 9.987288279370945e-06,
+      "loss": 0.54,
+      "step": 345
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.543418185542019,
+      "learning_rate": 9.987113624957993e-06,
+      "loss": 0.586,
+      "step": 346
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.269728838035235,
+      "learning_rate": 9.986937780421675e-06,
+      "loss": 0.5388,
+      "step": 347
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.427212880401611,
+      "learning_rate": 9.986760745803961e-06,
+      "loss": 0.5772,
+      "step": 348
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.75785623464889,
+      "learning_rate": 9.986582521147093e-06,
+      "loss": 0.5383,
+      "step": 349
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.917303082957225,
+      "learning_rate": 9.986403106493603e-06,
+      "loss": 0.5156,
+      "step": 350
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.298446282361764,
+      "learning_rate": 9.986222501886311e-06,
+      "loss": 0.5486,
+      "step": 351
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.764076443994568,
+      "learning_rate": 9.986040707368312e-06,
+      "loss": 0.5904,
+      "step": 352
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4753588553302286,
+      "learning_rate": 9.985857722982991e-06,
+      "loss": 0.5751,
+      "step": 353
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.206012679524437,
+      "learning_rate": 9.985673548774015e-06,
+      "loss": 0.491,
+      "step": 354
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.609916604632889,
+      "learning_rate": 9.985488184785337e-06,
+      "loss": 0.567,
+      "step": 355
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.8057742610396685,
+      "learning_rate": 9.985301631061188e-06,
+      "loss": 0.6369,
+      "step": 356
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.233662085062618,
+      "learning_rate": 9.985113887646093e-06,
+      "loss": 0.571,
+      "step": 357
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.763738084839761,
+      "learning_rate": 9.98492495458485e-06,
+      "loss": 0.561,
+      "step": 358
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.287610144203753,
+      "learning_rate": 9.984734831922546e-06,
+      "loss": 0.5604,
+      "step": 359
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.8853425282602165,
+      "learning_rate": 9.984543519704558e-06,
+      "loss": 0.5798,
+      "step": 360
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9998892982915681,
+      "learning_rate": 9.984351017976533e-06,
+      "loss": 0.6312,
+      "step": 361
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.005855156759343,
+      "learning_rate": 9.984157326784414e-06,
+      "loss": 0.612,
+      "step": 362
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.421510220434119,
+      "learning_rate": 9.983962446174423e-06,
+      "loss": 0.5596,
+      "step": 363
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.420281715177006,
+      "learning_rate": 9.983766376193063e-06,
+      "loss": 0.5929,
+      "step": 364
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.577327704606224,
+      "learning_rate": 9.983569116887129e-06,
+      "loss": 0.6193,
+      "step": 365
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.549613676857449,
+      "learning_rate": 9.983370668303691e-06,
+      "loss": 0.5266,
+      "step": 366
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.176602900555059,
+      "learning_rate": 9.983171030490108e-06,
+      "loss": 0.5386,
+      "step": 367
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.053423992976838,
+      "learning_rate": 9.982970203494023e-06,
+      "loss": 0.6397,
+      "step": 368
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6027947669367055,
+      "learning_rate": 9.982768187363359e-06,
+      "loss": 0.5973,
+      "step": 369
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.995753317876349,
+      "learning_rate": 9.982564982146329e-06,
+      "loss": 0.6276,
+      "step": 370
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3903018175244346,
+      "learning_rate": 9.982360587891418e-06,
+      "loss": 0.5643,
+      "step": 371
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.972729191772038,
+      "learning_rate": 9.982155004647412e-06,
+      "loss": 0.5165,
+      "step": 372
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.13611535744167,
+      "learning_rate": 9.981948232463364e-06,
+      "loss": 0.5269,
+      "step": 373
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.120440940286446,
+      "learning_rate": 9.981740271388622e-06,
+      "loss": 0.5889,
+      "step": 374
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.31747381169351,
+      "learning_rate": 9.981531121472811e-06,
+      "loss": 0.5407,
+      "step": 375
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.435769878856964,
+      "learning_rate": 9.981320782765847e-06,
+      "loss": 0.5202,
+      "step": 376
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.281763425719071,
+      "learning_rate": 9.98110925531792e-06,
+      "loss": 0.5953,
+      "step": 377
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.96231008950992,
+      "learning_rate": 9.980896539179512e-06,
+      "loss": 0.4878,
+      "step": 378
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.227847515340797,
+      "learning_rate": 9.980682634401386e-06,
+      "loss": 0.5803,
+      "step": 379
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.851879314521378,
+      "learning_rate": 9.980467541034584e-06,
+      "loss": 0.5999,
+      "step": 380
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.300392466987065,
+      "learning_rate": 9.98025125913044e-06,
+      "loss": 0.5141,
+      "step": 381
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.747166873032592,
+      "learning_rate": 9.980033788740568e-06,
+      "loss": 0.5022,
+      "step": 382
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.556397550568694,
+      "learning_rate": 9.979815129916861e-06,
+      "loss": 0.5584,
+      "step": 383
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.998458221640642,
+      "learning_rate": 9.979595282711503e-06,
+      "loss": 0.6028,
+      "step": 384
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.17924806581665,
+      "learning_rate": 9.979374247176955e-06,
+      "loss": 0.5783,
+      "step": 385
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 17.97507330314053,
+      "learning_rate": 9.97915202336597e-06,
+      "loss": 0.5192,
+      "step": 386
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 76.40610787653104,
+      "learning_rate": 9.978928611331576e-06,
+      "loss": 0.7398,
+      "step": 387
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.9553664677850966,
+      "learning_rate": 9.978704011127086e-06,
+      "loss": 0.5868,
+      "step": 388
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.433542875032972,
+      "learning_rate": 9.978478222806102e-06,
+      "loss": 0.5241,
+      "step": 389
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4497632830171998,
+      "learning_rate": 9.978251246422506e-06,
+      "loss": 0.5404,
+      "step": 390
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.423266334199798,
+      "learning_rate": 9.978023082030462e-06,
+      "loss": 0.6168,
+      "step": 391
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.691295365122946,
+      "learning_rate": 9.97779372968442e-06,
+      "loss": 0.5628,
+      "step": 392
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.393704924655381,
+      "learning_rate": 9.977563189439113e-06,
+      "loss": 0.6399,
+      "step": 393
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.2426386253161965,
+      "learning_rate": 9.977331461349553e-06,
+      "loss": 0.5966,
+      "step": 394
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.795146200147255,
+      "learning_rate": 9.977098545471047e-06,
+      "loss": 0.5219,
+      "step": 395
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.2474805325938485,
+      "learning_rate": 9.976864441859171e-06,
+      "loss": 0.5436,
+      "step": 396
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.453237967535938,
+      "learning_rate": 9.976629150569794e-06,
+      "loss": 0.5401,
+      "step": 397
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.727000388294842,
+      "learning_rate": 9.976392671659068e-06,
+      "loss": 0.5593,
+      "step": 398
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.017743922902207,
+      "learning_rate": 9.976155005183423e-06,
+      "loss": 0.5012,
+      "step": 399
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.922427043907916,
+      "learning_rate": 9.975916151199578e-06,
+      "loss": 0.5488,
+      "step": 400
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.193971439556997,
+      "learning_rate": 9.97567610976453e-06,
+      "loss": 0.5601,
+      "step": 401
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.183013893789854,
+      "learning_rate": 9.975434880935567e-06,
+      "loss": 0.5536,
+      "step": 402
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.29503952869622,
+      "learning_rate": 9.975192464770253e-06,
+      "loss": 0.6191,
+      "step": 403
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 24.980136083194964,
+      "learning_rate": 9.974948861326436e-06,
+      "loss": 0.5004,
+      "step": 404
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.823715892383603,
+      "learning_rate": 9.974704070662254e-06,
+      "loss": 0.5701,
+      "step": 405
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.669098554552117,
+      "learning_rate": 9.974458092836121e-06,
+      "loss": 0.626,
+      "step": 406
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.403529086133226,
+      "learning_rate": 9.974210927906737e-06,
+      "loss": 0.6251,
+      "step": 407
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 19.085284398792762,
+      "learning_rate": 9.973962575933087e-06,
+      "loss": 0.5815,
+      "step": 408
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.675478786974509,
+      "learning_rate": 9.973713036974438e-06,
+      "loss": 0.56,
+      "step": 409
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 24.41717595984447,
+      "learning_rate": 9.973462311090336e-06,
+      "loss": 0.5049,
+      "step": 410
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.9275479334233583,
+      "learning_rate": 9.973210398340619e-06,
+      "loss": 0.5306,
+      "step": 411
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.740054884165736,
+      "learning_rate": 9.972957298785397e-06,
+      "loss": 0.5695,
+      "step": 412
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.886581944733458,
+      "learning_rate": 9.972703012485078e-06,
+      "loss": 0.5416,
+      "step": 413
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.451230677600706,
+      "learning_rate": 9.97244753950034e-06,
+      "loss": 0.5513,
+      "step": 414
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.4106616327947785,
+      "learning_rate": 9.972190879892147e-06,
+      "loss": 0.5703,
+      "step": 415
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.2227033393436075,
+      "learning_rate": 9.971933033721753e-06,
+      "loss": 0.4569,
+      "step": 416
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7364138065420724,
+      "learning_rate": 9.971674001050687e-06,
+      "loss": 0.6168,
+      "step": 417
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.8710080238062123,
+      "learning_rate": 9.971413781940765e-06,
+      "loss": 0.5538,
+      "step": 418
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.318556608870995,
+      "learning_rate": 9.971152376454087e-06,
+      "loss": 0.5485,
+      "step": 419
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.719307350175902,
+      "learning_rate": 9.970889784653034e-06,
+      "loss": 0.5245,
+      "step": 420
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.7330916466222375,
+      "learning_rate": 9.97062600660027e-06,
+      "loss": 0.6126,
+      "step": 421
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.08925692657126,
+      "learning_rate": 9.970361042358744e-06,
+      "loss": 0.5518,
+      "step": 422
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.16970228245275,
+      "learning_rate": 9.970094891991687e-06,
+      "loss": 0.5307,
+      "step": 423
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.8627664718348136,
+      "learning_rate": 9.969827555562613e-06,
+      "loss": 0.4506,
+      "step": 424
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2947397616707408,
+      "learning_rate": 9.969559033135319e-06,
+      "loss": 0.5967,
+      "step": 425
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.6201928864767106,
+      "learning_rate": 9.969289324773883e-06,
+      "loss": 0.5736,
+      "step": 426
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.815248401333858,
+      "learning_rate": 9.969018430542671e-06,
+      "loss": 0.5802,
+      "step": 427
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.004485568750944,
+      "learning_rate": 9.968746350506329e-06,
+      "loss": 0.5775,
+      "step": 428
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.704335190330221,
+      "learning_rate": 9.968473084729784e-06,
+      "loss": 0.5376,
+      "step": 429
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.9861569760139535,
+      "learning_rate": 9.96819863327825e-06,
+      "loss": 0.6122,
+      "step": 430
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.425122906174862,
+      "learning_rate": 9.967922996217221e-06,
+      "loss": 0.5331,
+      "step": 431
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.041352599640688,
+      "learning_rate": 9.967646173612474e-06,
+      "loss": 0.5604,
+      "step": 432
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.676873179811385,
+      "learning_rate": 9.967368165530073e-06,
+      "loss": 0.5242,
+      "step": 433
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5229882995413666,
+      "learning_rate": 9.967088972036359e-06,
+      "loss": 0.5934,
+      "step": 434
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.468033042033904,
+      "learning_rate": 9.966808593197959e-06,
+      "loss": 0.5668,
+      "step": 435
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.660991670922934,
+      "learning_rate": 9.966527029081784e-06,
+      "loss": 0.5903,
+      "step": 436
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.027582230620482,
+      "learning_rate": 9.966244279755025e-06,
+      "loss": 0.5132,
+      "step": 437
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.5369347748644797,
+      "learning_rate": 9.965960345285159e-06,
+      "loss": 0.5666,
+      "step": 438
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.919423928073976,
+      "learning_rate": 9.965675225739942e-06,
+      "loss": 0.4819,
+      "step": 439
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.4651246232001522,
+      "learning_rate": 9.965388921187414e-06,
+      "loss": 0.6452,
+      "step": 440
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.388608256521621,
+      "learning_rate": 9.9651014316959e-06,
+      "loss": 0.5998,
+      "step": 441
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.053078232940991,
+      "learning_rate": 9.964812757334007e-06,
+      "loss": 0.4974,
+      "step": 442
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.123262357535189,
+      "learning_rate": 9.964522898170626e-06,
+      "loss": 0.5628,
+      "step": 443
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.539523724774265,
+      "learning_rate": 9.964231854274927e-06,
+      "loss": 0.5831,
+      "step": 444
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.9495433062629757,
+      "learning_rate": 9.963939625716362e-06,
+      "loss": 0.5828,
+      "step": 445
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.135710610667281,
+      "learning_rate": 9.963646212564671e-06,
+      "loss": 0.6071,
+      "step": 446
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.32678915276538,
+      "learning_rate": 9.963351614889873e-06,
+      "loss": 0.5225,
+      "step": 447
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.612260867483246,
+      "learning_rate": 9.963055832762271e-06,
+      "loss": 0.4865,
+      "step": 448
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3523418741101116,
+      "learning_rate": 9.962758866252452e-06,
+      "loss": 0.5919,
+      "step": 449
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.143943921614427,
+      "learning_rate": 9.962460715431284e-06,
+      "loss": 0.5693,
+      "step": 450
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.128468274509935,
+      "learning_rate": 9.962161380369914e-06,
+      "loss": 0.5716,
+      "step": 451
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.788152077789797,
+      "learning_rate": 9.96186086113978e-06,
+      "loss": 0.5488,
+      "step": 452
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.080940897498937,
+      "learning_rate": 9.961559157812595e-06,
+      "loss": 0.5785,
+      "step": 453
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.472735791366464,
+      "learning_rate": 9.961256270460356e-06,
+      "loss": 0.4859,
+      "step": 454
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.903287430622969,
+      "learning_rate": 9.960952199155347e-06,
+      "loss": 0.5579,
+      "step": 455
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.984449748097196,
+      "learning_rate": 9.960646943970129e-06,
+      "loss": 0.5159,
+      "step": 456
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.966278326217929,
+      "learning_rate": 9.960340504977551e-06,
+      "loss": 0.5584,
+      "step": 457
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.650412649797251,
+      "learning_rate": 9.960032882250738e-06,
+      "loss": 0.5129,
+      "step": 458
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.39622849374572,
+      "learning_rate": 9.959724075863104e-06,
+      "loss": 0.5258,
+      "step": 459
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.82484788426156,
+      "learning_rate": 9.959414085888341e-06,
+      "loss": 0.5558,
+      "step": 460
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.517670970751855,
+      "learning_rate": 9.959102912400424e-06,
+      "loss": 0.5234,
+      "step": 461
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.241218754059866,
+      "learning_rate": 9.958790555473613e-06,
+      "loss": 0.575,
+      "step": 462
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.189469831714092,
+      "learning_rate": 9.95847701518245e-06,
+      "loss": 0.518,
+      "step": 463
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.7829893942251958,
+      "learning_rate": 9.958162291601756e-06,
+      "loss": 0.5785,
+      "step": 464
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 15.14139924867607,
+      "learning_rate": 9.957846384806637e-06,
+      "loss": 0.578,
+      "step": 465
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.371977540454606,
+      "learning_rate": 9.95752929487248e-06,
+      "loss": 0.5256,
+      "step": 466
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.91306727990095,
+      "learning_rate": 9.957211021874956e-06,
+      "loss": 0.5204,
+      "step": 467
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.762184003956692,
+      "learning_rate": 9.956891565890018e-06,
+      "loss": 0.5162,
+      "step": 468
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 37.036080451162356,
+      "learning_rate": 9.956570926993901e-06,
+      "loss": 0.6089,
+      "step": 469
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.176496312557425,
+      "learning_rate": 9.95624910526312e-06,
+      "loss": 0.5389,
+      "step": 470
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.7558850724783435,
+      "learning_rate": 9.95592610077448e-06,
+      "loss": 0.5073,
+      "step": 471
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.022910849061221,
+      "learning_rate": 9.955601913605056e-06,
+      "loss": 0.569,
+      "step": 472
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.875416740436149,
+      "learning_rate": 9.955276543832215e-06,
+      "loss": 0.5834,
+      "step": 473
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.587277400956277,
+      "learning_rate": 9.954949991533604e-06,
+      "loss": 0.5381,
+      "step": 474
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.592422517923324,
+      "learning_rate": 9.95462225678715e-06,
+      "loss": 0.5224,
+      "step": 475
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.5018458506338,
+      "learning_rate": 9.954293339671064e-06,
+      "loss": 0.4911,
+      "step": 476
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.7353185470682,
+      "learning_rate": 9.953963240263837e-06,
+      "loss": 0.5976,
+      "step": 477
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.194882887644598,
+      "learning_rate": 9.953631958644248e-06,
+      "loss": 0.5558,
+      "step": 478
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.27504738362757,
+      "learning_rate": 9.95329949489135e-06,
+      "loss": 0.6246,
+      "step": 479
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.69488845848846,
+      "learning_rate": 9.952965849084483e-06,
+      "loss": 0.5306,
+      "step": 480
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.416052859208481,
+      "learning_rate": 9.95263102130327e-06,
+      "loss": 0.5838,
+      "step": 481
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 21.10202108481693,
+      "learning_rate": 9.95229501162761e-06,
+      "loss": 0.5465,
+      "step": 482
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.238713550155264,
+      "learning_rate": 9.951957820137694e-06,
+      "loss": 0.5924,
+      "step": 483
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.8693431581359015,
+      "learning_rate": 9.951619446913985e-06,
+      "loss": 0.4921,
+      "step": 484
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.442278284934623,
+      "learning_rate": 9.951279892037234e-06,
+      "loss": 0.5607,
+      "step": 485
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.303385560990992,
+      "learning_rate": 9.95093915558847e-06,
+      "loss": 0.4815,
+      "step": 486
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.889007228039267,
+      "learning_rate": 9.95059723764901e-06,
+      "loss": 0.5399,
+      "step": 487
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.484839171670657,
+      "learning_rate": 9.950254138300447e-06,
+      "loss": 0.5257,
+      "step": 488
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.2829726469629525,
+      "learning_rate": 9.949909857624658e-06,
+      "loss": 0.5247,
+      "step": 489
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3916596577356257,
+      "learning_rate": 9.949564395703803e-06,
+      "loss": 0.5987,
+      "step": 490
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.54294780997308,
+      "learning_rate": 9.949217752620322e-06,
+      "loss": 0.4398,
+      "step": 491
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.430483003061391,
+      "learning_rate": 9.94886992845694e-06,
+      "loss": 0.4554,
+      "step": 492
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.17501578410412,
+      "learning_rate": 9.948520923296656e-06,
+      "loss": 0.5612,
+      "step": 493
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.210116717247629,
+      "learning_rate": 9.948170737222763e-06,
+      "loss": 0.4608,
+      "step": 494
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.839233500167328,
+      "learning_rate": 9.947819370318825e-06,
+      "loss": 0.5149,
+      "step": 495
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.312165482182679,
+      "learning_rate": 9.947466822668696e-06,
+      "loss": 0.5531,
+      "step": 496
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.142061200802827,
+      "learning_rate": 9.947113094356505e-06,
+      "loss": 0.5297,
+      "step": 497
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.761292045141527,
+      "learning_rate": 9.946758185466666e-06,
+      "loss": 0.4934,
+      "step": 498
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.403910455524825,
+      "learning_rate": 9.946402096083874e-06,
+      "loss": 0.5157,
+      "step": 499
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.405641677429245,
+      "learning_rate": 9.946044826293106e-06,
+      "loss": 0.5634,
+      "step": 500
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.8194310098991187,
+      "learning_rate": 9.945686376179624e-06,
+      "loss": 0.5445,
+      "step": 501
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 38.425864109082276,
+      "learning_rate": 9.945326745828963e-06,
+      "loss": 0.5513,
+      "step": 502
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.9515881552304775,
+      "learning_rate": 9.94496593532695e-06,
+      "loss": 0.5305,
+      "step": 503
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.493571976371911,
+      "learning_rate": 9.944603944759685e-06,
+      "loss": 0.574,
+      "step": 504
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.82799872516371,
+      "learning_rate": 9.944240774213557e-06,
+      "loss": 0.5904,
+      "step": 505
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.2276612841893595,
+      "learning_rate": 9.94387642377523e-06,
+      "loss": 0.5128,
+      "step": 506
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.702210728139473,
+      "learning_rate": 9.943510893531652e-06,
+      "loss": 0.5199,
+      "step": 507
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.794873692784183,
+      "learning_rate": 9.943144183570058e-06,
+      "loss": 0.5198,
+      "step": 508
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4072298689375455,
+      "learning_rate": 9.942776293977951e-06,
+      "loss": 0.4993,
+      "step": 509
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.332745324166785,
+      "learning_rate": 9.942407224843133e-06,
+      "loss": 0.5545,
+      "step": 510
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.55004954134649,
+      "learning_rate": 9.942036976253672e-06,
+      "loss": 0.5997,
+      "step": 511
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.206188009168057,
+      "learning_rate": 9.94166554829793e-06,
+      "loss": 0.5935,
+      "step": 512
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4444611292914382,
+      "learning_rate": 9.941292941064537e-06,
+      "loss": 0.6056,
+      "step": 513
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.7204819579365225,
+      "learning_rate": 9.94091915464242e-06,
+      "loss": 0.4414,
+      "step": 514
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.402768022663327,
+      "learning_rate": 9.940544189120771e-06,
+      "loss": 0.584,
+      "step": 515
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.35147570407809,
+      "learning_rate": 9.940168044589077e-06,
+      "loss": 0.5615,
+      "step": 516
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.49686249047898,
+      "learning_rate": 9.939790721137102e-06,
+      "loss": 0.5217,
+      "step": 517
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.925563072629076,
+      "learning_rate": 9.939412218854888e-06,
+      "loss": 0.5512,
+      "step": 518
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.9920214507875516,
+      "learning_rate": 9.93903253783276e-06,
+      "loss": 0.5575,
+      "step": 519
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.6918470588438503,
+      "learning_rate": 9.938651678161326e-06,
+      "loss": 0.5752,
+      "step": 520
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.177063291589921,
+      "learning_rate": 9.938269639931475e-06,
+      "loss": 0.4518,
+      "step": 521
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.2388400852759354,
+      "learning_rate": 9.937886423234375e-06,
+      "loss": 0.5069,
+      "step": 522
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.231148903388963,
+      "learning_rate": 9.937502028161477e-06,
+      "loss": 0.565,
+      "step": 523
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.741163756474268,
+      "learning_rate": 9.937116454804514e-06,
+      "loss": 0.5567,
+      "step": 524
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.810736561247086,
+      "learning_rate": 9.936729703255498e-06,
+      "loss": 0.5338,
+      "step": 525
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.9573468748875085,
+      "learning_rate": 9.936341773606723e-06,
+      "loss": 0.5974,
+      "step": 526
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1815857805896552,
+      "learning_rate": 9.935952665950767e-06,
+      "loss": 0.5748,
+      "step": 527
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.2078161301120725,
+      "learning_rate": 9.935562380380484e-06,
+      "loss": 0.6156,
+      "step": 528
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.885381572037831,
+      "learning_rate": 9.935170916989013e-06,
+      "loss": 0.5052,
+      "step": 529
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.010733199988919,
+      "learning_rate": 9.93477827586977e-06,
+      "loss": 0.5554,
+      "step": 530
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.906623802002172,
+      "learning_rate": 9.934384457116457e-06,
+      "loss": 0.5387,
+      "step": 531
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.742285914933382,
+      "learning_rate": 9.933989460823056e-06,
+      "loss": 0.5041,
+      "step": 532
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.579152600869417,
+      "learning_rate": 9.933593287083827e-06,
+      "loss": 0.4479,
+      "step": 533
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.94041457861351,
+      "learning_rate": 9.933195935993312e-06,
+      "loss": 0.4963,
+      "step": 534
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1186838038708187,
+      "learning_rate": 9.932797407646337e-06,
+      "loss": 0.6729,
+      "step": 535
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.756980529013317,
+      "learning_rate": 9.932397702138006e-06,
+      "loss": 0.5748,
+      "step": 536
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.8141388610014735,
+      "learning_rate": 9.931996819563702e-06,
+      "loss": 0.5499,
+      "step": 537
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.927671633192791,
+      "learning_rate": 9.931594760019096e-06,
+      "loss": 0.5137,
+      "step": 538
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.027226445048273,
+      "learning_rate": 9.931191523600133e-06,
+      "loss": 0.5611,
+      "step": 539
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.139606981287645,
+      "learning_rate": 9.930787110403041e-06,
+      "loss": 0.5127,
+      "step": 540
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.8469075994415,
+      "learning_rate": 9.930381520524328e-06,
+      "loss": 0.4482,
+      "step": 541
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.573927185515536,
+      "learning_rate": 9.929974754060787e-06,
+      "loss": 0.5306,
+      "step": 542
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.778324482664255,
+      "learning_rate": 9.929566811109486e-06,
+      "loss": 0.527,
+      "step": 543
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.368202506906533,
+      "learning_rate": 9.929157691767778e-06,
+      "loss": 0.586,
+      "step": 544
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5497527803665485,
+      "learning_rate": 9.928747396133293e-06,
+      "loss": 0.5883,
+      "step": 545
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.601161511377555,
+      "learning_rate": 9.928335924303947e-06,
+      "loss": 0.6103,
+      "step": 546
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.333273134443389,
+      "learning_rate": 9.927923276377931e-06,
+      "loss": 0.5505,
+      "step": 547
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.572097925871523,
+      "learning_rate": 9.927509452453719e-06,
+      "loss": 0.5195,
+      "step": 548
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.007479301700096,
+      "learning_rate": 9.927094452630069e-06,
+      "loss": 0.5682,
+      "step": 549
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.43888721008406,
+      "learning_rate": 9.92667827700601e-06,
+      "loss": 0.5791,
+      "step": 550
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.669373539947307,
+      "learning_rate": 9.926260925680867e-06,
+      "loss": 0.5931,
+      "step": 551
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.35161640969527,
+      "learning_rate": 9.925842398754228e-06,
+      "loss": 0.5498,
+      "step": 552
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.821031768997714,
+      "learning_rate": 9.925422696325976e-06,
+      "loss": 0.4989,
+      "step": 553
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.803093896117229,
+      "learning_rate": 9.925001818496265e-06,
+      "loss": 0.495,
+      "step": 554
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.7372122948306594,
+      "learning_rate": 9.924579765365536e-06,
+      "loss": 0.4642,
+      "step": 555
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.22790687064754,
+      "learning_rate": 9.924156537034506e-06,
+      "loss": 0.5745,
+      "step": 556
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.885752461260323,
+      "learning_rate": 9.923732133604176e-06,
+      "loss": 0.5384,
+      "step": 557
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.185110929736898,
+      "learning_rate": 9.923306555175823e-06,
+      "loss": 0.5681,
+      "step": 558
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.598286716656606,
+      "learning_rate": 9.922879801851006e-06,
+      "loss": 0.5891,
+      "step": 559
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.810599452571767,
+      "learning_rate": 9.92245187373157e-06,
+      "loss": 0.5219,
+      "step": 560
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.724038116973454,
+      "learning_rate": 9.92202277091963e-06,
+      "loss": 0.5384,
+      "step": 561
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.441695506213235,
+      "learning_rate": 9.921592493517594e-06,
+      "loss": 0.5176,
+      "step": 562
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.924128962116078,
+      "learning_rate": 9.921161041628136e-06,
+      "loss": 0.5723,
+      "step": 563
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 12.46739574351716,
+      "learning_rate": 9.92072841535422e-06,
+      "loss": 0.5612,
+      "step": 564
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.427351452341228,
+      "learning_rate": 9.92029461479909e-06,
+      "loss": 0.5887,
+      "step": 565
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.532083896994779,
+      "learning_rate": 9.919859640066267e-06,
+      "loss": 0.5178,
+      "step": 566
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.1371605887183955,
+      "learning_rate": 9.919423491259551e-06,
+      "loss": 0.5823,
+      "step": 567
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.184819561892654,
+      "learning_rate": 9.918986168483027e-06,
+      "loss": 0.5019,
+      "step": 568
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.210768141780824,
+      "learning_rate": 9.918547671841057e-06,
+      "loss": 0.4505,
+      "step": 569
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.9585646183745014,
+      "learning_rate": 9.918108001438283e-06,
+      "loss": 0.524,
+      "step": 570
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.23145460828766,
+      "learning_rate": 9.91766715737963e-06,
+      "loss": 0.5179,
+      "step": 571
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.109189185774883,
+      "learning_rate": 9.917225139770296e-06,
+      "loss": 0.5237,
+      "step": 572
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.53740489178462,
+      "learning_rate": 9.91678194871577e-06,
+      "loss": 0.5456,
+      "step": 573
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.625836491341964,
+      "learning_rate": 9.916337584321811e-06,
+      "loss": 0.5439,
+      "step": 574
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.5869139235426974,
+      "learning_rate": 9.915892046694464e-06,
+      "loss": 0.4693,
+      "step": 575
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.531480559045545,
+      "learning_rate": 9.91544533594005e-06,
+      "loss": 0.4974,
+      "step": 576
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.3007762464713,
+      "learning_rate": 9.914997452165174e-06,
+      "loss": 0.5189,
+      "step": 577
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.544427727020317,
+      "learning_rate": 9.914548395476716e-06,
+      "loss": 0.4669,
+      "step": 578
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8801104744521044,
+      "learning_rate": 9.914098165981842e-06,
+      "loss": 0.5696,
+      "step": 579
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.675778780433523,
+      "learning_rate": 9.913646763787993e-06,
+      "loss": 0.4808,
+      "step": 580
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1657480701779526,
+      "learning_rate": 9.913194189002892e-06,
+      "loss": 0.4516,
+      "step": 581
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.6081283291210084,
+      "learning_rate": 9.912740441734541e-06,
+      "loss": 0.5338,
+      "step": 582
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.745027646474163,
+      "learning_rate": 9.912285522091223e-06,
+      "loss": 0.4911,
+      "step": 583
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.626144645448106,
+      "learning_rate": 9.911829430181496e-06,
+      "loss": 0.6033,
+      "step": 584
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.651174341401112,
+      "learning_rate": 9.911372166114208e-06,
+      "loss": 0.5535,
+      "step": 585
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7779520848639043,
+      "learning_rate": 9.910913729998477e-06,
+      "loss": 0.534,
+      "step": 586
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.423725218344855,
+      "learning_rate": 9.910454121943702e-06,
+      "loss": 0.5701,
+      "step": 587
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5759692673081775,
+      "learning_rate": 9.909993342059567e-06,
+      "loss": 0.5489,
+      "step": 588
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.332349137902474,
+      "learning_rate": 9.909531390456033e-06,
+      "loss": 0.5432,
+      "step": 589
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.766215498944896,
+      "learning_rate": 9.909068267243336e-06,
+      "loss": 0.5496,
+      "step": 590
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.083394180119648,
+      "learning_rate": 9.908603972531998e-06,
+      "loss": 0.4667,
+      "step": 591
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4094821376927027,
+      "learning_rate": 9.908138506432818e-06,
+      "loss": 0.4751,
+      "step": 592
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.266811046766004,
+      "learning_rate": 9.907671869056877e-06,
+      "loss": 0.4794,
+      "step": 593
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.5505237603017274,
+      "learning_rate": 9.907204060515527e-06,
+      "loss": 0.5629,
+      "step": 594
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5817292625362755,
+      "learning_rate": 9.906735080920414e-06,
+      "loss": 0.4803,
+      "step": 595
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.6517680254580345,
+      "learning_rate": 9.906264930383448e-06,
+      "loss": 0.5156,
+      "step": 596
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.343118404434846,
+      "learning_rate": 9.905793609016828e-06,
+      "loss": 0.5109,
+      "step": 597
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.609198127219048,
+      "learning_rate": 9.905321116933031e-06,
+      "loss": 0.4311,
+      "step": 598
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.5812105911319945,
+      "learning_rate": 9.904847454244812e-06,
+      "loss": 0.4916,
+      "step": 599
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.925080604783244,
+      "learning_rate": 9.904372621065206e-06,
+      "loss": 0.5354,
+      "step": 600
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.13565805887717,
+      "learning_rate": 9.903896617507526e-06,
+      "loss": 0.4798,
+      "step": 601
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.125308376097846,
+      "learning_rate": 9.903419443685366e-06,
+      "loss": 0.493,
+      "step": 602
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.084159587359883,
+      "learning_rate": 9.902941099712597e-06,
+      "loss": 0.5659,
+      "step": 603
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.323557550991385,
+      "learning_rate": 9.902461585703375e-06,
+      "loss": 0.4837,
+      "step": 604
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.535405732848617,
+      "learning_rate": 9.901980901772128e-06,
+      "loss": 0.6851,
+      "step": 605
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.262480316198713,
+      "learning_rate": 9.901499048033565e-06,
+      "loss": 0.5067,
+      "step": 606
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.34390922034351,
+      "learning_rate": 9.901016024602678e-06,
+      "loss": 0.4621,
+      "step": 607
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.715226457537914,
+      "learning_rate": 9.900531831594736e-06,
+      "loss": 0.5174,
+      "step": 608
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7075486442286483,
+      "learning_rate": 9.900046469125282e-06,
+      "loss": 0.4409,
+      "step": 609
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.904796834259952,
+      "learning_rate": 9.899559937310148e-06,
+      "loss": 0.5131,
+      "step": 610
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.509641059876668,
+      "learning_rate": 9.89907223626544e-06,
+      "loss": 0.532,
+      "step": 611
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.209679464684351,
+      "learning_rate": 9.898583366107539e-06,
+      "loss": 0.4677,
+      "step": 612
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.933183484830707,
+      "learning_rate": 9.89809332695311e-06,
+      "loss": 0.4909,
+      "step": 613
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.893680000922045,
+      "learning_rate": 9.897602118919098e-06,
+      "loss": 0.5391,
+      "step": 614
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.656170294091131,
+      "learning_rate": 9.897109742122721e-06,
+      "loss": 0.5014,
+      "step": 615
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2500005642874297,
+      "learning_rate": 9.896616196681484e-06,
+      "loss": 0.4587,
+      "step": 616
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.073729219471031,
+      "learning_rate": 9.896121482713164e-06,
+      "loss": 0.5421,
+      "step": 617
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.646425931337879,
+      "learning_rate": 9.895625600335818e-06,
+      "loss": 0.5574,
+      "step": 618
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.770649998881895,
+      "learning_rate": 9.895128549667787e-06,
+      "loss": 0.551,
+      "step": 619
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.40653331681264,
+      "learning_rate": 9.894630330827687e-06,
+      "loss": 0.5449,
+      "step": 620
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.047124561041556,
+      "learning_rate": 9.894130943934407e-06,
+      "loss": 0.5449,
+      "step": 621
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.3979120619217724,
+      "learning_rate": 9.893630389107126e-06,
+      "loss": 0.5053,
+      "step": 622
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.24392975104173,
+      "learning_rate": 9.893128666465294e-06,
+      "loss": 0.5408,
+      "step": 623
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.312126626133832,
+      "learning_rate": 9.892625776128643e-06,
+      "loss": 0.5006,
+      "step": 624
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.624726989020237,
+      "learning_rate": 9.892121718217182e-06,
+      "loss": 0.5051,
+      "step": 625
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5084826780752922,
+      "learning_rate": 9.8916164928512e-06,
+      "loss": 0.5617,
+      "step": 626
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.08802545580933,
+      "learning_rate": 9.891110100151263e-06,
+      "loss": 0.4669,
+      "step": 627
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7991357765738936,
+      "learning_rate": 9.890602540238217e-06,
+      "loss": 0.6139,
+      "step": 628
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.106317324372554,
+      "learning_rate": 9.890093813233186e-06,
+      "loss": 0.5279,
+      "step": 629
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.388096239970439,
+      "learning_rate": 9.88958391925757e-06,
+      "loss": 0.5153,
+      "step": 630
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.6288708344242675,
+      "learning_rate": 9.889072858433055e-06,
+      "loss": 0.5134,
+      "step": 631
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.695461982141519,
+      "learning_rate": 9.888560630881595e-06,
+      "loss": 0.5242,
+      "step": 632
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 30.0131411060209,
+      "learning_rate": 9.888047236725432e-06,
+      "loss": 0.581,
+      "step": 633
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.606268323912727,
+      "learning_rate": 9.88753267608708e-06,
+      "loss": 0.4976,
+      "step": 634
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.353250058895615,
+      "learning_rate": 9.887016949089334e-06,
+      "loss": 0.5704,
+      "step": 635
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.635169200376543,
+      "learning_rate": 9.886500055855268e-06,
+      "loss": 0.5162,
+      "step": 636
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.13072867816422,
+      "learning_rate": 9.885981996508231e-06,
+      "loss": 0.5224,
+      "step": 637
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 16.54032089434867,
+      "learning_rate": 9.885462771171855e-06,
+      "loss": 0.492,
+      "step": 638
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.758542400369512,
+      "learning_rate": 9.884942379970045e-06,
+      "loss": 0.4997,
+      "step": 639
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.292338012969868,
+      "learning_rate": 9.88442082302699e-06,
+      "loss": 0.5088,
+      "step": 640
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.176996531959761,
+      "learning_rate": 9.883898100467154e-06,
+      "loss": 0.508,
+      "step": 641
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.170658827741615,
+      "learning_rate": 9.883374212415275e-06,
+      "loss": 0.5805,
+      "step": 642
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 16.08527631282028,
+      "learning_rate": 9.882849158996379e-06,
+      "loss": 0.5335,
+      "step": 643
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.008707277307974,
+      "learning_rate": 9.882322940335763e-06,
+      "loss": 0.5142,
+      "step": 644
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.623512705310939,
+      "learning_rate": 9.881795556558999e-06,
+      "loss": 0.5146,
+      "step": 645
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.562073534618671,
+      "learning_rate": 9.88126700779195e-06,
+      "loss": 0.4305,
+      "step": 646
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.690223869286257,
+      "learning_rate": 9.880737294160742e-06,
+      "loss": 0.5352,
+      "step": 647
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.053045750281363,
+      "learning_rate": 9.88020641579179e-06,
+      "loss": 0.5143,
+      "step": 648
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.601959324167704,
+      "learning_rate": 9.87967437281178e-06,
+      "loss": 0.601,
+      "step": 649
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.558555880657527,
+      "learning_rate": 9.87914116534768e-06,
+      "loss": 0.5101,
+      "step": 650
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 12.502418823271269,
+      "learning_rate": 9.878606793526734e-06,
+      "loss": 0.5924,
+      "step": 651
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.04391956463187,
+      "learning_rate": 9.878071257476461e-06,
+      "loss": 0.5547,
+      "step": 652
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 14.458689757815653,
+      "learning_rate": 9.877534557324667e-06,
+      "loss": 0.5453,
+      "step": 653
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 20.770908660485713,
+      "learning_rate": 9.876996693199427e-06,
+      "loss": 0.5064,
+      "step": 654
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 18.346453583587707,
+      "learning_rate": 9.876457665229097e-06,
+      "loss": 0.4713,
+      "step": 655
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 18.4140593359916,
+      "learning_rate": 9.875917473542312e-06,
+      "loss": 0.5388,
+      "step": 656
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.173438515812045,
+      "learning_rate": 9.87537611826798e-06,
+      "loss": 0.4746,
+      "step": 657
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.30619313890996,
+      "learning_rate": 9.874833599535291e-06,
+      "loss": 0.5464,
+      "step": 658
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 18.13300877582824,
+      "learning_rate": 9.874289917473714e-06,
+      "loss": 0.5992,
+      "step": 659
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.320432046519848,
+      "learning_rate": 9.87374507221299e-06,
+      "loss": 0.5648,
+      "step": 660
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 13.883790827000775,
+      "learning_rate": 9.873199063883143e-06,
+      "loss": 0.5144,
+      "step": 661
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.45884275378918,
+      "learning_rate": 9.87265189261447e-06,
+      "loss": 0.4594,
+      "step": 662
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 63.491235740310195,
+      "learning_rate": 9.87210355853755e-06,
+      "loss": 0.5031,
+      "step": 663
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.430148460386038,
+      "learning_rate": 9.871554061783237e-06,
+      "loss": 0.5022,
+      "step": 664
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.686431902582455,
+      "learning_rate": 9.871003402482659e-06,
+      "loss": 0.5012,
+      "step": 665
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0250635275888587,
+      "learning_rate": 9.870451580767231e-06,
+      "loss": 0.5774,
+      "step": 666
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.979856779461892,
+      "learning_rate": 9.869898596768636e-06,
+      "loss": 0.5581,
+      "step": 667
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.80957149558432,
+      "learning_rate": 9.86934445061884e-06,
+      "loss": 0.4554,
+      "step": 668
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.988047936466621,
+      "learning_rate": 9.868789142450082e-06,
+      "loss": 0.483,
+      "step": 669
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.979814305779161,
+      "learning_rate": 9.868232672394881e-06,
+      "loss": 0.5401,
+      "step": 670
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9421015399324086,
+      "learning_rate": 9.867675040586035e-06,
+      "loss": 0.6044,
+      "step": 671
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.010016322414344,
+      "learning_rate": 9.867116247156614e-06,
+      "loss": 0.5579,
+      "step": 672
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.781902249603746,
+      "learning_rate": 9.866556292239971e-06,
+      "loss": 0.5244,
+      "step": 673
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 13.891502023559045,
+      "learning_rate": 9.86599517596973e-06,
+      "loss": 0.5633,
+      "step": 674
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 13.743974019919886,
+      "learning_rate": 9.8654328984798e-06,
+      "loss": 0.6186,
+      "step": 675
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.625371747978907,
+      "learning_rate": 9.864869459904363e-06,
+      "loss": 0.5263,
+      "step": 676
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.043557936978345,
+      "learning_rate": 9.86430486037787e-06,
+      "loss": 0.5079,
+      "step": 677
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.116079680258187,
+      "learning_rate": 9.863739100035067e-06,
+      "loss": 0.527,
+      "step": 678
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.481538300642201,
+      "learning_rate": 9.863172179010961e-06,
+      "loss": 0.5342,
+      "step": 679
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.961600898164992,
+      "learning_rate": 9.862604097440844e-06,
+      "loss": 0.5557,
+      "step": 680
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3280503018656313,
+      "learning_rate": 9.86203485546028e-06,
+      "loss": 0.5895,
+      "step": 681
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7210374051949073,
+      "learning_rate": 9.861464453205116e-06,
+      "loss": 0.5743,
+      "step": 682
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.68813229866644,
+      "learning_rate": 9.860892890811471e-06,
+      "loss": 0.534,
+      "step": 683
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.68134022468961,
+      "learning_rate": 9.860320168415743e-06,
+      "loss": 0.557,
+      "step": 684
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.426027011654915,
+      "learning_rate": 9.859746286154607e-06,
+      "loss": 0.5494,
+      "step": 685
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.475755033871641,
+      "learning_rate": 9.859171244165012e-06,
+      "loss": 0.5052,
+      "step": 686
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.58400149547184,
+      "learning_rate": 9.858595042584187e-06,
+      "loss": 0.5089,
+      "step": 687
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.313122634967122,
+      "learning_rate": 9.85801768154964e-06,
+      "loss": 0.4508,
+      "step": 688
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.720476563801436,
+      "learning_rate": 9.857439161199145e-06,
+      "loss": 0.54,
+      "step": 689
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.813813923597374,
+      "learning_rate": 9.856859481670764e-06,
+      "loss": 0.5166,
+      "step": 690
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.684267968911751,
+      "learning_rate": 9.856278643102832e-06,
+      "loss": 0.5784,
+      "step": 691
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.029274110439998,
+      "learning_rate": 9.855696645633961e-06,
+      "loss": 0.5453,
+      "step": 692
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.8839013487100775,
+      "learning_rate": 9.855113489403037e-06,
+      "loss": 0.4918,
+      "step": 693
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.436241031380622,
+      "learning_rate": 9.854529174549223e-06,
+      "loss": 0.4878,
+      "step": 694
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 15.02819330236376,
+      "learning_rate": 9.853943701211963e-06,
+      "loss": 0.4293,
+      "step": 695
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.124831595735831,
+      "learning_rate": 9.853357069530971e-06,
+      "loss": 0.5196,
+      "step": 696
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 10.52263192661969,
+      "learning_rate": 9.852769279646244e-06,
+      "loss": 0.4525,
+      "step": 697
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.930252106173418,
+      "learning_rate": 9.852180331698049e-06,
+      "loss": 0.607,
+      "step": 698
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0712702208275253,
+      "learning_rate": 9.851590225826932e-06,
+      "loss": 0.5198,
+      "step": 699
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.033243356219366,
+      "learning_rate": 9.85099896217372e-06,
+      "loss": 0.5763,
+      "step": 700
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.71111428920958,
+      "learning_rate": 9.850406540879507e-06,
+      "loss": 0.5813,
+      "step": 701
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.7961348610146874,
+      "learning_rate": 9.849812962085671e-06,
+      "loss": 0.5163,
+      "step": 702
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.362292165708552,
+      "learning_rate": 9.849218225933863e-06,
+      "loss": 0.5099,
+      "step": 703
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2801737711113885,
+      "learning_rate": 9.84862233256601e-06,
+      "loss": 0.5309,
+      "step": 704
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.220140805904522,
+      "learning_rate": 9.848025282124318e-06,
+      "loss": 0.5631,
+      "step": 705
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.2775615041512305,
+      "learning_rate": 9.847427074751263e-06,
+      "loss": 0.5167,
+      "step": 706
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.19893490380027,
+      "learning_rate": 9.846827710589603e-06,
+      "loss": 0.5367,
+      "step": 707
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.4200877440363975,
+      "learning_rate": 9.846227189782373e-06,
+      "loss": 0.4694,
+      "step": 708
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.620338343514876,
+      "learning_rate": 9.845625512472878e-06,
+      "loss": 0.6533,
+      "step": 709
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.061990276536358,
+      "learning_rate": 9.845022678804701e-06,
+      "loss": 0.4643,
+      "step": 710
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.207194936014409,
+      "learning_rate": 9.844418688921706e-06,
+      "loss": 0.4875,
+      "step": 711
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.233681285735088,
+      "learning_rate": 9.843813542968026e-06,
+      "loss": 0.5611,
+      "step": 712
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.616369076571687,
+      "learning_rate": 9.84320724108807e-06,
+      "loss": 0.6273,
+      "step": 713
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.8340648055401685,
+      "learning_rate": 9.842599783426535e-06,
+      "loss": 0.5299,
+      "step": 714
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.525619751874274,
+      "learning_rate": 9.841991170128374e-06,
+      "loss": 0.5209,
+      "step": 715
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.314254955355116,
+      "learning_rate": 9.841381401338834e-06,
+      "loss": 0.5193,
+      "step": 716
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.18568175879109,
+      "learning_rate": 9.840770477203425e-06,
+      "loss": 0.5226,
+      "step": 717
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.677135827470922,
+      "learning_rate": 9.840158397867941e-06,
+      "loss": 0.4774,
+      "step": 718
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.753441855106366,
+      "learning_rate": 9.839545163478447e-06,
+      "loss": 0.5458,
+      "step": 719
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.203843810911018,
+      "learning_rate": 9.838930774181285e-06,
+      "loss": 0.4839,
+      "step": 720
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.892317856373333,
+      "learning_rate": 9.838315230123075e-06,
+      "loss": 0.4867,
+      "step": 721
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.31786553547229,
+      "learning_rate": 9.837698531450706e-06,
+      "loss": 0.5101,
+      "step": 722
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.443977045226496,
+      "learning_rate": 9.83708067831135e-06,
+      "loss": 0.5572,
+      "step": 723
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.16533402252394,
+      "learning_rate": 9.83646167085245e-06,
+      "loss": 0.5338,
+      "step": 724
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.887288758018006,
+      "learning_rate": 9.835841509221726e-06,
+      "loss": 0.6768,
+      "step": 725
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.594369636005012,
+      "learning_rate": 9.835220193567174e-06,
+      "loss": 0.5006,
+      "step": 726
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.51957434757884,
+      "learning_rate": 9.834597724037064e-06,
+      "loss": 0.5504,
+      "step": 727
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.454462066223111,
+      "learning_rate": 9.833974100779941e-06,
+      "loss": 0.5405,
+      "step": 728
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.064262643929269,
+      "learning_rate": 9.83334932394463e-06,
+      "loss": 0.4458,
+      "step": 729
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.752775898192587,
+      "learning_rate": 9.832723393680222e-06,
+      "loss": 0.4601,
+      "step": 730
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.936468302399266,
+      "learning_rate": 9.832096310136092e-06,
+      "loss": 0.5311,
+      "step": 731
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.360525655282427,
+      "learning_rate": 9.831468073461887e-06,
+      "loss": 0.4816,
+      "step": 732
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.0176685371297935,
+      "learning_rate": 9.830838683807528e-06,
+      "loss": 0.5417,
+      "step": 733
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.514836400874859,
+      "learning_rate": 9.830208141323213e-06,
+      "loss": 0.5128,
+      "step": 734
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.758510369096224,
+      "learning_rate": 9.829576446159417e-06,
+      "loss": 0.4935,
+      "step": 735
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5013410566723069,
+      "learning_rate": 9.828943598466883e-06,
+      "loss": 0.5248,
+      "step": 736
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.348990979609816,
+      "learning_rate": 9.828309598396637e-06,
+      "loss": 0.516,
+      "step": 737
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 29.129971138645967,
+      "learning_rate": 9.827674446099976e-06,
+      "loss": 0.5207,
+      "step": 738
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.526371152545476,
+      "learning_rate": 9.827038141728468e-06,
+      "loss": 0.5335,
+      "step": 739
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6670033126500097,
+      "learning_rate": 9.826400685433968e-06,
+      "loss": 0.6124,
+      "step": 740
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.373677727252735,
+      "learning_rate": 9.825762077368595e-06,
+      "loss": 0.466,
+      "step": 741
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.00457018880989,
+      "learning_rate": 9.825122317684745e-06,
+      "loss": 0.5213,
+      "step": 742
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.733369309715239,
+      "learning_rate": 9.824481406535091e-06,
+      "loss": 0.5042,
+      "step": 743
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.161045734813862,
+      "learning_rate": 9.823839344072582e-06,
+      "loss": 0.5707,
+      "step": 744
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.77282703489,
+      "learning_rate": 9.823196130450435e-06,
+      "loss": 0.4855,
+      "step": 745
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.2276626171490035,
+      "learning_rate": 9.82255176582215e-06,
+      "loss": 0.471,
+      "step": 746
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.1738098659931495,
+      "learning_rate": 9.821906250341495e-06,
+      "loss": 0.4817,
+      "step": 747
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4556155074130797,
+      "learning_rate": 9.82125958416252e-06,
+      "loss": 0.5673,
+      "step": 748
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.456034104175465,
+      "learning_rate": 9.820611767439544e-06,
+      "loss": 0.574,
+      "step": 749
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.506052462033091,
+      "learning_rate": 9.819962800327156e-06,
+      "loss": 0.5002,
+      "step": 750
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.337344732370875,
+      "learning_rate": 9.819312682980233e-06,
+      "loss": 0.4941,
+      "step": 751
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4983573784444206,
+      "learning_rate": 9.818661415553916e-06,
+      "loss": 0.538,
+      "step": 752
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.616459918764729,
+      "learning_rate": 9.818008998203622e-06,
+      "loss": 0.4665,
+      "step": 753
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.253474267847464,
+      "learning_rate": 9.817355431085046e-06,
+      "loss": 0.5577,
+      "step": 754
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.373331948133087,
+      "learning_rate": 9.816700714354152e-06,
+      "loss": 0.5312,
+      "step": 755
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.112520842056831,
+      "learning_rate": 9.816044848167183e-06,
+      "loss": 0.492,
+      "step": 756
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.15500594420659,
+      "learning_rate": 9.815387832680657e-06,
+      "loss": 0.4916,
+      "step": 757
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.081118709276497,
+      "learning_rate": 9.814729668051359e-06,
+      "loss": 0.5086,
+      "step": 758
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.663504968063285,
+      "learning_rate": 9.814070354436359e-06,
+      "loss": 0.5734,
+      "step": 759
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.387022808134051,
+      "learning_rate": 9.813409891992988e-06,
+      "loss": 0.529,
+      "step": 760
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.252531397331252,
+      "learning_rate": 9.812748280878868e-06,
+      "loss": 0.5126,
+      "step": 761
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.054063325891665,
+      "learning_rate": 9.812085521251878e-06,
+      "loss": 0.6161,
+      "step": 762
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.017717325337717,
+      "learning_rate": 9.811421613270182e-06,
+      "loss": 0.5249,
+      "step": 763
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.291713515871051,
+      "learning_rate": 9.810756557092216e-06,
+      "loss": 0.5483,
+      "step": 764
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.713839365567865,
+      "learning_rate": 9.810090352876684e-06,
+      "loss": 0.5649,
+      "step": 765
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.044709804666683,
+      "learning_rate": 9.809423000782575e-06,
+      "loss": 0.5536,
+      "step": 766
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.171019087009112,
+      "learning_rate": 9.808754500969142e-06,
+      "loss": 0.5182,
+      "step": 767
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 12.221848222519958,
+      "learning_rate": 9.808084853595914e-06,
+      "loss": 0.4465,
+      "step": 768
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9743160236815003,
+      "learning_rate": 9.8074140588227e-06,
+      "loss": 0.6468,
+      "step": 769
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.737601455245831,
+      "learning_rate": 9.806742116809576e-06,
+      "loss": 0.4127,
+      "step": 770
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.004688670159399,
+      "learning_rate": 9.806069027716891e-06,
+      "loss": 0.4685,
+      "step": 771
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.905781085608231,
+      "learning_rate": 9.805394791705276e-06,
+      "loss": 0.5359,
+      "step": 772
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.164173924814828,
+      "learning_rate": 9.80471940893563e-06,
+      "loss": 0.5359,
+      "step": 773
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 12.121106440426251,
+      "learning_rate": 9.804042879569122e-06,
+      "loss": 0.5244,
+      "step": 774
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.412524198999384,
+      "learning_rate": 9.803365203767202e-06,
+      "loss": 0.5314,
+      "step": 775
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 32.74738974475791,
+      "learning_rate": 9.802686381691588e-06,
+      "loss": 0.525,
+      "step": 776
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.456815522964599,
+      "learning_rate": 9.802006413504278e-06,
+      "loss": 0.5605,
+      "step": 777
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 10.31379052838033,
+      "learning_rate": 9.801325299367536e-06,
+      "loss": 0.5107,
+      "step": 778
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3119613034360489,
+      "learning_rate": 9.800643039443903e-06,
+      "loss": 0.629,
+      "step": 779
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.190082028040901,
+      "learning_rate": 9.799959633896194e-06,
+      "loss": 0.4646,
+      "step": 780
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.322629782131484,
+      "learning_rate": 9.799275082887498e-06,
+      "loss": 0.4406,
+      "step": 781
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.719262487190556,
+      "learning_rate": 9.798589386581175e-06,
+      "loss": 0.5009,
+      "step": 782
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 19.88652721381345,
+      "learning_rate": 9.797902545140859e-06,
+      "loss": 0.5211,
+      "step": 783
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.623433813198425,
+      "learning_rate": 9.797214558730457e-06,
+      "loss": 0.5036,
+      "step": 784
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 22.255435855220746,
+      "learning_rate": 9.79652542751415e-06,
+      "loss": 0.4925,
+      "step": 785
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.725349884499909,
+      "learning_rate": 9.795835151656395e-06,
+      "loss": 0.4857,
+      "step": 786
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 30.34059618481862,
+      "learning_rate": 9.795143731321914e-06,
+      "loss": 0.4589,
+      "step": 787
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.393322313239448,
+      "learning_rate": 9.794451166675713e-06,
+      "loss": 0.5366,
+      "step": 788
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.598511781248794,
+      "learning_rate": 9.793757457883062e-06,
+      "loss": 0.5555,
+      "step": 789
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.347974641150477,
+      "learning_rate": 9.79306260510951e-06,
+      "loss": 0.5219,
+      "step": 790
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 16.725451309826198,
+      "learning_rate": 9.792366608520872e-06,
+      "loss": 0.4513,
+      "step": 791
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.870827706252377,
+      "learning_rate": 9.791669468283244e-06,
+      "loss": 0.5367,
+      "step": 792
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.74922967282771,
+      "learning_rate": 9.790971184562991e-06,
+      "loss": 0.5266,
+      "step": 793
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 13.050385361963912,
+      "learning_rate": 9.79027175752675e-06,
+      "loss": 0.5682,
+      "step": 794
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.582984528746119,
+      "learning_rate": 9.789571187341434e-06,
+      "loss": 0.4129,
+      "step": 795
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.747618613602189,
+      "learning_rate": 9.788869474174223e-06,
+      "loss": 0.4877,
+      "step": 796
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 13.38932581880064,
+      "learning_rate": 9.78816661819258e-06,
+      "loss": 0.4044,
+      "step": 797
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.899976226323329,
+      "learning_rate": 9.787462619564228e-06,
+      "loss": 0.4963,
+      "step": 798
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.027969300357336,
+      "learning_rate": 9.786757478457173e-06,
+      "loss": 0.4799,
+      "step": 799
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 20.999007283605376,
+      "learning_rate": 9.786051195039689e-06,
+      "loss": 0.5413,
+      "step": 800
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 22.9481995690256,
+      "learning_rate": 9.785343769480322e-06,
+      "loss": 0.4714,
+      "step": 801
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.584374020889029,
+      "learning_rate": 9.784635201947891e-06,
+      "loss": 0.603,
+      "step": 802
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 22.273286484600998,
+      "learning_rate": 9.783925492611492e-06,
+      "loss": 0.5038,
+      "step": 803
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.476164935957038,
+      "learning_rate": 9.783214641640486e-06,
+      "loss": 0.4806,
+      "step": 804
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.262811763396229,
+      "learning_rate": 9.782502649204513e-06,
+      "loss": 0.5485,
+      "step": 805
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.385920451633879,
+      "learning_rate": 9.78178951547348e-06,
+      "loss": 0.552,
+      "step": 806
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.138994710984688,
+      "learning_rate": 9.781075240617573e-06,
+      "loss": 0.416,
+      "step": 807
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 15.148131841525059,
+      "learning_rate": 9.780359824807241e-06,
+      "loss": 0.4693,
+      "step": 808
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.558825406708767,
+      "learning_rate": 9.779643268213216e-06,
+      "loss": 0.5794,
+      "step": 809
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 13.577652929895075,
+      "learning_rate": 9.778925571006494e-06,
+      "loss": 0.5525,
+      "step": 810
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.145771969980245,
+      "learning_rate": 9.778206733358346e-06,
+      "loss": 0.5393,
+      "step": 811
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 16.462393852171243,
+      "learning_rate": 9.777486755440316e-06,
+      "loss": 0.4696,
+      "step": 812
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 14.643324113786015,
+      "learning_rate": 9.77676563742422e-06,
+      "loss": 0.548,
+      "step": 813
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.601726630725164,
+      "learning_rate": 9.776043379482143e-06,
+      "loss": 0.548,
+      "step": 814
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.6499128862694095,
+      "learning_rate": 9.775319981786446e-06,
+      "loss": 0.4983,
+      "step": 815
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.6881479643495805,
+      "learning_rate": 9.774595444509761e-06,
+      "loss": 0.5441,
+      "step": 816
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.549244892714569,
+      "learning_rate": 9.773869767824989e-06,
+      "loss": 0.4928,
+      "step": 817
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.218888994788077,
+      "learning_rate": 9.773142951905307e-06,
+      "loss": 0.4887,
+      "step": 818
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.348482936814411,
+      "learning_rate": 9.772414996924165e-06,
+      "loss": 0.543,
+      "step": 819
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 15.358053921681456,
+      "learning_rate": 9.771685903055277e-06,
+      "loss": 0.5248,
+      "step": 820
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 12.646699743772382,
+      "learning_rate": 9.770955670472636e-06,
+      "loss": 0.4715,
+      "step": 821
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 10.287930497249548,
+      "learning_rate": 9.770224299350504e-06,
+      "loss": 0.4916,
+      "step": 822
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.787758581611556,
+      "learning_rate": 9.769491789863416e-06,
+      "loss": 0.5444,
+      "step": 823
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.0488161310080715,
+      "learning_rate": 9.768758142186177e-06,
+      "loss": 0.4586,
+      "step": 824
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.766785899769031,
+      "learning_rate": 9.768023356493865e-06,
+      "loss": 0.5112,
+      "step": 825
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 18.672668225849293,
+      "learning_rate": 9.76728743296183e-06,
+      "loss": 0.479,
+      "step": 826
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.61017933385382,
+      "learning_rate": 9.766550371765688e-06,
+      "loss": 0.4546,
+      "step": 827
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.376036842210855,
+      "learning_rate": 9.765812173081337e-06,
+      "loss": 0.544,
+      "step": 828
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 14.256196337499162,
+      "learning_rate": 9.765072837084939e-06,
+      "loss": 0.4662,
+      "step": 829
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.197074234680791,
+      "learning_rate": 9.764332363952927e-06,
+      "loss": 0.5129,
+      "step": 830
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.921441077837376,
+      "learning_rate": 9.763590753862007e-06,
+      "loss": 0.4645,
+      "step": 831
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.133830855777482,
+      "learning_rate": 9.762848006989159e-06,
+      "loss": 0.517,
+      "step": 832
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.994832116651285,
+      "learning_rate": 9.76210412351163e-06,
+      "loss": 0.505,
+      "step": 833
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 27.034126879001196,
+      "learning_rate": 9.761359103606943e-06,
+      "loss": 0.4556,
+      "step": 834
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 12.458462896324026,
+      "learning_rate": 9.760612947452885e-06,
+      "loss": 0.5222,
+      "step": 835
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.892885878951086,
+      "learning_rate": 9.75986565522752e-06,
+      "loss": 0.5285,
+      "step": 836
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 13.191607037361615,
+      "learning_rate": 9.759117227109185e-06,
+      "loss": 0.4993,
+      "step": 837
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.837371776396669,
+      "learning_rate": 9.75836766327648e-06,
+      "loss": 0.549,
+      "step": 838
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0790649656089317,
+      "learning_rate": 9.757616963908283e-06,
+      "loss": 0.6071,
+      "step": 839
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5902221088784447,
+      "learning_rate": 9.756865129183742e-06,
+      "loss": 0.5115,
+      "step": 840
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.221573041381752,
+      "learning_rate": 9.756112159282271e-06,
+      "loss": 0.4789,
+      "step": 841
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.345982771928576,
+      "learning_rate": 9.75535805438356e-06,
+      "loss": 0.5225,
+      "step": 842
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.97226195601666,
+      "learning_rate": 9.754602814667571e-06,
+      "loss": 0.5191,
+      "step": 843
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.719039718718786,
+      "learning_rate": 9.753846440314532e-06,
+      "loss": 0.5494,
+      "step": 844
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.995888783078644,
+      "learning_rate": 9.753088931504943e-06,
+      "loss": 0.5176,
+      "step": 845
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.477839636132996,
+      "learning_rate": 9.752330288419577e-06,
+      "loss": 0.5162,
+      "step": 846
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.937141881908445,
+      "learning_rate": 9.751570511239476e-06,
+      "loss": 0.5366,
+      "step": 847
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.6499242068346485,
+      "learning_rate": 9.750809600145955e-06,
+      "loss": 0.4225,
+      "step": 848
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.470924163046707,
+      "learning_rate": 9.750047555320592e-06,
+      "loss": 0.4744,
+      "step": 849
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 20.729489309591322,
+      "learning_rate": 9.749284376945248e-06,
+      "loss": 0.5235,
+      "step": 850
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.206104445753413,
+      "learning_rate": 9.748520065202045e-06,
+      "loss": 0.549,
+      "step": 851
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.996024416678358,
+      "learning_rate": 9.747754620273375e-06,
+      "loss": 0.5207,
+      "step": 852
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.756509478971665,
+      "learning_rate": 9.746988042341907e-06,
+      "loss": 0.4873,
+      "step": 853
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6667577910229412,
+      "learning_rate": 9.746220331590576e-06,
+      "loss": 0.6032,
+      "step": 854
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.321144216107207,
+      "learning_rate": 9.74545148820259e-06,
+      "loss": 0.5646,
+      "step": 855
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.756525752198923,
+      "learning_rate": 9.744681512361422e-06,
+      "loss": 0.5588,
+      "step": 856
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 15.060069574836112,
+      "learning_rate": 9.743910404250822e-06,
+      "loss": 0.5201,
+      "step": 857
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.79484498077948,
+      "learning_rate": 9.743138164054805e-06,
+      "loss": 0.5617,
+      "step": 858
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.992815571518326,
+      "learning_rate": 9.742364791957662e-06,
+      "loss": 0.5799,
+      "step": 859
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 29.121403884187195,
+      "learning_rate": 9.741590288143943e-06,
+      "loss": 0.5623,
+      "step": 860
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.69987443015057,
+      "learning_rate": 9.740814652798482e-06,
+      "loss": 0.477,
+      "step": 861
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.742826130276134,
+      "learning_rate": 9.740037886106375e-06,
+      "loss": 0.488,
+      "step": 862
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.547782809654253,
+      "learning_rate": 9.739259988252987e-06,
+      "loss": 0.5529,
+      "step": 863
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 13.819987930209724,
+      "learning_rate": 9.738480959423957e-06,
+      "loss": 0.4328,
+      "step": 864
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 24.107409132585683,
+      "learning_rate": 9.737700799805192e-06,
+      "loss": 0.5356,
+      "step": 865
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.393584479374104,
+      "learning_rate": 9.73691950958287e-06,
+      "loss": 0.5123,
+      "step": 866
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 12.584286152000336,
+      "learning_rate": 9.736137088943434e-06,
+      "loss": 0.6059,
+      "step": 867
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.856495424862972,
+      "learning_rate": 9.735353538073607e-06,
+      "loss": 0.4831,
+      "step": 868
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 12.593703950680343,
+      "learning_rate": 9.734568857160369e-06,
+      "loss": 0.5229,
+      "step": 869
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.152324176950547,
+      "learning_rate": 9.733783046390982e-06,
+      "loss": 0.5149,
+      "step": 870
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 13.167941468052158,
+      "learning_rate": 9.732996105952967e-06,
+      "loss": 0.4853,
+      "step": 871
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.086034172593884,
+      "learning_rate": 9.732208036034122e-06,
+      "loss": 0.5591,
+      "step": 872
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 21.855545261396625,
+      "learning_rate": 9.731418836822508e-06,
+      "loss": 0.4302,
+      "step": 873
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5717775569278518,
+      "learning_rate": 9.730628508506463e-06,
+      "loss": 0.6106,
+      "step": 874
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.55489582713876,
+      "learning_rate": 9.729837051274591e-06,
+      "loss": 0.5106,
+      "step": 875
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.580850307203784,
+      "learning_rate": 9.729044465315762e-06,
+      "loss": 0.4854,
+      "step": 876
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.766631456738243,
+      "learning_rate": 9.728250750819121e-06,
+      "loss": 0.5297,
+      "step": 877
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.776418950003466,
+      "learning_rate": 9.72745590797408e-06,
+      "loss": 0.5164,
+      "step": 878
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.545259173516069,
+      "learning_rate": 9.726659936970317e-06,
+      "loss": 0.5216,
+      "step": 879
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 15.778574263862074,
+      "learning_rate": 9.725862837997786e-06,
+      "loss": 0.5039,
+      "step": 880
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.968476918856887,
+      "learning_rate": 9.725064611246704e-06,
+      "loss": 0.5432,
+      "step": 881
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.438577548261287,
+      "learning_rate": 9.724265256907561e-06,
+      "loss": 0.457,
+      "step": 882
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.190698739460536,
+      "learning_rate": 9.723464775171115e-06,
+      "loss": 0.5612,
+      "step": 883
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 20.214417182139748,
+      "learning_rate": 9.72266316622839e-06,
+      "loss": 0.4754,
+      "step": 884
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.852847370402287,
+      "learning_rate": 9.721860430270685e-06,
+      "loss": 0.5359,
+      "step": 885
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.630754206135213,
+      "learning_rate": 9.721056567489564e-06,
+      "loss": 0.4577,
+      "step": 886
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.369630242698918,
+      "learning_rate": 9.720251578076857e-06,
+      "loss": 0.5232,
+      "step": 887
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 16.18115998435004,
+      "learning_rate": 9.719445462224673e-06,
+      "loss": 0.4429,
+      "step": 888
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.448366621332923,
+      "learning_rate": 9.718638220125377e-06,
+      "loss": 0.4862,
+      "step": 889
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 17.207228858508813,
+      "learning_rate": 9.717829851971612e-06,
+      "loss": 0.4875,
+      "step": 890
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 13.88082084059105,
+      "learning_rate": 9.717020357956286e-06,
+      "loss": 0.5104,
+      "step": 891
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.014011088821487,
+      "learning_rate": 9.716209738272578e-06,
+      "loss": 0.4785,
+      "step": 892
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 13.712477947897227,
+      "learning_rate": 9.715397993113934e-06,
+      "loss": 0.5376,
+      "step": 893
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 18.752971357216747,
+      "learning_rate": 9.714585122674065e-06,
+      "loss": 0.4823,
+      "step": 894
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.735169507732005,
+      "learning_rate": 9.713771127146956e-06,
+      "loss": 0.4684,
+      "step": 895
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.96261836252534,
+      "learning_rate": 9.712956006726861e-06,
+      "loss": 0.6214,
+      "step": 896
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 13.43002093592202,
+      "learning_rate": 9.712139761608299e-06,
+      "loss": 0.5111,
+      "step": 897
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 23.889816647359666,
+      "learning_rate": 9.711322391986055e-06,
+      "loss": 0.5114,
+      "step": 898
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 15.662217436735633,
+      "learning_rate": 9.710503898055191e-06,
+      "loss": 0.5261,
+      "step": 899
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 45.3605296309955,
+      "learning_rate": 9.70968428001103e-06,
+      "loss": 0.5732,
+      "step": 900
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 50.852085798689444,
+      "learning_rate": 9.708863538049165e-06,
+      "loss": 0.4737,
+      "step": 901
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 43.715623857255395,
+      "learning_rate": 9.708041672365456e-06,
+      "loss": 0.5511,
+      "step": 902
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 37.60185586232048,
+      "learning_rate": 9.707218683156036e-06,
+      "loss": 0.5121,
+      "step": 903
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 16.816524086331565,
+      "learning_rate": 9.706394570617304e-06,
+      "loss": 0.4939,
+      "step": 904
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 16.210929530145666,
+      "learning_rate": 9.70556933494592e-06,
+      "loss": 0.4645,
+      "step": 905
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 18.689524963072557,
+      "learning_rate": 9.704742976338824e-06,
+      "loss": 0.5059,
+      "step": 906
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.841447352543542,
+      "learning_rate": 9.703915494993215e-06,
+      "loss": 0.5622,
+      "step": 907
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 17.75242783926932,
+      "learning_rate": 9.703086891106564e-06,
+      "loss": 0.5275,
+      "step": 908
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 11.167141033512646,
+      "learning_rate": 9.702257164876607e-06,
+      "loss": 0.5325,
+      "step": 909
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 31.15109101486402,
+      "learning_rate": 9.701426316501353e-06,
+      "loss": 0.467,
+      "step": 910
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.695246263745162,
+      "learning_rate": 9.70059434617907e-06,
+      "loss": 0.5439,
+      "step": 911
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.536012405342726,
+      "learning_rate": 9.699761254108303e-06,
+      "loss": 0.4676,
+      "step": 912
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 21.668794105735902,
+      "learning_rate": 9.698927040487863e-06,
+      "loss": 0.4858,
+      "step": 913
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 12.878296454502214,
+      "learning_rate": 9.698091705516823e-06,
+      "loss": 0.5017,
+      "step": 914
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 24.274165668464406,
+      "learning_rate": 9.697255249394528e-06,
+      "loss": 0.548,
+      "step": 915
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.425196288348717,
+      "learning_rate": 9.696417672320588e-06,
+      "loss": 0.4773,
+      "step": 916
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.772683510988001,
+      "learning_rate": 9.695578974494886e-06,
+      "loss": 0.5596,
+      "step": 917
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.8902602475043535,
+      "learning_rate": 9.694739156117565e-06,
+      "loss": 0.4668,
+      "step": 918
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.582449753055394,
+      "learning_rate": 9.693898217389039e-06,
+      "loss": 0.5162,
+      "step": 919
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.566688566259629,
+      "learning_rate": 9.693056158509992e-06,
+      "loss": 0.5132,
+      "step": 920
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 11.461780628858964,
+      "learning_rate": 9.692212979681373e-06,
+      "loss": 0.5222,
+      "step": 921
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 14.430196224308371,
+      "learning_rate": 9.691368681104395e-06,
+      "loss": 0.4588,
+      "step": 922
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.565735553215955,
+      "learning_rate": 9.690523262980543e-06,
+      "loss": 0.585,
+      "step": 923
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5233216678417496,
+      "learning_rate": 9.689676725511567e-06,
+      "loss": 0.5545,
+      "step": 924
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.075990643732144,
+      "learning_rate": 9.688829068899483e-06,
+      "loss": 0.5755,
+      "step": 925
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.754480999082649,
+      "learning_rate": 9.687980293346578e-06,
+      "loss": 0.4018,
+      "step": 926
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7243256636212985,
+      "learning_rate": 9.687130399055402e-06,
+      "loss": 0.602,
+      "step": 927
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 11.197052395718842,
+      "learning_rate": 9.686279386228774e-06,
+      "loss": 0.575,
+      "step": 928
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 12.57692873688182,
+      "learning_rate": 9.68542725506978e-06,
+      "loss": 0.4828,
+      "step": 929
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.902770721370302,
+      "learning_rate": 9.684574005781772e-06,
+      "loss": 0.5751,
+      "step": 930
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.858692541245226,
+      "learning_rate": 9.683719638568368e-06,
+      "loss": 0.5199,
+      "step": 931
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.473739643251799,
+      "learning_rate": 9.682864153633455e-06,
+      "loss": 0.509,
+      "step": 932
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.181492014016197,
+      "learning_rate": 9.682007551181187e-06,
+      "loss": 0.5463,
+      "step": 933
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 11.600867816040104,
+      "learning_rate": 9.681149831415979e-06,
+      "loss": 0.4749,
+      "step": 934
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 35.214183659181856,
+      "learning_rate": 9.680290994542523e-06,
+      "loss": 0.5401,
+      "step": 935
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 13.442532256555586,
+      "learning_rate": 9.679431040765767e-06,
+      "loss": 0.4456,
+      "step": 936
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.342347588278624,
+      "learning_rate": 9.678569970290931e-06,
+      "loss": 0.4761,
+      "step": 937
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.615485266272987,
+      "learning_rate": 9.677707783323503e-06,
+      "loss": 0.4151,
+      "step": 938
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 25.390426672881006,
+      "learning_rate": 9.676844480069232e-06,
+      "loss": 0.4715,
+      "step": 939
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 15.493109965089273,
+      "learning_rate": 9.675980060734138e-06,
+      "loss": 0.5052,
+      "step": 940
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.100292914432225,
+      "learning_rate": 9.675114525524506e-06,
+      "loss": 0.6824,
+      "step": 941
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.483162995018544,
+      "learning_rate": 9.674247874646885e-06,
+      "loss": 0.4175,
+      "step": 942
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 12.187145777319943,
+      "learning_rate": 9.673380108308095e-06,
+      "loss": 0.5702,
+      "step": 943
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.744519195935023,
+      "learning_rate": 9.672511226715219e-06,
+      "loss": 0.479,
+      "step": 944
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.111245724851894,
+      "learning_rate": 9.671641230075605e-06,
+      "loss": 0.4795,
+      "step": 945
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.99869051160868,
+      "learning_rate": 9.670770118596871e-06,
+      "loss": 0.5166,
+      "step": 946
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.508050497310736,
+      "learning_rate": 9.669897892486896e-06,
+      "loss": 0.4977,
+      "step": 947
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.606441881501707,
+      "learning_rate": 9.66902455195383e-06,
+      "loss": 0.4531,
+      "step": 948
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.376672411208604,
+      "learning_rate": 9.668150097206085e-06,
+      "loss": 0.4468,
+      "step": 949
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 13.894571283434995,
+      "learning_rate": 9.667274528452343e-06,
+      "loss": 0.4533,
+      "step": 950
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.612885542180615,
+      "learning_rate": 9.66639784590155e-06,
+      "loss": 0.4775,
+      "step": 951
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 11.436885072264781,
+      "learning_rate": 9.665520049762913e-06,
+      "loss": 0.5424,
+      "step": 952
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.655806164435728,
+      "learning_rate": 9.664641140245913e-06,
+      "loss": 0.5136,
+      "step": 953
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.848970929784373,
+      "learning_rate": 9.663761117560291e-06,
+      "loss": 0.5345,
+      "step": 954
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.698417495603636,
+      "learning_rate": 9.662879981916054e-06,
+      "loss": 0.5222,
+      "step": 955
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 15.421797293380854,
+      "learning_rate": 9.66199773352348e-06,
+      "loss": 0.5241,
+      "step": 956
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.115822593036338,
+      "learning_rate": 9.661114372593103e-06,
+      "loss": 0.4524,
+      "step": 957
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.571896761568674,
+      "learning_rate": 9.660229899335733e-06,
+      "loss": 0.4617,
+      "step": 958
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.417452189767106,
+      "learning_rate": 9.65934431396244e-06,
+      "loss": 0.4642,
+      "step": 959
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.052499318627627,
+      "learning_rate": 9.658457616684556e-06,
+      "loss": 0.4959,
+      "step": 960
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.975110874813396,
+      "learning_rate": 9.657569807713685e-06,
+      "loss": 0.5237,
+      "step": 961
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.308322530252969,
+      "learning_rate": 9.656680887261693e-06,
+      "loss": 0.549,
+      "step": 962
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.746257772958202,
+      "learning_rate": 9.655790855540711e-06,
+      "loss": 0.5246,
+      "step": 963
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.414572721844497,
+      "learning_rate": 9.654899712763138e-06,
+      "loss": 0.5269,
+      "step": 964
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.740559291017155,
+      "learning_rate": 9.654007459141634e-06,
+      "loss": 0.5955,
+      "step": 965
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.7349335479764045,
+      "learning_rate": 9.653114094889128e-06,
+      "loss": 0.4796,
+      "step": 966
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.17517218862611,
+      "learning_rate": 9.65221962021881e-06,
+      "loss": 0.4851,
+      "step": 967
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.561026067474659,
+      "learning_rate": 9.651324035344138e-06,
+      "loss": 0.4725,
+      "step": 968
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.469554318308493,
+      "learning_rate": 9.650427340478834e-06,
+      "loss": 0.3884,
+      "step": 969
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.654500014453655,
+      "learning_rate": 9.649529535836887e-06,
+      "loss": 0.5929,
+      "step": 970
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.545698394961423,
+      "learning_rate": 9.648630621632547e-06,
+      "loss": 0.5285,
+      "step": 971
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.390090082124935,
+      "learning_rate": 9.64773059808033e-06,
+      "loss": 0.5096,
+      "step": 972
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.671101828562971,
+      "learning_rate": 9.646829465395017e-06,
+      "loss": 0.506,
+      "step": 973
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.651252983049962,
+      "learning_rate": 9.645927223791655e-06,
+      "loss": 0.5248,
+      "step": 974
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 38.001153156132474,
+      "learning_rate": 9.645023873485557e-06,
+      "loss": 0.5017,
+      "step": 975
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.768413452609935,
+      "learning_rate": 9.644119414692294e-06,
+      "loss": 0.4405,
+      "step": 976
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.614448652239094,
+      "learning_rate": 9.64321384762771e-06,
+      "loss": 0.506,
+      "step": 977
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2978786037526,
+      "learning_rate": 9.642307172507907e-06,
+      "loss": 0.5895,
+      "step": 978
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.117243271611956,
+      "learning_rate": 9.641399389549253e-06,
+      "loss": 0.5168,
+      "step": 979
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.378478358941672,
+      "learning_rate": 9.640490498968383e-06,
+      "loss": 0.4806,
+      "step": 980
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 11.386725899629303,
+      "learning_rate": 9.639580500982192e-06,
+      "loss": 0.438,
+      "step": 981
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 13.497164177411026,
+      "learning_rate": 9.638669395807844e-06,
+      "loss": 0.5074,
+      "step": 982
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.950546208800471,
+      "learning_rate": 9.637757183662762e-06,
+      "loss": 0.4606,
+      "step": 983
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.121221602143258,
+      "learning_rate": 9.63684386476464e-06,
+      "loss": 0.5092,
+      "step": 984
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.691420671511301,
+      "learning_rate": 9.635929439331429e-06,
+      "loss": 0.4891,
+      "step": 985
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.732427775191928,
+      "learning_rate": 9.63501390758135e-06,
+      "loss": 0.4732,
+      "step": 986
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 14.880137884713431,
+      "learning_rate": 9.63409726973288e-06,
+      "loss": 0.5272,
+      "step": 987
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 26.117316451470153,
+      "learning_rate": 9.633179526004772e-06,
+      "loss": 0.4863,
+      "step": 988
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.956529232036459,
+      "learning_rate": 9.632260676616032e-06,
+      "loss": 0.4752,
+      "step": 989
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.884891082586968,
+      "learning_rate": 9.631340721785934e-06,
+      "loss": 0.5239,
+      "step": 990
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.772862650165771,
+      "learning_rate": 9.630419661734018e-06,
+      "loss": 0.4499,
+      "step": 991
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3421716897356615,
+      "learning_rate": 9.629497496680083e-06,
+      "loss": 0.6071,
+      "step": 992
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.8740654849553,
+      "learning_rate": 9.628574226844194e-06,
+      "loss": 0.5145,
+      "step": 993
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.454950418961442,
+      "learning_rate": 9.627649852446684e-06,
+      "loss": 0.5368,
+      "step": 994
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.864703255481245,
+      "learning_rate": 9.62672437370814e-06,
+      "loss": 0.4865,
+      "step": 995
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 14.206685489584821,
+      "learning_rate": 9.625797790849422e-06,
+      "loss": 0.5421,
+      "step": 996
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.930551276420024,
+      "learning_rate": 9.624870104091646e-06,
+      "loss": 0.5104,
+      "step": 997
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 18.32397406226489,
+      "learning_rate": 9.6239413136562e-06,
+      "loss": 0.5106,
+      "step": 998
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.391498741158296,
+      "learning_rate": 9.623011419764725e-06,
+      "loss": 0.5157,
+      "step": 999
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 15.5777223798632,
+      "learning_rate": 9.622080422639133e-06,
+      "loss": 0.4748,
+      "step": 1000
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 17.28888564039094,
+      "learning_rate": 9.621148322501597e-06,
+      "loss": 0.4778,
+      "step": 1001
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.073931568560706,
+      "learning_rate": 9.620215119574553e-06,
+      "loss": 0.5489,
+      "step": 1002
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.09411490613066,
+      "learning_rate": 9.619280814080699e-06,
+      "loss": 0.5796,
+      "step": 1003
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.777832177074167,
+      "learning_rate": 9.618345406242999e-06,
+      "loss": 0.5531,
+      "step": 1004
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.2771209435661,
+      "learning_rate": 9.617408896284678e-06,
+      "loss": 0.4164,
+      "step": 1005
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.89916495437376,
+      "learning_rate": 9.616471284429224e-06,
+      "loss": 0.5341,
+      "step": 1006
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.704181983550628,
+      "learning_rate": 9.61553257090039e-06,
+      "loss": 0.5798,
+      "step": 1007
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 11.99159650690447,
+      "learning_rate": 9.614592755922188e-06,
+      "loss": 0.4858,
+      "step": 1008
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.712328416918565,
+      "learning_rate": 9.613651839718896e-06,
+      "loss": 0.4904,
+      "step": 1009
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.167676535860376,
+      "learning_rate": 9.612709822515055e-06,
+      "loss": 0.5351,
+      "step": 1010
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.193673219410297,
+      "learning_rate": 9.611766704535467e-06,
+      "loss": 0.493,
+      "step": 1011
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.396400025521136,
+      "learning_rate": 9.610822486005197e-06,
+      "loss": 0.4715,
+      "step": 1012
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 25.390671045662877,
+      "learning_rate": 9.609877167149573e-06,
+      "loss": 0.4448,
+      "step": 1013
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.643329683236608,
+      "learning_rate": 9.608930748194188e-06,
+      "loss": 0.4433,
+      "step": 1014
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.913269817334127,
+      "learning_rate": 9.607983229364891e-06,
+      "loss": 0.5051,
+      "step": 1015
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4933239854141311,
+      "learning_rate": 9.607034610887801e-06,
+      "loss": 0.5946,
+      "step": 1016
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.439564321220899,
+      "learning_rate": 9.606084892989293e-06,
+      "loss": 0.4593,
+      "step": 1017
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.855778868291502,
+      "learning_rate": 9.605134075896012e-06,
+      "loss": 0.5164,
+      "step": 1018
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.31609909203796,
+      "learning_rate": 9.604182159834856e-06,
+      "loss": 0.5474,
+      "step": 1019
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.8263714222077665,
+      "learning_rate": 9.603229145032994e-06,
+      "loss": 0.4895,
+      "step": 1020
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.09627909077371,
+      "learning_rate": 9.60227503171785e-06,
+      "loss": 0.4989,
+      "step": 1021
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.129461801891572,
+      "learning_rate": 9.601319820117114e-06,
+      "loss": 0.5318,
+      "step": 1022
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.945517291440395,
+      "learning_rate": 9.600363510458738e-06,
+      "loss": 0.565,
+      "step": 1023
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.565288045092654,
+      "learning_rate": 9.599406102970936e-06,
+      "loss": 0.4606,
+      "step": 1024
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.363612497499618,
+      "learning_rate": 9.598447597882181e-06,
+      "loss": 0.471,
+      "step": 1025
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.732012732417542,
+      "learning_rate": 9.597487995421214e-06,
+      "loss": 0.5527,
+      "step": 1026
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.544149608971028,
+      "learning_rate": 9.596527295817032e-06,
+      "loss": 0.5231,
+      "step": 1027
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.294403608037943,
+      "learning_rate": 9.595565499298896e-06,
+      "loss": 0.5598,
+      "step": 1028
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.357992762543305,
+      "learning_rate": 9.59460260609633e-06,
+      "loss": 0.5294,
+      "step": 1029
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.575525972538733,
+      "learning_rate": 9.593638616439118e-06,
+      "loss": 0.553,
+      "step": 1030
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.427907302232283,
+      "learning_rate": 9.592673530557306e-06,
+      "loss": 0.4554,
+      "step": 1031
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.845801017811169,
+      "learning_rate": 9.5917073486812e-06,
+      "loss": 0.4894,
+      "step": 1032
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.095422199586442,
+      "learning_rate": 9.590740071041374e-06,
+      "loss": 0.5325,
+      "step": 1033
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.03481310098649,
+      "learning_rate": 9.589771697868655e-06,
+      "loss": 0.5044,
+      "step": 1034
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.572077976385639,
+      "learning_rate": 9.588802229394136e-06,
+      "loss": 0.4879,
+      "step": 1035
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.273984375834355,
+      "learning_rate": 9.587831665849172e-06,
+      "loss": 0.546,
+      "step": 1036
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.575272755532328,
+      "learning_rate": 9.586860007465378e-06,
+      "loss": 0.5081,
+      "step": 1037
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.0968051920280555,
+      "learning_rate": 9.585887254474629e-06,
+      "loss": 0.4548,
+      "step": 1038
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.801712233242212,
+      "learning_rate": 9.584913407109062e-06,
+      "loss": 0.5557,
+      "step": 1039
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.493174438591285,
+      "learning_rate": 9.583938465601076e-06,
+      "loss": 0.5228,
+      "step": 1040
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.402832559180858,
+      "learning_rate": 9.582962430183332e-06,
+      "loss": 0.4996,
+      "step": 1041
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.938401594696955,
+      "learning_rate": 9.58198530108875e-06,
+      "loss": 0.4577,
+      "step": 1042
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.487981594398974,
+      "learning_rate": 9.581007078550513e-06,
+      "loss": 0.5064,
+      "step": 1043
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.484219086972032,
+      "learning_rate": 9.580027762802062e-06,
+      "loss": 0.473,
+      "step": 1044
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.315061161357885,
+      "learning_rate": 9.5790473540771e-06,
+      "loss": 0.5334,
+      "step": 1045
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.5322895729665005,
+      "learning_rate": 9.578065852609595e-06,
+      "loss": 0.529,
+      "step": 1046
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 13.207571896179806,
+      "learning_rate": 9.577083258633769e-06,
+      "loss": 0.5557,
+      "step": 1047
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.521964515889232,
+      "learning_rate": 9.576099572384109e-06,
+      "loss": 0.5242,
+      "step": 1048
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.448904821309215,
+      "learning_rate": 9.57511479409536e-06,
+      "loss": 0.5384,
+      "step": 1049
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.43064147369242,
+      "learning_rate": 9.574128924002533e-06,
+      "loss": 0.4747,
+      "step": 1050
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 15.261634531459844,
+      "learning_rate": 9.573141962340893e-06,
+      "loss": 0.5531,
+      "step": 1051
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.594404980946875,
+      "learning_rate": 9.572153909345969e-06,
+      "loss": 0.5579,
+      "step": 1052
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2951880160872553,
+      "learning_rate": 9.571164765253548e-06,
+      "loss": 0.5905,
+      "step": 1053
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.967959098582858,
+      "learning_rate": 9.570174530299682e-06,
+      "loss": 0.449,
+      "step": 1054
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.209229361710845,
+      "learning_rate": 9.569183204720677e-06,
+      "loss": 0.5469,
+      "step": 1055
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5991579082735123,
+      "learning_rate": 9.568190788753106e-06,
+      "loss": 0.5855,
+      "step": 1056
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 12.485225456393009,
+      "learning_rate": 9.567197282633797e-06,
+      "loss": 0.5404,
+      "step": 1057
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 21.618086859322997,
+      "learning_rate": 9.566202686599842e-06,
+      "loss": 0.5854,
+      "step": 1058
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.657913890183312,
+      "learning_rate": 9.565207000888587e-06,
+      "loss": 0.482,
+      "step": 1059
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.128946752106357,
+      "learning_rate": 9.564210225737646e-06,
+      "loss": 0.5495,
+      "step": 1060
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.616417458629353,
+      "learning_rate": 9.563212361384889e-06,
+      "loss": 0.4993,
+      "step": 1061
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.669566267554182,
+      "learning_rate": 9.562213408068443e-06,
+      "loss": 0.5414,
+      "step": 1062
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 14.12899816958299,
+      "learning_rate": 9.5612133660267e-06,
+      "loss": 0.52,
+      "step": 1063
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.8725420281941,
+      "learning_rate": 9.560212235498312e-06,
+      "loss": 0.5401,
+      "step": 1064
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 13.348498480165699,
+      "learning_rate": 9.559210016722184e-06,
+      "loss": 0.4909,
+      "step": 1065
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 12.084312049489514,
+      "learning_rate": 9.558206709937487e-06,
+      "loss": 0.5346,
+      "step": 1066
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 20.37280723756399,
+      "learning_rate": 9.55720231538365e-06,
+      "loss": 0.5183,
+      "step": 1067
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 19.11220351762379,
+      "learning_rate": 9.556196833300362e-06,
+      "loss": 0.5342,
+      "step": 1068
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4490967555159884,
+      "learning_rate": 9.555190263927573e-06,
+      "loss": 0.5941,
+      "step": 1069
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.421074464349934,
+      "learning_rate": 9.554182607505484e-06,
+      "loss": 0.4688,
+      "step": 1070
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 22.886920966496817,
+      "learning_rate": 9.553173864274567e-06,
+      "loss": 0.4872,
+      "step": 1071
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.404152740919663,
+      "learning_rate": 9.552164034475547e-06,
+      "loss": 0.5532,
+      "step": 1072
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.000119071858734,
+      "learning_rate": 9.55115311834941e-06,
+      "loss": 0.5006,
+      "step": 1073
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.872974933987493,
+      "learning_rate": 9.5501411161374e-06,
+      "loss": 0.5558,
+      "step": 1074
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 14.205960834080486,
+      "learning_rate": 9.54912802808102e-06,
+      "loss": 0.5421,
+      "step": 1075
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.35192458399684,
+      "learning_rate": 9.548113854422036e-06,
+      "loss": 0.4714,
+      "step": 1076
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.953015845208357,
+      "learning_rate": 9.547098595402466e-06,
+      "loss": 0.4549,
+      "step": 1077
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.749068825555604,
+      "learning_rate": 9.546082251264595e-06,
+      "loss": 0.4494,
+      "step": 1078
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.85196381058159,
+      "learning_rate": 9.545064822250961e-06,
+      "loss": 0.4887,
+      "step": 1079
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.257674441147294,
+      "learning_rate": 9.544046308604365e-06,
+      "loss": 0.53,
+      "step": 1080
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 13.110346898007494,
+      "learning_rate": 9.54302671056786e-06,
+      "loss": 0.5329,
+      "step": 1081
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.55354975885972,
+      "learning_rate": 9.542006028384768e-06,
+      "loss": 0.6032,
+      "step": 1082
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.50876178795718,
+      "learning_rate": 9.540984262298662e-06,
+      "loss": 0.4947,
+      "step": 1083
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 19.730498895592994,
+      "learning_rate": 9.539961412553375e-06,
+      "loss": 0.4252,
+      "step": 1084
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 14.340339860583967,
+      "learning_rate": 9.538937479393002e-06,
+      "loss": 0.4132,
+      "step": 1085
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 16.063053486422145,
+      "learning_rate": 9.53791246306189e-06,
+      "loss": 0.5392,
+      "step": 1086
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 14.06749962551076,
+      "learning_rate": 9.536886363804656e-06,
+      "loss": 0.5039,
+      "step": 1087
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 29.67291797370467,
+      "learning_rate": 9.535859181866159e-06,
+      "loss": 0.5462,
+      "step": 1088
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 15.329685803565182,
+      "learning_rate": 9.534830917491532e-06,
+      "loss": 0.4985,
+      "step": 1089
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 19.075324096842866,
+      "learning_rate": 9.533801570926157e-06,
+      "loss": 0.5013,
+      "step": 1090
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5235331613291045,
+      "learning_rate": 9.532771142415678e-06,
+      "loss": 0.6327,
+      "step": 1091
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.445688998289315,
+      "learning_rate": 9.531739632205995e-06,
+      "loss": 0.5291,
+      "step": 1092
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 12.102119506325907,
+      "learning_rate": 9.530707040543269e-06,
+      "loss": 0.4105,
+      "step": 1093
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 13.881835638067553,
+      "learning_rate": 9.529673367673915e-06,
+      "loss": 0.5223,
+      "step": 1094
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 21.332438633665447,
+      "learning_rate": 9.52863861384461e-06,
+      "loss": 0.5049,
+      "step": 1095
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 15.306241186982106,
+      "learning_rate": 9.527602779302288e-06,
+      "loss": 0.5417,
+      "step": 1096
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 12.880054907370333,
+      "learning_rate": 9.526565864294138e-06,
+      "loss": 0.5006,
+      "step": 1097
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 12.092983785015301,
+      "learning_rate": 9.525527869067611e-06,
+      "loss": 0.542,
+      "step": 1098
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 40.13844163699883,
+      "learning_rate": 9.524488793870412e-06,
+      "loss": 0.507,
+      "step": 1099
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 17.48252409675383,
+      "learning_rate": 9.523448638950509e-06,
+      "loss": 0.5615,
+      "step": 1100
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 17.310083534129436,
+      "learning_rate": 9.52240740455612e-06,
+      "loss": 0.4874,
+      "step": 1101
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.913650280714354,
+      "learning_rate": 9.52136509093573e-06,
+      "loss": 0.512,
+      "step": 1102
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 70.98404672064694,
+      "learning_rate": 9.52032169833807e-06,
+      "loss": 0.5529,
+      "step": 1103
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3362033235559758,
+      "learning_rate": 9.519277227012142e-06,
+      "loss": 0.5562,
+      "step": 1104
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.634363072721465,
+      "learning_rate": 9.518231677207193e-06,
+      "loss": 0.5373,
+      "step": 1105
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.570022377384805,
+      "learning_rate": 9.517185049172734e-06,
+      "loss": 0.5299,
+      "step": 1106
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.484255226655126,
+      "learning_rate": 9.516137343158534e-06,
+      "loss": 0.4519,
+      "step": 1107
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.75876132800781,
+      "learning_rate": 9.515088559414614e-06,
+      "loss": 0.4879,
+      "step": 1108
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.31111043536622,
+      "learning_rate": 9.514038698191257e-06,
+      "loss": 0.564,
+      "step": 1109
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 18.094655990258737,
+      "learning_rate": 9.512987759739004e-06,
+      "loss": 0.5401,
+      "step": 1110
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 27.808618569493124,
+      "learning_rate": 9.511935744308647e-06,
+      "loss": 0.5109,
+      "step": 1111
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.374757585984023,
+      "learning_rate": 9.510882652151239e-06,
+      "loss": 0.494,
+      "step": 1112
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.348347471059297,
+      "learning_rate": 9.50982848351809e-06,
+      "loss": 0.4842,
+      "step": 1113
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 21.78758314573042,
+      "learning_rate": 9.508773238660771e-06,
+      "loss": 0.4963,
+      "step": 1114
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 16.570675836274084,
+      "learning_rate": 9.5077169178311e-06,
+      "loss": 0.5098,
+      "step": 1115
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 12.043897593336926,
+      "learning_rate": 9.506659521281156e-06,
+      "loss": 0.4684,
+      "step": 1116
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 24.500947650440683,
+      "learning_rate": 9.50560104926328e-06,
+      "loss": 0.4908,
+      "step": 1117
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 16.199625921416168,
+      "learning_rate": 9.504541502030064e-06,
+      "loss": 0.4788,
+      "step": 1118
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 11.41063748046671,
+      "learning_rate": 9.503480879834359e-06,
+      "loss": 0.503,
+      "step": 1119
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.888600469787498,
+      "learning_rate": 9.50241918292927e-06,
+      "loss": 0.479,
+      "step": 1120
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.679295591106444,
+      "learning_rate": 9.50135641156816e-06,
+      "loss": 0.5706,
+      "step": 1121
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.950250466484968,
+      "learning_rate": 9.50029256600465e-06,
+      "loss": 0.5513,
+      "step": 1122
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4317481671964705,
+      "learning_rate": 9.499227646492613e-06,
+      "loss": 0.6126,
+      "step": 1123
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.622349336745607,
+      "learning_rate": 9.498161653286185e-06,
+      "loss": 0.4738,
+      "step": 1124
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 16.76586830578393,
+      "learning_rate": 9.49709458663975e-06,
+      "loss": 0.5394,
+      "step": 1125
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 12.214051552427792,
+      "learning_rate": 9.496026446807955e-06,
+      "loss": 0.4995,
+      "step": 1126
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.834554774643155,
+      "learning_rate": 9.4949572340457e-06,
+      "loss": 0.5204,
+      "step": 1127
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3391901161989683,
+      "learning_rate": 9.493886948608145e-06,
+      "loss": 0.5949,
+      "step": 1128
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.965429750362886,
+      "learning_rate": 9.492815590750696e-06,
+      "loss": 0.4818,
+      "step": 1129
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.638768552658153,
+      "learning_rate": 9.491743160729025e-06,
+      "loss": 0.4746,
+      "step": 1130
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 20.998917418809803,
+      "learning_rate": 9.490669658799059e-06,
+      "loss": 0.458,
+      "step": 1131
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.772483885751827,
+      "learning_rate": 9.489595085216974e-06,
+      "loss": 0.5029,
+      "step": 1132
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2976514267690746,
+      "learning_rate": 9.488519440239207e-06,
+      "loss": 0.601,
+      "step": 1133
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2704440715181649,
+      "learning_rate": 9.487442724122451e-06,
+      "loss": 0.548,
+      "step": 1134
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.7014949776266475,
+      "learning_rate": 9.486364937123652e-06,
+      "loss": 0.5297,
+      "step": 1135
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.887373602383863,
+      "learning_rate": 9.485286079500014e-06,
+      "loss": 0.4901,
+      "step": 1136
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.13660300183015,
+      "learning_rate": 9.484206151508993e-06,
+      "loss": 0.5638,
+      "step": 1137
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.384853044396187,
+      "learning_rate": 9.483125153408305e-06,
+      "loss": 0.5568,
+      "step": 1138
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.124788597510183,
+      "learning_rate": 9.482043085455918e-06,
+      "loss": 0.4945,
+      "step": 1139
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.082399764504842,
+      "learning_rate": 9.480959947910056e-06,
+      "loss": 0.4598,
+      "step": 1140
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.430666870642827,
+      "learning_rate": 9.4798757410292e-06,
+      "loss": 0.4738,
+      "step": 1141
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.278841849403527,
+      "learning_rate": 9.478790465072084e-06,
+      "loss": 0.5093,
+      "step": 1142
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.188161386664499,
+      "learning_rate": 9.477704120297698e-06,
+      "loss": 0.5447,
+      "step": 1143
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.78192165695814,
+      "learning_rate": 9.476616706965288e-06,
+      "loss": 0.4969,
+      "step": 1144
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.404634921523575,
+      "learning_rate": 9.47552822533435e-06,
+      "loss": 0.5253,
+      "step": 1145
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.603742876299739,
+      "learning_rate": 9.474438675664644e-06,
+      "loss": 0.439,
+      "step": 1146
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.754442333008274,
+      "learning_rate": 9.473348058216176e-06,
+      "loss": 0.6074,
+      "step": 1147
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4845035373501168,
+      "learning_rate": 9.472256373249214e-06,
+      "loss": 0.5831,
+      "step": 1148
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 14.84216765317522,
+      "learning_rate": 9.471163621024276e-06,
+      "loss": 0.4571,
+      "step": 1149
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.476034281613416,
+      "learning_rate": 9.470069801802134e-06,
+      "loss": 0.4339,
+      "step": 1150
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.158037781109141,
+      "learning_rate": 9.468974915843821e-06,
+      "loss": 0.5052,
+      "step": 1151
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 13.808539573827598,
+      "learning_rate": 9.467878963410617e-06,
+      "loss": 0.4187,
+      "step": 1152
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.732525064777297,
+      "learning_rate": 9.466781944764061e-06,
+      "loss": 0.5091,
+      "step": 1153
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 13.449421927175727,
+      "learning_rate": 9.465683860165944e-06,
+      "loss": 0.4832,
+      "step": 1154
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 12.040813943991704,
+      "learning_rate": 9.464584709878313e-06,
+      "loss": 0.4905,
+      "step": 1155
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.197044034350128,
+      "learning_rate": 9.463484494163471e-06,
+      "loss": 0.5736,
+      "step": 1156
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.1008694563348635,
+      "learning_rate": 9.46238321328397e-06,
+      "loss": 0.5175,
+      "step": 1157
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.381097566024325,
+      "learning_rate": 9.46128086750262e-06,
+      "loss": 0.5122,
+      "step": 1158
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 19.49755797574385,
+      "learning_rate": 9.460177457082487e-06,
+      "loss": 0.4547,
+      "step": 1159
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.487516320315352,
+      "learning_rate": 9.459072982286886e-06,
+      "loss": 0.539,
+      "step": 1160
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.4708071861341665,
+      "learning_rate": 9.45796744337939e-06,
+      "loss": 0.5097,
+      "step": 1161
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.382554656801205,
+      "learning_rate": 9.456860840623823e-06,
+      "loss": 0.4783,
+      "step": 1162
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.415161911444637,
+      "learning_rate": 9.455753174284266e-06,
+      "loss": 0.4847,
+      "step": 1163
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.6552989318957696,
+      "learning_rate": 9.454644444625053e-06,
+      "loss": 0.5033,
+      "step": 1164
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.762539319774012,
+      "learning_rate": 9.453534651910766e-06,
+      "loss": 0.4729,
+      "step": 1165
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.663151663527047,
+      "learning_rate": 9.452423796406251e-06,
+      "loss": 0.5824,
+      "step": 1166
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.496513900587899,
+      "learning_rate": 9.4513118783766e-06,
+      "loss": 0.5202,
+      "step": 1167
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.562112292268546,
+      "learning_rate": 9.450198898087161e-06,
+      "loss": 0.5058,
+      "step": 1168
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 16.00484709337884,
+      "learning_rate": 9.449084855803535e-06,
+      "loss": 0.5562,
+      "step": 1169
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.186457089968793,
+      "learning_rate": 9.447969751791577e-06,
+      "loss": 0.4983,
+      "step": 1170
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.8674554583478824,
+      "learning_rate": 9.446853586317394e-06,
+      "loss": 0.4991,
+      "step": 1171
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.278365093353962,
+      "learning_rate": 9.44573635964735e-06,
+      "loss": 0.4157,
+      "step": 1172
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.504340407671717,
+      "learning_rate": 9.44461807204806e-06,
+      "loss": 0.4415,
+      "step": 1173
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.772964990203814,
+      "learning_rate": 9.443498723786385e-06,
+      "loss": 0.4992,
+      "step": 1174
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.371786659077868,
+      "learning_rate": 9.442378315129456e-06,
+      "loss": 0.4924,
+      "step": 1175
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.772384794343003,
+      "learning_rate": 9.44125684634464e-06,
+      "loss": 0.5338,
+      "step": 1176
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.0027632949778145,
+      "learning_rate": 9.440134317699566e-06,
+      "loss": 0.5139,
+      "step": 1177
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.5320730609913555,
+      "learning_rate": 9.439010729462114e-06,
+      "loss": 0.49,
+      "step": 1178
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 15.528473681485258,
+      "learning_rate": 9.437886081900415e-06,
+      "loss": 0.5454,
+      "step": 1179
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.779623624851191,
+      "learning_rate": 9.436760375282858e-06,
+      "loss": 0.477,
+      "step": 1180
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.465588216824085,
+      "learning_rate": 9.43563360987808e-06,
+      "loss": 0.5556,
+      "step": 1181
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.9901778560404204,
+      "learning_rate": 9.43450578595497e-06,
+      "loss": 0.4973,
+      "step": 1182
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.369657437048644,
+      "learning_rate": 9.433376903782674e-06,
+      "loss": 0.484,
+      "step": 1183
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.573164597079299,
+      "learning_rate": 9.432246963630587e-06,
+      "loss": 0.509,
+      "step": 1184
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.718706189079023,
+      "learning_rate": 9.431115965768358e-06,
+      "loss": 0.5407,
+      "step": 1185
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.587573465877018,
+      "learning_rate": 9.429983910465887e-06,
+      "loss": 0.5279,
+      "step": 1186
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.56803726156193,
+      "learning_rate": 9.42885079799333e-06,
+      "loss": 0.4432,
+      "step": 1187
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.408489394442615,
+      "learning_rate": 9.427716628621089e-06,
+      "loss": 0.488,
+      "step": 1188
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.216768419757177,
+      "learning_rate": 9.426581402619824e-06,
+      "loss": 0.4735,
+      "step": 1189
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.690796080628148,
+      "learning_rate": 9.425445120260445e-06,
+      "loss": 0.5133,
+      "step": 1190
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.500302447006344,
+      "learning_rate": 9.424307781814112e-06,
+      "loss": 0.4458,
+      "step": 1191
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.140438425923382,
+      "learning_rate": 9.423169387552244e-06,
+      "loss": 0.4466,
+      "step": 1192
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 109.35465265953393,
+      "learning_rate": 9.422029937746501e-06,
+      "loss": 0.4934,
+      "step": 1193
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.428846090829191,
+      "learning_rate": 9.420889432668805e-06,
+      "loss": 0.4307,
+      "step": 1194
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.178199227168074,
+      "learning_rate": 9.419747872591325e-06,
+      "loss": 0.5049,
+      "step": 1195
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.580007863227983,
+      "learning_rate": 9.418605257786483e-06,
+      "loss": 0.5324,
+      "step": 1196
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.440081274596023,
+      "learning_rate": 9.417461588526951e-06,
+      "loss": 0.5224,
+      "step": 1197
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.688288925680701,
+      "learning_rate": 9.416316865085652e-06,
+      "loss": 0.5391,
+      "step": 1198
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 26.749340384732633,
+      "learning_rate": 9.415171087735767e-06,
+      "loss": 0.6168,
+      "step": 1199
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.054109476209597,
+      "learning_rate": 9.414024256750723e-06,
+      "loss": 0.4629,
+      "step": 1200
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.746156797597203,
+      "learning_rate": 9.412876372404199e-06,
+      "loss": 0.4176,
+      "step": 1201
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.681696287132947,
+      "learning_rate": 9.411727434970121e-06,
+      "loss": 0.5633,
+      "step": 1202
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.28539852767563,
+      "learning_rate": 9.41057744472268e-06,
+      "loss": 0.4613,
+      "step": 1203
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 13.711917720746209,
+      "learning_rate": 9.409426401936301e-06,
+      "loss": 0.4442,
+      "step": 1204
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.5023204341207625,
+      "learning_rate": 9.408274306885675e-06,
+      "loss": 0.4755,
+      "step": 1205
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.821841677189189,
+      "learning_rate": 9.407121159845731e-06,
+      "loss": 0.591,
+      "step": 1206
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.015376271851194,
+      "learning_rate": 9.405966961091661e-06,
+      "loss": 0.4842,
+      "step": 1207
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.891568675672287,
+      "learning_rate": 9.404811710898901e-06,
+      "loss": 0.4926,
+      "step": 1208
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.398481246278861,
+      "learning_rate": 9.40365540954314e-06,
+      "loss": 0.4738,
+      "step": 1209
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 11.027631774660325,
+      "learning_rate": 9.402498057300318e-06,
+      "loss": 0.4855,
+      "step": 1210
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.146261185798257,
+      "learning_rate": 9.401339654446623e-06,
+      "loss": 0.517,
+      "step": 1211
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.719155777515191,
+      "learning_rate": 9.400180201258498e-06,
+      "loss": 0.4783,
+      "step": 1212
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.847939979981822,
+      "learning_rate": 9.399019698012633e-06,
+      "loss": 0.4748,
+      "step": 1213
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.0608940120737875,
+      "learning_rate": 9.397858144985971e-06,
+      "loss": 0.5065,
+      "step": 1214
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.169037447311178,
+      "learning_rate": 9.396695542455706e-06,
+      "loss": 0.5056,
+      "step": 1215
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.836362224048936,
+      "learning_rate": 9.395531890699276e-06,
+      "loss": 0.4778,
+      "step": 1216
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.8270914611718325,
+      "learning_rate": 9.394367189994382e-06,
+      "loss": 0.4475,
+      "step": 1217
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.930380586962998,
+      "learning_rate": 9.393201440618963e-06,
+      "loss": 0.4487,
+      "step": 1218
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.894721021732552,
+      "learning_rate": 9.392034642851214e-06,
+      "loss": 0.4866,
+      "step": 1219
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.9290797705479696,
+      "learning_rate": 9.390866796969577e-06,
+      "loss": 0.5473,
+      "step": 1220
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.551576807084982,
+      "learning_rate": 9.389697903252753e-06,
+      "loss": 0.5245,
+      "step": 1221
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4580537146312427,
+      "learning_rate": 9.38852796197968e-06,
+      "loss": 0.5506,
+      "step": 1222
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.637019427426143,
+      "learning_rate": 9.387356973429556e-06,
+      "loss": 0.5582,
+      "step": 1223
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 133.17072410620233,
+      "learning_rate": 9.386184937881823e-06,
+      "loss": 0.5031,
+      "step": 1224
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.289333595540723,
+      "learning_rate": 9.385011855616177e-06,
+      "loss": 0.5461,
+      "step": 1225
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1518179433045899,
+      "learning_rate": 9.383837726912562e-06,
+      "loss": 0.5717,
+      "step": 1226
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.415458624396937,
+      "learning_rate": 9.382662552051169e-06,
+      "loss": 0.5198,
+      "step": 1227
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.791004043249624,
+      "learning_rate": 9.381486331312445e-06,
+      "loss": 0.5071,
+      "step": 1228
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.516965112775349,
+      "learning_rate": 9.380309064977082e-06,
+      "loss": 0.5226,
+      "step": 1229
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.983982392867575,
+      "learning_rate": 9.379130753326021e-06,
+      "loss": 0.5467,
+      "step": 1230
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 14.968260400206024,
+      "learning_rate": 9.377951396640455e-06,
+      "loss": 0.5243,
+      "step": 1231
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.951433375725924,
+      "learning_rate": 9.376770995201826e-06,
+      "loss": 0.485,
+      "step": 1232
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.308378916802709,
+      "learning_rate": 9.375589549291824e-06,
+      "loss": 0.4047,
+      "step": 1233
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.387222935729364,
+      "learning_rate": 9.374407059192387e-06,
+      "loss": 0.5099,
+      "step": 1234
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.39090304980783,
+      "learning_rate": 9.373223525185708e-06,
+      "loss": 0.4994,
+      "step": 1235
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.792575889175409,
+      "learning_rate": 9.372038947554223e-06,
+      "loss": 0.491,
+      "step": 1236
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.890886024495494,
+      "learning_rate": 9.370853326580619e-06,
+      "loss": 0.5356,
+      "step": 1237
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.544413370834462,
+      "learning_rate": 9.369666662547833e-06,
+      "loss": 0.537,
+      "step": 1238
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.944888106340925,
+      "learning_rate": 9.368478955739048e-06,
+      "loss": 0.5212,
+      "step": 1239
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.611748915546245,
+      "learning_rate": 9.367290206437703e-06,
+      "loss": 0.5074,
+      "step": 1240
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.066969104008949,
+      "learning_rate": 9.366100414927477e-06,
+      "loss": 0.4506,
+      "step": 1241
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.358228570081846,
+      "learning_rate": 9.364909581492301e-06,
+      "loss": 0.4557,
+      "step": 1242
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.596153540626799,
+      "learning_rate": 9.363717706416355e-06,
+      "loss": 0.5476,
+      "step": 1243
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.11737375065248,
+      "learning_rate": 9.362524789984072e-06,
+      "loss": 0.496,
+      "step": 1244
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.601918223107908,
+      "learning_rate": 9.361330832480125e-06,
+      "loss": 0.492,
+      "step": 1245
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.880687324489626,
+      "learning_rate": 9.36013583418944e-06,
+      "loss": 0.5879,
+      "step": 1246
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.901975176886403,
+      "learning_rate": 9.358939795397193e-06,
+      "loss": 0.5006,
+      "step": 1247
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 19.719699031692258,
+      "learning_rate": 9.357742716388806e-06,
+      "loss": 0.5052,
+      "step": 1248
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.8108688984226795,
+      "learning_rate": 9.356544597449947e-06,
+      "loss": 0.4733,
+      "step": 1249
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.044041620409056,
+      "learning_rate": 9.355345438866538e-06,
+      "loss": 0.5132,
+      "step": 1250
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.631959109480548,
+      "learning_rate": 9.354145240924746e-06,
+      "loss": 0.4502,
+      "step": 1251
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.402779074828729,
+      "learning_rate": 9.352944003910982e-06,
+      "loss": 0.5161,
+      "step": 1252
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.23455391436101,
+      "learning_rate": 9.351741728111911e-06,
+      "loss": 0.5736,
+      "step": 1253
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 14.325165586992664,
+      "learning_rate": 9.350538413814447e-06,
+      "loss": 0.4566,
+      "step": 1254
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.618766106100697,
+      "learning_rate": 9.349334061305744e-06,
+      "loss": 0.479,
+      "step": 1255
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.136575730529583,
+      "learning_rate": 9.34812867087321e-06,
+      "loss": 0.5113,
+      "step": 1256
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.318632817107247,
+      "learning_rate": 9.346922242804499e-06,
+      "loss": 0.46,
+      "step": 1257
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 27.942137813601036,
+      "learning_rate": 9.345714777387512e-06,
+      "loss": 0.4353,
+      "step": 1258
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.889311760078206,
+      "learning_rate": 9.3445062749104e-06,
+      "loss": 0.5285,
+      "step": 1259
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.633695497243732,
+      "learning_rate": 9.343296735661559e-06,
+      "loss": 0.5377,
+      "step": 1260
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.353230421451321,
+      "learning_rate": 9.342086159929629e-06,
+      "loss": 0.5278,
+      "step": 1261
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.570593972695121,
+      "learning_rate": 9.34087454800351e-06,
+      "loss": 0.5418,
+      "step": 1262
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.885950918833172,
+      "learning_rate": 9.339661900172331e-06,
+      "loss": 0.4488,
+      "step": 1263
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.087655818071346,
+      "learning_rate": 9.338448216725487e-06,
+      "loss": 0.5166,
+      "step": 1264
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.002019785944465,
+      "learning_rate": 9.337233497952605e-06,
+      "loss": 0.5122,
+      "step": 1265
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.347678680075843,
+      "learning_rate": 9.336017744143567e-06,
+      "loss": 0.4832,
+      "step": 1266
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.315419658681437,
+      "learning_rate": 9.3348009555885e-06,
+      "loss": 0.5239,
+      "step": 1267
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.1751391191781915,
+      "learning_rate": 9.333583132577777e-06,
+      "loss": 0.4924,
+      "step": 1268
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 13.288391393508343,
+      "learning_rate": 9.332364275402021e-06,
+      "loss": 0.5311,
+      "step": 1269
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 14.379948110651485,
+      "learning_rate": 9.3311443843521e-06,
+      "loss": 0.4869,
+      "step": 1270
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.008173431188924,
+      "learning_rate": 9.329923459719126e-06,
+      "loss": 0.5492,
+      "step": 1271
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.77186653032091,
+      "learning_rate": 9.328701501794461e-06,
+      "loss": 0.4814,
+      "step": 1272
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.506693441736103,
+      "learning_rate": 9.327478510869714e-06,
+      "loss": 0.5048,
+      "step": 1273
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.553521406030673,
+      "learning_rate": 9.32625448723674e-06,
+      "loss": 0.5693,
+      "step": 1274
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 12.25930794171845,
+      "learning_rate": 9.325029431187635e-06,
+      "loss": 0.5286,
+      "step": 1275
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.648351646028999,
+      "learning_rate": 9.323803343014749e-06,
+      "loss": 0.5684,
+      "step": 1276
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.9545806803530255,
+      "learning_rate": 9.322576223010677e-06,
+      "loss": 0.5414,
+      "step": 1277
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.825722571190724,
+      "learning_rate": 9.321348071468255e-06,
+      "loss": 0.5144,
+      "step": 1278
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.3784505295665435,
+      "learning_rate": 9.320118888680571e-06,
+      "loss": 0.4771,
+      "step": 1279
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.5405349256888465,
+      "learning_rate": 9.318888674940959e-06,
+      "loss": 0.4757,
+      "step": 1280
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.342385898634505,
+      "learning_rate": 9.31765743054299e-06,
+      "loss": 0.5489,
+      "step": 1281
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.415000434486711,
+      "learning_rate": 9.316425155780495e-06,
+      "loss": 0.5,
+      "step": 1282
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.156838461066658,
+      "learning_rate": 9.315191850947541e-06,
+      "loss": 0.4618,
+      "step": 1283
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.825326923772595,
+      "learning_rate": 9.31395751633844e-06,
+      "loss": 0.4166,
+      "step": 1284
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.87659249356095,
+      "learning_rate": 9.31272215224776e-06,
+      "loss": 0.4669,
+      "step": 1285
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.922852404300426,
+      "learning_rate": 9.311485758970302e-06,
+      "loss": 0.5317,
+      "step": 1286
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4768638154291598,
+      "learning_rate": 9.310248336801122e-06,
+      "loss": 0.648,
+      "step": 1287
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 11.38832393080718,
+      "learning_rate": 9.309009886035518e-06,
+      "loss": 0.5366,
+      "step": 1288
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.239412145060281,
+      "learning_rate": 9.307770406969032e-06,
+      "loss": 0.5133,
+      "step": 1289
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 15.805255782105718,
+      "learning_rate": 9.306529899897452e-06,
+      "loss": 0.4593,
+      "step": 1290
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.796085656656423,
+      "learning_rate": 9.305288365116816e-06,
+      "loss": 0.4784,
+      "step": 1291
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.274255387475334,
+      "learning_rate": 9.3040458029234e-06,
+      "loss": 0.5172,
+      "step": 1292
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.798273524576829,
+      "learning_rate": 9.302802213613729e-06,
+      "loss": 0.4621,
+      "step": 1293
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.288690697391453,
+      "learning_rate": 9.301557597484576e-06,
+      "loss": 0.4842,
+      "step": 1294
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.8770426847690396,
+      "learning_rate": 9.300311954832952e-06,
+      "loss": 0.5147,
+      "step": 1295
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.047214225580251,
+      "learning_rate": 9.299065285956119e-06,
+      "loss": 0.4828,
+      "step": 1296
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.3005555788526566,
+      "learning_rate": 9.297817591151581e-06,
+      "loss": 0.4923,
+      "step": 1297
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.4169056299224767,
+      "learning_rate": 9.296568870717087e-06,
+      "loss": 0.5853,
+      "step": 1298
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.253705735502439,
+      "learning_rate": 9.295319124950632e-06,
+      "loss": 0.5037,
+      "step": 1299
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.209245106970996,
+      "learning_rate": 9.294068354150456e-06,
+      "loss": 0.4758,
+      "step": 1300
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.179265305096352,
+      "learning_rate": 9.292816558615041e-06,
+      "loss": 0.5894,
+      "step": 1301
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.132730942693958,
+      "learning_rate": 9.291563738643116e-06,
+      "loss": 0.5405,
+      "step": 1302
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.909888067177052,
+      "learning_rate": 9.290309894533653e-06,
+      "loss": 0.4801,
+      "step": 1303
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.6696918608622395,
+      "learning_rate": 9.289055026585868e-06,
+      "loss": 0.5368,
+      "step": 1304
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.910752022917877,
+      "learning_rate": 9.287799135099224e-06,
+      "loss": 0.5216,
+      "step": 1305
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.591670736943786,
+      "learning_rate": 9.286542220373427e-06,
+      "loss": 0.4645,
+      "step": 1306
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.001286122705727,
+      "learning_rate": 9.285284282708424e-06,
+      "loss": 0.4849,
+      "step": 1307
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.457533753209876,
+      "learning_rate": 9.284025322404412e-06,
+      "loss": 0.5403,
+      "step": 1308
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.59171730433206,
+      "learning_rate": 9.282765339761828e-06,
+      "loss": 0.4629,
+      "step": 1309
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.260917557912551,
+      "learning_rate": 9.281504335081353e-06,
+      "loss": 0.5182,
+      "step": 1310
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.035950094463464,
+      "learning_rate": 9.280242308663913e-06,
+      "loss": 0.5026,
+      "step": 1311
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 23.57997518406638,
+      "learning_rate": 9.278979260810678e-06,
+      "loss": 0.5071,
+      "step": 1312
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.805392352865697,
+      "learning_rate": 9.277715191823063e-06,
+      "loss": 0.5124,
+      "step": 1313
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.094834601072279,
+      "learning_rate": 9.276450102002723e-06,
+      "loss": 0.469,
+      "step": 1314
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.2480198835074745,
+      "learning_rate": 9.275183991651557e-06,
+      "loss": 0.5067,
+      "step": 1315
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.680331829212345,
+      "learning_rate": 9.273916861071715e-06,
+      "loss": 0.477,
+      "step": 1316
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.814176784875257,
+      "learning_rate": 9.272648710565576e-06,
+      "loss": 0.4616,
+      "step": 1317
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.846380753115313,
+      "learning_rate": 9.27137954043578e-06,
+      "loss": 0.4458,
+      "step": 1318
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.2327803151047485,
+      "learning_rate": 9.270109350985197e-06,
+      "loss": 0.4799,
+      "step": 1319
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.059192562362302,
+      "learning_rate": 9.268838142516943e-06,
+      "loss": 0.4759,
+      "step": 1320
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.171157650721041,
+      "learning_rate": 9.267565915334382e-06,
+      "loss": 0.4414,
+      "step": 1321
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.583272495093906,
+      "learning_rate": 9.266292669741118e-06,
+      "loss": 0.4254,
+      "step": 1322
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.389893165740359,
+      "learning_rate": 9.265018406040996e-06,
+      "loss": 0.5756,
+      "step": 1323
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.575761734631283,
+      "learning_rate": 9.263743124538107e-06,
+      "loss": 0.4619,
+      "step": 1324
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.25550130146345,
+      "learning_rate": 9.262466825536783e-06,
+      "loss": 0.5331,
+      "step": 1325
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.5416439321992543,
+      "learning_rate": 9.2611895093416e-06,
+      "loss": 0.6031,
+      "step": 1326
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 432.8468035549292,
+      "learning_rate": 9.259911176257376e-06,
+      "loss": 0.5404,
+      "step": 1327
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.9738607073254535,
+      "learning_rate": 9.258631826589175e-06,
+      "loss": 0.4947,
+      "step": 1328
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.808425389380071,
+      "learning_rate": 9.257351460642295e-06,
+      "loss": 0.4982,
+      "step": 1329
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.206610541934726,
+      "learning_rate": 9.256070078722287e-06,
+      "loss": 0.4533,
+      "step": 1330
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 19.938371406381457,
+      "learning_rate": 9.254787681134939e-06,
+      "loss": 0.5684,
+      "step": 1331
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.969815014557337,
+      "learning_rate": 9.253504268186279e-06,
+      "loss": 0.4971,
+      "step": 1332
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.817235283433717,
+      "learning_rate": 9.252219840182583e-06,
+      "loss": 0.5319,
+      "step": 1333
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.009054167956684,
+      "learning_rate": 9.250934397430367e-06,
+      "loss": 0.5103,
+      "step": 1334
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.111311389869556,
+      "learning_rate": 9.249647940236386e-06,
+      "loss": 0.4554,
+      "step": 1335
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.560099183024048,
+      "learning_rate": 9.248360468907643e-06,
+      "loss": 0.4995,
+      "step": 1336
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.3007882277831535,
+      "learning_rate": 9.247071983751377e-06,
+      "loss": 0.512,
+      "step": 1337
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.3980511252078824,
+      "learning_rate": 9.245782485075072e-06,
+      "loss": 0.5961,
+      "step": 1338
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.210564132326873,
+      "learning_rate": 9.244491973186455e-06,
+      "loss": 0.4883,
+      "step": 1339
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.877034351394793,
+      "learning_rate": 9.243200448393492e-06,
+      "loss": 0.5011,
+      "step": 1340
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.618686010126334,
+      "learning_rate": 9.241907911004394e-06,
+      "loss": 0.4622,
+      "step": 1341
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.134539077221495,
+      "learning_rate": 9.24061436132761e-06,
+      "loss": 0.5205,
+      "step": 1342
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.978256327467286,
+      "learning_rate": 9.239319799671834e-06,
+      "loss": 0.5275,
+      "step": 1343
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.88898782930768,
+      "learning_rate": 9.238024226345998e-06,
+      "loss": 0.4374,
+      "step": 1344
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.630642165713237,
+      "learning_rate": 9.236727641659277e-06,
+      "loss": 0.433,
+      "step": 1345
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.166345229500069,
+      "learning_rate": 9.23543004592109e-06,
+      "loss": 0.4965,
+      "step": 1346
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.413745135294686,
+      "learning_rate": 9.234131439441095e-06,
+      "loss": 0.4838,
+      "step": 1347
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.133048250121696,
+      "learning_rate": 9.232831822529188e-06,
+      "loss": 0.5104,
+      "step": 1348
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.409959052970513,
+      "learning_rate": 9.231531195495509e-06,
+      "loss": 0.5096,
+      "step": 1349
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.184602726058528,
+      "learning_rate": 9.230229558650443e-06,
+      "loss": 0.5189,
+      "step": 1350
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.878181493119966,
+      "learning_rate": 9.22892691230461e-06,
+      "loss": 0.4721,
+      "step": 1351
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4103836403338035,
+      "learning_rate": 9.227623256768872e-06,
+      "loss": 0.5871,
+      "step": 1352
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.842425968646203,
+      "learning_rate": 9.226318592354336e-06,
+      "loss": 0.4418,
+      "step": 1353
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.684193545252756,
+      "learning_rate": 9.225012919372344e-06,
+      "loss": 0.4831,
+      "step": 1354
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.97009100245864,
+      "learning_rate": 9.223706238134485e-06,
+      "loss": 0.5447,
+      "step": 1355
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.009236065615713,
+      "learning_rate": 9.22239854895258e-06,
+      "loss": 0.5551,
+      "step": 1356
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 12.52474315220209,
+      "learning_rate": 9.2210898521387e-06,
+      "loss": 0.5544,
+      "step": 1357
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.703976585154395,
+      "learning_rate": 9.21978014800515e-06,
+      "loss": 0.4802,
+      "step": 1358
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.407600571493279,
+      "learning_rate": 9.218469436864476e-06,
+      "loss": 0.4556,
+      "step": 1359
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.450928062456577,
+      "learning_rate": 9.21715771902947e-06,
+      "loss": 0.4431,
+      "step": 1360
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 12.78456757442652,
+      "learning_rate": 9.215844994813155e-06,
+      "loss": 0.4795,
+      "step": 1361
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.175992059230888,
+      "learning_rate": 9.2145312645288e-06,
+      "loss": 0.5194,
+      "step": 1362
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4169141164909331,
+      "learning_rate": 9.21321652848992e-06,
+      "loss": 0.5349,
+      "step": 1363
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.635598628320136,
+      "learning_rate": 9.211900787010254e-06,
+      "loss": 0.5161,
+      "step": 1364
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.548005769959011,
+      "learning_rate": 9.210584040403794e-06,
+      "loss": 0.4893,
+      "step": 1365
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.08315589592544,
+      "learning_rate": 9.209266288984769e-06,
+      "loss": 0.4958,
+      "step": 1366
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.442191153430398,
+      "learning_rate": 9.207947533067647e-06,
+      "loss": 0.5006,
+      "step": 1367
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.585391293094771,
+      "learning_rate": 9.20662777296713e-06,
+      "loss": 0.5233,
+      "step": 1368
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.204065675044848,
+      "learning_rate": 9.205307008998176e-06,
+      "loss": 0.4624,
+      "step": 1369
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2806049031540891,
+      "learning_rate": 9.203985241475962e-06,
+      "loss": 0.595,
+      "step": 1370
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.352622672113265,
+      "learning_rate": 9.202662470715918e-06,
+      "loss": 0.5225,
+      "step": 1371
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.334240719838116,
+      "learning_rate": 9.201338697033709e-06,
+      "loss": 0.5094,
+      "step": 1372
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.632978657724268,
+      "learning_rate": 9.200013920745241e-06,
+      "loss": 0.5751,
+      "step": 1373
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.489048249694532,
+      "learning_rate": 9.198688142166656e-06,
+      "loss": 0.5159,
+      "step": 1374
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.565864612282921,
+      "learning_rate": 9.19736136161434e-06,
+      "loss": 0.5048,
+      "step": 1375
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.3780326025128735,
+      "learning_rate": 9.196033579404914e-06,
+      "loss": 0.4955,
+      "step": 1376
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.028199694992104,
+      "learning_rate": 9.19470479585524e-06,
+      "loss": 0.5177,
+      "step": 1377
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 42.70731937340343,
+      "learning_rate": 9.193375011282419e-06,
+      "loss": 0.4977,
+      "step": 1378
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.283314069450698,
+      "learning_rate": 9.19204422600379e-06,
+      "loss": 0.472,
+      "step": 1379
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.118304235174843,
+      "learning_rate": 9.190712440336928e-06,
+      "loss": 0.5267,
+      "step": 1380
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.789021112676851,
+      "learning_rate": 9.189379654599657e-06,
+      "loss": 0.4493,
+      "step": 1381
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 46.48333479375923,
+      "learning_rate": 9.188045869110026e-06,
+      "loss": 0.5548,
+      "step": 1382
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.390426175962068,
+      "learning_rate": 9.186711084186333e-06,
+      "loss": 0.5132,
+      "step": 1383
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 32.556600225291454,
+      "learning_rate": 9.185375300147111e-06,
+      "loss": 0.5115,
+      "step": 1384
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 135.4053692137494,
+      "learning_rate": 9.184038517311127e-06,
+      "loss": 0.4669,
+      "step": 1385
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.909026970033961,
+      "learning_rate": 9.182700735997395e-06,
+      "loss": 0.5414,
+      "step": 1386
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.70572219305469,
+      "learning_rate": 9.181361956525159e-06,
+      "loss": 0.483,
+      "step": 1387
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.65336434885517,
+      "learning_rate": 9.180022179213908e-06,
+      "loss": 0.5283,
+      "step": 1388
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 13.297981626923628,
+      "learning_rate": 9.178681404383365e-06,
+      "loss": 0.5025,
+      "step": 1389
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.384550788634928,
+      "learning_rate": 9.177339632353493e-06,
+      "loss": 0.4998,
+      "step": 1390
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.64499420238818,
+      "learning_rate": 9.175996863444489e-06,
+      "loss": 0.5028,
+      "step": 1391
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.943869543564151,
+      "learning_rate": 9.174653097976794e-06,
+      "loss": 0.5442,
+      "step": 1392
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.121333353354556,
+      "learning_rate": 9.173308336271083e-06,
+      "loss": 0.4628,
+      "step": 1393
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 23.292822590467964,
+      "learning_rate": 9.171962578648268e-06,
+      "loss": 0.5015,
+      "step": 1394
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.9757121130550495,
+      "learning_rate": 9.170615825429503e-06,
+      "loss": 0.4969,
+      "step": 1395
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 17.249292714427384,
+      "learning_rate": 9.169268076936175e-06,
+      "loss": 0.4712,
+      "step": 1396
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2734440972042236,
+      "learning_rate": 9.167919333489912e-06,
+      "loss": 0.6234,
+      "step": 1397
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.782569494948452,
+      "learning_rate": 9.166569595412576e-06,
+      "loss": 0.404,
+      "step": 1398
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4479841334576389,
+      "learning_rate": 9.165218863026268e-06,
+      "loss": 0.5553,
+      "step": 1399
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.92983083804337,
+      "learning_rate": 9.163867136653327e-06,
+      "loss": 0.5269,
+      "step": 1400
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.155494255078668,
+      "learning_rate": 9.162514416616329e-06,
+      "loss": 0.4839,
+      "step": 1401
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 12.35721953235503,
+      "learning_rate": 9.161160703238085e-06,
+      "loss": 0.4795,
+      "step": 1402
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 12.941667281890753,
+      "learning_rate": 9.15980599684165e-06,
+      "loss": 0.4851,
+      "step": 1403
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 14.503102980911349,
+      "learning_rate": 9.158450297750303e-06,
+      "loss": 0.4826,
+      "step": 1404
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.211804589954603,
+      "learning_rate": 9.157093606287573e-06,
+      "loss": 0.4227,
+      "step": 1405
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 27.628510822181873,
+      "learning_rate": 9.155735922777219e-06,
+      "loss": 0.4102,
+      "step": 1406
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 17.349167792670606,
+      "learning_rate": 9.154377247543236e-06,
+      "loss": 0.5061,
+      "step": 1407
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.707863982326407,
+      "learning_rate": 9.153017580909863e-06,
+      "loss": 0.4977,
+      "step": 1408
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.346047472364499,
+      "learning_rate": 9.151656923201566e-06,
+      "loss": 0.527,
+      "step": 1409
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.923685917646882,
+      "learning_rate": 9.150295274743054e-06,
+      "loss": 0.5107,
+      "step": 1410
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.143608794943074,
+      "learning_rate": 9.148932635859269e-06,
+      "loss": 0.5016,
+      "step": 1411
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.373911371978839,
+      "learning_rate": 9.147569006875392e-06,
+      "loss": 0.4922,
+      "step": 1412
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 19.063058102702172,
+      "learning_rate": 9.146204388116839e-06,
+      "loss": 0.493,
+      "step": 1413
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 14.428239187320592,
+      "learning_rate": 9.14483877990926e-06,
+      "loss": 0.4821,
+      "step": 1414
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.338104783299679,
+      "learning_rate": 9.143472182578548e-06,
+      "loss": 0.4802,
+      "step": 1415
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 21.121469105400216,
+      "learning_rate": 9.142104596450823e-06,
+      "loss": 0.4995,
+      "step": 1416
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.063409727002606,
+      "learning_rate": 9.140736021852446e-06,
+      "loss": 0.5163,
+      "step": 1417
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 17.601861145528414,
+      "learning_rate": 9.139366459110016e-06,
+      "loss": 0.5531,
+      "step": 1418
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.249191826667966,
+      "learning_rate": 9.137995908550363e-06,
+      "loss": 0.4765,
+      "step": 1419
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.801433064129984,
+      "learning_rate": 9.136624370500555e-06,
+      "loss": 0.4782,
+      "step": 1420
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.178305509487918,
+      "learning_rate": 9.135251845287895e-06,
+      "loss": 0.534,
+      "step": 1421
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.799888631283141,
+      "learning_rate": 9.133878333239923e-06,
+      "loss": 0.5095,
+      "step": 1422
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.7539426481581994,
+      "learning_rate": 9.132503834684412e-06,
+      "loss": 0.5573,
+      "step": 1423
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 14.248648711886878,
+      "learning_rate": 9.131128349949375e-06,
+      "loss": 0.5536,
+      "step": 1424
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.205449838269684,
+      "learning_rate": 9.129751879363053e-06,
+      "loss": 0.4763,
+      "step": 1425
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.1855540235029896,
+      "learning_rate": 9.12837442325393e-06,
+      "loss": 0.4613,
+      "step": 1426
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.312733988116978,
+      "learning_rate": 9.126995981950719e-06,
+      "loss": 0.4672,
+      "step": 1427
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.969378071207799,
+      "learning_rate": 9.125616555782374e-06,
+      "loss": 0.4944,
+      "step": 1428
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 13.560939353860247,
+      "learning_rate": 9.124236145078079e-06,
+      "loss": 0.4893,
+      "step": 1429
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.727310868847268,
+      "learning_rate": 9.122854750167255e-06,
+      "loss": 0.5298,
+      "step": 1430
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 16.87160167062892,
+      "learning_rate": 9.121472371379559e-06,
+      "loss": 0.4133,
+      "step": 1431
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 23.57780058146975,
+      "learning_rate": 9.120089009044876e-06,
+      "loss": 0.4868,
+      "step": 1432
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.680148847443938,
+      "learning_rate": 9.11870466349334e-06,
+      "loss": 0.5909,
+      "step": 1433
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.731360824656868,
+      "learning_rate": 9.117319335055304e-06,
+      "loss": 0.4931,
+      "step": 1434
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.812148184396323,
+      "learning_rate": 9.115933024061366e-06,
+      "loss": 0.4759,
+      "step": 1435
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.985552656479306,
+      "learning_rate": 9.114545730842352e-06,
+      "loss": 0.5127,
+      "step": 1436
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 13.059552253300287,
+      "learning_rate": 9.113157455729326e-06,
+      "loss": 0.5119,
+      "step": 1437
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.334915407973871,
+      "learning_rate": 9.111768199053588e-06,
+      "loss": 0.4913,
+      "step": 1438
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 14.119529773665557,
+      "learning_rate": 9.110377961146668e-06,
+      "loss": 0.566,
+      "step": 1439
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.846150868201904,
+      "learning_rate": 9.108986742340331e-06,
+      "loss": 0.5034,
+      "step": 1440
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.08892686335765,
+      "learning_rate": 9.10759454296658e-06,
+      "loss": 0.5015,
+      "step": 1441
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 23.45613332108298,
+      "learning_rate": 9.106201363357645e-06,
+      "loss": 0.489,
+      "step": 1442
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 19.827004993588893,
+      "learning_rate": 9.104807203845997e-06,
+      "loss": 0.5275,
+      "step": 1443
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 12.511753529328168,
+      "learning_rate": 9.103412064764338e-06,
+      "loss": 0.503,
+      "step": 1444
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 12.107480886845112,
+      "learning_rate": 9.102015946445602e-06,
+      "loss": 0.5262,
+      "step": 1445
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.878004419485872,
+      "learning_rate": 9.10061884922296e-06,
+      "loss": 0.5212,
+      "step": 1446
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 28.116267478321486,
+      "learning_rate": 9.099220773429812e-06,
+      "loss": 0.5442,
+      "step": 1447
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.443314897524829,
+      "learning_rate": 9.097821719399795e-06,
+      "loss": 0.4505,
+      "step": 1448
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 16.647970507174,
+      "learning_rate": 9.096421687466782e-06,
+      "loss": 0.4994,
+      "step": 1449
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.233590711197776,
+      "learning_rate": 9.095020677964874e-06,
+      "loss": 0.55,
+      "step": 1450
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.899181100425423,
+      "learning_rate": 9.093618691228407e-06,
+      "loss": 0.4766,
+      "step": 1451
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 40.80967423407616,
+      "learning_rate": 9.092215727591952e-06,
+      "loss": 0.4784,
+      "step": 1452
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.42039925823618,
+      "learning_rate": 9.090811787390313e-06,
+      "loss": 0.4862,
+      "step": 1453
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.785200126996847,
+      "learning_rate": 9.089406870958522e-06,
+      "loss": 0.4558,
+      "step": 1454
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.511216865150638,
+      "learning_rate": 9.08800097863185e-06,
+      "loss": 0.5197,
+      "step": 1455
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.224894069969645,
+      "learning_rate": 9.0865941107458e-06,
+      "loss": 0.529,
+      "step": 1456
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 12.899155665057501,
+      "learning_rate": 9.085186267636103e-06,
+      "loss": 0.5472,
+      "step": 1457
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.877000259025259,
+      "learning_rate": 9.08377744963873e-06,
+      "loss": 0.6346,
+      "step": 1458
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.915888978915937,
+      "learning_rate": 9.08236765708988e-06,
+      "loss": 0.4406,
+      "step": 1459
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 25.72000585285448,
+      "learning_rate": 9.080956890325985e-06,
+      "loss": 0.493,
+      "step": 1460
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.194341404954631,
+      "learning_rate": 9.079545149683708e-06,
+      "loss": 0.5405,
+      "step": 1461
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.408975067927627,
+      "learning_rate": 9.078132435499951e-06,
+      "loss": 0.5278,
+      "step": 1462
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.360398106935369,
+      "learning_rate": 9.07671874811184e-06,
+      "loss": 0.4664,
+      "step": 1463
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.896914999939522,
+      "learning_rate": 9.075304087856738e-06,
+      "loss": 0.4741,
+      "step": 1464
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.094189745324796,
+      "learning_rate": 9.073888455072239e-06,
+      "loss": 0.4509,
+      "step": 1465
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.516996394659442,
+      "learning_rate": 9.07247185009617e-06,
+      "loss": 0.5,
+      "step": 1466
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.261252757580161,
+      "learning_rate": 9.071054273266586e-06,
+      "loss": 0.5135,
+      "step": 1467
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.49978403359483,
+      "learning_rate": 9.069635724921784e-06,
+      "loss": 0.5813,
+      "step": 1468
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.39739633279772,
+      "learning_rate": 9.068216205400278e-06,
+      "loss": 0.5693,
+      "step": 1469
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.957432278613746,
+      "learning_rate": 9.066795715040825e-06,
+      "loss": 0.513,
+      "step": 1470
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.254103134007421,
+      "learning_rate": 9.065374254182414e-06,
+      "loss": 0.5117,
+      "step": 1471
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 13.077983048144374,
+      "learning_rate": 9.063951823164256e-06,
+      "loss": 0.4606,
+      "step": 1472
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.576507268153216,
+      "learning_rate": 9.062528422325804e-06,
+      "loss": 0.4501,
+      "step": 1473
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.424099464896427,
+      "learning_rate": 9.061104052006737e-06,
+      "loss": 0.5181,
+      "step": 1474
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 17.378660237187464,
+      "learning_rate": 9.059678712546964e-06,
+      "loss": 0.5073,
+      "step": 1475
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.308392613715438,
+      "learning_rate": 9.05825240428663e-06,
+      "loss": 0.5583,
+      "step": 1476
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.918691571190655,
+      "learning_rate": 9.056825127566107e-06,
+      "loss": 0.4525,
+      "step": 1477
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.115829345213592,
+      "learning_rate": 9.055396882726001e-06,
+      "loss": 0.4572,
+      "step": 1478
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.085009054783822,
+      "learning_rate": 9.053967670107148e-06,
+      "loss": 0.5051,
+      "step": 1479
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.90600339656031,
+      "learning_rate": 9.052537490050615e-06,
+      "loss": 0.4836,
+      "step": 1480
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.584471229361751,
+      "learning_rate": 9.051106342897699e-06,
+      "loss": 0.5384,
+      "step": 1481
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.53480885264879,
+      "learning_rate": 9.049674228989929e-06,
+      "loss": 0.4941,
+      "step": 1482
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.847990377012414,
+      "learning_rate": 9.048241148669066e-06,
+      "loss": 0.4394,
+      "step": 1483
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.875381513647627,
+      "learning_rate": 9.046807102277096e-06,
+      "loss": 0.4773,
+      "step": 1484
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.2767679503740545,
+      "learning_rate": 9.045372090156244e-06,
+      "loss": 0.4396,
+      "step": 1485
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.458650884658711,
+      "learning_rate": 9.043936112648956e-06,
+      "loss": 0.498,
+      "step": 1486
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.265900380200062,
+      "learning_rate": 9.042499170097917e-06,
+      "loss": 0.4541,
+      "step": 1487
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.182825735743462,
+      "learning_rate": 9.041061262846037e-06,
+      "loss": 0.4927,
+      "step": 1488
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.4337311328873446,
+      "learning_rate": 9.039622391236457e-06,
+      "loss": 0.4895,
+      "step": 1489
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.583484486706033,
+      "learning_rate": 9.038182555612551e-06,
+      "loss": 0.4796,
+      "step": 1490
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.696687945240696,
+      "learning_rate": 9.03674175631792e-06,
+      "loss": 0.4277,
+      "step": 1491
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.00230277684635,
+      "learning_rate": 9.035299993696396e-06,
+      "loss": 0.4199,
+      "step": 1492
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.193402853176698,
+      "learning_rate": 9.03385726809204e-06,
+      "loss": 0.5018,
+      "step": 1493
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.081424735682337,
+      "learning_rate": 9.032413579849146e-06,
+      "loss": 0.544,
+      "step": 1494
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.503721000668175,
+      "learning_rate": 9.030968929312231e-06,
+      "loss": 0.5325,
+      "step": 1495
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.593979988459677,
+      "learning_rate": 9.02952331682605e-06,
+      "loss": 0.5785,
+      "step": 1496
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.474528584345009,
+      "learning_rate": 9.028076742735583e-06,
+      "loss": 0.4555,
+      "step": 1497
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.817703541843584,
+      "learning_rate": 9.026629207386038e-06,
+      "loss": 0.5077,
+      "step": 1498
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.616431368330579,
+      "learning_rate": 9.025180711122857e-06,
+      "loss": 0.4743,
+      "step": 1499
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.151542388569003,
+      "learning_rate": 9.023731254291706e-06,
+      "loss": 0.446,
+      "step": 1500
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.627918830248438,
+      "learning_rate": 9.022280837238484e-06,
+      "loss": 0.4623,
+      "step": 1501
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.129737175707645,
+      "learning_rate": 9.02082946030932e-06,
+      "loss": 0.4728,
+      "step": 1502
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.372478521838489,
+      "learning_rate": 9.019377123850566e-06,
+      "loss": 0.5239,
+      "step": 1503
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.2072584269996787,
+      "learning_rate": 9.017923828208812e-06,
+      "loss": 0.491,
+      "step": 1504
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.561788126108043,
+      "learning_rate": 9.01646957373087e-06,
+      "loss": 0.445,
+      "step": 1505
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.40397455301525,
+      "learning_rate": 9.015014360763781e-06,
+      "loss": 0.5685,
+      "step": 1506
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.402329650156023,
+      "learning_rate": 9.013558189654819e-06,
+      "loss": 0.6057,
+      "step": 1507
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.201592731475415,
+      "learning_rate": 9.012101060751484e-06,
+      "loss": 0.406,
+      "step": 1508
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.6707778332294816,
+      "learning_rate": 9.010642974401505e-06,
+      "loss": 0.4499,
+      "step": 1509
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.5218108138367095,
+      "learning_rate": 9.009183930952837e-06,
+      "loss": 0.5088,
+      "step": 1510
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.041990522030021,
+      "learning_rate": 9.007723930753667e-06,
+      "loss": 0.5896,
+      "step": 1511
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.519231233376105,
+      "learning_rate": 9.006262974152411e-06,
+      "loss": 0.5319,
+      "step": 1512
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 13.890619553144052,
+      "learning_rate": 9.00480106149771e-06,
+      "loss": 0.499,
+      "step": 1513
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 12.764277952653995,
+      "learning_rate": 9.003338193138433e-06,
+      "loss": 0.5098,
+      "step": 1514
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.337403526654539,
+      "learning_rate": 9.00187436942368e-06,
+      "loss": 0.5164,
+      "step": 1515
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.478355267798711,
+      "learning_rate": 9.000409590702777e-06,
+      "loss": 0.4604,
+      "step": 1516
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.984036265618275,
+      "learning_rate": 8.998943857325277e-06,
+      "loss": 0.4696,
+      "step": 1517
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.925662389882426,
+      "learning_rate": 8.997477169640965e-06,
+      "loss": 0.5376,
+      "step": 1518
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.4826423798942665,
+      "learning_rate": 8.996009527999849e-06,
+      "loss": 0.4367,
+      "step": 1519
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.728759318804837,
+      "learning_rate": 8.994540932752168e-06,
+      "loss": 0.6066,
+      "step": 1520
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.687180636863638,
+      "learning_rate": 8.993071384248386e-06,
+      "loss": 0.713,
+      "step": 1521
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.036612061401911,
+      "learning_rate": 8.991600882839195e-06,
+      "loss": 0.5503,
+      "step": 1522
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.4033612859676925,
+      "learning_rate": 8.990129428875517e-06,
+      "loss": 0.5558,
+      "step": 1523
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.339142560411955,
+      "learning_rate": 8.988657022708497e-06,
+      "loss": 0.4918,
+      "step": 1524
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 21.437009577271816,
+      "learning_rate": 8.987183664689511e-06,
+      "loss": 0.5511,
+      "step": 1525
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1805261051584033,
+      "learning_rate": 8.985709355170162e-06,
+      "loss": 0.4578,
+      "step": 1526
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.299927412331658,
+      "learning_rate": 8.984234094502278e-06,
+      "loss": 0.5639,
+      "step": 1527
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.021996371802004,
+      "learning_rate": 8.982757883037913e-06,
+      "loss": 0.5398,
+      "step": 1528
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.300686914145199,
+      "learning_rate": 8.98128072112935e-06,
+      "loss": 0.5288,
+      "step": 1529
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.183426579181348,
+      "learning_rate": 8.979802609129101e-06,
+      "loss": 0.5156,
+      "step": 1530
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.721015940813186,
+      "learning_rate": 8.978323547389899e-06,
+      "loss": 0.5001,
+      "step": 1531
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.025467704964178,
+      "learning_rate": 8.97684353626471e-06,
+      "loss": 0.4918,
+      "step": 1532
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.179572041345102,
+      "learning_rate": 8.97536257610672e-06,
+      "loss": 0.4668,
+      "step": 1533
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.718165624468269,
+      "learning_rate": 8.973880667269348e-06,
+      "loss": 0.4776,
+      "step": 1534
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.705962047252945,
+      "learning_rate": 8.972397810106235e-06,
+      "loss": 0.5,
+      "step": 1535
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.8257552670085317,
+      "learning_rate": 8.970914004971249e-06,
+      "loss": 0.4914,
+      "step": 1536
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.61466705105209,
+      "learning_rate": 8.969429252218487e-06,
+      "loss": 0.5056,
+      "step": 1537
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.477905846502025,
+      "learning_rate": 8.967943552202268e-06,
+      "loss": 0.6105,
+      "step": 1538
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 28.237532498731063,
+      "learning_rate": 8.966456905277138e-06,
+      "loss": 0.5358,
+      "step": 1539
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.7084151387854067,
+      "learning_rate": 8.964969311797873e-06,
+      "loss": 0.4807,
+      "step": 1540
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.90764865438277,
+      "learning_rate": 8.96348077211947e-06,
+      "loss": 0.5164,
+      "step": 1541
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.09509778196222,
+      "learning_rate": 8.961991286597154e-06,
+      "loss": 0.4562,
+      "step": 1542
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.936573492749508,
+      "learning_rate": 8.960500855586374e-06,
+      "loss": 0.4729,
+      "step": 1543
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.977975310854192,
+      "learning_rate": 8.959009479442811e-06,
+      "loss": 0.5089,
+      "step": 1544
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.813690494260851,
+      "learning_rate": 8.95751715852236e-06,
+      "loss": 0.4874,
+      "step": 1545
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.7836356658002654,
+      "learning_rate": 8.956023893181152e-06,
+      "loss": 0.486,
+      "step": 1546
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.845656914896341,
+      "learning_rate": 8.954529683775539e-06,
+      "loss": 0.4486,
+      "step": 1547
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.5431160236267876,
+      "learning_rate": 8.953034530662096e-06,
+      "loss": 0.4564,
+      "step": 1548
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.295600230765116,
+      "learning_rate": 8.951538434197631e-06,
+      "loss": 0.5581,
+      "step": 1549
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.8005098960015475,
+      "learning_rate": 8.950041394739169e-06,
+      "loss": 0.5365,
+      "step": 1550
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.4124557518316845,
+      "learning_rate": 8.948543412643962e-06,
+      "loss": 0.5371,
+      "step": 1551
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.7992116211150027,
+      "learning_rate": 8.947044488269489e-06,
+      "loss": 0.472,
+      "step": 1552
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.4404926818713246,
+      "learning_rate": 8.945544621973454e-06,
+      "loss": 0.4439,
+      "step": 1553
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 19.724220377121565,
+      "learning_rate": 8.944043814113785e-06,
+      "loss": 0.4997,
+      "step": 1554
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5006398971090862,
+      "learning_rate": 8.942542065048631e-06,
+      "loss": 0.5852,
+      "step": 1555
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.9479806675799196,
+      "learning_rate": 8.94103937513637e-06,
+      "loss": 0.4397,
+      "step": 1556
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.240396290372782,
+      "learning_rate": 8.939535744735608e-06,
+      "loss": 0.5704,
+      "step": 1557
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.483672322966049,
+      "learning_rate": 8.938031174205166e-06,
+      "loss": 0.5407,
+      "step": 1558
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.620883873344044,
+      "learning_rate": 8.936525663904097e-06,
+      "loss": 0.5232,
+      "step": 1559
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.9346802210421945,
+      "learning_rate": 8.935019214191673e-06,
+      "loss": 0.5494,
+      "step": 1560
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.200113489173202,
+      "learning_rate": 8.933511825427393e-06,
+      "loss": 0.4709,
+      "step": 1561
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.8291109058451522,
+      "learning_rate": 8.932003497970983e-06,
+      "loss": 0.5548,
+      "step": 1562
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.258140076890803,
+      "learning_rate": 8.930494232182385e-06,
+      "loss": 0.5561,
+      "step": 1563
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.728808823924346,
+      "learning_rate": 8.928984028421777e-06,
+      "loss": 0.519,
+      "step": 1564
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.744608648630482,
+      "learning_rate": 8.927472887049546e-06,
+      "loss": 0.5023,
+      "step": 1565
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 16.546278846576048,
+      "learning_rate": 8.925960808426313e-06,
+      "loss": 0.4632,
+      "step": 1566
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.536346941375556,
+      "learning_rate": 8.924447792912921e-06,
+      "loss": 0.4559,
+      "step": 1567
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.440776909911878,
+      "learning_rate": 8.922933840870436e-06,
+      "loss": 0.4829,
+      "step": 1568
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.081884084987932,
+      "learning_rate": 8.921418952660146e-06,
+      "loss": 0.5551,
+      "step": 1569
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6872539177515735,
+      "learning_rate": 8.919903128643561e-06,
+      "loss": 0.5538,
+      "step": 1570
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.723160141797946,
+      "learning_rate": 8.918386369182423e-06,
+      "loss": 0.4926,
+      "step": 1571
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.012526672899946,
+      "learning_rate": 8.916868674638686e-06,
+      "loss": 0.4464,
+      "step": 1572
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.214614306187443,
+      "learning_rate": 8.915350045374533e-06,
+      "loss": 0.5044,
+      "step": 1573
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 11.104318510025482,
+      "learning_rate": 8.91383048175237e-06,
+      "loss": 0.4585,
+      "step": 1574
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3764786205247317,
+      "learning_rate": 8.912309984134826e-06,
+      "loss": 0.6087,
+      "step": 1575
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.49536406566925,
+      "learning_rate": 8.91078855288475e-06,
+      "loss": 0.4914,
+      "step": 1576
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.943308936776226,
+      "learning_rate": 8.909266188365218e-06,
+      "loss": 0.4914,
+      "step": 1577
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.9575791586285884,
+      "learning_rate": 8.907742890939526e-06,
+      "loss": 0.4861,
+      "step": 1578
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.578798004886123,
+      "learning_rate": 8.906218660971192e-06,
+      "loss": 0.5015,
+      "step": 1579
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.298115173163746,
+      "learning_rate": 8.90469349882396e-06,
+      "loss": 0.4087,
+      "step": 1580
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.414973434589644,
+      "learning_rate": 8.903167404861792e-06,
+      "loss": 0.4824,
+      "step": 1581
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.485907870921125,
+      "learning_rate": 8.901640379448875e-06,
+      "loss": 0.5567,
+      "step": 1582
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.755053685400907,
+      "learning_rate": 8.90011242294962e-06,
+      "loss": 0.465,
+      "step": 1583
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.07057115155756,
+      "learning_rate": 8.898583535728656e-06,
+      "loss": 0.4254,
+      "step": 1584
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 17.26107606581141,
+      "learning_rate": 8.89705371815084e-06,
+      "loss": 0.5965,
+      "step": 1585
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.81838509404585,
+      "learning_rate": 8.89552297058124e-06,
+      "loss": 0.5313,
+      "step": 1586
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.794821587600955,
+      "learning_rate": 8.893991293385159e-06,
+      "loss": 0.4032,
+      "step": 1587
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.444793442133021,
+      "learning_rate": 8.892458686928115e-06,
+      "loss": 0.4633,
+      "step": 1588
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.114230069577213,
+      "learning_rate": 8.890925151575846e-06,
+      "loss": 0.5664,
+      "step": 1589
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.173006355354904,
+      "learning_rate": 8.889390687694318e-06,
+      "loss": 0.5422,
+      "step": 1590
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.398966619540447,
+      "learning_rate": 8.887855295649713e-06,
+      "loss": 0.4815,
+      "step": 1591
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 69.17134608264283,
+      "learning_rate": 8.886318975808436e-06,
+      "loss": 0.5056,
+      "step": 1592
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.04047377674635,
+      "learning_rate": 8.884781728537117e-06,
+      "loss": 0.5805,
+      "step": 1593
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.145511927930641,
+      "learning_rate": 8.883243554202603e-06,
+      "loss": 0.4789,
+      "step": 1594
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.963253899583847,
+      "learning_rate": 8.88170445317196e-06,
+      "loss": 0.577,
+      "step": 1595
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.081286073706801,
+      "learning_rate": 8.880164425812483e-06,
+      "loss": 0.4723,
+      "step": 1596
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.186068393302075,
+      "learning_rate": 8.878623472491683e-06,
+      "loss": 0.4703,
+      "step": 1597
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.392687277500229,
+      "learning_rate": 8.877081593577289e-06,
+      "loss": 0.4525,
+      "step": 1598
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 32.20000426959304,
+      "learning_rate": 8.875538789437259e-06,
+      "loss": 0.4454,
+      "step": 1599
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.224367896613412,
+      "learning_rate": 8.873995060439765e-06,
+      "loss": 0.5762,
+      "step": 1600
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.634713571904084,
+      "learning_rate": 8.872450406953202e-06,
+      "loss": 0.4568,
+      "step": 1601
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.355223764585945,
+      "learning_rate": 8.870904829346187e-06,
+      "loss": 0.4946,
+      "step": 1602
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.1180019370521315,
+      "learning_rate": 8.869358327987556e-06,
+      "loss": 0.4824,
+      "step": 1603
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.0992207440796005,
+      "learning_rate": 8.867810903246364e-06,
+      "loss": 0.4444,
+      "step": 1604
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.1724872996723015,
+      "learning_rate": 8.86626255549189e-06,
+      "loss": 0.4056,
+      "step": 1605
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.809846539413385,
+      "learning_rate": 8.86471328509363e-06,
+      "loss": 0.489,
+      "step": 1606
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.3616182603951925,
+      "learning_rate": 8.8631630924213e-06,
+      "loss": 0.5177,
+      "step": 1607
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.97609224174291,
+      "learning_rate": 8.861611977844843e-06,
+      "loss": 0.5472,
+      "step": 1608
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.3523603447103945,
+      "learning_rate": 8.860059941734412e-06,
+      "loss": 0.4696,
+      "step": 1609
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.712176575715722,
+      "learning_rate": 8.858506984460383e-06,
+      "loss": 0.5528,
+      "step": 1610
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.44691845074557,
+      "learning_rate": 8.856953106393357e-06,
+      "loss": 0.5045,
+      "step": 1611
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.040833230502797,
+      "learning_rate": 8.85539830790415e-06,
+      "loss": 0.4487,
+      "step": 1612
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.621001085871318,
+      "learning_rate": 8.853842589363796e-06,
+      "loss": 0.5862,
+      "step": 1613
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.557456278573803,
+      "learning_rate": 8.852285951143556e-06,
+      "loss": 0.5322,
+      "step": 1614
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.6880220450133026,
+      "learning_rate": 8.850728393614903e-06,
+      "loss": 0.5186,
+      "step": 1615
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.288556227132453,
+      "learning_rate": 8.849169917149532e-06,
+      "loss": 0.4823,
+      "step": 1616
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 11.296741569900654,
+      "learning_rate": 8.847610522119358e-06,
+      "loss": 0.5524,
+      "step": 1617
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.972526387373982,
+      "learning_rate": 8.846050208896512e-06,
+      "loss": 0.7018,
+      "step": 1618
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.158975955011825,
+      "learning_rate": 8.844488977853351e-06,
+      "loss": 0.4879,
+      "step": 1619
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.6125019554649045,
+      "learning_rate": 8.842926829362447e-06,
+      "loss": 0.555,
+      "step": 1620
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.780515089934748,
+      "learning_rate": 8.841363763796585e-06,
+      "loss": 0.48,
+      "step": 1621
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.6402425437624655,
+      "learning_rate": 8.839799781528778e-06,
+      "loss": 0.4894,
+      "step": 1622
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.701134920357182,
+      "learning_rate": 8.838234882932254e-06,
+      "loss": 0.4408,
+      "step": 1623
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.8214848962619845,
+      "learning_rate": 8.836669068380461e-06,
+      "loss": 0.4687,
+      "step": 1624
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.927607111756089,
+      "learning_rate": 8.835102338247063e-06,
+      "loss": 0.4023,
+      "step": 1625
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.163116439516577,
+      "learning_rate": 8.833534692905945e-06,
+      "loss": 0.4706,
+      "step": 1626
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.154094327152368,
+      "learning_rate": 8.83196613273121e-06,
+      "loss": 0.5008,
+      "step": 1627
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.772041329013814,
+      "learning_rate": 8.830396658097176e-06,
+      "loss": 0.4576,
+      "step": 1628
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.144494067258229,
+      "learning_rate": 8.828826269378385e-06,
+      "loss": 0.5608,
+      "step": 1629
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.795100358442994,
+      "learning_rate": 8.827254966949594e-06,
+      "loss": 0.4369,
+      "step": 1630
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.524220784243382,
+      "learning_rate": 8.825682751185775e-06,
+      "loss": 0.4601,
+      "step": 1631
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.3840387303231445,
+      "learning_rate": 8.824109622462123e-06,
+      "loss": 0.4671,
+      "step": 1632
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.266428674581977,
+      "learning_rate": 8.82253558115405e-06,
+      "loss": 0.4906,
+      "step": 1633
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.623526914571225,
+      "learning_rate": 8.820960627637185e-06,
+      "loss": 0.5836,
+      "step": 1634
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.326559272938898,
+      "learning_rate": 8.819384762287372e-06,
+      "loss": 0.5281,
+      "step": 1635
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.205320946490698,
+      "learning_rate": 8.817807985480677e-06,
+      "loss": 0.4087,
+      "step": 1636
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7248459229059117,
+      "learning_rate": 8.816230297593382e-06,
+      "loss": 0.6174,
+      "step": 1637
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.766487058133429,
+      "learning_rate": 8.814651699001984e-06,
+      "loss": 0.4711,
+      "step": 1638
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.560501904242485,
+      "learning_rate": 8.813072190083201e-06,
+      "loss": 0.5029,
+      "step": 1639
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.781784477683787,
+      "learning_rate": 8.811491771213965e-06,
+      "loss": 0.52,
+      "step": 1640
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.027907580841462,
+      "learning_rate": 8.809910442771427e-06,
+      "loss": 0.4953,
+      "step": 1641
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.8979379211286567,
+      "learning_rate": 8.808328205132957e-06,
+      "loss": 0.4604,
+      "step": 1642
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.450293775066662,
+      "learning_rate": 8.806745058676137e-06,
+      "loss": 0.4645,
+      "step": 1643
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.4094134569507615,
+      "learning_rate": 8.80516100377877e-06,
+      "loss": 0.5406,
+      "step": 1644
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.811636257719224,
+      "learning_rate": 8.803576040818874e-06,
+      "loss": 0.5321,
+      "step": 1645
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.487177067160405,
+      "learning_rate": 8.801990170174685e-06,
+      "loss": 0.5319,
+      "step": 1646
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.6953686344707206,
+      "learning_rate": 8.800403392224652e-06,
+      "loss": 0.5106,
+      "step": 1647
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.783852864513696,
+      "learning_rate": 8.798815707347449e-06,
+      "loss": 0.4406,
+      "step": 1648
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.002935297200702,
+      "learning_rate": 8.797227115921953e-06,
+      "loss": 0.5236,
+      "step": 1649
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.585681638649321,
+      "learning_rate": 8.795637618327269e-06,
+      "loss": 0.5207,
+      "step": 1650
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.5058702588937574,
+      "learning_rate": 8.794047214942716e-06,
+      "loss": 0.4787,
+      "step": 1651
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0621273299799614,
+      "learning_rate": 8.792455906147825e-06,
+      "loss": 0.4954,
+      "step": 1652
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.3863423632919316,
+      "learning_rate": 8.790863692322345e-06,
+      "loss": 0.4834,
+      "step": 1653
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.086800459201511,
+      "learning_rate": 8.789270573846243e-06,
+      "loss": 0.4625,
+      "step": 1654
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.018283759382251,
+      "learning_rate": 8.7876765510997e-06,
+      "loss": 0.5181,
+      "step": 1655
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.9126111585435,
+      "learning_rate": 8.786081624463113e-06,
+      "loss": 0.4736,
+      "step": 1656
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.6098207226673504,
+      "learning_rate": 8.784485794317094e-06,
+      "loss": 0.5197,
+      "step": 1657
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 25.434575790810822,
+      "learning_rate": 8.782889061042473e-06,
+      "loss": 0.4489,
+      "step": 1658
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.623123257587492,
+      "learning_rate": 8.781291425020291e-06,
+      "loss": 0.4399,
+      "step": 1659
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.588996656431235,
+      "learning_rate": 8.779692886631812e-06,
+      "loss": 0.4834,
+      "step": 1660
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.141390110432851,
+      "learning_rate": 8.778093446258506e-06,
+      "loss": 0.4516,
+      "step": 1661
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.9457607614863255,
+      "learning_rate": 8.776493104282069e-06,
+      "loss": 0.5136,
+      "step": 1662
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.359451779538328,
+      "learning_rate": 8.774891861084399e-06,
+      "loss": 0.5075,
+      "step": 1663
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0228773282199333,
+      "learning_rate": 8.773289717047621e-06,
+      "loss": 0.5019,
+      "step": 1664
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.1570443551814265,
+      "learning_rate": 8.771686672554068e-06,
+      "loss": 0.5236,
+      "step": 1665
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.122739043388691,
+      "learning_rate": 8.770082727986292e-06,
+      "loss": 0.5183,
+      "step": 1666
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.616537409213944,
+      "learning_rate": 8.768477883727056e-06,
+      "loss": 0.5108,
+      "step": 1667
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.688798977017811,
+      "learning_rate": 8.76687214015934e-06,
+      "loss": 0.5319,
+      "step": 1668
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3875133199985785,
+      "learning_rate": 8.765265497666339e-06,
+      "loss": 0.719,
+      "step": 1669
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.675669693868851,
+      "learning_rate": 8.763657956631461e-06,
+      "loss": 0.5414,
+      "step": 1670
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.7141885523597375,
+      "learning_rate": 8.762049517438332e-06,
+      "loss": 0.5163,
+      "step": 1671
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.006016801245012,
+      "learning_rate": 8.760440180470784e-06,
+      "loss": 0.525,
+      "step": 1672
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.873245554602478,
+      "learning_rate": 8.75882994611287e-06,
+      "loss": 0.4427,
+      "step": 1673
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.400958463737394,
+      "learning_rate": 8.75721881474886e-06,
+      "loss": 0.5145,
+      "step": 1674
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.8153539939755574,
+      "learning_rate": 8.75560678676323e-06,
+      "loss": 0.4316,
+      "step": 1675
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.672317338727773,
+      "learning_rate": 8.753993862540678e-06,
+      "loss": 0.5213,
+      "step": 1676
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.698555107641,
+      "learning_rate": 8.752380042466107e-06,
+      "loss": 0.4416,
+      "step": 1677
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.236479927148768,
+      "learning_rate": 8.75076532692464e-06,
+      "loss": 0.5224,
+      "step": 1678
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.088855734919785,
+      "learning_rate": 8.749149716301612e-06,
+      "loss": 0.4271,
+      "step": 1679
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.606439945243697,
+      "learning_rate": 8.747533210982576e-06,
+      "loss": 0.4871,
+      "step": 1680
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.2344410052785335,
+      "learning_rate": 8.745915811353288e-06,
+      "loss": 0.5627,
+      "step": 1681
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 11.05021026541072,
+      "learning_rate": 8.744297517799728e-06,
+      "loss": 0.4676,
+      "step": 1682
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 29.41510511226427,
+      "learning_rate": 8.742678330708085e-06,
+      "loss": 0.5105,
+      "step": 1683
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.784851741938727,
+      "learning_rate": 8.741058250464758e-06,
+      "loss": 0.5444,
+      "step": 1684
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.374545347808052,
+      "learning_rate": 8.739437277456366e-06,
+      "loss": 0.4593,
+      "step": 1685
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.706560139265927,
+      "learning_rate": 8.737815412069737e-06,
+      "loss": 0.4056,
+      "step": 1686
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.152341776308717,
+      "learning_rate": 8.73619265469191e-06,
+      "loss": 0.4901,
+      "step": 1687
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.116323451961182,
+      "learning_rate": 8.734569005710142e-06,
+      "loss": 0.5244,
+      "step": 1688
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.837048672011069,
+      "learning_rate": 8.732944465511899e-06,
+      "loss": 0.428,
+      "step": 1689
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.685629218608238,
+      "learning_rate": 8.731319034484862e-06,
+      "loss": 0.5249,
+      "step": 1690
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.938563631714166,
+      "learning_rate": 8.729692713016921e-06,
+      "loss": 0.5015,
+      "step": 1691
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.318960008574061,
+      "learning_rate": 8.728065501496183e-06,
+      "loss": 0.5402,
+      "step": 1692
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 13.79651926512532,
+      "learning_rate": 8.726437400310965e-06,
+      "loss": 0.5039,
+      "step": 1693
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.627296378205132,
+      "learning_rate": 8.724808409849795e-06,
+      "loss": 0.5966,
+      "step": 1694
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.423552605889969,
+      "learning_rate": 8.723178530501417e-06,
+      "loss": 0.4478,
+      "step": 1695
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 13.005001217358673,
+      "learning_rate": 8.721547762654786e-06,
+      "loss": 0.5774,
+      "step": 1696
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.527322593397208,
+      "learning_rate": 8.719916106699063e-06,
+      "loss": 0.5402,
+      "step": 1697
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 11.381746489586043,
+      "learning_rate": 8.718283563023628e-06,
+      "loss": 0.4364,
+      "step": 1698
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.644291536834517,
+      "learning_rate": 8.716650132018075e-06,
+      "loss": 0.4915,
+      "step": 1699
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.505490353424833,
+      "learning_rate": 8.7150158140722e-06,
+      "loss": 0.5699,
+      "step": 1700
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.310694577414711,
+      "learning_rate": 8.71338060957602e-06,
+      "loss": 0.5051,
+      "step": 1701
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.898145353850273,
+      "learning_rate": 8.711744518919757e-06,
+      "loss": 0.4829,
+      "step": 1702
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.500599759507134,
+      "learning_rate": 8.710107542493846e-06,
+      "loss": 0.4641,
+      "step": 1703
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.785609001946518,
+      "learning_rate": 8.708469680688941e-06,
+      "loss": 0.4454,
+      "step": 1704
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 12.207839154157819,
+      "learning_rate": 8.706830933895893e-06,
+      "loss": 0.5116,
+      "step": 1705
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.763470957268615,
+      "learning_rate": 8.705191302505777e-06,
+      "loss": 0.4912,
+      "step": 1706
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.8536668164014145,
+      "learning_rate": 8.703550786909873e-06,
+      "loss": 0.4294,
+      "step": 1707
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 13.736500341565748,
+      "learning_rate": 8.701909387499674e-06,
+      "loss": 0.4742,
+      "step": 1708
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 13.736247885515326,
+      "learning_rate": 8.70026710466688e-06,
+      "loss": 0.5239,
+      "step": 1709
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.15627137113051,
+      "learning_rate": 8.69862393880341e-06,
+      "loss": 0.4705,
+      "step": 1710
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.305562459014393,
+      "learning_rate": 8.696979890301383e-06,
+      "loss": 0.4959,
+      "step": 1711
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.845415504914943,
+      "learning_rate": 8.695334959553139e-06,
+      "loss": 0.4682,
+      "step": 1712
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.425886614853946,
+      "learning_rate": 8.69368914695122e-06,
+      "loss": 0.4691,
+      "step": 1713
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7985279231382951,
+      "learning_rate": 8.692042452888385e-06,
+      "loss": 0.6462,
+      "step": 1714
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.273808183626967,
+      "learning_rate": 8.6903948777576e-06,
+      "loss": 0.4762,
+      "step": 1715
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.1320508650919425,
+      "learning_rate": 8.68874642195204e-06,
+      "loss": 0.5079,
+      "step": 1716
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.4699204319018175,
+      "learning_rate": 8.687097085865096e-06,
+      "loss": 0.5072,
+      "step": 1717
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.6980225365146056,
+      "learning_rate": 8.685446869890359e-06,
+      "loss": 0.5331,
+      "step": 1718
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.626394132804195,
+      "learning_rate": 8.683795774421644e-06,
+      "loss": 0.4791,
+      "step": 1719
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.094627104804033,
+      "learning_rate": 8.68214379985296e-06,
+      "loss": 0.4471,
+      "step": 1720
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.1343135022707225,
+      "learning_rate": 8.680490946578538e-06,
+      "loss": 0.4632,
+      "step": 1721
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.642830257804902,
+      "learning_rate": 8.678837214992815e-06,
+      "loss": 0.5063,
+      "step": 1722
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.4526371088508965,
+      "learning_rate": 8.677182605490435e-06,
+      "loss": 0.4812,
+      "step": 1723
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.5996428755776635,
+      "learning_rate": 8.675527118466254e-06,
+      "loss": 0.4852,
+      "step": 1724
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.248127966248044,
+      "learning_rate": 8.673870754315336e-06,
+      "loss": 0.4586,
+      "step": 1725
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.732716959695629,
+      "learning_rate": 8.672213513432958e-06,
+      "loss": 0.5593,
+      "step": 1726
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.174808278174984,
+      "learning_rate": 8.670555396214604e-06,
+      "loss": 0.4781,
+      "step": 1727
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.9326768689224925,
+      "learning_rate": 8.668896403055963e-06,
+      "loss": 0.5074,
+      "step": 1728
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.063687075272308,
+      "learning_rate": 8.667236534352936e-06,
+      "loss": 0.5015,
+      "step": 1729
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.289384214794713,
+      "learning_rate": 8.665575790501638e-06,
+      "loss": 0.5447,
+      "step": 1730
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.288055654048188,
+      "learning_rate": 8.663914171898388e-06,
+      "loss": 0.5276,
+      "step": 1731
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.591013441382628,
+      "learning_rate": 8.66225167893971e-06,
+      "loss": 0.4584,
+      "step": 1732
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.608858885699101,
+      "learning_rate": 8.660588312022345e-06,
+      "loss": 0.5112,
+      "step": 1733
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.480236393874353,
+      "learning_rate": 8.658924071543235e-06,
+      "loss": 0.4786,
+      "step": 1734
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6480562704402328,
+      "learning_rate": 8.657258957899536e-06,
+      "loss": 0.6693,
+      "step": 1735
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.712964484884782,
+      "learning_rate": 8.655592971488612e-06,
+      "loss": 0.3909,
+      "step": 1736
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.9523993818905465,
+      "learning_rate": 8.653926112708028e-06,
+      "loss": 0.5562,
+      "step": 1737
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.88565232528944,
+      "learning_rate": 8.652258381955567e-06,
+      "loss": 0.557,
+      "step": 1738
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.257852941197796,
+      "learning_rate": 8.650589779629218e-06,
+      "loss": 0.4583,
+      "step": 1739
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.668374004936056,
+      "learning_rate": 8.64892030612717e-06,
+      "loss": 0.5121,
+      "step": 1740
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.699757509249108,
+      "learning_rate": 8.647249961847828e-06,
+      "loss": 0.577,
+      "step": 1741
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.65447756940695,
+      "learning_rate": 8.645578747189803e-06,
+      "loss": 0.5308,
+      "step": 1742
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3051734868593994,
+      "learning_rate": 8.643906662551912e-06,
+      "loss": 0.5594,
+      "step": 1743
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.242595266781202,
+      "learning_rate": 8.642233708333183e-06,
+      "loss": 0.4674,
+      "step": 1744
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.372166397375121,
+      "learning_rate": 8.640559884932848e-06,
+      "loss": 0.5248,
+      "step": 1745
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 15.961298152122554,
+      "learning_rate": 8.638885192750348e-06,
+      "loss": 0.5336,
+      "step": 1746
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 16.388903237462902,
+      "learning_rate": 8.637209632185333e-06,
+      "loss": 0.4675,
+      "step": 1747
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.90907103741499,
+      "learning_rate": 8.635533203637653e-06,
+      "loss": 0.5293,
+      "step": 1748
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.94809111533093,
+      "learning_rate": 8.633855907507375e-06,
+      "loss": 0.5483,
+      "step": 1749
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.432071700735126,
+      "learning_rate": 8.632177744194766e-06,
+      "loss": 0.5178,
+      "step": 1750
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.914964138186451,
+      "learning_rate": 8.630498714100306e-06,
+      "loss": 0.5118,
+      "step": 1751
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.786472124812631,
+      "learning_rate": 8.628818817624672e-06,
+      "loss": 0.4803,
+      "step": 1752
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.543855766819282,
+      "learning_rate": 8.62713805516876e-06,
+      "loss": 0.5887,
+      "step": 1753
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.255353077316545,
+      "learning_rate": 8.625456427133665e-06,
+      "loss": 0.4387,
+      "step": 1754
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.5496368362857282,
+      "learning_rate": 8.623773933920689e-06,
+      "loss": 0.4464,
+      "step": 1755
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.136869128886629,
+      "learning_rate": 8.622090575931341e-06,
+      "loss": 0.5017,
+      "step": 1756
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 20.960506743289706,
+      "learning_rate": 8.62040635356734e-06,
+      "loss": 0.4815,
+      "step": 1757
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.923631606303919,
+      "learning_rate": 8.618721267230606e-06,
+      "loss": 0.4363,
+      "step": 1758
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.336558220678053,
+      "learning_rate": 8.61703531732327e-06,
+      "loss": 0.4891,
+      "step": 1759
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.9102619509329015,
+      "learning_rate": 8.615348504247663e-06,
+      "loss": 0.4527,
+      "step": 1760
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.356492946306187,
+      "learning_rate": 8.613660828406328e-06,
+      "loss": 0.5535,
+      "step": 1761
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.588172444952468,
+      "learning_rate": 8.61197229020201e-06,
+      "loss": 0.4982,
+      "step": 1762
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.8934242683224305,
+      "learning_rate": 8.610282890037664e-06,
+      "loss": 0.5157,
+      "step": 1763
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.92502392325219,
+      "learning_rate": 8.608592628316446e-06,
+      "loss": 0.4622,
+      "step": 1764
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.901789152839644,
+      "learning_rate": 8.606901505441719e-06,
+      "loss": 0.5146,
+      "step": 1765
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 14.072053580569666,
+      "learning_rate": 8.605209521817053e-06,
+      "loss": 0.501,
+      "step": 1766
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.903189522576522,
+      "learning_rate": 8.603516677846222e-06,
+      "loss": 0.5097,
+      "step": 1767
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.053744418915311,
+      "learning_rate": 8.601822973933207e-06,
+      "loss": 0.4541,
+      "step": 1768
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.612330055149999,
+      "learning_rate": 8.60012841048219e-06,
+      "loss": 0.4175,
+      "step": 1769
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.844650730344002,
+      "learning_rate": 8.598432987897565e-06,
+      "loss": 0.4639,
+      "step": 1770
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.197881919392901,
+      "learning_rate": 8.596736706583926e-06,
+      "loss": 0.507,
+      "step": 1771
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 16.36907618351591,
+      "learning_rate": 8.595039566946072e-06,
+      "loss": 0.5325,
+      "step": 1772
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.788018738524345,
+      "learning_rate": 8.593341569389008e-06,
+      "loss": 0.4754,
+      "step": 1773
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.279323201105887,
+      "learning_rate": 8.591642714317945e-06,
+      "loss": 0.4949,
+      "step": 1774
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 12.212077955354141,
+      "learning_rate": 8.589943002138296e-06,
+      "loss": 0.4978,
+      "step": 1775
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.761324304834908,
+      "learning_rate": 8.588242433255679e-06,
+      "loss": 0.468,
+      "step": 1776
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.141129198540113,
+      "learning_rate": 8.586541008075919e-06,
+      "loss": 0.5107,
+      "step": 1777
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.902380163912766,
+      "learning_rate": 8.584838727005045e-06,
+      "loss": 0.4897,
+      "step": 1778
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.7474945763245,
+      "learning_rate": 8.583135590449285e-06,
+      "loss": 0.4152,
+      "step": 1779
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 19.861544170191262,
+      "learning_rate": 8.581431598815077e-06,
+      "loss": 0.6019,
+      "step": 1780
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.502514597191615,
+      "learning_rate": 8.57972675250906e-06,
+      "loss": 0.4737,
+      "step": 1781
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.494484292370675,
+      "learning_rate": 8.57802105193808e-06,
+      "loss": 0.4464,
+      "step": 1782
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.050161142457534,
+      "learning_rate": 8.576314497509185e-06,
+      "loss": 0.5978,
+      "step": 1783
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.759088505694082,
+      "learning_rate": 8.574607089629627e-06,
+      "loss": 0.5524,
+      "step": 1784
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.147084885891129,
+      "learning_rate": 8.572898828706858e-06,
+      "loss": 0.4786,
+      "step": 1785
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.399932466749341,
+      "learning_rate": 8.571189715148539e-06,
+      "loss": 0.5037,
+      "step": 1786
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.074990533664426,
+      "learning_rate": 8.569479749362531e-06,
+      "loss": 0.4467,
+      "step": 1787
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.207518606558148,
+      "learning_rate": 8.567768931756903e-06,
+      "loss": 0.4336,
+      "step": 1788
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.003562436795217,
+      "learning_rate": 8.566057262739922e-06,
+      "loss": 0.442,
+      "step": 1789
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.510809088126551,
+      "learning_rate": 8.56434474272006e-06,
+      "loss": 0.4981,
+      "step": 1790
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.564979457302885,
+      "learning_rate": 8.56263137210599e-06,
+      "loss": 0.4587,
+      "step": 1791
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.7226634684707145,
+      "learning_rate": 8.560917151306594e-06,
+      "loss": 0.4504,
+      "step": 1792
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.781576613288939,
+      "learning_rate": 8.559202080730952e-06,
+      "loss": 0.4877,
+      "step": 1793
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.775262744022146,
+      "learning_rate": 8.557486160788348e-06,
+      "loss": 0.5346,
+      "step": 1794
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 28.022989280439724,
+      "learning_rate": 8.55576939188827e-06,
+      "loss": 0.492,
+      "step": 1795
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.935354467138271,
+      "learning_rate": 8.554051774440404e-06,
+      "loss": 0.4636,
+      "step": 1796
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.599607442048166,
+      "learning_rate": 8.552333308854645e-06,
+      "loss": 0.4865,
+      "step": 1797
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.659096205944559,
+      "learning_rate": 8.550613995541084e-06,
+      "loss": 0.5302,
+      "step": 1798
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 13.627743035687987,
+      "learning_rate": 8.54889383491002e-06,
+      "loss": 0.4651,
+      "step": 1799
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.434794239492384,
+      "learning_rate": 8.54717282737195e-06,
+      "loss": 0.4465,
+      "step": 1800
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.042855575229791,
+      "learning_rate": 8.545450973337575e-06,
+      "loss": 0.5001,
+      "step": 1801
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.450688593313572,
+      "learning_rate": 8.5437282732178e-06,
+      "loss": 0.502,
+      "step": 1802
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.868801112505036,
+      "learning_rate": 8.542004727423724e-06,
+      "loss": 0.5314,
+      "step": 1803
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 11.093308302781576,
+      "learning_rate": 8.540280336366662e-06,
+      "loss": 0.5238,
+      "step": 1804
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 12.58213146531138,
+      "learning_rate": 8.538555100458114e-06,
+      "loss": 0.5106,
+      "step": 1805
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.252357638272064,
+      "learning_rate": 8.536829020109796e-06,
+      "loss": 0.4776,
+      "step": 1806
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.762650358637841,
+      "learning_rate": 8.535102095733615e-06,
+      "loss": 0.5282,
+      "step": 1807
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.116321501069386,
+      "learning_rate": 8.533374327741687e-06,
+      "loss": 0.5367,
+      "step": 1808
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.5679847521469314,
+      "learning_rate": 8.531645716546322e-06,
+      "loss": 0.4385,
+      "step": 1809
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.333155217016179,
+      "learning_rate": 8.529916262560039e-06,
+      "loss": 0.4781,
+      "step": 1810
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.0902949407806775,
+      "learning_rate": 8.528185966195553e-06,
+      "loss": 0.4163,
+      "step": 1811
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.0236505254688915,
+      "learning_rate": 8.52645482786578e-06,
+      "loss": 0.4678,
+      "step": 1812
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 14.221946029093298,
+      "learning_rate": 8.524722847983843e-06,
+      "loss": 0.4408,
+      "step": 1813
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.949250880093887,
+      "learning_rate": 8.522990026963057e-06,
+      "loss": 0.4727,
+      "step": 1814
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.110156870110036,
+      "learning_rate": 8.521256365216941e-06,
+      "loss": 0.4803,
+      "step": 1815
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 13.628590696226684,
+      "learning_rate": 8.519521863159219e-06,
+      "loss": 0.5122,
+      "step": 1816
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.637886203011313,
+      "learning_rate": 8.51778652120381e-06,
+      "loss": 0.4607,
+      "step": 1817
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.209315918323017,
+      "learning_rate": 8.516050339764837e-06,
+      "loss": 0.5505,
+      "step": 1818
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.218812448646503,
+      "learning_rate": 8.51431331925662e-06,
+      "loss": 0.4536,
+      "step": 1819
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.460554472978976,
+      "learning_rate": 8.512575460093683e-06,
+      "loss": 0.4345,
+      "step": 1820
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.5248961250705495,
+      "learning_rate": 8.510836762690746e-06,
+      "loss": 0.474,
+      "step": 1821
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.023876765171489,
+      "learning_rate": 8.509097227462732e-06,
+      "loss": 0.5274,
+      "step": 1822
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.280240523966562,
+      "learning_rate": 8.507356854824767e-06,
+      "loss": 0.4937,
+      "step": 1823
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.953024984379047,
+      "learning_rate": 8.505615645192167e-06,
+      "loss": 0.5576,
+      "step": 1824
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.249564398263031,
+      "learning_rate": 8.503873598980456e-06,
+      "loss": 0.5054,
+      "step": 1825
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.17611958988426,
+      "learning_rate": 8.502130716605357e-06,
+      "loss": 0.4489,
+      "step": 1826
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.069451800631331,
+      "learning_rate": 8.50038699848279e-06,
+      "loss": 0.4492,
+      "step": 1827
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.062277612239777,
+      "learning_rate": 8.498642445028874e-06,
+      "loss": 0.4546,
+      "step": 1828
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4205387302616734,
+      "learning_rate": 8.496897056659931e-06,
+      "loss": 0.6011,
+      "step": 1829
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 13.201516928975844,
+      "learning_rate": 8.495150833792478e-06,
+      "loss": 0.5119,
+      "step": 1830
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.601077429374877,
+      "learning_rate": 8.493403776843234e-06,
+      "loss": 0.5678,
+      "step": 1831
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.349263316708461,
+      "learning_rate": 8.491655886229118e-06,
+      "loss": 0.5139,
+      "step": 1832
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 23.935170692534157,
+      "learning_rate": 8.489907162367243e-06,
+      "loss": 0.4797,
+      "step": 1833
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2310350281218412,
+      "learning_rate": 8.488157605674924e-06,
+      "loss": 0.5893,
+      "step": 1834
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.0531365364577745,
+      "learning_rate": 8.486407216569677e-06,
+      "loss": 0.4764,
+      "step": 1835
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.179423152698777,
+      "learning_rate": 8.484655995469214e-06,
+      "loss": 0.4813,
+      "step": 1836
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.704398625161359,
+      "learning_rate": 8.482903942791445e-06,
+      "loss": 0.4769,
+      "step": 1837
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.316044246996018,
+      "learning_rate": 8.48115105895448e-06,
+      "loss": 0.5835,
+      "step": 1838
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.885186855756579,
+      "learning_rate": 8.479397344376626e-06,
+      "loss": 0.4991,
+      "step": 1839
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 12.359084841215594,
+      "learning_rate": 8.477642799476387e-06,
+      "loss": 0.4813,
+      "step": 1840
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.423745871490641,
+      "learning_rate": 8.475887424672471e-06,
+      "loss": 0.4536,
+      "step": 1841
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2671694765264407,
+      "learning_rate": 8.474131220383779e-06,
+      "loss": 0.6233,
+      "step": 1842
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.084892532420517,
+      "learning_rate": 8.47237418702941e-06,
+      "loss": 0.4998,
+      "step": 1843
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 12.93300981373885,
+      "learning_rate": 8.470616325028663e-06,
+      "loss": 0.4361,
+      "step": 1844
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.623951776019602,
+      "learning_rate": 8.468857634801033e-06,
+      "loss": 0.5137,
+      "step": 1845
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.602910540911669,
+      "learning_rate": 8.467098116766213e-06,
+      "loss": 0.5938,
+      "step": 1846
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.850617537819085,
+      "learning_rate": 8.465337771344094e-06,
+      "loss": 0.469,
+      "step": 1847
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 15.208655893721819,
+      "learning_rate": 8.463576598954765e-06,
+      "loss": 0.5211,
+      "step": 1848
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 17.16457890158877,
+      "learning_rate": 8.461814600018512e-06,
+      "loss": 0.5364,
+      "step": 1849
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 13.101825803438341,
+      "learning_rate": 8.460051774955817e-06,
+      "loss": 0.475,
+      "step": 1850
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.550805611292844,
+      "learning_rate": 8.45828812418736e-06,
+      "loss": 0.5078,
+      "step": 1851
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 14.448961568010011,
+      "learning_rate": 8.456523648134018e-06,
+      "loss": 0.462,
+      "step": 1852
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.933825881135313,
+      "learning_rate": 8.454758347216866e-06,
+      "loss": 0.5176,
+      "step": 1853
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 73.74426371723455,
+      "learning_rate": 8.452992221857175e-06,
+      "loss": 0.4896,
+      "step": 1854
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 14.565198831120773,
+      "learning_rate": 8.451225272476413e-06,
+      "loss": 0.4857,
+      "step": 1855
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 13.397318925358146,
+      "learning_rate": 8.44945749949624e-06,
+      "loss": 0.4839,
+      "step": 1856
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.387171058174955,
+      "learning_rate": 8.447688903338521e-06,
+      "loss": 0.4969,
+      "step": 1857
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5556548758571396,
+      "learning_rate": 8.445919484425315e-06,
+      "loss": 0.6326,
+      "step": 1858
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.428192186207272,
+      "learning_rate": 8.444149243178873e-06,
+      "loss": 0.4953,
+      "step": 1859
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.18242170561202,
+      "learning_rate": 8.442378180021644e-06,
+      "loss": 0.4722,
+      "step": 1860
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.2718846612402706,
+      "learning_rate": 8.440606295376274e-06,
+      "loss": 0.443,
+      "step": 1861
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 31.299233178472303,
+      "learning_rate": 8.438833589665606e-06,
+      "loss": 0.4877,
+      "step": 1862
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 14.312259855990655,
+      "learning_rate": 8.43706006331268e-06,
+      "loss": 0.4551,
+      "step": 1863
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.117869103732973,
+      "learning_rate": 8.435285716740726e-06,
+      "loss": 0.5438,
+      "step": 1864
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.687136956511441,
+      "learning_rate": 8.433510550373177e-06,
+      "loss": 0.5039,
+      "step": 1865
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.086505593927276,
+      "learning_rate": 8.431734564633654e-06,
+      "loss": 0.482,
+      "step": 1866
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 17.851117512392186,
+      "learning_rate": 8.429957759945981e-06,
+      "loss": 0.5071,
+      "step": 1867
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.240297693730779,
+      "learning_rate": 8.428180136734176e-06,
+      "loss": 0.5413,
+      "step": 1868
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.473930899582503,
+      "learning_rate": 8.426401695422444e-06,
+      "loss": 0.4375,
+      "step": 1869
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.6846512656853,
+      "learning_rate": 8.424622436435199e-06,
+      "loss": 0.4365,
+      "step": 1870
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 13.679776280358043,
+      "learning_rate": 8.422842360197037e-06,
+      "loss": 0.5022,
+      "step": 1871
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 150.99814688875188,
+      "learning_rate": 8.421061467132761e-06,
+      "loss": 0.473,
+      "step": 1872
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 12.383248393204392,
+      "learning_rate": 8.419279757667356e-06,
+      "loss": 0.519,
+      "step": 1873
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 16.41083440802398,
+      "learning_rate": 8.417497232226013e-06,
+      "loss": 0.4275,
+      "step": 1874
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.215802619605748,
+      "learning_rate": 8.41571389123411e-06,
+      "loss": 0.4904,
+      "step": 1875
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.132820859585701,
+      "learning_rate": 8.413929735117229e-06,
+      "loss": 0.4951,
+      "step": 1876
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.437918597614514,
+      "learning_rate": 8.412144764301133e-06,
+      "loss": 0.4751,
+      "step": 1877
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.584245146489485,
+      "learning_rate": 8.410358979211792e-06,
+      "loss": 0.4664,
+      "step": 1878
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 11.412091687789838,
+      "learning_rate": 8.408572380275365e-06,
+      "loss": 0.422,
+      "step": 1879
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.384494138680635,
+      "learning_rate": 8.406784967918203e-06,
+      "loss": 0.4766,
+      "step": 1880
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.106123974759556,
+      "learning_rate": 8.404996742566855e-06,
+      "loss": 0.4358,
+      "step": 1881
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.07143594434362,
+      "learning_rate": 8.403207704648064e-06,
+      "loss": 0.5811,
+      "step": 1882
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.226107909620538,
+      "learning_rate": 8.401417854588763e-06,
+      "loss": 0.5203,
+      "step": 1883
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.044694800131651,
+      "learning_rate": 8.399627192816082e-06,
+      "loss": 0.4824,
+      "step": 1884
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.049425244878882,
+      "learning_rate": 8.397835719757343e-06,
+      "loss": 0.5795,
+      "step": 1885
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.192805016853793,
+      "learning_rate": 8.396043435840064e-06,
+      "loss": 0.5078,
+      "step": 1886
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.214471471189285,
+      "learning_rate": 8.394250341491958e-06,
+      "loss": 0.4712,
+      "step": 1887
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 12.350337528536405,
+      "learning_rate": 8.392456437140923e-06,
+      "loss": 0.4444,
+      "step": 1888
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.81283897269889,
+      "learning_rate": 8.39066172321506e-06,
+      "loss": 0.4932,
+      "step": 1889
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.287695319714385,
+      "learning_rate": 8.388866200142656e-06,
+      "loss": 0.4708,
+      "step": 1890
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.109766842376803,
+      "learning_rate": 8.387069868352196e-06,
+      "loss": 0.4238,
+      "step": 1891
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.712429590629011,
+      "learning_rate": 8.385272728272356e-06,
+      "loss": 0.4536,
+      "step": 1892
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.013873095311785,
+      "learning_rate": 8.383474780332002e-06,
+      "loss": 0.4801,
+      "step": 1893
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.849634504896088,
+      "learning_rate": 8.381676024960202e-06,
+      "loss": 0.4695,
+      "step": 1894
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.673773777896566,
+      "learning_rate": 8.379876462586202e-06,
+      "loss": 0.5303,
+      "step": 1895
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.223613553587142,
+      "learning_rate": 8.378076093639458e-06,
+      "loss": 0.4541,
+      "step": 1896
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.140242999750354,
+      "learning_rate": 8.376274918549605e-06,
+      "loss": 0.4952,
+      "step": 1897
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.101780926044955,
+      "learning_rate": 8.374472937746474e-06,
+      "loss": 0.4243,
+      "step": 1898
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.443413843434282,
+      "learning_rate": 8.372670151660092e-06,
+      "loss": 0.4903,
+      "step": 1899
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.087927991425739,
+      "learning_rate": 8.370866560720671e-06,
+      "loss": 0.4677,
+      "step": 1900
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.292840443964737,
+      "learning_rate": 8.369062165358626e-06,
+      "loss": 0.551,
+      "step": 1901
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.371315197490912,
+      "learning_rate": 8.367256966004552e-06,
+      "loss": 0.4542,
+      "step": 1902
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.408560536206505,
+      "learning_rate": 8.365450963089242e-06,
+      "loss": 0.4907,
+      "step": 1903
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.912018416396802,
+      "learning_rate": 8.363644157043682e-06,
+      "loss": 0.5776,
+      "step": 1904
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 59.97594752450321,
+      "learning_rate": 8.361836548299047e-06,
+      "loss": 0.477,
+      "step": 1905
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 11.05013248106383,
+      "learning_rate": 8.360028137286701e-06,
+      "loss": 0.5311,
+      "step": 1906
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.186335230654786,
+      "learning_rate": 8.358218924438208e-06,
+      "loss": 0.4646,
+      "step": 1907
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.867807096724809,
+      "learning_rate": 8.356408910185317e-06,
+      "loss": 0.4589,
+      "step": 1908
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.581858110443558,
+      "learning_rate": 8.354598094959965e-06,
+      "loss": 0.4849,
+      "step": 1909
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.212448272381275,
+      "learning_rate": 8.352786479194288e-06,
+      "loss": 0.4546,
+      "step": 1910
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.466302170262632,
+      "learning_rate": 8.350974063320608e-06,
+      "loss": 0.4207,
+      "step": 1911
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.770867287582647,
+      "learning_rate": 8.349160847771442e-06,
+      "loss": 0.4561,
+      "step": 1912
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.13976262285489,
+      "learning_rate": 8.347346832979491e-06,
+      "loss": 0.4789,
+      "step": 1913
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.063852300104583,
+      "learning_rate": 8.345532019377655e-06,
+      "loss": 0.5,
+      "step": 1914
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.668937730923807,
+      "learning_rate": 8.343716407399019e-06,
+      "loss": 0.4153,
+      "step": 1915
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.095286181012478,
+      "learning_rate": 8.341899997476859e-06,
+      "loss": 0.5318,
+      "step": 1916
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.9239370408539074,
+      "learning_rate": 8.340082790044644e-06,
+      "loss": 0.4303,
+      "step": 1917
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.428503042912719,
+      "learning_rate": 8.338264785536032e-06,
+      "loss": 0.4307,
+      "step": 1918
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 13.175547894527906,
+      "learning_rate": 8.33644598438487e-06,
+      "loss": 0.4566,
+      "step": 1919
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.91154551936458,
+      "learning_rate": 8.334626387025196e-06,
+      "loss": 0.3902,
+      "step": 1920
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.185033562289264,
+      "learning_rate": 8.332805993891241e-06,
+      "loss": 0.4078,
+      "step": 1921
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.529501153304833,
+      "learning_rate": 8.330984805417419e-06,
+      "loss": 0.5096,
+      "step": 1922
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 16.859744204279707,
+      "learning_rate": 8.32916282203834e-06,
+      "loss": 0.5028,
+      "step": 1923
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.090610679074903,
+      "learning_rate": 8.327340044188803e-06,
+      "loss": 0.4525,
+      "step": 1924
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2425053351042594,
+      "learning_rate": 8.325516472303792e-06,
+      "loss": 0.5785,
+      "step": 1925
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.319463705588319,
+      "learning_rate": 8.323692106818486e-06,
+      "loss": 0.4281,
+      "step": 1926
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.479883960634091,
+      "learning_rate": 8.321866948168252e-06,
+      "loss": 0.4025,
+      "step": 1927
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.564403959032149,
+      "learning_rate": 8.320040996788641e-06,
+      "loss": 0.4877,
+      "step": 1928
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.55778089581736,
+      "learning_rate": 8.3182142531154e-06,
+      "loss": 0.4558,
+      "step": 1929
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.088171622900685,
+      "learning_rate": 8.316386717584463e-06,
+      "loss": 0.4797,
+      "step": 1930
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.388169511492117,
+      "learning_rate": 8.314558390631954e-06,
+      "loss": 0.4528,
+      "step": 1931
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.697441883225904,
+      "learning_rate": 8.312729272694184e-06,
+      "loss": 0.5191,
+      "step": 1932
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.525395999412714,
+      "learning_rate": 8.31089936420765e-06,
+      "loss": 0.5245,
+      "step": 1933
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.53724236000614,
+      "learning_rate": 8.309068665609043e-06,
+      "loss": 0.4804,
+      "step": 1934
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.5059799080755116,
+      "learning_rate": 8.307237177335238e-06,
+      "loss": 0.4682,
+      "step": 1935
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.9493644428581867,
+      "learning_rate": 8.305404899823308e-06,
+      "loss": 0.4159,
+      "step": 1936
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.584174767082249,
+      "learning_rate": 8.303571833510498e-06,
+      "loss": 0.5067,
+      "step": 1937
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.873649077967587,
+      "learning_rate": 8.301737978834257e-06,
+      "loss": 0.4753,
+      "step": 1938
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 11.409539324638367,
+      "learning_rate": 8.299903336232212e-06,
+      "loss": 0.5078,
+      "step": 1939
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.826054198120511,
+      "learning_rate": 8.298067906142182e-06,
+      "loss": 0.4995,
+      "step": 1940
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.586393167785583,
+      "learning_rate": 8.296231689002176e-06,
+      "loss": 0.5092,
+      "step": 1941
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.5385528280024623,
+      "learning_rate": 8.294394685250386e-06,
+      "loss": 0.4935,
+      "step": 1942
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.793228514442019,
+      "learning_rate": 8.292556895325195e-06,
+      "loss": 0.4801,
+      "step": 1943
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.812104874920025,
+      "learning_rate": 8.29071831966517e-06,
+      "loss": 0.4289,
+      "step": 1944
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.450796437518967,
+      "learning_rate": 8.288878958709072e-06,
+      "loss": 0.4563,
+      "step": 1945
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.6190766930547618,
+      "learning_rate": 8.287038812895845e-06,
+      "loss": 0.5445,
+      "step": 1946
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.100854100676929,
+      "learning_rate": 8.28519788266462e-06,
+      "loss": 0.4698,
+      "step": 1947
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1853678960862974,
+      "learning_rate": 8.283356168454715e-06,
+      "loss": 0.4519,
+      "step": 1948
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.803452130620062,
+      "learning_rate": 8.281513670705637e-06,
+      "loss": 0.4749,
+      "step": 1949
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.627115010571595,
+      "learning_rate": 8.27967038985708e-06,
+      "loss": 0.554,
+      "step": 1950
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.8611967206334916,
+      "learning_rate": 8.277826326348921e-06,
+      "loss": 0.5131,
+      "step": 1951
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.6442814803644947,
+      "learning_rate": 8.275981480621232e-06,
+      "loss": 0.4688,
+      "step": 1952
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5180105780469244,
+      "learning_rate": 8.274135853114263e-06,
+      "loss": 0.4834,
+      "step": 1953
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.0693467847055547,
+      "learning_rate": 8.272289444268452e-06,
+      "loss": 0.4519,
+      "step": 1954
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.362765111180697,
+      "learning_rate": 8.270442254524432e-06,
+      "loss": 0.5201,
+      "step": 1955
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.3921777883462862,
+      "learning_rate": 8.268594284323008e-06,
+      "loss": 0.4216,
+      "step": 1956
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.500942070745847,
+      "learning_rate": 8.266745534105185e-06,
+      "loss": 0.4955,
+      "step": 1957
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.255584062934611,
+      "learning_rate": 8.264896004312147e-06,
+      "loss": 0.498,
+      "step": 1958
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.2127037395125395,
+      "learning_rate": 8.26304569538526e-06,
+      "loss": 0.521,
+      "step": 1959
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.700747247072509,
+      "learning_rate": 8.261194607766089e-06,
+      "loss": 0.447,
+      "step": 1960
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4751087350130125,
+      "learning_rate": 8.259342741896376e-06,
+      "loss": 0.6339,
+      "step": 1961
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.2214553913778934,
+      "learning_rate": 8.257490098218044e-06,
+      "loss": 0.4278,
+      "step": 1962
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.338134168787052,
+      "learning_rate": 8.255636677173212e-06,
+      "loss": 0.4849,
+      "step": 1963
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1859409457637344,
+      "learning_rate": 8.253782479204178e-06,
+      "loss": 0.4765,
+      "step": 1964
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5117292554240382,
+      "learning_rate": 8.251927504753427e-06,
+      "loss": 0.485,
+      "step": 1965
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.4926681944018276,
+      "learning_rate": 8.25007175426363e-06,
+      "loss": 0.5457,
+      "step": 1966
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.502478848056594,
+      "learning_rate": 8.248215228177645e-06,
+      "loss": 0.4623,
+      "step": 1967
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.578073808552931,
+      "learning_rate": 8.24635792693851e-06,
+      "loss": 0.5076,
+      "step": 1968
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.4154922910540093,
+      "learning_rate": 8.244499850989453e-06,
+      "loss": 0.5006,
+      "step": 1969
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.909656340943361,
+      "learning_rate": 8.242641000773882e-06,
+      "loss": 0.5305,
+      "step": 1970
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8961743673901017,
+      "learning_rate": 8.240781376735395e-06,
+      "loss": 0.4628,
+      "step": 1971
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.917908945541841,
+      "learning_rate": 8.23892097931777e-06,
+      "loss": 0.4598,
+      "step": 1972
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.741156802336023,
+      "learning_rate": 8.237059808964975e-06,
+      "loss": 0.4473,
+      "step": 1973
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.032668712719807,
+      "learning_rate": 8.235197866121155e-06,
+      "loss": 0.5037,
+      "step": 1974
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.0185136848240717,
+      "learning_rate": 8.233335151230645e-06,
+      "loss": 0.472,
+      "step": 1975
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 12.429646674056704,
+      "learning_rate": 8.231471664737967e-06,
+      "loss": 0.4228,
+      "step": 1976
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.228851602991045,
+      "learning_rate": 8.229607407087816e-06,
+      "loss": 0.5216,
+      "step": 1977
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.5526744296993145,
+      "learning_rate": 8.227742378725082e-06,
+      "loss": 0.4672,
+      "step": 1978
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.882425841669333,
+      "learning_rate": 8.225876580094833e-06,
+      "loss": 0.498,
+      "step": 1979
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.087778652514507,
+      "learning_rate": 8.224010011642326e-06,
+      "loss": 0.4754,
+      "step": 1980
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.187336347503646,
+      "learning_rate": 8.222142673812995e-06,
+      "loss": 0.5261,
+      "step": 1981
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.523961548911623,
+      "learning_rate": 8.220274567052461e-06,
+      "loss": 0.5126,
+      "step": 1982
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.5052179445481246,
+      "learning_rate": 8.21840569180653e-06,
+      "loss": 0.5107,
+      "step": 1983
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.0612608496042757,
+      "learning_rate": 8.216536048521188e-06,
+      "loss": 0.4819,
+      "step": 1984
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.6089297594096084,
+      "learning_rate": 8.21466563764261e-06,
+      "loss": 0.4708,
+      "step": 1985
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.026456647855598,
+      "learning_rate": 8.212794459617144e-06,
+      "loss": 0.4199,
+      "step": 1986
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.656058538823349,
+      "learning_rate": 8.210922514891333e-06,
+      "loss": 0.5044,
+      "step": 1987
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.9403317306778742,
+      "learning_rate": 8.209049803911895e-06,
+      "loss": 0.4356,
+      "step": 1988
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.5577057754780164,
+      "learning_rate": 8.207176327125735e-06,
+      "loss": 0.5261,
+      "step": 1989
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.407363825377123,
+      "learning_rate": 8.205302084979937e-06,
+      "loss": 0.4201,
+      "step": 1990
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.504710866212139,
+      "learning_rate": 8.203427077921768e-06,
+      "loss": 0.5056,
+      "step": 1991
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.571440377051897,
+      "learning_rate": 8.201551306398685e-06,
+      "loss": 0.4444,
+      "step": 1992
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5763005484626396,
+      "learning_rate": 8.199674770858315e-06,
+      "loss": 0.5007,
+      "step": 1993
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.670484043552327,
+      "learning_rate": 8.197797471748478e-06,
+      "loss": 0.5201,
+      "step": 1994
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.552146469875878,
+      "learning_rate": 8.19591940951717e-06,
+      "loss": 0.5054,
+      "step": 1995
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.3374418475358905,
+      "learning_rate": 8.194040584612573e-06,
+      "loss": 0.5143,
+      "step": 1996
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.353304581907365,
+      "learning_rate": 8.19216099748305e-06,
+      "loss": 0.5117,
+      "step": 1997
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.4034030542110356,
+      "learning_rate": 8.19028064857714e-06,
+      "loss": 0.5334,
+      "step": 1998
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.3089755476199203,
+      "learning_rate": 8.188399538343574e-06,
+      "loss": 0.5027,
+      "step": 1999
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.868066876994381,
+      "learning_rate": 8.18651766723126e-06,
+      "loss": 0.481,
+      "step": 2000
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.767863292317751,
+      "learning_rate": 8.184635035689283e-06,
+      "loss": 0.4768,
+      "step": 2001
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.372190543989447,
+      "learning_rate": 8.182751644166915e-06,
+      "loss": 0.3999,
+      "step": 2002
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.446942437485215,
+      "learning_rate": 8.180867493113611e-06,
+      "loss": 0.5254,
+      "step": 2003
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.214872549516012,
+      "learning_rate": 8.178982582979001e-06,
+      "loss": 0.5101,
+      "step": 2004
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.269443902190875,
+      "learning_rate": 8.177096914212901e-06,
+      "loss": 0.4826,
+      "step": 2005
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.19947193220622,
+      "learning_rate": 8.175210487265307e-06,
+      "loss": 0.4632,
+      "step": 2006
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.6315303257785265,
+      "learning_rate": 8.173323302586393e-06,
+      "loss": 0.5393,
+      "step": 2007
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.239807957951117,
+      "learning_rate": 8.171435360626519e-06,
+      "loss": 0.5029,
+      "step": 2008
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.537615747047689,
+      "learning_rate": 8.16954666183622e-06,
+      "loss": 0.4415,
+      "step": 2009
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.150769027654825,
+      "learning_rate": 8.167657206666217e-06,
+      "loss": 0.5134,
+      "step": 2010
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.5656512067934027,
+      "learning_rate": 8.165766995567408e-06,
+      "loss": 0.5854,
+      "step": 2011
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.013288406417272,
+      "learning_rate": 8.16387602899087e-06,
+      "loss": 0.5072,
+      "step": 2012
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.8652192956738816,
+      "learning_rate": 8.161984307387869e-06,
+      "loss": 0.5515,
+      "step": 2013
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5103800302150816,
+      "learning_rate": 8.16009183120984e-06,
+      "loss": 0.4655,
+      "step": 2014
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.4720729120187475,
+      "learning_rate": 8.158198600908404e-06,
+      "loss": 0.4939,
+      "step": 2015
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.958453184624063,
+      "learning_rate": 8.156304616935362e-06,
+      "loss": 0.5417,
+      "step": 2016
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.846866877369375,
+      "learning_rate": 8.15440987974269e-06,
+      "loss": 0.6142,
+      "step": 2017
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.76397675643439,
+      "learning_rate": 8.152514389782553e-06,
+      "loss": 0.4306,
+      "step": 2018
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.308392708126946,
+      "learning_rate": 8.150618147507288e-06,
+      "loss": 0.5295,
+      "step": 2019
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.486768825300372,
+      "learning_rate": 8.148721153369413e-06,
+      "loss": 0.4603,
+      "step": 2020
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.716408944567494,
+      "learning_rate": 8.146823407821625e-06,
+      "loss": 0.4265,
+      "step": 2021
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.3296597656415257,
+      "learning_rate": 8.144924911316803e-06,
+      "loss": 0.4787,
+      "step": 2022
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.990574664500636,
+      "learning_rate": 8.143025664308005e-06,
+      "loss": 0.4943,
+      "step": 2023
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.1312951518334695,
+      "learning_rate": 8.141125667248463e-06,
+      "loss": 0.5317,
+      "step": 2024
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.681245116193845,
+      "learning_rate": 8.139224920591598e-06,
+      "loss": 0.4425,
+      "step": 2025
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.8267613626899046,
+      "learning_rate": 8.137323424790996e-06,
+      "loss": 0.4611,
+      "step": 2026
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7696190903946976,
+      "learning_rate": 8.135421180300435e-06,
+      "loss": 0.5465,
+      "step": 2027
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.4065612451213854,
+      "learning_rate": 8.133518187573864e-06,
+      "loss": 0.4431,
+      "step": 2028
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.7174488705796107,
+      "learning_rate": 8.131614447065411e-06,
+      "loss": 0.4446,
+      "step": 2029
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.050472334301364,
+      "learning_rate": 8.129709959229389e-06,
+      "loss": 0.4785,
+      "step": 2030
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.4351775185835813,
+      "learning_rate": 8.127804724520278e-06,
+      "loss": 0.5031,
+      "step": 2031
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.007100437839703,
+      "learning_rate": 8.125898743392748e-06,
+      "loss": 0.4808,
+      "step": 2032
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5586077632724835,
+      "learning_rate": 8.12399201630164e-06,
+      "loss": 0.6114,
+      "step": 2033
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.376235176776049,
+      "learning_rate": 8.122084543701974e-06,
+      "loss": 0.4648,
+      "step": 2034
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.6721041804343337,
+      "learning_rate": 8.12017632604895e-06,
+      "loss": 0.4845,
+      "step": 2035
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.681413460958966,
+      "learning_rate": 8.118267363797943e-06,
+      "loss": 0.5184,
+      "step": 2036
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.731296360387705,
+      "learning_rate": 8.116357657404508e-06,
+      "loss": 0.4976,
+      "step": 2037
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.569076338376435,
+      "learning_rate": 8.114447207324377e-06,
+      "loss": 0.4644,
+      "step": 2038
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.791417564080911,
+      "learning_rate": 8.11253601401346e-06,
+      "loss": 0.5108,
+      "step": 2039
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.444381450521485,
+      "learning_rate": 8.110624077927841e-06,
+      "loss": 0.4431,
+      "step": 2040
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.435879462004735,
+      "learning_rate": 8.108711399523787e-06,
+      "loss": 0.4465,
+      "step": 2041
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.582886714480457,
+      "learning_rate": 8.106797979257738e-06,
+      "loss": 0.4837,
+      "step": 2042
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.756466978791274,
+      "learning_rate": 8.10488381758631e-06,
+      "loss": 0.45,
+      "step": 2043
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.009929127273303,
+      "learning_rate": 8.1029689149663e-06,
+      "loss": 0.5163,
+      "step": 2044
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.113618177217519,
+      "learning_rate": 8.101053271854681e-06,
+      "loss": 0.4631,
+      "step": 2045
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.933276270302843,
+      "learning_rate": 8.099136888708598e-06,
+      "loss": 0.507,
+      "step": 2046
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.4842137529285875,
+      "learning_rate": 8.09721976598538e-06,
+      "loss": 0.5072,
+      "step": 2047
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8269996026018434,
+      "learning_rate": 8.095301904142526e-06,
+      "loss": 0.4863,
+      "step": 2048
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.024216530452258,
+      "learning_rate": 8.093383303637715e-06,
+      "loss": 0.4922,
+      "step": 2049
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.7200555831604305,
+      "learning_rate": 8.091463964928801e-06,
+      "loss": 0.4707,
+      "step": 2050
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.834822972876296,
+      "learning_rate": 8.089543888473814e-06,
+      "loss": 0.4926,
+      "step": 2051
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.426336026688418,
+      "learning_rate": 8.08762307473096e-06,
+      "loss": 0.467,
+      "step": 2052
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.5505786325782305,
+      "learning_rate": 8.085701524158623e-06,
+      "loss": 0.5343,
+      "step": 2053
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.977518940312848,
+      "learning_rate": 8.083779237215363e-06,
+      "loss": 0.4399,
+      "step": 2054
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.311429715090669,
+      "learning_rate": 8.08185621435991e-06,
+      "loss": 0.4652,
+      "step": 2055
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.255482934946208,
+      "learning_rate": 8.079932456051173e-06,
+      "loss": 0.4484,
+      "step": 2056
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.1025038690054565,
+      "learning_rate": 8.078007962748243e-06,
+      "loss": 0.454,
+      "step": 2057
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.484045824534379,
+      "learning_rate": 8.076082734910374e-06,
+      "loss": 0.4777,
+      "step": 2058
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.70323585591674,
+      "learning_rate": 8.074156772997005e-06,
+      "loss": 0.4541,
+      "step": 2059
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.660005945017618,
+      "learning_rate": 8.072230077467749e-06,
+      "loss": 0.5198,
+      "step": 2060
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.804770453324357,
+      "learning_rate": 8.070302648782389e-06,
+      "loss": 0.4886,
+      "step": 2061
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.038516545225295,
+      "learning_rate": 8.068374487400885e-06,
+      "loss": 0.4563,
+      "step": 2062
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.090637895257619,
+      "learning_rate": 8.066445593783377e-06,
+      "loss": 0.4443,
+      "step": 2063
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.150128018373972,
+      "learning_rate": 8.064515968390175e-06,
+      "loss": 0.4921,
+      "step": 2064
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.375706286153132,
+      "learning_rate": 8.062585611681758e-06,
+      "loss": 0.4992,
+      "step": 2065
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.8341800901534477,
+      "learning_rate": 8.060654524118793e-06,
+      "loss": 0.4824,
+      "step": 2066
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.327680655176633,
+      "learning_rate": 8.058722706162112e-06,
+      "loss": 0.4786,
+      "step": 2067
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.437113712214887,
+      "learning_rate": 8.05679015827272e-06,
+      "loss": 0.5071,
+      "step": 2068
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.543795078075457,
+      "learning_rate": 8.054856880911806e-06,
+      "loss": 0.4857,
+      "step": 2069
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.009692575470058,
+      "learning_rate": 8.052922874540722e-06,
+      "loss": 0.5398,
+      "step": 2070
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.408972246639082,
+      "learning_rate": 8.050988139621001e-06,
+      "loss": 0.445,
+      "step": 2071
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.3962576168139185,
+      "learning_rate": 8.049052676614346e-06,
+      "loss": 0.4225,
+      "step": 2072
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.576044474025559,
+      "learning_rate": 8.047116485982638e-06,
+      "loss": 0.4897,
+      "step": 2073
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.573494057779335,
+      "learning_rate": 8.045179568187925e-06,
+      "loss": 0.4982,
+      "step": 2074
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.952367989162258,
+      "learning_rate": 8.043241923692436e-06,
+      "loss": 0.417,
+      "step": 2075
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.4992245110181988,
+      "learning_rate": 8.041303552958567e-06,
+      "loss": 0.472,
+      "step": 2076
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.292086418026972,
+      "learning_rate": 8.039364456448892e-06,
+      "loss": 0.5095,
+      "step": 2077
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.679280125003664,
+      "learning_rate": 8.037424634626157e-06,
+      "loss": 0.4253,
+      "step": 2078
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.922984905593826,
+      "learning_rate": 8.035484087953278e-06,
+      "loss": 0.538,
+      "step": 2079
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.818958788630032,
+      "learning_rate": 8.03354281689335e-06,
+      "loss": 0.4862,
+      "step": 2080
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.589804268321271,
+      "learning_rate": 8.031600821909636e-06,
+      "loss": 0.5003,
+      "step": 2081
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.516644248174846,
+      "learning_rate": 8.029658103465571e-06,
+      "loss": 0.5076,
+      "step": 2082
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.502339125545945,
+      "learning_rate": 8.027714662024768e-06,
+      "loss": 0.5039,
+      "step": 2083
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.663590284462899,
+      "learning_rate": 8.025770498051007e-06,
+      "loss": 0.4998,
+      "step": 2084
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.695407621501322,
+      "learning_rate": 8.023825612008243e-06,
+      "loss": 0.4742,
+      "step": 2085
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.374660107052677,
+      "learning_rate": 8.021880004360604e-06,
+      "loss": 0.4294,
+      "step": 2086
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2727153203765817,
+      "learning_rate": 8.019933675572389e-06,
+      "loss": 0.5777,
+      "step": 2087
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.6615253900440425,
+      "learning_rate": 8.01798662610807e-06,
+      "loss": 0.5448,
+      "step": 2088
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.530246365631273,
+      "learning_rate": 8.016038856432287e-06,
+      "loss": 0.4614,
+      "step": 2089
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.065180473846315,
+      "learning_rate": 8.01409036700986e-06,
+      "loss": 0.4679,
+      "step": 2090
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.371048369007121,
+      "learning_rate": 8.012141158305773e-06,
+      "loss": 0.5205,
+      "step": 2091
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.587327698869069,
+      "learning_rate": 8.010191230785184e-06,
+      "loss": 0.5486,
+      "step": 2092
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.821258332716041,
+      "learning_rate": 8.008240584913426e-06,
+      "loss": 0.508,
+      "step": 2093
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.011982498831261,
+      "learning_rate": 8.006289221155998e-06,
+      "loss": 0.6289,
+      "step": 2094
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.878147474332953,
+      "learning_rate": 8.004337139978575e-06,
+      "loss": 0.5477,
+      "step": 2095
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.282814793944137,
+      "learning_rate": 8.002384341846998e-06,
+      "loss": 0.5185,
+      "step": 2096
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.342754365287702,
+      "learning_rate": 8.000430827227285e-06,
+      "loss": 0.543,
+      "step": 2097
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.445318457807897,
+      "learning_rate": 7.998476596585622e-06,
+      "loss": 0.492,
+      "step": 2098
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.047950773267699,
+      "learning_rate": 7.996521650388365e-06,
+      "loss": 0.4311,
+      "step": 2099
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.51031738832105,
+      "learning_rate": 7.994565989102042e-06,
+      "loss": 0.4892,
+      "step": 2100
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.7394656305504292,
+      "learning_rate": 7.992609613193351e-06,
+      "loss": 0.45,
+      "step": 2101
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.070023113346108,
+      "learning_rate": 7.990652523129163e-06,
+      "loss": 0.5129,
+      "step": 2102
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.837699230467814,
+      "learning_rate": 7.988694719376516e-06,
+      "loss": 0.5216,
+      "step": 2103
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.500655472133036,
+      "learning_rate": 7.986736202402617e-06,
+      "loss": 0.4732,
+      "step": 2104
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.040471246124025,
+      "learning_rate": 7.98477697267485e-06,
+      "loss": 0.5288,
+      "step": 2105
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.184036126587419,
+      "learning_rate": 7.982817030660766e-06,
+      "loss": 0.4802,
+      "step": 2106
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.128219390586616,
+      "learning_rate": 7.980856376828078e-06,
+      "loss": 0.4946,
+      "step": 2107
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.718767210908153,
+      "learning_rate": 7.978895011644686e-06,
+      "loss": 0.5292,
+      "step": 2108
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.639627820588555,
+      "learning_rate": 7.97693293557864e-06,
+      "loss": 0.4575,
+      "step": 2109
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.486877771308906,
+      "learning_rate": 7.974970149098175e-06,
+      "loss": 0.4639,
+      "step": 2110
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.861783287525592,
+      "learning_rate": 7.973006652671686e-06,
+      "loss": 0.4708,
+      "step": 2111
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.028232489728966,
+      "learning_rate": 7.971042446767747e-06,
+      "loss": 0.444,
+      "step": 2112
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.772587799046671,
+      "learning_rate": 7.96907753185509e-06,
+      "loss": 0.4705,
+      "step": 2113
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.84025197652857,
+      "learning_rate": 7.967111908402623e-06,
+      "loss": 0.499,
+      "step": 2114
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.356241538060642,
+      "learning_rate": 7.965145576879425e-06,
+      "loss": 0.5044,
+      "step": 2115
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.807750504920291,
+      "learning_rate": 7.963178537754735e-06,
+      "loss": 0.4851,
+      "step": 2116
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.020002449227239,
+      "learning_rate": 7.961210791497972e-06,
+      "loss": 0.4761,
+      "step": 2117
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.8999533187777,
+      "learning_rate": 7.959242338578717e-06,
+      "loss": 0.4742,
+      "step": 2118
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 31.523979780068863,
+      "learning_rate": 7.957273179466719e-06,
+      "loss": 0.5224,
+      "step": 2119
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.55498993934344,
+      "learning_rate": 7.955303314631899e-06,
+      "loss": 0.4993,
+      "step": 2120
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.090650539062269,
+      "learning_rate": 7.953332744544344e-06,
+      "loss": 0.3998,
+      "step": 2121
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.727608535891911,
+      "learning_rate": 7.951361469674312e-06,
+      "loss": 0.5469,
+      "step": 2122
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.446840171682057,
+      "learning_rate": 7.949389490492228e-06,
+      "loss": 0.5006,
+      "step": 2123
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.166113422618472,
+      "learning_rate": 7.947416807468682e-06,
+      "loss": 0.5453,
+      "step": 2124
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.692659368011256,
+      "learning_rate": 7.945443421074436e-06,
+      "loss": 0.4698,
+      "step": 2125
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.379013079725918,
+      "learning_rate": 7.943469331780418e-06,
+      "loss": 0.4651,
+      "step": 2126
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.360336372036317,
+      "learning_rate": 7.941494540057723e-06,
+      "loss": 0.5045,
+      "step": 2127
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.8798164474051116,
+      "learning_rate": 7.93951904637762e-06,
+      "loss": 0.4779,
+      "step": 2128
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.586178382440598,
+      "learning_rate": 7.937542851211533e-06,
+      "loss": 0.4801,
+      "step": 2129
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.256108382607711,
+      "learning_rate": 7.935565955031064e-06,
+      "loss": 0.5116,
+      "step": 2130
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 11.640907416460164,
+      "learning_rate": 7.933588358307978e-06,
+      "loss": 0.5381,
+      "step": 2131
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.954041253881225,
+      "learning_rate": 7.93161006151421e-06,
+      "loss": 0.5127,
+      "step": 2132
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.240094312523461,
+      "learning_rate": 7.929631065121859e-06,
+      "loss": 0.5015,
+      "step": 2133
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3686261349287927,
+      "learning_rate": 7.927651369603192e-06,
+      "loss": 0.6195,
+      "step": 2134
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.316009964017145,
+      "learning_rate": 7.925670975430644e-06,
+      "loss": 0.4647,
+      "step": 2135
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.770136480265396,
+      "learning_rate": 7.923689883076813e-06,
+      "loss": 0.3973,
+      "step": 2136
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 24.42306528823531,
+      "learning_rate": 7.92170809301447e-06,
+      "loss": 0.4305,
+      "step": 2137
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.8943321634780705,
+      "learning_rate": 7.919725605716547e-06,
+      "loss": 0.4623,
+      "step": 2138
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.982860020072937,
+      "learning_rate": 7.917742421656142e-06,
+      "loss": 0.4728,
+      "step": 2139
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.35216111816546,
+      "learning_rate": 7.915758541306523e-06,
+      "loss": 0.5555,
+      "step": 2140
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.736581155957992,
+      "learning_rate": 7.913773965141126e-06,
+      "loss": 0.4504,
+      "step": 2141
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.785217131705203,
+      "learning_rate": 7.911788693633546e-06,
+      "loss": 0.5016,
+      "step": 2142
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.90801967389207,
+      "learning_rate": 7.90980272725755e-06,
+      "loss": 0.4722,
+      "step": 2143
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 22.730499328387808,
+      "learning_rate": 7.907816066487065e-06,
+      "loss": 0.5073,
+      "step": 2144
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.78894965584396,
+      "learning_rate": 7.90582871179619e-06,
+      "loss": 0.4589,
+      "step": 2145
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.261887595308605,
+      "learning_rate": 7.903840663659186e-06,
+      "loss": 0.4717,
+      "step": 2146
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.548216292368708,
+      "learning_rate": 7.90185192255048e-06,
+      "loss": 0.4711,
+      "step": 2147
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.565504178454162,
+      "learning_rate": 7.899862488944664e-06,
+      "loss": 0.4094,
+      "step": 2148
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.49578837097961,
+      "learning_rate": 7.897872363316496e-06,
+      "loss": 0.3931,
+      "step": 2149
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.45909877112897,
+      "learning_rate": 7.895881546140903e-06,
+      "loss": 0.5676,
+      "step": 2150
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.328625233123079,
+      "learning_rate": 7.893890037892966e-06,
+      "loss": 0.512,
+      "step": 2151
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.042434989107127,
+      "learning_rate": 7.891897839047944e-06,
+      "loss": 0.4762,
+      "step": 2152
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.744368777257508,
+      "learning_rate": 7.889904950081252e-06,
+      "loss": 0.4937,
+      "step": 2153
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.674849416026325,
+      "learning_rate": 7.887911371468474e-06,
+      "loss": 0.5436,
+      "step": 2154
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.013054728693937,
+      "learning_rate": 7.885917103685354e-06,
+      "loss": 0.4946,
+      "step": 2155
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.551857759603399,
+      "learning_rate": 7.883922147207805e-06,
+      "loss": 0.4562,
+      "step": 2156
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.031393922642984,
+      "learning_rate": 7.881926502511906e-06,
+      "loss": 0.4254,
+      "step": 2157
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.71574286855643,
+      "learning_rate": 7.879930170073891e-06,
+      "loss": 0.4783,
+      "step": 2158
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.616650024329564,
+      "learning_rate": 7.877933150370168e-06,
+      "loss": 0.5097,
+      "step": 2159
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.863964779273687,
+      "learning_rate": 7.875935443877305e-06,
+      "loss": 0.405,
+      "step": 2160
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.196974992329872,
+      "learning_rate": 7.873937051072037e-06,
+      "loss": 0.4858,
+      "step": 2161
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.16667125838661,
+      "learning_rate": 7.871937972431253e-06,
+      "loss": 0.5635,
+      "step": 2162
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.719646713991938,
+      "learning_rate": 7.869938208432016e-06,
+      "loss": 0.4877,
+      "step": 2163
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.1988541600288,
+      "learning_rate": 7.867937759551551e-06,
+      "loss": 0.5041,
+      "step": 2164
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.0627969480865875,
+      "learning_rate": 7.865936626267244e-06,
+      "loss": 0.4481,
+      "step": 2165
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.260221441621175,
+      "learning_rate": 7.86393480905664e-06,
+      "loss": 0.4045,
+      "step": 2166
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.323204477134403,
+      "learning_rate": 7.86193230839746e-06,
+      "loss": 0.4774,
+      "step": 2167
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.634169001136357,
+      "learning_rate": 7.859929124767574e-06,
+      "loss": 0.4664,
+      "step": 2168
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.307725787834439,
+      "learning_rate": 7.857925258645022e-06,
+      "loss": 0.508,
+      "step": 2169
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.75387980271713,
+      "learning_rate": 7.85592071050801e-06,
+      "loss": 0.5495,
+      "step": 2170
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.4334479081515505,
+      "learning_rate": 7.853915480834897e-06,
+      "loss": 0.5304,
+      "step": 2171
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.610796011653907,
+      "learning_rate": 7.851909570104215e-06,
+      "loss": 0.4874,
+      "step": 2172
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.46111470450064,
+      "learning_rate": 7.849902978794653e-06,
+      "loss": 0.4471,
+      "step": 2173
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.3407789685569433,
+      "learning_rate": 7.847895707385063e-06,
+      "loss": 0.4945,
+      "step": 2174
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.6494318295919395,
+      "learning_rate": 7.845887756354458e-06,
+      "loss": 0.4904,
+      "step": 2175
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.124591521678854,
+      "learning_rate": 7.843879126182017e-06,
+      "loss": 0.5073,
+      "step": 2176
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.254035371712426,
+      "learning_rate": 7.841869817347079e-06,
+      "loss": 0.4834,
+      "step": 2177
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.093151501267945,
+      "learning_rate": 7.839859830329141e-06,
+      "loss": 0.4956,
+      "step": 2178
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.940168631288694,
+      "learning_rate": 7.837849165607872e-06,
+      "loss": 0.4748,
+      "step": 2179
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.687471595529828,
+      "learning_rate": 7.835837823663093e-06,
+      "loss": 0.415,
+      "step": 2180
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.697200784902237,
+      "learning_rate": 7.833825804974791e-06,
+      "loss": 0.5021,
+      "step": 2181
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.271349965181905,
+      "learning_rate": 7.831813110023113e-06,
+      "loss": 0.5292,
+      "step": 2182
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 18.93701093774531,
+      "learning_rate": 7.829799739288364e-06,
+      "loss": 0.478,
+      "step": 2183
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 17.014321834641205,
+      "learning_rate": 7.827785693251023e-06,
+      "loss": 0.5284,
+      "step": 2184
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.866402131474855,
+      "learning_rate": 7.825770972391713e-06,
+      "loss": 0.4753,
+      "step": 2185
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.236777398818386,
+      "learning_rate": 7.82375557719123e-06,
+      "loss": 0.4717,
+      "step": 2186
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.504396751582203,
+      "learning_rate": 7.821739508130528e-06,
+      "loss": 0.6661,
+      "step": 2187
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.9820391702594624,
+      "learning_rate": 7.819722765690723e-06,
+      "loss": 0.4298,
+      "step": 2188
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.7361601999847975,
+      "learning_rate": 7.817705350353085e-06,
+      "loss": 0.4839,
+      "step": 2189
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.338259722341007,
+      "learning_rate": 7.81568726259905e-06,
+      "loss": 0.5113,
+      "step": 2190
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.59208814913395,
+      "learning_rate": 7.813668502910217e-06,
+      "loss": 0.4855,
+      "step": 2191
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.172904759245342,
+      "learning_rate": 7.811649071768342e-06,
+      "loss": 0.4979,
+      "step": 2192
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1817078946551491,
+      "learning_rate": 7.809628969655338e-06,
+      "loss": 0.5781,
+      "step": 2193
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.7623441585259623,
+      "learning_rate": 7.807608197053284e-06,
+      "loss": 0.5057,
+      "step": 2194
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.863023361224732,
+      "learning_rate": 7.805586754444416e-06,
+      "loss": 0.4687,
+      "step": 2195
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.692465100577142,
+      "learning_rate": 7.803564642311133e-06,
+      "loss": 0.4898,
+      "step": 2196
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.376862031115437,
+      "learning_rate": 7.801541861135988e-06,
+      "loss": 0.4951,
+      "step": 2197
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.115803755422277,
+      "learning_rate": 7.799518411401698e-06,
+      "loss": 0.4707,
+      "step": 2198
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.331988467054175,
+      "learning_rate": 7.797494293591138e-06,
+      "loss": 0.4712,
+      "step": 2199
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.645756969254886,
+      "learning_rate": 7.795469508187343e-06,
+      "loss": 0.4239,
+      "step": 2200
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.764241281048958,
+      "learning_rate": 7.79344405567351e-06,
+      "loss": 0.4836,
+      "step": 2201
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.45291251264358,
+      "learning_rate": 7.791417936532987e-06,
+      "loss": 0.4405,
+      "step": 2202
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.530735790795023,
+      "learning_rate": 7.78939115124929e-06,
+      "loss": 0.4865,
+      "step": 2203
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.895862206372964,
+      "learning_rate": 7.787363700306092e-06,
+      "loss": 0.5127,
+      "step": 2204
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.7095273962399045,
+      "learning_rate": 7.78533558418722e-06,
+      "loss": 0.4805,
+      "step": 2205
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.300751484560304,
+      "learning_rate": 7.783306803376664e-06,
+      "loss": 0.5849,
+      "step": 2206
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.410188735106823,
+      "learning_rate": 7.781277358358573e-06,
+      "loss": 0.444,
+      "step": 2207
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.003925819727987,
+      "learning_rate": 7.77924724961725e-06,
+      "loss": 0.4389,
+      "step": 2208
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.498274220337695,
+      "learning_rate": 7.777216477637162e-06,
+      "loss": 0.3882,
+      "step": 2209
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.211067056646403,
+      "learning_rate": 7.775185042902932e-06,
+      "loss": 0.5186,
+      "step": 2210
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.372237635794461,
+      "learning_rate": 7.773152945899343e-06,
+      "loss": 0.4679,
+      "step": 2211
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 11.969806409589777,
+      "learning_rate": 7.77112018711133e-06,
+      "loss": 0.443,
+      "step": 2212
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.113006032377814,
+      "learning_rate": 7.76908676702399e-06,
+      "loss": 0.5581,
+      "step": 2213
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.381632691074837,
+      "learning_rate": 7.767052686122582e-06,
+      "loss": 0.453,
+      "step": 2214
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.903511901982955,
+      "learning_rate": 7.765017944892513e-06,
+      "loss": 0.459,
+      "step": 2215
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.167734325882224,
+      "learning_rate": 7.762982543819358e-06,
+      "loss": 0.4724,
+      "step": 2216
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.524264263964921,
+      "learning_rate": 7.760946483388843e-06,
+      "loss": 0.5182,
+      "step": 2217
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.2427223109487247,
+      "learning_rate": 7.75890976408685e-06,
+      "loss": 0.4283,
+      "step": 2218
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.9112451067848384,
+      "learning_rate": 7.756872386399424e-06,
+      "loss": 0.4845,
+      "step": 2219
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.023423941003666,
+      "learning_rate": 7.754834350812765e-06,
+      "loss": 0.4259,
+      "step": 2220
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.550337240890039,
+      "learning_rate": 7.752795657813227e-06,
+      "loss": 0.4707,
+      "step": 2221
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.832586977104632,
+      "learning_rate": 7.750756307887323e-06,
+      "loss": 0.4975,
+      "step": 2222
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.6828741998903833,
+      "learning_rate": 7.748716301521726e-06,
+      "loss": 0.4354,
+      "step": 2223
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.718182301539495,
+      "learning_rate": 7.746675639203258e-06,
+      "loss": 0.4857,
+      "step": 2224
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.541075844528042,
+      "learning_rate": 7.744634321418907e-06,
+      "loss": 0.4612,
+      "step": 2225
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.164285514763037,
+      "learning_rate": 7.74259234865581e-06,
+      "loss": 0.5385,
+      "step": 2226
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.385217029859917,
+      "learning_rate": 7.74054972140126e-06,
+      "loss": 0.4892,
+      "step": 2227
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.4354853703148183,
+      "learning_rate": 7.738506440142713e-06,
+      "loss": 0.474,
+      "step": 2228
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.3111682565596974,
+      "learning_rate": 7.736462505367775e-06,
+      "loss": 0.5294,
+      "step": 2229
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4776405368640575,
+      "learning_rate": 7.734417917564212e-06,
+      "loss": 0.5849,
+      "step": 2230
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.123442443517847,
+      "learning_rate": 7.732372677219942e-06,
+      "loss": 0.5043,
+      "step": 2231
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.7629193807938774,
+      "learning_rate": 7.730326784823042e-06,
+      "loss": 0.5866,
+      "step": 2232
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.8303656972756643,
+      "learning_rate": 7.72828024086174e-06,
+      "loss": 0.4491,
+      "step": 2233
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.5550750408239176,
+      "learning_rate": 7.726233045824427e-06,
+      "loss": 0.5077,
+      "step": 2234
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.013021933844967,
+      "learning_rate": 7.724185200199642e-06,
+      "loss": 0.4422,
+      "step": 2235
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.074469155291931,
+      "learning_rate": 7.722136704476086e-06,
+      "loss": 0.4465,
+      "step": 2236
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.5485749830158526,
+      "learning_rate": 7.720087559142607e-06,
+      "loss": 0.3921,
+      "step": 2237
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.818457222940472,
+      "learning_rate": 7.718037764688215e-06,
+      "loss": 0.4346,
+      "step": 2238
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.527702045899896,
+      "learning_rate": 7.71598732160207e-06,
+      "loss": 0.4758,
+      "step": 2239
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.171578178258161,
+      "learning_rate": 7.713936230373491e-06,
+      "loss": 0.4446,
+      "step": 2240
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.858507846635064,
+      "learning_rate": 7.711884491491953e-06,
+      "loss": 0.4672,
+      "step": 2241
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.4540914075896705,
+      "learning_rate": 7.709832105447075e-06,
+      "loss": 0.4755,
+      "step": 2242
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.677163104081454,
+      "learning_rate": 7.707779072728643e-06,
+      "loss": 0.4744,
+      "step": 2243
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.754609000067211,
+      "learning_rate": 7.705725393826591e-06,
+      "loss": 0.4983,
+      "step": 2244
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 11.448159110835741,
+      "learning_rate": 7.703671069231007e-06,
+      "loss": 0.4673,
+      "step": 2245
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.853395701847896,
+      "learning_rate": 7.701616099432137e-06,
+      "loss": 0.464,
+      "step": 2246
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.6108452016247705,
+      "learning_rate": 7.699560484920378e-06,
+      "loss": 0.4559,
+      "step": 2247
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.792459428507114,
+      "learning_rate": 7.697504226186279e-06,
+      "loss": 0.4398,
+      "step": 2248
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.248093122888479,
+      "learning_rate": 7.695447323720547e-06,
+      "loss": 0.5419,
+      "step": 2249
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.736821857064174,
+      "learning_rate": 7.693389778014038e-06,
+      "loss": 0.4855,
+      "step": 2250
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.515647200063691,
+      "learning_rate": 7.691331589557767e-06,
+      "loss": 0.4689,
+      "step": 2251
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.7439235957701436,
+      "learning_rate": 7.6892727588429e-06,
+      "loss": 0.4595,
+      "step": 2252
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.486528988819889,
+      "learning_rate": 7.687213286360754e-06,
+      "loss": 0.4133,
+      "step": 2253
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.944870414467476,
+      "learning_rate": 7.685153172602799e-06,
+      "loss": 0.4275,
+      "step": 2254
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.8091008053728275,
+      "learning_rate": 7.683092418060664e-06,
+      "loss": 0.4873,
+      "step": 2255
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.3534539575238926,
+      "learning_rate": 7.681031023226126e-06,
+      "loss": 0.4834,
+      "step": 2256
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.660781671805133,
+      "learning_rate": 7.678968988591114e-06,
+      "loss": 0.4692,
+      "step": 2257
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.204042381836633,
+      "learning_rate": 7.676906314647713e-06,
+      "loss": 0.4332,
+      "step": 2258
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.092228284133106,
+      "learning_rate": 7.674843001888156e-06,
+      "loss": 0.5518,
+      "step": 2259
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.432305954559312,
+      "learning_rate": 7.672779050804834e-06,
+      "loss": 0.4905,
+      "step": 2260
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.303047100994622,
+      "learning_rate": 7.670714461890287e-06,
+      "loss": 0.4818,
+      "step": 2261
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.8671256128651255,
+      "learning_rate": 7.668649235637209e-06,
+      "loss": 0.423,
+      "step": 2262
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.106252559946471,
+      "learning_rate": 7.666583372538443e-06,
+      "loss": 0.4982,
+      "step": 2263
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.3922877286906736,
+      "learning_rate": 7.664516873086987e-06,
+      "loss": 0.4858,
+      "step": 2264
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.430306247078682,
+      "learning_rate": 7.662449737775992e-06,
+      "loss": 0.4979,
+      "step": 2265
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.739244484901206,
+      "learning_rate": 7.660381967098753e-06,
+      "loss": 0.459,
+      "step": 2266
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.287673453029182,
+      "learning_rate": 7.658313561548727e-06,
+      "loss": 0.4728,
+      "step": 2267
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.200883218055887,
+      "learning_rate": 7.656244521619516e-06,
+      "loss": 0.4403,
+      "step": 2268
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.241313989189559,
+      "learning_rate": 7.654174847804876e-06,
+      "loss": 0.5128,
+      "step": 2269
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.454916603722773,
+      "learning_rate": 7.652104540598712e-06,
+      "loss": 0.4542,
+      "step": 2270
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.846682446655618,
+      "learning_rate": 7.650033600495085e-06,
+      "loss": 0.4976,
+      "step": 2271
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.570160536200375,
+      "learning_rate": 7.647962027988198e-06,
+      "loss": 0.5302,
+      "step": 2272
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.69400521704816,
+      "learning_rate": 7.645889823572414e-06,
+      "loss": 0.5013,
+      "step": 2273
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.132253518412157,
+      "learning_rate": 7.64381698774224e-06,
+      "loss": 0.4776,
+      "step": 2274
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.248818058467961,
+      "learning_rate": 7.641743520992343e-06,
+      "loss": 0.5039,
+      "step": 2275
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9206915459343854,
+      "learning_rate": 7.63966942381753e-06,
+      "loss": 0.4484,
+      "step": 2276
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.018739013453988,
+      "learning_rate": 7.637594696712764e-06,
+      "loss": 0.486,
+      "step": 2277
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.060724434668855,
+      "learning_rate": 7.635519340173158e-06,
+      "loss": 0.4728,
+      "step": 2278
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0903544711181343,
+      "learning_rate": 7.633443354693972e-06,
+      "loss": 0.5149,
+      "step": 2279
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.938529300876123,
+      "learning_rate": 7.631366740770622e-06,
+      "loss": 0.5119,
+      "step": 2280
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.3174035234796326,
+      "learning_rate": 7.629289498898668e-06,
+      "loss": 0.454,
+      "step": 2281
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.745714741978283,
+      "learning_rate": 7.627211629573823e-06,
+      "loss": 0.4894,
+      "step": 2282
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.440185290054798,
+      "learning_rate": 7.62513313329195e-06,
+      "loss": 0.4784,
+      "step": 2283
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.5165780264646433,
+      "learning_rate": 7.623054010549059e-06,
+      "loss": 0.4701,
+      "step": 2284
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4499738306862464,
+      "learning_rate": 7.620974261841314e-06,
+      "loss": 0.4827,
+      "step": 2285
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.6224948144255533,
+      "learning_rate": 7.618893887665023e-06,
+      "loss": 0.4011,
+      "step": 2286
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.793566931630369,
+      "learning_rate": 7.616812888516647e-06,
+      "loss": 0.4877,
+      "step": 2287
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.391115502612298,
+      "learning_rate": 7.6147312648927965e-06,
+      "loss": 0.4667,
+      "step": 2288
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 13.835996460213295,
+      "learning_rate": 7.612649017290226e-06,
+      "loss": 0.5398,
+      "step": 2289
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2422443584194647,
+      "learning_rate": 7.610566146205845e-06,
+      "loss": 0.4862,
+      "step": 2290
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.6073575472593338,
+      "learning_rate": 7.608482652136708e-06,
+      "loss": 0.445,
+      "step": 2291
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.410911447507514,
+      "learning_rate": 7.606398535580022e-06,
+      "loss": 0.4626,
+      "step": 2292
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.716683746671777,
+      "learning_rate": 7.604313797033139e-06,
+      "loss": 0.4388,
+      "step": 2293
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.5714470414665596,
+      "learning_rate": 7.602228436993556e-06,
+      "loss": 0.5326,
+      "step": 2294
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.2992389994539955,
+      "learning_rate": 7.60014245595893e-06,
+      "loss": 0.5278,
+      "step": 2295
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.8073232961435317,
+      "learning_rate": 7.598055854427052e-06,
+      "loss": 0.3928,
+      "step": 2296
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2828805534073893,
+      "learning_rate": 7.595968632895872e-06,
+      "loss": 0.5716,
+      "step": 2297
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.351403044743196,
+      "learning_rate": 7.593880791863484e-06,
+      "loss": 0.5161,
+      "step": 2298
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6287241846561997,
+      "learning_rate": 7.59179233182813e-06,
+      "loss": 0.4746,
+      "step": 2299
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8255561895653836,
+      "learning_rate": 7.589703253288196e-06,
+      "loss": 0.4746,
+      "step": 2300
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.6237136435664787,
+      "learning_rate": 7.587613556742224e-06,
+      "loss": 0.5696,
+      "step": 2301
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.6381770416095205,
+      "learning_rate": 7.585523242688894e-06,
+      "loss": 0.4796,
+      "step": 2302
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4757415396026063,
+      "learning_rate": 7.583432311627042e-06,
+      "loss": 0.4885,
+      "step": 2303
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.8351873207164493,
+      "learning_rate": 7.581340764055642e-06,
+      "loss": 0.499,
+      "step": 2304
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.5355110284374742,
+      "learning_rate": 7.579248600473827e-06,
+      "loss": 0.4702,
+      "step": 2305
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.452357047864653,
+      "learning_rate": 7.577155821380866e-06,
+      "loss": 0.5849,
+      "step": 2306
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.314047411421253,
+      "learning_rate": 7.57506242727618e-06,
+      "loss": 0.3686,
+      "step": 2307
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.573902745993604,
+      "learning_rate": 7.5729684186593344e-06,
+      "loss": 0.4979,
+      "step": 2308
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8156751812625647,
+      "learning_rate": 7.570873796030046e-06,
+      "loss": 0.4316,
+      "step": 2309
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.600049244185723,
+      "learning_rate": 7.5687785598881734e-06,
+      "loss": 0.5268,
+      "step": 2310
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.5808200959525296,
+      "learning_rate": 7.566682710733724e-06,
+      "loss": 0.4497,
+      "step": 2311
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9107101813019423,
+      "learning_rate": 7.564586249066849e-06,
+      "loss": 0.5138,
+      "step": 2312
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.1324272698959206,
+      "learning_rate": 7.562489175387847e-06,
+      "loss": 0.5506,
+      "step": 2313
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.5583265096612626,
+      "learning_rate": 7.5603914901971655e-06,
+      "loss": 0.4833,
+      "step": 2314
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.217738882956242,
+      "learning_rate": 7.5582931939953945e-06,
+      "loss": 0.4615,
+      "step": 2315
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.723701861670058,
+      "learning_rate": 7.55619428728327e-06,
+      "loss": 0.4541,
+      "step": 2316
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7545212931361975,
+      "learning_rate": 7.554094770561675e-06,
+      "loss": 0.4732,
+      "step": 2317
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5956854653926493,
+      "learning_rate": 7.5519946443316375e-06,
+      "loss": 0.5931,
+      "step": 2318
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.698105812583981,
+      "learning_rate": 7.5498939090943325e-06,
+      "loss": 0.4715,
+      "step": 2319
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.582070491072131,
+      "learning_rate": 7.547792565351076e-06,
+      "loss": 0.4373,
+      "step": 2320
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.326872809435797,
+      "learning_rate": 7.545690613603333e-06,
+      "loss": 0.4178,
+      "step": 2321
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9036913839925713,
+      "learning_rate": 7.543588054352715e-06,
+      "loss": 0.4681,
+      "step": 2322
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.012533937870758,
+      "learning_rate": 7.541484888100974e-06,
+      "loss": 0.5056,
+      "step": 2323
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.433188318684449,
+      "learning_rate": 7.53938111535001e-06,
+      "loss": 0.4928,
+      "step": 2324
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0515687308988624,
+      "learning_rate": 7.537276736601864e-06,
+      "loss": 0.4644,
+      "step": 2325
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3941245715115476,
+      "learning_rate": 7.535171752358727e-06,
+      "loss": 0.6249,
+      "step": 2326
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.400350993669927,
+      "learning_rate": 7.533066163122932e-06,
+      "loss": 0.4655,
+      "step": 2327
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.9743790153201415,
+      "learning_rate": 7.5309599693969556e-06,
+      "loss": 0.4487,
+      "step": 2328
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.7286994813150485,
+      "learning_rate": 7.528853171683419e-06,
+      "loss": 0.4254,
+      "step": 2329
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.625766798890735,
+      "learning_rate": 7.526745770485087e-06,
+      "loss": 0.4968,
+      "step": 2330
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8421209314991422,
+      "learning_rate": 7.524637766304872e-06,
+      "loss": 0.467,
+      "step": 2331
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.300125231132899,
+      "learning_rate": 7.522529159645824e-06,
+      "loss": 0.4586,
+      "step": 2332
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.874058698185691,
+      "learning_rate": 7.520419951011142e-06,
+      "loss": 0.4456,
+      "step": 2333
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.771212291214637,
+      "learning_rate": 7.518310140904168e-06,
+      "loss": 0.4834,
+      "step": 2334
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.948613403980211,
+      "learning_rate": 7.516199729828385e-06,
+      "loss": 0.4857,
+      "step": 2335
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.149792528040112,
+      "learning_rate": 7.51408871828742e-06,
+      "loss": 0.5548,
+      "step": 2336
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.524187215245263,
+      "learning_rate": 7.511977106785047e-06,
+      "loss": 0.4868,
+      "step": 2337
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.165713280904329,
+      "learning_rate": 7.509864895825177e-06,
+      "loss": 0.5515,
+      "step": 2338
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.5807550343802252,
+      "learning_rate": 7.5077520859118725e-06,
+      "loss": 0.5035,
+      "step": 2339
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.3115251734617095,
+      "learning_rate": 7.505638677549327e-06,
+      "loss": 0.5427,
+      "step": 2340
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.3351459670309946,
+      "learning_rate": 7.503524671241891e-06,
+      "loss": 0.512,
+      "step": 2341
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.637112794821689,
+      "learning_rate": 7.501410067494044e-06,
+      "loss": 0.5465,
+      "step": 2342
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.841462445033107,
+      "learning_rate": 7.499294866810416e-06,
+      "loss": 0.5146,
+      "step": 2343
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.1341016899078467,
+      "learning_rate": 7.497179069695781e-06,
+      "loss": 0.4752,
+      "step": 2344
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.411183226226576,
+      "learning_rate": 7.49506267665505e-06,
+      "loss": 0.4508,
+      "step": 2345
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.631883951717825,
+      "learning_rate": 7.492945688193278e-06,
+      "loss": 0.5189,
+      "step": 2346
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.519052420568145,
+      "learning_rate": 7.490828104815663e-06,
+      "loss": 0.4944,
+      "step": 2347
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.5069326027306844,
+      "learning_rate": 7.488709927027544e-06,
+      "loss": 0.4421,
+      "step": 2348
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.147271464713538,
+      "learning_rate": 7.486591155334404e-06,
+      "loss": 0.5075,
+      "step": 2349
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.916707981575327,
+      "learning_rate": 7.484471790241865e-06,
+      "loss": 0.5159,
+      "step": 2350
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3412114822708805,
+      "learning_rate": 7.482351832255691e-06,
+      "loss": 0.5657,
+      "step": 2351
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.940698847921472,
+      "learning_rate": 7.480231281881789e-06,
+      "loss": 0.4997,
+      "step": 2352
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.1639305801434627,
+      "learning_rate": 7.4781101396262064e-06,
+      "loss": 0.5072,
+      "step": 2353
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6802727470418577,
+      "learning_rate": 7.475988405995133e-06,
+      "loss": 0.524,
+      "step": 2354
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8229294585406794,
+      "learning_rate": 7.473866081494896e-06,
+      "loss": 0.4526,
+      "step": 2355
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.328084063388005,
+      "learning_rate": 7.471743166631969e-06,
+      "loss": 0.4862,
+      "step": 2356
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.15578375667564,
+      "learning_rate": 7.469619661912962e-06,
+      "loss": 0.4907,
+      "step": 2357
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.786440732658542,
+      "learning_rate": 7.467495567844628e-06,
+      "loss": 0.44,
+      "step": 2358
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9863262528581376,
+      "learning_rate": 7.465370884933863e-06,
+      "loss": 0.5004,
+      "step": 2359
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.616551227698675,
+      "learning_rate": 7.463245613687695e-06,
+      "loss": 0.4941,
+      "step": 2360
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6790899021916617,
+      "learning_rate": 7.461119754613302e-06,
+      "loss": 0.4546,
+      "step": 2361
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.6973709898783294,
+      "learning_rate": 7.458993308217999e-06,
+      "loss": 0.4428,
+      "step": 2362
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3537657484537213,
+      "learning_rate": 7.456866275009238e-06,
+      "loss": 0.6041,
+      "step": 2363
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.7908744390982507,
+      "learning_rate": 7.4547386554946165e-06,
+      "loss": 0.5042,
+      "step": 2364
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.617095600969997,
+      "learning_rate": 7.4526104501818665e-06,
+      "loss": 0.4904,
+      "step": 2365
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.0657080777062142,
+      "learning_rate": 7.450481659578862e-06,
+      "loss": 0.4887,
+      "step": 2366
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.4908255036798472,
+      "learning_rate": 7.44835228419362e-06,
+      "loss": 0.4522,
+      "step": 2367
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.520778842206725,
+      "learning_rate": 7.44622232453429e-06,
+      "loss": 0.4988,
+      "step": 2368
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.701702868806101,
+      "learning_rate": 7.444091781109169e-06,
+      "loss": 0.4542,
+      "step": 2369
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.796542487044817,
+      "learning_rate": 7.441960654426688e-06,
+      "loss": 0.459,
+      "step": 2370
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2967643453209285,
+      "learning_rate": 7.4398289449954165e-06,
+      "loss": 0.5814,
+      "step": 2371
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.303351148039858,
+      "learning_rate": 7.437696653324067e-06,
+      "loss": 0.417,
+      "step": 2372
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.679285475377113,
+      "learning_rate": 7.4355637799214884e-06,
+      "loss": 0.446,
+      "step": 2373
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.3864743657630445,
+      "learning_rate": 7.433430325296669e-06,
+      "loss": 0.472,
+      "step": 2374
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.5446376520745755,
+      "learning_rate": 7.431296289958736e-06,
+      "loss": 0.4379,
+      "step": 2375
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.117399289277465,
+      "learning_rate": 7.429161674416953e-06,
+      "loss": 0.5216,
+      "step": 2376
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.944676639685801,
+      "learning_rate": 7.427026479180727e-06,
+      "loss": 0.4441,
+      "step": 2377
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.794498971384045,
+      "learning_rate": 7.4248907047596005e-06,
+      "loss": 0.5262,
+      "step": 2378
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.6051392071934245,
+      "learning_rate": 7.422754351663252e-06,
+      "loss": 0.5123,
+      "step": 2379
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.391459230053948,
+      "learning_rate": 7.4206174204015e-06,
+      "loss": 0.4023,
+      "step": 2380
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2943844882663575,
+      "learning_rate": 7.418479911484303e-06,
+      "loss": 0.5447,
+      "step": 2381
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5665395781799307,
+      "learning_rate": 7.416341825421755e-06,
+      "loss": 0.4606,
+      "step": 2382
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.271612990798521,
+      "learning_rate": 7.414203162724087e-06,
+      "loss": 0.4694,
+      "step": 2383
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.0382331565958105,
+      "learning_rate": 7.41206392390167e-06,
+      "loss": 0.484,
+      "step": 2384
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.5697827937076445,
+      "learning_rate": 7.409924109465011e-06,
+      "loss": 0.4507,
+      "step": 2385
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4761784504394564,
+      "learning_rate": 7.4077837199247545e-06,
+      "loss": 0.466,
+      "step": 2386
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.205812400079637,
+      "learning_rate": 7.405642755791684e-06,
+      "loss": 0.4937,
+      "step": 2387
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.056025525539817,
+      "learning_rate": 7.403501217576716e-06,
+      "loss": 0.4878,
+      "step": 2388
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3889560841604407,
+      "learning_rate": 7.401359105790909e-06,
+      "loss": 0.4498,
+      "step": 2389
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.2738953265966377,
+      "learning_rate": 7.399216420945453e-06,
+      "loss": 0.4348,
+      "step": 2390
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.613248927475757,
+      "learning_rate": 7.39707316355168e-06,
+      "loss": 0.425,
+      "step": 2391
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.996498917868073,
+      "learning_rate": 7.394929334121057e-06,
+      "loss": 0.4822,
+      "step": 2392
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.128055419570635,
+      "learning_rate": 7.392784933165186e-06,
+      "loss": 0.424,
+      "step": 2393
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.165272945381678,
+      "learning_rate": 7.390639961195805e-06,
+      "loss": 0.5441,
+      "step": 2394
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.127464113392247,
+      "learning_rate": 7.3884944187247895e-06,
+      "loss": 0.4873,
+      "step": 2395
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.947277047673751,
+      "learning_rate": 7.386348306264153e-06,
+      "loss": 0.5021,
+      "step": 2396
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.38805023443246,
+      "learning_rate": 7.384201624326042e-06,
+      "loss": 0.4678,
+      "step": 2397
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4039732258795277,
+      "learning_rate": 7.38205437342274e-06,
+      "loss": 0.4571,
+      "step": 2398
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.550805545956856,
+      "learning_rate": 7.379906554066667e-06,
+      "loss": 0.4614,
+      "step": 2399
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.0738208257797357,
+      "learning_rate": 7.3777581667703764e-06,
+      "loss": 0.4821,
+      "step": 2400
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6846303795490853,
+      "learning_rate": 7.3756092120465615e-06,
+      "loss": 0.45,
+      "step": 2401
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4463112035762338,
+      "learning_rate": 7.373459690408046e-06,
+      "loss": 0.4696,
+      "step": 2402
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.898412590499899,
+      "learning_rate": 7.371309602367792e-06,
+      "loss": 0.4397,
+      "step": 2403
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.3678424224651757,
+      "learning_rate": 7.369158948438895e-06,
+      "loss": 0.4646,
+      "step": 2404
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5468587515720715,
+      "learning_rate": 7.367007729134589e-06,
+      "loss": 0.5007,
+      "step": 2405
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.84200309342235,
+      "learning_rate": 7.3648559449682364e-06,
+      "loss": 0.4899,
+      "step": 2406
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.1385236052087917,
+      "learning_rate": 7.36270359645334e-06,
+      "loss": 0.5357,
+      "step": 2407
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.0117544221422285,
+      "learning_rate": 7.360550684103538e-06,
+      "loss": 0.4323,
+      "step": 2408
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.278077874312744,
+      "learning_rate": 7.358397208432597e-06,
+      "loss": 0.4423,
+      "step": 2409
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.739929815210438,
+      "learning_rate": 7.356243169954426e-06,
+      "loss": 0.4289,
+      "step": 2410
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6890506583928087,
+      "learning_rate": 7.35408856918306e-06,
+      "loss": 0.4918,
+      "step": 2411
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.5527607087240387,
+      "learning_rate": 7.3519334066326744e-06,
+      "loss": 0.4567,
+      "step": 2412
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7259827107013255,
+      "learning_rate": 7.349777682817576e-06,
+      "loss": 0.4772,
+      "step": 2413
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.01317696317138,
+      "learning_rate": 7.347621398252207e-06,
+      "loss": 0.5182,
+      "step": 2414
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.901459310104945,
+      "learning_rate": 7.345464553451141e-06,
+      "loss": 0.3925,
+      "step": 2415
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.798720250285504,
+      "learning_rate": 7.343307148929089e-06,
+      "loss": 0.503,
+      "step": 2416
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.764285321686053,
+      "learning_rate": 7.341149185200891e-06,
+      "loss": 0.4157,
+      "step": 2417
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.822440712892292,
+      "learning_rate": 7.338990662781524e-06,
+      "loss": 0.4875,
+      "step": 2418
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.5520900723640674,
+      "learning_rate": 7.336831582186097e-06,
+      "loss": 0.4654,
+      "step": 2419
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.1943192699259804,
+      "learning_rate": 7.334671943929853e-06,
+      "loss": 0.4738,
+      "step": 2420
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9235056902629255,
+      "learning_rate": 7.332511748528167e-06,
+      "loss": 0.5514,
+      "step": 2421
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5964164022569878,
+      "learning_rate": 7.330350996496547e-06,
+      "loss": 0.5358,
+      "step": 2422
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2529298418572723,
+      "learning_rate": 7.3281896883506355e-06,
+      "loss": 0.5752,
+      "step": 2423
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.547050492474515,
+      "learning_rate": 7.3260278246062045e-06,
+      "loss": 0.5059,
+      "step": 2424
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.565889330482098,
+      "learning_rate": 7.323865405779163e-06,
+      "loss": 0.4515,
+      "step": 2425
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3993677637499897,
+      "learning_rate": 7.321702432385546e-06,
+      "loss": 0.4807,
+      "step": 2426
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.890712382095551,
+      "learning_rate": 7.31953890494153e-06,
+      "loss": 0.5415,
+      "step": 2427
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.5513493780949887,
+      "learning_rate": 7.317374823963415e-06,
+      "loss": 0.5037,
+      "step": 2428
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.369435061019334,
+      "learning_rate": 7.315210189967637e-06,
+      "loss": 0.4041,
+      "step": 2429
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6106424036222062,
+      "learning_rate": 7.313045003470766e-06,
+      "loss": 0.4101,
+      "step": 2430
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.58224470044946,
+      "learning_rate": 7.310879264989498e-06,
+      "loss": 0.4305,
+      "step": 2431
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.9394626167088385,
+      "learning_rate": 7.308712975040667e-06,
+      "loss": 0.4888,
+      "step": 2432
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.503333802951347,
+      "learning_rate": 7.306546134141234e-06,
+      "loss": 0.48,
+      "step": 2433
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3862982817127865,
+      "learning_rate": 7.304378742808296e-06,
+      "loss": 0.5569,
+      "step": 2434
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.174515425988138,
+      "learning_rate": 7.3022108015590755e-06,
+      "loss": 0.4866,
+      "step": 2435
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.341559137287811,
+      "learning_rate": 7.30004231091093e-06,
+      "loss": 0.3901,
+      "step": 2436
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.4811153408009345,
+      "learning_rate": 7.29787327138135e-06,
+      "loss": 0.5211,
+      "step": 2437
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.3739158207486137,
+      "learning_rate": 7.2957036834879505e-06,
+      "loss": 0.4942,
+      "step": 2438
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.7242955373101747,
+      "learning_rate": 7.2935335477484844e-06,
+      "loss": 0.4635,
+      "step": 2439
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.770147307770478,
+      "learning_rate": 7.291362864680831e-06,
+      "loss": 0.5273,
+      "step": 2440
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.908437433983117,
+      "learning_rate": 7.289191634803002e-06,
+      "loss": 0.4769,
+      "step": 2441
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.200020574832539,
+      "learning_rate": 7.28701985863314e-06,
+      "loss": 0.5732,
+      "step": 2442
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.6926356792365813,
+      "learning_rate": 7.284847536689514e-06,
+      "loss": 0.4559,
+      "step": 2443
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.7538166653297127,
+      "learning_rate": 7.282674669490528e-06,
+      "loss": 0.5405,
+      "step": 2444
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.4560967316352516,
+      "learning_rate": 7.280501257554717e-06,
+      "loss": 0.4932,
+      "step": 2445
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.7161029463295336,
+      "learning_rate": 7.278327301400738e-06,
+      "loss": 0.4319,
+      "step": 2446
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.345377974682153,
+      "learning_rate": 7.276152801547389e-06,
+      "loss": 0.5685,
+      "step": 2447
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.2537992393948,
+      "learning_rate": 7.273977758513588e-06,
+      "loss": 0.4742,
+      "step": 2448
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.267840225793219,
+      "learning_rate": 7.271802172818387e-06,
+      "loss": 0.4522,
+      "step": 2449
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.9102798015901823,
+      "learning_rate": 7.269626044980968e-06,
+      "loss": 0.5225,
+      "step": 2450
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.132715424658986,
+      "learning_rate": 7.267449375520644e-06,
+      "loss": 0.39,
+      "step": 2451
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 13.300986338579055,
+      "learning_rate": 7.2652721649568516e-06,
+      "loss": 0.523,
+      "step": 2452
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.450007266556247,
+      "learning_rate": 7.2630944138091595e-06,
+      "loss": 0.5306,
+      "step": 2453
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.908865235441064,
+      "learning_rate": 7.260916122597267e-06,
+      "loss": 0.5516,
+      "step": 2454
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.942873889334752,
+      "learning_rate": 7.258737291840999e-06,
+      "loss": 0.4663,
+      "step": 2455
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.583131320200942,
+      "learning_rate": 7.256557922060315e-06,
+      "loss": 0.5487,
+      "step": 2456
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.199426091843973,
+      "learning_rate": 7.254378013775296e-06,
+      "loss": 0.5279,
+      "step": 2457
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.850121215742621,
+      "learning_rate": 7.2521975675061555e-06,
+      "loss": 0.4969,
+      "step": 2458
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.240548874900549,
+      "learning_rate": 7.250016583773232e-06,
+      "loss": 0.4426,
+      "step": 2459
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 16.928179558543754,
+      "learning_rate": 7.247835063097e-06,
+      "loss": 0.5458,
+      "step": 2460
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 22.699289561346102,
+      "learning_rate": 7.245653005998053e-06,
+      "loss": 0.5332,
+      "step": 2461
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.047685134894629,
+      "learning_rate": 7.243470412997118e-06,
+      "loss": 0.4457,
+      "step": 2462
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 17.866870740156013,
+      "learning_rate": 7.24128728461505e-06,
+      "loss": 0.475,
+      "step": 2463
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.2274055359494023,
+      "learning_rate": 7.239103621372825e-06,
+      "loss": 0.4988,
+      "step": 2464
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3161080676896624,
+      "learning_rate": 7.2369194237915565e-06,
+      "loss": 0.5564,
+      "step": 2465
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.130710110155509,
+      "learning_rate": 7.234734692392479e-06,
+      "loss": 0.555,
+      "step": 2466
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7000968887129133,
+      "learning_rate": 7.232549427696958e-06,
+      "loss": 0.4387,
+      "step": 2467
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 13.615869986831212,
+      "learning_rate": 7.23036363022648e-06,
+      "loss": 0.4546,
+      "step": 2468
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.047224633925523,
+      "learning_rate": 7.228177300502668e-06,
+      "loss": 0.4678,
+      "step": 2469
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.5798114342673992,
+      "learning_rate": 7.225990439047265e-06,
+      "loss": 0.4365,
+      "step": 2470
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.920529402273849,
+      "learning_rate": 7.223803046382142e-06,
+      "loss": 0.4861,
+      "step": 2471
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.208049628186236,
+      "learning_rate": 7.2216151230293e-06,
+      "loss": 0.5203,
+      "step": 2472
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.46167637221668,
+      "learning_rate": 7.219426669510862e-06,
+      "loss": 0.5041,
+      "step": 2473
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.303352840461777,
+      "learning_rate": 7.2172376863490824e-06,
+      "loss": 0.4607,
+      "step": 2474
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.4698930047824708,
+      "learning_rate": 7.215048174066337e-06,
+      "loss": 0.4473,
+      "step": 2475
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.249019520549649,
+      "learning_rate": 7.212858133185132e-06,
+      "loss": 0.4515,
+      "step": 2476
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1874611045886068,
+      "learning_rate": 7.210667564228099e-06,
+      "loss": 0.5608,
+      "step": 2477
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.500591819340289,
+      "learning_rate": 7.208476467717992e-06,
+      "loss": 0.4921,
+      "step": 2478
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.2456384618481753,
+      "learning_rate": 7.206284844177695e-06,
+      "loss": 0.4683,
+      "step": 2479
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.266524121256204,
+      "learning_rate": 7.204092694130218e-06,
+      "loss": 0.4988,
+      "step": 2480
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8134156725042487,
+      "learning_rate": 7.201900018098692e-06,
+      "loss": 0.4699,
+      "step": 2481
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.045493583913286,
+      "learning_rate": 7.19970681660638e-06,
+      "loss": 0.4907,
+      "step": 2482
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4213431954863918,
+      "learning_rate": 7.197513090176663e-06,
+      "loss": 0.5486,
+      "step": 2483
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.543827736508919,
+      "learning_rate": 7.195318839333054e-06,
+      "loss": 0.4544,
+      "step": 2484
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.4460178828026455,
+      "learning_rate": 7.193124064599188e-06,
+      "loss": 0.5397,
+      "step": 2485
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.13067526898253,
+      "learning_rate": 7.190928766498828e-06,
+      "loss": 0.543,
+      "step": 2486
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.655323900086479,
+      "learning_rate": 7.188732945555854e-06,
+      "loss": 0.4816,
+      "step": 2487
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.159541709150244,
+      "learning_rate": 7.186536602294278e-06,
+      "loss": 0.497,
+      "step": 2488
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.733146261504955,
+      "learning_rate": 7.184339737238235e-06,
+      "loss": 0.478,
+      "step": 2489
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.944714026763166,
+      "learning_rate": 7.182142350911986e-06,
+      "loss": 0.4398,
+      "step": 2490
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.5159824289025154,
+      "learning_rate": 7.179944443839913e-06,
+      "loss": 0.4895,
+      "step": 2491
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2306893866020683,
+      "learning_rate": 7.177746016546525e-06,
+      "loss": 0.5594,
+      "step": 2492
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9144531074063504,
+      "learning_rate": 7.175547069556451e-06,
+      "loss": 0.5251,
+      "step": 2493
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.4025611349858624,
+      "learning_rate": 7.17334760339445e-06,
+      "loss": 0.3815,
+      "step": 2494
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.163767397372717,
+      "learning_rate": 7.171147618585401e-06,
+      "loss": 0.4447,
+      "step": 2495
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.862737775664096,
+      "learning_rate": 7.16894711565431e-06,
+      "loss": 0.5153,
+      "step": 2496
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.520721774046978,
+      "learning_rate": 7.1667460951263e-06,
+      "loss": 0.4075,
+      "step": 2497
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.9311274655638675,
+      "learning_rate": 7.164544557526626e-06,
+      "loss": 0.5436,
+      "step": 2498
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.1469865371162267,
+      "learning_rate": 7.162342503380659e-06,
+      "loss": 0.4243,
+      "step": 2499
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.427717836716093,
+      "learning_rate": 7.160139933213899e-06,
+      "loss": 0.3954,
+      "step": 2500
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.069241649688368,
+      "learning_rate": 7.1579368475519665e-06,
+      "loss": 0.4758,
+      "step": 2501
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.682528917077807,
+      "learning_rate": 7.155733246920603e-06,
+      "loss": 0.5123,
+      "step": 2502
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.282706188952751,
+      "learning_rate": 7.153529131845678e-06,
+      "loss": 0.4466,
+      "step": 2503
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.174855962183863,
+      "learning_rate": 7.151324502853179e-06,
+      "loss": 0.494,
+      "step": 2504
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.778926872876768,
+      "learning_rate": 7.149119360469218e-06,
+      "loss": 0.3703,
+      "step": 2505
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.167661947319722,
+      "learning_rate": 7.146913705220031e-06,
+      "loss": 0.4402,
+      "step": 2506
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.67117642766957,
+      "learning_rate": 7.144707537631974e-06,
+      "loss": 0.4509,
+      "step": 2507
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.167825265498907,
+      "learning_rate": 7.142500858231526e-06,
+      "loss": 0.4666,
+      "step": 2508
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.545973808294085,
+      "learning_rate": 7.140293667545289e-06,
+      "loss": 0.5098,
+      "step": 2509
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.069696407357052,
+      "learning_rate": 7.138085966099986e-06,
+      "loss": 0.4161,
+      "step": 2510
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.335668373189224,
+      "learning_rate": 7.135877754422462e-06,
+      "loss": 0.523,
+      "step": 2511
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.009478497883999,
+      "learning_rate": 7.133669033039685e-06,
+      "loss": 0.4811,
+      "step": 2512
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7001275513313785,
+      "learning_rate": 7.131459802478742e-06,
+      "loss": 0.3922,
+      "step": 2513
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7703645069479585,
+      "learning_rate": 7.129250063266844e-06,
+      "loss": 0.4212,
+      "step": 2514
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.550220398959434,
+      "learning_rate": 7.127039815931323e-06,
+      "loss": 0.499,
+      "step": 2515
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.65764183758504,
+      "learning_rate": 7.124829060999632e-06,
+      "loss": 0.4994,
+      "step": 2516
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9428146673129807,
+      "learning_rate": 7.122617798999343e-06,
+      "loss": 0.4783,
+      "step": 2517
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.350277750916895,
+      "learning_rate": 7.120406030458151e-06,
+      "loss": 0.4727,
+      "step": 2518
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.074620205019584,
+      "learning_rate": 7.118193755903875e-06,
+      "loss": 0.4587,
+      "step": 2519
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.147323801255245,
+      "learning_rate": 7.115980975864449e-06,
+      "loss": 0.5663,
+      "step": 2520
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.986781005598588,
+      "learning_rate": 7.113767690867932e-06,
+      "loss": 0.4779,
+      "step": 2521
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.093697721263707,
+      "learning_rate": 7.111553901442499e-06,
+      "loss": 0.5608,
+      "step": 2522
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.360996968374214,
+      "learning_rate": 7.109339608116449e-06,
+      "loss": 0.4529,
+      "step": 2523
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7226960540930536,
+      "learning_rate": 7.107124811418201e-06,
+      "loss": 0.4583,
+      "step": 2524
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.157622868312493,
+      "learning_rate": 7.1049095118762935e-06,
+      "loss": 0.537,
+      "step": 2525
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.4151379421395025,
+      "learning_rate": 7.102693710019386e-06,
+      "loss": 0.4528,
+      "step": 2526
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.352865024462654,
+      "learning_rate": 7.100477406376255e-06,
+      "loss": 0.51,
+      "step": 2527
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.491197550580401,
+      "learning_rate": 7.098260601475798e-06,
+      "loss": 0.4745,
+      "step": 2528
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.673017389117518,
+      "learning_rate": 7.096043295847035e-06,
+      "loss": 0.439,
+      "step": 2529
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.117805873415266,
+      "learning_rate": 7.0938254900191e-06,
+      "loss": 0.4942,
+      "step": 2530
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.208658959351665,
+      "learning_rate": 7.0916071845212546e-06,
+      "loss": 0.4647,
+      "step": 2531
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.984412937025655,
+      "learning_rate": 7.089388379882869e-06,
+      "loss": 0.5581,
+      "step": 2532
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9007183279521063,
+      "learning_rate": 7.087169076633442e-06,
+      "loss": 0.5013,
+      "step": 2533
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.48976722203762,
+      "learning_rate": 7.084949275302586e-06,
+      "loss": 0.5164,
+      "step": 2534
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.246883197610956,
+      "learning_rate": 7.082728976420033e-06,
+      "loss": 0.5332,
+      "step": 2535
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.246441603180961,
+      "learning_rate": 7.080508180515636e-06,
+      "loss": 0.5152,
+      "step": 2536
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7679058375365506,
+      "learning_rate": 7.078286888119364e-06,
+      "loss": 0.4669,
+      "step": 2537
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9939204605539587,
+      "learning_rate": 7.076065099761307e-06,
+      "loss": 0.4893,
+      "step": 2538
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.434791090843588,
+      "learning_rate": 7.0738428159716695e-06,
+      "loss": 0.4548,
+      "step": 2539
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.319598766866957,
+      "learning_rate": 7.0716200372807795e-06,
+      "loss": 0.3992,
+      "step": 2540
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.64989431744513,
+      "learning_rate": 7.0693967642190784e-06,
+      "loss": 0.4575,
+      "step": 2541
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.951892777578616,
+      "learning_rate": 7.0671729973171275e-06,
+      "loss": 0.4577,
+      "step": 2542
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.497582046138382,
+      "learning_rate": 7.064948737105607e-06,
+      "loss": 0.3853,
+      "step": 2543
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.771240642969287,
+      "learning_rate": 7.0627239841153124e-06,
+      "loss": 0.476,
+      "step": 2544
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.6398774430586376,
+      "learning_rate": 7.060498738877159e-06,
+      "loss": 0.4996,
+      "step": 2545
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.69946825518343,
+      "learning_rate": 7.058273001922179e-06,
+      "loss": 0.4385,
+      "step": 2546
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.5087073478259616,
+      "learning_rate": 7.056046773781521e-06,
+      "loss": 0.5168,
+      "step": 2547
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7703037449211636,
+      "learning_rate": 7.05382005498645e-06,
+      "loss": 0.4804,
+      "step": 2548
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7462161805338607,
+      "learning_rate": 7.051592846068351e-06,
+      "loss": 0.4503,
+      "step": 2549
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6226172722161563,
+      "learning_rate": 7.049365147558726e-06,
+      "loss": 0.4589,
+      "step": 2550
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.4481356288883025,
+      "learning_rate": 7.047136959989193e-06,
+      "loss": 0.4107,
+      "step": 2551
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.801266433403592,
+      "learning_rate": 7.044908283891481e-06,
+      "loss": 0.4935,
+      "step": 2552
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.068944555113061,
+      "learning_rate": 7.042679119797443e-06,
+      "loss": 0.4857,
+      "step": 2553
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.584399671310988,
+      "learning_rate": 7.040449468239048e-06,
+      "loss": 0.4576,
+      "step": 2554
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 89.62009189447063,
+      "learning_rate": 7.038219329748377e-06,
+      "loss": 0.5017,
+      "step": 2555
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.08707089089859,
+      "learning_rate": 7.0359887048576325e-06,
+      "loss": 0.4956,
+      "step": 2556
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.529183891188903,
+      "learning_rate": 7.033757594099126e-06,
+      "loss": 0.4326,
+      "step": 2557
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.6028692682055605,
+      "learning_rate": 7.031525998005293e-06,
+      "loss": 0.53,
+      "step": 2558
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7971696327190503,
+      "learning_rate": 7.029293917108678e-06,
+      "loss": 0.5071,
+      "step": 2559
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.487357698147262,
+      "learning_rate": 7.027061351941948e-06,
+      "loss": 0.403,
+      "step": 2560
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.2582944488532712,
+      "learning_rate": 7.0248283030378775e-06,
+      "loss": 0.4743,
+      "step": 2561
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7277108087876063,
+      "learning_rate": 7.022594770929363e-06,
+      "loss": 0.4105,
+      "step": 2562
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.279027233181638,
+      "learning_rate": 7.0203607561494125e-06,
+      "loss": 0.5217,
+      "step": 2563
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.8943965921958297,
+      "learning_rate": 7.018126259231151e-06,
+      "loss": 0.5714,
+      "step": 2564
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.499912385775661,
+      "learning_rate": 7.01589128070782e-06,
+      "loss": 0.4685,
+      "step": 2565
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.634490660377617,
+      "learning_rate": 7.013655821112774e-06,
+      "loss": 0.4815,
+      "step": 2566
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.7637684660022455,
+      "learning_rate": 7.01141988097948e-06,
+      "loss": 0.4504,
+      "step": 2567
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.7158219227317977,
+      "learning_rate": 7.009183460841523e-06,
+      "loss": 0.4582,
+      "step": 2568
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.298136652730904,
+      "learning_rate": 7.006946561232602e-06,
+      "loss": 0.5172,
+      "step": 2569
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.296329536122423,
+      "learning_rate": 7.004709182686531e-06,
+      "loss": 0.5034,
+      "step": 2570
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.531712317033271,
+      "learning_rate": 7.002471325737236e-06,
+      "loss": 0.4295,
+      "step": 2571
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.391637753698668,
+      "learning_rate": 7.000232990918758e-06,
+      "loss": 0.4927,
+      "step": 2572
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.180925132182002,
+      "learning_rate": 6.997994178765255e-06,
+      "loss": 0.4871,
+      "step": 2573
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.303999549390409,
+      "learning_rate": 6.995754889810995e-06,
+      "loss": 0.4929,
+      "step": 2574
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.3592218086160037,
+      "learning_rate": 6.993515124590362e-06,
+      "loss": 0.481,
+      "step": 2575
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.9474698849426706,
+      "learning_rate": 6.9912748836378505e-06,
+      "loss": 0.4834,
+      "step": 2576
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.190630069585373,
+      "learning_rate": 6.989034167488073e-06,
+      "loss": 0.509,
+      "step": 2577
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.831245770172945,
+      "learning_rate": 6.986792976675753e-06,
+      "loss": 0.5488,
+      "step": 2578
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2584833943129166,
+      "learning_rate": 6.98455131173573e-06,
+      "loss": 0.5645,
+      "step": 2579
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.6062386734017196,
+      "learning_rate": 6.982309173202951e-06,
+      "loss": 0.5275,
+      "step": 2580
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.740900755627528,
+      "learning_rate": 6.98006656161248e-06,
+      "loss": 0.546,
+      "step": 2581
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 16.033685579990035,
+      "learning_rate": 6.9778234774994935e-06,
+      "loss": 0.5056,
+      "step": 2582
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.817444108062227,
+      "learning_rate": 6.97557992139928e-06,
+      "loss": 0.4031,
+      "step": 2583
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.7861456121972408,
+      "learning_rate": 6.9733358938472425e-06,
+      "loss": 0.4846,
+      "step": 2584
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.9977890308450332,
+      "learning_rate": 6.971091395378895e-06,
+      "loss": 0.4232,
+      "step": 2585
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.4247895174454985,
+      "learning_rate": 6.96884642652986e-06,
+      "loss": 0.429,
+      "step": 2586
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3825941655559477,
+      "learning_rate": 6.966600987835882e-06,
+      "loss": 0.62,
+      "step": 2587
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.201280933879752,
+      "learning_rate": 6.964355079832808e-06,
+      "loss": 0.4298,
+      "step": 2588
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.560599717145375,
+      "learning_rate": 6.962108703056602e-06,
+      "loss": 0.4791,
+      "step": 2589
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.4151280085291935,
+      "learning_rate": 6.95986185804334e-06,
+      "loss": 0.4078,
+      "step": 2590
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.751651247825529,
+      "learning_rate": 6.957614545329207e-06,
+      "loss": 0.468,
+      "step": 2591
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.4408467067611945,
+      "learning_rate": 6.9553667654504995e-06,
+      "loss": 0.4543,
+      "step": 2592
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 11.750691414016371,
+      "learning_rate": 6.953118518943629e-06,
+      "loss": 0.4532,
+      "step": 2593
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.756542615531776,
+      "learning_rate": 6.950869806345117e-06,
+      "loss": 0.4617,
+      "step": 2594
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1192898662299482,
+      "learning_rate": 6.948620628191596e-06,
+      "loss": 0.5645,
+      "step": 2595
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.420109002475213,
+      "learning_rate": 6.9463709850198055e-06,
+      "loss": 0.4584,
+      "step": 2596
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.393486846886865,
+      "learning_rate": 6.944120877366605e-06,
+      "loss": 0.4592,
+      "step": 2597
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.636295957254785,
+      "learning_rate": 6.941870305768954e-06,
+      "loss": 0.4767,
+      "step": 2598
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.395552489517268,
+      "learning_rate": 6.939619270763933e-06,
+      "loss": 0.4649,
+      "step": 2599
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.602451829478677,
+      "learning_rate": 6.937367772888726e-06,
+      "loss": 0.4931,
+      "step": 2600
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.37539369439954,
+      "learning_rate": 6.935115812680629e-06,
+      "loss": 0.5108,
+      "step": 2601
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.113245454140188,
+      "learning_rate": 6.932863390677052e-06,
+      "loss": 0.4491,
+      "step": 2602
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.5929787021613375,
+      "learning_rate": 6.93061050741551e-06,
+      "loss": 0.4919,
+      "step": 2603
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.731494332202367,
+      "learning_rate": 6.928357163433631e-06,
+      "loss": 0.4634,
+      "step": 2604
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.75513019726705,
+      "learning_rate": 6.926103359269152e-06,
+      "loss": 0.4525,
+      "step": 2605
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.562060049884234,
+      "learning_rate": 6.9238490954599215e-06,
+      "loss": 0.5108,
+      "step": 2606
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.474435969949516,
+      "learning_rate": 6.921594372543894e-06,
+      "loss": 0.5554,
+      "step": 2607
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.190822645139277,
+      "learning_rate": 6.91933919105914e-06,
+      "loss": 0.5453,
+      "step": 2608
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.384564796257071,
+      "learning_rate": 6.917083551543832e-06,
+      "loss": 0.5438,
+      "step": 2609
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.8306014508925017,
+      "learning_rate": 6.914827454536255e-06,
+      "loss": 0.428,
+      "step": 2610
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.965405593412001,
+      "learning_rate": 6.912570900574805e-06,
+      "loss": 0.616,
+      "step": 2611
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.0653542434579135,
+      "learning_rate": 6.910313890197983e-06,
+      "loss": 0.4735,
+      "step": 2612
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.590459889582739,
+      "learning_rate": 6.908056423944404e-06,
+      "loss": 0.4413,
+      "step": 2613
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 26.201162635074574,
+      "learning_rate": 6.905798502352788e-06,
+      "loss": 0.5028,
+      "step": 2614
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.153301014924827,
+      "learning_rate": 6.903540125961966e-06,
+      "loss": 0.4381,
+      "step": 2615
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.043632976805833,
+      "learning_rate": 6.9012812953108746e-06,
+      "loss": 0.4538,
+      "step": 2616
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.920155585033736,
+      "learning_rate": 6.8990220109385595e-06,
+      "loss": 0.4753,
+      "step": 2617
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.7248487419543044,
+      "learning_rate": 6.896762273384179e-06,
+      "loss": 0.4228,
+      "step": 2618
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.510265667654125,
+      "learning_rate": 6.894502083186993e-06,
+      "loss": 0.4925,
+      "step": 2619
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.166504893131973,
+      "learning_rate": 6.892241440886377e-06,
+      "loss": 0.5124,
+      "step": 2620
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.151127580555455,
+      "learning_rate": 6.889980347021804e-06,
+      "loss": 0.5123,
+      "step": 2621
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8269478317390186,
+      "learning_rate": 6.887718802132864e-06,
+      "loss": 0.4582,
+      "step": 2622
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3073494768057674,
+      "learning_rate": 6.885456806759253e-06,
+      "loss": 0.5866,
+      "step": 2623
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.750272131537456,
+      "learning_rate": 6.88319436144077e-06,
+      "loss": 0.4874,
+      "step": 2624
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7393839217074505,
+      "learning_rate": 6.880931466717327e-06,
+      "loss": 0.3877,
+      "step": 2625
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.707357918916511,
+      "learning_rate": 6.878668123128939e-06,
+      "loss": 0.4752,
+      "step": 2626
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5067168084540215,
+      "learning_rate": 6.876404331215728e-06,
+      "loss": 0.5904,
+      "step": 2627
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.538787638451028,
+      "learning_rate": 6.874140091517927e-06,
+      "loss": 0.4801,
+      "step": 2628
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.435453555584165,
+      "learning_rate": 6.871875404575874e-06,
+      "loss": 0.4794,
+      "step": 2629
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.94482919738694,
+      "learning_rate": 6.8696102709300104e-06,
+      "loss": 0.4962,
+      "step": 2630
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.989348944456226,
+      "learning_rate": 6.867344691120891e-06,
+      "loss": 0.4813,
+      "step": 2631
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.077974254547494,
+      "learning_rate": 6.865078665689169e-06,
+      "loss": 0.4352,
+      "step": 2632
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.929602877981309,
+      "learning_rate": 6.86281219517561e-06,
+      "loss": 0.5044,
+      "step": 2633
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7794802641800005,
+      "learning_rate": 6.860545280121086e-06,
+      "loss": 0.464,
+      "step": 2634
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.8820943903575693,
+      "learning_rate": 6.858277921066568e-06,
+      "loss": 0.484,
+      "step": 2635
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.942733930052468,
+      "learning_rate": 6.856010118553143e-06,
+      "loss": 0.4804,
+      "step": 2636
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0416033016526267,
+      "learning_rate": 6.8537418731219955e-06,
+      "loss": 0.5735,
+      "step": 2637
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.392032665513893,
+      "learning_rate": 6.851473185314419e-06,
+      "loss": 0.4148,
+      "step": 2638
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.30508253906485,
+      "learning_rate": 6.849204055671815e-06,
+      "loss": 0.4815,
+      "step": 2639
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.76620153612275,
+      "learning_rate": 6.846934484735685e-06,
+      "loss": 0.4613,
+      "step": 2640
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4042593511934442,
+      "learning_rate": 6.844664473047641e-06,
+      "loss": 0.473,
+      "step": 2641
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.2144097064836137,
+      "learning_rate": 6.8423940211493964e-06,
+      "loss": 0.4952,
+      "step": 2642
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6414567024726314,
+      "learning_rate": 6.840123129582773e-06,
+      "loss": 0.5099,
+      "step": 2643
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0920252857656245,
+      "learning_rate": 6.837851798889694e-06,
+      "loss": 0.451,
+      "step": 2644
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.073033550987081,
+      "learning_rate": 6.83558002961219e-06,
+      "loss": 0.4978,
+      "step": 2645
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.663003090004529,
+      "learning_rate": 6.833307822292393e-06,
+      "loss": 0.4602,
+      "step": 2646
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.653286777601565,
+      "learning_rate": 6.831035177472545e-06,
+      "loss": 0.4853,
+      "step": 2647
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.406204248708831,
+      "learning_rate": 6.82876209569499e-06,
+      "loss": 0.4933,
+      "step": 2648
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.94126719179757,
+      "learning_rate": 6.826488577502173e-06,
+      "loss": 0.402,
+      "step": 2649
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.236495961773163,
+      "learning_rate": 6.824214623436645e-06,
+      "loss": 0.4123,
+      "step": 2650
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.4983167598309715,
+      "learning_rate": 6.821940234041065e-06,
+      "loss": 0.4863,
+      "step": 2651
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.365559063673132,
+      "learning_rate": 6.819665409858188e-06,
+      "loss": 0.5593,
+      "step": 2652
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.526755912636174,
+      "learning_rate": 6.817390151430884e-06,
+      "loss": 0.4294,
+      "step": 2653
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 17.18988335039718,
+      "learning_rate": 6.815114459302115e-06,
+      "loss": 0.4283,
+      "step": 2654
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.6461214322071704,
+      "learning_rate": 6.8128383340149516e-06,
+      "loss": 0.4337,
+      "step": 2655
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.053656956856703,
+      "learning_rate": 6.81056177611257e-06,
+      "loss": 0.4957,
+      "step": 2656
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.7726857858812695,
+      "learning_rate": 6.808284786138246e-06,
+      "loss": 0.479,
+      "step": 2657
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6193265523656386,
+      "learning_rate": 6.806007364635361e-06,
+      "loss": 0.4444,
+      "step": 2658
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.791467929122126,
+      "learning_rate": 6.803729512147396e-06,
+      "loss": 0.4993,
+      "step": 2659
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.739295639482101,
+      "learning_rate": 6.801451229217938e-06,
+      "loss": 0.4635,
+      "step": 2660
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.52218733029149,
+      "learning_rate": 6.799172516390678e-06,
+      "loss": 0.5663,
+      "step": 2661
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6392401882982703,
+      "learning_rate": 6.7968933742094035e-06,
+      "loss": 0.4145,
+      "step": 2662
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.830439049912536,
+      "learning_rate": 6.794613803218012e-06,
+      "loss": 0.5252,
+      "step": 2663
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4767301364663012,
+      "learning_rate": 6.792333803960497e-06,
+      "loss": 0.4029,
+      "step": 2664
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4096789682837514,
+      "learning_rate": 6.790053376980958e-06,
+      "loss": 0.4777,
+      "step": 2665
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8997601813328835,
+      "learning_rate": 6.787772522823596e-06,
+      "loss": 0.4425,
+      "step": 2666
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.877382753823121,
+      "learning_rate": 6.7854912420327114e-06,
+      "loss": 0.4772,
+      "step": 2667
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.7273241582245613,
+      "learning_rate": 6.783209535152712e-06,
+      "loss": 0.4881,
+      "step": 2668
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5765791607966126,
+      "learning_rate": 6.780927402728099e-06,
+      "loss": 0.5036,
+      "step": 2669
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9779269179484147,
+      "learning_rate": 6.778644845303482e-06,
+      "loss": 0.4409,
+      "step": 2670
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7333926579117813,
+      "learning_rate": 6.776361863423572e-06,
+      "loss": 0.4434,
+      "step": 2671
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.7413579779823234,
+      "learning_rate": 6.774078457633177e-06,
+      "loss": 0.4753,
+      "step": 2672
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.3693737489658044,
+      "learning_rate": 6.771794628477209e-06,
+      "loss": 0.4619,
+      "step": 2673
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.365353032449093,
+      "learning_rate": 6.769510376500677e-06,
+      "loss": 0.4195,
+      "step": 2674
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.382994192876083,
+      "learning_rate": 6.767225702248699e-06,
+      "loss": 0.4237,
+      "step": 2675
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.852593872063646,
+      "learning_rate": 6.764940606266485e-06,
+      "loss": 0.5618,
+      "step": 2676
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.902473712492849,
+      "learning_rate": 6.762655089099353e-06,
+      "loss": 0.4808,
+      "step": 2677
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.385107298871977,
+      "learning_rate": 6.760369151292719e-06,
+      "loss": 0.5603,
+      "step": 2678
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9931059468018706,
+      "learning_rate": 6.758082793392094e-06,
+      "loss": 0.5248,
+      "step": 2679
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9078021827484686,
+      "learning_rate": 6.755796015943097e-06,
+      "loss": 0.4673,
+      "step": 2680
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.063288847436795,
+      "learning_rate": 6.753508819491443e-06,
+      "loss": 0.4333,
+      "step": 2681
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.9265000427314765,
+      "learning_rate": 6.751221204582948e-06,
+      "loss": 0.4893,
+      "step": 2682
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.3807908021344044,
+      "learning_rate": 6.748933171763527e-06,
+      "loss": 0.4696,
+      "step": 2683
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.415496291709117,
+      "learning_rate": 6.746644721579198e-06,
+      "loss": 0.4686,
+      "step": 2684
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.424958446614556,
+      "learning_rate": 6.744355854576074e-06,
+      "loss": 0.4904,
+      "step": 2685
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.938663623029467,
+      "learning_rate": 6.742066571300368e-06,
+      "loss": 0.4533,
+      "step": 2686
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.5134535953486945,
+      "learning_rate": 6.739776872298398e-06,
+      "loss": 0.4413,
+      "step": 2687
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.575249574934905,
+      "learning_rate": 6.7374867581165745e-06,
+      "loss": 0.5178,
+      "step": 2688
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.442213835426479,
+      "learning_rate": 6.735196229301409e-06,
+      "loss": 0.5173,
+      "step": 2689
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9244530376008386,
+      "learning_rate": 6.732905286399516e-06,
+      "loss": 0.4261,
+      "step": 2690
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8923016878278243,
+      "learning_rate": 6.730613929957601e-06,
+      "loss": 0.5049,
+      "step": 2691
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.700886026041967,
+      "learning_rate": 6.728322160522476e-06,
+      "loss": 0.5345,
+      "step": 2692
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.131206645849946,
+      "learning_rate": 6.726029978641046e-06,
+      "loss": 0.4321,
+      "step": 2693
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3231796218617438,
+      "learning_rate": 6.723737384860319e-06,
+      "loss": 0.5587,
+      "step": 2694
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.23720195648198,
+      "learning_rate": 6.721444379727398e-06,
+      "loss": 0.4699,
+      "step": 2695
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.067988253870886,
+      "learning_rate": 6.719150963789484e-06,
+      "loss": 0.4491,
+      "step": 2696
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.792634355597209,
+      "learning_rate": 6.7168571375938775e-06,
+      "loss": 0.4494,
+      "step": 2697
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.7008036062053256,
+      "learning_rate": 6.714562901687978e-06,
+      "loss": 0.4924,
+      "step": 2698
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.170969714281913,
+      "learning_rate": 6.712268256619279e-06,
+      "loss": 0.4247,
+      "step": 2699
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.560364176622954,
+      "learning_rate": 6.7099732029353735e-06,
+      "loss": 0.4192,
+      "step": 2700
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 10.472714094273426,
+      "learning_rate": 6.707677741183957e-06,
+      "loss": 0.4725,
+      "step": 2701
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.1451821467443475,
+      "learning_rate": 6.705381871912813e-06,
+      "loss": 0.5135,
+      "step": 2702
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.2289625758079215,
+      "learning_rate": 6.703085595669832e-06,
+      "loss": 0.4566,
+      "step": 2703
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8351475757162126,
+      "learning_rate": 6.70078891300299e-06,
+      "loss": 0.394,
+      "step": 2704
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7606016174901455,
+      "learning_rate": 6.698491824460371e-06,
+      "loss": 0.4341,
+      "step": 2705
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.930043687640907,
+      "learning_rate": 6.6961943305901515e-06,
+      "loss": 0.4231,
+      "step": 2706
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 17.15976904357539,
+      "learning_rate": 6.693896431940605e-06,
+      "loss": 0.9852,
+      "step": 2707
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.857941503136434,
+      "learning_rate": 6.6915981290601e-06,
+      "loss": 0.5087,
+      "step": 2708
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6352471493517857,
+      "learning_rate": 6.6892994224971044e-06,
+      "loss": 0.5104,
+      "step": 2709
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.391006285927482,
+      "learning_rate": 6.6870003128001784e-06,
+      "loss": 0.5035,
+      "step": 2710
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.589500913637127,
+      "learning_rate": 6.684700800517981e-06,
+      "loss": 0.4585,
+      "step": 2711
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.365026576219805,
+      "learning_rate": 6.682400886199272e-06,
+      "loss": 0.4772,
+      "step": 2712
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.490439023431818,
+      "learning_rate": 6.680100570392898e-06,
+      "loss": 0.5004,
+      "step": 2713
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1036526206907173,
+      "learning_rate": 6.677799853647804e-06,
+      "loss": 0.4785,
+      "step": 2714
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.4725433148499447,
+      "learning_rate": 6.675498736513036e-06,
+      "loss": 0.4204,
+      "step": 2715
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.750064706379368,
+      "learning_rate": 6.673197219537731e-06,
+      "loss": 0.5236,
+      "step": 2716
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.923850515142398,
+      "learning_rate": 6.670895303271123e-06,
+      "loss": 0.5273,
+      "step": 2717
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.204699151378934,
+      "learning_rate": 6.668592988262539e-06,
+      "loss": 0.5328,
+      "step": 2718
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.654279067061114,
+      "learning_rate": 6.666290275061404e-06,
+      "loss": 0.4557,
+      "step": 2719
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9806359355838428,
+      "learning_rate": 6.663987164217236e-06,
+      "loss": 0.4571,
+      "step": 2720
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.730724760495364,
+      "learning_rate": 6.66168365627965e-06,
+      "loss": 0.4758,
+      "step": 2721
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.565231152658073,
+      "learning_rate": 6.659379751798352e-06,
+      "loss": 0.5295,
+      "step": 2722
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.6104070435410693,
+      "learning_rate": 6.657075451323148e-06,
+      "loss": 0.4436,
+      "step": 2723
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.834102785072117,
+      "learning_rate": 6.654770755403934e-06,
+      "loss": 0.5049,
+      "step": 2724
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.2652921551846097,
+      "learning_rate": 6.652465664590702e-06,
+      "loss": 0.4132,
+      "step": 2725
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.202111885046156,
+      "learning_rate": 6.650160179433539e-06,
+      "loss": 0.4889,
+      "step": 2726
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.784382570564905,
+      "learning_rate": 6.647854300482626e-06,
+      "loss": 0.434,
+      "step": 2727
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0026001243955185,
+      "learning_rate": 6.645548028288234e-06,
+      "loss": 0.5168,
+      "step": 2728
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.922247361610575,
+      "learning_rate": 6.643241363400732e-06,
+      "loss": 0.4567,
+      "step": 2729
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.6272641005123125,
+      "learning_rate": 6.640934306370585e-06,
+      "loss": 0.4556,
+      "step": 2730
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.772712535106437,
+      "learning_rate": 6.6386268577483474e-06,
+      "loss": 0.5463,
+      "step": 2731
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.38711761039972,
+      "learning_rate": 6.636319018084667e-06,
+      "loss": 0.5496,
+      "step": 2732
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.984509874288353,
+      "learning_rate": 6.634010787930284e-06,
+      "loss": 0.4662,
+      "step": 2733
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1742101561885017,
+      "learning_rate": 6.631702167836037e-06,
+      "loss": 0.4679,
+      "step": 2734
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.682229800713317,
+      "learning_rate": 6.629393158352854e-06,
+      "loss": 0.5045,
+      "step": 2735
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6638278667558954,
+      "learning_rate": 6.627083760031755e-06,
+      "loss": 0.4134,
+      "step": 2736
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.4168407905465195,
+      "learning_rate": 6.624773973423856e-06,
+      "loss": 0.4827,
+      "step": 2737
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.037439292160289,
+      "learning_rate": 6.622463799080362e-06,
+      "loss": 0.4877,
+      "step": 2738
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2624109053429897,
+      "learning_rate": 6.620153237552572e-06,
+      "loss": 0.5612,
+      "step": 2739
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.513823717854631,
+      "learning_rate": 6.617842289391881e-06,
+      "loss": 0.4468,
+      "step": 2740
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.593256282675752,
+      "learning_rate": 6.615530955149768e-06,
+      "loss": 0.4998,
+      "step": 2741
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.1171043512366605,
+      "learning_rate": 6.613219235377815e-06,
+      "loss": 0.4787,
+      "step": 2742
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.900379386048598,
+      "learning_rate": 6.610907130627685e-06,
+      "loss": 0.3699,
+      "step": 2743
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.394613372515317,
+      "learning_rate": 6.60859464145114e-06,
+      "loss": 0.5168,
+      "step": 2744
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.217660764172856,
+      "learning_rate": 6.606281768400032e-06,
+      "loss": 0.4124,
+      "step": 2745
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.916730129757846,
+      "learning_rate": 6.603968512026305e-06,
+      "loss": 0.5176,
+      "step": 2746
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.302544562951499,
+      "learning_rate": 6.601654872881993e-06,
+      "loss": 0.4461,
+      "step": 2747
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.583520799382047,
+      "learning_rate": 6.5993408515192224e-06,
+      "loss": 0.4112,
+      "step": 2748
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.622447817236105,
+      "learning_rate": 6.59702644849021e-06,
+      "loss": 0.4569,
+      "step": 2749
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 14.327070902254965,
+      "learning_rate": 6.5947116643472646e-06,
+      "loss": 0.4354,
+      "step": 2750
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.266120893302219,
+      "learning_rate": 6.592396499642786e-06,
+      "loss": 0.4705,
+      "step": 2751
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.911482668587954,
+      "learning_rate": 6.5900809549292646e-06,
+      "loss": 0.4877,
+      "step": 2752
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.1065688896443655,
+      "learning_rate": 6.587765030759279e-06,
+      "loss": 0.4363,
+      "step": 2753
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.30583973820975,
+      "learning_rate": 6.585448727685505e-06,
+      "loss": 0.4262,
+      "step": 2754
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1605210794029632,
+      "learning_rate": 6.583132046260701e-06,
+      "loss": 0.524,
+      "step": 2755
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.583254289165055,
+      "learning_rate": 6.580814987037719e-06,
+      "loss": 0.5538,
+      "step": 2756
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.897575167215788,
+      "learning_rate": 6.578497550569504e-06,
+      "loss": 0.452,
+      "step": 2757
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.633504636764331,
+      "learning_rate": 6.576179737409087e-06,
+      "loss": 0.4644,
+      "step": 2758
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.925804311796853,
+      "learning_rate": 6.573861548109589e-06,
+      "loss": 0.4791,
+      "step": 2759
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.528698096542152,
+      "learning_rate": 6.571542983224224e-06,
+      "loss": 0.4503,
+      "step": 2760
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.019478417165658,
+      "learning_rate": 6.569224043306293e-06,
+      "loss": 0.4652,
+      "step": 2761
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.897658191236235,
+      "learning_rate": 6.566904728909188e-06,
+      "loss": 0.4387,
+      "step": 2762
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.914205376551238,
+      "learning_rate": 6.564585040586388e-06,
+      "loss": 0.4214,
+      "step": 2763
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 26.90624661216952,
+      "learning_rate": 6.562264978891462e-06,
+      "loss": 0.4742,
+      "step": 2764
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.4542601911009205,
+      "learning_rate": 6.559944544378072e-06,
+      "loss": 0.5071,
+      "step": 2765
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 20.323793108299714,
+      "learning_rate": 6.557623737599963e-06,
+      "loss": 0.4653,
+      "step": 2766
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.130791431956046,
+      "learning_rate": 6.555302559110973e-06,
+      "loss": 0.4572,
+      "step": 2767
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.554891748820894,
+      "learning_rate": 6.552981009465027e-06,
+      "loss": 0.6201,
+      "step": 2768
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.87098358450737,
+      "learning_rate": 6.550659089216138e-06,
+      "loss": 0.4579,
+      "step": 2769
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.068972908832782,
+      "learning_rate": 6.548336798918411e-06,
+      "loss": 0.5396,
+      "step": 2770
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.827314715195966,
+      "learning_rate": 6.5460141391260345e-06,
+      "loss": 0.45,
+      "step": 2771
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 11.243824523464562,
+      "learning_rate": 6.543691110393289e-06,
+      "loss": 0.4345,
+      "step": 2772
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.203276108272592,
+      "learning_rate": 6.541367713274539e-06,
+      "loss": 0.5186,
+      "step": 2773
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.386941785964687,
+      "learning_rate": 6.539043948324239e-06,
+      "loss": 0.5822,
+      "step": 2774
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.838100108215123,
+      "learning_rate": 6.536719816096936e-06,
+      "loss": 0.4712,
+      "step": 2775
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.719951804297019,
+      "learning_rate": 6.534395317147257e-06,
+      "loss": 0.4603,
+      "step": 2776
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.897280460439331,
+      "learning_rate": 6.532070452029922e-06,
+      "loss": 0.4738,
+      "step": 2777
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3966527219859393,
+      "learning_rate": 6.529745221299731e-06,
+      "loss": 0.6114,
+      "step": 2778
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.436220128884038,
+      "learning_rate": 6.527419625511581e-06,
+      "loss": 0.4946,
+      "step": 2779
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 25.73745791775027,
+      "learning_rate": 6.52509366522045e-06,
+      "loss": 0.3347,
+      "step": 2780
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 15.190280469048854,
+      "learning_rate": 6.522767340981406e-06,
+      "loss": 0.4625,
+      "step": 2781
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2932950851439577,
+      "learning_rate": 6.5204406533496e-06,
+      "loss": 0.5365,
+      "step": 2782
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 12.829908481643743,
+      "learning_rate": 6.518113602880275e-06,
+      "loss": 0.497,
+      "step": 2783
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.175146064295874,
+      "learning_rate": 6.5157861901287545e-06,
+      "loss": 0.4694,
+      "step": 2784
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 16.411835261163326,
+      "learning_rate": 6.513458415650452e-06,
+      "loss": 0.5201,
+      "step": 2785
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.1997373048831514,
+      "learning_rate": 6.51113028000087e-06,
+      "loss": 0.525,
+      "step": 2786
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 13.510608987435019,
+      "learning_rate": 6.508801783735591e-06,
+      "loss": 0.5129,
+      "step": 2787
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 46.12006136789878,
+      "learning_rate": 6.5064729274102875e-06,
+      "loss": 0.5015,
+      "step": 2788
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.514114891752724,
+      "learning_rate": 6.504143711580718e-06,
+      "loss": 0.4637,
+      "step": 2789
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 19.662999277017793,
+      "learning_rate": 6.501814136802725e-06,
+      "loss": 0.5133,
+      "step": 2790
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.990169314082399,
+      "learning_rate": 6.499484203632238e-06,
+      "loss": 0.4741,
+      "step": 2791
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.690199786685389,
+      "learning_rate": 6.497153912625271e-06,
+      "loss": 0.5041,
+      "step": 2792
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.318113545468492,
+      "learning_rate": 6.494823264337925e-06,
+      "loss": 0.4537,
+      "step": 2793
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.4449557626080916,
+      "learning_rate": 6.492492259326384e-06,
+      "loss": 0.3757,
+      "step": 2794
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 15.016525407445265,
+      "learning_rate": 6.490160898146919e-06,
+      "loss": 0.4966,
+      "step": 2795
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 13.048025426642337,
+      "learning_rate": 6.487829181355887e-06,
+      "loss": 0.4937,
+      "step": 2796
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.3644151964490465,
+      "learning_rate": 6.4854971095097245e-06,
+      "loss": 0.4087,
+      "step": 2797
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.052544094906661,
+      "learning_rate": 6.483164683164958e-06,
+      "loss": 0.4557,
+      "step": 2798
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.395910897467711,
+      "learning_rate": 6.480831902878198e-06,
+      "loss": 0.5157,
+      "step": 2799
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 11.946570993195593,
+      "learning_rate": 6.47849876920614e-06,
+      "loss": 0.4455,
+      "step": 2800
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.9682337963206935,
+      "learning_rate": 6.476165282705559e-06,
+      "loss": 0.4763,
+      "step": 2801
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 12.55711060984555,
+      "learning_rate": 6.473831443933318e-06,
+      "loss": 0.385,
+      "step": 2802
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.912547954730923,
+      "learning_rate": 6.471497253446366e-06,
+      "loss": 0.4995,
+      "step": 2803
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.345932591804477,
+      "learning_rate": 6.469162711801731e-06,
+      "loss": 0.4921,
+      "step": 2804
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.083865790119315,
+      "learning_rate": 6.466827819556529e-06,
+      "loss": 0.4579,
+      "step": 2805
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.677919737187017,
+      "learning_rate": 6.464492577267958e-06,
+      "loss": 0.4584,
+      "step": 2806
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 13.30620290685418,
+      "learning_rate": 6.4621569854933e-06,
+      "loss": 0.4596,
+      "step": 2807
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.296907954068159,
+      "learning_rate": 6.459821044789918e-06,
+      "loss": 0.4988,
+      "step": 2808
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 33.04980072630393,
+      "learning_rate": 6.4574847557152595e-06,
+      "loss": 0.4587,
+      "step": 2809
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.7201287175221935,
+      "learning_rate": 6.45514811882686e-06,
+      "loss": 0.4357,
+      "step": 2810
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.302298262811125,
+      "learning_rate": 6.452811134682329e-06,
+      "loss": 0.5603,
+      "step": 2811
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.509496070364374,
+      "learning_rate": 6.45047380383937e-06,
+      "loss": 0.4915,
+      "step": 2812
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 15.623023107917717,
+      "learning_rate": 6.448136126855755e-06,
+      "loss": 0.5302,
+      "step": 2813
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 16.87577088574106,
+      "learning_rate": 6.4457981042893525e-06,
+      "loss": 0.5389,
+      "step": 2814
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.378613929374242,
+      "learning_rate": 6.443459736698106e-06,
+      "loss": 0.4677,
+      "step": 2815
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.793061140847403,
+      "learning_rate": 6.441121024640043e-06,
+      "loss": 0.4414,
+      "step": 2816
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 16.861273084357705,
+      "learning_rate": 6.438781968673271e-06,
+      "loss": 0.4633,
+      "step": 2817
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 27.011331304719207,
+      "learning_rate": 6.436442569355984e-06,
+      "loss": 0.4701,
+      "step": 2818
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.70671248446562,
+      "learning_rate": 6.4341028272464545e-06,
+      "loss": 0.4488,
+      "step": 2819
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.547948149137019,
+      "learning_rate": 6.4317627429030385e-06,
+      "loss": 0.4944,
+      "step": 2820
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.668611592824467,
+      "learning_rate": 6.429422316884174e-06,
+      "loss": 0.386,
+      "step": 2821
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 10.367443276145769,
+      "learning_rate": 6.427081549748377e-06,
+      "loss": 0.5022,
+      "step": 2822
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.768215903741776,
+      "learning_rate": 6.4247404420542495e-06,
+      "loss": 0.3762,
+      "step": 2823
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 10.087593458796759,
+      "learning_rate": 6.422398994360473e-06,
+      "loss": 0.4221,
+      "step": 2824
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.9940429562708175,
+      "learning_rate": 6.420057207225807e-06,
+      "loss": 0.5322,
+      "step": 2825
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.998240436246018,
+      "learning_rate": 6.417715081209098e-06,
+      "loss": 0.4671,
+      "step": 2826
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 10.367729028584408,
+      "learning_rate": 6.415372616869269e-06,
+      "loss": 0.4909,
+      "step": 2827
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.7039831743808,
+      "learning_rate": 6.413029814765326e-06,
+      "loss": 0.4222,
+      "step": 2828
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.812814161463242,
+      "learning_rate": 6.410686675456354e-06,
+      "loss": 0.397,
+      "step": 2829
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.167465800082248,
+      "learning_rate": 6.408343199501519e-06,
+      "loss": 0.5039,
+      "step": 2830
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.224493902731253,
+      "learning_rate": 6.4059993874600665e-06,
+      "loss": 0.4704,
+      "step": 2831
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.032891669575696,
+      "learning_rate": 6.4036552398913244e-06,
+      "loss": 0.4686,
+      "step": 2832
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 11.38400033097673,
+      "learning_rate": 6.401310757354698e-06,
+      "loss": 0.4795,
+      "step": 2833
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.645819897136621,
+      "learning_rate": 6.398965940409676e-06,
+      "loss": 0.4452,
+      "step": 2834
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.188581229438318,
+      "learning_rate": 6.396620789615825e-06,
+      "loss": 0.5087,
+      "step": 2835
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.871289230179782,
+      "learning_rate": 6.39427530553279e-06,
+      "loss": 0.4458,
+      "step": 2836
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.4049003430802145,
+      "learning_rate": 6.391929488720294e-06,
+      "loss": 0.5149,
+      "step": 2837
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 10.797600214784543,
+      "learning_rate": 6.3895833397381464e-06,
+      "loss": 0.5331,
+      "step": 2838
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 10.833469750964092,
+      "learning_rate": 6.38723685914623e-06,
+      "loss": 0.4709,
+      "step": 2839
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 20.736665990322223,
+      "learning_rate": 6.3848900475045086e-06,
+      "loss": 0.4775,
+      "step": 2840
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 14.638327172300627,
+      "learning_rate": 6.382542905373022e-06,
+      "loss": 0.5171,
+      "step": 2841
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.702794886538009,
+      "learning_rate": 6.3801954333118955e-06,
+      "loss": 0.5315,
+      "step": 2842
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.925505891023349,
+      "learning_rate": 6.377847631881326e-06,
+      "loss": 0.4176,
+      "step": 2843
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.628373974834523,
+      "learning_rate": 6.3754995016415936e-06,
+      "loss": 0.448,
+      "step": 2844
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.734274034482985,
+      "learning_rate": 6.373151043153056e-06,
+      "loss": 0.4636,
+      "step": 2845
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.02214068032078,
+      "learning_rate": 6.370802256976148e-06,
+      "loss": 0.46,
+      "step": 2846
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.225247930226508,
+      "learning_rate": 6.368453143671383e-06,
+      "loss": 0.5111,
+      "step": 2847
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.876982908781532,
+      "learning_rate": 6.366103703799351e-06,
+      "loss": 0.4639,
+      "step": 2848
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.788792257363526,
+      "learning_rate": 6.363753937920725e-06,
+      "loss": 0.5428,
+      "step": 2849
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.585629500633913,
+      "learning_rate": 6.3614038465962514e-06,
+      "loss": 0.4506,
+      "step": 2850
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 19.438511110730275,
+      "learning_rate": 6.359053430386755e-06,
+      "loss": 0.426,
+      "step": 2851
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.928408966575904,
+      "learning_rate": 6.356702689853137e-06,
+      "loss": 0.4141,
+      "step": 2852
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.621456090744973,
+      "learning_rate": 6.354351625556381e-06,
+      "loss": 0.4445,
+      "step": 2853
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.5093117084897054,
+      "learning_rate": 6.3520002380575395e-06,
+      "loss": 0.5141,
+      "step": 2854
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.82002362526606,
+      "learning_rate": 6.349648527917752e-06,
+      "loss": 0.4355,
+      "step": 2855
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.172747252242528,
+      "learning_rate": 6.347296495698225e-06,
+      "loss": 0.5529,
+      "step": 2856
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.5994608789766995,
+      "learning_rate": 6.344944141960251e-06,
+      "loss": 0.4565,
+      "step": 2857
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 31.020871409318158,
+      "learning_rate": 6.342591467265193e-06,
+      "loss": 0.4096,
+      "step": 2858
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2352114791426783,
+      "learning_rate": 6.3402384721744935e-06,
+      "loss": 0.5142,
+      "step": 2859
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.1856316205273822,
+      "learning_rate": 6.33788515724967e-06,
+      "loss": 0.5133,
+      "step": 2860
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.749681326450956,
+      "learning_rate": 6.335531523052316e-06,
+      "loss": 0.4016,
+      "step": 2861
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 120.5668977647027,
+      "learning_rate": 6.333177570144103e-06,
+      "loss": 0.4847,
+      "step": 2862
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.707753304881009,
+      "learning_rate": 6.330823299086777e-06,
+      "loss": 0.4409,
+      "step": 2863
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 10.447715442702926,
+      "learning_rate": 6.328468710442163e-06,
+      "loss": 0.4939,
+      "step": 2864
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.226041478807944,
+      "learning_rate": 6.326113804772157e-06,
+      "loss": 0.4415,
+      "step": 2865
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.833078021697118,
+      "learning_rate": 6.3237585826387325e-06,
+      "loss": 0.4589,
+      "step": 2866
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.486664846964552,
+      "learning_rate": 6.321403044603941e-06,
+      "loss": 0.4958,
+      "step": 2867
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.413022349163082,
+      "learning_rate": 6.319047191229906e-06,
+      "loss": 0.4783,
+      "step": 2868
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.099621573316986,
+      "learning_rate": 6.3166910230788284e-06,
+      "loss": 0.5262,
+      "step": 2869
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.572350286280132,
+      "learning_rate": 6.314334540712983e-06,
+      "loss": 0.444,
+      "step": 2870
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.686426869676033,
+      "learning_rate": 6.311977744694722e-06,
+      "loss": 0.4338,
+      "step": 2871
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.834913841187811,
+      "learning_rate": 6.309620635586466e-06,
+      "loss": 0.4183,
+      "step": 2872
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.081157674525627,
+      "learning_rate": 6.307263213950721e-06,
+      "loss": 0.4527,
+      "step": 2873
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.439799046523862,
+      "learning_rate": 6.304905480350055e-06,
+      "loss": 0.4303,
+      "step": 2874
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.352210517561995,
+      "learning_rate": 6.302547435347121e-06,
+      "loss": 0.4769,
+      "step": 2875
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.048311095692171,
+      "learning_rate": 6.300189079504644e-06,
+      "loss": 0.4039,
+      "step": 2876
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 15.943869898815723,
+      "learning_rate": 6.297830413385415e-06,
+      "loss": 0.4409,
+      "step": 2877
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.017159949879932,
+      "learning_rate": 6.295471437552309e-06,
+      "loss": 0.5156,
+      "step": 2878
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.167930754050777,
+      "learning_rate": 6.293112152568274e-06,
+      "loss": 0.4864,
+      "step": 2879
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.434436766618643,
+      "learning_rate": 6.290752558996325e-06,
+      "loss": 0.4384,
+      "step": 2880
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.8966942400788023,
+      "learning_rate": 6.288392657399555e-06,
+      "loss": 0.4764,
+      "step": 2881
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.304484480744687,
+      "learning_rate": 6.286032448341133e-06,
+      "loss": 0.4454,
+      "step": 2882
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.025010823083057,
+      "learning_rate": 6.283671932384296e-06,
+      "loss": 0.4403,
+      "step": 2883
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.814595715705873,
+      "learning_rate": 6.281311110092358e-06,
+      "loss": 0.4136,
+      "step": 2884
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 19.38526408510347,
+      "learning_rate": 6.278949982028705e-06,
+      "loss": 0.4122,
+      "step": 2885
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.674326131668105,
+      "learning_rate": 6.276588548756795e-06,
+      "loss": 0.434,
+      "step": 2886
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.7460754247759183,
+      "learning_rate": 6.274226810840161e-06,
+      "loss": 0.4332,
+      "step": 2887
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.511335587793365,
+      "learning_rate": 6.271864768842406e-06,
+      "loss": 0.493,
+      "step": 2888
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2983960605419433,
+      "learning_rate": 6.2695024233272095e-06,
+      "loss": 0.5474,
+      "step": 2889
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.195212234202778,
+      "learning_rate": 6.267139774858318e-06,
+      "loss": 0.4485,
+      "step": 2890
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.9996118883143,
+      "learning_rate": 6.264776823999556e-06,
+      "loss": 0.4593,
+      "step": 2891
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 14.903877953451142,
+      "learning_rate": 6.262413571314814e-06,
+      "loss": 0.5282,
+      "step": 2892
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 129.60616215718593,
+      "learning_rate": 6.260050017368062e-06,
+      "loss": 0.448,
+      "step": 2893
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.992329053609898,
+      "learning_rate": 6.2576861627233355e-06,
+      "loss": 0.3721,
+      "step": 2894
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.775458184546797,
+      "learning_rate": 6.255322007944744e-06,
+      "loss": 0.521,
+      "step": 2895
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.276729067301927,
+      "learning_rate": 6.2529575535964685e-06,
+      "loss": 0.5539,
+      "step": 2896
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.387276350721105,
+      "learning_rate": 6.250592800242762e-06,
+      "loss": 0.4548,
+      "step": 2897
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.599231443871725,
+      "learning_rate": 6.248227748447949e-06,
+      "loss": 0.4904,
+      "step": 2898
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.804613201083422,
+      "learning_rate": 6.245862398776427e-06,
+      "loss": 0.506,
+      "step": 2899
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.42167919671319,
+      "learning_rate": 6.243496751792658e-06,
+      "loss": 0.4713,
+      "step": 2900
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.849394671621753,
+      "learning_rate": 6.241130808061182e-06,
+      "loss": 0.5078,
+      "step": 2901
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.8874500944831825,
+      "learning_rate": 6.238764568146606e-06,
+      "loss": 0.4847,
+      "step": 2902
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 16.206077161020335,
+      "learning_rate": 6.236398032613609e-06,
+      "loss": 0.4706,
+      "step": 2903
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.811333191992275,
+      "learning_rate": 6.234031202026942e-06,
+      "loss": 0.5102,
+      "step": 2904
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.271011927450286,
+      "learning_rate": 6.231664076951421e-06,
+      "loss": 0.4578,
+      "step": 2905
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.9928845148593854,
+      "learning_rate": 6.229296657951942e-06,
+      "loss": 0.4449,
+      "step": 2906
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 19.619239275853378,
+      "learning_rate": 6.226928945593459e-06,
+      "loss": 0.4396,
+      "step": 2907
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.51387948735736,
+      "learning_rate": 6.224560940441006e-06,
+      "loss": 0.492,
+      "step": 2908
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.58547597251968,
+      "learning_rate": 6.222192643059681e-06,
+      "loss": 0.4752,
+      "step": 2909
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.127400337743388,
+      "learning_rate": 6.219824054014656e-06,
+      "loss": 0.4914,
+      "step": 2910
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.298296148806085,
+      "learning_rate": 6.217455173871168e-06,
+      "loss": 0.4532,
+      "step": 2911
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.266291317185697,
+      "learning_rate": 6.215086003194528e-06,
+      "loss": 0.5083,
+      "step": 2912
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.063603729960131,
+      "learning_rate": 6.212716542550112e-06,
+      "loss": 0.4731,
+      "step": 2913
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.798648517513329,
+      "learning_rate": 6.21034679250337e-06,
+      "loss": 0.4797,
+      "step": 2914
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.163752400320914,
+      "learning_rate": 6.207976753619817e-06,
+      "loss": 0.442,
+      "step": 2915
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.554722972050721,
+      "learning_rate": 6.205606426465036e-06,
+      "loss": 0.4609,
+      "step": 2916
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.858743791737207,
+      "learning_rate": 6.203235811604687e-06,
+      "loss": 0.4178,
+      "step": 2917
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.430806749111668,
+      "learning_rate": 6.200864909604488e-06,
+      "loss": 0.463,
+      "step": 2918
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.683070536151628,
+      "learning_rate": 6.198493721030234e-06,
+      "loss": 0.4217,
+      "step": 2919
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.751872888152386,
+      "learning_rate": 6.196122246447779e-06,
+      "loss": 0.5221,
+      "step": 2920
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.220702100998107,
+      "learning_rate": 6.193750486423057e-06,
+      "loss": 0.4912,
+      "step": 2921
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.584599976744937,
+      "learning_rate": 6.191378441522061e-06,
+      "loss": 0.4837,
+      "step": 2922
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.298815818935616,
+      "learning_rate": 6.1890061123108565e-06,
+      "loss": 0.521,
+      "step": 2923
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.805676013351562,
+      "learning_rate": 6.186633499355576e-06,
+      "loss": 0.4255,
+      "step": 2924
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.917810138462583,
+      "learning_rate": 6.184260603222416e-06,
+      "loss": 0.4925,
+      "step": 2925
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.903642368695931,
+      "learning_rate": 6.181887424477646e-06,
+      "loss": 0.4293,
+      "step": 2926
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 12.539622236503787,
+      "learning_rate": 6.179513963687601e-06,
+      "loss": 0.4939,
+      "step": 2927
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 19.066186474789436,
+      "learning_rate": 6.177140221418684e-06,
+      "loss": 0.4281,
+      "step": 2928
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.1219456875319365,
+      "learning_rate": 6.174766198237362e-06,
+      "loss": 0.5037,
+      "step": 2929
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 29.955314239193036,
+      "learning_rate": 6.17239189471017e-06,
+      "loss": 0.526,
+      "step": 2930
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 16.007468067120975,
+      "learning_rate": 6.1700173114037146e-06,
+      "loss": 0.4886,
+      "step": 2931
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.564053407660757,
+      "learning_rate": 6.167642448884663e-06,
+      "loss": 0.5325,
+      "step": 2932
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.240686494865597,
+      "learning_rate": 6.165267307719754e-06,
+      "loss": 0.5792,
+      "step": 2933
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.253415449361312,
+      "learning_rate": 6.1628918884757895e-06,
+      "loss": 0.5987,
+      "step": 2934
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 24.433133552177132,
+      "learning_rate": 6.160516191719639e-06,
+      "loss": 0.4968,
+      "step": 2935
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.9931852423106236,
+      "learning_rate": 6.158140218018236e-06,
+      "loss": 0.462,
+      "step": 2936
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 19.236499732538242,
+      "learning_rate": 6.155763967938585e-06,
+      "loss": 0.4212,
+      "step": 2937
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.274228560984213,
+      "learning_rate": 6.153387442047753e-06,
+      "loss": 0.5129,
+      "step": 2938
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.294532477232685,
+      "learning_rate": 6.151010640912873e-06,
+      "loss": 0.569,
+      "step": 2939
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.491773618404347,
+      "learning_rate": 6.1486335651011455e-06,
+      "loss": 0.4585,
+      "step": 2940
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.56092831572986,
+      "learning_rate": 6.146256215179834e-06,
+      "loss": 0.435,
+      "step": 2941
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 11.236197360791412,
+      "learning_rate": 6.143878591716266e-06,
+      "loss": 0.4664,
+      "step": 2942
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.667814557033001,
+      "learning_rate": 6.141500695277841e-06,
+      "loss": 0.4689,
+      "step": 2943
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.74286595299248,
+      "learning_rate": 6.139122526432017e-06,
+      "loss": 0.4937,
+      "step": 2944
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.1687008557766445,
+      "learning_rate": 6.136744085746322e-06,
+      "loss": 0.49,
+      "step": 2945
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.098445146363322,
+      "learning_rate": 6.134365373788344e-06,
+      "loss": 0.4851,
+      "step": 2946
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2617298511860449,
+      "learning_rate": 6.13198639112574e-06,
+      "loss": 0.5173,
+      "step": 2947
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.057634896107505,
+      "learning_rate": 6.129607138326229e-06,
+      "loss": 0.5721,
+      "step": 2948
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.514425756558621,
+      "learning_rate": 6.127227615957594e-06,
+      "loss": 0.4605,
+      "step": 2949
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.322004129411983,
+      "learning_rate": 6.124847824587684e-06,
+      "loss": 0.5908,
+      "step": 2950
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.698888176923222,
+      "learning_rate": 6.1224677647844124e-06,
+      "loss": 0.4521,
+      "step": 2951
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.897884546136255,
+      "learning_rate": 6.120087437115757e-06,
+      "loss": 0.4381,
+      "step": 2952
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.502939995314409,
+      "learning_rate": 6.117706842149758e-06,
+      "loss": 0.5154,
+      "step": 2953
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.389267912671945,
+      "learning_rate": 6.115325980454517e-06,
+      "loss": 0.4302,
+      "step": 2954
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.7781321062543447,
+      "learning_rate": 6.112944852598205e-06,
+      "loss": 0.4931,
+      "step": 2955
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.893630618320507,
+      "learning_rate": 6.1105634591490525e-06,
+      "loss": 0.4351,
+      "step": 2956
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.101882233375057,
+      "learning_rate": 6.108181800675358e-06,
+      "loss": 0.3645,
+      "step": 2957
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.667989651168263,
+      "learning_rate": 6.105799877745475e-06,
+      "loss": 0.4367,
+      "step": 2958
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.842490529575291,
+      "learning_rate": 6.103417690927827e-06,
+      "loss": 0.4476,
+      "step": 2959
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.933347081976434,
+      "learning_rate": 6.101035240790897e-06,
+      "loss": 0.3655,
+      "step": 2960
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.3566911180389605,
+      "learning_rate": 6.098652527903235e-06,
+      "loss": 0.4941,
+      "step": 2961
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.994200167373279,
+      "learning_rate": 6.096269552833448e-06,
+      "loss": 0.4327,
+      "step": 2962
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.863694621041805,
+      "learning_rate": 6.09388631615021e-06,
+      "loss": 0.4955,
+      "step": 2963
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.045503656825029,
+      "learning_rate": 6.091502818422258e-06,
+      "loss": 0.484,
+      "step": 2964
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.017810884325633,
+      "learning_rate": 6.0891190602183845e-06,
+      "loss": 0.5009,
+      "step": 2965
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.515713266760821,
+      "learning_rate": 6.0867350421074524e-06,
+      "loss": 0.4287,
+      "step": 2966
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.32627328553966,
+      "learning_rate": 6.084350764658381e-06,
+      "loss": 0.4568,
+      "step": 2967
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 10.132788183300953,
+      "learning_rate": 6.081966228440156e-06,
+      "loss": 0.4727,
+      "step": 2968
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 12.956433016385434,
+      "learning_rate": 6.079581434021819e-06,
+      "loss": 0.4863,
+      "step": 2969
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.482706597331589,
+      "learning_rate": 6.077196381972482e-06,
+      "loss": 0.5453,
+      "step": 2970
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.695678259087538,
+      "learning_rate": 6.074811072861307e-06,
+      "loss": 0.3629,
+      "step": 2971
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.313988620049725,
+      "learning_rate": 6.072425507257528e-06,
+      "loss": 0.3593,
+      "step": 2972
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.535246856373871,
+      "learning_rate": 6.070039685730433e-06,
+      "loss": 0.4588,
+      "step": 2973
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.974839937793737,
+      "learning_rate": 6.067653608849376e-06,
+      "loss": 0.4814,
+      "step": 2974
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 69.96755068662458,
+      "learning_rate": 6.065267277183767e-06,
+      "loss": 0.4312,
+      "step": 2975
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.180487673274635,
+      "learning_rate": 6.0628806913030815e-06,
+      "loss": 0.4752,
+      "step": 2976
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.563480804634143,
+      "learning_rate": 6.060493851776852e-06,
+      "loss": 0.4193,
+      "step": 2977
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.128795744027954,
+      "learning_rate": 6.058106759174677e-06,
+      "loss": 0.3979,
+      "step": 2978
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.203556351127735,
+      "learning_rate": 6.0557194140662064e-06,
+      "loss": 0.443,
+      "step": 2979
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.689772728688631,
+      "learning_rate": 6.053331817021159e-06,
+      "loss": 0.4923,
+      "step": 2980
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.22409546078687,
+      "learning_rate": 6.05094396860931e-06,
+      "loss": 0.4475,
+      "step": 2981
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.347506414721554,
+      "learning_rate": 6.048555869400493e-06,
+      "loss": 0.4423,
+      "step": 2982
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.315507120057155,
+      "learning_rate": 6.0461675199646065e-06,
+      "loss": 0.4338,
+      "step": 2983
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 10.523601489962932,
+      "learning_rate": 6.043778920871601e-06,
+      "loss": 0.4318,
+      "step": 2984
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.406587377530553,
+      "learning_rate": 6.0413900726914955e-06,
+      "loss": 0.4481,
+      "step": 2985
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.437395983043228,
+      "learning_rate": 6.039000975994362e-06,
+      "loss": 0.4691,
+      "step": 2986
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.753539845800579,
+      "learning_rate": 6.036611631350333e-06,
+      "loss": 0.4003,
+      "step": 2987
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.453974796678845,
+      "learning_rate": 6.034222039329604e-06,
+      "loss": 0.413,
+      "step": 2988
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.52419402041998,
+      "learning_rate": 6.031832200502422e-06,
+      "loss": 0.4512,
+      "step": 2989
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.097790472489987,
+      "learning_rate": 6.029442115439102e-06,
+      "loss": 0.5015,
+      "step": 2990
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 10.216568186887628,
+      "learning_rate": 6.02705178471001e-06,
+      "loss": 0.4749,
+      "step": 2991
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 13.386485668775096,
+      "learning_rate": 6.024661208885576e-06,
+      "loss": 0.4326,
+      "step": 2992
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 16.246062311818296,
+      "learning_rate": 6.022270388536287e-06,
+      "loss": 0.447,
+      "step": 2993
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.8761523724036016,
+      "learning_rate": 6.019879324232684e-06,
+      "loss": 0.4688,
+      "step": 2994
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.590549779223677,
+      "learning_rate": 6.017488016545372e-06,
+      "loss": 0.4403,
+      "step": 2995
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.081755614938339,
+      "learning_rate": 6.015096466045012e-06,
+      "loss": 0.4681,
+      "step": 2996
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.453224836260443,
+      "learning_rate": 6.012704673302323e-06,
+      "loss": 0.4298,
+      "step": 2997
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.630456567721023,
+      "learning_rate": 6.010312638888082e-06,
+      "loss": 0.5031,
+      "step": 2998
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.1757028398868545,
+      "learning_rate": 6.007920363373124e-06,
+      "loss": 0.4027,
+      "step": 2999
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.383145004097821,
+      "learning_rate": 6.005527847328338e-06,
+      "loss": 0.4884,
+      "step": 3000
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.420703968434824,
+      "learning_rate": 6.003135091324677e-06,
+      "loss": 0.426,
+      "step": 3001
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.395177362597513,
+      "learning_rate": 6.0007420959331465e-06,
+      "loss": 0.4361,
+      "step": 3002
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 12.976079561168797,
+      "learning_rate": 5.998348861724808e-06,
+      "loss": 0.5099,
+      "step": 3003
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.645064769081378,
+      "learning_rate": 5.995955389270784e-06,
+      "loss": 0.3893,
+      "step": 3004
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.445898048616861,
+      "learning_rate": 5.993561679142253e-06,
+      "loss": 0.3685,
+      "step": 3005
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 17.122033836583807,
+      "learning_rate": 5.991167731910448e-06,
+      "loss": 0.379,
+      "step": 3006
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.388934948867373,
+      "learning_rate": 5.98877354814666e-06,
+      "loss": 0.4928,
+      "step": 3007
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.602511908593512,
+      "learning_rate": 5.986379128422236e-06,
+      "loss": 0.4612,
+      "step": 3008
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.319162821256345,
+      "learning_rate": 5.9839844733085806e-06,
+      "loss": 0.4506,
+      "step": 3009
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 10.287549852979044,
+      "learning_rate": 5.981589583377155e-06,
+      "loss": 0.3952,
+      "step": 3010
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.361216336811319,
+      "learning_rate": 5.97919445919947e-06,
+      "loss": 0.4621,
+      "step": 3011
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.1208449239265,
+      "learning_rate": 5.976799101347102e-06,
+      "loss": 0.4737,
+      "step": 3012
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.6287622580700205,
+      "learning_rate": 5.9744035103916775e-06,
+      "loss": 0.4348,
+      "step": 3013
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.240008806756864,
+      "learning_rate": 5.972007686904878e-06,
+      "loss": 0.468,
+      "step": 3014
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.601990239797793,
+      "learning_rate": 5.969611631458444e-06,
+      "loss": 0.4018,
+      "step": 3015
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.564357271598327,
+      "learning_rate": 5.967215344624168e-06,
+      "loss": 0.4803,
+      "step": 3016
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 14.972478710890037,
+      "learning_rate": 5.964818826973903e-06,
+      "loss": 0.474,
+      "step": 3017
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.423299847576006,
+      "learning_rate": 5.9624220790795474e-06,
+      "loss": 0.4762,
+      "step": 3018
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.5295518993204995,
+      "learning_rate": 5.960025101513063e-06,
+      "loss": 0.4957,
+      "step": 3019
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.077293027793658,
+      "learning_rate": 5.957627894846465e-06,
+      "loss": 0.4197,
+      "step": 3020
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.4222217448199177,
+      "learning_rate": 5.955230459651822e-06,
+      "loss": 0.4742,
+      "step": 3021
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.972283015825895,
+      "learning_rate": 5.952832796501256e-06,
+      "loss": 0.491,
+      "step": 3022
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 14.03138382708002,
+      "learning_rate": 5.950434905966945e-06,
+      "loss": 0.4811,
+      "step": 3023
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1859448600123303,
+      "learning_rate": 5.948036788621121e-06,
+      "loss": 0.5771,
+      "step": 3024
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 23.586624891973372,
+      "learning_rate": 5.945638445036069e-06,
+      "loss": 0.452,
+      "step": 3025
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.575839725218991,
+      "learning_rate": 5.9432398757841315e-06,
+      "loss": 0.4491,
+      "step": 3026
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2751361252434745,
+      "learning_rate": 5.940841081437702e-06,
+      "loss": 0.5689,
+      "step": 3027
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.626488534885077,
+      "learning_rate": 5.9384420625692266e-06,
+      "loss": 0.4316,
+      "step": 3028
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.1366265256247265,
+      "learning_rate": 5.9360428197512056e-06,
+      "loss": 0.4276,
+      "step": 3029
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.883829318000987,
+      "learning_rate": 5.933643353556195e-06,
+      "loss": 0.4926,
+      "step": 3030
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.4098243410612716,
+      "learning_rate": 5.931243664556803e-06,
+      "loss": 0.4011,
+      "step": 3031
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.896959075466838,
+      "learning_rate": 5.9288437533256915e-06,
+      "loss": 0.4683,
+      "step": 3032
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.886693300823188,
+      "learning_rate": 5.926443620435572e-06,
+      "loss": 0.627,
+      "step": 3033
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.108506864501156,
+      "learning_rate": 5.924043266459214e-06,
+      "loss": 0.4845,
+      "step": 3034
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.8393924733125395,
+      "learning_rate": 5.921642691969436e-06,
+      "loss": 0.5092,
+      "step": 3035
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 29.411254033657546,
+      "learning_rate": 5.919241897539111e-06,
+      "loss": 0.4501,
+      "step": 3036
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.560448452029831,
+      "learning_rate": 5.916840883741164e-06,
+      "loss": 0.4848,
+      "step": 3037
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.642124747659159,
+      "learning_rate": 5.91443965114857e-06,
+      "loss": 0.4386,
+      "step": 3038
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2141450607545425,
+      "learning_rate": 5.912038200334362e-06,
+      "loss": 0.5436,
+      "step": 3039
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 15.123352826719152,
+      "learning_rate": 5.9096365318716195e-06,
+      "loss": 0.428,
+      "step": 3040
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.770042914816767,
+      "learning_rate": 5.9072346463334765e-06,
+      "loss": 0.4803,
+      "step": 3041
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.913405825811076,
+      "learning_rate": 5.904832544293118e-06,
+      "loss": 0.4608,
+      "step": 3042
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.972297140527095,
+      "learning_rate": 5.902430226323781e-06,
+      "loss": 0.4749,
+      "step": 3043
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.366406699894982,
+      "learning_rate": 5.900027692998755e-06,
+      "loss": 0.4251,
+      "step": 3044
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.614672933492882,
+      "learning_rate": 5.8976249448913784e-06,
+      "loss": 0.4149,
+      "step": 3045
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.955260843285969,
+      "learning_rate": 5.895221982575042e-06,
+      "loss": 0.494,
+      "step": 3046
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.730007061925838,
+      "learning_rate": 5.892818806623189e-06,
+      "loss": 0.3844,
+      "step": 3047
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.816388027808729,
+      "learning_rate": 5.890415417609312e-06,
+      "loss": 0.4758,
+      "step": 3048
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.510906777338831,
+      "learning_rate": 5.8880118161069556e-06,
+      "loss": 0.4753,
+      "step": 3049
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.334214220022214,
+      "learning_rate": 5.885608002689714e-06,
+      "loss": 0.4528,
+      "step": 3050
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.108381841817872,
+      "learning_rate": 5.883203977931236e-06,
+      "loss": 0.3675,
+      "step": 3051
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.808269688565124,
+      "learning_rate": 5.880799742405212e-06,
+      "loss": 0.4496,
+      "step": 3052
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.014758805013585,
+      "learning_rate": 5.8783952966853905e-06,
+      "loss": 0.4421,
+      "step": 3053
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 13.325225442197393,
+      "learning_rate": 5.875990641345568e-06,
+      "loss": 0.4498,
+      "step": 3054
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 11.027122161639543,
+      "learning_rate": 5.87358577695959e-06,
+      "loss": 0.4113,
+      "step": 3055
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.003579691132278,
+      "learning_rate": 5.871180704101354e-06,
+      "loss": 0.4459,
+      "step": 3056
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.305201862857531,
+      "learning_rate": 5.868775423344806e-06,
+      "loss": 0.5017,
+      "step": 3057
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.622884414503929,
+      "learning_rate": 5.866369935263939e-06,
+      "loss": 0.4444,
+      "step": 3058
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.524586138913282,
+      "learning_rate": 5.8639642404328e-06,
+      "loss": 0.506,
+      "step": 3059
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.826045078682481,
+      "learning_rate": 5.861558339425481e-06,
+      "loss": 0.4714,
+      "step": 3060
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.703327001043192,
+      "learning_rate": 5.859152232816131e-06,
+      "loss": 0.4246,
+      "step": 3061
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.895368051133767,
+      "learning_rate": 5.856745921178937e-06,
+      "loss": 0.457,
+      "step": 3062
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.933247989064422,
+      "learning_rate": 5.854339405088143e-06,
+      "loss": 0.4776,
+      "step": 3063
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.457887112796842,
+      "learning_rate": 5.851932685118037e-06,
+      "loss": 0.4871,
+      "step": 3064
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.895334093952295,
+      "learning_rate": 5.8495257618429615e-06,
+      "loss": 0.4061,
+      "step": 3065
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.639967382847509,
+      "learning_rate": 5.8471186358373025e-06,
+      "loss": 0.4711,
+      "step": 3066
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.67272662542438,
+      "learning_rate": 5.844711307675493e-06,
+      "loss": 0.524,
+      "step": 3067
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.549650819979633,
+      "learning_rate": 5.842303777932022e-06,
+      "loss": 0.4642,
+      "step": 3068
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 11.534854335683184,
+      "learning_rate": 5.8398960471814196e-06,
+      "loss": 0.4714,
+      "step": 3069
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 11.34067912607852,
+      "learning_rate": 5.8374881159982645e-06,
+      "loss": 0.4238,
+      "step": 3070
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.3661023885875405,
+      "learning_rate": 5.835079984957187e-06,
+      "loss": 0.4671,
+      "step": 3071
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.909384356509735,
+      "learning_rate": 5.832671654632861e-06,
+      "loss": 0.4224,
+      "step": 3072
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.963602018061618,
+      "learning_rate": 5.830263125600011e-06,
+      "loss": 0.4146,
+      "step": 3073
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.645261360097426,
+      "learning_rate": 5.827854398433407e-06,
+      "loss": 0.4449,
+      "step": 3074
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.172568981876063,
+      "learning_rate": 5.8254454737078665e-06,
+      "loss": 0.4587,
+      "step": 3075
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 16.8571769690598,
+      "learning_rate": 5.8230363519982565e-06,
+      "loss": 0.4594,
+      "step": 3076
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.482528579650665,
+      "learning_rate": 5.820627033879488e-06,
+      "loss": 0.4195,
+      "step": 3077
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 11.199140684334441,
+      "learning_rate": 5.818217519926518e-06,
+      "loss": 0.4516,
+      "step": 3078
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.378680061356225,
+      "learning_rate": 5.815807810714356e-06,
+      "loss": 0.5173,
+      "step": 3079
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.083222178669987,
+      "learning_rate": 5.813397906818052e-06,
+      "loss": 0.4285,
+      "step": 3080
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.450696444483291,
+      "learning_rate": 5.810987808812705e-06,
+      "loss": 0.4665,
+      "step": 3081
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.041495777521601,
+      "learning_rate": 5.808577517273459e-06,
+      "loss": 0.4449,
+      "step": 3082
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.750091877057997,
+      "learning_rate": 5.806167032775508e-06,
+      "loss": 0.4641,
+      "step": 3083
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.551963841233672,
+      "learning_rate": 5.803756355894087e-06,
+      "loss": 0.4508,
+      "step": 3084
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 15.36025155679563,
+      "learning_rate": 5.8013454872044816e-06,
+      "loss": 0.4417,
+      "step": 3085
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.392521546269121,
+      "learning_rate": 5.798934427282019e-06,
+      "loss": 0.5054,
+      "step": 3086
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.372660424430705,
+      "learning_rate": 5.796523176702074e-06,
+      "loss": 0.4856,
+      "step": 3087
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.768462398963454,
+      "learning_rate": 5.794111736040064e-06,
+      "loss": 0.5417,
+      "step": 3088
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.717548903726133,
+      "learning_rate": 5.79170010587146e-06,
+      "loss": 0.4305,
+      "step": 3089
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.035969281525046,
+      "learning_rate": 5.7892882867717705e-06,
+      "loss": 0.4787,
+      "step": 3090
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.81313869060724,
+      "learning_rate": 5.786876279316551e-06,
+      "loss": 0.526,
+      "step": 3091
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.19821449774543,
+      "learning_rate": 5.784464084081402e-06,
+      "loss": 0.5263,
+      "step": 3092
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.06840073370822,
+      "learning_rate": 5.782051701641969e-06,
+      "loss": 0.3849,
+      "step": 3093
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.374673519137275,
+      "learning_rate": 5.779639132573942e-06,
+      "loss": 0.4842,
+      "step": 3094
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.43732725047366,
+      "learning_rate": 5.7772263774530565e-06,
+      "loss": 0.4515,
+      "step": 3095
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.467800868152375,
+      "learning_rate": 5.7748134368550935e-06,
+      "loss": 0.4214,
+      "step": 3096
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.01038279518574,
+      "learning_rate": 5.772400311355872e-06,
+      "loss": 0.4347,
+      "step": 3097
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.074339124672884,
+      "learning_rate": 5.7699870015312646e-06,
+      "loss": 0.4683,
+      "step": 3098
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.039062793877982,
+      "learning_rate": 5.767573507957178e-06,
+      "loss": 0.4439,
+      "step": 3099
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.917505861340575,
+      "learning_rate": 5.76515983120957e-06,
+      "loss": 0.4443,
+      "step": 3100
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.843045981194791,
+      "learning_rate": 5.762745971864441e-06,
+      "loss": 0.4158,
+      "step": 3101
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.043465873075565,
+      "learning_rate": 5.760331930497831e-06,
+      "loss": 0.4972,
+      "step": 3102
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.517738193249536,
+      "learning_rate": 5.757917707685829e-06,
+      "loss": 0.5144,
+      "step": 3103
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 13.37729348296037,
+      "learning_rate": 5.7555033040045625e-06,
+      "loss": 0.4687,
+      "step": 3104
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.414470789802739,
+      "learning_rate": 5.753088720030206e-06,
+      "loss": 0.4717,
+      "step": 3105
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.791109326288943,
+      "learning_rate": 5.7506739563389705e-06,
+      "loss": 0.4175,
+      "step": 3106
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.865592504916592,
+      "learning_rate": 5.7482590135071194e-06,
+      "loss": 0.4792,
+      "step": 3107
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.334393345982736,
+      "learning_rate": 5.745843892110952e-06,
+      "loss": 0.4627,
+      "step": 3108
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.000214262766905,
+      "learning_rate": 5.743428592726813e-06,
+      "loss": 0.4532,
+      "step": 3109
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.043919600037595,
+      "learning_rate": 5.741013115931087e-06,
+      "loss": 0.4832,
+      "step": 3110
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.399708475505839,
+      "learning_rate": 5.738597462300205e-06,
+      "loss": 0.577,
+      "step": 3111
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.3845451784220995,
+      "learning_rate": 5.736181632410635e-06,
+      "loss": 0.4082,
+      "step": 3112
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.765795252000268,
+      "learning_rate": 5.733765626838894e-06,
+      "loss": 0.4195,
+      "step": 3113
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.236385503238355,
+      "learning_rate": 5.731349446161533e-06,
+      "loss": 0.47,
+      "step": 3114
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.706118423609485,
+      "learning_rate": 5.728933090955151e-06,
+      "loss": 0.4786,
+      "step": 3115
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.119660145827664,
+      "learning_rate": 5.726516561796387e-06,
+      "loss": 0.4838,
+      "step": 3116
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.537871335432183,
+      "learning_rate": 5.7240998592619165e-06,
+      "loss": 0.5145,
+      "step": 3117
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 21.038663859852814,
+      "learning_rate": 5.7216829839284646e-06,
+      "loss": 0.4717,
+      "step": 3118
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 12.840097205113555,
+      "learning_rate": 5.719265936372793e-06,
+      "loss": 0.4899,
+      "step": 3119
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.2076129392836465,
+      "learning_rate": 5.716848717171706e-06,
+      "loss": 0.5292,
+      "step": 3120
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.39907907781943,
+      "learning_rate": 5.714431326902047e-06,
+      "loss": 0.4762,
+      "step": 3121
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 12.618395041778575,
+      "learning_rate": 5.7120137661407005e-06,
+      "loss": 0.4632,
+      "step": 3122
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.451566632969732,
+      "learning_rate": 5.709596035464592e-06,
+      "loss": 0.4813,
+      "step": 3123
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.416503003580778,
+      "learning_rate": 5.70717813545069e-06,
+      "loss": 0.5047,
+      "step": 3124
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 35.65028924243356,
+      "learning_rate": 5.704760066676003e-06,
+      "loss": 0.4634,
+      "step": 3125
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.474458387051392,
+      "learning_rate": 5.702341829717575e-06,
+      "loss": 0.4493,
+      "step": 3126
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 13.452484435768707,
+      "learning_rate": 5.699923425152495e-06,
+      "loss": 0.5138,
+      "step": 3127
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.983730366911048,
+      "learning_rate": 5.697504853557888e-06,
+      "loss": 0.4153,
+      "step": 3128
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.9315589684090515,
+      "learning_rate": 5.695086115510924e-06,
+      "loss": 0.4539,
+      "step": 3129
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 23.148437575675043,
+      "learning_rate": 5.69266721158881e-06,
+      "loss": 0.5023,
+      "step": 3130
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.789955312540621,
+      "learning_rate": 5.69024814236879e-06,
+      "loss": 0.4387,
+      "step": 3131
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.658923412725847,
+      "learning_rate": 5.687828908428153e-06,
+      "loss": 0.4704,
+      "step": 3132
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.219025318694479,
+      "learning_rate": 5.685409510344223e-06,
+      "loss": 0.4534,
+      "step": 3133
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.564538956915076,
+      "learning_rate": 5.682989948694363e-06,
+      "loss": 0.4089,
+      "step": 3134
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.167202023758512,
+      "learning_rate": 5.680570224055979e-06,
+      "loss": 0.5171,
+      "step": 3135
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.651144185075979,
+      "learning_rate": 5.678150337006512e-06,
+      "loss": 0.4182,
+      "step": 3136
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.456699787212029,
+      "learning_rate": 5.675730288123442e-06,
+      "loss": 0.4599,
+      "step": 3137
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.057024384405909,
+      "learning_rate": 5.673310077984293e-06,
+      "loss": 0.3857,
+      "step": 3138
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2932104838601441,
+      "learning_rate": 5.670889707166618e-06,
+      "loss": 0.5601,
+      "step": 3139
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.059070323209951,
+      "learning_rate": 5.668469176248017e-06,
+      "loss": 0.4636,
+      "step": 3140
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.9412784999811357,
+      "learning_rate": 5.666048485806124e-06,
+      "loss": 0.4786,
+      "step": 3141
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.1548373543652737,
+      "learning_rate": 5.663627636418611e-06,
+      "loss": 0.4784,
+      "step": 3142
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.181993005099054,
+      "learning_rate": 5.661206628663189e-06,
+      "loss": 0.5575,
+      "step": 3143
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.9533606093609426,
+      "learning_rate": 5.658785463117607e-06,
+      "loss": 0.4793,
+      "step": 3144
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.8632699036884715,
+      "learning_rate": 5.656364140359653e-06,
+      "loss": 0.5806,
+      "step": 3145
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6685592634425714,
+      "learning_rate": 5.6539426609671485e-06,
+      "loss": 0.3926,
+      "step": 3146
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.063058798954238,
+      "learning_rate": 5.651521025517954e-06,
+      "loss": 0.5079,
+      "step": 3147
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.164566714346429,
+      "learning_rate": 5.64909923458997e-06,
+      "loss": 0.4597,
+      "step": 3148
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.9349549252412657,
+      "learning_rate": 5.646677288761132e-06,
+      "loss": 0.5664,
+      "step": 3149
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.889440916712922,
+      "learning_rate": 5.644255188609411e-06,
+      "loss": 0.4477,
+      "step": 3150
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.5492922835504976,
+      "learning_rate": 5.6418329347128165e-06,
+      "loss": 0.4897,
+      "step": 3151
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5571806458707342,
+      "learning_rate": 5.639410527649395e-06,
+      "loss": 0.5251,
+      "step": 3152
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.635909692338665,
+      "learning_rate": 5.6369879679972285e-06,
+      "loss": 0.4029,
+      "step": 3153
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.345980058289814,
+      "learning_rate": 5.634565256334435e-06,
+      "loss": 0.5002,
+      "step": 3154
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.6397571867578837,
+      "learning_rate": 5.632142393239174e-06,
+      "loss": 0.471,
+      "step": 3155
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.3704828441445147,
+      "learning_rate": 5.629719379289633e-06,
+      "loss": 0.3915,
+      "step": 3156
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 23.507070834633065,
+      "learning_rate": 5.627296215064039e-06,
+      "loss": 0.4643,
+      "step": 3157
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.957406963770855,
+      "learning_rate": 5.624872901140657e-06,
+      "loss": 0.3978,
+      "step": 3158
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 11.218637599985948,
+      "learning_rate": 5.622449438097785e-06,
+      "loss": 0.4338,
+      "step": 3159
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.282680276785911,
+      "learning_rate": 5.620025826513758e-06,
+      "loss": 0.5381,
+      "step": 3160
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.2077390147945515,
+      "learning_rate": 5.6176020669669464e-06,
+      "loss": 0.4575,
+      "step": 3161
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.664140470623212,
+      "learning_rate": 5.615178160035755e-06,
+      "loss": 0.4852,
+      "step": 3162
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.866206035415342,
+      "learning_rate": 5.612754106298625e-06,
+      "loss": 0.4993,
+      "step": 3163
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.7056785463742274,
+      "learning_rate": 5.61032990633403e-06,
+      "loss": 0.4652,
+      "step": 3164
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.247469840658331,
+      "learning_rate": 5.607905560720481e-06,
+      "loss": 0.5389,
+      "step": 3165
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.8592977303152116,
+      "learning_rate": 5.605481070036523e-06,
+      "loss": 0.4525,
+      "step": 3166
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.0220763494375933,
+      "learning_rate": 5.603056434860739e-06,
+      "loss": 0.3894,
+      "step": 3167
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.249027478167296,
+      "learning_rate": 5.600631655771739e-06,
+      "loss": 0.5713,
+      "step": 3168
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.4263642422843046,
+      "learning_rate": 5.5982067333481735e-06,
+      "loss": 0.4534,
+      "step": 3169
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.606751129313969,
+      "learning_rate": 5.5957816681687246e-06,
+      "loss": 0.4616,
+      "step": 3170
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.884304660456333,
+      "learning_rate": 5.5933564608121095e-06,
+      "loss": 0.5166,
+      "step": 3171
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6970126005004829,
+      "learning_rate": 5.590931111857079e-06,
+      "loss": 0.5386,
+      "step": 3172
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.154039772658649,
+      "learning_rate": 5.5885056218824185e-06,
+      "loss": 0.4867,
+      "step": 3173
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.0363520755483,
+      "learning_rate": 5.586079991466945e-06,
+      "loss": 0.4555,
+      "step": 3174
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.127212776221932,
+      "learning_rate": 5.58365422118951e-06,
+      "loss": 0.4436,
+      "step": 3175
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.647537751906127,
+      "learning_rate": 5.581228311629e-06,
+      "loss": 0.4517,
+      "step": 3176
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.317199225642473,
+      "learning_rate": 5.578802263364333e-06,
+      "loss": 0.4455,
+      "step": 3177
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.6251371340360707,
+      "learning_rate": 5.5763760769744594e-06,
+      "loss": 0.54,
+      "step": 3178
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.3214395582701854,
+      "learning_rate": 5.573949753038364e-06,
+      "loss": 0.4648,
+      "step": 3179
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.297957349116306,
+      "learning_rate": 5.571523292135067e-06,
+      "loss": 0.4989,
+      "step": 3180
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.976242648602837,
+      "learning_rate": 5.569096694843613e-06,
+      "loss": 0.4671,
+      "step": 3181
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.802481147927998,
+      "learning_rate": 5.566669961743088e-06,
+      "loss": 0.4788,
+      "step": 3182
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.06989713775913,
+      "learning_rate": 5.564243093412605e-06,
+      "loss": 0.4598,
+      "step": 3183
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.489522273614001,
+      "learning_rate": 5.561816090431315e-06,
+      "loss": 0.4766,
+      "step": 3184
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.1403164438120665,
+      "learning_rate": 5.559388953378393e-06,
+      "loss": 0.4475,
+      "step": 3185
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.89871877495417,
+      "learning_rate": 5.556961682833052e-06,
+      "loss": 0.4424,
+      "step": 3186
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.2063029776138885,
+      "learning_rate": 5.5545342793745345e-06,
+      "loss": 0.4882,
+      "step": 3187
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.376312124732114,
+      "learning_rate": 5.552106743582115e-06,
+      "loss": 0.53,
+      "step": 3188
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.554156742740164,
+      "learning_rate": 5.549679076035104e-06,
+      "loss": 0.4648,
+      "step": 3189
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.550521146322817,
+      "learning_rate": 5.5472512773128326e-06,
+      "loss": 0.4514,
+      "step": 3190
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.6665951731813227,
+      "learning_rate": 5.544823347994677e-06,
+      "loss": 0.4079,
+      "step": 3191
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.409750586423111,
+      "learning_rate": 5.542395288660032e-06,
+      "loss": 0.4182,
+      "step": 3192
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.780824638495456,
+      "learning_rate": 5.5399670998883305e-06,
+      "loss": 0.426,
+      "step": 3193
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3605253066651224,
+      "learning_rate": 5.537538782259037e-06,
+      "loss": 0.5659,
+      "step": 3194
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.005783623385917,
+      "learning_rate": 5.535110336351642e-06,
+      "loss": 0.4581,
+      "step": 3195
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.0520683998268225,
+      "learning_rate": 5.5326817627456695e-06,
+      "loss": 0.4219,
+      "step": 3196
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.130545303079728,
+      "learning_rate": 5.5302530620206765e-06,
+      "loss": 0.4916,
+      "step": 3197
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2227916493186013,
+      "learning_rate": 5.527824234756243e-06,
+      "loss": 0.551,
+      "step": 3198
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.863211075529677,
+      "learning_rate": 5.5253952815319865e-06,
+      "loss": 0.4407,
+      "step": 3199
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.756551073754293,
+      "learning_rate": 5.522966202927551e-06,
+      "loss": 0.4191,
+      "step": 3200
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.56040429887214,
+      "learning_rate": 5.52053699952261e-06,
+      "loss": 0.445,
+      "step": 3201
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.0316646358828825,
+      "learning_rate": 5.518107671896871e-06,
+      "loss": 0.5828,
+      "step": 3202
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.409697557185427,
+      "learning_rate": 5.515678220630064e-06,
+      "loss": 0.4597,
+      "step": 3203
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.987591006317606,
+      "learning_rate": 5.513248646301957e-06,
+      "loss": 0.4235,
+      "step": 3204
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.000521828200423,
+      "learning_rate": 5.510818949492337e-06,
+      "loss": 0.4584,
+      "step": 3205
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.053772872980277,
+      "learning_rate": 5.508389130781031e-06,
+      "loss": 0.4009,
+      "step": 3206
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.221718470839906,
+      "learning_rate": 5.505959190747889e-06,
+      "loss": 0.4596,
+      "step": 3207
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.0661858479859765,
+      "learning_rate": 5.503529129972792e-06,
+      "loss": 0.4706,
+      "step": 3208
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.998377295945082,
+      "learning_rate": 5.501098949035648e-06,
+      "loss": 0.4746,
+      "step": 3209
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.9352148526871265,
+      "learning_rate": 5.498668648516395e-06,
+      "loss": 0.4232,
+      "step": 3210
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.2745708213396005,
+      "learning_rate": 5.496238228994997e-06,
+      "loss": 0.4596,
+      "step": 3211
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.5177934557602395,
+      "learning_rate": 5.493807691051451e-06,
+      "loss": 0.5331,
+      "step": 3212
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.363429555229731,
+      "learning_rate": 5.4913770352657814e-06,
+      "loss": 0.4343,
+      "step": 3213
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.585552795595535,
+      "learning_rate": 5.488946262218037e-06,
+      "loss": 0.4672,
+      "step": 3214
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.127516266899229,
+      "learning_rate": 5.486515372488294e-06,
+      "loss": 0.3972,
+      "step": 3215
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.732876608185631,
+      "learning_rate": 5.484084366656666e-06,
+      "loss": 0.5211,
+      "step": 3216
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.764826690842418,
+      "learning_rate": 5.481653245303281e-06,
+      "loss": 0.411,
+      "step": 3217
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.601291358917296,
+      "learning_rate": 5.479222009008306e-06,
+      "loss": 0.4699,
+      "step": 3218
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.095162361491254,
+      "learning_rate": 5.476790658351927e-06,
+      "loss": 0.4565,
+      "step": 3219
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.618087612634257,
+      "learning_rate": 5.474359193914363e-06,
+      "loss": 0.4525,
+      "step": 3220
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.389409260587045,
+      "learning_rate": 5.471927616275856e-06,
+      "loss": 0.433,
+      "step": 3221
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.865514228573659,
+      "learning_rate": 5.4694959260166795e-06,
+      "loss": 0.4686,
+      "step": 3222
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.805259582049252,
+      "learning_rate": 5.46706412371713e-06,
+      "loss": 0.4209,
+      "step": 3223
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.579052577552933,
+      "learning_rate": 5.464632209957532e-06,
+      "loss": 0.4378,
+      "step": 3224
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.466297555880539,
+      "learning_rate": 5.462200185318236e-06,
+      "loss": 0.4411,
+      "step": 3225
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.273337063617493,
+      "learning_rate": 5.459768050379622e-06,
+      "loss": 0.4215,
+      "step": 3226
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.0711764515032725,
+      "learning_rate": 5.457335805722092e-06,
+      "loss": 0.5188,
+      "step": 3227
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.757428239371602,
+      "learning_rate": 5.454903451926077e-06,
+      "loss": 0.4578,
+      "step": 3228
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.411607230427831,
+      "learning_rate": 5.452470989572031e-06,
+      "loss": 0.4217,
+      "step": 3229
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.433383694292181,
+      "learning_rate": 5.45003841924044e-06,
+      "loss": 0.4538,
+      "step": 3230
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.925206867271547,
+      "learning_rate": 5.44760574151181e-06,
+      "loss": 0.4763,
+      "step": 3231
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.112728685718616,
+      "learning_rate": 5.4451729569666735e-06,
+      "loss": 0.4749,
+      "step": 3232
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.744233646869358,
+      "learning_rate": 5.4427400661855916e-06,
+      "loss": 0.4184,
+      "step": 3233
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.840975923901068,
+      "learning_rate": 5.440307069749147e-06,
+      "loss": 0.416,
+      "step": 3234
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.694575746940418,
+      "learning_rate": 5.437873968237948e-06,
+      "loss": 0.461,
+      "step": 3235
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.246738936081393,
+      "learning_rate": 5.4354407622326324e-06,
+      "loss": 0.4307,
+      "step": 3236
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.398998426939157,
+      "learning_rate": 5.4330074523138586e-06,
+      "loss": 0.3538,
+      "step": 3237
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.15310211778712,
+      "learning_rate": 5.430574039062312e-06,
+      "loss": 0.5053,
+      "step": 3238
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.0256818057927894,
+      "learning_rate": 5.428140523058698e-06,
+      "loss": 0.4252,
+      "step": 3239
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.379003183699032,
+      "learning_rate": 5.425706904883753e-06,
+      "loss": 0.4589,
+      "step": 3240
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.1879335713915,
+      "learning_rate": 5.4232731851182355e-06,
+      "loss": 0.4863,
+      "step": 3241
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.705671182181169,
+      "learning_rate": 5.420839364342925e-06,
+      "loss": 0.4858,
+      "step": 3242
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.488484698737066,
+      "learning_rate": 5.418405443138631e-06,
+      "loss": 0.3787,
+      "step": 3243
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.579148638539216,
+      "learning_rate": 5.415971422086182e-06,
+      "loss": 0.4243,
+      "step": 3244
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.3155983631965125,
+      "learning_rate": 5.413537301766433e-06,
+      "loss": 0.4531,
+      "step": 3245
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 19.742292378614664,
+      "learning_rate": 5.411103082760259e-06,
+      "loss": 0.4864,
+      "step": 3246
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.792634384931771,
+      "learning_rate": 5.408668765648565e-06,
+      "loss": 0.5039,
+      "step": 3247
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.374334102062656,
+      "learning_rate": 5.406234351012275e-06,
+      "loss": 0.4756,
+      "step": 3248
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.155682520256152,
+      "learning_rate": 5.4037998394323356e-06,
+      "loss": 0.4069,
+      "step": 3249
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.325556722529026,
+      "learning_rate": 5.401365231489718e-06,
+      "loss": 0.4651,
+      "step": 3250
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 12.561341732055103,
+      "learning_rate": 5.398930527765416e-06,
+      "loss": 0.4859,
+      "step": 3251
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 28.47595791948495,
+      "learning_rate": 5.396495728840448e-06,
+      "loss": 0.4391,
+      "step": 3252
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1696913615848277,
+      "learning_rate": 5.3940608352958534e-06,
+      "loss": 0.4805,
+      "step": 3253
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.6818854932250598,
+      "learning_rate": 5.391625847712695e-06,
+      "loss": 0.4323,
+      "step": 3254
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.8075845104537915,
+      "learning_rate": 5.389190766672057e-06,
+      "loss": 0.4603,
+      "step": 3255
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.665670113898421,
+      "learning_rate": 5.386755592755044e-06,
+      "loss": 0.4566,
+      "step": 3256
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.9516943925407477,
+      "learning_rate": 5.384320326542789e-06,
+      "loss": 0.4552,
+      "step": 3257
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.091935876677221,
+      "learning_rate": 5.381884968616442e-06,
+      "loss": 0.4509,
+      "step": 3258
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.63573354157472,
+      "learning_rate": 5.379449519557175e-06,
+      "loss": 0.4148,
+      "step": 3259
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.5816478696515395,
+      "learning_rate": 5.377013979946183e-06,
+      "loss": 0.5229,
+      "step": 3260
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.262931556483842,
+      "learning_rate": 5.374578350364686e-06,
+      "loss": 0.4633,
+      "step": 3261
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.859426386570577,
+      "learning_rate": 5.372142631393916e-06,
+      "loss": 0.4557,
+      "step": 3262
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3176677048114953,
+      "learning_rate": 5.369706823615138e-06,
+      "loss": 0.5563,
+      "step": 3263
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.791295119135724,
+      "learning_rate": 5.367270927609627e-06,
+      "loss": 0.441,
+      "step": 3264
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4060839149248263,
+      "learning_rate": 5.364834943958688e-06,
+      "loss": 0.5696,
+      "step": 3265
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.6067184379781727,
+      "learning_rate": 5.362398873243644e-06,
+      "loss": 0.4033,
+      "step": 3266
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.160678843813218,
+      "learning_rate": 5.359962716045836e-06,
+      "loss": 0.4208,
+      "step": 3267
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.26679962276524,
+      "learning_rate": 5.35752647294663e-06,
+      "loss": 0.4455,
+      "step": 3268
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.798644079797333,
+      "learning_rate": 5.355090144527408e-06,
+      "loss": 0.4663,
+      "step": 3269
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.861094812045507,
+      "learning_rate": 5.352653731369576e-06,
+      "loss": 0.4877,
+      "step": 3270
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.9589362656198666,
+      "learning_rate": 5.350217234054559e-06,
+      "loss": 0.5565,
+      "step": 3271
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.0391228827973937,
+      "learning_rate": 5.3477806531638025e-06,
+      "loss": 0.5136,
+      "step": 3272
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.149651375953562,
+      "learning_rate": 5.345343989278771e-06,
+      "loss": 0.4026,
+      "step": 3273
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.921150674950357,
+      "learning_rate": 5.342907242980948e-06,
+      "loss": 0.4502,
+      "step": 3274
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.765230244320097,
+      "learning_rate": 5.340470414851839e-06,
+      "loss": 0.4853,
+      "step": 3275
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.481198134182762,
+      "learning_rate": 5.338033505472968e-06,
+      "loss": 0.3824,
+      "step": 3276
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.769270456440631,
+      "learning_rate": 5.335596515425879e-06,
+      "loss": 0.4569,
+      "step": 3277
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.019814958256495,
+      "learning_rate": 5.333159445292135e-06,
+      "loss": 0.5026,
+      "step": 3278
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.4954519176525025,
+      "learning_rate": 5.330722295653315e-06,
+      "loss": 0.443,
+      "step": 3279
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.9237392661841275,
+      "learning_rate": 5.328285067091022e-06,
+      "loss": 0.4735,
+      "step": 3280
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.9428035995371125,
+      "learning_rate": 5.3258477601868745e-06,
+      "loss": 0.3763,
+      "step": 3281
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.3629913060942727,
+      "learning_rate": 5.3234103755225115e-06,
+      "loss": 0.5532,
+      "step": 3282
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 17.176192576684052,
+      "learning_rate": 5.32097291367959e-06,
+      "loss": 0.5333,
+      "step": 3283
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.081856887425179,
+      "learning_rate": 5.318535375239784e-06,
+      "loss": 0.5695,
+      "step": 3284
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.0722452641289366,
+      "learning_rate": 5.316097760784786e-06,
+      "loss": 0.4614,
+      "step": 3285
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.365557311997831,
+      "learning_rate": 5.313660070896311e-06,
+      "loss": 0.4329,
+      "step": 3286
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.263080487688134,
+      "learning_rate": 5.3112223061560865e-06,
+      "loss": 0.4472,
+      "step": 3287
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.374427471569894,
+      "learning_rate": 5.30878446714586e-06,
+      "loss": 0.453,
+      "step": 3288
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.508626132987856,
+      "learning_rate": 5.3063465544473965e-06,
+      "loss": 0.3698,
+      "step": 3289
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4049375358944123,
+      "learning_rate": 5.30390856864248e-06,
+      "loss": 0.4769,
+      "step": 3290
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.712612801149028,
+      "learning_rate": 5.301470510312909e-06,
+      "loss": 0.4495,
+      "step": 3291
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.097102779736724,
+      "learning_rate": 5.299032380040504e-06,
+      "loss": 0.4283,
+      "step": 3292
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.75776519062649,
+      "learning_rate": 5.296594178407096e-06,
+      "loss": 0.4164,
+      "step": 3293
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.08760899435055,
+      "learning_rate": 5.294155905994539e-06,
+      "loss": 0.4637,
+      "step": 3294
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.766519620726101,
+      "learning_rate": 5.291717563384702e-06,
+      "loss": 0.4387,
+      "step": 3295
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.0430251977363705,
+      "learning_rate": 5.289279151159467e-06,
+      "loss": 0.4561,
+      "step": 3296
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.380963646955826,
+      "learning_rate": 5.286840669900742e-06,
+      "loss": 0.464,
+      "step": 3297
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.4721056471701806,
+      "learning_rate": 5.284402120190439e-06,
+      "loss": 0.4235,
+      "step": 3298
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.7011138622413067,
+      "learning_rate": 5.281963502610497e-06,
+      "loss": 0.4309,
+      "step": 3299
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.9534246719248185,
+      "learning_rate": 5.2795248177428646e-06,
+      "loss": 0.4663,
+      "step": 3300
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.759452463026602,
+      "learning_rate": 5.2770860661695115e-06,
+      "loss": 0.4636,
+      "step": 3301
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.5500426992803087,
+      "learning_rate": 5.274647248472419e-06,
+      "loss": 0.5299,
+      "step": 3302
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5934580441323676,
+      "learning_rate": 5.272208365233585e-06,
+      "loss": 0.4794,
+      "step": 3303
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.033740734255481,
+      "learning_rate": 5.269769417035026e-06,
+      "loss": 0.5054,
+      "step": 3304
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.6379714799954668,
+      "learning_rate": 5.267330404458769e-06,
+      "loss": 0.3783,
+      "step": 3305
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.5875397094240573,
+      "learning_rate": 5.264891328086862e-06,
+      "loss": 0.4506,
+      "step": 3306
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.385035323002471,
+      "learning_rate": 5.262452188501365e-06,
+      "loss": 0.4453,
+      "step": 3307
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.584980069027644,
+      "learning_rate": 5.260012986284354e-06,
+      "loss": 0.4003,
+      "step": 3308
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.772230307358214,
+      "learning_rate": 5.257573722017917e-06,
+      "loss": 0.4062,
+      "step": 3309
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.049980850076305,
+      "learning_rate": 5.2551343962841615e-06,
+      "loss": 0.4964,
+      "step": 3310
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.5467979341422895,
+      "learning_rate": 5.252695009665206e-06,
+      "loss": 0.4641,
+      "step": 3311
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.859719462182213,
+      "learning_rate": 5.250255562743188e-06,
+      "loss": 0.3876,
+      "step": 3312
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.047033734465715,
+      "learning_rate": 5.247816056100255e-06,
+      "loss": 0.4758,
+      "step": 3313
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.240647672560057,
+      "learning_rate": 5.245376490318566e-06,
+      "loss": 0.4309,
+      "step": 3314
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3777127749030087,
+      "learning_rate": 5.242936865980304e-06,
+      "loss": 0.4516,
+      "step": 3315
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.6658855171079057,
+      "learning_rate": 5.240497183667657e-06,
+      "loss": 0.4586,
+      "step": 3316
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.9630513085929864,
+      "learning_rate": 5.238057443962834e-06,
+      "loss": 0.4859,
+      "step": 3317
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.5330773653344494,
+      "learning_rate": 5.235617647448048e-06,
+      "loss": 0.4297,
+      "step": 3318
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 21.25727874186291,
+      "learning_rate": 5.233177794705536e-06,
+      "loss": 0.4126,
+      "step": 3319
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.769918163210967,
+      "learning_rate": 5.230737886317541e-06,
+      "loss": 0.4863,
+      "step": 3320
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.4990919755753653,
+      "learning_rate": 5.2282979228663236e-06,
+      "loss": 0.4412,
+      "step": 3321
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.892564054651963,
+      "learning_rate": 5.225857904934155e-06,
+      "loss": 0.4076,
+      "step": 3322
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.186331258285111,
+      "learning_rate": 5.2234178331033215e-06,
+      "loss": 0.4439,
+      "step": 3323
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.139353996842082,
+      "learning_rate": 5.22097770795612e-06,
+      "loss": 0.5027,
+      "step": 3324
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.530008368315678,
+      "learning_rate": 5.218537530074862e-06,
+      "loss": 0.4458,
+      "step": 3325
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.962023616624714,
+      "learning_rate": 5.21609730004187e-06,
+      "loss": 0.4584,
+      "step": 3326
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.3513286606755646,
+      "learning_rate": 5.213657018439482e-06,
+      "loss": 0.436,
+      "step": 3327
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.4170556046407374,
+      "learning_rate": 5.211216685850042e-06,
+      "loss": 0.4536,
+      "step": 3328
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 22.31515906934759,
+      "learning_rate": 5.208776302855915e-06,
+      "loss": 0.4659,
+      "step": 3329
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 20.357757128154105,
+      "learning_rate": 5.20633587003947e-06,
+      "loss": 0.4537,
+      "step": 3330
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.8639261122811344,
+      "learning_rate": 5.203895387983093e-06,
+      "loss": 0.4983,
+      "step": 3331
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.0484410183988553,
+      "learning_rate": 5.2014548572691796e-06,
+      "loss": 0.4693,
+      "step": 3332
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.036665199129306,
+      "learning_rate": 5.199014278480135e-06,
+      "loss": 0.3851,
+      "step": 3333
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.1032911895418964,
+      "learning_rate": 5.19657365219838e-06,
+      "loss": 0.4369,
+      "step": 3334
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.020590575955939,
+      "learning_rate": 5.194132979006346e-06,
+      "loss": 0.4796,
+      "step": 3335
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.8025778690321035,
+      "learning_rate": 5.191692259486475e-06,
+      "loss": 0.5066,
+      "step": 3336
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.1991070049848505,
+      "learning_rate": 5.189251494221218e-06,
+      "loss": 0.4471,
+      "step": 3337
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.210517795523312,
+      "learning_rate": 5.186810683793038e-06,
+      "loss": 0.4184,
+      "step": 3338
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.433575860469508,
+      "learning_rate": 5.184369828784409e-06,
+      "loss": 0.4968,
+      "step": 3339
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.297485175785062,
+      "learning_rate": 5.181928929777817e-06,
+      "loss": 0.3983,
+      "step": 3340
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 12.269897665984946,
+      "learning_rate": 5.1794879873557605e-06,
+      "loss": 0.4311,
+      "step": 3341
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.592431741408581,
+      "learning_rate": 5.177047002100739e-06,
+      "loss": 0.5051,
+      "step": 3342
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.899059396345337,
+      "learning_rate": 5.174605974595274e-06,
+      "loss": 0.4218,
+      "step": 3343
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.497862822273135,
+      "learning_rate": 5.172164905421888e-06,
+      "loss": 0.4616,
+      "step": 3344
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.662840702861832,
+      "learning_rate": 5.169723795163119e-06,
+      "loss": 0.4173,
+      "step": 3345
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.9742356443026807,
+      "learning_rate": 5.167282644401512e-06,
+      "loss": 0.3888,
+      "step": 3346
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6869484046200975,
+      "learning_rate": 5.164841453719623e-06,
+      "loss": 0.4747,
+      "step": 3347
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 14.391452105445113,
+      "learning_rate": 5.162400223700017e-06,
+      "loss": 0.473,
+      "step": 3348
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.7068099547723525,
+      "learning_rate": 5.159958954925267e-06,
+      "loss": 0.4785,
+      "step": 3349
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.1050363365132787,
+      "learning_rate": 5.157517647977958e-06,
+      "loss": 0.5281,
+      "step": 3350
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.330353150023166,
+      "learning_rate": 5.155076303440686e-06,
+      "loss": 0.4592,
+      "step": 3351
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.705451851741254,
+      "learning_rate": 5.152634921896046e-06,
+      "loss": 0.4859,
+      "step": 3352
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.0708551705973224,
+      "learning_rate": 5.1501935039266524e-06,
+      "loss": 0.4347,
+      "step": 3353
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.6008086476195897,
+      "learning_rate": 5.147752050115127e-06,
+      "loss": 0.4852,
+      "step": 3354
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.35196073413082,
+      "learning_rate": 5.145310561044091e-06,
+      "loss": 0.4481,
+      "step": 3355
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6751035264760574,
+      "learning_rate": 5.142869037296188e-06,
+      "loss": 0.4832,
+      "step": 3356
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.90243119613261,
+      "learning_rate": 5.1404274794540565e-06,
+      "loss": 0.398,
+      "step": 3357
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.796291533480356,
+      "learning_rate": 5.137985888100352e-06,
+      "loss": 0.4277,
+      "step": 3358
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.791838010266537,
+      "learning_rate": 5.135544263817735e-06,
+      "loss": 0.4246,
+      "step": 3359
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.547302138630425,
+      "learning_rate": 5.133102607188875e-06,
+      "loss": 0.4722,
+      "step": 3360
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.452572250905354,
+      "learning_rate": 5.130660918796446e-06,
+      "loss": 0.4756,
+      "step": 3361
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.576033157228893,
+      "learning_rate": 5.128219199223132e-06,
+      "loss": 0.473,
+      "step": 3362
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.1144084907418708,
+      "learning_rate": 5.125777449051627e-06,
+      "loss": 0.4815,
+      "step": 3363
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.240967014170178,
+      "learning_rate": 5.123335668864627e-06,
+      "loss": 0.5644,
+      "step": 3364
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.890279233273213,
+      "learning_rate": 5.120893859244838e-06,
+      "loss": 0.4148,
+      "step": 3365
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.346384067330006,
+      "learning_rate": 5.1184520207749735e-06,
+      "loss": 0.445,
+      "step": 3366
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.867170831572803,
+      "learning_rate": 5.116010154037753e-06,
+      "loss": 0.3849,
+      "step": 3367
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.337960371320644,
+      "learning_rate": 5.113568259615901e-06,
+      "loss": 0.4872,
+      "step": 3368
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.906170119967276,
+      "learning_rate": 5.111126338092153e-06,
+      "loss": 0.4266,
+      "step": 3369
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.5116594669574335,
+      "learning_rate": 5.1086843900492476e-06,
+      "loss": 0.4774,
+      "step": 3370
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.5027918612460565,
+      "learning_rate": 5.106242416069932e-06,
+      "loss": 0.4677,
+      "step": 3371
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.938188817263368,
+      "learning_rate": 5.103800416736957e-06,
+      "loss": 0.437,
+      "step": 3372
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.0124811809810375,
+      "learning_rate": 5.101358392633078e-06,
+      "loss": 0.4875,
+      "step": 3373
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.506166327783403,
+      "learning_rate": 5.098916344341062e-06,
+      "loss": 0.4918,
+      "step": 3374
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2820857996999695,
+      "learning_rate": 5.096474272443678e-06,
+      "loss": 0.59,
+      "step": 3375
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.173620473422902,
+      "learning_rate": 5.094032177523703e-06,
+      "loss": 0.451,
+      "step": 3376
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.6933139236618375,
+      "learning_rate": 5.091590060163916e-06,
+      "loss": 0.4567,
+      "step": 3377
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.172780238157196,
+      "learning_rate": 5.089147920947101e-06,
+      "loss": 0.4621,
+      "step": 3378
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.907892624629719,
+      "learning_rate": 5.086705760456053e-06,
+      "loss": 0.502,
+      "step": 3379
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.487782358215405,
+      "learning_rate": 5.084263579273566e-06,
+      "loss": 0.532,
+      "step": 3380
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.051137053257438,
+      "learning_rate": 5.081821377982446e-06,
+      "loss": 0.399,
+      "step": 3381
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.859558571389233,
+      "learning_rate": 5.079379157165493e-06,
+      "loss": 0.4624,
+      "step": 3382
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.5097424993680546,
+      "learning_rate": 5.076936917405522e-06,
+      "loss": 0.4135,
+      "step": 3383
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.059271148389168,
+      "learning_rate": 5.074494659285346e-06,
+      "loss": 0.4256,
+      "step": 3384
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.08453940537398,
+      "learning_rate": 5.072052383387787e-06,
+      "loss": 0.4586,
+      "step": 3385
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7070257113029568,
+      "learning_rate": 5.069610090295667e-06,
+      "loss": 0.3907,
+      "step": 3386
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.5096620019079543,
+      "learning_rate": 5.067167780591815e-06,
+      "loss": 0.4839,
+      "step": 3387
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.309144822229189,
+      "learning_rate": 5.064725454859064e-06,
+      "loss": 0.5119,
+      "step": 3388
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.240076456281339,
+      "learning_rate": 5.062283113680249e-06,
+      "loss": 0.3928,
+      "step": 3389
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.413247312515577,
+      "learning_rate": 5.05984075763821e-06,
+      "loss": 0.486,
+      "step": 3390
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.784550943300344,
+      "learning_rate": 5.0573983873157885e-06,
+      "loss": 0.4729,
+      "step": 3391
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.358901926133858,
+      "learning_rate": 5.0549560032958335e-06,
+      "loss": 0.4562,
+      "step": 3392
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7249252171028173,
+      "learning_rate": 5.052513606161192e-06,
+      "loss": 0.4634,
+      "step": 3393
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.6424462382218405,
+      "learning_rate": 5.05007119649472e-06,
+      "loss": 0.4088,
+      "step": 3394
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.8784175764397717,
+      "learning_rate": 5.047628774879273e-06,
+      "loss": 0.4778,
+      "step": 3395
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.221259237159651,
+      "learning_rate": 5.0451863418977074e-06,
+      "loss": 0.4523,
+      "step": 3396
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.4938120375126993,
+      "learning_rate": 5.042743898132886e-06,
+      "loss": 0.4433,
+      "step": 3397
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.752674577893305,
+      "learning_rate": 5.040301444167672e-06,
+      "loss": 0.5248,
+      "step": 3398
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7552701164392324,
+      "learning_rate": 5.037858980584934e-06,
+      "loss": 0.5008,
+      "step": 3399
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.733078586853428,
+      "learning_rate": 5.035416507967541e-06,
+      "loss": 0.5384,
+      "step": 3400
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.167652804430579,
+      "learning_rate": 5.032974026898363e-06,
+      "loss": 0.6161,
+      "step": 3401
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.3638938074112876,
+      "learning_rate": 5.030531537960272e-06,
+      "loss": 0.4794,
+      "step": 3402
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.5496377118684292,
+      "learning_rate": 5.028089041736142e-06,
+      "loss": 0.5091,
+      "step": 3403
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.008969015618689,
+      "learning_rate": 5.025646538808852e-06,
+      "loss": 0.4277,
+      "step": 3404
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.5026658681671448,
+      "learning_rate": 5.023204029761279e-06,
+      "loss": 0.4501,
+      "step": 3405
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.1698791553949257,
+      "learning_rate": 5.020761515176304e-06,
+      "loss": 0.4124,
+      "step": 3406
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.139961506890189,
+      "learning_rate": 5.018318995636807e-06,
+      "loss": 0.422,
+      "step": 3407
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.5418882889432277,
+      "learning_rate": 5.015876471725669e-06,
+      "loss": 0.4943,
+      "step": 3408
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.97062695454358,
+      "learning_rate": 5.013433944025775e-06,
+      "loss": 0.4516,
+      "step": 3409
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.58842963321321,
+      "learning_rate": 5.0109914131200085e-06,
+      "loss": 0.4406,
+      "step": 3410
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.241393409859594,
+      "learning_rate": 5.008548879591255e-06,
+      "loss": 0.527,
+      "step": 3411
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.624821053524552,
+      "learning_rate": 5.006106344022399e-06,
+      "loss": 0.4587,
+      "step": 3412
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.8728721755382445,
+      "learning_rate": 5.003663806996326e-06,
+      "loss": 0.4271,
+      "step": 3413
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.3522566835872807,
+      "learning_rate": 5.001221269095923e-06,
+      "loss": 0.4548,
+      "step": 3414
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.482307662634816,
+      "learning_rate": 4.998778730904077e-06,
+      "loss": 0.477,
+      "step": 3415
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.170252465197291,
+      "learning_rate": 4.9963361930036755e-06,
+      "loss": 0.4839,
+      "step": 3416
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7121381836810534,
+      "learning_rate": 4.993893655977603e-06,
+      "loss": 0.4861,
+      "step": 3417
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2954829777667467,
+      "learning_rate": 4.991451120408747e-06,
+      "loss": 0.5517,
+      "step": 3418
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.957209305995384,
+      "learning_rate": 4.989008586879993e-06,
+      "loss": 0.4631,
+      "step": 3419
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.286742550451276,
+      "learning_rate": 4.986566055974225e-06,
+      "loss": 0.4693,
+      "step": 3420
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.191885250935636,
+      "learning_rate": 4.984123528274332e-06,
+      "loss": 0.4561,
+      "step": 3421
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6663191754587747,
+      "learning_rate": 4.981681004363195e-06,
+      "loss": 0.4831,
+      "step": 3422
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3930109168328557,
+      "learning_rate": 4.979238484823697e-06,
+      "loss": 0.4218,
+      "step": 3423
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.283414588613568,
+      "learning_rate": 4.976795970238723e-06,
+      "loss": 0.4411,
+      "step": 3424
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1801747095590076,
+      "learning_rate": 4.974353461191149e-06,
+      "loss": 0.3749,
+      "step": 3425
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.62708561780288,
+      "learning_rate": 4.97191095826386e-06,
+      "loss": 0.3877,
+      "step": 3426
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3634553758177157,
+      "learning_rate": 4.9694684620397315e-06,
+      "loss": 0.3894,
+      "step": 3427
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.557484764878095,
+      "learning_rate": 4.967025973101638e-06,
+      "loss": 0.4785,
+      "step": 3428
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.365239611851695,
+      "learning_rate": 4.96458349203246e-06,
+      "loss": 0.485,
+      "step": 3429
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.050010411987575,
+      "learning_rate": 4.962141019415066e-06,
+      "loss": 0.4137,
+      "step": 3430
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.3825976652597225,
+      "learning_rate": 4.9596985558323285e-06,
+      "loss": 0.4809,
+      "step": 3431
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3565590900118,
+      "learning_rate": 4.957256101867117e-06,
+      "loss": 0.4765,
+      "step": 3432
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2770063943017262,
+      "learning_rate": 4.954813658102294e-06,
+      "loss": 0.5567,
+      "step": 3433
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.935390010925133,
+      "learning_rate": 4.95237122512073e-06,
+      "loss": 0.4746,
+      "step": 3434
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.3716757944283295,
+      "learning_rate": 4.949928803505281e-06,
+      "loss": 0.4408,
+      "step": 3435
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8882367873943666,
+      "learning_rate": 4.94748639383881e-06,
+      "loss": 0.4884,
+      "step": 3436
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.3926998299187745,
+      "learning_rate": 4.94504399670417e-06,
+      "loss": 0.4838,
+      "step": 3437
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.577280721413775,
+      "learning_rate": 4.942601612684212e-06,
+      "loss": 0.4813,
+      "step": 3438
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.6345878639724227,
+      "learning_rate": 4.940159242361792e-06,
+      "loss": 0.4442,
+      "step": 3439
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.8389289827307134,
+      "learning_rate": 4.937716886319752e-06,
+      "loss": 0.4319,
+      "step": 3440
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.153677529000984,
+      "learning_rate": 4.935274545140938e-06,
+      "loss": 0.5238,
+      "step": 3441
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.4125762357007385,
+      "learning_rate": 4.932832219408187e-06,
+      "loss": 0.448,
+      "step": 3442
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.0153865510135063,
+      "learning_rate": 4.930389909704334e-06,
+      "loss": 0.4648,
+      "step": 3443
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.4607241572899525,
+      "learning_rate": 4.927947616612216e-06,
+      "loss": 0.4211,
+      "step": 3444
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.608041559920161,
+      "learning_rate": 4.925505340714654e-06,
+      "loss": 0.4555,
+      "step": 3445
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.629418764208967,
+      "learning_rate": 4.92306308259448e-06,
+      "loss": 0.395,
+      "step": 3446
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.71131896726482,
+      "learning_rate": 4.920620842834509e-06,
+      "loss": 0.4477,
+      "step": 3447
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.381204380253442,
+      "learning_rate": 4.918178622017557e-06,
+      "loss": 0.5077,
+      "step": 3448
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.0009703084447716,
+      "learning_rate": 4.915736420726435e-06,
+      "loss": 0.4198,
+      "step": 3449
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.1031666761428283,
+      "learning_rate": 4.913294239543948e-06,
+      "loss": 0.484,
+      "step": 3450
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.641347489466055,
+      "learning_rate": 4.9108520790529e-06,
+      "loss": 0.501,
+      "step": 3451
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.4244278328985156,
+      "learning_rate": 4.908409939836088e-06,
+      "loss": 0.4523,
+      "step": 3452
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.546748770485189,
+      "learning_rate": 4.905967822476298e-06,
+      "loss": 0.4223,
+      "step": 3453
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.35592321881806,
+      "learning_rate": 4.903525727556323e-06,
+      "loss": 0.4751,
+      "step": 3454
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.3165390379995285,
+      "learning_rate": 4.901083655658938e-06,
+      "loss": 0.4951,
+      "step": 3455
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.0046155185822014,
+      "learning_rate": 4.8986416073669235e-06,
+      "loss": 0.4488,
+      "step": 3456
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.5446113442596725,
+      "learning_rate": 4.896199583263046e-06,
+      "loss": 0.4214,
+      "step": 3457
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 14.525826216912545,
+      "learning_rate": 4.89375758393007e-06,
+      "loss": 0.5256,
+      "step": 3458
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2675446046835244,
+      "learning_rate": 4.891315609950753e-06,
+      "loss": 0.6005,
+      "step": 3459
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.834310777039866,
+      "learning_rate": 4.888873661907847e-06,
+      "loss": 0.3408,
+      "step": 3460
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.8742551774174125,
+      "learning_rate": 4.8864317403841e-06,
+      "loss": 0.4435,
+      "step": 3461
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.219946019862945,
+      "learning_rate": 4.88398984596225e-06,
+      "loss": 0.4313,
+      "step": 3462
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.6458496770574573,
+      "learning_rate": 4.881547979225027e-06,
+      "loss": 0.434,
+      "step": 3463
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.833995588738102,
+      "learning_rate": 4.879106140755164e-06,
+      "loss": 0.542,
+      "step": 3464
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.028555726158057,
+      "learning_rate": 4.876664331135374e-06,
+      "loss": 0.4186,
+      "step": 3465
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.3753813061826565,
+      "learning_rate": 4.874222550948375e-06,
+      "loss": 0.3891,
+      "step": 3466
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.3051854712314426,
+      "learning_rate": 4.871780800776869e-06,
+      "loss": 0.4434,
+      "step": 3467
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.725181173749494,
+      "learning_rate": 4.8693390812035555e-06,
+      "loss": 0.4406,
+      "step": 3468
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.357822365901962,
+      "learning_rate": 4.866897392811127e-06,
+      "loss": 0.4371,
+      "step": 3469
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6877054952492574,
+      "learning_rate": 4.8644557361822655e-06,
+      "loss": 0.4448,
+      "step": 3470
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.7100074004978705,
+      "learning_rate": 4.862014111899649e-06,
+      "loss": 0.4414,
+      "step": 3471
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.1444855497419875,
+      "learning_rate": 4.859572520545946e-06,
+      "loss": 0.4122,
+      "step": 3472
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.699209978669724,
+      "learning_rate": 4.857130962703814e-06,
+      "loss": 0.4115,
+      "step": 3473
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.167948988219005,
+      "learning_rate": 4.85468943895591e-06,
+      "loss": 0.5134,
+      "step": 3474
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.054866792012111,
+      "learning_rate": 4.852247949884875e-06,
+      "loss": 0.4394,
+      "step": 3475
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.760939822803174,
+      "learning_rate": 4.849806496073348e-06,
+      "loss": 0.3993,
+      "step": 3476
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.7559279808200188,
+      "learning_rate": 4.8473650781039565e-06,
+      "loss": 0.5151,
+      "step": 3477
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.8757257256668805,
+      "learning_rate": 4.844923696559316e-06,
+      "loss": 0.5334,
+      "step": 3478
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.0021347551935254,
+      "learning_rate": 4.8424823520220425e-06,
+      "loss": 0.4195,
+      "step": 3479
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.4106709025296054,
+      "learning_rate": 4.840041045074733e-06,
+      "loss": 0.5534,
+      "step": 3480
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.992383492305769,
+      "learning_rate": 4.8375997762999844e-06,
+      "loss": 0.4749,
+      "step": 3481
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5171999896043895,
+      "learning_rate": 4.835158546280379e-06,
+      "loss": 0.5981,
+      "step": 3482
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.875873049237427,
+      "learning_rate": 4.83271735559849e-06,
+      "loss": 0.5574,
+      "step": 3483
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.236480791126966,
+      "learning_rate": 4.830276204836884e-06,
+      "loss": 0.4705,
+      "step": 3484
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.2456375566861184,
+      "learning_rate": 4.827835094578113e-06,
+      "loss": 0.4734,
+      "step": 3485
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.903724623353245,
+      "learning_rate": 4.825394025404728e-06,
+      "loss": 0.5015,
+      "step": 3486
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.153910391826314,
+      "learning_rate": 4.822952997899263e-06,
+      "loss": 0.4516,
+      "step": 3487
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7390976593294263,
+      "learning_rate": 4.820512012644242e-06,
+      "loss": 0.4871,
+      "step": 3488
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.668034220044751,
+      "learning_rate": 4.8180710702221835e-06,
+      "loss": 0.4366,
+      "step": 3489
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.389067934098476,
+      "learning_rate": 4.8156301712155915e-06,
+      "loss": 0.4785,
+      "step": 3490
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.793780583605442,
+      "learning_rate": 4.813189316206965e-06,
+      "loss": 0.4415,
+      "step": 3491
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.904876710735061,
+      "learning_rate": 4.8107485057787846e-06,
+      "loss": 0.4349,
+      "step": 3492
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.150573629415088,
+      "learning_rate": 4.808307740513526e-06,
+      "loss": 0.4682,
+      "step": 3493
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.073998108765457,
+      "learning_rate": 4.8058670209936554e-06,
+      "loss": 0.4255,
+      "step": 3494
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.980940730043345,
+      "learning_rate": 4.80342634780162e-06,
+      "loss": 0.4555,
+      "step": 3495
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.513436261901973,
+      "learning_rate": 4.800985721519866e-06,
+      "loss": 0.4799,
+      "step": 3496
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.044406585000728,
+      "learning_rate": 4.798545142730824e-06,
+      "loss": 0.4575,
+      "step": 3497
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 23.838611026002955,
+      "learning_rate": 4.7961046120169075e-06,
+      "loss": 0.5083,
+      "step": 3498
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.464785884601695,
+      "learning_rate": 4.7936641299605304e-06,
+      "loss": 0.474,
+      "step": 3499
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 18.869656414745627,
+      "learning_rate": 4.791223697144085e-06,
+      "loss": 0.4329,
+      "step": 3500
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.012713917989065,
+      "learning_rate": 4.788783314149959e-06,
+      "loss": 0.4908,
+      "step": 3501
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.153512107442264,
+      "learning_rate": 4.786342981560521e-06,
+      "loss": 0.4977,
+      "step": 3502
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.136883813680489,
+      "learning_rate": 4.78390269995813e-06,
+      "loss": 0.4626,
+      "step": 3503
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.828426853825977,
+      "learning_rate": 4.7814624699251395e-06,
+      "loss": 0.5404,
+      "step": 3504
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.287874335497635,
+      "learning_rate": 4.779022292043881e-06,
+      "loss": 0.471,
+      "step": 3505
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.841635754919218,
+      "learning_rate": 4.77658216689668e-06,
+      "loss": 0.4742,
+      "step": 3506
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.390665406607505,
+      "learning_rate": 4.774142095065847e-06,
+      "loss": 0.5069,
+      "step": 3507
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.2450440516247485,
+      "learning_rate": 4.771702077133677e-06,
+      "loss": 0.4492,
+      "step": 3508
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.482684831989166,
+      "learning_rate": 4.76926211368246e-06,
+      "loss": 0.4567,
+      "step": 3509
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.9931534306242527,
+      "learning_rate": 4.766822205294465e-06,
+      "loss": 0.5023,
+      "step": 3510
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.0534445668670958,
+      "learning_rate": 4.764382352551954e-06,
+      "loss": 0.453,
+      "step": 3511
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.19558116585608,
+      "learning_rate": 4.761942556037169e-06,
+      "loss": 0.4593,
+      "step": 3512
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.5384557717650473,
+      "learning_rate": 4.759502816332343e-06,
+      "loss": 0.4345,
+      "step": 3513
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.215280363167575,
+      "learning_rate": 4.757063134019697e-06,
+      "loss": 0.5998,
+      "step": 3514
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.44681201668645,
+      "learning_rate": 4.754623509681434e-06,
+      "loss": 0.4065,
+      "step": 3515
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.620920828336688,
+      "learning_rate": 4.752183943899747e-06,
+      "loss": 0.4442,
+      "step": 3516
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.985622264758341,
+      "learning_rate": 4.749744437256813e-06,
+      "loss": 0.4817,
+      "step": 3517
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.0293660522348484,
+      "learning_rate": 4.747304990334794e-06,
+      "loss": 0.4616,
+      "step": 3518
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.6476129275878977,
+      "learning_rate": 4.74486560371584e-06,
+      "loss": 0.4505,
+      "step": 3519
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9244168520030143,
+      "learning_rate": 4.742426277982083e-06,
+      "loss": 0.4416,
+      "step": 3520
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0611069597390574,
+      "learning_rate": 4.739987013715648e-06,
+      "loss": 0.5044,
+      "step": 3521
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.316638499942588,
+      "learning_rate": 4.737547811498636e-06,
+      "loss": 0.4032,
+      "step": 3522
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.8644422214268297,
+      "learning_rate": 4.735108671913138e-06,
+      "loss": 0.4901,
+      "step": 3523
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.390755803162155,
+      "learning_rate": 4.732669595541232e-06,
+      "loss": 0.4043,
+      "step": 3524
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.588097898209697,
+      "learning_rate": 4.730230582964975e-06,
+      "loss": 0.44,
+      "step": 3525
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.852318814887911,
+      "learning_rate": 4.727791634766416e-06,
+      "loss": 0.4509,
+      "step": 3526
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.402263480213148,
+      "learning_rate": 4.725352751527583e-06,
+      "loss": 0.4865,
+      "step": 3527
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.466337588194331,
+      "learning_rate": 4.72291393383049e-06,
+      "loss": 0.4737,
+      "step": 3528
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.140726653614173,
+      "learning_rate": 4.720475182257137e-06,
+      "loss": 0.4926,
+      "step": 3529
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.453446035964061,
+      "learning_rate": 4.718036497389506e-06,
+      "loss": 0.4336,
+      "step": 3530
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.208028076463462,
+      "learning_rate": 4.715597879809562e-06,
+      "loss": 0.4315,
+      "step": 3531
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2282949751536858,
+      "learning_rate": 4.713159330099261e-06,
+      "loss": 0.5179,
+      "step": 3532
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.6535437494079477,
+      "learning_rate": 4.710720848840532e-06,
+      "loss": 0.4213,
+      "step": 3533
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.4255539553471235,
+      "learning_rate": 4.7082824366153e-06,
+      "loss": 0.4841,
+      "step": 3534
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.104363131189795,
+      "learning_rate": 4.7058440940054625e-06,
+      "loss": 0.4456,
+      "step": 3535
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.3264529913533143,
+      "learning_rate": 4.703405821592905e-06,
+      "loss": 0.4624,
+      "step": 3536
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.263774946068872,
+      "learning_rate": 4.7009676199594985e-06,
+      "loss": 0.4387,
+      "step": 3537
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.2520655007153954,
+      "learning_rate": 4.69852948968709e-06,
+      "loss": 0.4522,
+      "step": 3538
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.720797670736403,
+      "learning_rate": 4.6960914313575205e-06,
+      "loss": 0.4249,
+      "step": 3539
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.2293799596105424,
+      "learning_rate": 4.693653445552605e-06,
+      "loss": 0.4194,
+      "step": 3540
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.284295098013576,
+      "learning_rate": 4.691215532854141e-06,
+      "loss": 0.3881,
+      "step": 3541
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.8329973408280282,
+      "learning_rate": 4.688777693843915e-06,
+      "loss": 0.4622,
+      "step": 3542
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5504158785385314,
+      "learning_rate": 4.68633992910369e-06,
+      "loss": 0.4835,
+      "step": 3543
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.991906721170582,
+      "learning_rate": 4.683902239215215e-06,
+      "loss": 0.5291,
+      "step": 3544
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.125463922956898,
+      "learning_rate": 4.681464624760219e-06,
+      "loss": 0.4237,
+      "step": 3545
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 13.947853773046074,
+      "learning_rate": 4.679027086320412e-06,
+      "loss": 0.4548,
+      "step": 3546
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.153086345969848,
+      "learning_rate": 4.676589624477491e-06,
+      "loss": 0.4703,
+      "step": 3547
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.305823470890065,
+      "learning_rate": 4.674152239813126e-06,
+      "loss": 0.4695,
+      "step": 3548
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.899897516440773,
+      "learning_rate": 4.671714932908979e-06,
+      "loss": 0.3633,
+      "step": 3549
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.75380547454389,
+      "learning_rate": 4.669277704346687e-06,
+      "loss": 0.4496,
+      "step": 3550
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.588039909086197,
+      "learning_rate": 4.666840554707866e-06,
+      "loss": 0.4735,
+      "step": 3551
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.048772352324333,
+      "learning_rate": 4.664403484574122e-06,
+      "loss": 0.5019,
+      "step": 3552
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.212233422467621,
+      "learning_rate": 4.661966494527032e-06,
+      "loss": 0.4051,
+      "step": 3553
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.3540716855398864,
+      "learning_rate": 4.659529585148162e-06,
+      "loss": 0.4256,
+      "step": 3554
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7140547875838186,
+      "learning_rate": 4.657092757019055e-06,
+      "loss": 0.4162,
+      "step": 3555
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.665371427384759,
+      "learning_rate": 4.654656010721231e-06,
+      "loss": 0.4506,
+      "step": 3556
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.6544569022344486,
+      "learning_rate": 4.652219346836199e-06,
+      "loss": 0.4794,
+      "step": 3557
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.9876235667536752,
+      "learning_rate": 4.649782765945442e-06,
+      "loss": 0.5115,
+      "step": 3558
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.095931963052549,
+      "learning_rate": 4.6473462686304255e-06,
+      "loss": 0.4694,
+      "step": 3559
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.9133563055814085,
+      "learning_rate": 4.6449098554725945e-06,
+      "loss": 0.4443,
+      "step": 3560
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.130199727341025,
+      "learning_rate": 4.642473527053372e-06,
+      "loss": 0.492,
+      "step": 3561
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.7056252491219555,
+      "learning_rate": 4.640037283954165e-06,
+      "loss": 0.4036,
+      "step": 3562
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.8756960968885674,
+      "learning_rate": 4.6376011267563576e-06,
+      "loss": 0.4642,
+      "step": 3563
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.374028069167655,
+      "learning_rate": 4.635165056041314e-06,
+      "loss": 0.4738,
+      "step": 3564
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.456154783867747,
+      "learning_rate": 4.632729072390376e-06,
+      "loss": 0.4613,
+      "step": 3565
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.646974499538116,
+      "learning_rate": 4.630293176384865e-06,
+      "loss": 0.4082,
+      "step": 3566
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.4369401726325983,
+      "learning_rate": 4.627857368606086e-06,
+      "loss": 0.4994,
+      "step": 3567
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.427581206120353,
+      "learning_rate": 4.625421649635317e-06,
+      "loss": 0.4739,
+      "step": 3568
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.110166471095125,
+      "learning_rate": 4.622986020053818e-06,
+      "loss": 0.4779,
+      "step": 3569
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.7541408821356703,
+      "learning_rate": 4.620550480442828e-06,
+      "loss": 0.3812,
+      "step": 3570
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.822937875550511,
+      "learning_rate": 4.618115031383559e-06,
+      "loss": 0.4627,
+      "step": 3571
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.0805632944686683,
+      "learning_rate": 4.615679673457213e-06,
+      "loss": 0.4992,
+      "step": 3572
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.4835279501143415,
+      "learning_rate": 4.613244407244956e-06,
+      "loss": 0.4542,
+      "step": 3573
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.832094378239397,
+      "learning_rate": 4.610809233327946e-06,
+      "loss": 0.4584,
+      "step": 3574
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2386887127766082,
+      "learning_rate": 4.608374152287307e-06,
+      "loss": 0.547,
+      "step": 3575
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.806360868548605,
+      "learning_rate": 4.605939164704147e-06,
+      "loss": 0.4643,
+      "step": 3576
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.997720747786057,
+      "learning_rate": 4.603504271159553e-06,
+      "loss": 0.4532,
+      "step": 3577
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.2957150311531964,
+      "learning_rate": 4.601069472234584e-06,
+      "loss": 0.4619,
+      "step": 3578
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.649172866262382,
+      "learning_rate": 4.598634768510284e-06,
+      "loss": 0.4372,
+      "step": 3579
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3373489891189552,
+      "learning_rate": 4.596200160567667e-06,
+      "loss": 0.5865,
+      "step": 3580
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.024451249224027,
+      "learning_rate": 4.593765648987727e-06,
+      "loss": 0.391,
+      "step": 3581
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.3412811684671477,
+      "learning_rate": 4.591331234351436e-06,
+      "loss": 0.456,
+      "step": 3582
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.7626667516222256,
+      "learning_rate": 4.588896917239741e-06,
+      "loss": 0.3939,
+      "step": 3583
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.4364873392986612,
+      "learning_rate": 4.5864626982335685e-06,
+      "loss": 0.4549,
+      "step": 3584
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.1055795871425915,
+      "learning_rate": 4.584028577913819e-06,
+      "loss": 0.4614,
+      "step": 3585
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1312700284967037,
+      "learning_rate": 4.58159455686137e-06,
+      "loss": 0.4969,
+      "step": 3586
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.522486460425019,
+      "learning_rate": 4.579160635657076e-06,
+      "loss": 0.505,
+      "step": 3587
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 17.064758804235545,
+      "learning_rate": 4.576726814881765e-06,
+      "loss": 0.4407,
+      "step": 3588
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3221708935160172,
+      "learning_rate": 4.574293095116248e-06,
+      "loss": 0.5853,
+      "step": 3589
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.492522266300199,
+      "learning_rate": 4.5718594769413045e-06,
+      "loss": 0.4729,
+      "step": 3590
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.20958806292807,
+      "learning_rate": 4.56942596093769e-06,
+      "loss": 0.4889,
+      "step": 3591
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4163001256298524,
+      "learning_rate": 4.566992547686142e-06,
+      "loss": 0.4267,
+      "step": 3592
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.388778521911487,
+      "learning_rate": 4.564559237767368e-06,
+      "loss": 0.4623,
+      "step": 3593
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.944930482158423,
+      "learning_rate": 4.562126031762053e-06,
+      "loss": 0.4673,
+      "step": 3594
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.410210767571355,
+      "learning_rate": 4.5596929302508566e-06,
+      "loss": 0.5191,
+      "step": 3595
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.328477899141395,
+      "learning_rate": 4.557259933814409e-06,
+      "loss": 0.3894,
+      "step": 3596
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.1610464189136445,
+      "learning_rate": 4.554827043033328e-06,
+      "loss": 0.4368,
+      "step": 3597
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2219190631737484,
+      "learning_rate": 4.552394258488192e-06,
+      "loss": 0.5273,
+      "step": 3598
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.53375942750944,
+      "learning_rate": 4.549961580759561e-06,
+      "loss": 0.4417,
+      "step": 3599
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2157975588817882,
+      "learning_rate": 4.547529010427971e-06,
+      "loss": 0.5519,
+      "step": 3600
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.087248834398162,
+      "learning_rate": 4.5450965480739245e-06,
+      "loss": 0.4797,
+      "step": 3601
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3323876344802354,
+      "learning_rate": 4.54266419427791e-06,
+      "loss": 0.5814,
+      "step": 3602
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.4293596943243267,
+      "learning_rate": 4.5402319496203796e-06,
+      "loss": 0.4677,
+      "step": 3603
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.3907581821304524,
+      "learning_rate": 4.537799814681765e-06,
+      "loss": 0.4952,
+      "step": 3604
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 27.195689806612886,
+      "learning_rate": 4.535367790042471e-06,
+      "loss": 0.4623,
+      "step": 3605
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.5527307043628285,
+      "learning_rate": 4.532935876282871e-06,
+      "loss": 0.4476,
+      "step": 3606
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3124562313135855,
+      "learning_rate": 4.530504073983322e-06,
+      "loss": 0.6166,
+      "step": 3607
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6438443670303857,
+      "learning_rate": 4.528072383724144e-06,
+      "loss": 0.4315,
+      "step": 3608
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0726134382175725,
+      "learning_rate": 4.525640806085638e-06,
+      "loss": 0.5123,
+      "step": 3609
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6558606207230753,
+      "learning_rate": 4.523209341648075e-06,
+      "loss": 0.3796,
+      "step": 3610
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.890507619464277,
+      "learning_rate": 4.520777990991696e-06,
+      "loss": 0.5054,
+      "step": 3611
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.4124917559136767,
+      "learning_rate": 4.518346754696721e-06,
+      "loss": 0.4584,
+      "step": 3612
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.879553414196396,
+      "learning_rate": 4.515915633343335e-06,
+      "loss": 0.393,
+      "step": 3613
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2004468339785968,
+      "learning_rate": 4.5134846275117065e-06,
+      "loss": 0.5976,
+      "step": 3614
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8007658352487934,
+      "learning_rate": 4.5110537377819665e-06,
+      "loss": 0.4512,
+      "step": 3615
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.9969787394457663,
+      "learning_rate": 4.50862296473422e-06,
+      "loss": 0.4689,
+      "step": 3616
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.9921251545900076,
+      "learning_rate": 4.5061923089485505e-06,
+      "loss": 0.4661,
+      "step": 3617
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.041834346511043,
+      "learning_rate": 4.503761771005004e-06,
+      "loss": 0.418,
+      "step": 3618
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.693491790877917,
+      "learning_rate": 4.501331351483607e-06,
+      "loss": 0.3758,
+      "step": 3619
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.070403620321998,
+      "learning_rate": 4.498901050964353e-06,
+      "loss": 0.4571,
+      "step": 3620
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3269682285479902,
+      "learning_rate": 4.496470870027209e-06,
+      "loss": 0.5596,
+      "step": 3621
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.919025404219239,
+      "learning_rate": 4.494040809252112e-06,
+      "loss": 0.4293,
+      "step": 3622
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.059555164955259,
+      "learning_rate": 4.491610869218969e-06,
+      "loss": 0.4819,
+      "step": 3623
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.3111506942735636,
+      "learning_rate": 4.489181050507664e-06,
+      "loss": 0.4553,
+      "step": 3624
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.048236455299816,
+      "learning_rate": 4.486751353698047e-06,
+      "loss": 0.3941,
+      "step": 3625
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.829447067519785,
+      "learning_rate": 4.4843217793699365e-06,
+      "loss": 0.5054,
+      "step": 3626
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.883516326739647,
+      "learning_rate": 4.481892328103131e-06,
+      "loss": 0.3779,
+      "step": 3627
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.844861984948812,
+      "learning_rate": 4.47946300047739e-06,
+      "loss": 0.4639,
+      "step": 3628
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.212595259719701,
+      "learning_rate": 4.477033797072451e-06,
+      "loss": 0.3977,
+      "step": 3629
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.839727917115151,
+      "learning_rate": 4.474604718468016e-06,
+      "loss": 0.4317,
+      "step": 3630
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.450431118581267,
+      "learning_rate": 4.472175765243758e-06,
+      "loss": 0.464,
+      "step": 3631
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.077424784882286,
+      "learning_rate": 4.469746937979325e-06,
+      "loss": 0.4681,
+      "step": 3632
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.1701420017884745,
+      "learning_rate": 4.4673182372543305e-06,
+      "loss": 0.4798,
+      "step": 3633
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.6906097518137866,
+      "learning_rate": 4.464889663648359e-06,
+      "loss": 0.4627,
+      "step": 3634
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.046488246071865,
+      "learning_rate": 4.462461217740965e-06,
+      "loss": 0.4384,
+      "step": 3635
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.058887673260045,
+      "learning_rate": 4.46003290011167e-06,
+      "loss": 0.4799,
+      "step": 3636
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.614535493773436,
+      "learning_rate": 4.45760471133997e-06,
+      "loss": 0.4443,
+      "step": 3637
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.582446944665605,
+      "learning_rate": 4.455176652005325e-06,
+      "loss": 0.4284,
+      "step": 3638
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.483392356907345,
+      "learning_rate": 4.452748722687168e-06,
+      "loss": 0.3582,
+      "step": 3639
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.290613832679188,
+      "learning_rate": 4.4503209239648995e-06,
+      "loss": 0.4562,
+      "step": 3640
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.754892929241506,
+      "learning_rate": 4.447893256417885e-06,
+      "loss": 0.473,
+      "step": 3641
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.935003859721646,
+      "learning_rate": 4.445465720625467e-06,
+      "loss": 0.4454,
+      "step": 3642
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7707915972103616,
+      "learning_rate": 4.443038317166951e-06,
+      "loss": 0.4685,
+      "step": 3643
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.505217139008351,
+      "learning_rate": 4.440611046621608e-06,
+      "loss": 0.5013,
+      "step": 3644
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.1161057654199564,
+      "learning_rate": 4.438183909568687e-06,
+      "loss": 0.4878,
+      "step": 3645
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.870952892683315,
+      "learning_rate": 4.435756906587395e-06,
+      "loss": 0.4676,
+      "step": 3646
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.4913246651433125,
+      "learning_rate": 4.433330038256914e-06,
+      "loss": 0.4599,
+      "step": 3647
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.72607318514973,
+      "learning_rate": 4.430903305156389e-06,
+      "loss": 0.4127,
+      "step": 3648
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.005872535105403,
+      "learning_rate": 4.4284767078649345e-06,
+      "loss": 0.4233,
+      "step": 3649
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.3677120183549696,
+      "learning_rate": 4.4260502469616365e-06,
+      "loss": 0.3943,
+      "step": 3650
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.5162997676667143,
+      "learning_rate": 4.423623923025541e-06,
+      "loss": 0.4082,
+      "step": 3651
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.178985082861699,
+      "learning_rate": 4.421197736635669e-06,
+      "loss": 0.412,
+      "step": 3652
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 14.477750216243061,
+      "learning_rate": 4.418771688371003e-06,
+      "loss": 0.5137,
+      "step": 3653
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.233838056149836,
+      "learning_rate": 4.4163457788104905e-06,
+      "loss": 0.5534,
+      "step": 3654
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.922280358920338,
+      "learning_rate": 4.413920008533057e-06,
+      "loss": 0.4505,
+      "step": 3655
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.8067760976042635,
+      "learning_rate": 4.411494378117583e-06,
+      "loss": 0.489,
+      "step": 3656
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8003119663702747,
+      "learning_rate": 4.409068888142923e-06,
+      "loss": 0.4217,
+      "step": 3657
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.866391730923889,
+      "learning_rate": 4.406643539187893e-06,
+      "loss": 0.3683,
+      "step": 3658
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.1739151640845185,
+      "learning_rate": 4.404218331831277e-06,
+      "loss": 0.49,
+      "step": 3659
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1844224348561958,
+      "learning_rate": 4.401793266651829e-06,
+      "loss": 0.5543,
+      "step": 3660
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8016563765552527,
+      "learning_rate": 4.399368344228262e-06,
+      "loss": 0.4667,
+      "step": 3661
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.425879411029741,
+      "learning_rate": 4.396943565139262e-06,
+      "loss": 0.4275,
+      "step": 3662
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.083956558982024,
+      "learning_rate": 4.3945189299634775e-06,
+      "loss": 0.473,
+      "step": 3663
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.586302234383876,
+      "learning_rate": 4.3920944392795206e-06,
+      "loss": 0.4508,
+      "step": 3664
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.0047529969942084,
+      "learning_rate": 4.389670093665972e-06,
+      "loss": 0.432,
+      "step": 3665
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.3782134489602105,
+      "learning_rate": 4.387245893701376e-06,
+      "loss": 0.503,
+      "step": 3666
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.184324921621955,
+      "learning_rate": 4.3848218399642454e-06,
+      "loss": 0.4479,
+      "step": 3667
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.079358196536918,
+      "learning_rate": 4.382397933033054e-06,
+      "loss": 0.4686,
+      "step": 3668
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.192506832573775,
+      "learning_rate": 4.3799741734862424e-06,
+      "loss": 0.4596,
+      "step": 3669
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.1529257194915834,
+      "learning_rate": 4.377550561902217e-06,
+      "loss": 0.5427,
+      "step": 3670
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.2221924724097963,
+      "learning_rate": 4.375127098859343e-06,
+      "loss": 0.4757,
+      "step": 3671
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.345586262988627,
+      "learning_rate": 4.372703784935962e-06,
+      "loss": 0.3744,
+      "step": 3672
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.80529402242681,
+      "learning_rate": 4.37028062071037e-06,
+      "loss": 0.4115,
+      "step": 3673
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.1655249544247033,
+      "learning_rate": 4.367857606760828e-06,
+      "loss": 0.4583,
+      "step": 3674
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.344405582861784,
+      "learning_rate": 4.3654347436655655e-06,
+      "loss": 0.4396,
+      "step": 3675
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.201644393759148,
+      "learning_rate": 4.363012032002772e-06,
+      "loss": 0.4743,
+      "step": 3676
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.286190166757507,
+      "learning_rate": 4.360589472350607e-06,
+      "loss": 0.4432,
+      "step": 3677
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.328887891816517,
+      "learning_rate": 4.358167065287186e-06,
+      "loss": 0.4665,
+      "step": 3678
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.09657255677466,
+      "learning_rate": 4.3557448113905894e-06,
+      "loss": 0.4167,
+      "step": 3679
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.356866312441596,
+      "learning_rate": 4.3533227112388694e-06,
+      "loss": 0.5428,
+      "step": 3680
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.559404359490227,
+      "learning_rate": 4.35090076541003e-06,
+      "loss": 0.4833,
+      "step": 3681
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.9347378630454957,
+      "learning_rate": 4.348478974482047e-06,
+      "loss": 0.4567,
+      "step": 3682
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.0611278272803855,
+      "learning_rate": 4.346057339032854e-06,
+      "loss": 0.4312,
+      "step": 3683
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.6304892091145695,
+      "learning_rate": 4.3436358596403476e-06,
+      "loss": 0.5078,
+      "step": 3684
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.780309014631191,
+      "learning_rate": 4.3412145368823934e-06,
+      "loss": 0.4932,
+      "step": 3685
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.486716513964562,
+      "learning_rate": 4.338793371336812e-06,
+      "loss": 0.4246,
+      "step": 3686
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.3010646883317905,
+      "learning_rate": 4.336372363581391e-06,
+      "loss": 0.4222,
+      "step": 3687
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4046670620123414,
+      "learning_rate": 4.3339515141938795e-06,
+      "loss": 0.5594,
+      "step": 3688
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.510420916282598,
+      "learning_rate": 4.331530823751984e-06,
+      "loss": 0.425,
+      "step": 3689
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.270868980735197,
+      "learning_rate": 4.329110292833383e-06,
+      "loss": 0.5166,
+      "step": 3690
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.2253763390533923,
+      "learning_rate": 4.326689922015709e-06,
+      "loss": 0.4495,
+      "step": 3691
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5563899126097223,
+      "learning_rate": 4.324269711876559e-06,
+      "loss": 0.5175,
+      "step": 3692
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.2519443604917377,
+      "learning_rate": 4.321849662993491e-06,
+      "loss": 0.4851,
+      "step": 3693
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.962968745083644,
+      "learning_rate": 4.319429775944021e-06,
+      "loss": 0.3481,
+      "step": 3694
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8017135791269934,
+      "learning_rate": 4.317010051305639e-06,
+      "loss": 0.4358,
+      "step": 3695
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.813601825032154,
+      "learning_rate": 4.314590489655778e-06,
+      "loss": 0.4951,
+      "step": 3696
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.906634301008147,
+      "learning_rate": 4.3121710915718474e-06,
+      "loss": 0.3819,
+      "step": 3697
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 11.113220627126022,
+      "learning_rate": 4.309751857631211e-06,
+      "loss": 0.4649,
+      "step": 3698
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.9539136184721326,
+      "learning_rate": 4.3073327884111915e-06,
+      "loss": 0.448,
+      "step": 3699
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.8079398349205285,
+      "learning_rate": 4.304913884489078e-06,
+      "loss": 0.5251,
+      "step": 3700
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.809916124103518,
+      "learning_rate": 4.302495146442112e-06,
+      "loss": 0.4297,
+      "step": 3701
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.7283355053578866,
+      "learning_rate": 4.300076574847507e-06,
+      "loss": 0.4976,
+      "step": 3702
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.8822465858160426,
+      "learning_rate": 4.297658170282427e-06,
+      "loss": 0.4829,
+      "step": 3703
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.3182148679641714,
+      "learning_rate": 4.2952399333239985e-06,
+      "loss": 0.438,
+      "step": 3704
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.9196694711338833,
+      "learning_rate": 4.2928218645493105e-06,
+      "loss": 0.4983,
+      "step": 3705
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.448278511388973,
+      "learning_rate": 4.290403964535408e-06,
+      "loss": 0.5016,
+      "step": 3706
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.080095031982521,
+      "learning_rate": 4.287986233859301e-06,
+      "loss": 0.4241,
+      "step": 3707
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.054066037293088,
+      "learning_rate": 4.285568673097955e-06,
+      "loss": 0.4011,
+      "step": 3708
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.095032667614165,
+      "learning_rate": 4.283151282828295e-06,
+      "loss": 0.5172,
+      "step": 3709
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.4706088869054237,
+      "learning_rate": 4.280734063627208e-06,
+      "loss": 0.3854,
+      "step": 3710
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.460625969113091,
+      "learning_rate": 4.2783170160715346e-06,
+      "loss": 0.4206,
+      "step": 3711
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.928018662756119,
+      "learning_rate": 4.275900140738084e-06,
+      "loss": 0.4158,
+      "step": 3712
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.8811917846404276,
+      "learning_rate": 4.273483438203616e-06,
+      "loss": 0.4266,
+      "step": 3713
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.100769638870864,
+      "learning_rate": 4.27106690904485e-06,
+      "loss": 0.4353,
+      "step": 3714
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.457538457042008,
+      "learning_rate": 4.268650553838468e-06,
+      "loss": 0.5121,
+      "step": 3715
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.917510206711214,
+      "learning_rate": 4.2662343731611065e-06,
+      "loss": 0.4845,
+      "step": 3716
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.224251413796545,
+      "learning_rate": 4.2638183675893655e-06,
+      "loss": 0.4281,
+      "step": 3717
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.75078392617113,
+      "learning_rate": 4.261402537699797e-06,
+      "loss": 0.4972,
+      "step": 3718
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.276124561260907,
+      "learning_rate": 4.258986884068913e-06,
+      "loss": 0.4608,
+      "step": 3719
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.023113067498944,
+      "learning_rate": 4.256571407273189e-06,
+      "loss": 0.5003,
+      "step": 3720
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.7639214701540804,
+      "learning_rate": 4.254156107889049e-06,
+      "loss": 0.4742,
+      "step": 3721
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.8077414587185223,
+      "learning_rate": 4.251740986492882e-06,
+      "loss": 0.4822,
+      "step": 3722
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.41816501884602,
+      "learning_rate": 4.249326043661032e-06,
+      "loss": 0.499,
+      "step": 3723
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.467159527683841,
+      "learning_rate": 4.246911279969797e-06,
+      "loss": 0.4309,
+      "step": 3724
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.644577999164011,
+      "learning_rate": 4.244496695995438e-06,
+      "loss": 0.5314,
+      "step": 3725
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.479901681264197,
+      "learning_rate": 4.242082292314172e-06,
+      "loss": 0.4539,
+      "step": 3726
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.084483768623457,
+      "learning_rate": 4.23966806950217e-06,
+      "loss": 0.446,
+      "step": 3727
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.3706273889282485,
+      "learning_rate": 4.2372540281355614e-06,
+      "loss": 0.4237,
+      "step": 3728
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.074422438592611,
+      "learning_rate": 4.2348401687904305e-06,
+      "loss": 0.4096,
+      "step": 3729
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.561739359933813,
+      "learning_rate": 4.232426492042824e-06,
+      "loss": 0.4529,
+      "step": 3730
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.9143420312390447,
+      "learning_rate": 4.230012998468738e-06,
+      "loss": 0.4447,
+      "step": 3731
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.7307871671537125,
+      "learning_rate": 4.227599688644129e-06,
+      "loss": 0.4299,
+      "step": 3732
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.3124886170339365,
+      "learning_rate": 4.22518656314491e-06,
+      "loss": 0.3715,
+      "step": 3733
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.042866287318259,
+      "learning_rate": 4.2227736225469434e-06,
+      "loss": 0.4493,
+      "step": 3734
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.625237889074587,
+      "learning_rate": 4.220360867426059e-06,
+      "loss": 0.531,
+      "step": 3735
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.035402015602318,
+      "learning_rate": 4.217948298358031e-06,
+      "loss": 0.4937,
+      "step": 3736
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.100650195189811,
+      "learning_rate": 4.215535915918599e-06,
+      "loss": 0.4963,
+      "step": 3737
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.002068530550297,
+      "learning_rate": 4.2131237206834506e-06,
+      "loss": 0.4514,
+      "step": 3738
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.8512357998123425,
+      "learning_rate": 4.21071171322823e-06,
+      "loss": 0.3758,
+      "step": 3739
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.9706212876280618,
+      "learning_rate": 4.2082998941285405e-06,
+      "loss": 0.455,
+      "step": 3740
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.925160611815083,
+      "learning_rate": 4.205888263959935e-06,
+      "loss": 0.4452,
+      "step": 3741
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.460936886482379,
+      "learning_rate": 4.2034768232979286e-06,
+      "loss": 0.423,
+      "step": 3742
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.693464740018308,
+      "learning_rate": 4.201065572717984e-06,
+      "loss": 0.3508,
+      "step": 3743
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.2977808098530543,
+      "learning_rate": 4.19865451279552e-06,
+      "loss": 0.416,
+      "step": 3744
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.949640480079124,
+      "learning_rate": 4.196243644105915e-06,
+      "loss": 0.4342,
+      "step": 3745
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.115900430881847,
+      "learning_rate": 4.193832967224493e-06,
+      "loss": 0.5142,
+      "step": 3746
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6471569624513993,
+      "learning_rate": 4.191422482726542e-06,
+      "loss": 0.4338,
+      "step": 3747
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.4609638393392315,
+      "learning_rate": 4.1890121911872975e-06,
+      "loss": 0.4472,
+      "step": 3748
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.512341882136168,
+      "learning_rate": 4.18660209318195e-06,
+      "loss": 0.4591,
+      "step": 3749
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.30351029899018,
+      "learning_rate": 4.184192189285647e-06,
+      "loss": 0.413,
+      "step": 3750
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.422912584568043,
+      "learning_rate": 4.181782480073482e-06,
+      "loss": 0.4381,
+      "step": 3751
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.255355689820745,
+      "learning_rate": 4.179372966120514e-06,
+      "loss": 0.4618,
+      "step": 3752
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.7357101275621725,
+      "learning_rate": 4.176963648001746e-06,
+      "loss": 0.4832,
+      "step": 3753
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.910943432851932,
+      "learning_rate": 4.1745545262921334e-06,
+      "loss": 0.4549,
+      "step": 3754
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.466433753037585,
+      "learning_rate": 4.172145601566594e-06,
+      "loss": 0.4296,
+      "step": 3755
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.1377886032315687,
+      "learning_rate": 4.169736874399991e-06,
+      "loss": 0.4878,
+      "step": 3756
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.435219589425918,
+      "learning_rate": 4.16732834536714e-06,
+      "loss": 0.4481,
+      "step": 3757
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.205351584041125,
+      "learning_rate": 4.164920015042816e-06,
+      "loss": 0.419,
+      "step": 3758
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.7306223133844325,
+      "learning_rate": 4.162511884001736e-06,
+      "loss": 0.4446,
+      "step": 3759
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3161335503631997,
+      "learning_rate": 4.160103952818582e-06,
+      "loss": 0.4518,
+      "step": 3760
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.998832938898976,
+      "learning_rate": 4.1576962220679796e-06,
+      "loss": 0.5471,
+      "step": 3761
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.144881721483335,
+      "learning_rate": 4.1552886923245075e-06,
+      "loss": 0.4373,
+      "step": 3762
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 11.807120142824246,
+      "learning_rate": 4.152881364162701e-06,
+      "loss": 0.4269,
+      "step": 3763
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.196372374604001,
+      "learning_rate": 4.150474238157039e-06,
+      "loss": 0.4393,
+      "step": 3764
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.075919598805602,
+      "learning_rate": 4.148067314881964e-06,
+      "loss": 0.4154,
+      "step": 3765
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.802354117209687,
+      "learning_rate": 4.14566059491186e-06,
+      "loss": 0.4626,
+      "step": 3766
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.5181329680198337,
+      "learning_rate": 4.143254078821065e-06,
+      "loss": 0.392,
+      "step": 3767
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.618790600780879,
+      "learning_rate": 4.140847767183872e-06,
+      "loss": 0.4955,
+      "step": 3768
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.253947578786492,
+      "learning_rate": 4.138441660574518e-06,
+      "loss": 0.5044,
+      "step": 3769
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.004823198878446,
+      "learning_rate": 4.136035759567202e-06,
+      "loss": 0.4908,
+      "step": 3770
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.920748463492911,
+      "learning_rate": 4.1336300647360635e-06,
+      "loss": 0.4411,
+      "step": 3771
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.042136379001271,
+      "learning_rate": 4.131224576655195e-06,
+      "loss": 0.4198,
+      "step": 3772
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.691848950251611,
+      "learning_rate": 4.1288192958986464e-06,
+      "loss": 0.419,
+      "step": 3773
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.5028573514819126,
+      "learning_rate": 4.12641422304041e-06,
+      "loss": 0.3994,
+      "step": 3774
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.4277369782924043,
+      "learning_rate": 4.124009358654433e-06,
+      "loss": 0.3919,
+      "step": 3775
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8204239078453863,
+      "learning_rate": 4.121604703314612e-06,
+      "loss": 0.5225,
+      "step": 3776
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.892782365351085,
+      "learning_rate": 4.119200257594789e-06,
+      "loss": 0.4549,
+      "step": 3777
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.5938108099082258,
+      "learning_rate": 4.116796022068767e-06,
+      "loss": 0.5055,
+      "step": 3778
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.307273099932867,
+      "learning_rate": 4.1143919973102856e-06,
+      "loss": 0.4042,
+      "step": 3779
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.118192768755623,
+      "learning_rate": 4.111988183893046e-06,
+      "loss": 0.4717,
+      "step": 3780
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6645901761303006,
+      "learning_rate": 4.10958458239069e-06,
+      "loss": 0.4744,
+      "step": 3781
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.6274248632772688,
+      "learning_rate": 4.107181193376812e-06,
+      "loss": 0.4243,
+      "step": 3782
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.944124399338622,
+      "learning_rate": 4.104778017424961e-06,
+      "loss": 0.3698,
+      "step": 3783
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.716516990297118,
+      "learning_rate": 4.102375055108624e-06,
+      "loss": 0.4686,
+      "step": 3784
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.919077800281757,
+      "learning_rate": 4.099972307001248e-06,
+      "loss": 0.4612,
+      "step": 3785
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.56710923583181,
+      "learning_rate": 4.097569773676221e-06,
+      "loss": 0.4254,
+      "step": 3786
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.2556409000249342,
+      "learning_rate": 4.0951674557068824e-06,
+      "loss": 0.4384,
+      "step": 3787
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.0533035373997905,
+      "learning_rate": 4.092765353666524e-06,
+      "loss": 0.4396,
+      "step": 3788
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.400837710023985,
+      "learning_rate": 4.09036346812838e-06,
+      "loss": 0.3972,
+      "step": 3789
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.530494071447724,
+      "learning_rate": 4.087961799665639e-06,
+      "loss": 0.4477,
+      "step": 3790
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.191102419148896,
+      "learning_rate": 4.085560348851431e-06,
+      "loss": 0.5014,
+      "step": 3791
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.522334971938472,
+      "learning_rate": 4.083159116258838e-06,
+      "loss": 0.4815,
+      "step": 3792
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6877758985199396,
+      "learning_rate": 4.080758102460891e-06,
+      "loss": 0.4611,
+      "step": 3793
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.37841878880807,
+      "learning_rate": 4.078357308030565e-06,
+      "loss": 0.4809,
+      "step": 3794
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.064971415371922,
+      "learning_rate": 4.075956733540787e-06,
+      "loss": 0.4598,
+      "step": 3795
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.699938349323046,
+      "learning_rate": 4.073556379564429e-06,
+      "loss": 0.4672,
+      "step": 3796
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.044675633798052,
+      "learning_rate": 4.07115624667431e-06,
+      "loss": 0.4538,
+      "step": 3797
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.8069917767795776,
+      "learning_rate": 4.0687563354431986e-06,
+      "loss": 0.4338,
+      "step": 3798
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.894886480147642,
+      "learning_rate": 4.066356646443806e-06,
+      "loss": 0.4628,
+      "step": 3799
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.9780328737289157,
+      "learning_rate": 4.063957180248796e-06,
+      "loss": 0.4856,
+      "step": 3800
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.070128793278779,
+      "learning_rate": 4.061557937430777e-06,
+      "loss": 0.5372,
+      "step": 3801
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.388383629547504,
+      "learning_rate": 4.0591589185623e-06,
+      "loss": 0.5216,
+      "step": 3802
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.093885847147076,
+      "learning_rate": 4.05676012421587e-06,
+      "loss": 0.4383,
+      "step": 3803
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.833820887662677,
+      "learning_rate": 4.05436155496393e-06,
+      "loss": 0.4603,
+      "step": 3804
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.2681093337286358,
+      "learning_rate": 4.05196321137888e-06,
+      "loss": 0.4555,
+      "step": 3805
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.210998529514009,
+      "learning_rate": 4.049565094033057e-06,
+      "loss": 0.4544,
+      "step": 3806
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.290992340049877,
+      "learning_rate": 4.0471672034987446e-06,
+      "loss": 0.4026,
+      "step": 3807
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.415609489644776,
+      "learning_rate": 4.044769540348179e-06,
+      "loss": 0.4378,
+      "step": 3808
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.760407663262526,
+      "learning_rate": 4.042372105153535e-06,
+      "loss": 0.4597,
+      "step": 3809
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.844833424338622,
+      "learning_rate": 4.039974898486938e-06,
+      "loss": 0.4378,
+      "step": 3810
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.609878177192518,
+      "learning_rate": 4.037577920920455e-06,
+      "loss": 0.4371,
+      "step": 3811
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.063551689916742,
+      "learning_rate": 4.035181173026099e-06,
+      "loss": 0.4554,
+      "step": 3812
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.095990008225165,
+      "learning_rate": 4.032784655375832e-06,
+      "loss": 0.4336,
+      "step": 3813
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.589989040144424,
+      "learning_rate": 4.030388368541557e-06,
+      "loss": 0.3961,
+      "step": 3814
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.968738490214082,
+      "learning_rate": 4.027992313095124e-06,
+      "loss": 0.4436,
+      "step": 3815
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1874661783453915,
+      "learning_rate": 4.025596489608326e-06,
+      "loss": 0.5373,
+      "step": 3816
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.586353906262547,
+      "learning_rate": 4.0232008986528985e-06,
+      "loss": 0.4975,
+      "step": 3817
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.4645308538466,
+      "learning_rate": 4.0208055408005316e-06,
+      "loss": 0.4693,
+      "step": 3818
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.6754331202646413,
+      "learning_rate": 4.018410416622848e-06,
+      "loss": 0.4635,
+      "step": 3819
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.29712552204372,
+      "learning_rate": 4.016015526691421e-06,
+      "loss": 0.4155,
+      "step": 3820
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.2708343902202897,
+      "learning_rate": 4.013620871577765e-06,
+      "loss": 0.4009,
+      "step": 3821
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.6925404903174863,
+      "learning_rate": 4.011226451853341e-06,
+      "loss": 0.4177,
+      "step": 3822
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.33518671856447,
+      "learning_rate": 4.0088322680895545e-06,
+      "loss": 0.443,
+      "step": 3823
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.42611930784702,
+      "learning_rate": 4.0064383208577475e-06,
+      "loss": 0.4592,
+      "step": 3824
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.064120276902632,
+      "learning_rate": 4.0040446107292165e-06,
+      "loss": 0.4299,
+      "step": 3825
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.52094255201696,
+      "learning_rate": 4.001651138275194e-06,
+      "loss": 0.3608,
+      "step": 3826
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.7214776779876364,
+      "learning_rate": 3.999257904066855e-06,
+      "loss": 0.3773,
+      "step": 3827
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.891657857256591,
+      "learning_rate": 3.996864908675325e-06,
+      "loss": 0.5148,
+      "step": 3828
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.0822886419417035,
+      "learning_rate": 3.994472152671662e-06,
+      "loss": 0.4498,
+      "step": 3829
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.5296903332642313,
+      "learning_rate": 3.992079636626878e-06,
+      "loss": 0.4994,
+      "step": 3830
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.315017312131069,
+      "learning_rate": 3.989687361111919e-06,
+      "loss": 0.5009,
+      "step": 3831
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.009823787285301,
+      "learning_rate": 3.9872953266976774e-06,
+      "loss": 0.5108,
+      "step": 3832
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4213315323951665,
+      "learning_rate": 3.9849035339549904e-06,
+      "loss": 0.4418,
+      "step": 3833
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.9316200526211755,
+      "learning_rate": 3.982511983454629e-06,
+      "loss": 0.4361,
+      "step": 3834
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.226637870072047,
+      "learning_rate": 3.980120675767319e-06,
+      "loss": 0.3685,
+      "step": 3835
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.098428565683648,
+      "learning_rate": 3.977729611463717e-06,
+      "loss": 0.446,
+      "step": 3836
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.206832987115077,
+      "learning_rate": 3.9753387911144245e-06,
+      "loss": 0.4422,
+      "step": 3837
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.682644375850624,
+      "learning_rate": 3.972948215289992e-06,
+      "loss": 0.5019,
+      "step": 3838
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.819916263655875,
+      "learning_rate": 3.970557884560899e-06,
+      "loss": 0.5358,
+      "step": 3839
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.8543266558686375,
+      "learning_rate": 3.968167799497579e-06,
+      "loss": 0.4445,
+      "step": 3840
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.8166683220953255,
+      "learning_rate": 3.9657779606703985e-06,
+      "loss": 0.4785,
+      "step": 3841
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.7611725251145485,
+      "learning_rate": 3.963388368649667e-06,
+      "loss": 0.4448,
+      "step": 3842
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.332272397440074,
+      "learning_rate": 3.960999024005639e-06,
+      "loss": 0.4881,
+      "step": 3843
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.99467430813858,
+      "learning_rate": 3.958609927308505e-06,
+      "loss": 0.4048,
+      "step": 3844
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.977211615799983,
+      "learning_rate": 3.9562210791283994e-06,
+      "loss": 0.4232,
+      "step": 3845
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.7038739925427553,
+      "learning_rate": 3.953832480035396e-06,
+      "loss": 0.4704,
+      "step": 3846
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.607936277232311,
+      "learning_rate": 3.951444130599507e-06,
+      "loss": 0.388,
+      "step": 3847
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.8754332494342587,
+      "learning_rate": 3.949056031390691e-06,
+      "loss": 0.4067,
+      "step": 3848
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.44140849446421,
+      "learning_rate": 3.946668182978841e-06,
+      "loss": 0.4482,
+      "step": 3849
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.732917779973226,
+      "learning_rate": 3.944280585933794e-06,
+      "loss": 0.4214,
+      "step": 3850
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.41208164454842,
+      "learning_rate": 3.9418932408253256e-06,
+      "loss": 0.4502,
+      "step": 3851
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.653226027970726,
+      "learning_rate": 3.9395061482231474e-06,
+      "loss": 0.4408,
+      "step": 3852
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.959581666976122,
+      "learning_rate": 3.93711930869692e-06,
+      "loss": 0.4843,
+      "step": 3853
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.42181232619998,
+      "learning_rate": 3.934732722816234e-06,
+      "loss": 0.4754,
+      "step": 3854
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.213007837532748,
+      "learning_rate": 3.932346391150627e-06,
+      "loss": 0.4678,
+      "step": 3855
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.588453871587705,
+      "learning_rate": 3.929960314269569e-06,
+      "loss": 0.4584,
+      "step": 3856
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.250239785513397,
+      "learning_rate": 3.927574492742473e-06,
+      "loss": 0.4564,
+      "step": 3857
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.812969321742089,
+      "learning_rate": 3.925188927138694e-06,
+      "loss": 0.4453,
+      "step": 3858
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.40932519580338,
+      "learning_rate": 3.922803618027519e-06,
+      "loss": 0.4488,
+      "step": 3859
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7899936831543393,
+      "learning_rate": 3.920418565978182e-06,
+      "loss": 0.4117,
+      "step": 3860
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.8866379970939215,
+      "learning_rate": 3.918033771559846e-06,
+      "loss": 0.4553,
+      "step": 3861
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 12.864507745898237,
+      "learning_rate": 3.91564923534162e-06,
+      "loss": 0.4683,
+      "step": 3862
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.8245837484009333,
+      "learning_rate": 3.913264957892549e-06,
+      "loss": 0.4757,
+      "step": 3863
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 13.744252201125041,
+      "learning_rate": 3.9108809397816154e-06,
+      "loss": 0.4266,
+      "step": 3864
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.630564614036765,
+      "learning_rate": 3.908497181577743e-06,
+      "loss": 0.5326,
+      "step": 3865
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.082092079702226,
+      "learning_rate": 3.90611368384979e-06,
+      "loss": 0.3823,
+      "step": 3866
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.536295524373419,
+      "learning_rate": 3.9037304471665525e-06,
+      "loss": 0.4667,
+      "step": 3867
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.903252018663176,
+      "learning_rate": 3.901347472096767e-06,
+      "loss": 0.37,
+      "step": 3868
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.900086867129049,
+      "learning_rate": 3.898964759209105e-06,
+      "loss": 0.427,
+      "step": 3869
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.8482026680460026,
+      "learning_rate": 3.896582309072175e-06,
+      "loss": 0.4362,
+      "step": 3870
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.593407478794892,
+      "learning_rate": 3.894200122254528e-06,
+      "loss": 0.4486,
+      "step": 3871
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.428488499865246,
+      "learning_rate": 3.891818199324644e-06,
+      "loss": 0.4899,
+      "step": 3872
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.395171462252501,
+      "learning_rate": 3.889436540850948e-06,
+      "loss": 0.4333,
+      "step": 3873
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.1575532850275114,
+      "learning_rate": 3.887055147401797e-06,
+      "loss": 0.4321,
+      "step": 3874
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.605102161149329,
+      "learning_rate": 3.884674019545484e-06,
+      "loss": 0.4422,
+      "step": 3875
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.7286572460990937,
+      "learning_rate": 3.882293157850245e-06,
+      "loss": 0.4364,
+      "step": 3876
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.659070305451452,
+      "learning_rate": 3.879912562884244e-06,
+      "loss": 0.446,
+      "step": 3877
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.679684806880758,
+      "learning_rate": 3.877532235215589e-06,
+      "loss": 0.3922,
+      "step": 3878
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8155908540711883,
+      "learning_rate": 3.875152175412319e-06,
+      "loss": 0.4801,
+      "step": 3879
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.120111832659367,
+      "learning_rate": 3.872772384042408e-06,
+      "loss": 0.4474,
+      "step": 3880
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1073271249620888,
+      "learning_rate": 3.870392861673774e-06,
+      "loss": 0.4901,
+      "step": 3881
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7946935367066894,
+      "learning_rate": 3.86801360887426e-06,
+      "loss": 0.416,
+      "step": 3882
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.472519954368805,
+      "learning_rate": 3.8656346262116565e-06,
+      "loss": 0.5222,
+      "step": 3883
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 11.888011337517117,
+      "learning_rate": 3.8632559142536795e-06,
+      "loss": 0.4729,
+      "step": 3884
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.823315182217647,
+      "learning_rate": 3.8608774735679835e-06,
+      "loss": 0.4506,
+      "step": 3885
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.8900950332405815,
+      "learning_rate": 3.858499304722161e-06,
+      "loss": 0.4581,
+      "step": 3886
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.040592637631079,
+      "learning_rate": 3.856121408283735e-06,
+      "loss": 0.4304,
+      "step": 3887
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.432579107233802,
+      "learning_rate": 3.853743784820169e-06,
+      "loss": 0.4949,
+      "step": 3888
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.3878343634866326,
+      "learning_rate": 3.851366434898857e-06,
+      "loss": 0.3906,
+      "step": 3889
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.532765857145093,
+      "learning_rate": 3.8489893590871275e-06,
+      "loss": 0.439,
+      "step": 3890
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2204393937906575,
+      "learning_rate": 3.846612557952248e-06,
+      "loss": 0.5284,
+      "step": 3891
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.42920845323881,
+      "learning_rate": 3.844236032061415e-06,
+      "loss": 0.461,
+      "step": 3892
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2324256957638833,
+      "learning_rate": 3.841859781981765e-06,
+      "loss": 0.5673,
+      "step": 3893
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.554493836809629,
+      "learning_rate": 3.839483808280364e-06,
+      "loss": 0.4703,
+      "step": 3894
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.325686909821763,
+      "learning_rate": 3.837108111524213e-06,
+      "loss": 0.3804,
+      "step": 3895
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.64493375771464,
+      "learning_rate": 3.834732692280248e-06,
+      "loss": 0.4347,
+      "step": 3896
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.31644062137482,
+      "learning_rate": 3.832357551115337e-06,
+      "loss": 0.4471,
+      "step": 3897
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.903948417974065,
+      "learning_rate": 3.829982688596287e-06,
+      "loss": 0.447,
+      "step": 3898
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.570794961224365,
+      "learning_rate": 3.827608105289832e-06,
+      "loss": 0.401,
+      "step": 3899
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.224796085871885,
+      "learning_rate": 3.82523380176264e-06,
+      "loss": 0.4128,
+      "step": 3900
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.996710572629503,
+      "learning_rate": 3.8228597785813186e-06,
+      "loss": 0.4227,
+      "step": 3901
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.017857467008076,
+      "learning_rate": 3.8204860363124e-06,
+      "loss": 0.4734,
+      "step": 3902
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.0340383974600345,
+      "learning_rate": 3.818112575522355e-06,
+      "loss": 0.3988,
+      "step": 3903
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.063504383674613,
+      "learning_rate": 3.815739396777587e-06,
+      "loss": 0.4533,
+      "step": 3904
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 11.546315356747908,
+      "learning_rate": 3.813366500644426e-06,
+      "loss": 0.3868,
+      "step": 3905
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.4027131804055455,
+      "learning_rate": 3.810993887689145e-06,
+      "loss": 0.4346,
+      "step": 3906
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.6435181230051685,
+      "learning_rate": 3.8086215584779395e-06,
+      "loss": 0.4405,
+      "step": 3907
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.2694779293584,
+      "learning_rate": 3.806249513576945e-06,
+      "loss": 0.4424,
+      "step": 3908
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.5783562057844054,
+      "learning_rate": 3.803877753552223e-06,
+      "loss": 0.4632,
+      "step": 3909
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3843775139673493,
+      "learning_rate": 3.801506278969769e-06,
+      "loss": 0.5327,
+      "step": 3910
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.8359043089334834,
+      "learning_rate": 3.7991350903955138e-06,
+      "loss": 0.5134,
+      "step": 3911
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.2393513672950855,
+      "learning_rate": 3.7967641883953144e-06,
+      "loss": 0.5469,
+      "step": 3912
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.765021958793624,
+      "learning_rate": 3.7943935735349647e-06,
+      "loss": 0.3794,
+      "step": 3913
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.164424688518929,
+      "learning_rate": 3.7920232463801866e-06,
+      "loss": 0.5157,
+      "step": 3914
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.8468160581979056,
+      "learning_rate": 3.789653207496631e-06,
+      "loss": 0.4694,
+      "step": 3915
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 12.44937779827826,
+      "learning_rate": 3.7872834574498894e-06,
+      "loss": 0.4222,
+      "step": 3916
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.165782554198708,
+      "learning_rate": 3.7849139968054728e-06,
+      "loss": 0.407,
+      "step": 3917
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.681458829866862,
+      "learning_rate": 3.782544826128833e-06,
+      "loss": 0.5007,
+      "step": 3918
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.7520348590162493,
+      "learning_rate": 3.780175945985346e-06,
+      "loss": 0.5048,
+      "step": 3919
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.6402701549190333,
+      "learning_rate": 3.77780735694032e-06,
+      "loss": 0.3853,
+      "step": 3920
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.9457507661649633,
+      "learning_rate": 3.7754390595589962e-06,
+      "loss": 0.4238,
+      "step": 3921
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.7949855926788367,
+      "learning_rate": 3.7730710544065413e-06,
+      "loss": 0.5078,
+      "step": 3922
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2221028319353835,
+      "learning_rate": 3.77070334204806e-06,
+      "loss": 0.5746,
+      "step": 3923
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.586762219397688,
+      "learning_rate": 3.7683359230485794e-06,
+      "loss": 0.378,
+      "step": 3924
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.3795651934166724,
+      "learning_rate": 3.76596879797306e-06,
+      "loss": 0.366,
+      "step": 3925
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.355652889916826,
+      "learning_rate": 3.7636019673863926e-06,
+      "loss": 0.4032,
+      "step": 3926
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.007924862278221,
+      "learning_rate": 3.761235431853395e-06,
+      "loss": 0.4933,
+      "step": 3927
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.336640886607247,
+      "learning_rate": 3.75886919193882e-06,
+      "loss": 0.4459,
+      "step": 3928
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.8036769742169887,
+      "learning_rate": 3.756503248207344e-06,
+      "loss": 0.4854,
+      "step": 3929
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.179796315765809,
+      "learning_rate": 3.7541376012235753e-06,
+      "loss": 0.5288,
+      "step": 3930
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.235615642290056,
+      "learning_rate": 3.751772251552052e-06,
+      "loss": 0.5509,
+      "step": 3931
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.93599352931071,
+      "learning_rate": 3.7494071997572386e-06,
+      "loss": 0.4108,
+      "step": 3932
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.151237142602862,
+      "learning_rate": 3.747042446403533e-06,
+      "loss": 0.4672,
+      "step": 3933
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.906394689752917,
+      "learning_rate": 3.744677992055259e-06,
+      "loss": 0.4637,
+      "step": 3934
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.416352854911965,
+      "learning_rate": 3.7423138372766654e-06,
+      "loss": 0.5008,
+      "step": 3935
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.75958120818296,
+      "learning_rate": 3.7399499826319397e-06,
+      "loss": 0.4597,
+      "step": 3936
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.8896375972382327,
+      "learning_rate": 3.737586428685186e-06,
+      "loss": 0.4212,
+      "step": 3937
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.8731968571840034,
+      "learning_rate": 3.735223176000446e-06,
+      "loss": 0.4093,
+      "step": 3938
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.862717593071882,
+      "learning_rate": 3.7328602251416835e-06,
+      "loss": 0.441,
+      "step": 3939
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.9742213521317544,
+      "learning_rate": 3.7304975766727913e-06,
+      "loss": 0.4165,
+      "step": 3940
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 12.01955298356408,
+      "learning_rate": 3.7281352311575947e-06,
+      "loss": 0.4028,
+      "step": 3941
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9754500304432985,
+      "learning_rate": 3.7257731891598402e-06,
+      "loss": 0.4317,
+      "step": 3942
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1195693288507145,
+      "learning_rate": 3.7234114512432064e-06,
+      "loss": 0.5294,
+      "step": 3943
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 11.569465769621749,
+      "learning_rate": 3.721050017971298e-06,
+      "loss": 0.4284,
+      "step": 3944
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.0491076696329795,
+      "learning_rate": 3.7186888899076427e-06,
+      "loss": 0.4152,
+      "step": 3945
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.705513185628955,
+      "learning_rate": 3.7163280676157055e-06,
+      "loss": 0.3982,
+      "step": 3946
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.52724256847801,
+      "learning_rate": 3.713967551658868e-06,
+      "loss": 0.4308,
+      "step": 3947
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.2774077231419465,
+      "learning_rate": 3.7116073426004463e-06,
+      "loss": 0.475,
+      "step": 3948
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.038592634621072,
+      "learning_rate": 3.709247441003678e-06,
+      "loss": 0.4843,
+      "step": 3949
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.223512489428968,
+      "learning_rate": 3.7068878474317273e-06,
+      "loss": 0.4629,
+      "step": 3950
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.919465917965498,
+      "learning_rate": 3.704528562447691e-06,
+      "loss": 0.4421,
+      "step": 3951
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.4969707531047955,
+      "learning_rate": 3.7021695866145846e-06,
+      "loss": 0.48,
+      "step": 3952
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1035028946611682,
+      "learning_rate": 3.699810920495358e-06,
+      "loss": 0.4849,
+      "step": 3953
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.0430254104808,
+      "learning_rate": 3.697452564652879e-06,
+      "loss": 0.4463,
+      "step": 3954
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.691028679130676,
+      "learning_rate": 3.6950945196499454e-06,
+      "loss": 0.4156,
+      "step": 3955
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.017633651799711,
+      "learning_rate": 3.6927367860492813e-06,
+      "loss": 0.4553,
+      "step": 3956
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.5971774035056305,
+      "learning_rate": 3.6903793644135333e-06,
+      "loss": 0.4096,
+      "step": 3957
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.376116494017634,
+      "learning_rate": 3.688022255305279e-06,
+      "loss": 0.463,
+      "step": 3958
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.851463620455795,
+      "learning_rate": 3.685665459287018e-06,
+      "loss": 0.411,
+      "step": 3959
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.385108898146386,
+      "learning_rate": 3.683308976921173e-06,
+      "loss": 0.4514,
+      "step": 3960
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3557357766478106,
+      "learning_rate": 3.6809528087700956e-06,
+      "loss": 0.4226,
+      "step": 3961
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.370694491986353,
+      "learning_rate": 3.6785969553960598e-06,
+      "loss": 0.4686,
+      "step": 3962
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.441701649246758,
+      "learning_rate": 3.6762414173612683e-06,
+      "loss": 0.3916,
+      "step": 3963
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.068758494708756,
+      "learning_rate": 3.6738861952278456e-06,
+      "loss": 0.493,
+      "step": 3964
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.9212312352478937,
+      "learning_rate": 3.6715312895578388e-06,
+      "loss": 0.4666,
+      "step": 3965
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.644703293804203,
+      "learning_rate": 3.669176700913224e-06,
+      "loss": 0.4602,
+      "step": 3966
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.067670912612484,
+      "learning_rate": 3.6668224298558973e-06,
+      "loss": 0.3769,
+      "step": 3967
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.074620071399767,
+      "learning_rate": 3.6644684769476857e-06,
+      "loss": 0.4589,
+      "step": 3968
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.785863552947082,
+      "learning_rate": 3.6621148427503326e-06,
+      "loss": 0.4504,
+      "step": 3969
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.594619990847038,
+      "learning_rate": 3.6597615278255073e-06,
+      "loss": 0.4988,
+      "step": 3970
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.993625777801499,
+      "learning_rate": 3.6574085327348074e-06,
+      "loss": 0.4303,
+      "step": 3971
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.0432593125304725,
+      "learning_rate": 3.655055858039749e-06,
+      "loss": 0.5025,
+      "step": 3972
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.284346288589723,
+      "learning_rate": 3.6527035043017756e-06,
+      "loss": 0.4497,
+      "step": 3973
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.731856515998285,
+      "learning_rate": 3.6503514720822507e-06,
+      "loss": 0.4395,
+      "step": 3974
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.2937291167291463,
+      "learning_rate": 3.6479997619424605e-06,
+      "loss": 0.5068,
+      "step": 3975
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1884893327885775,
+      "learning_rate": 3.645648374443621e-06,
+      "loss": 0.5428,
+      "step": 3976
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.199422002868311,
+      "learning_rate": 3.6432973101468638e-06,
+      "loss": 0.4483,
+      "step": 3977
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2396888660069516,
+      "learning_rate": 3.640946569613247e-06,
+      "loss": 0.4094,
+      "step": 3978
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.6751073395562734,
+      "learning_rate": 3.6385961534037506e-06,
+      "loss": 0.4748,
+      "step": 3979
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.458976524699675,
+      "learning_rate": 3.6362460620792755e-06,
+      "loss": 0.443,
+      "step": 3980
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.34339235804711,
+      "learning_rate": 3.6338962962006504e-06,
+      "loss": 0.4501,
+      "step": 3981
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.791445218336222,
+      "learning_rate": 3.6315468563286204e-06,
+      "loss": 0.5081,
+      "step": 3982
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.170620791205664,
+      "learning_rate": 3.6291977430238544e-06,
+      "loss": 0.4685,
+      "step": 3983
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.685833756703963,
+      "learning_rate": 3.626848956846947e-06,
+      "loss": 0.4652,
+      "step": 3984
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7807688779558157,
+      "learning_rate": 3.624500498358407e-06,
+      "loss": 0.4613,
+      "step": 3985
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.444401965086056,
+      "learning_rate": 3.6221523681186755e-06,
+      "loss": 0.4849,
+      "step": 3986
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.108764509632119,
+      "learning_rate": 3.619804566688108e-06,
+      "loss": 0.4622,
+      "step": 3987
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.707945294768867,
+      "learning_rate": 3.6174570946269794e-06,
+      "loss": 0.4689,
+      "step": 3988
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.907174698943255,
+      "learning_rate": 3.615109952495495e-06,
+      "loss": 0.4331,
+      "step": 3989
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.605965764594416,
+      "learning_rate": 3.612763140853771e-06,
+      "loss": 0.4421,
+      "step": 3990
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.9770040486925007,
+      "learning_rate": 3.6104166602618552e-06,
+      "loss": 0.4087,
+      "step": 3991
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.300240674285796,
+      "learning_rate": 3.6080705112797077e-06,
+      "loss": 0.4427,
+      "step": 3992
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.9492693362347153,
+      "learning_rate": 3.605724694467212e-06,
+      "loss": 0.4301,
+      "step": 3993
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.906306608948889,
+      "learning_rate": 3.6033792103841763e-06,
+      "loss": 0.3949,
+      "step": 3994
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.811142046787412,
+      "learning_rate": 3.601034059590324e-06,
+      "loss": 0.4986,
+      "step": 3995
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.8544245758084106,
+      "learning_rate": 3.5986892426453024e-06,
+      "loss": 0.4653,
+      "step": 3996
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.925799766472331,
+      "learning_rate": 3.596344760108678e-06,
+      "loss": 0.4193,
+      "step": 3997
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.502867745035287,
+      "learning_rate": 3.594000612539935e-06,
+      "loss": 0.4533,
+      "step": 3998
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.428192361402524,
+      "learning_rate": 3.591656800498483e-06,
+      "loss": 0.4336,
+      "step": 3999
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.569962374830117,
+      "learning_rate": 3.5893133245436475e-06,
+      "loss": 0.464,
+      "step": 4000
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7408903621180176,
+      "learning_rate": 3.5869701852346763e-06,
+      "loss": 0.3715,
+      "step": 4001
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 20.385413649027072,
+      "learning_rate": 3.584627383130733e-06,
+      "loss": 0.4159,
+      "step": 4002
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.788405223074752,
+      "learning_rate": 3.5822849187909027e-06,
+      "loss": 0.4541,
+      "step": 4003
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.3787999623014033,
+      "learning_rate": 3.579942792774195e-06,
+      "loss": 0.5147,
+      "step": 4004
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.851494378698466,
+      "learning_rate": 3.5776010056395294e-06,
+      "loss": 0.5066,
+      "step": 4005
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.035042938643484,
+      "learning_rate": 3.5752595579457526e-06,
+      "loss": 0.3936,
+      "step": 4006
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.8187556215828,
+      "learning_rate": 3.5729184502516255e-06,
+      "loss": 0.4363,
+      "step": 4007
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.3644227028526674,
+      "learning_rate": 3.5705776831158277e-06,
+      "loss": 0.428,
+      "step": 4008
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.347480314351162,
+      "learning_rate": 3.5682372570969628e-06,
+      "loss": 0.4233,
+      "step": 4009
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.113569062661106,
+      "learning_rate": 3.565897172753545e-06,
+      "loss": 0.3856,
+      "step": 4010
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.265739931296165,
+      "learning_rate": 3.563557430644017e-06,
+      "loss": 0.4245,
+      "step": 4011
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.1991291856913096,
+      "learning_rate": 3.561218031326731e-06,
+      "loss": 0.4459,
+      "step": 4012
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.077278530620917,
+      "learning_rate": 3.5588789753599594e-06,
+      "loss": 0.4508,
+      "step": 4013
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.144920789452597,
+      "learning_rate": 3.5565402633018963e-06,
+      "loss": 0.4683,
+      "step": 4014
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.784236318720534,
+      "learning_rate": 3.554201895710648e-06,
+      "loss": 0.4209,
+      "step": 4015
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.603804583771765,
+      "learning_rate": 3.5518638731442454e-06,
+      "loss": 0.47,
+      "step": 4016
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.918355727188468,
+      "learning_rate": 3.5495261961606332e-06,
+      "loss": 0.457,
+      "step": 4017
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1344735466323284,
+      "learning_rate": 3.547188865317671e-06,
+      "loss": 0.5404,
+      "step": 4018
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.2138200586442682,
+      "learning_rate": 3.5448518811731424e-06,
+      "loss": 0.4468,
+      "step": 4019
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.4180797151960425,
+      "learning_rate": 3.5425152442847405e-06,
+      "loss": 0.4581,
+      "step": 4020
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.62687617142585,
+      "learning_rate": 3.540178955210084e-06,
+      "loss": 0.5304,
+      "step": 4021
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.556292400368309,
+      "learning_rate": 3.537843014506702e-06,
+      "loss": 0.4269,
+      "step": 4022
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.4310153502267564,
+      "learning_rate": 3.535507422732043e-06,
+      "loss": 0.4044,
+      "step": 4023
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.8490371471988003,
+      "learning_rate": 3.5331721804434728e-06,
+      "loss": 0.4572,
+      "step": 4024
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.392016365796824,
+      "learning_rate": 3.5308372881982687e-06,
+      "loss": 0.4306,
+      "step": 4025
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.3600933707934075,
+      "learning_rate": 3.5285027465536353e-06,
+      "loss": 0.5201,
+      "step": 4026
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.7862054681419726,
+      "learning_rate": 3.5261685560666836e-06,
+      "loss": 0.5053,
+      "step": 4027
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1972869741681706,
+      "learning_rate": 3.523834717294442e-06,
+      "loss": 0.5238,
+      "step": 4028
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6851024667049597,
+      "learning_rate": 3.5215012307938622e-06,
+      "loss": 0.4111,
+      "step": 4029
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.687887414271254,
+      "learning_rate": 3.5191680971218024e-06,
+      "loss": 0.4187,
+      "step": 4030
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.522284973057458,
+      "learning_rate": 3.5168353168350433e-06,
+      "loss": 0.4383,
+      "step": 4031
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.648043688463352,
+      "learning_rate": 3.514502890490279e-06,
+      "loss": 0.4824,
+      "step": 4032
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.66733954413022,
+      "learning_rate": 3.512170818644115e-06,
+      "loss": 0.4491,
+      "step": 4033
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.356066978837546,
+      "learning_rate": 3.509839101853082e-06,
+      "loss": 0.4103,
+      "step": 4034
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.116187965764531,
+      "learning_rate": 3.507507740673617e-06,
+      "loss": 0.4113,
+      "step": 4035
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.5508064043655843,
+      "learning_rate": 3.505176735662077e-06,
+      "loss": 0.4832,
+      "step": 4036
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.14323474333763,
+      "learning_rate": 3.502846087374731e-06,
+      "loss": 0.4411,
+      "step": 4037
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.4688436849602327,
+      "learning_rate": 3.5005157963677625e-06,
+      "loss": 0.3867,
+      "step": 4038
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.965346364036287,
+      "learning_rate": 3.4981858631972764e-06,
+      "loss": 0.4535,
+      "step": 4039
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 19.753306391292035,
+      "learning_rate": 3.4958562884192827e-06,
+      "loss": 0.4672,
+      "step": 4040
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.212642947465992,
+      "learning_rate": 3.493527072589714e-06,
+      "loss": 0.4556,
+      "step": 4041
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.325872056679902,
+      "learning_rate": 3.4911982162644115e-06,
+      "loss": 0.4697,
+      "step": 4042
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4481126102536233,
+      "learning_rate": 3.488869719999131e-06,
+      "loss": 0.4224,
+      "step": 4043
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.136604087663385,
+      "learning_rate": 3.4865415843495485e-06,
+      "loss": 0.4373,
+      "step": 4044
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.354065009877868,
+      "learning_rate": 3.4842138098712463e-06,
+      "loss": 0.3881,
+      "step": 4045
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6080057548778397,
+      "learning_rate": 3.481886397119727e-06,
+      "loss": 0.4476,
+      "step": 4046
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.865163645571122,
+      "learning_rate": 3.479559346650401e-06,
+      "loss": 0.4976,
+      "step": 4047
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.2944238927361624,
+      "learning_rate": 3.4772326590185957e-06,
+      "loss": 0.4015,
+      "step": 4048
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.400152120338766,
+      "learning_rate": 3.4749063347795514e-06,
+      "loss": 0.4205,
+      "step": 4049
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.8344291810765423,
+      "learning_rate": 3.4725803744884197e-06,
+      "loss": 0.4682,
+      "step": 4050
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.2424943774541988,
+      "learning_rate": 3.4702547787002706e-06,
+      "loss": 0.4721,
+      "step": 4051
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.334924833238669,
+      "learning_rate": 3.4679295479700814e-06,
+      "loss": 0.4279,
+      "step": 4052
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.141393375792844,
+      "learning_rate": 3.465604682852744e-06,
+      "loss": 0.4881,
+      "step": 4053
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.643999399469282,
+      "learning_rate": 3.4632801839030656e-06,
+      "loss": 0.4852,
+      "step": 4054
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0040675409439754,
+      "learning_rate": 3.4609560516757602e-06,
+      "loss": 0.4923,
+      "step": 4055
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8225482656878365,
+      "learning_rate": 3.4586322867254628e-06,
+      "loss": 0.4855,
+      "step": 4056
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.21572189714957,
+      "learning_rate": 3.456308889606714e-06,
+      "loss": 0.381,
+      "step": 4057
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.625519854635449,
+      "learning_rate": 3.4539858608739667e-06,
+      "loss": 0.4962,
+      "step": 4058
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.399521888680049,
+      "learning_rate": 3.451663201081591e-06,
+      "loss": 0.4338,
+      "step": 4059
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.495567250438975,
+      "learning_rate": 3.4493409107838618e-06,
+      "loss": 0.3367,
+      "step": 4060
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8580752812641745,
+      "learning_rate": 3.447018990534975e-06,
+      "loss": 0.4818,
+      "step": 4061
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.5408231594346247,
+      "learning_rate": 3.4446974408890295e-06,
+      "loss": 0.3608,
+      "step": 4062
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.239499267516946,
+      "learning_rate": 3.442376262400038e-06,
+      "loss": 0.4519,
+      "step": 4063
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.55063718568584,
+      "learning_rate": 3.44005545562193e-06,
+      "loss": 0.5071,
+      "step": 4064
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.133845257128605,
+      "learning_rate": 3.437735021108539e-06,
+      "loss": 0.4319,
+      "step": 4065
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.707557216301683,
+      "learning_rate": 3.435414959413614e-06,
+      "loss": 0.4029,
+      "step": 4066
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.023728042445161,
+      "learning_rate": 3.433095271090815e-06,
+      "loss": 0.3984,
+      "step": 4067
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.4128677451251845,
+      "learning_rate": 3.430775956693707e-06,
+      "loss": 0.4201,
+      "step": 4068
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.9119567478435466,
+      "learning_rate": 3.428457016775777e-06,
+      "loss": 0.4666,
+      "step": 4069
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.0184967300669365,
+      "learning_rate": 3.4261384518904116e-06,
+      "loss": 0.4419,
+      "step": 4070
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7946789774822434,
+      "learning_rate": 3.423820262590915e-06,
+      "loss": 0.3978,
+      "step": 4071
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.314063419130091,
+      "learning_rate": 3.4215024494304982e-06,
+      "loss": 0.4918,
+      "step": 4072
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0664962220843424,
+      "learning_rate": 3.4191850129622816e-06,
+      "loss": 0.5281,
+      "step": 4073
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.063403964501148,
+      "learning_rate": 3.4168679537393013e-06,
+      "loss": 0.4897,
+      "step": 4074
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.9832900125881534,
+      "learning_rate": 3.4145512723144967e-06,
+      "loss": 0.3958,
+      "step": 4075
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.6376765553586967,
+      "learning_rate": 3.412234969240722e-06,
+      "loss": 0.3613,
+      "step": 4076
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.614262544507862,
+      "learning_rate": 3.409919045070739e-06,
+      "loss": 0.4669,
+      "step": 4077
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.471682755427336,
+      "learning_rate": 3.407603500357215e-06,
+      "loss": 0.4645,
+      "step": 4078
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.856667770090094,
+      "learning_rate": 3.4052883356527367e-06,
+      "loss": 0.4205,
+      "step": 4079
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.9295399918661236,
+      "learning_rate": 3.40297355150979e-06,
+      "loss": 0.4011,
+      "step": 4080
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.8220652952320067,
+      "learning_rate": 3.4006591484807784e-06,
+      "loss": 0.4016,
+      "step": 4081
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.096546809974063,
+      "learning_rate": 3.398345127118008e-06,
+      "loss": 0.5163,
+      "step": 4082
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.295085389338741,
+      "learning_rate": 3.3960314879736954e-06,
+      "loss": 0.3845,
+      "step": 4083
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.078405828436352,
+      "learning_rate": 3.3937182315999682e-06,
+      "loss": 0.4138,
+      "step": 4084
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.819176125989123,
+      "learning_rate": 3.391405358548859e-06,
+      "loss": 0.3911,
+      "step": 4085
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.088530077992319,
+      "learning_rate": 3.3890928693723156e-06,
+      "loss": 0.5482,
+      "step": 4086
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.9535732122171208,
+      "learning_rate": 3.3867807646221872e-06,
+      "loss": 0.4568,
+      "step": 4087
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 11.693660561605986,
+      "learning_rate": 3.3844690448502325e-06,
+      "loss": 0.4372,
+      "step": 4088
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7542625662105626,
+      "learning_rate": 3.3821577106081215e-06,
+      "loss": 0.42,
+      "step": 4089
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.133255499297002,
+      "learning_rate": 3.379846762447428e-06,
+      "loss": 0.3861,
+      "step": 4090
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.8231785986383025,
+      "learning_rate": 3.37753620091964e-06,
+      "loss": 0.4052,
+      "step": 4091
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.357325787470231,
+      "learning_rate": 3.375226026576146e-06,
+      "loss": 0.4904,
+      "step": 4092
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.593308397152471,
+      "learning_rate": 3.372916239968246e-06,
+      "loss": 0.4301,
+      "step": 4093
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.7230942151229764,
+      "learning_rate": 3.370606841647148e-06,
+      "loss": 0.4815,
+      "step": 4094
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.907366375839802,
+      "learning_rate": 3.3682978321639654e-06,
+      "loss": 0.4591,
+      "step": 4095
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.359502140478121,
+      "learning_rate": 3.365989212069717e-06,
+      "loss": 0.5102,
+      "step": 4096
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.161584237022204,
+      "learning_rate": 3.363680981915336e-06,
+      "loss": 0.4307,
+      "step": 4097
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.8795737103518575,
+      "learning_rate": 3.3613731422516534e-06,
+      "loss": 0.4453,
+      "step": 4098
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.9846780106631647,
+      "learning_rate": 3.359065693629415e-06,
+      "loss": 0.4085,
+      "step": 4099
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7881602507502006,
+      "learning_rate": 3.3567586365992687e-06,
+      "loss": 0.4043,
+      "step": 4100
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.670572860799817,
+      "learning_rate": 3.3544519717117683e-06,
+      "loss": 0.4712,
+      "step": 4101
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.3892969376016806,
+      "learning_rate": 3.3521456995173774e-06,
+      "loss": 0.4163,
+      "step": 4102
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.5974569106702914,
+      "learning_rate": 3.349839820566462e-06,
+      "loss": 0.4476,
+      "step": 4103
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 15.369537354780407,
+      "learning_rate": 3.347534335409299e-06,
+      "loss": 0.5016,
+      "step": 4104
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.441457117957543,
+      "learning_rate": 3.3452292445960676e-06,
+      "loss": 0.4207,
+      "step": 4105
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.83868381301182,
+      "learning_rate": 3.3429245486768538e-06,
+      "loss": 0.402,
+      "step": 4106
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.938646553091695,
+      "learning_rate": 3.3406202482016493e-06,
+      "loss": 0.4418,
+      "step": 4107
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.979804355797596,
+      "learning_rate": 3.3383163437203507e-06,
+      "loss": 0.4139,
+      "step": 4108
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.3365716750619128,
+      "learning_rate": 3.3360128357827647e-06,
+      "loss": 0.4409,
+      "step": 4109
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.44924014142774,
+      "learning_rate": 3.3337097249385973e-06,
+      "loss": 0.4408,
+      "step": 4110
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.9480677648974996,
+      "learning_rate": 3.331407011737462e-06,
+      "loss": 0.4514,
+      "step": 4111
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.1013596536723944,
+      "learning_rate": 3.329104696728879e-06,
+      "loss": 0.4645,
+      "step": 4112
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.4205428794364803,
+      "learning_rate": 3.3268027804622695e-06,
+      "loss": 0.5026,
+      "step": 4113
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.2625849021849,
+      "learning_rate": 3.324501263486965e-06,
+      "loss": 0.4282,
+      "step": 4114
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.483131019450267,
+      "learning_rate": 3.322200146352198e-06,
+      "loss": 0.4547,
+      "step": 4115
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.511018207964915,
+      "learning_rate": 3.319899429607104e-06,
+      "loss": 0.4586,
+      "step": 4116
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.329424894334045,
+      "learning_rate": 3.31759911380073e-06,
+      "loss": 0.4567,
+      "step": 4117
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.8689168702447008,
+      "learning_rate": 3.315299199482019e-06,
+      "loss": 0.4458,
+      "step": 4118
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.374575116351191,
+      "learning_rate": 3.3129996871998237e-06,
+      "loss": 0.446,
+      "step": 4119
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.8174257575364745,
+      "learning_rate": 3.310700577502899e-06,
+      "loss": 0.4529,
+      "step": 4120
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.1905396767409777,
+      "learning_rate": 3.3084018709399005e-06,
+      "loss": 0.4846,
+      "step": 4121
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.2684225263685573,
+      "learning_rate": 3.3061035680593967e-06,
+      "loss": 0.4467,
+      "step": 4122
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.819086736697005,
+      "learning_rate": 3.3038056694098485e-06,
+      "loss": 0.3351,
+      "step": 4123
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.169389399500903,
+      "learning_rate": 3.3015081755396305e-06,
+      "loss": 0.4266,
+      "step": 4124
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.6418206259424655,
+      "learning_rate": 3.2992110869970127e-06,
+      "loss": 0.4635,
+      "step": 4125
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2632004760848792,
+      "learning_rate": 3.2969144043301704e-06,
+      "loss": 0.5505,
+      "step": 4126
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.021828557340513,
+      "learning_rate": 3.294618128087188e-06,
+      "loss": 0.4407,
+      "step": 4127
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.5059832414780914,
+      "learning_rate": 3.292322258816044e-06,
+      "loss": 0.504,
+      "step": 4128
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3972618637035412,
+      "learning_rate": 3.2900267970646273e-06,
+      "loss": 0.5657,
+      "step": 4129
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.783531084916077,
+      "learning_rate": 3.2877317433807243e-06,
+      "loss": 0.492,
+      "step": 4130
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.557317078228341,
+      "learning_rate": 3.285437098312024e-06,
+      "loss": 0.4335,
+      "step": 4131
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.9531228885242755,
+      "learning_rate": 3.2831428624061246e-06,
+      "loss": 0.4381,
+      "step": 4132
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.967109994450408,
+      "learning_rate": 3.2808490362105178e-06,
+      "loss": 0.4152,
+      "step": 4133
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8485199355576016,
+      "learning_rate": 3.278555620272604e-06,
+      "loss": 0.4971,
+      "step": 4134
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6443138578268917,
+      "learning_rate": 3.2762626151396832e-06,
+      "loss": 0.4597,
+      "step": 4135
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.667453266928034,
+      "learning_rate": 3.2739700213589543e-06,
+      "loss": 0.4726,
+      "step": 4136
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.8378470931115167,
+      "learning_rate": 3.2716778394775255e-06,
+      "loss": 0.3917,
+      "step": 4137
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5455667890253646,
+      "learning_rate": 3.269386070042399e-06,
+      "loss": 0.4655,
+      "step": 4138
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.970105256220969,
+      "learning_rate": 3.2670947136004856e-06,
+      "loss": 0.4013,
+      "step": 4139
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 14.130624816058997,
+      "learning_rate": 3.264803770698592e-06,
+      "loss": 0.4824,
+      "step": 4140
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.767995597014194,
+      "learning_rate": 3.262513241883427e-06,
+      "loss": 0.4123,
+      "step": 4141
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.067819550933766,
+      "learning_rate": 3.260223127701604e-06,
+      "loss": 0.4174,
+      "step": 4142
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.810689746904155,
+      "learning_rate": 3.2579334286996313e-06,
+      "loss": 0.4471,
+      "step": 4143
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.100034766633999,
+      "learning_rate": 3.2556441454239274e-06,
+      "loss": 0.4174,
+      "step": 4144
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.5966439255199,
+      "learning_rate": 3.2533552784208032e-06,
+      "loss": 0.4261,
+      "step": 4145
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.090498322363859,
+      "learning_rate": 3.2510668282364734e-06,
+      "loss": 0.4081,
+      "step": 4146
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.1893001675459978,
+      "learning_rate": 3.2487787954170536e-06,
+      "loss": 0.3901,
+      "step": 4147
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.830068279838418,
+      "learning_rate": 3.2464911805085577e-06,
+      "loss": 0.4452,
+      "step": 4148
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7752334817967057,
+      "learning_rate": 3.244203984056904e-06,
+      "loss": 0.4624,
+      "step": 4149
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 11.920636569242394,
+      "learning_rate": 3.2419172066079073e-06,
+      "loss": 0.4285,
+      "step": 4150
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4336455029864017,
+      "learning_rate": 3.2396308487072826e-06,
+      "loss": 0.4268,
+      "step": 4151
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.6445613135995427,
+      "learning_rate": 3.2373449109006476e-06,
+      "loss": 0.4519,
+      "step": 4152
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.988567928380297,
+      "learning_rate": 3.235059393733515e-06,
+      "loss": 0.4561,
+      "step": 4153
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3286880896559583,
+      "learning_rate": 3.2327742977513034e-06,
+      "loss": 0.4349,
+      "step": 4154
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.859370437972936,
+      "learning_rate": 3.2304896234993255e-06,
+      "loss": 0.427,
+      "step": 4155
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.814227704538722,
+      "learning_rate": 3.2282053715227934e-06,
+      "loss": 0.4425,
+      "step": 4156
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.049519594715914,
+      "learning_rate": 3.225921542366825e-06,
+      "loss": 0.3947,
+      "step": 4157
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.270789754081606,
+      "learning_rate": 3.223638136576429e-06,
+      "loss": 0.471,
+      "step": 4158
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.180175986416722,
+      "learning_rate": 3.221355154696518e-06,
+      "loss": 0.4119,
+      "step": 4159
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.8695469719514453,
+      "learning_rate": 3.2190725972719033e-06,
+      "loss": 0.4502,
+      "step": 4160
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.4934928216042325,
+      "learning_rate": 3.2167904648472896e-06,
+      "loss": 0.4214,
+      "step": 4161
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 14.465713287539895,
+      "learning_rate": 3.2145087579672894e-06,
+      "loss": 0.4326,
+      "step": 4162
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.804850285086972,
+      "learning_rate": 3.212227477176405e-06,
+      "loss": 0.3765,
+      "step": 4163
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.8894928684584356,
+      "learning_rate": 3.2099466230190436e-06,
+      "loss": 0.4364,
+      "step": 4164
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.586257041590815,
+      "learning_rate": 3.207666196039505e-06,
+      "loss": 0.5021,
+      "step": 4165
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.784656500382422,
+      "learning_rate": 3.2053861967819894e-06,
+      "loss": 0.4536,
+      "step": 4166
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.9891940845795726,
+      "learning_rate": 3.2031066257905973e-06,
+      "loss": 0.5142,
+      "step": 4167
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2457393398365733,
+      "learning_rate": 3.200827483609324e-06,
+      "loss": 0.5392,
+      "step": 4168
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.277139420881489,
+      "learning_rate": 3.1985487707820635e-06,
+      "loss": 0.5737,
+      "step": 4169
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.162792230031295,
+      "learning_rate": 3.196270487852607e-06,
+      "loss": 0.457,
+      "step": 4170
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.779906764390809,
+      "learning_rate": 3.1939926353646412e-06,
+      "loss": 0.4938,
+      "step": 4171
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.451719390696094,
+      "learning_rate": 3.191715213861756e-06,
+      "loss": 0.4721,
+      "step": 4172
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.568459318857308,
+      "learning_rate": 3.1894382238874307e-06,
+      "loss": 0.3876,
+      "step": 4173
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.727431008057013,
+      "learning_rate": 3.1871616659850493e-06,
+      "loss": 0.4209,
+      "step": 4174
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.140329912955659,
+      "learning_rate": 3.1848855406978872e-06,
+      "loss": 0.4011,
+      "step": 4175
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.1335665617801824,
+      "learning_rate": 3.182609848569118e-06,
+      "loss": 0.369,
+      "step": 4176
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.292738467481602,
+      "learning_rate": 3.1803345901418126e-06,
+      "loss": 0.4024,
+      "step": 4177
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.0084688940698974,
+      "learning_rate": 3.1780597659589363e-06,
+      "loss": 0.4461,
+      "step": 4178
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.145930763914243,
+      "learning_rate": 3.175785376563356e-06,
+      "loss": 0.4897,
+      "step": 4179
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6364328485092097,
+      "learning_rate": 3.1735114224978295e-06,
+      "loss": 0.4418,
+      "step": 4180
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.9085466026260534,
+      "learning_rate": 3.1712379043050116e-06,
+      "loss": 0.4193,
+      "step": 4181
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.1657466834124266,
+      "learning_rate": 3.168964822527456e-06,
+      "loss": 0.5386,
+      "step": 4182
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.8090696852334833,
+      "learning_rate": 3.166692177707607e-06,
+      "loss": 0.4494,
+      "step": 4183
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.654037777930578,
+      "learning_rate": 3.164419970387812e-06,
+      "loss": 0.3667,
+      "step": 4184
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.375401668600867,
+      "learning_rate": 3.162148201110308e-06,
+      "loss": 0.4186,
+      "step": 4185
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.733822214842829,
+      "learning_rate": 3.1598768704172283e-06,
+      "loss": 0.4368,
+      "step": 4186
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.203142104926271,
+      "learning_rate": 3.1576059788506052e-06,
+      "loss": 0.4143,
+      "step": 4187
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5106096053421543,
+      "learning_rate": 3.1553355269523594e-06,
+      "loss": 0.4308,
+      "step": 4188
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.180485644778623,
+      "learning_rate": 3.1530655152643157e-06,
+      "loss": 0.4143,
+      "step": 4189
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.312917387630994,
+      "learning_rate": 3.1507959443281872e-06,
+      "loss": 0.3748,
+      "step": 4190
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.675023479696254,
+      "learning_rate": 3.148526814685581e-06,
+      "loss": 0.4486,
+      "step": 4191
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.935785844122816,
+      "learning_rate": 3.1462581268780058e-06,
+      "loss": 0.462,
+      "step": 4192
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.3755335751767,
+      "learning_rate": 3.1439898814468585e-06,
+      "loss": 0.4368,
+      "step": 4193
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.319869731786069,
+      "learning_rate": 3.1417220789334326e-06,
+      "loss": 0.4003,
+      "step": 4194
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.9281100697810745,
+      "learning_rate": 3.1394547198789173e-06,
+      "loss": 0.4332,
+      "step": 4195
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.50118220624619,
+      "learning_rate": 3.1371878048243897e-06,
+      "loss": 0.4153,
+      "step": 4196
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.597950823402391,
+      "learning_rate": 3.134921334310832e-06,
+      "loss": 0.4746,
+      "step": 4197
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.823342534201827,
+      "learning_rate": 3.1326553088791107e-06,
+      "loss": 0.4348,
+      "step": 4198
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.7286051322202436,
+      "learning_rate": 3.1303897290699904e-06,
+      "loss": 0.3972,
+      "step": 4199
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.57304316688605,
+      "learning_rate": 3.1281245954241286e-06,
+      "loss": 0.3477,
+      "step": 4200
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.843591605815458,
+      "learning_rate": 3.1258599084820734e-06,
+      "loss": 0.4924,
+      "step": 4201
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.446061494156459,
+      "learning_rate": 3.1235956687842733e-06,
+      "loss": 0.4437,
+      "step": 4202
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.2842103653263885,
+      "learning_rate": 3.1213318768710634e-06,
+      "loss": 0.4514,
+      "step": 4203
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.078509101844589,
+      "learning_rate": 3.119068533282675e-06,
+      "loss": 0.4097,
+      "step": 4204
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.8761872484029896,
+      "learning_rate": 3.1168056385592318e-06,
+      "loss": 0.4196,
+      "step": 4205
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.9881416998584487,
+      "learning_rate": 3.114543193240748e-06,
+      "loss": 0.4511,
+      "step": 4206
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.2859200436626455,
+      "learning_rate": 3.112281197867136e-06,
+      "loss": 0.3775,
+      "step": 4207
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.3738781816485077,
+      "learning_rate": 3.1100196529781977e-06,
+      "loss": 0.397,
+      "step": 4208
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.993913074115008,
+      "learning_rate": 3.107758559113625e-06,
+      "loss": 0.4757,
+      "step": 4209
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.829888873644678,
+      "learning_rate": 3.1054979168130074e-06,
+      "loss": 0.405,
+      "step": 4210
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.029030500269111,
+      "learning_rate": 3.103237726615822e-06,
+      "loss": 0.4492,
+      "step": 4211
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.0634323904280745,
+      "learning_rate": 3.1009779890614418e-06,
+      "loss": 0.4344,
+      "step": 4212
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.68774751470287,
+      "learning_rate": 3.098718704689129e-06,
+      "loss": 0.4587,
+      "step": 4213
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.484813580150848,
+      "learning_rate": 3.0964598740380354e-06,
+      "loss": 0.4554,
+      "step": 4214
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.148140489452931,
+      "learning_rate": 3.0942014976472134e-06,
+      "loss": 0.4763,
+      "step": 4215
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.5920501665135247,
+      "learning_rate": 3.0919435760555972e-06,
+      "loss": 0.4823,
+      "step": 4216
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4204315391585736,
+      "learning_rate": 3.0896861098020192e-06,
+      "loss": 0.3781,
+      "step": 4217
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.902008321270663,
+      "learning_rate": 3.0874290994251988e-06,
+      "loss": 0.3859,
+      "step": 4218
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.3183353791392496,
+      "learning_rate": 3.085172545463747e-06,
+      "loss": 0.4651,
+      "step": 4219
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.349838259522261,
+      "learning_rate": 3.082916448456171e-06,
+      "loss": 0.4811,
+      "step": 4220
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 11.414992933107373,
+      "learning_rate": 3.080660808940862e-06,
+      "loss": 0.434,
+      "step": 4221
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2349218711913075,
+      "learning_rate": 3.078405627456107e-06,
+      "loss": 0.4148,
+      "step": 4222
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.484004309073225,
+      "learning_rate": 3.076150904540081e-06,
+      "loss": 0.447,
+      "step": 4223
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 13.8554465107056,
+      "learning_rate": 3.073896640730849e-06,
+      "loss": 0.4097,
+      "step": 4224
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.4628825410170725,
+      "learning_rate": 3.071642836566371e-06,
+      "loss": 0.4373,
+      "step": 4225
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.473157458929054,
+      "learning_rate": 3.0693894925844906e-06,
+      "loss": 0.4502,
+      "step": 4226
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.9053544485036134,
+      "learning_rate": 3.0671366093229495e-06,
+      "loss": 0.4394,
+      "step": 4227
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.850874630445069,
+      "learning_rate": 3.0648841873193725e-06,
+      "loss": 0.4164,
+      "step": 4228
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.574953364336203,
+      "learning_rate": 3.062632227111276e-06,
+      "loss": 0.4255,
+      "step": 4229
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2038555013274514,
+      "learning_rate": 3.060380729236069e-06,
+      "loss": 0.4083,
+      "step": 4230
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.87574092147822,
+      "learning_rate": 3.0581296942310464e-06,
+      "loss": 0.394,
+      "step": 4231
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.504919368964063,
+      "learning_rate": 3.0558791226333974e-06,
+      "loss": 0.4633,
+      "step": 4232
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.033485624101838,
+      "learning_rate": 3.053629014980196e-06,
+      "loss": 0.4561,
+      "step": 4233
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7090215797176054,
+      "learning_rate": 3.051379371808406e-06,
+      "loss": 0.4361,
+      "step": 4234
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.131954816212699,
+      "learning_rate": 3.0491301936548847e-06,
+      "loss": 0.4176,
+      "step": 4235
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.681442399990167,
+      "learning_rate": 3.0468814810563714e-06,
+      "loss": 0.4007,
+      "step": 4236
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.83941352189141,
+      "learning_rate": 3.044633234549502e-06,
+      "loss": 0.468,
+      "step": 4237
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.3232015998886277,
+      "learning_rate": 3.0423854546707966e-06,
+      "loss": 0.4862,
+      "step": 4238
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.9000610673829677,
+      "learning_rate": 3.0401381419566624e-06,
+      "loss": 0.4197,
+      "step": 4239
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2137154805579864,
+      "learning_rate": 3.0378912969433994e-06,
+      "loss": 0.5219,
+      "step": 4240
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.03641470545588,
+      "learning_rate": 3.035644920167193e-06,
+      "loss": 0.4653,
+      "step": 4241
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.669601911310437,
+      "learning_rate": 3.03339901216412e-06,
+      "loss": 0.4383,
+      "step": 4242
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.5535364004142087,
+      "learning_rate": 3.0311535734701413e-06,
+      "loss": 0.4829,
+      "step": 4243
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 24.805328316506102,
+      "learning_rate": 3.028908604621107e-06,
+      "loss": 0.5084,
+      "step": 4244
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.3106147611164465,
+      "learning_rate": 3.0266641061527584e-06,
+      "loss": 0.4117,
+      "step": 4245
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2639217482381109,
+      "learning_rate": 3.0244200786007204e-06,
+      "loss": 0.518,
+      "step": 4246
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.9907223549953996,
+      "learning_rate": 3.0221765225005078e-06,
+      "loss": 0.429,
+      "step": 4247
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.999994271494147,
+      "learning_rate": 3.0199334383875223e-06,
+      "loss": 0.4165,
+      "step": 4248
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.07551204483128,
+      "learning_rate": 3.0176908267970494e-06,
+      "loss": 0.5264,
+      "step": 4249
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.004382197632611,
+      "learning_rate": 3.015448688264271e-06,
+      "loss": 0.3438,
+      "step": 4250
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.722289554512023,
+      "learning_rate": 3.013207023324246e-06,
+      "loss": 0.4296,
+      "step": 4251
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.860736230168339,
+      "learning_rate": 3.010965832511928e-06,
+      "loss": 0.4479,
+      "step": 4252
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.783670247345435,
+      "learning_rate": 3.0087251163621516e-06,
+      "loss": 0.4088,
+      "step": 4253
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2907179369922095,
+      "learning_rate": 3.0064848754096398e-06,
+      "loss": 0.4316,
+      "step": 4254
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.5934671897463204,
+      "learning_rate": 3.004245110189007e-06,
+      "loss": 0.4279,
+      "step": 4255
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.7973344784980134,
+      "learning_rate": 3.002005821234746e-06,
+      "loss": 0.4924,
+      "step": 4256
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.26083950210593,
+      "learning_rate": 2.999767009081243e-06,
+      "loss": 0.4468,
+      "step": 4257
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.255745570860005,
+      "learning_rate": 2.9975286742627667e-06,
+      "loss": 0.4909,
+      "step": 4258
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.819666463823498,
+      "learning_rate": 2.9952908173134704e-06,
+      "loss": 0.5017,
+      "step": 4259
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.763800155319947,
+      "learning_rate": 2.993053438767399e-06,
+      "loss": 0.4321,
+      "step": 4260
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.5531038028014956,
+      "learning_rate": 2.9908165391584775e-06,
+      "loss": 0.4306,
+      "step": 4261
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.8285004096833473,
+      "learning_rate": 2.9885801190205215e-06,
+      "loss": 0.4659,
+      "step": 4262
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 12.687664967667553,
+      "learning_rate": 2.986344178887228e-06,
+      "loss": 0.4262,
+      "step": 4263
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.6207045711248687,
+      "learning_rate": 2.98410871929218e-06,
+      "loss": 0.3788,
+      "step": 4264
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.606990772419031,
+      "learning_rate": 2.98187374076885e-06,
+      "loss": 0.4473,
+      "step": 4265
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.727685269673953,
+      "learning_rate": 2.979639243850588e-06,
+      "loss": 0.4185,
+      "step": 4266
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.0124699786689035,
+      "learning_rate": 2.9774052290706386e-06,
+      "loss": 0.432,
+      "step": 4267
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.221983650996144,
+      "learning_rate": 2.975171696962124e-06,
+      "loss": 0.341,
+      "step": 4268
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.4573217062743313,
+      "learning_rate": 2.9729386480580535e-06,
+      "loss": 0.4124,
+      "step": 4269
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.8209838307559965,
+      "learning_rate": 2.9707060828913226e-06,
+      "loss": 0.4969,
+      "step": 4270
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1869242059904859,
+      "learning_rate": 2.9684740019947073e-06,
+      "loss": 0.5589,
+      "step": 4271
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.844529840321809,
+      "learning_rate": 2.9662424059008742e-06,
+      "loss": 0.4462,
+      "step": 4272
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1814585118069272,
+      "learning_rate": 2.96401129514237e-06,
+      "loss": 0.5118,
+      "step": 4273
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.49959202083849,
+      "learning_rate": 2.961780670251624e-06,
+      "loss": 0.4241,
+      "step": 4274
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.6489546747736,
+      "learning_rate": 2.959550531760954e-06,
+      "loss": 0.4209,
+      "step": 4275
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.031413034018773,
+      "learning_rate": 2.9573208802025577e-06,
+      "loss": 0.4754,
+      "step": 4276
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.6903598897626875,
+      "learning_rate": 2.9550917161085213e-06,
+      "loss": 0.435,
+      "step": 4277
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.6026312676211463,
+      "learning_rate": 2.9528630400108097e-06,
+      "loss": 0.4949,
+      "step": 4278
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.476945756331593,
+      "learning_rate": 2.950634852441274e-06,
+      "loss": 0.439,
+      "step": 4279
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.8043460578808395,
+      "learning_rate": 2.9484071539316494e-06,
+      "loss": 0.451,
+      "step": 4280
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.850384099935864,
+      "learning_rate": 2.9461799450135502e-06,
+      "loss": 0.4102,
+      "step": 4281
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.700714102998691,
+      "learning_rate": 2.943953226218481e-06,
+      "loss": 0.3756,
+      "step": 4282
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.31944729888374,
+      "learning_rate": 2.941726998077823e-06,
+      "loss": 0.4065,
+      "step": 4283
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.7162586105225186,
+      "learning_rate": 2.939501261122841e-06,
+      "loss": 0.4639,
+      "step": 4284
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.807495758188896,
+      "learning_rate": 2.9372760158846884e-06,
+      "loss": 0.5127,
+      "step": 4285
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.71399331007614,
+      "learning_rate": 2.935051262894394e-06,
+      "loss": 0.4317,
+      "step": 4286
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.190015487512295,
+      "learning_rate": 2.9328270026828737e-06,
+      "loss": 0.4286,
+      "step": 4287
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.016634790335939,
+      "learning_rate": 2.9306032357809232e-06,
+      "loss": 0.5187,
+      "step": 4288
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.579473369660197,
+      "learning_rate": 2.928379962719221e-06,
+      "loss": 0.4945,
+      "step": 4289
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.78229745485092,
+      "learning_rate": 2.926157184028331e-06,
+      "loss": 0.4599,
+      "step": 4290
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.323466247480093,
+      "learning_rate": 2.923934900238694e-06,
+      "loss": 0.4777,
+      "step": 4291
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.814833020144376,
+      "learning_rate": 2.921713111880637e-06,
+      "loss": 0.4136,
+      "step": 4292
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.457495168334471,
+      "learning_rate": 2.919491819484366e-06,
+      "loss": 0.5199,
+      "step": 4293
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.14076624129154,
+      "learning_rate": 2.9172710235799686e-06,
+      "loss": 0.4343,
+      "step": 4294
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.560090442871984,
+      "learning_rate": 2.915050724697417e-06,
+      "loss": 0.4318,
+      "step": 4295
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 10.667521784069349,
+      "learning_rate": 2.9128309233665585e-06,
+      "loss": 0.4769,
+      "step": 4296
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.871134562263709,
+      "learning_rate": 2.910611620117132e-06,
+      "loss": 0.4954,
+      "step": 4297
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.761334207200338,
+      "learning_rate": 2.908392815478748e-06,
+      "loss": 0.4229,
+      "step": 4298
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.60594305620509,
+      "learning_rate": 2.9061745099808997e-06,
+      "loss": 0.4777,
+      "step": 4299
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.356389774295121,
+      "learning_rate": 2.903956704152967e-06,
+      "loss": 0.4686,
+      "step": 4300
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.832368351611848,
+      "learning_rate": 2.901739398524202e-06,
+      "loss": 0.4261,
+      "step": 4301
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.44051124362793,
+      "learning_rate": 2.8995225936237466e-06,
+      "loss": 0.4674,
+      "step": 4302
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.327171468363169,
+      "learning_rate": 2.897306289980616e-06,
+      "loss": 0.4292,
+      "step": 4303
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.848054489487304,
+      "learning_rate": 2.8950904881237065e-06,
+      "loss": 0.4074,
+      "step": 4304
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.7426854888982626,
+      "learning_rate": 2.8928751885817995e-06,
+      "loss": 0.4426,
+      "step": 4305
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.9630468565067294,
+      "learning_rate": 2.890660391883553e-06,
+      "loss": 0.4283,
+      "step": 4306
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1947282656835776,
+      "learning_rate": 2.8884460985575043e-06,
+      "loss": 0.5508,
+      "step": 4307
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.419579978689222,
+      "learning_rate": 2.886232309132072e-06,
+      "loss": 0.4615,
+      "step": 4308
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.029674828211915,
+      "learning_rate": 2.8840190241355526e-06,
+      "loss": 0.4432,
+      "step": 4309
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.046092568138356,
+      "learning_rate": 2.881806244096127e-06,
+      "loss": 0.4131,
+      "step": 4310
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.70295951788894,
+      "learning_rate": 2.879593969541849e-06,
+      "loss": 0.4169,
+      "step": 4311
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.9876104565843082,
+      "learning_rate": 2.8773822010006593e-06,
+      "loss": 0.4699,
+      "step": 4312
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.241154178734816,
+      "learning_rate": 2.875170939000371e-06,
+      "loss": 0.4659,
+      "step": 4313
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.186116935460794,
+      "learning_rate": 2.8729601840686777e-06,
+      "loss": 0.3915,
+      "step": 4314
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.440808919833905,
+      "learning_rate": 2.870749936733157e-06,
+      "loss": 0.4459,
+      "step": 4315
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.255718688164043,
+      "learning_rate": 2.868540197521258e-06,
+      "loss": 0.4706,
+      "step": 4316
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.9417146766522637,
+      "learning_rate": 2.866330966960316e-06,
+      "loss": 0.4347,
+      "step": 4317
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.697306060566815,
+      "learning_rate": 2.8641222455775388e-06,
+      "loss": 0.4401,
+      "step": 4318
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.365444123149797,
+      "learning_rate": 2.861914033900015e-06,
+      "loss": 0.4505,
+      "step": 4319
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.3409113103619354,
+      "learning_rate": 2.859706332454713e-06,
+      "loss": 0.4882,
+      "step": 4320
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1306698484462991,
+      "learning_rate": 2.857499141768476e-06,
+      "loss": 0.5001,
+      "step": 4321
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.410574765447539,
+      "learning_rate": 2.8552924623680273e-06,
+      "loss": 0.4975,
+      "step": 4322
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.563479156474254,
+      "learning_rate": 2.8530862947799707e-06,
+      "loss": 0.3978,
+      "step": 4323
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.292523661253694,
+      "learning_rate": 2.8508806395307825e-06,
+      "loss": 0.4633,
+      "step": 4324
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.447121391229444,
+      "learning_rate": 2.848675497146823e-06,
+      "loss": 0.4792,
+      "step": 4325
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.3254022648522366,
+      "learning_rate": 2.846470868154324e-06,
+      "loss": 0.4306,
+      "step": 4326
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.314454355077301,
+      "learning_rate": 2.8442667530793975e-06,
+      "loss": 0.4715,
+      "step": 4327
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3589335789412338,
+      "learning_rate": 2.8420631524480356e-06,
+      "loss": 0.5793,
+      "step": 4328
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.290905484796616,
+      "learning_rate": 2.8398600667861032e-06,
+      "loss": 0.4674,
+      "step": 4329
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.110200939111929,
+      "learning_rate": 2.8376574966193432e-06,
+      "loss": 0.4259,
+      "step": 4330
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.372939338726948,
+      "learning_rate": 2.8354554424733776e-06,
+      "loss": 0.4123,
+      "step": 4331
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 25.30634865411913,
+      "learning_rate": 2.833253904873702e-06,
+      "loss": 0.4843,
+      "step": 4332
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.905317262299432,
+      "learning_rate": 2.831052884345693e-06,
+      "loss": 0.4428,
+      "step": 4333
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.246732088165109,
+      "learning_rate": 2.828852381414599e-06,
+      "loss": 0.4258,
+      "step": 4334
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.341307430365894,
+      "learning_rate": 2.8266523966055516e-06,
+      "loss": 0.3687,
+      "step": 4335
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.875475913022291,
+      "learning_rate": 2.824452930443551e-06,
+      "loss": 0.4731,
+      "step": 4336
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.099843840421281,
+      "learning_rate": 2.822253983453477e-06,
+      "loss": 0.4175,
+      "step": 4337
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.805262952802066,
+      "learning_rate": 2.8200555561600884e-06,
+      "loss": 0.4217,
+      "step": 4338
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.8417496624621927,
+      "learning_rate": 2.8178576490880137e-06,
+      "loss": 0.4693,
+      "step": 4339
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.315259502344015,
+      "learning_rate": 2.8156602627617647e-06,
+      "loss": 0.4518,
+      "step": 4340
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.701438759857595,
+      "learning_rate": 2.8134633977057236e-06,
+      "loss": 0.4394,
+      "step": 4341
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.963209728896554,
+      "learning_rate": 2.811267054444149e-06,
+      "loss": 0.4053,
+      "step": 4342
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.50805954808813,
+      "learning_rate": 2.8090712335011755e-06,
+      "loss": 0.4224,
+      "step": 4343
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.351187008453021,
+      "learning_rate": 2.806875935400811e-06,
+      "loss": 0.4946,
+      "step": 4344
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.7530632582233223,
+      "learning_rate": 2.804681160666947e-06,
+      "loss": 0.4078,
+      "step": 4345
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.631067963612568,
+      "learning_rate": 2.8024869098233386e-06,
+      "loss": 0.3993,
+      "step": 4346
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.8057556953557485,
+      "learning_rate": 2.8002931833936213e-06,
+      "loss": 0.4569,
+      "step": 4347
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.794464173769737,
+      "learning_rate": 2.798099981901309e-06,
+      "loss": 0.3834,
+      "step": 4348
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.773465212202362,
+      "learning_rate": 2.795907305869783e-06,
+      "loss": 0.4113,
+      "step": 4349
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.628533675439235,
+      "learning_rate": 2.793715155822305e-06,
+      "loss": 0.4923,
+      "step": 4350
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.011423465785355,
+      "learning_rate": 2.7915235322820096e-06,
+      "loss": 0.4717,
+      "step": 4351
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.12729451346462,
+      "learning_rate": 2.789332435771904e-06,
+      "loss": 0.4027,
+      "step": 4352
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1433392242730365,
+      "learning_rate": 2.7871418668148703e-06,
+      "loss": 0.494,
+      "step": 4353
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.336123069759631,
+      "learning_rate": 2.7849518259336637e-06,
+      "loss": 0.4031,
+      "step": 4354
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.9470886230195013,
+      "learning_rate": 2.7827623136509197e-06,
+      "loss": 0.4852,
+      "step": 4355
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.250727188240011,
+      "learning_rate": 2.78057333048914e-06,
+      "loss": 0.3852,
+      "step": 4356
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 12.560849701238379,
+      "learning_rate": 2.7783848769707016e-06,
+      "loss": 0.4303,
+      "step": 4357
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.212088990541189,
+      "learning_rate": 2.7761969536178598e-06,
+      "loss": 0.42,
+      "step": 4358
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.567913330313754,
+      "learning_rate": 2.7740095609527363e-06,
+      "loss": 0.4302,
+      "step": 4359
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.809994176497501,
+      "learning_rate": 2.771822699497333e-06,
+      "loss": 0.4466,
+      "step": 4360
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.0247448050094,
+      "learning_rate": 2.7696363697735213e-06,
+      "loss": 0.4645,
+      "step": 4361
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.960530562498612,
+      "learning_rate": 2.767450572303043e-06,
+      "loss": 0.4404,
+      "step": 4362
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.103447009062216,
+      "learning_rate": 2.765265307607522e-06,
+      "loss": 0.4861,
+      "step": 4363
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.246512240991744,
+      "learning_rate": 2.763080576208445e-06,
+      "loss": 0.446,
+      "step": 4364
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.448272992424974,
+      "learning_rate": 2.760896378627177e-06,
+      "loss": 0.4706,
+      "step": 4365
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.829550773135516,
+      "learning_rate": 2.758712715384954e-06,
+      "loss": 0.4737,
+      "step": 4366
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.2126655343411805,
+      "learning_rate": 2.756529587002883e-06,
+      "loss": 0.3792,
+      "step": 4367
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.437739359027148,
+      "learning_rate": 2.7543469940019486e-06,
+      "loss": 0.4204,
+      "step": 4368
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.212829142047479,
+      "learning_rate": 2.7521649369030006e-06,
+      "loss": 0.4356,
+      "step": 4369
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.9906867974911813,
+      "learning_rate": 2.7499834162267684e-06,
+      "loss": 0.4764,
+      "step": 4370
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 17.452710708997856,
+      "learning_rate": 2.7478024324938474e-06,
+      "loss": 0.4624,
+      "step": 4371
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.7540841766833495,
+      "learning_rate": 2.7456219862247047e-06,
+      "loss": 0.3688,
+      "step": 4372
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.268272033177547,
+      "learning_rate": 2.7434420779396864e-06,
+      "loss": 0.4398,
+      "step": 4373
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.52362885239163,
+      "learning_rate": 2.7412627081589993e-06,
+      "loss": 0.3822,
+      "step": 4374
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.8909266895296435,
+      "learning_rate": 2.739083877402734e-06,
+      "loss": 0.46,
+      "step": 4375
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.953537456540554,
+      "learning_rate": 2.7369055861908418e-06,
+      "loss": 0.439,
+      "step": 4376
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.12068434109332,
+      "learning_rate": 2.7347278350431505e-06,
+      "loss": 0.4173,
+      "step": 4377
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.715467588011136,
+      "learning_rate": 2.7325506244793583e-06,
+      "loss": 0.459,
+      "step": 4378
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.25848876181839,
+      "learning_rate": 2.730373955019031e-06,
+      "loss": 0.411,
+      "step": 4379
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.45368503260071,
+      "learning_rate": 2.728197827181614e-06,
+      "loss": 0.4415,
+      "step": 4380
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2061540680364977,
+      "learning_rate": 2.7260222414864145e-06,
+      "loss": 0.5208,
+      "step": 4381
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.794568301926648,
+      "learning_rate": 2.723847198452612e-06,
+      "loss": 0.4405,
+      "step": 4382
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.870938639791854,
+      "learning_rate": 2.7216726985992627e-06,
+      "loss": 0.4574,
+      "step": 4383
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.479505958164214,
+      "learning_rate": 2.7194987424452834e-06,
+      "loss": 0.4528,
+      "step": 4384
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 11.514641412860714,
+      "learning_rate": 2.7173253305094717e-06,
+      "loss": 0.4463,
+      "step": 4385
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.9313806145047487,
+      "learning_rate": 2.7151524633104874e-06,
+      "loss": 0.3647,
+      "step": 4386
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.60966959573462,
+      "learning_rate": 2.7129801413668626e-06,
+      "loss": 0.4488,
+      "step": 4387
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.496082134423018,
+      "learning_rate": 2.710808365197e-06,
+      "loss": 0.528,
+      "step": 4388
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.980087493830957,
+      "learning_rate": 2.70863713531917e-06,
+      "loss": 0.4756,
+      "step": 4389
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.446874548239667,
+      "learning_rate": 2.7064664522515172e-06,
+      "loss": 0.4268,
+      "step": 4390
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.00240919918048,
+      "learning_rate": 2.7042963165120516e-06,
+      "loss": 0.4837,
+      "step": 4391
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.7764154699034873,
+      "learning_rate": 2.702126728618652e-06,
+      "loss": 0.3895,
+      "step": 4392
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.053744569789017,
+      "learning_rate": 2.699957689089071e-06,
+      "loss": 0.4383,
+      "step": 4393
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.497921837348814,
+      "learning_rate": 2.6977891984409254e-06,
+      "loss": 0.4618,
+      "step": 4394
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1559608750588002,
+      "learning_rate": 2.695621257191705e-06,
+      "loss": 0.4985,
+      "step": 4395
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.174164904502262,
+      "learning_rate": 2.693453865858767e-06,
+      "loss": 0.4354,
+      "step": 4396
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.682379889659037,
+      "learning_rate": 2.6912870249593333e-06,
+      "loss": 0.4576,
+      "step": 4397
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 35.94340271548921,
+      "learning_rate": 2.6891207350105025e-06,
+      "loss": 0.4143,
+      "step": 4398
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.154744498229646,
+      "learning_rate": 2.6869549965292364e-06,
+      "loss": 0.4779,
+      "step": 4399
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.1675618714759293,
+      "learning_rate": 2.6847898100323645e-06,
+      "loss": 0.4336,
+      "step": 4400
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.161446751526091,
+      "learning_rate": 2.6826251760365887e-06,
+      "loss": 0.4038,
+      "step": 4401
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.2678951590518412,
+      "learning_rate": 2.680461095058472e-06,
+      "loss": 0.4443,
+      "step": 4402
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.550965756267644,
+      "learning_rate": 2.678297567614455e-06,
+      "loss": 0.4793,
+      "step": 4403
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.952472736138958,
+      "learning_rate": 2.6761345942208384e-06,
+      "loss": 0.3951,
+      "step": 4404
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.9276301742694,
+      "learning_rate": 2.6739721753937963e-06,
+      "loss": 0.4009,
+      "step": 4405
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1446247743952453,
+      "learning_rate": 2.6718103116493666e-06,
+      "loss": 0.5652,
+      "step": 4406
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.9940060241387165,
+      "learning_rate": 2.669649003503453e-06,
+      "loss": 0.474,
+      "step": 4407
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.6215339534667277,
+      "learning_rate": 2.6674882514718335e-06,
+      "loss": 0.426,
+      "step": 4408
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.529765672138517,
+      "learning_rate": 2.665328056070147e-06,
+      "loss": 0.4699,
+      "step": 4409
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.598942845530387,
+      "learning_rate": 2.663168417813903e-06,
+      "loss": 0.4835,
+      "step": 4410
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 16.0907759285874,
+      "learning_rate": 2.6610093372184764e-06,
+      "loss": 0.4519,
+      "step": 4411
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.56009402126643,
+      "learning_rate": 2.658850814799111e-06,
+      "loss": 0.4197,
+      "step": 4412
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.886579430888448,
+      "learning_rate": 2.6566928510709135e-06,
+      "loss": 0.427,
+      "step": 4413
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.965329351102854,
+      "learning_rate": 2.6545354465488593e-06,
+      "loss": 0.4249,
+      "step": 4414
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 28.381606758235975,
+      "learning_rate": 2.652378601747795e-06,
+      "loss": 0.4896,
+      "step": 4415
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.034779291478902,
+      "learning_rate": 2.650222317182426e-06,
+      "loss": 0.4481,
+      "step": 4416
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.931396386376095,
+      "learning_rate": 2.648066593367327e-06,
+      "loss": 0.4271,
+      "step": 4417
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.7103258570970015,
+      "learning_rate": 2.6459114308169414e-06,
+      "loss": 0.3913,
+      "step": 4418
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.829445458272366,
+      "learning_rate": 2.643756830045574e-06,
+      "loss": 0.4875,
+      "step": 4419
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.599875800177118,
+      "learning_rate": 2.6416027915674027e-06,
+      "loss": 0.4439,
+      "step": 4420
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.9642682234151705,
+      "learning_rate": 2.6394493158964628e-06,
+      "loss": 0.414,
+      "step": 4421
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.582960253276647,
+      "learning_rate": 2.637296403546661e-06,
+      "loss": 0.3911,
+      "step": 4422
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.390873301554247,
+      "learning_rate": 2.6351440550317665e-06,
+      "loss": 0.4043,
+      "step": 4423
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.533277350610243,
+      "learning_rate": 2.6329922708654124e-06,
+      "loss": 0.3888,
+      "step": 4424
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.150295216726759,
+      "learning_rate": 2.630841051561106e-06,
+      "loss": 0.4754,
+      "step": 4425
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 272.026482859545,
+      "learning_rate": 2.62869039763221e-06,
+      "loss": 0.53,
+      "step": 4426
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.047501765964948,
+      "learning_rate": 2.6265403095919547e-06,
+      "loss": 0.4107,
+      "step": 4427
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.213980613919312,
+      "learning_rate": 2.6243907879534397e-06,
+      "loss": 0.4918,
+      "step": 4428
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.305880841344576,
+      "learning_rate": 2.622241833229623e-06,
+      "loss": 0.4531,
+      "step": 4429
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.923891070606426,
+      "learning_rate": 2.6200934459333337e-06,
+      "loss": 0.4043,
+      "step": 4430
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.487964336511981,
+      "learning_rate": 2.617945626577262e-06,
+      "loss": 0.4389,
+      "step": 4431
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.980207956292584,
+      "learning_rate": 2.615798375673958e-06,
+      "loss": 0.4549,
+      "step": 4432
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.962797961768096,
+      "learning_rate": 2.613651693735848e-06,
+      "loss": 0.4402,
+      "step": 4433
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.628685374014429,
+      "learning_rate": 2.611505581275212e-06,
+      "loss": 0.3865,
+      "step": 4434
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.997354702589493,
+      "learning_rate": 2.6093600388041977e-06,
+      "loss": 0.3422,
+      "step": 4435
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.3507032031446435,
+      "learning_rate": 2.6072150668348175e-06,
+      "loss": 0.4204,
+      "step": 4436
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.872216878208923,
+      "learning_rate": 2.6050706658789437e-06,
+      "loss": 0.3605,
+      "step": 4437
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.144199736648732,
+      "learning_rate": 2.6029268364483206e-06,
+      "loss": 0.4193,
+      "step": 4438
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.865539887399338,
+      "learning_rate": 2.6007835790545493e-06,
+      "loss": 0.43,
+      "step": 4439
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.516676269724278,
+      "learning_rate": 2.598640894209093e-06,
+      "loss": 0.4787,
+      "step": 4440
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.09768123760752,
+      "learning_rate": 2.5964987824232857e-06,
+      "loss": 0.413,
+      "step": 4441
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0510665768935583,
+      "learning_rate": 2.594357244208317e-06,
+      "loss": 0.5387,
+      "step": 4442
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.866635429938676,
+      "learning_rate": 2.592216280075246e-06,
+      "loss": 0.4361,
+      "step": 4443
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.679483229501885,
+      "learning_rate": 2.590075890534991e-06,
+      "loss": 0.4692,
+      "step": 4444
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 11.926879515943446,
+      "learning_rate": 2.5879360760983306e-06,
+      "loss": 0.3772,
+      "step": 4445
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.510184199589595,
+      "learning_rate": 2.585796837275914e-06,
+      "loss": 0.4201,
+      "step": 4446
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.921457635212152,
+      "learning_rate": 2.5836581745782474e-06,
+      "loss": 0.5154,
+      "step": 4447
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.476561580027541,
+      "learning_rate": 2.581520088515699e-06,
+      "loss": 0.3951,
+      "step": 4448
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1988205705169679,
+      "learning_rate": 2.5793825795985027e-06,
+      "loss": 0.5207,
+      "step": 4449
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.519278380788434,
+      "learning_rate": 2.57724564833675e-06,
+      "loss": 0.4473,
+      "step": 4450
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0648684055111914,
+      "learning_rate": 2.575109295240401e-06,
+      "loss": 0.5368,
+      "step": 4451
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.4232756253397914,
+      "learning_rate": 2.5729735208192726e-06,
+      "loss": 0.4183,
+      "step": 4452
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.871397854742682,
+      "learning_rate": 2.570838325583047e-06,
+      "loss": 0.4544,
+      "step": 4453
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.288503909523776,
+      "learning_rate": 2.568703710041266e-06,
+      "loss": 0.4885,
+      "step": 4454
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.676419313709423,
+      "learning_rate": 2.566569674703332e-06,
+      "loss": 0.5154,
+      "step": 4455
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.893021268469413,
+      "learning_rate": 2.564436220078513e-06,
+      "loss": 0.4583,
+      "step": 4456
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.178885096044598,
+      "learning_rate": 2.5623033466759345e-06,
+      "loss": 0.3964,
+      "step": 4457
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.302400830747199,
+      "learning_rate": 2.5601710550045856e-06,
+      "loss": 0.4631,
+      "step": 4458
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 12.107473780230713,
+      "learning_rate": 2.5580393455733153e-06,
+      "loss": 0.4786,
+      "step": 4459
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1321055198611474,
+      "learning_rate": 2.5559082188908317e-06,
+      "loss": 0.5672,
+      "step": 4460
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.2708641438076835,
+      "learning_rate": 2.553777675465711e-06,
+      "loss": 0.4397,
+      "step": 4461
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.767372584261381,
+      "learning_rate": 2.551647715806381e-06,
+      "loss": 0.4031,
+      "step": 4462
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.253141579549303,
+      "learning_rate": 2.549518340421139e-06,
+      "loss": 0.402,
+      "step": 4463
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.9464314145639103,
+      "learning_rate": 2.547389549818137e-06,
+      "loss": 0.4202,
+      "step": 4464
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 31.907346259845525,
+      "learning_rate": 2.5452613445053847e-06,
+      "loss": 0.411,
+      "step": 4465
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.070057855700322,
+      "learning_rate": 2.5431337249907628e-06,
+      "loss": 0.4431,
+      "step": 4466
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.733684069782123,
+      "learning_rate": 2.541006691782001e-06,
+      "loss": 0.3722,
+      "step": 4467
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 25.448426624984773,
+      "learning_rate": 2.538880245386698e-06,
+      "loss": 0.4109,
+      "step": 4468
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.188092790602591,
+      "learning_rate": 2.536754386312307e-06,
+      "loss": 0.432,
+      "step": 4469
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0657586111854107,
+      "learning_rate": 2.5346291150661407e-06,
+      "loss": 0.5393,
+      "step": 4470
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.254217490570659,
+      "learning_rate": 2.5325044321553737e-06,
+      "loss": 0.4649,
+      "step": 4471
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.626629848650046,
+      "learning_rate": 2.5303803380870394e-06,
+      "loss": 0.4522,
+      "step": 4472
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.4184715142252413,
+      "learning_rate": 2.5282568333680337e-06,
+      "loss": 0.4436,
+      "step": 4473
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.868419182342017,
+      "learning_rate": 2.5261339185051063e-06,
+      "loss": 0.4389,
+      "step": 4474
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.502172559428933,
+      "learning_rate": 2.5240115940048686e-06,
+      "loss": 0.4089,
+      "step": 4475
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.708852952818557,
+      "learning_rate": 2.5218898603737952e-06,
+      "loss": 0.3746,
+      "step": 4476
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.452830154847227,
+      "learning_rate": 2.5197687181182105e-06,
+      "loss": 0.4767,
+      "step": 4477
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.047366366421238,
+      "learning_rate": 2.51764816774431e-06,
+      "loss": 0.4036,
+      "step": 4478
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.636987678485354,
+      "learning_rate": 2.515528209758137e-06,
+      "loss": 0.3874,
+      "step": 4479
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 11.944147760602071,
+      "learning_rate": 2.5134088446655958e-06,
+      "loss": 0.4867,
+      "step": 4480
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.44987372891256,
+      "learning_rate": 2.5112900729724565e-06,
+      "loss": 0.4286,
+      "step": 4481
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.706559371110334,
+      "learning_rate": 2.5091718951843384e-06,
+      "loss": 0.4185,
+      "step": 4482
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.152327862547834,
+      "learning_rate": 2.5070543118067244e-06,
+      "loss": 0.4392,
+      "step": 4483
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.456112949079817,
+      "learning_rate": 2.5049373233449532e-06,
+      "loss": 0.3905,
+      "step": 4484
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.19055494127556,
+      "learning_rate": 2.5028209303042207e-06,
+      "loss": 0.4376,
+      "step": 4485
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.420412062691334,
+      "learning_rate": 2.500705133189585e-06,
+      "loss": 0.459,
+      "step": 4486
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.4612743290333983,
+      "learning_rate": 2.4985899325059578e-06,
+      "loss": 0.5248,
+      "step": 4487
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.686372145597631,
+      "learning_rate": 2.496475328758111e-06,
+      "loss": 0.3779,
+      "step": 4488
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.155366349644389,
+      "learning_rate": 2.494361322450674e-06,
+      "loss": 0.414,
+      "step": 4489
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.712826376929609,
+      "learning_rate": 2.4922479140881283e-06,
+      "loss": 0.5018,
+      "step": 4490
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.523784571382723,
+      "learning_rate": 2.490135104174823e-06,
+      "loss": 0.4569,
+      "step": 4491
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1600711872342322,
+      "learning_rate": 2.4880228932149546e-06,
+      "loss": 0.517,
+      "step": 4492
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.4214046008623615,
+      "learning_rate": 2.485911281712582e-06,
+      "loss": 0.4492,
+      "step": 4493
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.115657271223559,
+      "learning_rate": 2.4838002701716184e-06,
+      "loss": 0.4548,
+      "step": 4494
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.662059536090669,
+      "learning_rate": 2.4816898590958338e-06,
+      "loss": 0.4475,
+      "step": 4495
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.481344999517694,
+      "learning_rate": 2.4795800489888593e-06,
+      "loss": 0.3711,
+      "step": 4496
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.391467438549771,
+      "learning_rate": 2.4774708403541768e-06,
+      "loss": 0.4438,
+      "step": 4497
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.528469680813858,
+      "learning_rate": 2.47536223369513e-06,
+      "loss": 0.4724,
+      "step": 4498
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.4375114190833433,
+      "learning_rate": 2.4732542295149143e-06,
+      "loss": 0.3832,
+      "step": 4499
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.308241352733989,
+      "learning_rate": 2.471146828316582e-06,
+      "loss": 0.4379,
+      "step": 4500
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.4481875596949845,
+      "learning_rate": 2.4690400306030453e-06,
+      "loss": 0.4491,
+      "step": 4501
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.158016159275677,
+      "learning_rate": 2.4669338368770674e-06,
+      "loss": 0.4347,
+      "step": 4502
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.304517886631863,
+      "learning_rate": 2.4648282476412727e-06,
+      "loss": 0.4053,
+      "step": 4503
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.412040272731639,
+      "learning_rate": 2.4627232633981368e-06,
+      "loss": 0.4673,
+      "step": 4504
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.499352600200761,
+      "learning_rate": 2.460618884649993e-06,
+      "loss": 0.4697,
+      "step": 4505
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.643633887993586,
+      "learning_rate": 2.4585151118990286e-06,
+      "loss": 0.4103,
+      "step": 4506
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.798532353278695,
+      "learning_rate": 2.4564119456472856e-06,
+      "loss": 0.3388,
+      "step": 4507
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.767016457125586,
+      "learning_rate": 2.4543093863966675e-06,
+      "loss": 0.4759,
+      "step": 4508
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.2930411129763115,
+      "learning_rate": 2.452207434648926e-06,
+      "loss": 0.4265,
+      "step": 4509
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.266145038678747,
+      "learning_rate": 2.450106090905669e-06,
+      "loss": 0.4678,
+      "step": 4510
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.509971414357076,
+      "learning_rate": 2.4480053556683638e-06,
+      "loss": 0.4519,
+      "step": 4511
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.55741596874126,
+      "learning_rate": 2.445905229438325e-06,
+      "loss": 0.4723,
+      "step": 4512
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.199393472643776,
+      "learning_rate": 2.4438057127167307e-06,
+      "loss": 0.3809,
+      "step": 4513
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.364261609779675,
+      "learning_rate": 2.441706806004607e-06,
+      "loss": 0.4307,
+      "step": 4514
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.270565185147571,
+      "learning_rate": 2.439608509802836e-06,
+      "loss": 0.4114,
+      "step": 4515
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.329635907436552,
+      "learning_rate": 2.437510824612155e-06,
+      "loss": 0.4027,
+      "step": 4516
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.783217983060687,
+      "learning_rate": 2.4354137509331526e-06,
+      "loss": 0.4066,
+      "step": 4517
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.266490816961862,
+      "learning_rate": 2.4333172892662784e-06,
+      "loss": 0.4658,
+      "step": 4518
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.124930718978582,
+      "learning_rate": 2.4312214401118282e-06,
+      "loss": 0.4626,
+      "step": 4519
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.9660690622702095,
+      "learning_rate": 2.4291262039699544e-06,
+      "loss": 0.4937,
+      "step": 4520
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.305849853422665,
+      "learning_rate": 2.4270315813406664e-06,
+      "loss": 0.4454,
+      "step": 4521
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.698637588666971,
+      "learning_rate": 2.424937572723821e-06,
+      "loss": 0.4753,
+      "step": 4522
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.586873932601037,
+      "learning_rate": 2.422844178619135e-06,
+      "loss": 0.4649,
+      "step": 4523
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.092495400873869,
+      "learning_rate": 2.4207513995261745e-06,
+      "loss": 0.4092,
+      "step": 4524
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.3905903417965355,
+      "learning_rate": 2.418659235944357e-06,
+      "loss": 0.3599,
+      "step": 4525
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.656486039008912,
+      "learning_rate": 2.41656768837296e-06,
+      "loss": 0.4161,
+      "step": 4526
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.8032244113971356,
+      "learning_rate": 2.4144767573111074e-06,
+      "loss": 0.4164,
+      "step": 4527
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.687471682575265,
+      "learning_rate": 2.412386443257779e-06,
+      "loss": 0.48,
+      "step": 4528
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.466082416505405,
+      "learning_rate": 2.410296746711806e-06,
+      "loss": 0.4881,
+      "step": 4529
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.772450169476161,
+      "learning_rate": 2.408207668171872e-06,
+      "loss": 0.3472,
+      "step": 4530
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.307142647415721,
+      "learning_rate": 2.406119208136517e-06,
+      "loss": 0.4412,
+      "step": 4531
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 16.53462810148362,
+      "learning_rate": 2.4040313671041276e-06,
+      "loss": 0.4462,
+      "step": 4532
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.730368474049808,
+      "learning_rate": 2.4019441455729494e-06,
+      "loss": 0.492,
+      "step": 4533
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.6306245641289125,
+      "learning_rate": 2.3998575440410733e-06,
+      "loss": 0.4854,
+      "step": 4534
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.6512486764189913,
+      "learning_rate": 2.3977715630064442e-06,
+      "loss": 0.4513,
+      "step": 4535
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.175449154591465,
+      "learning_rate": 2.395686202966864e-06,
+      "loss": 0.4322,
+      "step": 4536
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 24.026666226977785,
+      "learning_rate": 2.393601464419977e-06,
+      "loss": 0.4016,
+      "step": 4537
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.131896151842517,
+      "learning_rate": 2.3915173478632916e-06,
+      "loss": 0.3715,
+      "step": 4538
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.129544868409922,
+      "learning_rate": 2.389433853794156e-06,
+      "loss": 0.4685,
+      "step": 4539
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.333273276297152,
+      "learning_rate": 2.3873509827097756e-06,
+      "loss": 0.4946,
+      "step": 4540
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.398083876655644,
+      "learning_rate": 2.385268735107207e-06,
+      "loss": 0.3865,
+      "step": 4541
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.045336038096149,
+      "learning_rate": 2.3831871114833537e-06,
+      "loss": 0.5099,
+      "step": 4542
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.184042356912598,
+      "learning_rate": 2.3811061123349783e-06,
+      "loss": 0.5092,
+      "step": 4543
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.186751493810831,
+      "learning_rate": 2.3790257381586882e-06,
+      "loss": 0.4478,
+      "step": 4544
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.306576889678205,
+      "learning_rate": 2.376945989450941e-06,
+      "loss": 0.4906,
+      "step": 4545
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.917671598270758,
+      "learning_rate": 2.3748668667080516e-06,
+      "loss": 0.4311,
+      "step": 4546
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.709369186023426,
+      "learning_rate": 2.372788370426179e-06,
+      "loss": 0.4034,
+      "step": 4547
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.028036620918392,
+      "learning_rate": 2.3707105011013326e-06,
+      "loss": 0.4585,
+      "step": 4548
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.153543576664654,
+      "learning_rate": 2.368633259229379e-06,
+      "loss": 0.4197,
+      "step": 4549
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.799360530377895,
+      "learning_rate": 2.366556645306029e-06,
+      "loss": 0.3663,
+      "step": 4550
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.829285559413873,
+      "learning_rate": 2.3644806598268446e-06,
+      "loss": 0.4434,
+      "step": 4551
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.146776232131839,
+      "learning_rate": 2.3624053032872385e-06,
+      "loss": 0.3757,
+      "step": 4552
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.729286985052793,
+      "learning_rate": 2.360330576182471e-06,
+      "loss": 0.3892,
+      "step": 4553
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.133138191483974,
+      "learning_rate": 2.3582564790076586e-06,
+      "loss": 0.47,
+      "step": 4554
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 15.039260666038281,
+      "learning_rate": 2.356183012257759e-06,
+      "loss": 0.5098,
+      "step": 4555
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.564456473143087,
+      "learning_rate": 2.3541101764275883e-06,
+      "loss": 0.4014,
+      "step": 4556
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.030094906203111,
+      "learning_rate": 2.352037972011805e-06,
+      "loss": 0.4528,
+      "step": 4557
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.959836056600525,
+      "learning_rate": 2.3499663995049165e-06,
+      "loss": 0.4613,
+      "step": 4558
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.7974239036737965,
+      "learning_rate": 2.3478954594012884e-06,
+      "loss": 0.4467,
+      "step": 4559
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.6174852007861995,
+      "learning_rate": 2.3458251521951236e-06,
+      "loss": 0.4046,
+      "step": 4560
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.9198416112721945,
+      "learning_rate": 2.3437554783804838e-06,
+      "loss": 0.3758,
+      "step": 4561
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.414526658694877,
+      "learning_rate": 2.341686438451274e-06,
+      "loss": 0.4066,
+      "step": 4562
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.366281314133032,
+      "learning_rate": 2.3396180329012488e-06,
+      "loss": 0.46,
+      "step": 4563
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.94041075223834,
+      "learning_rate": 2.3375502622240115e-06,
+      "loss": 0.4072,
+      "step": 4564
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.28504549223454,
+      "learning_rate": 2.3354831269130133e-06,
+      "loss": 0.4463,
+      "step": 4565
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.68620075171947,
+      "learning_rate": 2.3334166274615582e-06,
+      "loss": 0.399,
+      "step": 4566
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.307452964421236,
+      "learning_rate": 2.331350764362793e-06,
+      "loss": 0.4453,
+      "step": 4567
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.037296210984081,
+      "learning_rate": 2.3292855381097134e-06,
+      "loss": 0.4667,
+      "step": 4568
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.935676980907501,
+      "learning_rate": 2.3272209491951674e-06,
+      "loss": 0.4206,
+      "step": 4569
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.223737573691994,
+      "learning_rate": 2.325156998111844e-06,
+      "loss": 0.4843,
+      "step": 4570
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.426790935369577,
+      "learning_rate": 2.323093685352289e-06,
+      "loss": 0.4588,
+      "step": 4571
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.597001989728083,
+      "learning_rate": 2.3210310114088876e-06,
+      "loss": 0.4185,
+      "step": 4572
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.297678665704269,
+      "learning_rate": 2.3189689767738737e-06,
+      "loss": 0.4298,
+      "step": 4573
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.8997794861676125,
+      "learning_rate": 2.3169075819393355e-06,
+      "loss": 0.4147,
+      "step": 4574
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.999493847655554,
+      "learning_rate": 2.3148468273972013e-06,
+      "loss": 0.4574,
+      "step": 4575
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.1911402605255725,
+      "learning_rate": 2.3127867136392487e-06,
+      "loss": 0.4347,
+      "step": 4576
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.165114242751637,
+      "learning_rate": 2.3107272411571024e-06,
+      "loss": 0.4886,
+      "step": 4577
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.972705143940993,
+      "learning_rate": 2.3086684104422333e-06,
+      "loss": 0.4216,
+      "step": 4578
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.926095429899442,
+      "learning_rate": 2.306610221985964e-06,
+      "loss": 0.4322,
+      "step": 4579
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.489777705254749,
+      "learning_rate": 2.3045526762794546e-06,
+      "loss": 0.4706,
+      "step": 4580
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.204448492801335,
+      "learning_rate": 2.3024957738137227e-06,
+      "loss": 0.4188,
+      "step": 4581
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.802353210661846,
+      "learning_rate": 2.3004395150796244e-06,
+      "loss": 0.3693,
+      "step": 4582
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.5014865840948906,
+      "learning_rate": 2.298383900567863e-06,
+      "loss": 0.4722,
+      "step": 4583
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.7536278404125865,
+      "learning_rate": 2.2963289307689933e-06,
+      "loss": 0.3982,
+      "step": 4584
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.249145251195088,
+      "learning_rate": 2.294274606173411e-06,
+      "loss": 0.4896,
+      "step": 4585
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.197223032904191,
+      "learning_rate": 2.2922209272713598e-06,
+      "loss": 0.5714,
+      "step": 4586
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1648321594125401,
+      "learning_rate": 2.290167894552928e-06,
+      "loss": 0.4932,
+      "step": 4587
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.463588051263578,
+      "learning_rate": 2.2881155085080505e-06,
+      "loss": 0.4655,
+      "step": 4588
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.667121994737903,
+      "learning_rate": 2.28606376962651e-06,
+      "loss": 0.4012,
+      "step": 4589
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.090926567888159,
+      "learning_rate": 2.2840126783979307e-06,
+      "loss": 0.4878,
+      "step": 4590
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.6082281652327035,
+      "learning_rate": 2.281962235311787e-06,
+      "loss": 0.4537,
+      "step": 4591
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.7634535940592695,
+      "learning_rate": 2.279912440857395e-06,
+      "loss": 0.4316,
+      "step": 4592
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.169632495348908,
+      "learning_rate": 2.277863295523914e-06,
+      "loss": 0.4203,
+      "step": 4593
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.690919204361506,
+      "learning_rate": 2.2758147998003575e-06,
+      "loss": 0.4343,
+      "step": 4594
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 29.483921318362853,
+      "learning_rate": 2.273766954175572e-06,
+      "loss": 0.4786,
+      "step": 4595
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.751363819404805,
+      "learning_rate": 2.2717197591382595e-06,
+      "loss": 0.3788,
+      "step": 4596
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.5950124046572824,
+      "learning_rate": 2.26967321517696e-06,
+      "loss": 0.4025,
+      "step": 4597
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.9530483590891334,
+      "learning_rate": 2.26762732278006e-06,
+      "loss": 0.4013,
+      "step": 4598
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.24545903136202,
+      "learning_rate": 2.2655820824357904e-06,
+      "loss": 0.4219,
+      "step": 4599
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.563241387138226,
+      "learning_rate": 2.2635374946322257e-06,
+      "loss": 0.4323,
+      "step": 4600
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.882516292669679,
+      "learning_rate": 2.2614935598572896e-06,
+      "loss": 0.453,
+      "step": 4601
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.078845839984108,
+      "learning_rate": 2.2594502785987434e-06,
+      "loss": 0.3548,
+      "step": 4602
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.250606849565479,
+      "learning_rate": 2.257407651344193e-06,
+      "loss": 0.4155,
+      "step": 4603
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.420533821994416,
+      "learning_rate": 2.255365678581095e-06,
+      "loss": 0.3934,
+      "step": 4604
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.886899516945829,
+      "learning_rate": 2.2533243607967416e-06,
+      "loss": 0.4555,
+      "step": 4605
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.86707533288496,
+      "learning_rate": 2.2512836984782753e-06,
+      "loss": 0.4299,
+      "step": 4606
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 26.974018865404727,
+      "learning_rate": 2.249243692112678e-06,
+      "loss": 0.4535,
+      "step": 4607
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.997298008021307,
+      "learning_rate": 2.2472043421867734e-06,
+      "loss": 0.4571,
+      "step": 4608
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.293712151444383,
+      "learning_rate": 2.2451656491872357e-06,
+      "loss": 0.4524,
+      "step": 4609
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.9687203949358665,
+      "learning_rate": 2.243127613600577e-06,
+      "loss": 0.4269,
+      "step": 4610
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.966021983752036,
+      "learning_rate": 2.241090235913152e-06,
+      "loss": 0.3689,
+      "step": 4611
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.989376092084924,
+      "learning_rate": 2.2390535166111603e-06,
+      "loss": 0.3786,
+      "step": 4612
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.968203036394358,
+      "learning_rate": 2.237017456180643e-06,
+      "loss": 0.3963,
+      "step": 4613
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.036780402997223,
+      "learning_rate": 2.234982055107488e-06,
+      "loss": 0.4976,
+      "step": 4614
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 10.410156394099785,
+      "learning_rate": 2.2329473138774192e-06,
+      "loss": 0.5123,
+      "step": 4615
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 10.948106820397602,
+      "learning_rate": 2.23091323297601e-06,
+      "loss": 0.4463,
+      "step": 4616
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.800167084295893,
+      "learning_rate": 2.2288798128886722e-06,
+      "loss": 0.3882,
+      "step": 4617
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 12.135761369676365,
+      "learning_rate": 2.2268470541006574e-06,
+      "loss": 0.411,
+      "step": 4618
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.420117036253311,
+      "learning_rate": 2.2248149570970672e-06,
+      "loss": 0.4848,
+      "step": 4619
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1551565530720522,
+      "learning_rate": 2.2227835223628387e-06,
+      "loss": 0.5533,
+      "step": 4620
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 23.697108492652568,
+      "learning_rate": 2.220752750382752e-06,
+      "loss": 0.3574,
+      "step": 4621
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.9671962894469255,
+      "learning_rate": 2.218722641641431e-06,
+      "loss": 0.4243,
+      "step": 4622
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.353322658454584,
+      "learning_rate": 2.216693196623338e-06,
+      "loss": 0.3787,
+      "step": 4623
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 13.83502814193874,
+      "learning_rate": 2.2146644158127827e-06,
+      "loss": 0.3777,
+      "step": 4624
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 10.693264209919535,
+      "learning_rate": 2.212636299693909e-06,
+      "loss": 0.3688,
+      "step": 4625
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.086954309172666,
+      "learning_rate": 2.2106088487507104e-06,
+      "loss": 0.4273,
+      "step": 4626
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.085475982982225,
+      "learning_rate": 2.2085820634670148e-06,
+      "loss": 0.38,
+      "step": 4627
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.61605925906109,
+      "learning_rate": 2.206555944326492e-06,
+      "loss": 0.4852,
+      "step": 4628
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.11558863841789,
+      "learning_rate": 2.2045304918126574e-06,
+      "loss": 0.4336,
+      "step": 4629
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.854839164108935,
+      "learning_rate": 2.202505706408862e-06,
+      "loss": 0.4531,
+      "step": 4630
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.452764270677095,
+      "learning_rate": 2.2004815885983026e-06,
+      "loss": 0.4723,
+      "step": 4631
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.205039843815364,
+      "learning_rate": 2.1984581388640136e-06,
+      "loss": 0.4369,
+      "step": 4632
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 11.060438635453224,
+      "learning_rate": 2.1964353576888687e-06,
+      "loss": 0.4104,
+      "step": 4633
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.634745901049915,
+      "learning_rate": 2.194413245555585e-06,
+      "loss": 0.4607,
+      "step": 4634
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.584523491819987,
+      "learning_rate": 2.1923918029467165e-06,
+      "loss": 0.4561,
+      "step": 4635
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 35.25405147532874,
+      "learning_rate": 2.190371030344664e-06,
+      "loss": 0.398,
+      "step": 4636
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.715199306268774,
+      "learning_rate": 2.1883509282316613e-06,
+      "loss": 0.45,
+      "step": 4637
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.231615177859496,
+      "learning_rate": 2.1863314970897837e-06,
+      "loss": 0.4158,
+      "step": 4638
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.552721319654022,
+      "learning_rate": 2.184312737400951e-06,
+      "loss": 0.5289,
+      "step": 4639
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.34897053490368,
+      "learning_rate": 2.182294649646916e-06,
+      "loss": 0.4341,
+      "step": 4640
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.437392369300287,
+      "learning_rate": 2.1802772343092787e-06,
+      "loss": 0.4257,
+      "step": 4641
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 20.056076856889685,
+      "learning_rate": 2.178260491869472e-06,
+      "loss": 0.4329,
+      "step": 4642
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.844687318809765,
+      "learning_rate": 2.1762444228087708e-06,
+      "loss": 0.3964,
+      "step": 4643
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.603557886156365,
+      "learning_rate": 2.1742290276082893e-06,
+      "loss": 0.4116,
+      "step": 4644
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 16.2170742957076,
+      "learning_rate": 2.1722143067489797e-06,
+      "loss": 0.384,
+      "step": 4645
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.220376346298095,
+      "learning_rate": 2.170200260711637e-06,
+      "loss": 0.4162,
+      "step": 4646
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 13.467417263954843,
+      "learning_rate": 2.1681868899768916e-06,
+      "loss": 0.4421,
+      "step": 4647
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.269462910136216,
+      "learning_rate": 2.166174195025211e-06,
+      "loss": 0.4284,
+      "step": 4648
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.004013252437515,
+      "learning_rate": 2.1641621763369086e-06,
+      "loss": 0.4043,
+      "step": 4649
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.560257411420245,
+      "learning_rate": 2.1621508343921275e-06,
+      "loss": 0.416,
+      "step": 4650
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.020505422365288,
+      "learning_rate": 2.1601401696708583e-06,
+      "loss": 0.4375,
+      "step": 4651
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.201869938330477,
+      "learning_rate": 2.1581301826529232e-06,
+      "loss": 0.4585,
+      "step": 4652
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.889826556007674,
+      "learning_rate": 2.1561208738179833e-06,
+      "loss": 0.4516,
+      "step": 4653
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.619533407874556,
+      "learning_rate": 2.154112243645543e-06,
+      "loss": 0.4502,
+      "step": 4654
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.023614709868001,
+      "learning_rate": 2.152104292614939e-06,
+      "loss": 0.4321,
+      "step": 4655
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.210272671405171,
+      "learning_rate": 2.1500970212053492e-06,
+      "loss": 0.3809,
+      "step": 4656
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.441683276674121,
+      "learning_rate": 2.148090429895787e-06,
+      "loss": 0.4349,
+      "step": 4657
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.767996467563625,
+      "learning_rate": 2.1460845191651036e-06,
+      "loss": 0.4178,
+      "step": 4658
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 15.117822916745567,
+      "learning_rate": 2.1440792894919932e-06,
+      "loss": 0.4851,
+      "step": 4659
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.285538550767005,
+      "learning_rate": 2.1420747413549805e-06,
+      "loss": 0.4687,
+      "step": 4660
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.930910529415885,
+      "learning_rate": 2.140070875232428e-06,
+      "loss": 0.4365,
+      "step": 4661
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.070733697446871,
+      "learning_rate": 2.138067691602543e-06,
+      "loss": 0.414,
+      "step": 4662
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.177455473795536,
+      "learning_rate": 2.1360651909433596e-06,
+      "loss": 0.4108,
+      "step": 4663
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.0297085019450165,
+      "learning_rate": 2.134063373732759e-06,
+      "loss": 0.4308,
+      "step": 4664
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.3707977985592805,
+      "learning_rate": 2.1320622404484503e-06,
+      "loss": 0.3829,
+      "step": 4665
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.208764243613324,
+      "learning_rate": 2.130061791567984e-06,
+      "loss": 0.3999,
+      "step": 4666
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.44023721421205,
+      "learning_rate": 2.1280620275687487e-06,
+      "loss": 0.4692,
+      "step": 4667
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.36590258757543,
+      "learning_rate": 2.1260629489279662e-06,
+      "loss": 0.4444,
+      "step": 4668
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 12.946079389636454,
+      "learning_rate": 2.1240645561226964e-06,
+      "loss": 0.4797,
+      "step": 4669
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.261322914908133,
+      "learning_rate": 2.122066849629834e-06,
+      "loss": 0.3824,
+      "step": 4670
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.135524966736439,
+      "learning_rate": 2.1200698299261106e-06,
+      "loss": 0.4369,
+      "step": 4671
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.396645568956617,
+      "learning_rate": 2.118073497488098e-06,
+      "loss": 0.4547,
+      "step": 4672
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.597303668527404,
+      "learning_rate": 2.1160778527921953e-06,
+      "loss": 0.4345,
+      "step": 4673
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.8010262912373305,
+      "learning_rate": 2.114082896314648e-06,
+      "loss": 0.4077,
+      "step": 4674
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.933496550250718,
+      "learning_rate": 2.112088628531529e-06,
+      "loss": 0.4337,
+      "step": 4675
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.899411637486555,
+      "learning_rate": 2.1100950499187483e-06,
+      "loss": 0.4384,
+      "step": 4676
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 12.286007187349762,
+      "learning_rate": 2.108102160952057e-06,
+      "loss": 0.4037,
+      "step": 4677
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.896029011362263,
+      "learning_rate": 2.106109962107035e-06,
+      "loss": 0.4439,
+      "step": 4678
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.518024639038835,
+      "learning_rate": 2.1041184538590996e-06,
+      "loss": 0.4726,
+      "step": 4679
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.87395562555042,
+      "learning_rate": 2.1021276366835052e-06,
+      "loss": 0.419,
+      "step": 4680
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 12.02543737973597,
+      "learning_rate": 2.1001375110553373e-06,
+      "loss": 0.368,
+      "step": 4681
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.734900180804966,
+      "learning_rate": 2.0981480774495226e-06,
+      "loss": 0.413,
+      "step": 4682
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.542658841616869,
+      "learning_rate": 2.0961593363408154e-06,
+      "loss": 0.3926,
+      "step": 4683
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.464696121585491,
+      "learning_rate": 2.0941712882038116e-06,
+      "loss": 0.4628,
+      "step": 4684
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.402154659081515,
+      "learning_rate": 2.0921839335129374e-06,
+      "loss": 0.4044,
+      "step": 4685
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2150549324737538,
+      "learning_rate": 2.090197272742452e-06,
+      "loss": 0.56,
+      "step": 4686
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.893079830497851,
+      "learning_rate": 2.088211306366455e-06,
+      "loss": 0.4764,
+      "step": 4687
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.499264425402562,
+      "learning_rate": 2.086226034858874e-06,
+      "loss": 0.4563,
+      "step": 4688
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 18.238104301765745,
+      "learning_rate": 2.084241458693476e-06,
+      "loss": 0.4623,
+      "step": 4689
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.503142197456145,
+      "learning_rate": 2.08225757834386e-06,
+      "loss": 0.3976,
+      "step": 4690
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.057851564852268,
+      "learning_rate": 2.080274394283457e-06,
+      "loss": 0.4715,
+      "step": 4691
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.898048866615284,
+      "learning_rate": 2.0782919069855336e-06,
+      "loss": 0.4023,
+      "step": 4692
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.596477407889757,
+      "learning_rate": 2.076310116923188e-06,
+      "loss": 0.4809,
+      "step": 4693
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.451967831340048,
+      "learning_rate": 2.0743290245693586e-06,
+      "loss": 0.4034,
+      "step": 4694
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.600355110987674,
+      "learning_rate": 2.0723486303968104e-06,
+      "loss": 0.4291,
+      "step": 4695
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.61573714106362,
+      "learning_rate": 2.070368934878142e-06,
+      "loss": 0.4572,
+      "step": 4696
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.264408809762894,
+      "learning_rate": 2.0683899384857913e-06,
+      "loss": 0.4349,
+      "step": 4697
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.935668233463539,
+      "learning_rate": 2.066411641692022e-06,
+      "loss": 0.4133,
+      "step": 4698
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.690459249479889,
+      "learning_rate": 2.0644340449689374e-06,
+      "loss": 0.4289,
+      "step": 4699
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.421532254128499,
+      "learning_rate": 2.0624571487884692e-06,
+      "loss": 0.3993,
+      "step": 4700
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.313477050650719,
+      "learning_rate": 2.0604809536223814e-06,
+      "loss": 0.4838,
+      "step": 4701
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.683965706031938,
+      "learning_rate": 2.058505459942276e-06,
+      "loss": 0.4449,
+      "step": 4702
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.534991503362,
+      "learning_rate": 2.056530668219583e-06,
+      "loss": 0.4444,
+      "step": 4703
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.414881440044616,
+      "learning_rate": 2.0545565789255656e-06,
+      "loss": 0.4789,
+      "step": 4704
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.302244049254322,
+      "learning_rate": 2.05258319253132e-06,
+      "loss": 0.3543,
+      "step": 4705
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 12.63342888608853,
+      "learning_rate": 2.050610509507773e-06,
+      "loss": 0.481,
+      "step": 4706
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.080221815325946,
+      "learning_rate": 2.048638530325689e-06,
+      "loss": 0.4668,
+      "step": 4707
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.660873750762749,
+      "learning_rate": 2.046667255455656e-06,
+      "loss": 0.3461,
+      "step": 4708
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.846244225820388,
+      "learning_rate": 2.0446966853681026e-06,
+      "loss": 0.4274,
+      "step": 4709
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.238483083439394,
+      "learning_rate": 2.0427268205332833e-06,
+      "loss": 0.457,
+      "step": 4710
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.306997425363786,
+      "learning_rate": 2.040757661421284e-06,
+      "loss": 0.4035,
+      "step": 4711
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.168392162710345,
+      "learning_rate": 2.038789208502029e-06,
+      "loss": 0.3801,
+      "step": 4712
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.511669164384748,
+      "learning_rate": 2.036821462245266e-06,
+      "loss": 0.4544,
+      "step": 4713
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.6854478787518623,
+      "learning_rate": 2.034854423120578e-06,
+      "loss": 0.4399,
+      "step": 4714
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.5988742302303782,
+      "learning_rate": 2.032888091597379e-06,
+      "loss": 0.4744,
+      "step": 4715
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.8593936874751247,
+      "learning_rate": 2.0309224681449118e-06,
+      "loss": 0.4042,
+      "step": 4716
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 21.22083884726939,
+      "learning_rate": 2.028957553232256e-06,
+      "loss": 0.4897,
+      "step": 4717
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.901116152699715,
+      "learning_rate": 2.0269933473283137e-06,
+      "loss": 0.415,
+      "step": 4718
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.2067091800392715,
+      "learning_rate": 2.025029850901827e-06,
+      "loss": 0.3993,
+      "step": 4719
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.7798294752347745,
+      "learning_rate": 2.023067064421362e-06,
+      "loss": 0.467,
+      "step": 4720
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.366255035940329,
+      "learning_rate": 2.0211049883553162e-06,
+      "loss": 0.4067,
+      "step": 4721
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.9227577268451377,
+      "learning_rate": 2.019143623171922e-06,
+      "loss": 0.4161,
+      "step": 4722
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.10171748518996,
+      "learning_rate": 2.0171829693392353e-06,
+      "loss": 0.4384,
+      "step": 4723
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 12.501246466865606,
+      "learning_rate": 2.01522302732515e-06,
+      "loss": 0.4197,
+      "step": 4724
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.912877247628547,
+      "learning_rate": 2.0132637975973832e-06,
+      "loss": 0.4512,
+      "step": 4725
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3154409210532136,
+      "learning_rate": 2.0113052806234868e-06,
+      "loss": 0.5244,
+      "step": 4726
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.250235380106761,
+      "learning_rate": 2.0093474768708395e-06,
+      "loss": 0.4354,
+      "step": 4727
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.106157840111753,
+      "learning_rate": 2.0073903868066495e-06,
+      "loss": 0.4321,
+      "step": 4728
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.181806441718165,
+      "learning_rate": 2.0054340108979598e-06,
+      "loss": 0.4835,
+      "step": 4729
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.168744364264677,
+      "learning_rate": 2.003478349611637e-06,
+      "loss": 0.3424,
+      "step": 4730
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.076747303519995,
+      "learning_rate": 2.001523403414379e-06,
+      "loss": 0.4094,
+      "step": 4731
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9906013701034028,
+      "learning_rate": 1.999569172772716e-06,
+      "loss": 0.4802,
+      "step": 4732
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.8285734828874447,
+      "learning_rate": 1.997615658153002e-06,
+      "loss": 0.449,
+      "step": 4733
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.8344524000735345,
+      "learning_rate": 1.9956628600214267e-06,
+      "loss": 0.4346,
+      "step": 4734
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 13.735080444926478,
+      "learning_rate": 1.9937107788440035e-06,
+      "loss": 0.3805,
+      "step": 4735
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.264136044196781,
+      "learning_rate": 1.9917594150865744e-06,
+      "loss": 0.3887,
+      "step": 4736
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.9343279246266087,
+      "learning_rate": 1.9898087692148164e-06,
+      "loss": 0.3166,
+      "step": 4737
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.337342481763093,
+      "learning_rate": 1.987858841694229e-06,
+      "loss": 0.399,
+      "step": 4738
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.4241735249813416,
+      "learning_rate": 1.9859096329901422e-06,
+      "loss": 0.3852,
+      "step": 4739
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.7575853911919115,
+      "learning_rate": 1.983961143567715e-06,
+      "loss": 0.4134,
+      "step": 4740
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.446791940639822,
+      "learning_rate": 1.9820133738919324e-06,
+      "loss": 0.409,
+      "step": 4741
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.796601320489815,
+      "learning_rate": 1.980066324427613e-06,
+      "loss": 0.5017,
+      "step": 4742
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.945047528469804,
+      "learning_rate": 1.9781199956393963e-06,
+      "loss": 0.4521,
+      "step": 4743
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.874048737161528,
+      "learning_rate": 1.9761743879917585e-06,
+      "loss": 0.4128,
+      "step": 4744
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9873889417082498,
+      "learning_rate": 1.9742295019489952e-06,
+      "loss": 0.4534,
+      "step": 4745
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.639668549414138,
+      "learning_rate": 1.9722853379752333e-06,
+      "loss": 0.4191,
+      "step": 4746
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.6971203170595976,
+      "learning_rate": 1.9703418965344303e-06,
+      "loss": 0.417,
+      "step": 4747
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 20.39474374729187,
+      "learning_rate": 1.968399178090366e-06,
+      "loss": 0.4511,
+      "step": 4748
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.04790137541763,
+      "learning_rate": 1.966457183106652e-06,
+      "loss": 0.4986,
+      "step": 4749
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.6056859627574367,
+      "learning_rate": 1.9645159120467237e-06,
+      "loss": 0.4548,
+      "step": 4750
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.224561677316631,
+      "learning_rate": 1.962575365373845e-06,
+      "loss": 0.4647,
+      "step": 4751
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.127696206630603,
+      "learning_rate": 1.96063554355111e-06,
+      "loss": 0.4852,
+      "step": 4752
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.88517042410116,
+      "learning_rate": 1.9586964470414342e-06,
+      "loss": 0.4456,
+      "step": 4753
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.66025798061986,
+      "learning_rate": 1.9567580763075666e-06,
+      "loss": 0.4512,
+      "step": 4754
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.417383929690043,
+      "learning_rate": 1.954820431812077e-06,
+      "loss": 0.4331,
+      "step": 4755
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.786966869059034,
+      "learning_rate": 1.9528835140173634e-06,
+      "loss": 0.369,
+      "step": 4756
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.571297753914703,
+      "learning_rate": 1.9509473233856546e-06,
+      "loss": 0.4642,
+      "step": 4757
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.8844154931955295,
+      "learning_rate": 1.9490118603789987e-06,
+      "loss": 0.4878,
+      "step": 4758
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.044764298217436,
+      "learning_rate": 1.9470771254592772e-06,
+      "loss": 0.4965,
+      "step": 4759
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 14.163962241445091,
+      "learning_rate": 1.945143119088195e-06,
+      "loss": 0.4583,
+      "step": 4760
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.51364752690002,
+      "learning_rate": 1.9432098417272797e-06,
+      "loss": 0.465,
+      "step": 4761
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.102156915083765,
+      "learning_rate": 1.941277293837891e-06,
+      "loss": 0.4587,
+      "step": 4762
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.632436543277974,
+      "learning_rate": 1.9393454758812075e-06,
+      "loss": 0.3707,
+      "step": 4763
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1629162868207483,
+      "learning_rate": 1.937414388318243e-06,
+      "loss": 0.4965,
+      "step": 4764
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.2781670246465335,
+      "learning_rate": 1.9354840316098283e-06,
+      "loss": 0.4051,
+      "step": 4765
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 14.332777738099624,
+      "learning_rate": 1.9335544062166227e-06,
+      "loss": 0.4171,
+      "step": 4766
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 21.23823239045441,
+      "learning_rate": 1.9316255125991145e-06,
+      "loss": 0.3867,
+      "step": 4767
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.054718638541873,
+      "learning_rate": 1.929697351217611e-06,
+      "loss": 0.4448,
+      "step": 4768
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2764466745029368,
+      "learning_rate": 1.9277699225322517e-06,
+      "loss": 0.5246,
+      "step": 4769
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 17.999192601220216,
+      "learning_rate": 1.9258432270029947e-06,
+      "loss": 0.4023,
+      "step": 4770
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.10815055361747,
+      "learning_rate": 1.923917265089626e-06,
+      "loss": 0.433,
+      "step": 4771
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.907788718316416,
+      "learning_rate": 1.9219920372517587e-06,
+      "loss": 0.367,
+      "step": 4772
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.395287114142184,
+      "learning_rate": 1.920067543948829e-06,
+      "loss": 0.4652,
+      "step": 4773
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.80292021307846,
+      "learning_rate": 1.918143785640093e-06,
+      "loss": 0.3906,
+      "step": 4774
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.4021567765020295,
+      "learning_rate": 1.91622076278464e-06,
+      "loss": 0.3641,
+      "step": 4775
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.016041904029297,
+      "learning_rate": 1.9142984758413774e-06,
+      "loss": 0.4161,
+      "step": 4776
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.3499910086681997,
+      "learning_rate": 1.912376925269041e-06,
+      "loss": 0.3903,
+      "step": 4777
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.137413852974838,
+      "learning_rate": 1.9104561115261886e-06,
+      "loss": 0.4768,
+      "step": 4778
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 18.316567464908562,
+      "learning_rate": 1.9085360350712007e-06,
+      "loss": 0.3942,
+      "step": 4779
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.117784915751582,
+      "learning_rate": 1.9066166963622867e-06,
+      "loss": 0.4866,
+      "step": 4780
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.983007106790981,
+      "learning_rate": 1.904698095857474e-06,
+      "loss": 0.4503,
+      "step": 4781
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.499873402984663,
+      "learning_rate": 1.9027802340146207e-06,
+      "loss": 0.4329,
+      "step": 4782
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.676540606912331,
+      "learning_rate": 1.9008631112914022e-06,
+      "loss": 0.4497,
+      "step": 4783
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.376980655451133,
+      "learning_rate": 1.8989467281453206e-06,
+      "loss": 0.4122,
+      "step": 4784
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.388283138070941,
+      "learning_rate": 1.8970310850337015e-06,
+      "loss": 0.425,
+      "step": 4785
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.168339183284806,
+      "learning_rate": 1.8951161824136904e-06,
+      "loss": 0.4904,
+      "step": 4786
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.79316188216168,
+      "learning_rate": 1.8932020207422641e-06,
+      "loss": 0.4421,
+      "step": 4787
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.7934992858567425,
+      "learning_rate": 1.8912886004762148e-06,
+      "loss": 0.3934,
+      "step": 4788
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.110173722074461,
+      "learning_rate": 1.8893759220721591e-06,
+      "loss": 0.398,
+      "step": 4789
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.777282508696908,
+      "learning_rate": 1.8874639859865419e-06,
+      "loss": 0.4125,
+      "step": 4790
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 12.09581046049798,
+      "learning_rate": 1.8855527926756228e-06,
+      "loss": 0.3868,
+      "step": 4791
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.297436907399447,
+      "learning_rate": 1.8836423425954925e-06,
+      "loss": 0.4236,
+      "step": 4792
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5108793158951612,
+      "learning_rate": 1.8817326362020588e-06,
+      "loss": 0.405,
+      "step": 4793
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.250977099788029,
+      "learning_rate": 1.8798236739510505e-06,
+      "loss": 0.4438,
+      "step": 4794
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 11.346898801866713,
+      "learning_rate": 1.8779154562980266e-06,
+      "loss": 0.4057,
+      "step": 4795
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.871366371493626,
+      "learning_rate": 1.8760079836983613e-06,
+      "loss": 0.4721,
+      "step": 4796
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.471619437773189,
+      "learning_rate": 1.8741012566072537e-06,
+      "loss": 0.4488,
+      "step": 4797
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.7314066834232316,
+      "learning_rate": 1.872195275479724e-06,
+      "loss": 0.4699,
+      "step": 4798
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.2041254733788405,
+      "learning_rate": 1.8702900407706136e-06,
+      "loss": 0.4332,
+      "step": 4799
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.20921586563363,
+      "learning_rate": 1.8683855529345902e-06,
+      "loss": 0.4201,
+      "step": 4800
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.4733403520552235,
+      "learning_rate": 1.8664818124261375e-06,
+      "loss": 0.4338,
+      "step": 4801
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.848958990117113,
+      "learning_rate": 1.8645788196995669e-06,
+      "loss": 0.4331,
+      "step": 4802
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.765926051250804,
+      "learning_rate": 1.8626765752090058e-06,
+      "loss": 0.4234,
+      "step": 4803
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.8996676426468473,
+      "learning_rate": 1.8607750794084039e-06,
+      "loss": 0.441,
+      "step": 4804
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.764143348480458,
+      "learning_rate": 1.858874332751537e-06,
+      "loss": 0.397,
+      "step": 4805
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.842235171457089,
+      "learning_rate": 1.8569743356919973e-06,
+      "loss": 0.4874,
+      "step": 4806
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.893625141478686,
+      "learning_rate": 1.8550750886831987e-06,
+      "loss": 0.3924,
+      "step": 4807
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.680053428108346,
+      "learning_rate": 1.8531765921783778e-06,
+      "loss": 0.4088,
+      "step": 4808
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.1158080825436985,
+      "learning_rate": 1.8512788466305899e-06,
+      "loss": 0.3828,
+      "step": 4809
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.916174875721781,
+      "learning_rate": 1.8493818524927143e-06,
+      "loss": 0.4532,
+      "step": 4810
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 18.113534331027083,
+      "learning_rate": 1.8474856102174471e-06,
+      "loss": 0.4893,
+      "step": 4811
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.054129068271102,
+      "learning_rate": 1.84559012025731e-06,
+      "loss": 0.3585,
+      "step": 4812
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.695444230783209,
+      "learning_rate": 1.843695383064641e-06,
+      "loss": 0.4027,
+      "step": 4813
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.753104053450278,
+      "learning_rate": 1.8418013990915966e-06,
+      "loss": 0.3926,
+      "step": 4814
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.585469979832173,
+      "learning_rate": 1.839908168790161e-06,
+      "loss": 0.4142,
+      "step": 4815
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.043318930118474,
+      "learning_rate": 1.8380156926121312e-06,
+      "loss": 0.4595,
+      "step": 4816
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.875248160862492,
+      "learning_rate": 1.8361239710091294e-06,
+      "loss": 0.4641,
+      "step": 4817
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9573754336354109,
+      "learning_rate": 1.834233004432594e-06,
+      "loss": 0.4947,
+      "step": 4818
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.896341779314252,
+      "learning_rate": 1.8323427933337855e-06,
+      "loss": 0.404,
+      "step": 4819
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 19.914681773464604,
+      "learning_rate": 1.830453338163783e-06,
+      "loss": 0.4253,
+      "step": 4820
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.844182332281133,
+      "learning_rate": 1.828564639373483e-06,
+      "loss": 0.4165,
+      "step": 4821
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.249741511633091,
+      "learning_rate": 1.8266766974136086e-06,
+      "loss": 0.3703,
+      "step": 4822
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.7952526084534,
+      "learning_rate": 1.824789512734696e-06,
+      "loss": 0.4048,
+      "step": 4823
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.329583271609569,
+      "learning_rate": 1.8229030857870995e-06,
+      "loss": 0.4291,
+      "step": 4824
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.424959433592154,
+      "learning_rate": 1.8210174170210005e-06,
+      "loss": 0.4428,
+      "step": 4825
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.8269199898893205,
+      "learning_rate": 1.8191325068863897e-06,
+      "loss": 0.4203,
+      "step": 4826
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.989762975483705,
+      "learning_rate": 1.817248355833085e-06,
+      "loss": 0.4354,
+      "step": 4827
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.670025855796599,
+      "learning_rate": 1.815364964310719e-06,
+      "loss": 0.4262,
+      "step": 4828
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.372761417585162,
+      "learning_rate": 1.8134823327687412e-06,
+      "loss": 0.4699,
+      "step": 4829
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.809787111876093,
+      "learning_rate": 1.8116004616564264e-06,
+      "loss": 0.4748,
+      "step": 4830
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.421935693195781,
+      "learning_rate": 1.809719351422861e-06,
+      "loss": 0.4401,
+      "step": 4831
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.311108126361975,
+      "learning_rate": 1.8078390025169534e-06,
+      "loss": 0.4088,
+      "step": 4832
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.2955653274612215,
+      "learning_rate": 1.8059594153874288e-06,
+      "loss": 0.4529,
+      "step": 4833
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.293111965725443,
+      "learning_rate": 1.8040805904828307e-06,
+      "loss": 0.3747,
+      "step": 4834
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.522337220497181,
+      "learning_rate": 1.8022025282515238e-06,
+      "loss": 0.4361,
+      "step": 4835
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.979629153985799,
+      "learning_rate": 1.8003252291416856e-06,
+      "loss": 0.3609,
+      "step": 4836
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.896113590177915,
+      "learning_rate": 1.7984486936013169e-06,
+      "loss": 0.4897,
+      "step": 4837
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.0867163934319795,
+      "learning_rate": 1.7965729220782325e-06,
+      "loss": 0.3431,
+      "step": 4838
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.891413896978916,
+      "learning_rate": 1.7946979150200638e-06,
+      "loss": 0.3887,
+      "step": 4839
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 23.492161178521677,
+      "learning_rate": 1.792823672874266e-06,
+      "loss": 0.444,
+      "step": 4840
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.968469908086001,
+      "learning_rate": 1.7909501960881055e-06,
+      "loss": 0.4346,
+      "step": 4841
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.057792926634205,
+      "learning_rate": 1.7890774851086685e-06,
+      "loss": 0.3707,
+      "step": 4842
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.0387093934563225,
+      "learning_rate": 1.787205540382858e-06,
+      "loss": 0.4249,
+      "step": 4843
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2843482238104174,
+      "learning_rate": 1.7853343623573926e-06,
+      "loss": 0.531,
+      "step": 4844
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.551262326249063,
+      "learning_rate": 1.7834639514788137e-06,
+      "loss": 0.4359,
+      "step": 4845
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0251030905347258,
+      "learning_rate": 1.7815943081934711e-06,
+      "loss": 0.4867,
+      "step": 4846
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2468523902854127,
+      "learning_rate": 1.7797254329475406e-06,
+      "loss": 0.5672,
+      "step": 4847
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.934355350139679,
+      "learning_rate": 1.7778573261870075e-06,
+      "loss": 0.4741,
+      "step": 4848
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.150816999915261,
+      "learning_rate": 1.7759899883576748e-06,
+      "loss": 0.4424,
+      "step": 4849
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.839801670311012,
+      "learning_rate": 1.7741234199051665e-06,
+      "loss": 0.3926,
+      "step": 4850
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.607223532359424,
+      "learning_rate": 1.7722576212749175e-06,
+      "loss": 0.4142,
+      "step": 4851
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.059926435240088,
+      "learning_rate": 1.770392592912184e-06,
+      "loss": 0.372,
+      "step": 4852
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.545666886737184,
+      "learning_rate": 1.7685283352620342e-06,
+      "loss": 0.4919,
+      "step": 4853
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.0488808107560805,
+      "learning_rate": 1.7666648487693544e-06,
+      "loss": 0.4825,
+      "step": 4854
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.1692074926481,
+      "learning_rate": 1.764802133878847e-06,
+      "loss": 0.4811,
+      "step": 4855
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.920473605852468,
+      "learning_rate": 1.7629401910350264e-06,
+      "loss": 0.4232,
+      "step": 4856
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.767986532710334,
+      "learning_rate": 1.7610790206822303e-06,
+      "loss": 0.3185,
+      "step": 4857
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.979722557519594,
+      "learning_rate": 1.7592186232646069e-06,
+      "loss": 0.4315,
+      "step": 4858
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.8077919084509935,
+      "learning_rate": 1.7573589992261185e-06,
+      "loss": 0.417,
+      "step": 4859
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 11.594822457211778,
+      "learning_rate": 1.755500149010549e-06,
+      "loss": 0.3718,
+      "step": 4860
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5021175619209115,
+      "learning_rate": 1.75364207306149e-06,
+      "loss": 0.5242,
+      "step": 4861
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 28.36490192142048,
+      "learning_rate": 1.7517847718223557e-06,
+      "loss": 0.4318,
+      "step": 4862
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 16.472748306922306,
+      "learning_rate": 1.7499282457363702e-06,
+      "loss": 0.463,
+      "step": 4863
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.323775283283541,
+      "learning_rate": 1.7480724952465733e-06,
+      "loss": 0.4751,
+      "step": 4864
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.40553883348144,
+      "learning_rate": 1.7462175207958238e-06,
+      "loss": 0.4378,
+      "step": 4865
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.00406488243665,
+      "learning_rate": 1.7443633228267898e-06,
+      "loss": 0.4043,
+      "step": 4866
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.197980994180602,
+      "learning_rate": 1.742509901781958e-06,
+      "loss": 0.4632,
+      "step": 4867
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.178142734831738,
+      "learning_rate": 1.740657258103627e-06,
+      "loss": 0.4512,
+      "step": 4868
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.118999819289833,
+      "learning_rate": 1.7388053922339103e-06,
+      "loss": 0.5419,
+      "step": 4869
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.840371729404817,
+      "learning_rate": 1.7369543046147398e-06,
+      "loss": 0.4184,
+      "step": 4870
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.533382509410105,
+      "learning_rate": 1.7351039956878546e-06,
+      "loss": 0.3232,
+      "step": 4871
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.302847347435059,
+      "learning_rate": 1.7332544658948159e-06,
+      "loss": 0.4842,
+      "step": 4872
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.041976038676482,
+      "learning_rate": 1.7314057156769925e-06,
+      "loss": 0.4303,
+      "step": 4873
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.571689330323869,
+      "learning_rate": 1.729557745475569e-06,
+      "loss": 0.4163,
+      "step": 4874
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.224163560404418,
+      "learning_rate": 1.7277105557315477e-06,
+      "loss": 0.4514,
+      "step": 4875
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 17.02968199332374,
+      "learning_rate": 1.725864146885739e-06,
+      "loss": 0.4865,
+      "step": 4876
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.9187941838507596,
+      "learning_rate": 1.7240185193787701e-06,
+      "loss": 0.3722,
+      "step": 4877
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.39109024770403,
+      "learning_rate": 1.7221736736510802e-06,
+      "loss": 0.4453,
+      "step": 4878
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 16.67848950841273,
+      "learning_rate": 1.7203296101429217e-06,
+      "loss": 0.3983,
+      "step": 4879
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.892557021435531,
+      "learning_rate": 1.718486329294365e-06,
+      "loss": 0.5005,
+      "step": 4880
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.117090748996902,
+      "learning_rate": 1.716643831545286e-06,
+      "loss": 0.4444,
+      "step": 4881
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.718355933694574,
+      "learning_rate": 1.7148021173353819e-06,
+      "loss": 0.3732,
+      "step": 4882
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.1676876210267,
+      "learning_rate": 1.7129611871041563e-06,
+      "loss": 0.4147,
+      "step": 4883
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.088968575755412,
+      "learning_rate": 1.7111210412909274e-06,
+      "loss": 0.5055,
+      "step": 4884
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.187910030590789,
+      "learning_rate": 1.7092816803348306e-06,
+      "loss": 0.469,
+      "step": 4885
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.215746213669689,
+      "learning_rate": 1.7074431046748075e-06,
+      "loss": 0.404,
+      "step": 4886
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.966191084117976,
+      "learning_rate": 1.7056053147496149e-06,
+      "loss": 0.4984,
+      "step": 4887
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.789996401071698,
+      "learning_rate": 1.7037683109978253e-06,
+      "loss": 0.4302,
+      "step": 4888
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.661061526998839,
+      "learning_rate": 1.7019320938578189e-06,
+      "loss": 0.4155,
+      "step": 4889
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.186022501292075,
+      "learning_rate": 1.7000966637677902e-06,
+      "loss": 0.3907,
+      "step": 4890
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.010487626288365,
+      "learning_rate": 1.6982620211657464e-06,
+      "loss": 0.4232,
+      "step": 4891
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.195608680303167,
+      "learning_rate": 1.6964281664895027e-06,
+      "loss": 0.4556,
+      "step": 4892
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.663817486152631,
+      "learning_rate": 1.6945951001766947e-06,
+      "loss": 0.4809,
+      "step": 4893
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.954266927858951,
+      "learning_rate": 1.6927628226647607e-06,
+      "loss": 0.5132,
+      "step": 4894
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.641869878517609,
+      "learning_rate": 1.6909313343909584e-06,
+      "loss": 0.4287,
+      "step": 4895
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.803206334854586,
+      "learning_rate": 1.6891006357923522e-06,
+      "loss": 0.3854,
+      "step": 4896
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 18.952405855327473,
+      "learning_rate": 1.687270727305817e-06,
+      "loss": 0.4007,
+      "step": 4897
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 12.207564145502667,
+      "learning_rate": 1.6854416093680458e-06,
+      "loss": 0.4533,
+      "step": 4898
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.330269943490551,
+      "learning_rate": 1.6836132824155355e-06,
+      "loss": 0.512,
+      "step": 4899
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.348833866728041,
+      "learning_rate": 1.6817857468846e-06,
+      "loss": 0.397,
+      "step": 4900
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.423152856202949,
+      "learning_rate": 1.6799590032113606e-06,
+      "loss": 0.4674,
+      "step": 4901
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.1333136289443475,
+      "learning_rate": 1.6781330518317512e-06,
+      "loss": 0.4347,
+      "step": 4902
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.823223427455497,
+      "learning_rate": 1.6763078931815159e-06,
+      "loss": 0.3793,
+      "step": 4903
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.455143246369038,
+      "learning_rate": 1.674483527696209e-06,
+      "loss": 0.4417,
+      "step": 4904
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.247116094823998,
+      "learning_rate": 1.6726599558111988e-06,
+      "loss": 0.4591,
+      "step": 4905
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.2975895748970405,
+      "learning_rate": 1.6708371779616617e-06,
+      "loss": 0.4598,
+      "step": 4906
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.650598619127493,
+      "learning_rate": 1.6690151945825817e-06,
+      "loss": 0.4714,
+      "step": 4907
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.184568054265106,
+      "learning_rate": 1.6671940061087611e-06,
+      "loss": 0.4474,
+      "step": 4908
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.388266695649065,
+      "learning_rate": 1.6653736129748038e-06,
+      "loss": 0.4171,
+      "step": 4909
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 18.78637382925074,
+      "learning_rate": 1.6635540156151308e-06,
+      "loss": 0.4044,
+      "step": 4910
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.804404357101728,
+      "learning_rate": 1.6617352144639692e-06,
+      "loss": 0.4127,
+      "step": 4911
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.298417331355293,
+      "learning_rate": 1.6599172099553573e-06,
+      "loss": 0.5037,
+      "step": 4912
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 22.9245769578651,
+      "learning_rate": 1.658100002523143e-06,
+      "loss": 0.4455,
+      "step": 4913
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.128143686680953,
+      "learning_rate": 1.6562835926009823e-06,
+      "loss": 0.5221,
+      "step": 4914
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.87987630690349,
+      "learning_rate": 1.6544679806223468e-06,
+      "loss": 0.4389,
+      "step": 4915
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.997562544096999,
+      "learning_rate": 1.652653167020511e-06,
+      "loss": 0.5207,
+      "step": 4916
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.6756898675156,
+      "learning_rate": 1.6508391522285598e-06,
+      "loss": 0.3716,
+      "step": 4917
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 11.629126002075862,
+      "learning_rate": 1.6490259366793931e-06,
+      "loss": 0.4882,
+      "step": 4918
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.511961143783831,
+      "learning_rate": 1.6472135208057128e-06,
+      "loss": 0.4109,
+      "step": 4919
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.787115792259728,
+      "learning_rate": 1.6454019050400367e-06,
+      "loss": 0.4362,
+      "step": 4920
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 17.995833660736523,
+      "learning_rate": 1.6435910898146862e-06,
+      "loss": 0.3676,
+      "step": 4921
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.40433157762816,
+      "learning_rate": 1.6417810755617918e-06,
+      "loss": 0.3838,
+      "step": 4922
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.033522241669954,
+      "learning_rate": 1.6399718627132987e-06,
+      "loss": 0.4589,
+      "step": 4923
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.0192212072313405,
+      "learning_rate": 1.6381634517009555e-06,
+      "loss": 0.4903,
+      "step": 4924
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.480625025617363,
+      "learning_rate": 1.6363558429563197e-06,
+      "loss": 0.3969,
+      "step": 4925
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.098339693090555,
+      "learning_rate": 1.6345490369107603e-06,
+      "loss": 0.4634,
+      "step": 4926
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 11.193848370471887,
+      "learning_rate": 1.6327430339954498e-06,
+      "loss": 0.4342,
+      "step": 4927
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 52.726233316108406,
+      "learning_rate": 1.6309378346413763e-06,
+      "loss": 0.3572,
+      "step": 4928
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.340193376500938,
+      "learning_rate": 1.6291334392793278e-06,
+      "loss": 0.4048,
+      "step": 4929
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.051044141667209,
+      "learning_rate": 1.6273298483399092e-06,
+      "loss": 0.3871,
+      "step": 4930
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 26.84123581530738,
+      "learning_rate": 1.6255270622535268e-06,
+      "loss": 0.4584,
+      "step": 4931
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 15.725892856004375,
+      "learning_rate": 1.6237250814503952e-06,
+      "loss": 0.3978,
+      "step": 4932
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.428397990579211,
+      "learning_rate": 1.6219239063605425e-06,
+      "loss": 0.413,
+      "step": 4933
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.529300400335759,
+      "learning_rate": 1.6201235374137975e-06,
+      "loss": 0.4498,
+      "step": 4934
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.814173658846859,
+      "learning_rate": 1.618323975039801e-06,
+      "loss": 0.4679,
+      "step": 4935
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 13.615970551633417,
+      "learning_rate": 1.6165252196679998e-06,
+      "loss": 0.4909,
+      "step": 4936
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.1922503619048,
+      "learning_rate": 1.614727271727647e-06,
+      "loss": 0.426,
+      "step": 4937
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.108320072326528,
+      "learning_rate": 1.612930131647807e-06,
+      "loss": 0.3736,
+      "step": 4938
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.12896918113066,
+      "learning_rate": 1.6111337998573456e-06,
+      "loss": 0.5228,
+      "step": 4939
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2678537297782337,
+      "learning_rate": 1.6093382767849424e-06,
+      "loss": 0.5425,
+      "step": 4940
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 12.649758765802616,
+      "learning_rate": 1.6075435628590785e-06,
+      "loss": 0.3955,
+      "step": 4941
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.478793926993321,
+      "learning_rate": 1.6057496585080434e-06,
+      "loss": 0.4447,
+      "step": 4942
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.320495025362742,
+      "learning_rate": 1.603956564159936e-06,
+      "loss": 0.4795,
+      "step": 4943
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.116434211852291,
+      "learning_rate": 1.6021642802426568e-06,
+      "loss": 0.4213,
+      "step": 4944
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 27.316673231305767,
+      "learning_rate": 1.6003728071839198e-06,
+      "loss": 0.3771,
+      "step": 4945
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 34.85074357696168,
+      "learning_rate": 1.598582145411239e-06,
+      "loss": 0.4135,
+      "step": 4946
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.0812517510222825,
+      "learning_rate": 1.5967922953519382e-06,
+      "loss": 0.4112,
+      "step": 4947
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.6985521819593163,
+      "learning_rate": 1.5950032574331464e-06,
+      "loss": 0.3631,
+      "step": 4948
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 11.654332732581707,
+      "learning_rate": 1.593215032081798e-06,
+      "loss": 0.4314,
+      "step": 4949
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.00186330803566,
+      "learning_rate": 1.5914276197246365e-06,
+      "loss": 0.397,
+      "step": 4950
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.64873041810846,
+      "learning_rate": 1.589641020788209e-06,
+      "loss": 0.3741,
+      "step": 4951
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.434797175090717,
+      "learning_rate": 1.5878552356988674e-06,
+      "loss": 0.4011,
+      "step": 4952
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.6503700571785704,
+      "learning_rate": 1.5860702648827736e-06,
+      "loss": 0.4605,
+      "step": 4953
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.212988445695651,
+      "learning_rate": 1.5842861087658895e-06,
+      "loss": 0.4288,
+      "step": 4954
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.8616210004920544,
+      "learning_rate": 1.5825027677739891e-06,
+      "loss": 0.4097,
+      "step": 4955
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.879344358693332,
+      "learning_rate": 1.5807202423326463e-06,
+      "loss": 0.4339,
+      "step": 4956
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.488216598804553,
+      "learning_rate": 1.5789385328672407e-06,
+      "loss": 0.4498,
+      "step": 4957
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.873919797969425,
+      "learning_rate": 1.5771576398029625e-06,
+      "loss": 0.4168,
+      "step": 4958
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.826562207919402,
+      "learning_rate": 1.5753775635648022e-06,
+      "loss": 0.4974,
+      "step": 4959
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 17.696820792300283,
+      "learning_rate": 1.5735983045775567e-06,
+      "loss": 0.4892,
+      "step": 4960
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 13.478672577520692,
+      "learning_rate": 1.5718198632658272e-06,
+      "loss": 0.4541,
+      "step": 4961
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.68228939997883,
+      "learning_rate": 1.5700422400540194e-06,
+      "loss": 0.4658,
+      "step": 4962
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.772958624788832,
+      "learning_rate": 1.5682654353663474e-06,
+      "loss": 0.4275,
+      "step": 4963
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 11.524743331608585,
+      "learning_rate": 1.5664894496268246e-06,
+      "loss": 0.428,
+      "step": 4964
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.532414852531679,
+      "learning_rate": 1.564714283259275e-06,
+      "loss": 0.4226,
+      "step": 4965
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.58502843547674,
+      "learning_rate": 1.562939936687322e-06,
+      "loss": 0.4098,
+      "step": 4966
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.281653260809573,
+      "learning_rate": 1.5611664103343933e-06,
+      "loss": 0.4738,
+      "step": 4967
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.0186521786628306,
+      "learning_rate": 1.5593937046237268e-06,
+      "loss": 0.4278,
+      "step": 4968
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.664192832521856,
+      "learning_rate": 1.5576218199783578e-06,
+      "loss": 0.418,
+      "step": 4969
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.110019828106071,
+      "learning_rate": 1.5558507568211289e-06,
+      "loss": 0.5295,
+      "step": 4970
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.346694597508153,
+      "learning_rate": 1.5540805155746863e-06,
+      "loss": 0.4122,
+      "step": 4971
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.4587951647396435,
+      "learning_rate": 1.5523110966614778e-06,
+      "loss": 0.4142,
+      "step": 4972
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 18.4464459571236,
+      "learning_rate": 1.5505425005037606e-06,
+      "loss": 0.5019,
+      "step": 4973
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.433047318968047,
+      "learning_rate": 1.5487747275235888e-06,
+      "loss": 0.4582,
+      "step": 4974
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.583302793967848,
+      "learning_rate": 1.5470077781428254e-06,
+      "loss": 0.4551,
+      "step": 4975
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.920971170604424,
+      "learning_rate": 1.5452416527831348e-06,
+      "loss": 0.396,
+      "step": 4976
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 15.03584059436499,
+      "learning_rate": 1.5434763518659817e-06,
+      "loss": 0.4102,
+      "step": 4977
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.455199832648352,
+      "learning_rate": 1.5417118758126408e-06,
+      "loss": 0.5157,
+      "step": 4978
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3688125223528294,
+      "learning_rate": 1.5399482250441827e-06,
+      "loss": 0.5449,
+      "step": 4979
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.913251830182728,
+      "learning_rate": 1.5381853999814882e-06,
+      "loss": 0.4201,
+      "step": 4980
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.102867226372371,
+      "learning_rate": 1.5364234010452355e-06,
+      "loss": 0.4513,
+      "step": 4981
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.336885961993578,
+      "learning_rate": 1.534662228655907e-06,
+      "loss": 0.4167,
+      "step": 4982
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.660676716155712,
+      "learning_rate": 1.5329018832337895e-06,
+      "loss": 0.4492,
+      "step": 4983
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.574517719236944,
+      "learning_rate": 1.5311423651989681e-06,
+      "loss": 0.449,
+      "step": 4984
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.337382435500911,
+      "learning_rate": 1.5293836749713386e-06,
+      "loss": 0.4748,
+      "step": 4985
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.953432348233978,
+      "learning_rate": 1.5276258129705918e-06,
+      "loss": 0.403,
+      "step": 4986
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.968175989762074,
+      "learning_rate": 1.5258687796162213e-06,
+      "loss": 0.4911,
+      "step": 4987
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 17.076921614142,
+      "learning_rate": 1.5241125753275294e-06,
+      "loss": 0.4593,
+      "step": 4988
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 21.29709436044458,
+      "learning_rate": 1.5223572005236127e-06,
+      "loss": 0.3512,
+      "step": 4989
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.886619188293746,
+      "learning_rate": 1.5206026556233761e-06,
+      "loss": 0.4307,
+      "step": 4990
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.704069095500053,
+      "learning_rate": 1.518848941045522e-06,
+      "loss": 0.4009,
+      "step": 4991
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 34.07310361399107,
+      "learning_rate": 1.5170960572085552e-06,
+      "loss": 0.3855,
+      "step": 4992
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.31798468522245,
+      "learning_rate": 1.5153440045307865e-06,
+      "loss": 0.4079,
+      "step": 4993
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.825764117579442,
+      "learning_rate": 1.5135927834303238e-06,
+      "loss": 0.4062,
+      "step": 4994
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.7469029121473705,
+      "learning_rate": 1.511842394325077e-06,
+      "loss": 0.4039,
+      "step": 4995
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.482634405815571,
+      "learning_rate": 1.51009283763276e-06,
+      "loss": 0.3795,
+      "step": 4996
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.924154988394382,
+      "learning_rate": 1.5083441137708837e-06,
+      "loss": 0.3926,
+      "step": 4997
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.011807234411183,
+      "learning_rate": 1.506596223156767e-06,
+      "loss": 0.4554,
+      "step": 4998
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2554318549154682,
+      "learning_rate": 1.5048491662075244e-06,
+      "loss": 0.5372,
+      "step": 4999
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.727035800579789,
+      "learning_rate": 1.5031029433400706e-06,
+      "loss": 0.4526,
+      "step": 5000
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9668285819469707,
+      "learning_rate": 1.5013575549711274e-06,
+      "loss": 0.4905,
+      "step": 5001
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.305456759648467,
+      "learning_rate": 1.499613001517211e-06,
+      "loss": 0.4602,
+      "step": 5002
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.64271793865173,
+      "learning_rate": 1.497869283394644e-06,
+      "loss": 0.3659,
+      "step": 5003
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.469227620734533,
+      "learning_rate": 1.4961264010195454e-06,
+      "loss": 0.4103,
+      "step": 5004
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.257764161899248,
+      "learning_rate": 1.4943843548078353e-06,
+      "loss": 0.4562,
+      "step": 5005
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.573707041627797,
+      "learning_rate": 1.4926431451752361e-06,
+      "loss": 0.4107,
+      "step": 5006
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.692833025163236,
+      "learning_rate": 1.4909027725372676e-06,
+      "loss": 0.4144,
+      "step": 5007
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1615584357344928,
+      "learning_rate": 1.4891632373092551e-06,
+      "loss": 0.5538,
+      "step": 5008
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 14.037220153784945,
+      "learning_rate": 1.4874245399063197e-06,
+      "loss": 0.4295,
+      "step": 5009
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.991245378434636,
+      "learning_rate": 1.4856866807433807e-06,
+      "loss": 0.4175,
+      "step": 5010
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.047828441359134,
+      "learning_rate": 1.483949660235165e-06,
+      "loss": 0.4211,
+      "step": 5011
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.10353848545864,
+      "learning_rate": 1.48221347879619e-06,
+      "loss": 0.3995,
+      "step": 5012
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.078817742004709,
+      "learning_rate": 1.4804781368407822e-06,
+      "loss": 0.4085,
+      "step": 5013
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.93248203287576,
+      "learning_rate": 1.4787436347830602e-06,
+      "loss": 0.4385,
+      "step": 5014
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.2386796232560195,
+      "learning_rate": 1.4770099730369447e-06,
+      "loss": 0.4073,
+      "step": 5015
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.731620232758024,
+      "learning_rate": 1.4752771520161584e-06,
+      "loss": 0.4584,
+      "step": 5016
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.705219357918715,
+      "learning_rate": 1.4735451721342203e-06,
+      "loss": 0.4428,
+      "step": 5017
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.374967374081161,
+      "learning_rate": 1.471814033804449e-06,
+      "loss": 0.4582,
+      "step": 5018
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.57975153571692,
+      "learning_rate": 1.4700837374399634e-06,
+      "loss": 0.4491,
+      "step": 5019
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.067021484946328,
+      "learning_rate": 1.468354283453679e-06,
+      "loss": 0.4487,
+      "step": 5020
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.771526750438657,
+      "learning_rate": 1.466625672258316e-06,
+      "loss": 0.4155,
+      "step": 5021
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.463939821230833,
+      "learning_rate": 1.4648979042663853e-06,
+      "loss": 0.4636,
+      "step": 5022
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 11.856508054285326,
+      "learning_rate": 1.4631709798902055e-06,
+      "loss": 0.4192,
+      "step": 5023
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.846500647665819,
+      "learning_rate": 1.4614448995418868e-06,
+      "loss": 0.4378,
+      "step": 5024
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.0311535692938625,
+      "learning_rate": 1.4597196636333389e-06,
+      "loss": 0.3613,
+      "step": 5025
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2251153346105808,
+      "learning_rate": 1.4579952725762753e-06,
+      "loss": 0.5423,
+      "step": 5026
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.890956368901254,
+      "learning_rate": 1.4562717267822013e-06,
+      "loss": 0.4588,
+      "step": 5027
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.421942934626455,
+      "learning_rate": 1.454549026662425e-06,
+      "loss": 0.4605,
+      "step": 5028
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 19.777961898317834,
+      "learning_rate": 1.4528271726280518e-06,
+      "loss": 0.4693,
+      "step": 5029
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.164174319310765,
+      "learning_rate": 1.4511061650899822e-06,
+      "loss": 0.4213,
+      "step": 5030
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 181.86978784583428,
+      "learning_rate": 1.4493860044589187e-06,
+      "loss": 0.4255,
+      "step": 5031
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.21550242859919,
+      "learning_rate": 1.447666691145357e-06,
+      "loss": 0.4102,
+      "step": 5032
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.418637999509029,
+      "learning_rate": 1.4459482255595975e-06,
+      "loss": 0.4263,
+      "step": 5033
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.003351528918016,
+      "learning_rate": 1.4442306081117325e-06,
+      "loss": 0.4699,
+      "step": 5034
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.40828557545126,
+      "learning_rate": 1.4425138392116516e-06,
+      "loss": 0.4436,
+      "step": 5035
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.154325875127878,
+      "learning_rate": 1.4407979192690486e-06,
+      "loss": 0.3938,
+      "step": 5036
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 14.159195327938118,
+      "learning_rate": 1.439082848693406e-06,
+      "loss": 0.4731,
+      "step": 5037
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.848309158793154,
+      "learning_rate": 1.4373686278940107e-06,
+      "loss": 0.4148,
+      "step": 5038
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 14.440524526535384,
+      "learning_rate": 1.4356552572799432e-06,
+      "loss": 0.3897,
+      "step": 5039
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.117808057537764,
+      "learning_rate": 1.4339427372600811e-06,
+      "loss": 0.3786,
+      "step": 5040
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.339388721116429,
+      "learning_rate": 1.4322310682430996e-06,
+      "loss": 0.454,
+      "step": 5041
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.8734836291596215,
+      "learning_rate": 1.4305202506374695e-06,
+      "loss": 0.4473,
+      "step": 5042
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.825666242375968,
+      "learning_rate": 1.428810284851463e-06,
+      "loss": 0.482,
+      "step": 5043
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 55.66188834973104,
+      "learning_rate": 1.4271011712931443e-06,
+      "loss": 0.4147,
+      "step": 5044
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.982597282498034,
+      "learning_rate": 1.4253929103703745e-06,
+      "loss": 0.3752,
+      "step": 5045
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 12.134829373252874,
+      "learning_rate": 1.4236855024908153e-06,
+      "loss": 0.4382,
+      "step": 5046
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.638815280767644,
+      "learning_rate": 1.4219789480619183e-06,
+      "loss": 0.4752,
+      "step": 5047
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.848971597262186,
+      "learning_rate": 1.4202732474909397e-06,
+      "loss": 0.4162,
+      "step": 5048
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.618388263151876,
+      "learning_rate": 1.4185684011849243e-06,
+      "loss": 0.4437,
+      "step": 5049
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 11.510118289733407,
+      "learning_rate": 1.4168644095507155e-06,
+      "loss": 0.4332,
+      "step": 5050
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0487746346455564,
+      "learning_rate": 1.4151612729949566e-06,
+      "loss": 0.4938,
+      "step": 5051
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.430452058097448,
+      "learning_rate": 1.4134589919240815e-06,
+      "loss": 0.4029,
+      "step": 5052
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.594057396851826,
+      "learning_rate": 1.4117575667443224e-06,
+      "loss": 0.4821,
+      "step": 5053
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.324014851793637,
+      "learning_rate": 1.4100569978617068e-06,
+      "loss": 0.3887,
+      "step": 5054
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 42.432809571795296,
+      "learning_rate": 1.4083572856820564e-06,
+      "loss": 0.4324,
+      "step": 5055
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.378214528511265,
+      "learning_rate": 1.4066584306109937e-06,
+      "loss": 0.4643,
+      "step": 5056
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.132300953441496,
+      "learning_rate": 1.4049604330539285e-06,
+      "loss": 0.3994,
+      "step": 5057
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.704928136765426,
+      "learning_rate": 1.403263293416075e-06,
+      "loss": 0.5244,
+      "step": 5058
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.865145634010185,
+      "learning_rate": 1.4015670121024355e-06,
+      "loss": 0.4691,
+      "step": 5059
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.873181629234607,
+      "learning_rate": 1.3998715895178093e-06,
+      "loss": 0.445,
+      "step": 5060
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 28.163526929394067,
+      "learning_rate": 1.3981770260667948e-06,
+      "loss": 0.4537,
+      "step": 5061
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.531773599097027,
+      "learning_rate": 1.3964833221537794e-06,
+      "loss": 0.4047,
+      "step": 5062
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.100363959231573,
+      "learning_rate": 1.3947904781829497e-06,
+      "loss": 0.4087,
+      "step": 5063
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.537724691517257,
+      "learning_rate": 1.3930984945582837e-06,
+      "loss": 0.4265,
+      "step": 5064
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.603927719488869,
+      "learning_rate": 1.3914073716835557e-06,
+      "loss": 0.4516,
+      "step": 5065
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.544240876025431,
+      "learning_rate": 1.389717109962338e-06,
+      "loss": 0.4305,
+      "step": 5066
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.480323024762036,
+      "learning_rate": 1.3880277097979899e-06,
+      "loss": 0.3711,
+      "step": 5067
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.702112300801673,
+      "learning_rate": 1.3863391715936736e-06,
+      "loss": 0.4476,
+      "step": 5068
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1843117534125258,
+      "learning_rate": 1.3846514957523389e-06,
+      "loss": 0.5532,
+      "step": 5069
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.024133521290889,
+      "learning_rate": 1.3829646826767312e-06,
+      "loss": 0.5096,
+      "step": 5070
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.677878441041741,
+      "learning_rate": 1.3812787327693945e-06,
+      "loss": 0.435,
+      "step": 5071
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.956339524483958,
+      "learning_rate": 1.37959364643266e-06,
+      "loss": 0.4537,
+      "step": 5072
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 27.865882478510272,
+      "learning_rate": 1.377909424068659e-06,
+      "loss": 0.4076,
+      "step": 5073
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.615725433438222,
+      "learning_rate": 1.3762260660793125e-06,
+      "loss": 0.4513,
+      "step": 5074
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.99264071019721,
+      "learning_rate": 1.374543572866337e-06,
+      "loss": 0.4499,
+      "step": 5075
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.101540467122868,
+      "learning_rate": 1.3728619448312413e-06,
+      "loss": 0.4315,
+      "step": 5076
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.209007994735339,
+      "learning_rate": 1.3711811823753286e-06,
+      "loss": 0.381,
+      "step": 5077
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.545759366064242,
+      "learning_rate": 1.369501285899697e-06,
+      "loss": 0.4285,
+      "step": 5078
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.404897518354886,
+      "learning_rate": 1.3678222558052357e-06,
+      "loss": 0.4278,
+      "step": 5079
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.274680194953741,
+      "learning_rate": 1.366144092492626e-06,
+      "loss": 0.4319,
+      "step": 5080
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.252560613176315,
+      "learning_rate": 1.364466796362348e-06,
+      "loss": 0.3952,
+      "step": 5081
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.587293480920098,
+      "learning_rate": 1.362790367814668e-06,
+      "loss": 0.4472,
+      "step": 5082
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.9490719415231865,
+      "learning_rate": 1.3611148072496515e-06,
+      "loss": 0.4503,
+      "step": 5083
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.045786535039181,
+      "learning_rate": 1.3594401150671521e-06,
+      "loss": 0.3873,
+      "step": 5084
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.506321741753398,
+      "learning_rate": 1.3577662916668167e-06,
+      "loss": 0.4566,
+      "step": 5085
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.8837960273195815,
+      "learning_rate": 1.3560933374480877e-06,
+      "loss": 0.435,
+      "step": 5086
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.092460221328097,
+      "learning_rate": 1.3544212528101986e-06,
+      "loss": 0.5137,
+      "step": 5087
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.0347044119138,
+      "learning_rate": 1.3527500381521746e-06,
+      "loss": 0.448,
+      "step": 5088
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.930244687050622,
+      "learning_rate": 1.3510796938728331e-06,
+      "loss": 0.3529,
+      "step": 5089
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.329309031948746,
+      "learning_rate": 1.3494102203707848e-06,
+      "loss": 0.4177,
+      "step": 5090
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.620250468507199,
+      "learning_rate": 1.3477416180444337e-06,
+      "loss": 0.4571,
+      "step": 5091
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.955158178794182,
+      "learning_rate": 1.346073887291972e-06,
+      "loss": 0.4233,
+      "step": 5092
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.703983476991287,
+      "learning_rate": 1.3444070285113903e-06,
+      "loss": 0.4612,
+      "step": 5093
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.573998628513427,
+      "learning_rate": 1.342741042100465e-06,
+      "loss": 0.4237,
+      "step": 5094
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.575709227372103,
+      "learning_rate": 1.3410759284567654e-06,
+      "loss": 0.3889,
+      "step": 5095
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.122960875552425,
+      "learning_rate": 1.339411687977657e-06,
+      "loss": 0.3971,
+      "step": 5096
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 11.590237955023165,
+      "learning_rate": 1.3377483210602916e-06,
+      "loss": 0.4994,
+      "step": 5097
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.60530443659417,
+      "learning_rate": 1.3360858281016148e-06,
+      "loss": 0.4718,
+      "step": 5098
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.644522660749432,
+      "learning_rate": 1.3344242094983634e-06,
+      "loss": 0.4426,
+      "step": 5099
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.694368078447145,
+      "learning_rate": 1.332763465647064e-06,
+      "loss": 0.3835,
+      "step": 5100
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.313994520571706,
+      "learning_rate": 1.3311035969440394e-06,
+      "loss": 0.3112,
+      "step": 5101
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.283374724321883,
+      "learning_rate": 1.3294446037853976e-06,
+      "loss": 0.4149,
+      "step": 5102
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.196433765836003,
+      "learning_rate": 1.327786486567042e-06,
+      "loss": 0.3418,
+      "step": 5103
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.0053539570023124,
+      "learning_rate": 1.3261292456846648e-06,
+      "loss": 0.4472,
+      "step": 5104
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.455400762496257,
+      "learning_rate": 1.3244728815337472e-06,
+      "loss": 0.4125,
+      "step": 5105
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.852037869214728,
+      "learning_rate": 1.3228173945095668e-06,
+      "loss": 0.4249,
+      "step": 5106
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.08036676535131,
+      "learning_rate": 1.3211627850071872e-06,
+      "loss": 0.3972,
+      "step": 5107
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.6794212779192295,
+      "learning_rate": 1.3195090534214628e-06,
+      "loss": 0.3449,
+      "step": 5108
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.8932722797065846,
+      "learning_rate": 1.3178562001470414e-06,
+      "loss": 0.4247,
+      "step": 5109
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.23750522876511,
+      "learning_rate": 1.3162042255783591e-06,
+      "loss": 0.4437,
+      "step": 5110
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 69.25474528353405,
+      "learning_rate": 1.314553130109642e-06,
+      "loss": 0.3696,
+      "step": 5111
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.391821194467162,
+      "learning_rate": 1.3129029141349075e-06,
+      "loss": 0.4452,
+      "step": 5112
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.4782162064933555,
+      "learning_rate": 1.3112535780479607e-06,
+      "loss": 0.4021,
+      "step": 5113
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.155508232891005,
+      "learning_rate": 1.3096051222424023e-06,
+      "loss": 0.4565,
+      "step": 5114
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 20.104237032047223,
+      "learning_rate": 1.3079575471116157e-06,
+      "loss": 0.4195,
+      "step": 5115
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 20.615915478594456,
+      "learning_rate": 1.3063108530487812e-06,
+      "loss": 0.4159,
+      "step": 5116
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.457310352825753,
+      "learning_rate": 1.304665040446863e-06,
+      "loss": 0.4336,
+      "step": 5117
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.047668525756197,
+      "learning_rate": 1.3030201096986172e-06,
+      "loss": 0.5261,
+      "step": 5118
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 13.476392258921688,
+      "learning_rate": 1.301376061196592e-06,
+      "loss": 0.4793,
+      "step": 5119
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2967073747642202,
+      "learning_rate": 1.2997328953331194e-06,
+      "loss": 0.5636,
+      "step": 5120
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.272901466599276,
+      "learning_rate": 1.2980906125003273e-06,
+      "loss": 0.4419,
+      "step": 5121
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.032222843301946,
+      "learning_rate": 1.296449213090128e-06,
+      "loss": 0.4813,
+      "step": 5122
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 12.105906031826008,
+      "learning_rate": 1.2948086974942242e-06,
+      "loss": 0.5116,
+      "step": 5123
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.168090665655384,
+      "learning_rate": 1.293169066104109e-06,
+      "loss": 0.4129,
+      "step": 5124
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.982034978506171,
+      "learning_rate": 1.2915303193110619e-06,
+      "loss": 0.3922,
+      "step": 5125
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.729913984841636,
+      "learning_rate": 1.289892457506155e-06,
+      "loss": 0.3865,
+      "step": 5126
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.713053274031575,
+      "learning_rate": 1.2882554810802466e-06,
+      "loss": 0.4657,
+      "step": 5127
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.168304109565556,
+      "learning_rate": 1.286619390423982e-06,
+      "loss": 0.4183,
+      "step": 5128
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.869869976406264,
+      "learning_rate": 1.2849841859278017e-06,
+      "loss": 0.3842,
+      "step": 5129
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.898132562050409,
+      "learning_rate": 1.2833498679819261e-06,
+      "loss": 0.362,
+      "step": 5130
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 12.084542099062338,
+      "learning_rate": 1.2817164369763718e-06,
+      "loss": 0.4165,
+      "step": 5131
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 16.603211518560276,
+      "learning_rate": 1.280083893300939e-06,
+      "loss": 0.5093,
+      "step": 5132
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.411280345203798,
+      "learning_rate": 1.2784522373452174e-06,
+      "loss": 0.3904,
+      "step": 5133
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 29.509160866334966,
+      "learning_rate": 1.2768214694985848e-06,
+      "loss": 0.4202,
+      "step": 5134
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.066588868172394,
+      "learning_rate": 1.2751915901502055e-06,
+      "loss": 0.4333,
+      "step": 5135
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 13.68333296425764,
+      "learning_rate": 1.273562599689037e-06,
+      "loss": 0.4025,
+      "step": 5136
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.386057410407292,
+      "learning_rate": 1.271934498503819e-06,
+      "loss": 0.4212,
+      "step": 5137
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.350984280334277,
+      "learning_rate": 1.27030728698308e-06,
+      "loss": 0.3799,
+      "step": 5138
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.86664175747827,
+      "learning_rate": 1.2686809655151406e-06,
+      "loss": 0.3936,
+      "step": 5139
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 24.64420114527087,
+      "learning_rate": 1.2670555344881014e-06,
+      "loss": 0.4639,
+      "step": 5140
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 21.063636411691398,
+      "learning_rate": 1.265430994289859e-06,
+      "loss": 0.4485,
+      "step": 5141
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 48.66010014128539,
+      "learning_rate": 1.2638073453080919e-06,
+      "loss": 0.357,
+      "step": 5142
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.281473066130745,
+      "learning_rate": 1.2621845879302646e-06,
+      "loss": 0.3863,
+      "step": 5143
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.739456548021198,
+      "learning_rate": 1.2605627225436352e-06,
+      "loss": 0.4571,
+      "step": 5144
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.785867311520014,
+      "learning_rate": 1.2589417495352436e-06,
+      "loss": 0.411,
+      "step": 5145
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 19.675627993697038,
+      "learning_rate": 1.2573216692919176e-06,
+      "loss": 0.3885,
+      "step": 5146
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.8680192078513445,
+      "learning_rate": 1.255702482200274e-06,
+      "loss": 0.4342,
+      "step": 5147
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.959885106418071,
+      "learning_rate": 1.254084188646713e-06,
+      "loss": 0.34,
+      "step": 5148
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 23.189940167183142,
+      "learning_rate": 1.2524667890174264e-06,
+      "loss": 0.4411,
+      "step": 5149
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.612172820829139,
+      "learning_rate": 1.250850283698387e-06,
+      "loss": 0.4543,
+      "step": 5150
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.925550668017921,
+      "learning_rate": 1.2492346730753613e-06,
+      "loss": 0.386,
+      "step": 5151
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.26038152259737,
+      "learning_rate": 1.2476199575338955e-06,
+      "loss": 0.4198,
+      "step": 5152
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.848273321239681,
+      "learning_rate": 1.246006137459323e-06,
+      "loss": 0.407,
+      "step": 5153
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 11.15514675183654,
+      "learning_rate": 1.2443932132367698e-06,
+      "loss": 0.4604,
+      "step": 5154
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.164009496871939,
+      "learning_rate": 1.2427811852511396e-06,
+      "loss": 0.3594,
+      "step": 5155
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.873862003696068,
+      "learning_rate": 1.2411700538871291e-06,
+      "loss": 0.5251,
+      "step": 5156
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.445269651283365,
+      "learning_rate": 1.239559819529218e-06,
+      "loss": 0.4922,
+      "step": 5157
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.748615963388067,
+      "learning_rate": 1.2379504825616705e-06,
+      "loss": 0.4075,
+      "step": 5158
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.341510109603943,
+      "learning_rate": 1.2363420433685397e-06,
+      "loss": 0.4819,
+      "step": 5159
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3017595339647985,
+      "learning_rate": 1.234734502333661e-06,
+      "loss": 0.54,
+      "step": 5160
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.537145110478899,
+      "learning_rate": 1.2331278598406603e-06,
+      "loss": 0.4541,
+      "step": 5161
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.117118839961999,
+      "learning_rate": 1.2315221162729457e-06,
+      "loss": 0.4469,
+      "step": 5162
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.540606619211511,
+      "learning_rate": 1.2299172720137086e-06,
+      "loss": 0.3367,
+      "step": 5163
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 16.212249645585914,
+      "learning_rate": 1.2283133274459331e-06,
+      "loss": 0.4083,
+      "step": 5164
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1453264676266575,
+      "learning_rate": 1.2267102829523791e-06,
+      "loss": 0.541,
+      "step": 5165
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.24754173471259,
+      "learning_rate": 1.2251081389156015e-06,
+      "loss": 0.3938,
+      "step": 5166
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.920056837344895,
+      "learning_rate": 1.223506895717933e-06,
+      "loss": 0.4108,
+      "step": 5167
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.74138231909456,
+      "learning_rate": 1.2219065537414942e-06,
+      "loss": 0.4541,
+      "step": 5168
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.493548631984039,
+      "learning_rate": 1.2203071133681898e-06,
+      "loss": 0.3815,
+      "step": 5169
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 39.58590795316855,
+      "learning_rate": 1.2187085749797089e-06,
+      "loss": 0.4503,
+      "step": 5170
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 13.659835420170028,
+      "learning_rate": 1.2171109389575293e-06,
+      "loss": 0.4047,
+      "step": 5171
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.260350055767581,
+      "learning_rate": 1.2155142056829083e-06,
+      "loss": 0.4899,
+      "step": 5172
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.403695723549134,
+      "learning_rate": 1.2139183755368889e-06,
+      "loss": 0.3754,
+      "step": 5173
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 13.818665836170146,
+      "learning_rate": 1.2123234489003015e-06,
+      "loss": 0.4229,
+      "step": 5174
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.346491825637881,
+      "learning_rate": 1.2107294261537572e-06,
+      "loss": 0.4204,
+      "step": 5175
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.639800028249642,
+      "learning_rate": 1.2091363076776557e-06,
+      "loss": 0.3931,
+      "step": 5176
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.108770899807388,
+      "learning_rate": 1.207544093852177e-06,
+      "loss": 0.4301,
+      "step": 5177
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 18.90795717953795,
+      "learning_rate": 1.2059527850572843e-06,
+      "loss": 0.4846,
+      "step": 5178
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.544912438320002,
+      "learning_rate": 1.2043623816727312e-06,
+      "loss": 0.3919,
+      "step": 5179
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.472741349013857,
+      "learning_rate": 1.2027728840780488e-06,
+      "loss": 0.3832,
+      "step": 5180
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.692862650437506,
+      "learning_rate": 1.2011842926525546e-06,
+      "loss": 0.4579,
+      "step": 5181
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.07395618371269,
+      "learning_rate": 1.1995966077753496e-06,
+      "loss": 0.4485,
+      "step": 5182
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 11.765022936239674,
+      "learning_rate": 1.1980098298253167e-06,
+      "loss": 0.3811,
+      "step": 5183
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 21.99215706969242,
+      "learning_rate": 1.1964239591811277e-06,
+      "loss": 0.4788,
+      "step": 5184
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 11.826477716857916,
+      "learning_rate": 1.1948389962212308e-06,
+      "loss": 0.4322,
+      "step": 5185
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.096983445717911,
+      "learning_rate": 1.1932549413238643e-06,
+      "loss": 0.4465,
+      "step": 5186
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.026818728888764,
+      "learning_rate": 1.1916717948670449e-06,
+      "loss": 0.4769,
+      "step": 5187
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.27559121547826,
+      "learning_rate": 1.190089557228573e-06,
+      "loss": 0.4182,
+      "step": 5188
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 15.101294881154143,
+      "learning_rate": 1.1885082287860361e-06,
+      "loss": 0.4571,
+      "step": 5189
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.248892158745734,
+      "learning_rate": 1.1869278099167997e-06,
+      "loss": 0.4346,
+      "step": 5190
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.91246491225219,
+      "learning_rate": 1.1853483009980166e-06,
+      "loss": 0.4192,
+      "step": 5191
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 12.274843248254083,
+      "learning_rate": 1.1837697024066192e-06,
+      "loss": 0.4277,
+      "step": 5192
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 29.499369614079253,
+      "learning_rate": 1.1821920145193232e-06,
+      "loss": 0.4382,
+      "step": 5193
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2177193721271373,
+      "learning_rate": 1.180615237712629e-06,
+      "loss": 0.5203,
+      "step": 5194
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.000334475433054,
+      "learning_rate": 1.1790393723628157e-06,
+      "loss": 0.4305,
+      "step": 5195
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2452708378447501,
+      "learning_rate": 1.17746441884595e-06,
+      "loss": 0.5829,
+      "step": 5196
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 12.920142900737668,
+      "learning_rate": 1.1758903775378777e-06,
+      "loss": 0.4838,
+      "step": 5197
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.337086902962379,
+      "learning_rate": 1.1743172488142257e-06,
+      "loss": 0.4407,
+      "step": 5198
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.083253470462495,
+      "learning_rate": 1.1727450330504082e-06,
+      "loss": 0.371,
+      "step": 5199
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 14.57793039208313,
+      "learning_rate": 1.1711737306216147e-06,
+      "loss": 0.4439,
+      "step": 5200
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.370808857700116,
+      "learning_rate": 1.1696033419028242e-06,
+      "loss": 0.4751,
+      "step": 5201
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.85052341898441,
+      "learning_rate": 1.1680338672687913e-06,
+      "loss": 0.4178,
+      "step": 5202
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.62725604523496,
+      "learning_rate": 1.1664653070940562e-06,
+      "loss": 0.4499,
+      "step": 5203
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 11.785066592062842,
+      "learning_rate": 1.1648976617529383e-06,
+      "loss": 0.459,
+      "step": 5204
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.139168944944306,
+      "learning_rate": 1.1633309316195402e-06,
+      "loss": 0.4385,
+      "step": 5205
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 11.987025374891225,
+      "learning_rate": 1.1617651170677474e-06,
+      "loss": 0.3551,
+      "step": 5206
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.091033276487138,
+      "learning_rate": 1.1602002184712242e-06,
+      "loss": 0.4036,
+      "step": 5207
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.028489284680846,
+      "learning_rate": 1.158636236203417e-06,
+      "loss": 0.4398,
+      "step": 5208
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.596042492868288,
+      "learning_rate": 1.157073170637556e-06,
+      "loss": 0.4539,
+      "step": 5209
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 12.258058163483097,
+      "learning_rate": 1.1555110221466487e-06,
+      "loss": 0.4554,
+      "step": 5210
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.713158450557584,
+      "learning_rate": 1.1539497911034875e-06,
+      "loss": 0.3848,
+      "step": 5211
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.268720137388348,
+      "learning_rate": 1.1523894778806439e-06,
+      "loss": 0.4247,
+      "step": 5212
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.767642548762897,
+      "learning_rate": 1.1508300828504682e-06,
+      "loss": 0.4178,
+      "step": 5213
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.216853147507242,
+      "learning_rate": 1.1492716063850973e-06,
+      "loss": 0.4653,
+      "step": 5214
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 19.075985686223902,
+      "learning_rate": 1.1477140488564443e-06,
+      "loss": 0.4205,
+      "step": 5215
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 21.48094994487318,
+      "learning_rate": 1.1461574106362043e-06,
+      "loss": 0.3891,
+      "step": 5216
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.502345181472593,
+      "learning_rate": 1.1446016920958524e-06,
+      "loss": 0.4421,
+      "step": 5217
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 17.610433753667458,
+      "learning_rate": 1.1430468936066442e-06,
+      "loss": 0.4814,
+      "step": 5218
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1727431148548835,
+      "learning_rate": 1.1414930155396186e-06,
+      "loss": 0.492,
+      "step": 5219
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 12.525349032783764,
+      "learning_rate": 1.1399400582655911e-06,
+      "loss": 0.4542,
+      "step": 5220
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.544124345077458,
+      "learning_rate": 1.138388022155158e-06,
+      "loss": 0.4135,
+      "step": 5221
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.973888666074085,
+      "learning_rate": 1.1368369075786995e-06,
+      "loss": 0.3604,
+      "step": 5222
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 16.320872691536408,
+      "learning_rate": 1.1352867149063706e-06,
+      "loss": 0.4852,
+      "step": 5223
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 12.007480316573227,
+      "learning_rate": 1.1337374445081106e-06,
+      "loss": 0.5231,
+      "step": 5224
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.569547556532935,
+      "learning_rate": 1.132189096753637e-06,
+      "loss": 0.4011,
+      "step": 5225
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.438710607587812,
+      "learning_rate": 1.1306416720124459e-06,
+      "loss": 0.4146,
+      "step": 5226
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.600343731643779,
+      "learning_rate": 1.1290951706538145e-06,
+      "loss": 0.409,
+      "step": 5227
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.926458272949903,
+      "learning_rate": 1.1275495930467988e-06,
+      "loss": 0.397,
+      "step": 5228
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.639140339002207,
+      "learning_rate": 1.1260049395602368e-06,
+      "loss": 0.4193,
+      "step": 5229
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 14.94909099458609,
+      "learning_rate": 1.1244612105627434e-06,
+      "loss": 0.4862,
+      "step": 5230
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.995619107407077,
+      "learning_rate": 1.1229184064227117e-06,
+      "loss": 0.475,
+      "step": 5231
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 14.3032674304945,
+      "learning_rate": 1.1213765275083193e-06,
+      "loss": 0.4597,
+      "step": 5232
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.039345703647712,
+      "learning_rate": 1.1198355741875171e-06,
+      "loss": 0.443,
+      "step": 5233
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3641022482825955,
+      "learning_rate": 1.1182955468280404e-06,
+      "loss": 0.4717,
+      "step": 5234
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 20.235760096401627,
+      "learning_rate": 1.1167564457973989e-06,
+      "loss": 0.3645,
+      "step": 5235
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.450642579799381,
+      "learning_rate": 1.1152182714628828e-06,
+      "loss": 0.4189,
+      "step": 5236
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 19.491593805563863,
+      "learning_rate": 1.1136810241915635e-06,
+      "loss": 0.4447,
+      "step": 5237
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 20.41090531956967,
+      "learning_rate": 1.112144704350288e-06,
+      "loss": 0.4328,
+      "step": 5238
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.505361119913244,
+      "learning_rate": 1.1106093123056837e-06,
+      "loss": 0.4498,
+      "step": 5239
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1849355129116315,
+      "learning_rate": 1.109074848424156e-06,
+      "loss": 0.526,
+      "step": 5240
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.862573424603288,
+      "learning_rate": 1.1075413130718877e-06,
+      "loss": 0.4209,
+      "step": 5241
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 12.907877673102453,
+      "learning_rate": 1.106008706614843e-06,
+      "loss": 0.488,
+      "step": 5242
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.700250926595706,
+      "learning_rate": 1.1044770294187612e-06,
+      "loss": 0.3887,
+      "step": 5243
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 15.731433854698702,
+      "learning_rate": 1.1029462818491632e-06,
+      "loss": 0.3796,
+      "step": 5244
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.050652082431767,
+      "learning_rate": 1.1014164642713448e-06,
+      "loss": 0.3696,
+      "step": 5245
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.80600062796878,
+      "learning_rate": 1.0998875770503802e-06,
+      "loss": 0.4105,
+      "step": 5246
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 16.827187558745475,
+      "learning_rate": 1.0983596205511254e-06,
+      "loss": 0.434,
+      "step": 5247
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.94698906524837,
+      "learning_rate": 1.0968325951382087e-06,
+      "loss": 0.4323,
+      "step": 5248
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.989165068700943,
+      "learning_rate": 1.0953065011760417e-06,
+      "loss": 0.3795,
+      "step": 5249
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.315884966656627,
+      "learning_rate": 1.0937813390288093e-06,
+      "loss": 0.5044,
+      "step": 5250
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 19.3752936298468,
+      "learning_rate": 1.092257109060476e-06,
+      "loss": 0.4636,
+      "step": 5251
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.742785348959713,
+      "learning_rate": 1.0907338116347842e-06,
+      "loss": 0.4063,
+      "step": 5252
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.197864785968546,
+      "learning_rate": 1.089211447115251e-06,
+      "loss": 0.4279,
+      "step": 5253
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 42.55599065405896,
+      "learning_rate": 1.087690015865176e-06,
+      "loss": 0.4398,
+      "step": 5254
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 17.435792625543357,
+      "learning_rate": 1.0861695182476318e-06,
+      "loss": 0.4745,
+      "step": 5255
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.294964717592858,
+      "learning_rate": 1.0846499546254685e-06,
+      "loss": 0.4321,
+      "step": 5256
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.080013680961672,
+      "learning_rate": 1.0831313253613162e-06,
+      "loss": 0.4354,
+      "step": 5257
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.34560421743313,
+      "learning_rate": 1.0816136308175778e-06,
+      "loss": 0.445,
+      "step": 5258
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.427217639397913,
+      "learning_rate": 1.0800968713564387e-06,
+      "loss": 0.4083,
+      "step": 5259
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.464687147707914,
+      "learning_rate": 1.0785810473398562e-06,
+      "loss": 0.4389,
+      "step": 5260
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.30510846547853,
+      "learning_rate": 1.0770661591295662e-06,
+      "loss": 0.4677,
+      "step": 5261
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.060799733776654,
+      "learning_rate": 1.0755522070870806e-06,
+      "loss": 0.4414,
+      "step": 5262
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.122451529670742,
+      "learning_rate": 1.074039191573688e-06,
+      "loss": 0.4667,
+      "step": 5263
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.790589408615965,
+      "learning_rate": 1.072527112950456e-06,
+      "loss": 0.4312,
+      "step": 5264
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.188513790568354,
+      "learning_rate": 1.071015971578226e-06,
+      "loss": 0.4594,
+      "step": 5265
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.293406749758283,
+      "learning_rate": 1.0695057678176146e-06,
+      "loss": 0.4654,
+      "step": 5266
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.401803872182206,
+      "learning_rate": 1.067996502029019e-06,
+      "loss": 0.3949,
+      "step": 5267
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 14.206695496936929,
+      "learning_rate": 1.066488174572607e-06,
+      "loss": 0.4492,
+      "step": 5268
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 11.086724535628768,
+      "learning_rate": 1.0649807858083289e-06,
+      "loss": 0.3833,
+      "step": 5269
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.2338792586553,
+      "learning_rate": 1.0634743360959054e-06,
+      "loss": 0.4675,
+      "step": 5270
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 14.200366538782356,
+      "learning_rate": 1.0619688257948342e-06,
+      "loss": 0.4065,
+      "step": 5271
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 12.759278601939412,
+      "learning_rate": 1.0604642552643923e-06,
+      "loss": 0.3823,
+      "step": 5272
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.41425315419526,
+      "learning_rate": 1.0589606248636291e-06,
+      "loss": 0.4213,
+      "step": 5273
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.556358613651867,
+      "learning_rate": 1.0574579349513708e-06,
+      "loss": 0.4393,
+      "step": 5274
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 12.935282220300795,
+      "learning_rate": 1.0559561858862177e-06,
+      "loss": 0.4972,
+      "step": 5275
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.833717266170387,
+      "learning_rate": 1.0544553780265466e-06,
+      "loss": 0.3769,
+      "step": 5276
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 13.544550508415306,
+      "learning_rate": 1.0529555117305117e-06,
+      "loss": 0.3664,
+      "step": 5277
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 33.131326730641916,
+      "learning_rate": 1.0514565873560384e-06,
+      "loss": 0.4159,
+      "step": 5278
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 11.837387050213831,
+      "learning_rate": 1.0499586052608323e-06,
+      "loss": 0.3898,
+      "step": 5279
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.371320740145254,
+      "learning_rate": 1.04846156580237e-06,
+      "loss": 0.4659,
+      "step": 5280
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.046284978154636,
+      "learning_rate": 1.0469654693379027e-06,
+      "loss": 0.3309,
+      "step": 5281
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.520986960420434,
+      "learning_rate": 1.0454703162244623e-06,
+      "loss": 0.3909,
+      "step": 5282
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.199677140028825,
+      "learning_rate": 1.0439761068188482e-06,
+      "loss": 0.4164,
+      "step": 5283
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.653736334657397,
+      "learning_rate": 1.0424828414776405e-06,
+      "loss": 0.4403,
+      "step": 5284
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.427586901550289,
+      "learning_rate": 1.0409905205571914e-06,
+      "loss": 0.4399,
+      "step": 5285
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.473819357641133,
+      "learning_rate": 1.0394991444136265e-06,
+      "loss": 0.3068,
+      "step": 5286
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.3930709411111,
+      "learning_rate": 1.0380087134028482e-06,
+      "loss": 0.4176,
+      "step": 5287
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.1545566685279125,
+      "learning_rate": 1.0365192278805313e-06,
+      "loss": 0.3902,
+      "step": 5288
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.234105407720222,
+      "learning_rate": 1.0350306882021287e-06,
+      "loss": 0.4429,
+      "step": 5289
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.771297465224404,
+      "learning_rate": 1.0335430947228642e-06,
+      "loss": 0.4064,
+      "step": 5290
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 12.936314675266622,
+      "learning_rate": 1.032056447797734e-06,
+      "loss": 0.3565,
+      "step": 5291
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.408937629863596,
+      "learning_rate": 1.0305707477815152e-06,
+      "loss": 0.3856,
+      "step": 5292
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 15.813999713890535,
+      "learning_rate": 1.0290859950287512e-06,
+      "loss": 0.4508,
+      "step": 5293
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.966289492204469,
+      "learning_rate": 1.027602189893766e-06,
+      "loss": 0.4801,
+      "step": 5294
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 15.975350251276435,
+      "learning_rate": 1.0261193327306534e-06,
+      "loss": 0.3629,
+      "step": 5295
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 19.66338686977912,
+      "learning_rate": 1.0246374238932816e-06,
+      "loss": 0.4264,
+      "step": 5296
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.153297592693,
+      "learning_rate": 1.0231564637352931e-06,
+      "loss": 0.4036,
+      "step": 5297
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.814403688561901,
+      "learning_rate": 1.021676452610102e-06,
+      "loss": 0.4095,
+      "step": 5298
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 15.548988967711908,
+      "learning_rate": 1.0201973908709011e-06,
+      "loss": 0.4349,
+      "step": 5299
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 20.660775469645763,
+      "learning_rate": 1.0187192788706518e-06,
+      "loss": 0.3837,
+      "step": 5300
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 13.31224617058704,
+      "learning_rate": 1.017242116962089e-06,
+      "loss": 0.4714,
+      "step": 5301
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 12.529008240697914,
+      "learning_rate": 1.0157659054977237e-06,
+      "loss": 0.3807,
+      "step": 5302
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.68972179040758,
+      "learning_rate": 1.0142906448298379e-06,
+      "loss": 0.4383,
+      "step": 5303
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.764145934663592,
+      "learning_rate": 1.0128163353104885e-06,
+      "loss": 0.4468,
+      "step": 5304
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 14.013593383993157,
+      "learning_rate": 1.0113429772915035e-06,
+      "loss": 0.4578,
+      "step": 5305
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 52.972378245581666,
+      "learning_rate": 1.0098705711244833e-06,
+      "loss": 0.4233,
+      "step": 5306
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.003729375059885,
+      "learning_rate": 1.0083991171608054e-06,
+      "loss": 0.443,
+      "step": 5307
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.2713120334391,
+      "learning_rate": 1.0069286157516155e-06,
+      "loss": 0.4061,
+      "step": 5308
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.6198649402066,
+      "learning_rate": 1.0054590672478338e-06,
+      "loss": 0.4388,
+      "step": 5309
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 12.993781945864539,
+      "learning_rate": 1.0039904720001525e-06,
+      "loss": 0.4329,
+      "step": 5310
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 14.130819494909543,
+      "learning_rate": 1.0025228303590363e-06,
+      "loss": 0.395,
+      "step": 5311
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.844447558354887,
+      "learning_rate": 1.0010561426747246e-06,
+      "loss": 0.4606,
+      "step": 5312
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.733595716993202,
+      "learning_rate": 9.995904092972247e-07,
+      "loss": 0.4685,
+      "step": 5313
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.073245529057157,
+      "learning_rate": 9.98125630576322e-07,
+      "loss": 0.4309,
+      "step": 5314
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 13.480674402508539,
+      "learning_rate": 9.966618068615692e-07,
+      "loss": 0.4629,
+      "step": 5315
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.050572746636589,
+      "learning_rate": 9.951989385022914e-07,
+      "loss": 0.4178,
+      "step": 5316
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.916632033503475,
+      "learning_rate": 9.937370258475897e-07,
+      "loss": 0.3576,
+      "step": 5317
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 46.572830132848544,
+      "learning_rate": 9.922760692463323e-07,
+      "loss": 0.4319,
+      "step": 5318
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.253967507060426,
+      "learning_rate": 9.908160690471641e-07,
+      "loss": 0.3997,
+      "step": 5319
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.20017756424348,
+      "learning_rate": 9.893570255984968e-07,
+      "loss": 0.4085,
+      "step": 5320
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 13.998402059686661,
+      "learning_rate": 9.878989392485173e-07,
+      "loss": 0.4871,
+      "step": 5321
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 12.1145248316699,
+      "learning_rate": 9.86441810345183e-07,
+      "loss": 0.4949,
+      "step": 5322
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.155262799365879,
+      "learning_rate": 9.849856392362201e-07,
+      "loss": 0.4464,
+      "step": 5323
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.3441905632034965,
+      "learning_rate": 9.83530426269132e-07,
+      "loss": 0.4947,
+      "step": 5324
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 28.563408218828048,
+      "learning_rate": 9.820761717911897e-07,
+      "loss": 0.4785,
+      "step": 5325
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.342751265067179,
+      "learning_rate": 9.806228761494341e-07,
+      "loss": 0.4825,
+      "step": 5326
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 49.722750149805314,
+      "learning_rate": 9.79170539690682e-07,
+      "loss": 0.4701,
+      "step": 5327
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.53220764843252,
+      "learning_rate": 9.777191627615163e-07,
+      "loss": 0.4701,
+      "step": 5328
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 78.49228166441343,
+      "learning_rate": 9.762687457082954e-07,
+      "loss": 0.4688,
+      "step": 5329
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 29.672214054558914,
+      "learning_rate": 9.748192888771452e-07,
+      "loss": 0.407,
+      "step": 5330
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1171786664690846,
+      "learning_rate": 9.733707926139636e-07,
+      "loss": 0.5409,
+      "step": 5331
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.807406188305423,
+      "learning_rate": 9.719232572644189e-07,
+      "loss": 0.3971,
+      "step": 5332
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 13.134096030758567,
+      "learning_rate": 9.704766831739514e-07,
+      "loss": 0.4528,
+      "step": 5333
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.771346155826658,
+      "learning_rate": 9.690310706877698e-07,
+      "loss": 0.4757,
+      "step": 5334
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 11.64095514484084,
+      "learning_rate": 9.67586420150856e-07,
+      "loss": 0.4665,
+      "step": 5335
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.841373354329896,
+      "learning_rate": 9.661427319079603e-07,
+      "loss": 0.4369,
+      "step": 5336
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 23.246526163948307,
+      "learning_rate": 9.647000063036048e-07,
+      "loss": 0.4107,
+      "step": 5337
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 14.568157543815765,
+      "learning_rate": 9.63258243682081e-07,
+      "loss": 0.4377,
+      "step": 5338
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.636343147116894,
+      "learning_rate": 9.61817444387449e-07,
+      "loss": 0.4075,
+      "step": 5339
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.27642767753732,
+      "learning_rate": 9.603776087635435e-07,
+      "loss": 0.4763,
+      "step": 5340
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.859837587409997,
+      "learning_rate": 9.589387371539638e-07,
+      "loss": 0.5496,
+      "step": 5341
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 14.78797014793003,
+      "learning_rate": 9.575008299020838e-07,
+      "loss": 0.4167,
+      "step": 5342
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.809449159255145,
+      "learning_rate": 9.560638873510452e-07,
+      "loss": 0.4425,
+      "step": 5343
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.327736070113407,
+      "learning_rate": 9.546279098437584e-07,
+      "loss": 0.4135,
+      "step": 5344
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.317090082233126,
+      "learning_rate": 9.531928977229055e-07,
+      "loss": 0.3851,
+      "step": 5345
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.298722133531391,
+      "learning_rate": 9.517588513309356e-07,
+      "loss": 0.4234,
+      "step": 5346
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 141.77190223924276,
+      "learning_rate": 9.503257710100717e-07,
+      "loss": 0.5149,
+      "step": 5347
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 11.4353560413999,
+      "learning_rate": 9.488936571023022e-07,
+      "loss": 0.4062,
+      "step": 5348
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.947462834547668,
+      "learning_rate": 9.474625099493862e-07,
+      "loss": 0.4298,
+      "step": 5349
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.987739151364714,
+      "learning_rate": 9.460323298928531e-07,
+      "loss": 0.4521,
+      "step": 5350
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 26.889159092654957,
+      "learning_rate": 9.446031172739995e-07,
+      "loss": 0.4861,
+      "step": 5351
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 19.90265312681175,
+      "learning_rate": 9.431748724338946e-07,
+      "loss": 0.4991,
+      "step": 5352
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.186931072508379,
+      "learning_rate": 9.417475957133726e-07,
+      "loss": 0.347,
+      "step": 5353
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.623916292344132,
+      "learning_rate": 9.403212874530382e-07,
+      "loss": 0.4723,
+      "step": 5354
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.473699289582015,
+      "learning_rate": 9.388959479932658e-07,
+      "loss": 0.3714,
+      "step": 5355
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.991104637246604,
+      "learning_rate": 9.374715776741967e-07,
+      "loss": 0.4417,
+      "step": 5356
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 18.34837033996337,
+      "learning_rate": 9.360481768357449e-07,
+      "loss": 0.4074,
+      "step": 5357
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.168149168833336,
+      "learning_rate": 9.346257458175884e-07,
+      "loss": 0.4204,
+      "step": 5358
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 16.718114171699025,
+      "learning_rate": 9.332042849591749e-07,
+      "loss": 0.4634,
+      "step": 5359
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.54269076672843,
+      "learning_rate": 9.317837945997238e-07,
+      "loss": 0.3795,
+      "step": 5360
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 12.963376025408735,
+      "learning_rate": 9.303642750782177e-07,
+      "loss": 0.3941,
+      "step": 5361
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.727492164126,
+      "learning_rate": 9.28945726733414e-07,
+      "loss": 0.4527,
+      "step": 5362
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 17.435572675800458,
+      "learning_rate": 9.275281499038325e-07,
+      "loss": 0.4318,
+      "step": 5363
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.354787508507699,
+      "learning_rate": 9.261115449277619e-07,
+      "loss": 0.4289,
+      "step": 5364
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 16.05186992055341,
+      "learning_rate": 9.24695912143263e-07,
+      "loss": 0.5115,
+      "step": 5365
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.906203933418942,
+      "learning_rate": 9.232812518881618e-07,
+      "loss": 0.4823,
+      "step": 5366
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 18.532784926667624,
+      "learning_rate": 9.218675645000508e-07,
+      "loss": 0.3802,
+      "step": 5367
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 13.150237916566377,
+      "learning_rate": 9.20454850316293e-07,
+      "loss": 0.455,
+      "step": 5368
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 11.821623203235147,
+      "learning_rate": 9.190431096740172e-07,
+      "loss": 0.3826,
+      "step": 5369
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 57.15329305542216,
+      "learning_rate": 9.176323429101219e-07,
+      "loss": 0.3859,
+      "step": 5370
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.500499509022175,
+      "learning_rate": 9.162225503612704e-07,
+      "loss": 0.4201,
+      "step": 5371
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 11.458023785866128,
+      "learning_rate": 9.148137323638978e-07,
+      "loss": 0.3961,
+      "step": 5372
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 19.575331499971533,
+      "learning_rate": 9.13405889254203e-07,
+      "loss": 0.4376,
+      "step": 5373
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 14.570755550071617,
+      "learning_rate": 9.119990213681512e-07,
+      "loss": 0.4522,
+      "step": 5374
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 26.613532275491732,
+      "learning_rate": 9.105931290414799e-07,
+      "loss": 0.4517,
+      "step": 5375
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 23.17720518315366,
+      "learning_rate": 9.091882126096885e-07,
+      "loss": 0.4643,
+      "step": 5376
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.807516051988936,
+      "learning_rate": 9.077842724080482e-07,
+      "loss": 0.4292,
+      "step": 5377
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.669734192521455,
+      "learning_rate": 9.063813087715933e-07,
+      "loss": 0.4309,
+      "step": 5378
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.961110951705839,
+      "learning_rate": 9.049793220351272e-07,
+      "loss": 0.4317,
+      "step": 5379
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.45197552929036,
+      "learning_rate": 9.035783125332199e-07,
+      "loss": 0.3876,
+      "step": 5380
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 11.301394583195993,
+      "learning_rate": 9.021782806002055e-07,
+      "loss": 0.4791,
+      "step": 5381
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 11.663360197406467,
+      "learning_rate": 9.007792265701903e-07,
+      "loss": 0.3954,
+      "step": 5382
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.999743232797956,
+      "learning_rate": 8.993811507770428e-07,
+      "loss": 0.4135,
+      "step": 5383
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.125528599395487,
+      "learning_rate": 8.979840535543988e-07,
+      "loss": 0.4115,
+      "step": 5384
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 14.42854521659178,
+      "learning_rate": 8.965879352356632e-07,
+      "loss": 0.4148,
+      "step": 5385
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 23.552832492318323,
+      "learning_rate": 8.951927961540025e-07,
+      "loss": 0.4123,
+      "step": 5386
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.359465691146914,
+      "learning_rate": 8.937986366423551e-07,
+      "loss": 0.4304,
+      "step": 5387
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.943569126365619,
+      "learning_rate": 8.924054570334212e-07,
+      "loss": 0.4343,
+      "step": 5388
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 11.692673723055693,
+      "learning_rate": 8.910132576596697e-07,
+      "loss": 0.4283,
+      "step": 5389
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.886952991639738,
+      "learning_rate": 8.896220388533338e-07,
+      "loss": 0.4901,
+      "step": 5390
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.149600853929758,
+      "learning_rate": 8.882318009464124e-07,
+      "loss": 0.4219,
+      "step": 5391
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0314381093565235,
+      "learning_rate": 8.868425442706747e-07,
+      "loss": 0.49,
+      "step": 5392
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 11.238530004192857,
+      "learning_rate": 8.8545426915765e-07,
+      "loss": 0.4258,
+      "step": 5393
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 15.31924123839077,
+      "learning_rate": 8.840669759386355e-07,
+      "loss": 0.4736,
+      "step": 5394
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.941482194418099,
+      "learning_rate": 8.826806649446973e-07,
+      "loss": 0.4757,
+      "step": 5395
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 13.217421501621535,
+      "learning_rate": 8.812953365066607e-07,
+      "loss": 0.4531,
+      "step": 5396
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 20.988005390500383,
+      "learning_rate": 8.799109909551234e-07,
+      "loss": 0.4171,
+      "step": 5397
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.609200048094525,
+      "learning_rate": 8.785276286204436e-07,
+      "loss": 0.4233,
+      "step": 5398
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 19.681995742839995,
+      "learning_rate": 8.771452498327454e-07,
+      "loss": 0.3887,
+      "step": 5399
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.540920297998119,
+      "learning_rate": 8.757638549219216e-07,
+      "loss": 0.3774,
+      "step": 5400
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 11.855411721353846,
+      "learning_rate": 8.743834442176269e-07,
+      "loss": 0.3917,
+      "step": 5401
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 12.478983553372718,
+      "learning_rate": 8.730040180492822e-07,
+      "loss": 0.4504,
+      "step": 5402
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.894161809961167,
+      "learning_rate": 8.716255767460729e-07,
+      "loss": 0.5123,
+      "step": 5403
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0773921881952275,
+      "learning_rate": 8.702481206369485e-07,
+      "loss": 0.5369,
+      "step": 5404
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.460680299698193,
+      "learning_rate": 8.688716500506278e-07,
+      "loss": 0.4928,
+      "step": 5405
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 14.941661007730284,
+      "learning_rate": 8.674961653155889e-07,
+      "loss": 0.4299,
+      "step": 5406
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 12.470646639276142,
+      "learning_rate": 8.661216667600785e-07,
+      "loss": 0.4447,
+      "step": 5407
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.8558013261313215,
+      "learning_rate": 8.647481547121067e-07,
+      "loss": 0.3873,
+      "step": 5408
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.246643880539626,
+      "learning_rate": 8.633756294994461e-07,
+      "loss": 0.4499,
+      "step": 5409
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 11.355628594910776,
+      "learning_rate": 8.62004091449638e-07,
+      "loss": 0.4973,
+      "step": 5410
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 11.294057088049206,
+      "learning_rate": 8.606335408899841e-07,
+      "loss": 0.361,
+      "step": 5411
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 20.095617129834707,
+      "learning_rate": 8.592639781475537e-07,
+      "loss": 0.4067,
+      "step": 5412
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.774971684403115,
+      "learning_rate": 8.578954035491782e-07,
+      "loss": 0.3653,
+      "step": 5413
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 22.3761901239624,
+      "learning_rate": 8.565278174214541e-07,
+      "loss": 0.4877,
+      "step": 5414
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 14.323027601414552,
+      "learning_rate": 8.551612200907411e-07,
+      "loss": 0.379,
+      "step": 5415
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 12.014425899888368,
+      "learning_rate": 8.537956118831626e-07,
+      "loss": 0.3544,
+      "step": 5416
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 11.983591973531155,
+      "learning_rate": 8.524309931246094e-07,
+      "loss": 0.4074,
+      "step": 5417
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 13.098253115895954,
+      "learning_rate": 8.51067364140733e-07,
+      "loss": 0.426,
+      "step": 5418
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.278898689836112,
+      "learning_rate": 8.49704725256948e-07,
+      "loss": 0.4144,
+      "step": 5419
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.729294212717467,
+      "learning_rate": 8.483430767984358e-07,
+      "loss": 0.4441,
+      "step": 5420
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.560971225962868,
+      "learning_rate": 8.469824190901382e-07,
+      "loss": 0.421,
+      "step": 5421
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 14.92554607386126,
+      "learning_rate": 8.45622752456764e-07,
+      "loss": 0.449,
+      "step": 5422
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 12.751843861618113,
+      "learning_rate": 8.442640772227834e-07,
+      "loss": 0.4509,
+      "step": 5423
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.6499411260161745,
+      "learning_rate": 8.429063937124294e-07,
+      "loss": 0.4114,
+      "step": 5424
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.880109251472874,
+      "learning_rate": 8.415497022496994e-07,
+      "loss": 0.369,
+      "step": 5425
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.010374038016272,
+      "learning_rate": 8.401940031583523e-07,
+      "loss": 0.4386,
+      "step": 5426
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.3201261174502195,
+      "learning_rate": 8.388392967619152e-07,
+      "loss": 0.4033,
+      "step": 5427
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 13.96327718756873,
+      "learning_rate": 8.374855833836726e-07,
+      "loss": 0.4673,
+      "step": 5428
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.885553946939302,
+      "learning_rate": 8.361328633466737e-07,
+      "loss": 0.473,
+      "step": 5429
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 17.758939212542643,
+      "learning_rate": 8.347811369737335e-07,
+      "loss": 0.4137,
+      "step": 5430
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.630900636313524,
+      "learning_rate": 8.334304045874248e-07,
+      "loss": 0.4273,
+      "step": 5431
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.094728582479015,
+      "learning_rate": 8.320806665100889e-07,
+      "loss": 0.426,
+      "step": 5432
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.909052902383191,
+      "learning_rate": 8.307319230638255e-07,
+      "loss": 0.4948,
+      "step": 5433
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.188368797120107,
+      "learning_rate": 8.293841745704967e-07,
+      "loss": 0.4242,
+      "step": 5434
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 17.74081302461166,
+      "learning_rate": 8.280374213517323e-07,
+      "loss": 0.4164,
+      "step": 5435
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 46.89130617455893,
+      "learning_rate": 8.266916637289191e-07,
+      "loss": 0.4618,
+      "step": 5436
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.479381634809684,
+      "learning_rate": 8.25346902023208e-07,
+      "loss": 0.4521,
+      "step": 5437
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.354713366541333,
+      "learning_rate": 8.240031365555134e-07,
+      "loss": 0.3983,
+      "step": 5438
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.092164241310773,
+      "learning_rate": 8.226603676465094e-07,
+      "loss": 0.4554,
+      "step": 5439
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.119249123186006,
+      "learning_rate": 8.213185956166364e-07,
+      "loss": 0.5143,
+      "step": 5440
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.121095829200007,
+      "learning_rate": 8.199778207860925e-07,
+      "loss": 0.4416,
+      "step": 5441
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.375719252416167,
+      "learning_rate": 8.18638043474842e-07,
+      "loss": 0.4097,
+      "step": 5442
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.2311877621263765,
+      "learning_rate": 8.172992640026072e-07,
+      "loss": 0.4376,
+      "step": 5443
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.215199756003829,
+      "learning_rate": 8.159614826888734e-07,
+      "loss": 0.404,
+      "step": 5444
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.256454462751428,
+      "learning_rate": 8.146246998528912e-07,
+      "loss": 0.452,
+      "step": 5445
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.516756407954842,
+      "learning_rate": 8.132889158136681e-07,
+      "loss": 0.374,
+      "step": 5446
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.117380328944872,
+      "learning_rate": 8.119541308899742e-07,
+      "loss": 0.4284,
+      "step": 5447
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.562529523511456,
+      "learning_rate": 8.106203454003442e-07,
+      "loss": 0.3677,
+      "step": 5448
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.869982190998707,
+      "learning_rate": 8.092875596630722e-07,
+      "loss": 0.3814,
+      "step": 5449
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.372778564510488,
+      "learning_rate": 8.079557739962129e-07,
+      "loss": 0.4585,
+      "step": 5450
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.398518841347423,
+      "learning_rate": 8.066249887175837e-07,
+      "loss": 0.4737,
+      "step": 5451
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0279437760131922,
+      "learning_rate": 8.052952041447614e-07,
+      "loss": 0.4857,
+      "step": 5452
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.196641811159402,
+      "learning_rate": 8.039664205950876e-07,
+      "loss": 0.3688,
+      "step": 5453
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.303075909876602,
+      "learning_rate": 8.02638638385661e-07,
+      "loss": 0.4225,
+      "step": 5454
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.992061594200957,
+      "learning_rate": 8.013118578333451e-07,
+      "loss": 0.4252,
+      "step": 5455
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.462497695830738,
+      "learning_rate": 7.999860792547609e-07,
+      "loss": 0.5213,
+      "step": 5456
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.308784006368848,
+      "learning_rate": 7.986613029662915e-07,
+      "loss": 0.4743,
+      "step": 5457
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.457188862158105,
+      "learning_rate": 7.973375292840835e-07,
+      "loss": 0.3949,
+      "step": 5458
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 12.912844136822514,
+      "learning_rate": 7.960147585240396e-07,
+      "loss": 0.4804,
+      "step": 5459
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 20.159357480109776,
+      "learning_rate": 7.946929910018264e-07,
+      "loss": 0.4379,
+      "step": 5460
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.395285434474368,
+      "learning_rate": 7.933722270328698e-07,
+      "loss": 0.3389,
+      "step": 5461
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 11.39123696763879,
+      "learning_rate": 7.920524669323554e-07,
+      "loss": 0.459,
+      "step": 5462
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.572361982268138,
+      "learning_rate": 7.907337110152324e-07,
+      "loss": 0.376,
+      "step": 5463
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.425311075581576,
+      "learning_rate": 7.894159595962065e-07,
+      "loss": 0.3773,
+      "step": 5464
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.060761149606728,
+      "learning_rate": 7.88099212989748e-07,
+      "loss": 0.4358,
+      "step": 5465
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.917067035064541,
+      "learning_rate": 7.86783471510083e-07,
+      "loss": 0.4486,
+      "step": 5466
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.3500289954649105,
+      "learning_rate": 7.854687354711993e-07,
+      "loss": 0.4813,
+      "step": 5467
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 23.2102581882762,
+      "learning_rate": 7.841550051868469e-07,
+      "loss": 0.4012,
+      "step": 5468
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 12.970681113446565,
+      "learning_rate": 7.828422809705316e-07,
+      "loss": 0.5034,
+      "step": 5469
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 12.726628095879253,
+      "learning_rate": 7.815305631355241e-07,
+      "loss": 0.466,
+      "step": 5470
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.588893959587628,
+      "learning_rate": 7.802198519948517e-07,
+      "loss": 0.4592,
+      "step": 5471
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.701525593299186,
+      "learning_rate": 7.789101478613015e-07,
+      "loss": 0.4248,
+      "step": 5472
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1242544254659539,
+      "learning_rate": 7.776014510474211e-07,
+      "loss": 0.4944,
+      "step": 5473
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.148906692264298,
+      "learning_rate": 7.762937618655164e-07,
+      "loss": 0.4348,
+      "step": 5474
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.885614452350353,
+      "learning_rate": 7.74987080627656e-07,
+      "loss": 0.472,
+      "step": 5475
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.520510608247074,
+      "learning_rate": 7.73681407645665e-07,
+      "loss": 0.5044,
+      "step": 5476
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.961601093829659,
+      "learning_rate": 7.72376743231128e-07,
+      "loss": 0.3834,
+      "step": 5477
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.950733669095644,
+      "learning_rate": 7.71073087695392e-07,
+      "loss": 0.4509,
+      "step": 5478
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.714956890221021,
+      "learning_rate": 7.697704413495577e-07,
+      "loss": 0.4424,
+      "step": 5479
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.013422939968912,
+      "learning_rate": 7.684688045044919e-07,
+      "loss": 0.5245,
+      "step": 5480
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 24.399389602411276,
+      "learning_rate": 7.671681774708145e-07,
+      "loss": 0.4571,
+      "step": 5481
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 15.102592354188358,
+      "learning_rate": 7.658685605589067e-07,
+      "loss": 0.4404,
+      "step": 5482
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 24.226653396307476,
+      "learning_rate": 7.645699540789103e-07,
+      "loss": 0.401,
+      "step": 5483
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2530389737284793,
+      "learning_rate": 7.632723583407231e-07,
+      "loss": 0.5657,
+      "step": 5484
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.265824344163817,
+      "learning_rate": 7.619757736540034e-07,
+      "loss": 0.3777,
+      "step": 5485
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.832633498463991,
+      "learning_rate": 7.606802003281682e-07,
+      "loss": 0.422,
+      "step": 5486
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 13.691435534783574,
+      "learning_rate": 7.593856386723902e-07,
+      "loss": 0.4799,
+      "step": 5487
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.2553935562674745,
+      "learning_rate": 7.580920889956073e-07,
+      "loss": 0.3921,
+      "step": 5488
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.938674433190346,
+      "learning_rate": 7.56799551606508e-07,
+      "loss": 0.4451,
+      "step": 5489
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1597192554984277,
+      "learning_rate": 7.555080268135461e-07,
+      "loss": 0.4974,
+      "step": 5490
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.908251298624413,
+      "learning_rate": 7.542175149249298e-07,
+      "loss": 0.4622,
+      "step": 5491
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.112911582078047,
+      "learning_rate": 7.529280162486246e-07,
+      "loss": 0.4138,
+      "step": 5492
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.804774232295548,
+      "learning_rate": 7.516395310923585e-07,
+      "loss": 0.4225,
+      "step": 5493
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 36.71552951418733,
+      "learning_rate": 7.503520597636149e-07,
+      "loss": 0.445,
+      "step": 5494
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.42487675676883,
+      "learning_rate": 7.490656025696346e-07,
+      "loss": 0.4624,
+      "step": 5495
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.14948136468449,
+      "learning_rate": 7.477801598174183e-07,
+      "loss": 0.4288,
+      "step": 5496
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.59903812856341,
+      "learning_rate": 7.464957318137217e-07,
+      "loss": 0.4074,
+      "step": 5497
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1234741500879277,
+      "learning_rate": 7.452123188650629e-07,
+      "loss": 0.5009,
+      "step": 5498
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 19.074137615173424,
+      "learning_rate": 7.439299212777135e-07,
+      "loss": 0.3728,
+      "step": 5499
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.071198476218797,
+      "learning_rate": 7.426485393577054e-07,
+      "loss": 0.404,
+      "step": 5500
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 154.97052663751754,
+      "learning_rate": 7.413681734108275e-07,
+      "loss": 0.4382,
+      "step": 5501
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.052756220088788,
+      "learning_rate": 7.40088823742624e-07,
+      "loss": 0.3785,
+      "step": 5502
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.857804008926568,
+      "learning_rate": 7.388104906584015e-07,
+      "loss": 0.4741,
+      "step": 5503
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.871414569430011,
+      "learning_rate": 7.375331744632181e-07,
+      "loss": 0.4818,
+      "step": 5504
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.571135352751194,
+      "learning_rate": 7.362568754618943e-07,
+      "loss": 0.3936,
+      "step": 5505
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.838243795593858,
+      "learning_rate": 7.34981593959005e-07,
+      "loss": 0.4177,
+      "step": 5506
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.847496226071302,
+      "learning_rate": 7.337073302588832e-07,
+      "loss": 0.3739,
+      "step": 5507
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.256660253561318,
+      "learning_rate": 7.324340846656187e-07,
+      "loss": 0.4332,
+      "step": 5508
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.836791684332274,
+      "learning_rate": 7.31161857483057e-07,
+      "loss": 0.424,
+      "step": 5509
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.574437513010073,
+      "learning_rate": 7.29890649014805e-07,
+      "loss": 0.4879,
+      "step": 5510
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.273232357956443,
+      "learning_rate": 7.286204595642216e-07,
+      "loss": 0.3898,
+      "step": 5511
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 12.348302459052665,
+      "learning_rate": 7.273512894344231e-07,
+      "loss": 0.3724,
+      "step": 5512
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.039736219026485,
+      "learning_rate": 7.260831389282874e-07,
+      "loss": 0.3652,
+      "step": 5513
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.035678171431481,
+      "learning_rate": 7.248160083484423e-07,
+      "loss": 0.4277,
+      "step": 5514
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.685298921132938,
+      "learning_rate": 7.235498979972777e-07,
+      "loss": 0.3904,
+      "step": 5515
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.510818244015671,
+      "learning_rate": 7.222848081769374e-07,
+      "loss": 0.3888,
+      "step": 5516
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.001664522982363,
+      "learning_rate": 7.21020739189322e-07,
+      "loss": 0.4183,
+      "step": 5517
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.3508437427253845,
+      "learning_rate": 7.197576913360877e-07,
+      "loss": 0.4485,
+      "step": 5518
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.210353999241169,
+      "learning_rate": 7.184956649186475e-07,
+      "loss": 0.5199,
+      "step": 5519
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 21.3927844333565,
+      "learning_rate": 7.17234660238173e-07,
+      "loss": 0.4009,
+      "step": 5520
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1176423681492933,
+      "learning_rate": 7.159746775955889e-07,
+      "loss": 0.481,
+      "step": 5521
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.959402327989377,
+      "learning_rate": 7.147157172915764e-07,
+      "loss": 0.4296,
+      "step": 5522
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.153409803049946,
+      "learning_rate": 7.134577796265746e-07,
+      "loss": 0.4894,
+      "step": 5523
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 21.96269428148758,
+      "learning_rate": 7.122008649007761e-07,
+      "loss": 0.4399,
+      "step": 5524
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.381761205361828,
+      "learning_rate": 7.109449734141328e-07,
+      "loss": 0.4067,
+      "step": 5525
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 14.266455533689362,
+      "learning_rate": 7.096901054663491e-07,
+      "loss": 0.4407,
+      "step": 5526
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 19.85916593862597,
+      "learning_rate": 7.084362613568852e-07,
+      "loss": 0.536,
+      "step": 5527
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 14.701483063656545,
+      "learning_rate": 7.071834413849599e-07,
+      "loss": 0.4339,
+      "step": 5528
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 13.620920958147204,
+      "learning_rate": 7.059316458495446e-07,
+      "loss": 0.3993,
+      "step": 5529
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.3607575525173115,
+      "learning_rate": 7.046808750493683e-07,
+      "loss": 0.4513,
+      "step": 5530
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.131029464001363,
+      "learning_rate": 7.034311292829144e-07,
+      "loss": 0.4068,
+      "step": 5531
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.303302720631565,
+      "learning_rate": 7.021824088484202e-07,
+      "loss": 0.4474,
+      "step": 5532
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.091218172696102,
+      "learning_rate": 7.009347140438821e-07,
+      "loss": 0.3944,
+      "step": 5533
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.956084420562494,
+      "learning_rate": 6.996880451670484e-07,
+      "loss": 0.4214,
+      "step": 5534
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.378505163929583,
+      "learning_rate": 6.984424025154252e-07,
+      "loss": 0.4888,
+      "step": 5535
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.43662923810866,
+      "learning_rate": 6.971977863862717e-07,
+      "loss": 0.4234,
+      "step": 5536
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.400655742768366,
+      "learning_rate": 6.95954197076601e-07,
+      "loss": 0.3862,
+      "step": 5537
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.147481212659574,
+      "learning_rate": 6.947116348831856e-07,
+      "loss": 0.485,
+      "step": 5538
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.704196513212135,
+      "learning_rate": 6.934701001025479e-07,
+      "loss": 0.4243,
+      "step": 5539
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.044711986869853,
+      "learning_rate": 6.922295930309691e-07,
+      "loss": 0.4583,
+      "step": 5540
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.599515668062049,
+      "learning_rate": 6.909901139644831e-07,
+      "loss": 0.4241,
+      "step": 5541
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 19.41049883453156,
+      "learning_rate": 6.897516631988787e-07,
+      "loss": 0.4753,
+      "step": 5542
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.379022583398443,
+      "learning_rate": 6.885142410296991e-07,
+      "loss": 0.3576,
+      "step": 5543
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.666542407232468,
+      "learning_rate": 6.872778477522412e-07,
+      "loss": 0.3911,
+      "step": 5544
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.493434267439428,
+      "learning_rate": 6.860424836615604e-07,
+      "loss": 0.4475,
+      "step": 5545
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.392206695963077,
+      "learning_rate": 6.848081490524616e-07,
+      "loss": 0.4431,
+      "step": 5546
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.928015810422853,
+      "learning_rate": 6.835748442195062e-07,
+      "loss": 0.3892,
+      "step": 5547
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.121054453843085,
+      "learning_rate": 6.823425694570101e-07,
+      "loss": 0.3908,
+      "step": 5548
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 16.506259053732673,
+      "learning_rate": 6.811113250590429e-07,
+      "loss": 0.4186,
+      "step": 5549
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.609310226575656,
+      "learning_rate": 6.798811113194286e-07,
+      "loss": 0.4155,
+      "step": 5550
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 14.296086674112372,
+      "learning_rate": 6.786519285317455e-07,
+      "loss": 0.3894,
+      "step": 5551
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.827772775845816,
+      "learning_rate": 6.77423776989325e-07,
+      "loss": 0.489,
+      "step": 5552
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.082914081382292,
+      "learning_rate": 6.761966569852523e-07,
+      "loss": 0.4827,
+      "step": 5553
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 25.851530539859105,
+      "learning_rate": 6.749705688123665e-07,
+      "loss": 0.3872,
+      "step": 5554
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.137169488944476,
+      "learning_rate": 6.737455127632625e-07,
+      "loss": 0.4535,
+      "step": 5555
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4446082073226152,
+      "learning_rate": 6.725214891302873e-07,
+      "loss": 0.5627,
+      "step": 5556
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 30.26879326773742,
+      "learning_rate": 6.712984982055393e-07,
+      "loss": 0.4265,
+      "step": 5557
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.98885768070102,
+      "learning_rate": 6.700765402808751e-07,
+      "loss": 0.402,
+      "step": 5558
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.670090435308791,
+      "learning_rate": 6.688556156479021e-07,
+      "loss": 0.3914,
+      "step": 5559
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.643360809748664,
+      "learning_rate": 6.676357245979792e-07,
+      "loss": 0.4076,
+      "step": 5560
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.172960413519421,
+      "learning_rate": 6.664168674222238e-07,
+      "loss": 0.4046,
+      "step": 5561
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.234849693930766,
+      "learning_rate": 6.651990444115014e-07,
+      "loss": 0.4168,
+      "step": 5562
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 12.464880376582926,
+      "learning_rate": 6.639822558564346e-07,
+      "loss": 0.3797,
+      "step": 5563
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 28.060458943835744,
+      "learning_rate": 6.627665020473966e-07,
+      "loss": 0.4054,
+      "step": 5564
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 18.681770129167187,
+      "learning_rate": 6.61551783274515e-07,
+      "loss": 0.4515,
+      "step": 5565
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 19.642670452947392,
+      "learning_rate": 6.603380998276699e-07,
+      "loss": 0.4188,
+      "step": 5566
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 12.646286709544395,
+      "learning_rate": 6.591254519964924e-07,
+      "loss": 0.4279,
+      "step": 5567
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.285966288290398,
+      "learning_rate": 6.579138400703716e-07,
+      "loss": 0.4581,
+      "step": 5568
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 27.18120613476007,
+      "learning_rate": 6.567032643384441e-07,
+      "loss": 0.4479,
+      "step": 5569
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.970165211670107,
+      "learning_rate": 6.554937250896015e-07,
+      "loss": 0.4604,
+      "step": 5570
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.083272600231045,
+      "learning_rate": 6.542852226124891e-07,
+      "loss": 0.4189,
+      "step": 5571
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.233899474832337,
+      "learning_rate": 6.530777571955021e-07,
+      "loss": 0.3937,
+      "step": 5572
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 11.219075539303324,
+      "learning_rate": 6.518713291267914e-07,
+      "loss": 0.4151,
+      "step": 5573
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.182827857968582,
+      "learning_rate": 6.506659386942576e-07,
+      "loss": 0.4477,
+      "step": 5574
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 12.858608685374824,
+      "learning_rate": 6.494615861855536e-07,
+      "loss": 0.4441,
+      "step": 5575
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 13.343200874300361,
+      "learning_rate": 6.482582718880887e-07,
+      "loss": 0.412,
+      "step": 5576
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 19.636744994451625,
+      "learning_rate": 6.470559960890188e-07,
+      "loss": 0.4431,
+      "step": 5577
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 24.97897313528571,
+      "learning_rate": 6.458547590752562e-07,
+      "loss": 0.4364,
+      "step": 5578
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 20.482634387962907,
+      "learning_rate": 6.446545611334631e-07,
+      "loss": 0.4387,
+      "step": 5579
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 13.992366881230474,
+      "learning_rate": 6.434554025500533e-07,
+      "loss": 0.4134,
+      "step": 5580
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.465793524023145,
+      "learning_rate": 6.422572836111957e-07,
+      "loss": 0.4027,
+      "step": 5581
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.805258766383043,
+      "learning_rate": 6.410602046028075e-07,
+      "loss": 0.4268,
+      "step": 5582
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.343037177447048,
+      "learning_rate": 6.398641658105603e-07,
+      "loss": 0.3878,
+      "step": 5583
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.144721756354544,
+      "learning_rate": 6.386691675198764e-07,
+      "loss": 0.4519,
+      "step": 5584
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.88427942944416,
+      "learning_rate": 6.374752100159287e-07,
+      "loss": 0.4038,
+      "step": 5585
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 16.222589234824692,
+      "learning_rate": 6.362822935836449e-07,
+      "loss": 0.4105,
+      "step": 5586
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 21.00806368749471,
+      "learning_rate": 6.350904185077006e-07,
+      "loss": 0.4303,
+      "step": 5587
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.679518334885346,
+      "learning_rate": 6.338995850725255e-07,
+      "loss": 0.4337,
+      "step": 5588
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2148912408716686,
+      "learning_rate": 6.327097935622988e-07,
+      "loss": 0.5058,
+      "step": 5589
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.117702754912248,
+      "learning_rate": 6.315210442609515e-07,
+      "loss": 0.4407,
+      "step": 5590
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.923260996879476,
+      "learning_rate": 6.303333374521687e-07,
+      "loss": 0.3506,
+      "step": 5591
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 13.262861983087861,
+      "learning_rate": 6.291466734193818e-07,
+      "loss": 0.4489,
+      "step": 5592
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.180824567897123,
+      "learning_rate": 6.27961052445778e-07,
+      "loss": 0.3984,
+      "step": 5593
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.675765574744439,
+      "learning_rate": 6.26776474814293e-07,
+      "loss": 0.4685,
+      "step": 5594
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1748222223127798,
+      "learning_rate": 6.255929408076128e-07,
+      "loss": 0.5008,
+      "step": 5595
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.337558067729937,
+      "learning_rate": 6.244104507081778e-07,
+      "loss": 0.4225,
+      "step": 5596
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 18.087740993517055,
+      "learning_rate": 6.232290047981743e-07,
+      "loss": 0.4671,
+      "step": 5597
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.789449449626573,
+      "learning_rate": 6.220486033595452e-07,
+      "loss": 0.5024,
+      "step": 5598
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 12.990060295659438,
+      "learning_rate": 6.208692466739802e-07,
+      "loss": 0.4763,
+      "step": 5599
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 11.045174501733419,
+      "learning_rate": 6.196909350229197e-07,
+      "loss": 0.4317,
+      "step": 5600
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 15.685068973750885,
+      "learning_rate": 6.185136686875565e-07,
+      "loss": 0.3601,
+      "step": 5601
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.968061600331163,
+      "learning_rate": 6.173374479488315e-07,
+      "loss": 0.4069,
+      "step": 5602
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 14.901855997486388,
+      "learning_rate": 6.161622730874406e-07,
+      "loss": 0.4526,
+      "step": 5603
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.530263102298733,
+      "learning_rate": 6.149881443838252e-07,
+      "loss": 0.4611,
+      "step": 5604
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 11.326347898171674,
+      "learning_rate": 6.138150621181782e-07,
+      "loss": 0.457,
+      "step": 5605
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.351497553305427,
+      "learning_rate": 6.126430265704459e-07,
+      "loss": 0.4792,
+      "step": 5606
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.50017692252538,
+      "learning_rate": 6.114720380203204e-07,
+      "loss": 0.4717,
+      "step": 5607
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.374019015791695,
+      "learning_rate": 6.103020967472484e-07,
+      "loss": 0.4664,
+      "step": 5608
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2256092850377978,
+      "learning_rate": 6.091332030304231e-07,
+      "loss": 0.5165,
+      "step": 5609
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 24.563565165765784,
+      "learning_rate": 6.079653571487875e-07,
+      "loss": 0.3516,
+      "step": 5610
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.611685673779102,
+      "learning_rate": 6.067985593810389e-07,
+      "loss": 0.3764,
+      "step": 5611
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 13.915056033764408,
+      "learning_rate": 6.056328100056197e-07,
+      "loss": 0.3612,
+      "step": 5612
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.533147516021703,
+      "learning_rate": 6.044681093007249e-07,
+      "loss": 0.4666,
+      "step": 5613
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 13.976486293622372,
+      "learning_rate": 6.033044575442975e-07,
+      "loss": 0.4071,
+      "step": 5614
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.51258030395662,
+      "learning_rate": 6.021418550140307e-07,
+      "loss": 0.4768,
+      "step": 5615
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.669433300531895,
+      "learning_rate": 6.009803019873689e-07,
+      "loss": 0.406,
+      "step": 5616
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 58.62598027916851,
+      "learning_rate": 5.998197987415033e-07,
+      "loss": 0.4652,
+      "step": 5617
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.438668231595424,
+      "learning_rate": 5.98660345553378e-07,
+      "loss": 0.4624,
+      "step": 5618
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.280991923028768,
+      "learning_rate": 5.975019426996837e-07,
+      "loss": 0.5044,
+      "step": 5619
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.383664618119922,
+      "learning_rate": 5.963445904568599e-07,
+      "loss": 0.3666,
+      "step": 5620
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 17.08530571967915,
+      "learning_rate": 5.95188289101099e-07,
+      "loss": 0.4836,
+      "step": 5621
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.776040278474298,
+      "learning_rate": 5.940330389083399e-07,
+      "loss": 0.5028,
+      "step": 5622
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.611438755357886,
+      "learning_rate": 5.928788401542706e-07,
+      "loss": 0.4252,
+      "step": 5623
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.6247140275729794,
+      "learning_rate": 5.917256931143289e-07,
+      "loss": 0.3918,
+      "step": 5624
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 39.4693845621563,
+      "learning_rate": 5.905735980637001e-07,
+      "loss": 0.3979,
+      "step": 5625
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.758711497395522,
+      "learning_rate": 5.894225552773225e-07,
+      "loss": 0.4984,
+      "step": 5626
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 12.665761726618195,
+      "learning_rate": 5.882725650298787e-07,
+      "loss": 0.4493,
+      "step": 5627
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2716606821246144,
+      "learning_rate": 5.871236275958031e-07,
+      "loss": 0.5166,
+      "step": 5628
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 32.90641965638817,
+      "learning_rate": 5.85975743249278e-07,
+      "loss": 0.4303,
+      "step": 5629
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.647120613957467,
+      "learning_rate": 5.84828912264232e-07,
+      "loss": 0.402,
+      "step": 5630
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.04465802355464,
+      "learning_rate": 5.836831349143473e-07,
+      "loss": 0.5031,
+      "step": 5631
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 30.9031687896907,
+      "learning_rate": 5.825384114730498e-07,
+      "loss": 0.3993,
+      "step": 5632
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.1514732297621535,
+      "learning_rate": 5.81394742213518e-07,
+      "loss": 0.4274,
+      "step": 5633
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.150250806909699,
+      "learning_rate": 5.802521274086754e-07,
+      "loss": 0.4055,
+      "step": 5634
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.354934230114342,
+      "learning_rate": 5.791105673311959e-07,
+      "loss": 0.4529,
+      "step": 5635
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 13.724247803012537,
+      "learning_rate": 5.779700622535001e-07,
+      "loss": 0.4861,
+      "step": 5636
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 17.093153718017877,
+      "learning_rate": 5.768306124477574e-07,
+      "loss": 0.4123,
+      "step": 5637
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.234835527298902,
+      "learning_rate": 5.756922181858881e-07,
+      "loss": 0.4443,
+      "step": 5638
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.9940439274297175,
+      "learning_rate": 5.745548797395567e-07,
+      "loss": 0.4067,
+      "step": 5639
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.470793971543523,
+      "learning_rate": 5.73418597380177e-07,
+      "loss": 0.4463,
+      "step": 5640
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.262088099724101,
+      "learning_rate": 5.722833713789122e-07,
+      "loss": 0.3918,
+      "step": 5641
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 15.084521902574652,
+      "learning_rate": 5.711492020066711e-07,
+      "loss": 0.4437,
+      "step": 5642
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.452488994968562,
+      "learning_rate": 5.700160895341133e-07,
+      "loss": 0.4064,
+      "step": 5643
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 16.007820558833274,
+      "learning_rate": 5.688840342316431e-07,
+      "loss": 0.3816,
+      "step": 5644
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.8515851911073,
+      "learning_rate": 5.677530363694139e-07,
+      "loss": 0.3804,
+      "step": 5645
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.536552095346227,
+      "learning_rate": 5.666230962173275e-07,
+      "loss": 0.4057,
+      "step": 5646
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 11.59117474445883,
+      "learning_rate": 5.654942140450304e-07,
+      "loss": 0.465,
+      "step": 5647
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.086468860328539,
+      "learning_rate": 5.643663901219215e-07,
+      "loss": 0.3988,
+      "step": 5648
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.272839910671308,
+      "learning_rate": 5.632396247171429e-07,
+      "loss": 0.4426,
+      "step": 5649
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.882336577814514,
+      "learning_rate": 5.621139180995849e-07,
+      "loss": 0.4801,
+      "step": 5650
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.777267751360537,
+      "learning_rate": 5.609892705378878e-07,
+      "loss": 0.4407,
+      "step": 5651
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.178585486453405,
+      "learning_rate": 5.59865682300435e-07,
+      "loss": 0.4972,
+      "step": 5652
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.565976527795983,
+      "learning_rate": 5.587431536553611e-07,
+      "loss": 0.3992,
+      "step": 5653
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.176518295115487,
+      "learning_rate": 5.576216848705452e-07,
+      "loss": 0.4311,
+      "step": 5654
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 118.30770608181587,
+      "learning_rate": 5.565012762136135e-07,
+      "loss": 0.4091,
+      "step": 5655
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 17.948241797255406,
+      "learning_rate": 5.553819279519418e-07,
+      "loss": 0.4411,
+      "step": 5656
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.898729305891955,
+      "learning_rate": 5.542636403526503e-07,
+      "loss": 0.4756,
+      "step": 5657
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.941702124890432,
+      "learning_rate": 5.531464136826064e-07,
+      "loss": 0.3906,
+      "step": 5658
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 12.289334493421135,
+      "learning_rate": 5.52030248208425e-07,
+      "loss": 0.4139,
+      "step": 5659
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.592136752729822,
+      "learning_rate": 5.509151441964666e-07,
+      "loss": 0.3875,
+      "step": 5660
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 17.93025674198459,
+      "learning_rate": 5.498011019128407e-07,
+      "loss": 0.4285,
+      "step": 5661
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.020766477949305,
+      "learning_rate": 5.486881216234008e-07,
+      "loss": 0.3646,
+      "step": 5662
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 84.18799922370366,
+      "learning_rate": 5.475762035937499e-07,
+      "loss": 0.4056,
+      "step": 5663
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.437349271202502,
+      "learning_rate": 5.464653480892351e-07,
+      "loss": 0.4835,
+      "step": 5664
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.29269975766321,
+      "learning_rate": 5.45355555374949e-07,
+      "loss": 0.4377,
+      "step": 5665
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.721969389061686,
+      "learning_rate": 5.442468257157346e-07,
+      "loss": 0.3924,
+      "step": 5666
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.421541353933005,
+      "learning_rate": 5.431391593761764e-07,
+      "loss": 0.3613,
+      "step": 5667
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.937045483059979,
+      "learning_rate": 5.420325566206102e-07,
+      "loss": 0.4931,
+      "step": 5668
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.579533209791268,
+      "learning_rate": 5.409270177131143e-07,
+      "loss": 0.4375,
+      "step": 5669
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.613406449542559,
+      "learning_rate": 5.398225429175142e-07,
+      "loss": 0.4224,
+      "step": 5670
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 12.797931599574172,
+      "learning_rate": 5.387191324973806e-07,
+      "loss": 0.4917,
+      "step": 5671
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.60346318774248,
+      "learning_rate": 5.376167867160326e-07,
+      "loss": 0.329,
+      "step": 5672
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 13.170520445355194,
+      "learning_rate": 5.365155058365312e-07,
+      "loss": 0.3761,
+      "step": 5673
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.249832483312413,
+      "learning_rate": 5.35415290121688e-07,
+      "loss": 0.4405,
+      "step": 5674
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.567058518041789,
+      "learning_rate": 5.343161398340568e-07,
+      "loss": 0.434,
+      "step": 5675
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 10.647481148075773,
+      "learning_rate": 5.332180552359406e-07,
+      "loss": 0.3875,
+      "step": 5676
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.7095035020816125,
+      "learning_rate": 5.321210365893842e-07,
+      "loss": 0.3879,
+      "step": 5677
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.239415784637044,
+      "learning_rate": 5.310250841561792e-07,
+      "loss": 0.4681,
+      "step": 5678
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.866506980445486,
+      "learning_rate": 5.299301981978655e-07,
+      "loss": 0.4115,
+      "step": 5679
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 14.942293573647621,
+      "learning_rate": 5.28836378975725e-07,
+      "loss": 0.4461,
+      "step": 5680
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 18.9784239148245,
+      "learning_rate": 5.277436267507868e-07,
+      "loss": 0.3994,
+      "step": 5681
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.025737393460759,
+      "learning_rate": 5.26651941783825e-07,
+      "loss": 0.449,
+      "step": 5682
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 10.36701485000101,
+      "learning_rate": 5.255613243353575e-07,
+      "loss": 0.4616,
+      "step": 5683
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.033426528217516,
+      "learning_rate": 5.244717746656513e-07,
+      "loss": 0.4316,
+      "step": 5684
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.979353206368469,
+      "learning_rate": 5.233832930347149e-07,
+      "loss": 0.4736,
+      "step": 5685
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.487124641691687,
+      "learning_rate": 5.222958797023036e-07,
+      "loss": 0.4213,
+      "step": 5686
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.479140295090433,
+      "learning_rate": 5.212095349279179e-07,
+      "loss": 0.4897,
+      "step": 5687
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 14.612605987349252,
+      "learning_rate": 5.201242589708011e-07,
+      "loss": 0.4276,
+      "step": 5688
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 16.4484802234681,
+      "learning_rate": 5.190400520899452e-07,
+      "loss": 0.414,
+      "step": 5689
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.033005193540101,
+      "learning_rate": 5.179569145440827e-07,
+      "loss": 0.4161,
+      "step": 5690
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.195047327074397,
+      "learning_rate": 5.168748465916961e-07,
+      "loss": 0.4376,
+      "step": 5691
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.181609040873978,
+      "learning_rate": 5.15793848491008e-07,
+      "loss": 0.4019,
+      "step": 5692
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.950056359607112,
+      "learning_rate": 5.147139204999884e-07,
+      "loss": 0.414,
+      "step": 5693
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.482197610920505,
+      "learning_rate": 5.136350628763498e-07,
+      "loss": 0.3904,
+      "step": 5694
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.979316039537605,
+      "learning_rate": 5.125572758775505e-07,
+      "loss": 0.4267,
+      "step": 5695
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.746346609027996,
+      "learning_rate": 5.114805597607942e-07,
+      "loss": 0.4052,
+      "step": 5696
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.75101781866843,
+      "learning_rate": 5.104049147830281e-07,
+      "loss": 0.4353,
+      "step": 5697
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.191257214619394,
+      "learning_rate": 5.093303412009426e-07,
+      "loss": 0.3919,
+      "step": 5698
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.164935376401182,
+      "learning_rate": 5.082568392709758e-07,
+      "loss": 0.4103,
+      "step": 5699
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 27.3851666585,
+      "learning_rate": 5.071844092493051e-07,
+      "loss": 0.3764,
+      "step": 5700
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.973146889492861,
+      "learning_rate": 5.061130513918577e-07,
+      "loss": 0.461,
+      "step": 5701
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.467336534638455,
+      "learning_rate": 5.050427659543006e-07,
+      "loss": 0.4437,
+      "step": 5702
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1466089920999127,
+      "learning_rate": 5.039735531920453e-07,
+      "loss": 0.5031,
+      "step": 5703
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.949478849548411,
+      "learning_rate": 5.029054133602512e-07,
+      "loss": 0.4773,
+      "step": 5704
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.72311772445394,
+      "learning_rate": 5.018383467138177e-07,
+      "loss": 0.4232,
+      "step": 5705
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.80274400159251,
+      "learning_rate": 5.007723535073888e-07,
+      "loss": 0.4374,
+      "step": 5706
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.8926207686313745,
+      "learning_rate": 4.997074339953528e-07,
+      "loss": 0.4657,
+      "step": 5707
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 27.695058194721316,
+      "learning_rate": 4.986435884318413e-07,
+      "loss": 0.4128,
+      "step": 5708
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.430352442889728,
+      "learning_rate": 4.975808170707314e-07,
+      "loss": 0.4633,
+      "step": 5709
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.54661209309096,
+      "learning_rate": 4.965191201656417e-07,
+      "loss": 0.4296,
+      "step": 5710
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 12.390199704193106,
+      "learning_rate": 4.954584979699356e-07,
+      "loss": 0.362,
+      "step": 5711
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 79.10493654650362,
+      "learning_rate": 4.943989507367203e-07,
+      "loss": 0.4672,
+      "step": 5712
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 37.57192897845998,
+      "learning_rate": 4.933404787188439e-07,
+      "loss": 0.3725,
+      "step": 5713
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.227387249385961,
+      "learning_rate": 4.922830821689023e-07,
+      "loss": 0.3913,
+      "step": 5714
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.530733664575141,
+      "learning_rate": 4.912267613392307e-07,
+      "loss": 0.4059,
+      "step": 5715
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 12.794248794476655,
+      "learning_rate": 4.901715164819099e-07,
+      "loss": 0.4061,
+      "step": 5716
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 15.931633078597596,
+      "learning_rate": 4.891173478487626e-07,
+      "loss": 0.3957,
+      "step": 5717
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 12.677477570007873,
+      "learning_rate": 4.88064255691355e-07,
+      "loss": 0.4207,
+      "step": 5718
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.055336745677284,
+      "learning_rate": 4.870122402609984e-07,
+      "loss": 0.3826,
+      "step": 5719
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 15.305161504303738,
+      "learning_rate": 4.859613018087434e-07,
+      "loss": 0.4033,
+      "step": 5720
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 10.299566564219118,
+      "learning_rate": 4.849114405853872e-07,
+      "loss": 0.4242,
+      "step": 5721
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.560275551971317,
+      "learning_rate": 4.838626568414684e-07,
+      "loss": 0.4527,
+      "step": 5722
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.34603021045852,
+      "learning_rate": 4.828149508272661e-07,
+      "loss": 0.3828,
+      "step": 5723
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2624194968561506,
+      "learning_rate": 4.817683227928083e-07,
+      "loss": 0.4778,
+      "step": 5724
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.74923902802455,
+      "learning_rate": 4.807227729878583e-07,
+      "loss": 0.3932,
+      "step": 5725
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.131562637663482,
+      "learning_rate": 4.796783016619295e-07,
+      "loss": 0.4531,
+      "step": 5726
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.802769459316018,
+      "learning_rate": 4.786349090642711e-07,
+      "loss": 0.4297,
+      "step": 5727
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.03984817667625,
+      "learning_rate": 4.775925954438798e-07,
+      "loss": 0.4012,
+      "step": 5728
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.357935921487051,
+      "learning_rate": 4.7655136104949264e-07,
+      "loss": 0.4575,
+      "step": 5729
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.05842934765025,
+      "learning_rate": 4.7551120612958803e-07,
+      "loss": 0.4345,
+      "step": 5730
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 10.445141783594204,
+      "learning_rate": 4.7447213093239077e-07,
+      "loss": 0.328,
+      "step": 5731
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.431366136855896,
+      "learning_rate": 4.7343413570586415e-07,
+      "loss": 0.4746,
+      "step": 5732
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.549995759024563,
+      "learning_rate": 4.72397220697714e-07,
+      "loss": 0.3851,
+      "step": 5733
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.1655662236003606,
+      "learning_rate": 4.7136138615539116e-07,
+      "loss": 0.4136,
+      "step": 5734
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.755325717763183,
+      "learning_rate": 4.703266323260858e-07,
+      "loss": 0.4618,
+      "step": 5735
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 16.90871352422481,
+      "learning_rate": 4.692929594567325e-07,
+      "loss": 0.4791,
+      "step": 5736
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 18.251547897826573,
+      "learning_rate": 4.682603677940062e-07,
+      "loss": 0.4507,
+      "step": 5737
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 12.050366827926503,
+      "learning_rate": 4.672288575843226e-07,
+      "loss": 0.3917,
+      "step": 5738
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.534185912978316,
+      "learning_rate": 4.661984290738436e-07,
+      "loss": 0.3498,
+      "step": 5739
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.068298797850447,
+      "learning_rate": 4.651690825084687e-07,
+      "loss": 0.4057,
+      "step": 5740
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.56245523703993,
+      "learning_rate": 4.6414081813384193e-07,
+      "loss": 0.358,
+      "step": 5741
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 11.15730740651848,
+      "learning_rate": 4.63113636195347e-07,
+      "loss": 0.4554,
+      "step": 5742
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.9321532360970775,
+      "learning_rate": 4.6208753693810946e-07,
+      "loss": 0.3842,
+      "step": 5743
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 61.35444380576177,
+      "learning_rate": 4.6106252060699965e-07,
+      "loss": 0.4292,
+      "step": 5744
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.2972561593342276,
+      "learning_rate": 4.6003858744662564e-07,
+      "loss": 0.4482,
+      "step": 5745
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.084508613070108,
+      "learning_rate": 4.5901573770133935e-07,
+      "loss": 0.5078,
+      "step": 5746
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.7949200066925775,
+      "learning_rate": 4.5799397161523316e-07,
+      "loss": 0.4514,
+      "step": 5747
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 13.620322355015267,
+      "learning_rate": 4.569732894321405e-07,
+      "loss": 0.3886,
+      "step": 5748
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.945537482859955,
+      "learning_rate": 4.559536913956375e-07,
+      "loss": 0.4519,
+      "step": 5749
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 17.524320864370292,
+      "learning_rate": 4.549351777490402e-07,
+      "loss": 0.4646,
+      "step": 5750
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.610888430315229,
+      "learning_rate": 4.539177487354063e-07,
+      "loss": 0.4867,
+      "step": 5751
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.1163939932395195,
+      "learning_rate": 4.529014045975355e-07,
+      "loss": 0.4037,
+      "step": 5752
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.113875827951045,
+      "learning_rate": 4.5188614557796605e-07,
+      "loss": 0.3851,
+      "step": 5753
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.385745072195075,
+      "learning_rate": 4.5087197191898137e-07,
+      "loss": 0.4305,
+      "step": 5754
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.546124107484307,
+      "learning_rate": 4.498588838626017e-07,
+      "loss": 0.4615,
+      "step": 5755
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.409475452055824,
+      "learning_rate": 4.488468816505914e-07,
+      "loss": 0.4435,
+      "step": 5756
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.700102155429978,
+      "learning_rate": 4.4783596552445463e-07,
+      "loss": 0.3715,
+      "step": 5757
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 31.044687684877545,
+      "learning_rate": 4.468261357254339e-07,
+      "loss": 0.3872,
+      "step": 5758
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.734995082863173,
+      "learning_rate": 4.458173924945175e-07,
+      "loss": 0.4475,
+      "step": 5759
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 15.902255381996751,
+      "learning_rate": 4.4480973607242907e-07,
+      "loss": 0.429,
+      "step": 5760
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1850681851243217,
+      "learning_rate": 4.4380316669963787e-07,
+      "loss": 0.5049,
+      "step": 5761
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 11.563051576657616,
+      "learning_rate": 4.4279768461635066e-07,
+      "loss": 0.3898,
+      "step": 5762
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.028345858062199,
+      "learning_rate": 4.417932900625144e-07,
+      "loss": 0.3766,
+      "step": 5763
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 24.969022287012418,
+      "learning_rate": 4.4078998327781787e-07,
+      "loss": 0.4169,
+      "step": 5764
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.348139426378493,
+      "learning_rate": 4.397877645016896e-07,
+      "loss": 0.4384,
+      "step": 5765
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 15.008014953805748,
+      "learning_rate": 4.387866339732999e-07,
+      "loss": 0.4192,
+      "step": 5766
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.782074008351293,
+      "learning_rate": 4.377865919315577e-07,
+      "loss": 0.4543,
+      "step": 5767
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 12.992674157927052,
+      "learning_rate": 4.367876386151115e-07,
+      "loss": 0.4539,
+      "step": 5768
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 15.700736873013351,
+      "learning_rate": 4.3578977426235346e-07,
+      "loss": 0.399,
+      "step": 5769
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.565182926446994,
+      "learning_rate": 4.3479299911141195e-07,
+      "loss": 0.3982,
+      "step": 5770
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 14.067938724301634,
+      "learning_rate": 4.337973134001588e-07,
+      "loss": 0.5046,
+      "step": 5771
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.255399802036756,
+      "learning_rate": 4.328027173662025e-07,
+      "loss": 0.4484,
+      "step": 5772
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.167122074006211,
+      "learning_rate": 4.318092112468941e-07,
+      "loss": 0.4761,
+      "step": 5773
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.019541635289777,
+      "learning_rate": 4.308167952793235e-07,
+      "loss": 0.5185,
+      "step": 5774
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.127206797273236,
+      "learning_rate": 4.29825469700319e-07,
+      "loss": 0.4355,
+      "step": 5775
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.192695657851663,
+      "learning_rate": 4.2883523474645307e-07,
+      "loss": 0.4329,
+      "step": 5776
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 25.03825766016349,
+      "learning_rate": 4.278460906540333e-07,
+      "loss": 0.3992,
+      "step": 5777
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 12.467845012877225,
+      "learning_rate": 4.2685803765910807e-07,
+      "loss": 0.4476,
+      "step": 5778
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.656219505642666,
+      "learning_rate": 4.2587107599746825e-07,
+      "loss": 0.4481,
+      "step": 5779
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.57832767715027,
+      "learning_rate": 4.248852059046399e-07,
+      "loss": 0.4296,
+      "step": 5780
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.059078139845612,
+      "learning_rate": 4.239004276158926e-07,
+      "loss": 0.4153,
+      "step": 5781
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.0779925181147565,
+      "learning_rate": 4.229167413662327e-07,
+      "loss": 0.4088,
+      "step": 5782
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.622199452157603,
+      "learning_rate": 4.2193414739040593e-07,
+      "loss": 0.3877,
+      "step": 5783
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.694962440302269,
+      "learning_rate": 4.209526459229002e-07,
+      "loss": 0.3568,
+      "step": 5784
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.895018591157557,
+      "learning_rate": 4.199722371979398e-07,
+      "loss": 0.4204,
+      "step": 5785
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.850017061556132,
+      "learning_rate": 4.1899292144948923e-07,
+      "loss": 0.3747,
+      "step": 5786
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.219309258328249,
+      "learning_rate": 4.1801469891125156e-07,
+      "loss": 0.447,
+      "step": 5787
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.101422102388584,
+      "learning_rate": 4.17037569816669e-07,
+      "loss": 0.4256,
+      "step": 5788
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.687046241994253,
+      "learning_rate": 4.160615343989255e-07,
+      "loss": 0.4552,
+      "step": 5789
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.025507936986526,
+      "learning_rate": 4.150865928909409e-07,
+      "loss": 0.4493,
+      "step": 5790
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.3397577390930815,
+      "learning_rate": 4.141127455253729e-07,
+      "loss": 0.4176,
+      "step": 5791
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.497353435800689,
+      "learning_rate": 4.1313999253462356e-07,
+      "loss": 0.5068,
+      "step": 5792
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 16.3698569289533,
+      "learning_rate": 4.1216833415082815e-07,
+      "loss": 0.4403,
+      "step": 5793
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.28682631783735,
+      "learning_rate": 4.1119777060586397e-07,
+      "loss": 0.4417,
+      "step": 5794
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 11.377348347937348,
+      "learning_rate": 4.1022830213134577e-07,
+      "loss": 0.417,
+      "step": 5795
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.180122450446617,
+      "learning_rate": 4.092599289586263e-07,
+      "loss": 0.4661,
+      "step": 5796
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 17.800636002196796,
+      "learning_rate": 4.082926513187996e-07,
+      "loss": 0.3614,
+      "step": 5797
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0663976759232696,
+      "learning_rate": 4.0732646944269593e-07,
+      "loss": 0.5574,
+      "step": 5798
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.779870943763425,
+      "learning_rate": 4.063613835608837e-07,
+      "loss": 0.4356,
+      "step": 5799
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 23.986355881970134,
+      "learning_rate": 4.053973939036715e-07,
+      "loss": 0.4697,
+      "step": 5800
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.972932561980954,
+      "learning_rate": 4.044345007011047e-07,
+      "loss": 0.3772,
+      "step": 5801
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.461362181638089,
+      "learning_rate": 4.0347270418296945e-07,
+      "loss": 0.416,
+      "step": 5802
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.183477579521882,
+      "learning_rate": 4.025120045787867e-07,
+      "loss": 0.3976,
+      "step": 5803
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.308461113007686,
+      "learning_rate": 4.0155240211781966e-07,
+      "loss": 0.4567,
+      "step": 5804
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.092301587722812,
+      "learning_rate": 4.0059389702906627e-07,
+      "loss": 0.4107,
+      "step": 5805
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.865731334096125,
+      "learning_rate": 3.99636489541263e-07,
+      "loss": 0.3818,
+      "step": 5806
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.29541235815272,
+      "learning_rate": 3.986801798828871e-07,
+      "loss": 0.4302,
+      "step": 5807
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 26.220178910674,
+      "learning_rate": 3.9772496828215214e-07,
+      "loss": 0.4538,
+      "step": 5808
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.8420321802545665,
+      "learning_rate": 3.9677085496700796e-07,
+      "loss": 0.4201,
+      "step": 5809
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.53346726518796,
+      "learning_rate": 3.958178401651447e-07,
+      "loss": 0.4284,
+      "step": 5810
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.320379191639871,
+      "learning_rate": 3.9486592410398925e-07,
+      "loss": 0.4187,
+      "step": 5811
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 19.72833264662245,
+      "learning_rate": 3.9391510701070725e-07,
+      "loss": 0.4125,
+      "step": 5812
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.502274491886736,
+      "learning_rate": 3.9296538911219984e-07,
+      "loss": 0.4858,
+      "step": 5813
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 14.742654346493456,
+      "learning_rate": 3.9201677063510967e-07,
+      "loss": 0.4259,
+      "step": 5814
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 39.58727536396428,
+      "learning_rate": 3.91069251805814e-07,
+      "loss": 0.4317,
+      "step": 5815
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.328524105195852,
+      "learning_rate": 3.9012283285042696e-07,
+      "loss": 0.4619,
+      "step": 5816
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.450612251860516,
+      "learning_rate": 3.8917751399480396e-07,
+      "loss": 0.4735,
+      "step": 5817
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 16.320421472311793,
+      "learning_rate": 3.882332954645335e-07,
+      "loss": 0.4579,
+      "step": 5818
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.970528836564768,
+      "learning_rate": 3.872901774849458e-07,
+      "loss": 0.3822,
+      "step": 5819
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.942055003037975,
+      "learning_rate": 3.863481602811048e-07,
+      "loss": 0.4,
+      "step": 5820
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2167118776089723,
+      "learning_rate": 3.854072440778134e-07,
+      "loss": 0.5192,
+      "step": 5821
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.889310368873597,
+      "learning_rate": 3.8446742909961197e-07,
+      "loss": 0.4161,
+      "step": 5822
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 11.383990877006225,
+      "learning_rate": 3.8352871557077674e-07,
+      "loss": 0.4948,
+      "step": 5823
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2002827312773676,
+      "learning_rate": 3.8259110371532347e-07,
+      "loss": 0.5312,
+      "step": 5824
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 12.2864149092543,
+      "learning_rate": 3.8165459375700263e-07,
+      "loss": 0.4097,
+      "step": 5825
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2586840835496809,
+      "learning_rate": 3.807191859193021e-07,
+      "loss": 0.5256,
+      "step": 5826
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.218857571189341,
+      "learning_rate": 3.797848804254489e-07,
+      "loss": 0.371,
+      "step": 5827
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.718419251602715,
+      "learning_rate": 3.788516774984041e-07,
+      "loss": 0.4883,
+      "step": 5828
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 13.247463609512966,
+      "learning_rate": 3.7791957736086794e-07,
+      "loss": 0.4283,
+      "step": 5829
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.450029385520013,
+      "learning_rate": 3.769885802352768e-07,
+      "loss": 0.408,
+      "step": 5830
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 15.62022109198703,
+      "learning_rate": 3.760586863438015e-07,
+      "loss": 0.5181,
+      "step": 5831
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.420797661763059,
+      "learning_rate": 3.751298959083538e-07,
+      "loss": 0.412,
+      "step": 5832
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.478791512088389,
+      "learning_rate": 3.742022091505798e-07,
+      "loss": 0.5118,
+      "step": 5833
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.649552586048902,
+      "learning_rate": 3.7327562629186133e-07,
+      "loss": 0.3923,
+      "step": 5834
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.157657199957873,
+      "learning_rate": 3.723501475533187e-07,
+      "loss": 0.4114,
+      "step": 5835
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.028315649281295,
+      "learning_rate": 3.714257731558063e-07,
+      "loss": 0.3404,
+      "step": 5836
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.778623088036855,
+      "learning_rate": 3.705025033199189e-07,
+      "loss": 0.3864,
+      "step": 5837
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.520110896763575,
+      "learning_rate": 3.695803382659835e-07,
+      "loss": 0.4551,
+      "step": 5838
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.23156189014286,
+      "learning_rate": 3.6865927821406687e-07,
+      "loss": 0.3906,
+      "step": 5839
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.40454729061486,
+      "learning_rate": 3.677393233839693e-07,
+      "loss": 0.3828,
+      "step": 5840
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.267521077205595,
+      "learning_rate": 3.668204739952286e-07,
+      "loss": 0.4079,
+      "step": 5841
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.741427543154004,
+      "learning_rate": 3.6590273026711975e-07,
+      "loss": 0.4377,
+      "step": 5842
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 17.163739083812075,
+      "learning_rate": 3.649860924186521e-07,
+      "loss": 0.4158,
+      "step": 5843
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.04906459931671,
+      "learning_rate": 3.640705606685718e-07,
+      "loss": 0.5255,
+      "step": 5844
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.795166239455593,
+      "learning_rate": 3.6315613523536185e-07,
+      "loss": 0.5055,
+      "step": 5845
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.795142062949072,
+      "learning_rate": 3.622428163372382e-07,
+      "loss": 0.461,
+      "step": 5846
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 41.59665340527445,
+      "learning_rate": 3.6133060419215825e-07,
+      "loss": 0.4074,
+      "step": 5847
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 42.333353518364206,
+      "learning_rate": 3.6041949901780895e-07,
+      "loss": 0.4329,
+      "step": 5848
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.7609524991011565,
+      "learning_rate": 3.5950950103161906e-07,
+      "loss": 0.4373,
+      "step": 5849
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 12.048411841711832,
+      "learning_rate": 3.5860061045074836e-07,
+      "loss": 0.4571,
+      "step": 5850
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 13.386935904670116,
+      "learning_rate": 3.5769282749209375e-07,
+      "loss": 0.4024,
+      "step": 5851
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.21681420818885,
+      "learning_rate": 3.5678615237229043e-07,
+      "loss": 0.4239,
+      "step": 5852
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.144338260682893,
+      "learning_rate": 3.558805853077052e-07,
+      "loss": 0.5091,
+      "step": 5853
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 18.794005908617386,
+      "learning_rate": 3.5497612651444355e-07,
+      "loss": 0.4535,
+      "step": 5854
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.13700831213853,
+      "learning_rate": 3.540727762083451e-07,
+      "loss": 0.3585,
+      "step": 5855
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.497479375444295,
+      "learning_rate": 3.531705346049846e-07,
+      "loss": 0.3891,
+      "step": 5856
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.241630777726071,
+      "learning_rate": 3.522694019196732e-07,
+      "loss": 0.4747,
+      "step": 5857
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.436203674512253,
+      "learning_rate": 3.5136937836745546e-07,
+      "loss": 0.3904,
+      "step": 5858
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.43799922507736,
+      "learning_rate": 3.504704641631146e-07,
+      "loss": 0.4385,
+      "step": 5859
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 22.965251743508155,
+      "learning_rate": 3.495726595211668e-07,
+      "loss": 0.437,
+      "step": 5860
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.362678855286216,
+      "learning_rate": 3.4867596465586295e-07,
+      "loss": 0.4571,
+      "step": 5861
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 12.359486268755676,
+      "learning_rate": 3.477803797811913e-07,
+      "loss": 0.3913,
+      "step": 5862
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.048281823987576,
+      "learning_rate": 3.4688590511087304e-07,
+      "loss": 0.3852,
+      "step": 5863
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 11.87739033110838,
+      "learning_rate": 3.459925408583664e-07,
+      "loss": 0.4177,
+      "step": 5864
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 12.665808506976699,
+      "learning_rate": 3.451002872368625e-07,
+      "loss": 0.4198,
+      "step": 5865
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.8026966861875335,
+      "learning_rate": 3.442091444592888e-07,
+      "loss": 0.3369,
+      "step": 5866
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.023375766446867,
+      "learning_rate": 3.433191127383079e-07,
+      "loss": 0.3314,
+      "step": 5867
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 22.176753973495885,
+      "learning_rate": 3.42430192286316e-07,
+      "loss": 0.4332,
+      "step": 5868
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.801643402467804,
+      "learning_rate": 3.4154238331544575e-07,
+      "loss": 0.4571,
+      "step": 5869
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.011212173278045,
+      "learning_rate": 3.4065568603756263e-07,
+      "loss": 0.365,
+      "step": 5870
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.633075045903737,
+      "learning_rate": 3.397701006642673e-07,
+      "loss": 0.4836,
+      "step": 5871
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 21.004804027129715,
+      "learning_rate": 3.388856274068975e-07,
+      "loss": 0.3793,
+      "step": 5872
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.469106425141895,
+      "learning_rate": 3.3800226647652146e-07,
+      "loss": 0.4306,
+      "step": 5873
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 29.074676700270786,
+      "learning_rate": 3.371200180839462e-07,
+      "loss": 0.4601,
+      "step": 5874
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.521713636465842,
+      "learning_rate": 3.3623888243971113e-07,
+      "loss": 0.475,
+      "step": 5875
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 11.557612339071632,
+      "learning_rate": 3.35358859754088e-07,
+      "loss": 0.3847,
+      "step": 5876
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.665560747020367,
+      "learning_rate": 3.344799502370877e-07,
+      "loss": 0.4702,
+      "step": 5877
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 13.80683927724409,
+      "learning_rate": 3.3360215409845133e-07,
+      "loss": 0.4031,
+      "step": 5878
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.587454133285425,
+      "learning_rate": 3.327254715476569e-07,
+      "loss": 0.4477,
+      "step": 5879
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.211162935467746,
+      "learning_rate": 3.3184990279391536e-07,
+      "loss": 0.4184,
+      "step": 5880
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.108456532153015,
+      "learning_rate": 3.309754480461713e-07,
+      "loss": 0.4231,
+      "step": 5881
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.923728186089213,
+      "learning_rate": 3.3010210751310555e-07,
+      "loss": 0.4775,
+      "step": 5882
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.789054789791504,
+      "learning_rate": 3.2922988140313086e-07,
+      "loss": 0.4642,
+      "step": 5883
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.086120285412205,
+      "learning_rate": 3.283587699243962e-07,
+      "loss": 0.4471,
+      "step": 5884
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.684113000506246,
+      "learning_rate": 3.2748877328478314e-07,
+      "loss": 0.4535,
+      "step": 5885
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.373516612576686,
+      "learning_rate": 3.266198916919061e-07,
+      "loss": 0.4123,
+      "step": 5886
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.157874962369352,
+      "learning_rate": 3.2575212535311596e-07,
+      "loss": 0.4352,
+      "step": 5887
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.07206604021634,
+      "learning_rate": 3.248854744754953e-07,
+      "loss": 0.4601,
+      "step": 5888
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.371191084921188,
+      "learning_rate": 3.2401993926586326e-07,
+      "loss": 0.3942,
+      "step": 5889
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 15.629838693153198,
+      "learning_rate": 3.2315551993076955e-07,
+      "loss": 0.4213,
+      "step": 5890
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 11.473874375715154,
+      "learning_rate": 3.222922166764991e-07,
+      "loss": 0.5146,
+      "step": 5891
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3413898055516917,
+      "learning_rate": 3.2143002970907e-07,
+      "loss": 0.5721,
+      "step": 5892
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.48045174682581,
+      "learning_rate": 3.2056895923423426e-07,
+      "loss": 0.4209,
+      "step": 5893
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 13.150911596863523,
+      "learning_rate": 3.197090054574786e-07,
+      "loss": 0.3993,
+      "step": 5894
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.562272341104438,
+      "learning_rate": 3.1885016858402175e-07,
+      "loss": 0.4022,
+      "step": 5895
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.168984802124054,
+      "learning_rate": 3.179924488188146e-07,
+      "loss": 0.4049,
+      "step": 5896
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 11.696890917737706,
+      "learning_rate": 3.1713584636654584e-07,
+      "loss": 0.4386,
+      "step": 5897
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 13.105419591031598,
+      "learning_rate": 3.1628036143163344e-07,
+      "loss": 0.4056,
+      "step": 5898
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 11.736275011397376,
+      "learning_rate": 3.154259942182292e-07,
+      "loss": 0.3967,
+      "step": 5899
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.9574506291647955,
+      "learning_rate": 3.1457274493022105e-07,
+      "loss": 0.3852,
+      "step": 5900
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.479543251059091,
+      "learning_rate": 3.1372061377122565e-07,
+      "loss": 0.4248,
+      "step": 5901
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.270029517196845,
+      "learning_rate": 3.128696009445986e-07,
+      "loss": 0.437,
+      "step": 5902
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.739981247843087,
+      "learning_rate": 3.120197066534231e-07,
+      "loss": 0.4753,
+      "step": 5903
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.580745683221518,
+      "learning_rate": 3.1117093110051787e-07,
+      "loss": 0.3802,
+      "step": 5904
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.76151752955906,
+      "learning_rate": 3.103232744884349e-07,
+      "loss": 0.3253,
+      "step": 5905
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.660374946957707,
+      "learning_rate": 3.094767370194579e-07,
+      "loss": 0.4542,
+      "step": 5906
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.436438158758204,
+      "learning_rate": 3.0863131889560583e-07,
+      "loss": 0.5159,
+      "step": 5907
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.838535856965322,
+      "learning_rate": 3.077870203186284e-07,
+      "loss": 0.4172,
+      "step": 5908
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.708910380159401,
+      "learning_rate": 3.069438414900078e-07,
+      "loss": 0.4152,
+      "step": 5909
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0947214880228184,
+      "learning_rate": 3.0610178261096134e-07,
+      "loss": 0.5209,
+      "step": 5910
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.310474423995987,
+      "learning_rate": 3.0526084388243613e-07,
+      "loss": 0.3925,
+      "step": 5911
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.892581751437814,
+      "learning_rate": 3.0442102550511553e-07,
+      "loss": 0.3509,
+      "step": 5912
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.836463588081054,
+      "learning_rate": 3.0358232767941253e-07,
+      "loss": 0.4146,
+      "step": 5913
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.26926936529437,
+      "learning_rate": 3.0274475060547423e-07,
+      "loss": 0.5146,
+      "step": 5914
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.693172961761073,
+      "learning_rate": 3.019082944831786e-07,
+      "loss": 0.365,
+      "step": 5915
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.3409964378446535,
+      "learning_rate": 3.0107295951213753e-07,
+      "loss": 0.4116,
+      "step": 5916
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 25.805125235166113,
+      "learning_rate": 3.0023874589169667e-07,
+      "loss": 0.5167,
+      "step": 5917
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.493040299453499,
+      "learning_rate": 2.994056538209311e-07,
+      "loss": 0.4228,
+      "step": 5918
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.006246393576023,
+      "learning_rate": 2.985736834986497e-07,
+      "loss": 0.4168,
+      "step": 5919
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.208133491756821,
+      "learning_rate": 2.977428351233941e-07,
+      "loss": 0.4235,
+      "step": 5920
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.43967447959776,
+      "learning_rate": 2.969131088934374e-07,
+      "loss": 0.4695,
+      "step": 5921
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.946178976018677,
+      "learning_rate": 2.9608450500678566e-07,
+      "loss": 0.4563,
+      "step": 5922
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 18.403417902225954,
+      "learning_rate": 2.9525702366117683e-07,
+      "loss": 0.4507,
+      "step": 5923
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.459395665680696,
+      "learning_rate": 2.9443066505407956e-07,
+      "loss": 0.4302,
+      "step": 5924
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.093963549298736,
+      "learning_rate": 2.936054293826973e-07,
+      "loss": 0.4015,
+      "step": 5925
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.937180722602683,
+      "learning_rate": 2.927813168439636e-07,
+      "loss": 0.4148,
+      "step": 5926
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 37.43256228148086,
+      "learning_rate": 2.9195832763454393e-07,
+      "loss": 0.42,
+      "step": 5927
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.166699629278286,
+      "learning_rate": 2.9113646195083723e-07,
+      "loss": 0.3557,
+      "step": 5928
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.963191051335659,
+      "learning_rate": 2.903157199889711e-07,
+      "loss": 0.3823,
+      "step": 5929
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.885475335171903,
+      "learning_rate": 2.8949610194481e-07,
+      "loss": 0.4291,
+      "step": 5930
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 37.574632057038095,
+      "learning_rate": 2.886776080139447e-07,
+      "loss": 0.4378,
+      "step": 5931
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.581847989615499,
+      "learning_rate": 2.8786023839170287e-07,
+      "loss": 0.461,
+      "step": 5932
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.111300579536097,
+      "learning_rate": 2.8704399327314005e-07,
+      "loss": 0.3309,
+      "step": 5933
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.923693285473334,
+      "learning_rate": 2.8622887285304437e-07,
+      "loss": 0.4058,
+      "step": 5934
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.8306779466319245,
+      "learning_rate": 2.8541487732593687e-07,
+      "loss": 0.4386,
+      "step": 5935
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.658779694308569,
+      "learning_rate": 2.8460200688606885e-07,
+      "loss": 0.4602,
+      "step": 5936
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 31.434862952894317,
+      "learning_rate": 2.8379026172742297e-07,
+      "loss": 0.3659,
+      "step": 5937
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.751107030673094,
+      "learning_rate": 2.8297964204371477e-07,
+      "loss": 0.4335,
+      "step": 5938
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.931064161192655,
+      "learning_rate": 2.8217014802838894e-07,
+      "loss": 0.4217,
+      "step": 5939
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.830491024998155,
+      "learning_rate": 2.813617798746243e-07,
+      "loss": 0.4012,
+      "step": 5940
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.358509556948288,
+      "learning_rate": 2.805545377753288e-07,
+      "loss": 0.4796,
+      "step": 5941
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.6338949307507855,
+      "learning_rate": 2.7974842192314277e-07,
+      "loss": 0.4229,
+      "step": 5942
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.914309459169457,
+      "learning_rate": 2.7894343251043787e-07,
+      "loss": 0.4117,
+      "step": 5943
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.015595362723658,
+      "learning_rate": 2.7813956972931543e-07,
+      "loss": 0.4404,
+      "step": 5944
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.901879425882384,
+      "learning_rate": 2.7733683377161035e-07,
+      "loss": 0.4406,
+      "step": 5945
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.042872878868316,
+      "learning_rate": 2.7653522482888606e-07,
+      "loss": 0.3736,
+      "step": 5946
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.731108451343736,
+      "learning_rate": 2.75734743092439e-07,
+      "loss": 0.4247,
+      "step": 5947
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.784134408659243,
+      "learning_rate": 2.749353887532963e-07,
+      "loss": 0.4717,
+      "step": 5948
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 12.855547547684075,
+      "learning_rate": 2.7413716200221485e-07,
+      "loss": 0.4025,
+      "step": 5949
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.690992590756652,
+      "learning_rate": 2.7334006302968395e-07,
+      "loss": 0.4157,
+      "step": 5950
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.447659208624662,
+      "learning_rate": 2.725440920259215e-07,
+      "loss": 0.4379,
+      "step": 5951
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.657507798853742,
+      "learning_rate": 2.7174924918087987e-07,
+      "loss": 0.4562,
+      "step": 5952
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 41.18739517789081,
+      "learning_rate": 2.709555346842391e-07,
+      "loss": 0.3168,
+      "step": 5953
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.825166919859997,
+      "learning_rate": 2.701629487254104e-07,
+      "loss": 0.449,
+      "step": 5954
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.82649782666973,
+      "learning_rate": 2.6937149149353803e-07,
+      "loss": 0.4019,
+      "step": 5955
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.066950538402935,
+      "learning_rate": 2.685811631774926e-07,
+      "loss": 0.4304,
+      "step": 5956
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.938553033151722,
+      "learning_rate": 2.677919639658805e-07,
+      "loss": 0.482,
+      "step": 5957
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 50.26138556187903,
+      "learning_rate": 2.670038940470349e-07,
+      "loss": 0.4394,
+      "step": 5958
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.925414078694652,
+      "learning_rate": 2.662169536090192e-07,
+      "loss": 0.3938,
+      "step": 5959
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.700971743360954,
+      "learning_rate": 2.654311428396311e-07,
+      "loss": 0.3995,
+      "step": 5960
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 56.33366616847569,
+      "learning_rate": 2.6464646192639496e-07,
+      "loss": 0.4386,
+      "step": 5961
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.825047504067705,
+      "learning_rate": 2.6386291105656656e-07,
+      "loss": 0.455,
+      "step": 5962
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.370317165418127,
+      "learning_rate": 2.6308049041713247e-07,
+      "loss": 0.4483,
+      "step": 5963
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.018362355876949,
+      "learning_rate": 2.6229920019480934e-07,
+      "loss": 0.466,
+      "step": 5964
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.472306304143079,
+      "learning_rate": 2.615190405760443e-07,
+      "loss": 0.4138,
+      "step": 5965
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.048097489540707,
+      "learning_rate": 2.607400117470138e-07,
+      "loss": 0.458,
+      "step": 5966
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.6609356250245915,
+      "learning_rate": 2.599621138936265e-07,
+      "loss": 0.4238,
+      "step": 5967
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.057106471833475,
+      "learning_rate": 2.5918534720151823e-07,
+      "loss": 0.4579,
+      "step": 5968
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.213143784548258,
+      "learning_rate": 2.584097118560569e-07,
+      "loss": 0.448,
+      "step": 5969
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.323593317303844,
+      "learning_rate": 2.5763520804234e-07,
+      "loss": 0.4208,
+      "step": 5970
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2344997344719222,
+      "learning_rate": 2.5686183594519467e-07,
+      "loss": 0.5261,
+      "step": 5971
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.740790511037284,
+      "learning_rate": 2.560895957491788e-07,
+      "loss": 0.5076,
+      "step": 5972
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.430463530071608,
+      "learning_rate": 2.5531848763857894e-07,
+      "loss": 0.4736,
+      "step": 5973
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.23844208113187,
+      "learning_rate": 2.545485117974111e-07,
+      "loss": 0.3933,
+      "step": 5974
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0811573669767824,
+      "learning_rate": 2.537796684094246e-07,
+      "loss": 0.5218,
+      "step": 5975
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.58761368366624,
+      "learning_rate": 2.530119576580936e-07,
+      "loss": 0.4835,
+      "step": 5976
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 17.97062323608271,
+      "learning_rate": 2.522453797266261e-07,
+      "loss": 0.4003,
+      "step": 5977
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 56.48831233462823,
+      "learning_rate": 2.514799347979574e-07,
+      "loss": 0.4126,
+      "step": 5978
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.660000643558496,
+      "learning_rate": 2.507156230547525e-07,
+      "loss": 0.4777,
+      "step": 5979
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.445899400591786,
+      "learning_rate": 2.4995244467940773e-07,
+      "loss": 0.3921,
+      "step": 5980
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.31067687194062,
+      "learning_rate": 2.4919039985404626e-07,
+      "loss": 0.4439,
+      "step": 5981
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.727360803821629,
+      "learning_rate": 2.4842948876052373e-07,
+      "loss": 0.4434,
+      "step": 5982
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.528255476822902,
+      "learning_rate": 2.476697115804233e-07,
+      "loss": 0.4788,
+      "step": 5983
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.56690615952462,
+      "learning_rate": 2.4691106849505766e-07,
+      "loss": 0.4451,
+      "step": 5984
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.351468954746938,
+      "learning_rate": 2.4615355968546915e-07,
+      "loss": 0.4576,
+      "step": 5985
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.588075521537686,
+      "learning_rate": 2.4539718533242885e-07,
+      "loss": 0.4148,
+      "step": 5986
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.2692570735456306,
+      "learning_rate": 2.446419456164395e-07,
+      "loss": 0.3312,
+      "step": 5987
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 16.862239544064373,
+      "learning_rate": 2.4388784071772976e-07,
+      "loss": 0.3892,
+      "step": 5988
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.536035206226808,
+      "learning_rate": 2.431348708162595e-07,
+      "loss": 0.3907,
+      "step": 5989
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.9818460449923005,
+      "learning_rate": 2.4238303609171734e-07,
+      "loss": 0.4744,
+      "step": 5990
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 12.325482472388074,
+      "learning_rate": 2.4163233672352017e-07,
+      "loss": 0.3997,
+      "step": 5991
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.084774400080377,
+      "learning_rate": 2.408827728908164e-07,
+      "loss": 0.5107,
+      "step": 5992
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1040052633469488,
+      "learning_rate": 2.4013434477248023e-07,
+      "loss": 0.5045,
+      "step": 5993
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.79610948152564,
+      "learning_rate": 2.393870525471159e-07,
+      "loss": 0.3961,
+      "step": 5994
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.868330017388978,
+      "learning_rate": 2.386408963930592e-07,
+      "loss": 0.4462,
+      "step": 5995
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.92076172156508,
+      "learning_rate": 2.3789587648837043e-07,
+      "loss": 0.4445,
+      "step": 5996
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.742327209160975,
+      "learning_rate": 2.3715199301084235e-07,
+      "loss": 0.4055,
+      "step": 5997
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.544016624514466,
+      "learning_rate": 2.364092461379941e-07,
+      "loss": 0.3791,
+      "step": 5998
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.799585469594696,
+      "learning_rate": 2.35667636047075e-07,
+      "loss": 0.3983,
+      "step": 5999
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.601040485219471,
+      "learning_rate": 2.3492716291506235e-07,
+      "loss": 0.4707,
+      "step": 6000
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.679971576664796,
+      "learning_rate": 2.341878269186626e-07,
+      "loss": 0.3836,
+      "step": 6001
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 11.555158103521904,
+      "learning_rate": 2.3344962823431183e-07,
+      "loss": 0.4243,
+      "step": 6002
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.369498920543641,
+      "learning_rate": 2.327125670381719e-07,
+      "loss": 0.4431,
+      "step": 6003
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 20.738723382480593,
+      "learning_rate": 2.3197664350613547e-07,
+      "loss": 0.4553,
+      "step": 6004
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.900448518895032,
+      "learning_rate": 2.3124185781382368e-07,
+      "loss": 0.5062,
+      "step": 6005
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.671027022612743,
+      "learning_rate": 2.3050821013658465e-07,
+      "loss": 0.3903,
+      "step": 6006
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.248884464187572,
+      "learning_rate": 2.297757006494966e-07,
+      "loss": 0.5548,
+      "step": 6007
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 13.29356887073713,
+      "learning_rate": 2.2904432952736533e-07,
+      "loss": 0.4211,
+      "step": 6008
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 12.532186972980012,
+      "learning_rate": 2.2831409694472395e-07,
+      "loss": 0.4355,
+      "step": 6009
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.235767644856566,
+      "learning_rate": 2.2758500307583642e-07,
+      "loss": 0.4012,
+      "step": 6010
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 15.737942138769418,
+      "learning_rate": 2.26857048094693e-07,
+      "loss": 0.4185,
+      "step": 6011
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.60557061687433,
+      "learning_rate": 2.261302321750114e-07,
+      "loss": 0.4227,
+      "step": 6012
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 13.56688523077517,
+      "learning_rate": 2.254045554902412e-07,
+      "loss": 0.3822,
+      "step": 6013
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.90680573226255,
+      "learning_rate": 2.2468001821355502e-07,
+      "loss": 0.4469,
+      "step": 6014
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.152151691376813,
+      "learning_rate": 2.239566205178584e-07,
+      "loss": 0.4141,
+      "step": 6015
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.016338476327478,
+      "learning_rate": 2.2323436257578212e-07,
+      "loss": 0.4269,
+      "step": 6016
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0486243164892097,
+      "learning_rate": 2.2251324455968447e-07,
+      "loss": 0.501,
+      "step": 6017
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.301773411730899,
+      "learning_rate": 2.21793266641655e-07,
+      "loss": 0.4132,
+      "step": 6018
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.4305781111696705,
+      "learning_rate": 2.210744289935074e-07,
+      "loss": 0.3648,
+      "step": 6019
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.780243809139604,
+      "learning_rate": 2.2035673178678553e-07,
+      "loss": 0.4576,
+      "step": 6020
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1923403148339042,
+      "learning_rate": 2.196401751927596e-07,
+      "loss": 0.474,
+      "step": 6021
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.061521025676669,
+      "learning_rate": 2.1892475938242895e-07,
+      "loss": 0.4223,
+      "step": 6022
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.55886429224066,
+      "learning_rate": 2.1821048452652095e-07,
+      "loss": 0.3507,
+      "step": 6023
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 12.418535918653141,
+      "learning_rate": 2.1749735079548861e-07,
+      "loss": 0.3805,
+      "step": 6024
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.200482062165804,
+      "learning_rate": 2.1678535835951532e-07,
+      "loss": 0.3832,
+      "step": 6025
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.357742183138216,
+      "learning_rate": 2.160745073885101e-07,
+      "loss": 0.4424,
+      "step": 6026
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.233555603130945,
+      "learning_rate": 2.1536479805210952e-07,
+      "loss": 0.4291,
+      "step": 6027
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.1710118690525,
+      "learning_rate": 2.1465623051967975e-07,
+      "loss": 0.4057,
+      "step": 6028
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.9083133829225805,
+      "learning_rate": 2.1394880496031223e-07,
+      "loss": 0.411,
+      "step": 6029
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.75403303958984,
+      "learning_rate": 2.1324252154282743e-07,
+      "loss": 0.4844,
+      "step": 6030
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.624183726125615,
+      "learning_rate": 2.1253738043577221e-07,
+      "loss": 0.4423,
+      "step": 6031
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.664724592601209,
+      "learning_rate": 2.1183338180742142e-07,
+      "loss": 0.4139,
+      "step": 6032
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0565802086343434,
+      "learning_rate": 2.1113052582577675e-07,
+      "loss": 0.4605,
+      "step": 6033
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.592858971807619,
+      "learning_rate": 2.104288126585674e-07,
+      "loss": 0.4075,
+      "step": 6034
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.096021036684525,
+      "learning_rate": 2.0972824247325109e-07,
+      "loss": 0.442,
+      "step": 6035
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.629129948431016,
+      "learning_rate": 2.090288154370107e-07,
+      "loss": 0.48,
+      "step": 6036
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 13.036012427022493,
+      "learning_rate": 2.0833053171675722e-07,
+      "loss": 0.4638,
+      "step": 6037
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 13.541993179349438,
+      "learning_rate": 2.0763339147912954e-07,
+      "loss": 0.482,
+      "step": 6038
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 33.39607907012994,
+      "learning_rate": 2.0693739489049235e-07,
+      "loss": 0.3405,
+      "step": 6039
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.787505409632686,
+      "learning_rate": 2.0624254211693894e-07,
+      "loss": 0.4341,
+      "step": 6040
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.451348763693033,
+      "learning_rate": 2.0554883332428777e-07,
+      "loss": 0.4305,
+      "step": 6041
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 53.18736987286233,
+      "learning_rate": 2.048562686780864e-07,
+      "loss": 0.4924,
+      "step": 6042
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.357649108103432,
+      "learning_rate": 2.041648483436076e-07,
+      "loss": 0.4606,
+      "step": 6043
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.532069854911818,
+      "learning_rate": 2.034745724858511e-07,
+      "loss": 0.3895,
+      "step": 6044
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 12.587868492103368,
+      "learning_rate": 2.0278544126954514e-07,
+      "loss": 0.3911,
+      "step": 6045
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.234987106807465,
+      "learning_rate": 2.0209745485914368e-07,
+      "loss": 0.4138,
+      "step": 6046
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 12.342673172628077,
+      "learning_rate": 2.014106134188265e-07,
+      "loss": 0.4961,
+      "step": 6047
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.242794987818895,
+      "learning_rate": 2.0072491711250308e-07,
+      "loss": 0.3942,
+      "step": 6048
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.426404424257709,
+      "learning_rate": 2.0004036610380583e-07,
+      "loss": 0.4529,
+      "step": 6049
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.489751749543051,
+      "learning_rate": 1.9935696055609798e-07,
+      "loss": 0.4268,
+      "step": 6050
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 17.66782386662816,
+      "learning_rate": 1.986747006324652e-07,
+      "loss": 0.4749,
+      "step": 6051
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.56879124382153,
+      "learning_rate": 1.9799358649572275e-07,
+      "loss": 0.405,
+      "step": 6052
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.065688135132426,
+      "learning_rate": 1.9731361830841177e-07,
+      "loss": 0.4141,
+      "step": 6053
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 16.602856883407917,
+      "learning_rate": 1.966347962327997e-07,
+      "loss": 0.4122,
+      "step": 6054
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.831499194964778,
+      "learning_rate": 1.9595712043087968e-07,
+      "loss": 0.4748,
+      "step": 6055
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.848797752722055,
+      "learning_rate": 1.9528059106437237e-07,
+      "loss": 0.5357,
+      "step": 6056
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.19568554864068,
+      "learning_rate": 1.946052082947242e-07,
+      "loss": 0.4482,
+      "step": 6057
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.93631392675379,
+      "learning_rate": 1.9393097228310954e-07,
+      "loss": 0.4474,
+      "step": 6058
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.260907125757506,
+      "learning_rate": 1.932578831904258e-07,
+      "loss": 0.4447,
+      "step": 6059
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.256014806639852,
+      "learning_rate": 1.9258594117730123e-07,
+      "loss": 0.4063,
+      "step": 6060
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.163849213436904,
+      "learning_rate": 1.9191514640408692e-07,
+      "loss": 0.4479,
+      "step": 6061
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.329715900589596,
+      "learning_rate": 1.9124549903085987e-07,
+      "loss": 0.4495,
+      "step": 6062
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.034254586824161,
+      "learning_rate": 1.9057699921742613e-07,
+      "loss": 0.5243,
+      "step": 6063
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.7182537672694425,
+      "learning_rate": 1.8990964712331582e-07,
+      "loss": 0.3729,
+      "step": 6064
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.581959634756734,
+      "learning_rate": 1.8924344290778608e-07,
+      "loss": 0.4436,
+      "step": 6065
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.208452850194081,
+      "learning_rate": 1.8857838672981855e-07,
+      "loss": 0.4557,
+      "step": 6066
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.828221397572074,
+      "learning_rate": 1.8791447874812296e-07,
+      "loss": 0.4238,
+      "step": 6067
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.523347033792497,
+      "learning_rate": 1.872517191211337e-07,
+      "loss": 0.4285,
+      "step": 6068
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.313057016353419,
+      "learning_rate": 1.865901080070115e-07,
+      "loss": 0.4722,
+      "step": 6069
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.13967800304817,
+      "learning_rate": 1.8592964556364334e-07,
+      "loss": 0.4482,
+      "step": 6070
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.2306709715323105,
+      "learning_rate": 1.8527033194864207e-07,
+      "loss": 0.4854,
+      "step": 6071
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.59877598463886,
+      "learning_rate": 1.8461216731934517e-07,
+      "loss": 0.3817,
+      "step": 6072
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.07994446266774,
+      "learning_rate": 1.839551518328181e-07,
+      "loss": 0.4427,
+      "step": 6073
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.2372882074171825,
+      "learning_rate": 1.8329928564584932e-07,
+      "loss": 0.396,
+      "step": 6074
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 16.64621995554326,
+      "learning_rate": 1.8264456891495586e-07,
+      "loss": 0.4128,
+      "step": 6075
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.193948012152696,
+      "learning_rate": 1.8199100179637885e-07,
+      "loss": 0.4485,
+      "step": 6076
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.934371184769104,
+      "learning_rate": 1.813385844460852e-07,
+      "loss": 0.399,
+      "step": 6077
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 14.718352630750164,
+      "learning_rate": 1.806873170197676e-07,
+      "loss": 0.4267,
+      "step": 6078
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.869044575771374,
+      "learning_rate": 1.8003719967284395e-07,
+      "loss": 0.439,
+      "step": 6079
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.700500732923179,
+      "learning_rate": 1.7938823256045845e-07,
+      "loss": 0.3987,
+      "step": 6080
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.481911746281208,
+      "learning_rate": 1.7874041583748114e-07,
+      "loss": 0.4864,
+      "step": 6081
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.687644605406585,
+      "learning_rate": 1.78093749658505e-07,
+      "loss": 0.409,
+      "step": 6082
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.299199419817891,
+      "learning_rate": 1.7744823417785163e-07,
+      "loss": 0.4652,
+      "step": 6083
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.50805089776302,
+      "learning_rate": 1.7680386954956607e-07,
+      "loss": 0.3713,
+      "step": 6084
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 13.252272735615387,
+      "learning_rate": 1.7616065592742038e-07,
+      "loss": 0.3684,
+      "step": 6085
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 16.49320463551489,
+      "learning_rate": 1.7551859346490952e-07,
+      "loss": 0.4802,
+      "step": 6086
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.032024264991484,
+      "learning_rate": 1.7487768231525592e-07,
+      "loss": 0.3817,
+      "step": 6087
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.209413377151795,
+      "learning_rate": 1.7423792263140616e-07,
+      "loss": 0.4104,
+      "step": 6088
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.913212388741603,
+      "learning_rate": 1.735993145660325e-07,
+      "loss": 0.4581,
+      "step": 6089
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 14.43621099527534,
+      "learning_rate": 1.7296185827153144e-07,
+      "loss": 0.4174,
+      "step": 6090
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.672026295219572,
+      "learning_rate": 1.7232555390002682e-07,
+      "loss": 0.4634,
+      "step": 6091
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 18.816607288368626,
+      "learning_rate": 1.7169040160336436e-07,
+      "loss": 0.4519,
+      "step": 6092
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.33786473699606,
+      "learning_rate": 1.7105640153311787e-07,
+      "loss": 0.4667,
+      "step": 6093
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.753321000508387,
+      "learning_rate": 1.7042355384058407e-07,
+      "loss": 0.4346,
+      "step": 6094
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.774343435345806,
+      "learning_rate": 1.697918586767866e-07,
+      "loss": 0.376,
+      "step": 6095
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.52829607848003,
+      "learning_rate": 1.6916131619247267e-07,
+      "loss": 0.3897,
+      "step": 6096
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.986911068852063,
+      "learning_rate": 1.6853192653811356e-07,
+      "loss": 0.4449,
+      "step": 6097
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.982571705142476,
+      "learning_rate": 1.679036898639086e-07,
+      "loss": 0.4668,
+      "step": 6098
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.189999698332444,
+      "learning_rate": 1.6727660631977894e-07,
+      "loss": 0.4533,
+      "step": 6099
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 14.46312422642193,
+      "learning_rate": 1.666506760553721e-07,
+      "loss": 0.4804,
+      "step": 6100
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.950630066121511,
+      "learning_rate": 1.6602589922005917e-07,
+      "loss": 0.4365,
+      "step": 6101
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.690284345503232,
+      "learning_rate": 1.6540227596293644e-07,
+      "loss": 0.3894,
+      "step": 6102
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.292282705426501,
+      "learning_rate": 1.6477980643282654e-07,
+      "loss": 0.4882,
+      "step": 6103
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.871789321425036,
+      "learning_rate": 1.6415849077827395e-07,
+      "loss": 0.3604,
+      "step": 6104
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.211371475177253,
+      "learning_rate": 1.6353832914755065e-07,
+      "loss": 0.4463,
+      "step": 6105
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.146457691565257,
+      "learning_rate": 1.62919321688651e-07,
+      "loss": 0.4122,
+      "step": 6106
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.035170646792051,
+      "learning_rate": 1.6230146854929462e-07,
+      "loss": 0.4458,
+      "step": 6107
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.164358744721879,
+      "learning_rate": 1.6168476987692693e-07,
+      "loss": 0.4221,
+      "step": 6108
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.403206890944526,
+      "learning_rate": 1.6106922581871464e-07,
+      "loss": 0.4844,
+      "step": 6109
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.964761433693864,
+      "learning_rate": 1.6045483652155358e-07,
+      "loss": 0.4696,
+      "step": 6110
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.911398358921779,
+      "learning_rate": 1.598416021320598e-07,
+      "loss": 0.4484,
+      "step": 6111
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.843012016254029,
+      "learning_rate": 1.5922952279657566e-07,
+      "loss": 0.4043,
+      "step": 6112
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 18.483307704338824,
+      "learning_rate": 1.5861859866116824e-07,
+      "loss": 0.3827,
+      "step": 6113
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.836460207677469,
+      "learning_rate": 1.5800882987162648e-07,
+      "loss": 0.4279,
+      "step": 6114
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.709478875809834,
+      "learning_rate": 1.5740021657346728e-07,
+      "loss": 0.379,
+      "step": 6115
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.683578411386008,
+      "learning_rate": 1.5679275891192948e-07,
+      "loss": 0.4608,
+      "step": 6116
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.532536639875415,
+      "learning_rate": 1.5618645703197598e-07,
+      "loss": 0.4154,
+      "step": 6117
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.56188073742697,
+      "learning_rate": 1.5558131107829556e-07,
+      "loss": 0.3965,
+      "step": 6118
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.838740835058254,
+      "learning_rate": 1.5497732119529874e-07,
+      "loss": 0.4639,
+      "step": 6119
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.579710017900108,
+      "learning_rate": 1.5437448752712303e-07,
+      "loss": 0.3847,
+      "step": 6120
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.369471342909991,
+      "learning_rate": 1.537728102176278e-07,
+      "loss": 0.416,
+      "step": 6121
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.24045950028533,
+      "learning_rate": 1.5317228941039597e-07,
+      "loss": 0.5346,
+      "step": 6122
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.410840526854266,
+      "learning_rate": 1.5257292524873791e-07,
+      "loss": 0.389,
+      "step": 6123
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.143972820106212,
+      "learning_rate": 1.519747178756842e-07,
+      "loss": 0.4761,
+      "step": 6124
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.940472454062137,
+      "learning_rate": 1.5137766743399062e-07,
+      "loss": 0.457,
+      "step": 6125
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.283784516214541,
+      "learning_rate": 1.5078177406613825e-07,
+      "loss": 0.4376,
+      "step": 6126
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.065278075248987,
+      "learning_rate": 1.5018703791432997e-07,
+      "loss": 0.4197,
+      "step": 6127
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.239892905682847,
+      "learning_rate": 1.495934591204945e-07,
+      "loss": 0.4385,
+      "step": 6128
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.4470178701795655,
+      "learning_rate": 1.4900103782628183e-07,
+      "loss": 0.4671,
+      "step": 6129
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.18589751807776,
+      "learning_rate": 1.4840977417306836e-07,
+      "loss": 0.4816,
+      "step": 6130
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.324755842896137,
+      "learning_rate": 1.4781966830195282e-07,
+      "loss": 0.4315,
+      "step": 6131
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 19.09558742726461,
+      "learning_rate": 1.4723072035375708e-07,
+      "loss": 0.432,
+      "step": 6132
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.726508621554789,
+      "learning_rate": 1.4664293046902923e-07,
+      "loss": 0.4485,
+      "step": 6133
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.663620356403694,
+      "learning_rate": 1.4605629878803818e-07,
+      "loss": 0.4551,
+      "step": 6134
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.934062349993554,
+      "learning_rate": 1.454708254507775e-07,
+      "loss": 0.4236,
+      "step": 6135
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.838564535227337,
+      "learning_rate": 1.448865105969649e-07,
+      "loss": 0.4215,
+      "step": 6136
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 15.019324743734105,
+      "learning_rate": 1.4430335436603992e-07,
+      "loss": 0.3952,
+      "step": 6137
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 24.834864355586774,
+      "learning_rate": 1.437213568971685e-07,
+      "loss": 0.4209,
+      "step": 6138
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.774246720594802,
+      "learning_rate": 1.431405183292367e-07,
+      "loss": 0.3794,
+      "step": 6139
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 14.353543307368222,
+      "learning_rate": 1.4256083880085647e-07,
+      "loss": 0.4665,
+      "step": 6140
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.527545171894781,
+      "learning_rate": 1.4198231845036326e-07,
+      "loss": 0.4026,
+      "step": 6141
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.064523420757355,
+      "learning_rate": 1.4140495741581273e-07,
+      "loss": 0.4306,
+      "step": 6142
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 32.57352240275346,
+      "learning_rate": 1.4082875583498857e-07,
+      "loss": 0.4544,
+      "step": 6143
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 23.6531853769556,
+      "learning_rate": 1.402537138453941e-07,
+      "loss": 0.4804,
+      "step": 6144
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.710499275511275,
+      "learning_rate": 1.396798315842568e-07,
+      "loss": 0.4208,
+      "step": 6145
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.536440172262472,
+      "learning_rate": 1.391071091885293e-07,
+      "loss": 0.4149,
+      "step": 6146
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.432204853019314,
+      "learning_rate": 1.3853554679488455e-07,
+      "loss": 0.4924,
+      "step": 6147
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.580591845126987,
+      "learning_rate": 1.3796514453972066e-07,
+      "loss": 0.4878,
+      "step": 6148
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.605951389950999,
+      "learning_rate": 1.373959025591576e-07,
+      "loss": 0.4007,
+      "step": 6149
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.098101817067565,
+      "learning_rate": 1.3682782098903947e-07,
+      "loss": 0.4188,
+      "step": 6150
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 13.017511020270387,
+      "learning_rate": 1.362608999649334e-07,
+      "loss": 0.404,
+      "step": 6151
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 20.75751617456195,
+      "learning_rate": 1.35695139622129e-07,
+      "loss": 0.4719,
+      "step": 6152
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.311549792026235,
+      "learning_rate": 1.351305400956393e-07,
+      "loss": 0.4027,
+      "step": 6153
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.415796134660333,
+      "learning_rate": 1.3456710152019992e-07,
+      "loss": 0.4715,
+      "step": 6154
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.6957421292133725,
+      "learning_rate": 1.3400482403026937e-07,
+      "loss": 0.3633,
+      "step": 6155
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.150408492424024,
+      "learning_rate": 1.3344370776003034e-07,
+      "loss": 0.4589,
+      "step": 6156
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.674866785251966,
+      "learning_rate": 1.3288375284338683e-07,
+      "loss": 0.4759,
+      "step": 6157
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.450143856609761,
+      "learning_rate": 1.323249594139664e-07,
+      "loss": 0.3857,
+      "step": 6158
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.174677013537165,
+      "learning_rate": 1.3176732760511956e-07,
+      "loss": 0.4495,
+      "step": 6159
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.675806888912525,
+      "learning_rate": 1.3121085754991937e-07,
+      "loss": 0.4605,
+      "step": 6160
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.776563307604608,
+      "learning_rate": 1.3065554938116176e-07,
+      "loss": 0.4629,
+      "step": 6161
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.008889242790026,
+      "learning_rate": 1.301014032313641e-07,
+      "loss": 0.3937,
+      "step": 6162
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.537559925888587,
+      "learning_rate": 1.295484192327695e-07,
+      "loss": 0.4519,
+      "step": 6163
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.781368990509396,
+      "learning_rate": 1.2899659751734128e-07,
+      "loss": 0.4172,
+      "step": 6164
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.548193029786226,
+      "learning_rate": 1.2844593821676465e-07,
+      "loss": 0.3655,
+      "step": 6165
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 13.855774342748715,
+      "learning_rate": 1.2789644146245117e-07,
+      "loss": 0.3743,
+      "step": 6166
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.718233595876965,
+      "learning_rate": 1.2734810738553038e-07,
+      "loss": 0.4506,
+      "step": 6167
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.308908109534446,
+      "learning_rate": 1.2680093611685818e-07,
+      "loss": 0.4041,
+      "step": 6168
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.888002872018777,
+      "learning_rate": 1.2625492778701065e-07,
+      "loss": 0.435,
+      "step": 6169
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 11.316311473231387,
+      "learning_rate": 1.2571008252628747e-07,
+      "loss": 0.3485,
+      "step": 6170
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 14.294616480703894,
+      "learning_rate": 1.2516640046470964e-07,
+      "loss": 0.4262,
+      "step": 6171
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.907786672349816,
+      "learning_rate": 1.246238817320211e-07,
+      "loss": 0.4633,
+      "step": 6172
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.442265314643649,
+      "learning_rate": 1.2408252645769003e-07,
+      "loss": 0.4426,
+      "step": 6173
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.74199060534077,
+      "learning_rate": 1.2354233477090361e-07,
+      "loss": 0.4579,
+      "step": 6174
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.905669051502944,
+      "learning_rate": 1.230033068005737e-07,
+      "loss": 0.5288,
+      "step": 6175
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.652861926247441,
+      "learning_rate": 1.2246544267533357e-07,
+      "loss": 0.4272,
+      "step": 6176
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 16.146985015771364,
+      "learning_rate": 1.219287425235388e-07,
+      "loss": 0.432,
+      "step": 6177
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.919366802008169,
+      "learning_rate": 1.213932064732687e-07,
+      "loss": 0.3562,
+      "step": 6178
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 12.793186117279193,
+      "learning_rate": 1.2085883465232206e-07,
+      "loss": 0.4123,
+      "step": 6179
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.738207152620466,
+      "learning_rate": 1.2032562718822082e-07,
+      "loss": 0.4254,
+      "step": 6180
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.301695417673379,
+      "learning_rate": 1.1979358420821096e-07,
+      "loss": 0.4468,
+      "step": 6181
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 12.15240676581641,
+      "learning_rate": 1.1926270583925815e-07,
+      "loss": 0.4315,
+      "step": 6182
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.853076754472145,
+      "learning_rate": 1.1873299220805102e-07,
+      "loss": 0.481,
+      "step": 6183
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.878955791984587,
+      "learning_rate": 1.1820444344100069e-07,
+      "loss": 0.4651,
+      "step": 6184
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1265777992631434,
+      "learning_rate": 1.1767705966423959e-07,
+      "loss": 0.532,
+      "step": 6185
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 36.0739734141118,
+      "learning_rate": 1.1715084100362262e-07,
+      "loss": 0.4248,
+      "step": 6186
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.798952760578612,
+      "learning_rate": 1.1662578758472542e-07,
+      "loss": 0.4315,
+      "step": 6187
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 15.66556221203124,
+      "learning_rate": 1.1610189953284834e-07,
+      "loss": 0.3742,
+      "step": 6188
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.7103720766085315,
+      "learning_rate": 1.1557917697301136e-07,
+      "loss": 0.4285,
+      "step": 6189
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 15.82778755315649,
+      "learning_rate": 1.1505762002995579e-07,
+      "loss": 0.4252,
+      "step": 6190
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 11.551355240602343,
+      "learning_rate": 1.1453722882814656e-07,
+      "loss": 0.351,
+      "step": 6191
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.995026845216769,
+      "learning_rate": 1.1401800349176984e-07,
+      "loss": 0.4264,
+      "step": 6192
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.803779554303081,
+      "learning_rate": 1.1349994414473375e-07,
+      "loss": 0.4216,
+      "step": 6193
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.91791611313793,
+      "learning_rate": 1.1298305091066664e-07,
+      "loss": 0.3705,
+      "step": 6194
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.134301460968024,
+      "learning_rate": 1.1246732391292092e-07,
+      "loss": 0.4105,
+      "step": 6195
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.430889220347723,
+      "learning_rate": 1.1195276327456927e-07,
+      "loss": 0.3953,
+      "step": 6196
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.766552114911704,
+      "learning_rate": 1.1143936911840513e-07,
+      "loss": 0.3973,
+      "step": 6197
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.354382267756021,
+      "learning_rate": 1.1092714156694606e-07,
+      "loss": 0.4345,
+      "step": 6198
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.89816523297762,
+      "learning_rate": 1.1041608074242982e-07,
+      "loss": 0.4089,
+      "step": 6199
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.31942206584094,
+      "learning_rate": 1.0990618676681553e-07,
+      "loss": 0.3869,
+      "step": 6200
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.00723360202763,
+      "learning_rate": 1.0939745976178361e-07,
+      "loss": 0.4864,
+      "step": 6201
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.55530122855174,
+      "learning_rate": 1.0888989984873754e-07,
+      "loss": 0.3974,
+      "step": 6202
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 26.402458639949128,
+      "learning_rate": 1.0838350714880042e-07,
+      "loss": 0.4416,
+      "step": 6203
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.835441528544204,
+      "learning_rate": 1.0787828178281834e-07,
+      "loss": 0.4563,
+      "step": 6204
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0950243012630387,
+      "learning_rate": 1.0737422387135765e-07,
+      "loss": 0.4646,
+      "step": 6205
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2311540739899822,
+      "learning_rate": 1.0687133353470713e-07,
+      "loss": 0.5451,
+      "step": 6206
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.4556181216107085,
+      "learning_rate": 1.0636961089287522e-07,
+      "loss": 0.4073,
+      "step": 6207
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.72157523524437,
+      "learning_rate": 1.058690560655945e-07,
+      "loss": 0.4027,
+      "step": 6208
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 14.102804485824553,
+      "learning_rate": 1.0536966917231606e-07,
+      "loss": 0.4182,
+      "step": 6209
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.096964757122657,
+      "learning_rate": 1.048714503322129e-07,
+      "loss": 0.3686,
+      "step": 6210
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.325144212009119,
+      "learning_rate": 1.0437439966418161e-07,
+      "loss": 0.4216,
+      "step": 6211
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.601879498296856,
+      "learning_rate": 1.0387851728683729e-07,
+      "loss": 0.4356,
+      "step": 6212
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.988576442202231,
+      "learning_rate": 1.0338380331851694e-07,
+      "loss": 0.3926,
+      "step": 6213
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.032648911924355,
+      "learning_rate": 1.0289025787727947e-07,
+      "loss": 0.329,
+      "step": 6214
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.183799346618919,
+      "learning_rate": 1.0239788108090343e-07,
+      "loss": 0.4528,
+      "step": 6215
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 13.753660575879948,
+      "learning_rate": 1.0190667304689094e-07,
+      "loss": 0.4737,
+      "step": 6216
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 19.59175075494714,
+      "learning_rate": 1.014166338924627e-07,
+      "loss": 0.3934,
+      "step": 6217
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.35005613758522,
+      "learning_rate": 1.0092776373456181e-07,
+      "loss": 0.3916,
+      "step": 6218
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.560156293011428,
+      "learning_rate": 1.0044006268985218e-07,
+      "loss": 0.4236,
+      "step": 6219
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.892760656111012,
+      "learning_rate": 9.995353087471848e-08,
+      "loss": 0.4406,
+      "step": 6220
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.197428676206108,
+      "learning_rate": 9.946816840526674e-08,
+      "loss": 0.462,
+      "step": 6221
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 18.138688334081312,
+      "learning_rate": 9.898397539732318e-08,
+      "loss": 0.4115,
+      "step": 6222
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.734147240297599,
+      "learning_rate": 9.850095196643649e-08,
+      "loss": 0.4495,
+      "step": 6223
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.21041995832657,
+      "learning_rate": 9.801909822787448e-08,
+      "loss": 0.4348,
+      "step": 6224
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3672457935183118,
+      "learning_rate": 9.753841429662626e-08,
+      "loss": 0.4983,
+      "step": 6225
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.75535753004519,
+      "learning_rate": 9.705890028740284e-08,
+      "loss": 0.4493,
+      "step": 6226
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.2818425810044145,
+      "learning_rate": 9.658055631463547e-08,
+      "loss": 0.4811,
+      "step": 6227
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.5004438332178065,
+      "learning_rate": 9.610338249247509e-08,
+      "loss": 0.4306,
+      "step": 6228
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.856745748908933,
+      "learning_rate": 9.562737893479556e-08,
+      "loss": 0.4241,
+      "step": 6229
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.1205381727042685,
+      "learning_rate": 9.515254575518829e-08,
+      "loss": 0.3558,
+      "step": 6230
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.80451632731278,
+      "learning_rate": 9.467888306696982e-08,
+      "loss": 0.4796,
+      "step": 6231
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.8637371868357695,
+      "learning_rate": 9.420639098317252e-08,
+      "loss": 0.4235,
+      "step": 6232
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 15.059627688805241,
+      "learning_rate": 9.373506961655343e-08,
+      "loss": 0.4188,
+      "step": 6233
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.17776818776651,
+      "learning_rate": 9.326491907958812e-08,
+      "loss": 0.509,
+      "step": 6234
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.416916361852539,
+      "learning_rate": 9.279593948447296e-08,
+      "loss": 0.4201,
+      "step": 6235
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 14.180444136034877,
+      "learning_rate": 9.232813094312509e-08,
+      "loss": 0.4084,
+      "step": 6236
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.308438770242068,
+      "learning_rate": 9.186149356718244e-08,
+      "loss": 0.4343,
+      "step": 6237
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 11.439157840899536,
+      "learning_rate": 9.139602746800258e-08,
+      "loss": 0.3929,
+      "step": 6238
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.222587676307345,
+      "learning_rate": 9.0931732756665e-08,
+      "loss": 0.376,
+      "step": 6239
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.45966418695035,
+      "learning_rate": 9.046860954396886e-08,
+      "loss": 0.4553,
+      "step": 6240
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.453528185798511,
+      "learning_rate": 9.00066579404335e-08,
+      "loss": 0.4835,
+      "step": 6241
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 13.380526233604195,
+      "learning_rate": 8.954587805629856e-08,
+      "loss": 0.4695,
+      "step": 6242
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.544152201976444,
+      "learning_rate": 8.908627000152492e-08,
+      "loss": 0.4822,
+      "step": 6243
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1088655596431665,
+      "learning_rate": 8.862783388579265e-08,
+      "loss": 0.506,
+      "step": 6244
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.951002118295332,
+      "learning_rate": 8.817056981850369e-08,
+      "loss": 0.4107,
+      "step": 6245
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.224758466188311,
+      "learning_rate": 8.771447790877908e-08,
+      "loss": 0.4776,
+      "step": 6246
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.984385284862904,
+      "learning_rate": 8.725955826546006e-08,
+      "loss": 0.4134,
+      "step": 6247
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.255270867898895,
+      "learning_rate": 8.680581099710873e-08,
+      "loss": 0.46,
+      "step": 6248
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.900324054356084,
+      "learning_rate": 8.635323621200786e-08,
+      "loss": 0.4158,
+      "step": 6249
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.9917552263978,
+      "learning_rate": 8.590183401815833e-08,
+      "loss": 0.4184,
+      "step": 6250
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.765921449419827,
+      "learning_rate": 8.54516045232845e-08,
+      "loss": 0.4116,
+      "step": 6251
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.4262308851782,
+      "learning_rate": 8.500254783482765e-08,
+      "loss": 0.4079,
+      "step": 6252
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.755059067253178,
+      "learning_rate": 8.45546640599515e-08,
+      "loss": 0.38,
+      "step": 6253
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.210292720432568,
+      "learning_rate": 8.410795330553778e-08,
+      "loss": 0.4571,
+      "step": 6254
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.651356874701834,
+      "learning_rate": 8.366241567818956e-08,
+      "loss": 0.3776,
+      "step": 6255
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 12.314024890424601,
+      "learning_rate": 8.321805128423122e-08,
+      "loss": 0.4299,
+      "step": 6256
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.821411932237205,
+      "learning_rate": 8.277486022970404e-08,
+      "loss": 0.4901,
+      "step": 6257
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.457702541839718,
+      "learning_rate": 8.233284262037178e-08,
+      "loss": 0.4552,
+      "step": 6258
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.333146713571961,
+      "learning_rate": 8.189199856171725e-08,
+      "loss": 0.456,
+      "step": 6259
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3214667530385984,
+      "learning_rate": 8.145232815894355e-08,
+      "loss": 0.5036,
+      "step": 6260
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.47030316604993,
+      "learning_rate": 8.10138315169734e-08,
+      "loss": 0.4781,
+      "step": 6261
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.017416910082287,
+      "learning_rate": 8.057650874044976e-08,
+      "loss": 0.4112,
+      "step": 6262
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.568797617582446,
+      "learning_rate": 8.01403599337347e-08,
+      "loss": 0.3967,
+      "step": 6263
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.56097533380947,
+      "learning_rate": 7.970538520091108e-08,
+      "loss": 0.3991,
+      "step": 6264
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 15.603113132324735,
+      "learning_rate": 7.927158464578032e-08,
+      "loss": 0.4866,
+      "step": 6265
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.325758803233794,
+      "learning_rate": 7.883895837186573e-08,
+      "loss": 0.4307,
+      "step": 6266
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 11.429880250929648,
+      "learning_rate": 7.840750648240857e-08,
+      "loss": 0.3861,
+      "step": 6267
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.947449269433012,
+      "learning_rate": 7.797722908036987e-08,
+      "loss": 0.4308,
+      "step": 6268
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.00072491795342,
+      "learning_rate": 7.75481262684319e-08,
+      "loss": 0.375,
+      "step": 6269
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.121494868597713,
+      "learning_rate": 7.71201981489944e-08,
+      "loss": 0.4152,
+      "step": 6270
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 12.508415058362617,
+      "learning_rate": 7.669344482417907e-08,
+      "loss": 0.4033,
+      "step": 6271
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.011429860813396,
+      "learning_rate": 7.62678663958255e-08,
+      "loss": 0.4854,
+      "step": 6272
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 17.447920375317437,
+      "learning_rate": 7.584346296549417e-08,
+      "loss": 0.4521,
+      "step": 6273
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 11.407042500360957,
+      "learning_rate": 7.542023463446457e-08,
+      "loss": 0.4909,
+      "step": 6274
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.649233769420017,
+      "learning_rate": 7.499818150373539e-08,
+      "loss": 0.4889,
+      "step": 6275
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.00110288029275,
+      "learning_rate": 7.45773036740255e-08,
+      "loss": 0.4806,
+      "step": 6276
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.32964137723021,
+      "learning_rate": 7.415760124577287e-08,
+      "loss": 0.4439,
+      "step": 6277
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.758530169061703,
+      "learning_rate": 7.373907431913519e-08,
+      "loss": 0.4388,
+      "step": 6278
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.923175238872517,
+      "learning_rate": 7.332172299399032e-08,
+      "loss": 0.4391,
+      "step": 6279
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.721110009292348,
+      "learning_rate": 7.290554736993305e-08,
+      "loss": 0.4425,
+      "step": 6280
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.504487975282788,
+      "learning_rate": 7.24905475462817e-08,
+      "loss": 0.4039,
+      "step": 6281
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.116429704100573,
+      "learning_rate": 7.207672362207041e-08,
+      "loss": 0.4873,
+      "step": 6282
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.5708142677286085,
+      "learning_rate": 7.166407569605404e-08,
+      "loss": 0.4147,
+      "step": 6283
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.465201351311464,
+      "learning_rate": 7.125260386670718e-08,
+      "loss": 0.49,
+      "step": 6284
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.969689070164696,
+      "learning_rate": 7.084230823222294e-08,
+      "loss": 0.4904,
+      "step": 6285
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.687710035906269,
+      "learning_rate": 7.043318889051409e-08,
+      "loss": 0.4666,
+      "step": 6286
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.26145084273816,
+      "learning_rate": 7.002524593921367e-08,
+      "loss": 0.428,
+      "step": 6287
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.2681979010817015,
+      "learning_rate": 6.961847947567213e-08,
+      "loss": 0.3484,
+      "step": 6288
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.769069016244952,
+      "learning_rate": 6.921288959696015e-08,
+      "loss": 0.4604,
+      "step": 6289
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2406731506893431,
+      "learning_rate": 6.880847639986809e-08,
+      "loss": 0.4679,
+      "step": 6290
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 11.985781528201327,
+      "learning_rate": 6.840523998090431e-08,
+      "loss": 0.3882,
+      "step": 6291
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 11.29599953459447,
+      "learning_rate": 6.800318043629794e-08,
+      "loss": 0.4848,
+      "step": 6292
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.839041523855355,
+      "learning_rate": 6.760229786199556e-08,
+      "loss": 0.4722,
+      "step": 6293
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.99172585405479,
+      "learning_rate": 6.7202592353664e-08,
+      "loss": 0.393,
+      "step": 6294
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.314006183245606,
+      "learning_rate": 6.68040640066886e-08,
+      "loss": 0.4432,
+      "step": 6295
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.01781299250835,
+      "learning_rate": 6.64067129161744e-08,
+      "loss": 0.4178,
+      "step": 6296
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.394195086486468,
+      "learning_rate": 6.601053917694556e-08,
+      "loss": 0.4301,
+      "step": 6297
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.826779852745231,
+      "learning_rate": 6.561554288354421e-08,
+      "loss": 0.407,
+      "step": 6298
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.431016798834973,
+      "learning_rate": 6.522172413023165e-08,
+      "loss": 0.405,
+      "step": 6299
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4023904753945833,
+      "learning_rate": 6.482908301098934e-08,
+      "loss": 0.5305,
+      "step": 6300
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.634721821716822,
+      "learning_rate": 6.443761961951789e-08,
+      "loss": 0.3993,
+      "step": 6301
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 11.663410809290141,
+      "learning_rate": 6.404733404923424e-08,
+      "loss": 0.4235,
+      "step": 6302
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3037939125186047,
+      "learning_rate": 6.365822639327724e-08,
+      "loss": 0.5162,
+      "step": 6303
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.2548184881288,
+      "learning_rate": 6.327029674450314e-08,
+      "loss": 0.4164,
+      "step": 6304
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0718293017262177,
+      "learning_rate": 6.288354519548679e-08,
+      "loss": 0.4989,
+      "step": 6305
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.918522250962681,
+      "learning_rate": 6.24979718385238e-08,
+      "loss": 0.3669,
+      "step": 6306
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.3224170557502,
+      "learning_rate": 6.211357676562613e-08,
+      "loss": 0.5019,
+      "step": 6307
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.091087989153223,
+      "learning_rate": 6.173036006852595e-08,
+      "loss": 0.3646,
+      "step": 6308
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.599578542821639,
+      "learning_rate": 6.134832183867456e-08,
+      "loss": 0.4109,
+      "step": 6309
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.406878349537093,
+      "learning_rate": 6.09674621672407e-08,
+      "loss": 0.4126,
+      "step": 6310
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.02480753055882,
+      "learning_rate": 6.058778114511333e-08,
+      "loss": 0.4367,
+      "step": 6311
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.02759280320257,
+      "learning_rate": 6.02092788628983e-08,
+      "loss": 0.4175,
+      "step": 6312
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.260191815563298,
+      "learning_rate": 5.983195541092224e-08,
+      "loss": 0.4395,
+      "step": 6313
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 13.92889875208836,
+      "learning_rate": 5.945581087922925e-08,
+      "loss": 0.4362,
+      "step": 6314
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.624150660621997,
+      "learning_rate": 5.9080845357581965e-08,
+      "loss": 0.4116,
+      "step": 6315
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 29.043009347251978,
+      "learning_rate": 5.870705893546324e-08,
+      "loss": 0.394,
+      "step": 6316
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.652851887471437,
+      "learning_rate": 5.8334451702072304e-08,
+      "loss": 0.4503,
+      "step": 6317
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 22.906452452120956,
+      "learning_rate": 5.796302374632801e-08,
+      "loss": 0.4604,
+      "step": 6318
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 12.886603936450323,
+      "learning_rate": 5.759277515686834e-08,
+      "loss": 0.4265,
+      "step": 6319
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.728404051456792,
+      "learning_rate": 5.722370602204874e-08,
+      "loss": 0.4077,
+      "step": 6320
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.99216077297953,
+      "learning_rate": 5.685581642994431e-08,
+      "loss": 0.3951,
+      "step": 6321
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.542490385279293,
+      "learning_rate": 5.648910646834815e-08,
+      "loss": 0.436,
+      "step": 6322
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.633001852234321,
+      "learning_rate": 5.612357622477138e-08,
+      "loss": 0.4052,
+      "step": 6323
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.7343659676861,
+      "learning_rate": 5.5759225786444216e-08,
+      "loss": 0.4007,
+      "step": 6324
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 12.059719197207686,
+      "learning_rate": 5.5396055240314883e-08,
+      "loss": 0.4538,
+      "step": 6325
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.032819691692465,
+      "learning_rate": 5.503406467305128e-08,
+      "loss": 0.3745,
+      "step": 6326
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 11.469563669440324,
+      "learning_rate": 5.4673254171037635e-08,
+      "loss": 0.408,
+      "step": 6327
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 19.925974111131847,
+      "learning_rate": 5.4313623820377305e-08,
+      "loss": 0.4222,
+      "step": 6328
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 15.275587544538409,
+      "learning_rate": 5.3955173706894405e-08,
+      "loss": 0.4476,
+      "step": 6329
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.238281336076064,
+      "learning_rate": 5.359790391612718e-08,
+      "loss": 0.3873,
+      "step": 6330
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.246258780544184,
+      "learning_rate": 5.324181453333521e-08,
+      "loss": 0.3904,
+      "step": 6331
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.499224801930144,
+      "learning_rate": 5.288690564349608e-08,
+      "loss": 0.3831,
+      "step": 6332
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.596704315212283,
+      "learning_rate": 5.2533177331304806e-08,
+      "loss": 0.4389,
+      "step": 6333
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.980345955490508,
+      "learning_rate": 5.218062968117499e-08,
+      "loss": 0.4185,
+      "step": 6334
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.27548363822673,
+      "learning_rate": 5.182926277723821e-08,
+      "loss": 0.4503,
+      "step": 6335
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.070633468285622,
+      "learning_rate": 5.147907670334462e-08,
+      "loss": 0.4852,
+      "step": 6336
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.317264199415102,
+      "learning_rate": 5.113007154306293e-08,
+      "loss": 0.4181,
+      "step": 6337
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.664993413885735,
+      "learning_rate": 5.078224737967874e-08,
+      "loss": 0.4323,
+      "step": 6338
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.795743005034303,
+      "learning_rate": 5.043560429619787e-08,
+      "loss": 0.4317,
+      "step": 6339
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.874772361412588,
+      "learning_rate": 5.009014237534249e-08,
+      "loss": 0.5446,
+      "step": 6340
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.37146880660566,
+      "learning_rate": 4.974586169955386e-08,
+      "loss": 0.3951,
+      "step": 6341
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.1652422013938795,
+      "learning_rate": 4.940276235099073e-08,
+      "loss": 0.4307,
+      "step": 6342
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.362350513948233,
+      "learning_rate": 4.906084441152981e-08,
+      "loss": 0.396,
+      "step": 6343
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.835658017923685,
+      "learning_rate": 4.872010796276749e-08,
+      "loss": 0.4128,
+      "step": 6344
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.81766170852999,
+      "learning_rate": 4.838055308601652e-08,
+      "loss": 0.4167,
+      "step": 6345
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1282704089583009,
+      "learning_rate": 4.804217986230763e-08,
+      "loss": 0.4789,
+      "step": 6346
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 17.77406959076672,
+      "learning_rate": 4.7704988372390126e-08,
+      "loss": 0.4522,
+      "step": 6347
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.778476244250957,
+      "learning_rate": 4.7368978696731846e-08,
+      "loss": 0.5139,
+      "step": 6348
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.663569072693145,
+      "learning_rate": 4.703415091551755e-08,
+      "loss": 0.442,
+      "step": 6349
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.313865315046639,
+      "learning_rate": 4.670050510865165e-08,
+      "loss": 0.4034,
+      "step": 6350
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 14.883221809590651,
+      "learning_rate": 4.6368041355753234e-08,
+      "loss": 0.3708,
+      "step": 6351
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 16.184157583432206,
+      "learning_rate": 4.6036759736163286e-08,
+      "loss": 0.4693,
+      "step": 6352
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 11.538429776588618,
+      "learning_rate": 4.570666032893689e-08,
+      "loss": 0.3846,
+      "step": 6353
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.740438768383062,
+      "learning_rate": 4.537774321285105e-08,
+      "loss": 0.4036,
+      "step": 6354
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 11.094305721374937,
+      "learning_rate": 4.505000846639685e-08,
+      "loss": 0.3677,
+      "step": 6355
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.689872612810646,
+      "learning_rate": 4.4723456167785616e-08,
+      "loss": 0.437,
+      "step": 6356
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.2137954511819125,
+      "learning_rate": 4.4398086394945026e-08,
+      "loss": 0.4495,
+      "step": 6357
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 11.510387041526407,
+      "learning_rate": 4.407389922552185e-08,
+      "loss": 0.3892,
+      "step": 6358
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.357229057087444,
+      "learning_rate": 4.375089473687921e-08,
+      "loss": 0.4452,
+      "step": 6359
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.765084299397444,
+      "learning_rate": 4.3429073006099886e-08,
+      "loss": 0.4471,
+      "step": 6360
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.950980556609874,
+      "learning_rate": 4.310843410998244e-08,
+      "loss": 0.408,
+      "step": 6361
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1516825670888677,
+      "learning_rate": 4.278897812504457e-08,
+      "loss": 0.4804,
+      "step": 6362
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.353312920986031,
+      "learning_rate": 4.247070512752083e-08,
+      "loss": 0.4179,
+      "step": 6363
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.1134559667998545,
+      "learning_rate": 4.215361519336436e-08,
+      "loss": 0.4269,
+      "step": 6364
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.993329363549481,
+      "learning_rate": 4.183770839824464e-08,
+      "loss": 0.4467,
+      "step": 6365
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0782704789222695,
+      "learning_rate": 4.1522984817549685e-08,
+      "loss": 0.5193,
+      "step": 6366
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.941580645300886,
+      "learning_rate": 4.120944452638609e-08,
+      "loss": 0.4131,
+      "step": 6367
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 30.35268772377776,
+      "learning_rate": 4.089708759957567e-08,
+      "loss": 0.3911,
+      "step": 6368
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.811053553892459,
+      "learning_rate": 4.058591411165991e-08,
+      "loss": 0.3856,
+      "step": 6369
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.938666508186083,
+      "learning_rate": 4.027592413689663e-08,
+      "loss": 0.4051,
+      "step": 6370
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.736984749197012,
+      "learning_rate": 3.996711774926221e-08,
+      "loss": 0.412,
+      "step": 6371
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.947262093428573,
+      "learning_rate": 3.9659495022450476e-08,
+      "loss": 0.3723,
+      "step": 6372
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.598192862602188,
+      "learning_rate": 3.93530560298716e-08,
+      "loss": 0.4488,
+      "step": 6373
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.75004632959922,
+      "learning_rate": 3.9047800844654316e-08,
+      "loss": 0.4838,
+      "step": 6374
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.935638982323909,
+      "learning_rate": 3.8743729539645335e-08,
+      "loss": 0.4391,
+      "step": 6375
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.282839817522767,
+      "learning_rate": 3.8440842187406624e-08,
+      "loss": 0.4274,
+      "step": 6376
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 11.811873216177023,
+      "learning_rate": 3.813913886022091e-08,
+      "loss": 0.3425,
+      "step": 6377
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 17.406691558001874,
+      "learning_rate": 3.7838619630085614e-08,
+      "loss": 0.4302,
+      "step": 6378
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.44707879629809,
+      "learning_rate": 3.7539284568716695e-08,
+      "loss": 0.4069,
+      "step": 6379
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 53.576056872120446,
+      "learning_rate": 3.7241133747547566e-08,
+      "loss": 0.4504,
+      "step": 6380
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.456705337740436,
+      "learning_rate": 3.694416723772853e-08,
+      "loss": 0.4545,
+      "step": 6381
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.590957538849181,
+      "learning_rate": 3.6648385110127914e-08,
+      "loss": 0.4294,
+      "step": 6382
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.380673246140858,
+      "learning_rate": 3.635378743533036e-08,
+      "loss": 0.3266,
+      "step": 6383
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.869573716862622,
+      "learning_rate": 3.6060374283639623e-08,
+      "loss": 0.4105,
+      "step": 6384
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 19.764516354673134,
+      "learning_rate": 3.5768145725075254e-08,
+      "loss": 0.431,
+      "step": 6385
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.6059966910546,
+      "learning_rate": 3.547710182937425e-08,
+      "loss": 0.5175,
+      "step": 6386
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.706697493953652,
+      "learning_rate": 3.518724266599216e-08,
+      "loss": 0.4397,
+      "step": 6387
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.270126018500545,
+      "learning_rate": 3.4898568304099765e-08,
+      "loss": 0.4052,
+      "step": 6388
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0861909327135764,
+      "learning_rate": 3.461107881258696e-08,
+      "loss": 0.5001,
+      "step": 6389
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 11.83188469126037,
+      "learning_rate": 3.4324774260060534e-08,
+      "loss": 0.3851,
+      "step": 6390
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.051978433052861,
+      "learning_rate": 3.403965471484305e-08,
+      "loss": 0.4744,
+      "step": 6391
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 12.165182441355029,
+      "learning_rate": 3.3755720244975644e-08,
+      "loss": 0.4219,
+      "step": 6392
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 13.1365429976562,
+      "learning_rate": 3.347297091821688e-08,
+      "loss": 0.4167,
+      "step": 6393
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.189605261849392,
+      "learning_rate": 3.3191406802041693e-08,
+      "loss": 0.415,
+      "step": 6394
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.411579948771681,
+      "learning_rate": 3.2911027963642426e-08,
+      "loss": 0.4383,
+      "step": 6395
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.541251057207983,
+      "learning_rate": 3.2631834469928344e-08,
+      "loss": 0.4211,
+      "step": 6396
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.877095413449062,
+      "learning_rate": 3.235382638752671e-08,
+      "loss": 0.4256,
+      "step": 6397
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 13.256902337495772,
+      "learning_rate": 3.2077003782780556e-08,
+      "loss": 0.4473,
+      "step": 6398
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 12.187651781279879,
+      "learning_rate": 3.180136672175149e-08,
+      "loss": 0.4015,
+      "step": 6399
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.3717444832438055,
+      "learning_rate": 3.152691527021745e-08,
+      "loss": 0.4136,
+      "step": 6400
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1265465572278124,
+      "learning_rate": 3.125364949367271e-08,
+      "loss": 0.4865,
+      "step": 6401
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2124475643556294,
+      "learning_rate": 3.098156945732955e-08,
+      "loss": 0.4712,
+      "step": 6402
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.83264450950352,
+      "learning_rate": 3.0710675226117704e-08,
+      "loss": 0.3976,
+      "step": 6403
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.575853149319501,
+      "learning_rate": 3.044096686468323e-08,
+      "loss": 0.4171,
+      "step": 6404
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.290732658188067,
+      "learning_rate": 3.0172444437387984e-08,
+      "loss": 0.426,
+      "step": 6405
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.87231698407366,
+      "learning_rate": 2.990510800831348e-08,
+      "loss": 0.4058,
+      "step": 6406
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 12.767768156757501,
+      "learning_rate": 2.963895764125646e-08,
+      "loss": 0.4227,
+      "step": 6407
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.571654312087443,
+      "learning_rate": 2.9373993399730015e-08,
+      "loss": 0.3978,
+      "step": 6408
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.900859353435743,
+      "learning_rate": 2.9110215346966896e-08,
+      "loss": 0.3694,
+      "step": 6409
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.116910089222625,
+      "learning_rate": 2.8847623545913972e-08,
+      "loss": 0.421,
+      "step": 6410
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 15.75342509575384,
+      "learning_rate": 2.858621805923556e-08,
+      "loss": 0.4465,
+      "step": 6411
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.897233728633616,
+      "learning_rate": 2.8325998949314536e-08,
+      "loss": 0.3729,
+      "step": 6412
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.73804443388177,
+      "learning_rate": 2.8066966278248452e-08,
+      "loss": 0.391,
+      "step": 6413
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.894319376587845,
+      "learning_rate": 2.7809120107853972e-08,
+      "loss": 0.4669,
+      "step": 6414
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.472225806128964,
+      "learning_rate": 2.7552460499662425e-08,
+      "loss": 0.4556,
+      "step": 6415
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 16.35753675542616,
+      "learning_rate": 2.7296987514923712e-08,
+      "loss": 0.4654,
+      "step": 6416
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 12.960255954754041,
+      "learning_rate": 2.704270121460295e-08,
+      "loss": 0.4158,
+      "step": 6417
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.7894061718514624,
+      "learning_rate": 2.6789601659383825e-08,
+      "loss": 0.3923,
+      "step": 6418
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 13.719924428557155,
+      "learning_rate": 2.653768890966524e-08,
+      "loss": 0.4477,
+      "step": 6419
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.900415858421535,
+      "learning_rate": 2.6286963025564107e-08,
+      "loss": 0.3456,
+      "step": 6420
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 21.06817031022724,
+      "learning_rate": 2.603742406691312e-08,
+      "loss": 0.4639,
+      "step": 6421
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 16.21772980058928,
+      "learning_rate": 2.578907209326298e-08,
+      "loss": 0.3669,
+      "step": 6422
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.880611174455782,
+      "learning_rate": 2.5541907163879608e-08,
+      "loss": 0.3786,
+      "step": 6423
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.68457715615109,
+      "learning_rate": 2.5295929337746385e-08,
+      "loss": 0.4214,
+      "step": 6424
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.645144525956304,
+      "learning_rate": 2.505113867356357e-08,
+      "loss": 0.3484,
+      "step": 6425
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.147909665755614,
+      "learning_rate": 2.4807535229748327e-08,
+      "loss": 0.4256,
+      "step": 6426
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.524605171572896,
+      "learning_rate": 2.4565119064433595e-08,
+      "loss": 0.5083,
+      "step": 6427
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.612651645571024,
+      "learning_rate": 2.4323890235469215e-08,
+      "loss": 0.4247,
+      "step": 6428
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.193682687598692,
+      "learning_rate": 2.4083848800423027e-08,
+      "loss": 0.4329,
+      "step": 6429
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.813235195391314,
+      "learning_rate": 2.3844994816577538e-08,
+      "loss": 0.4039,
+      "step": 6430
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.969331476256643,
+      "learning_rate": 2.3607328340932713e-08,
+      "loss": 0.4301,
+      "step": 6431
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.189955892024086,
+      "learning_rate": 2.3370849430205956e-08,
+      "loss": 0.4373,
+      "step": 6432
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.793033310378792,
+      "learning_rate": 2.3135558140829906e-08,
+      "loss": 0.4075,
+      "step": 6433
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.8542222453754995,
+      "learning_rate": 2.290145452895465e-08,
+      "loss": 0.3903,
+      "step": 6434
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 13.904417722870559,
+      "learning_rate": 2.266853865044716e-08,
+      "loss": 0.442,
+      "step": 6435
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.861851640662365,
+      "learning_rate": 2.2436810560889088e-08,
+      "loss": 0.4072,
+      "step": 6436
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.510979629817108,
+      "learning_rate": 2.2206270315581203e-08,
+      "loss": 0.4119,
+      "step": 6437
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.119670279977187,
+      "learning_rate": 2.1976917969538934e-08,
+      "loss": 0.4636,
+      "step": 6438
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.1923880321675515,
+      "learning_rate": 2.174875357749462e-08,
+      "loss": 0.4003,
+      "step": 6439
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.84732399017021,
+      "learning_rate": 2.1521777193898032e-08,
+      "loss": 0.456,
+      "step": 6440
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 13.829982068583558,
+      "learning_rate": 2.129598887291473e-08,
+      "loss": 0.4334,
+      "step": 6441
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 11.99644636017591,
+      "learning_rate": 2.1071388668426063e-08,
+      "loss": 0.4352,
+      "step": 6442
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.436932653803079,
+      "learning_rate": 2.084797663403082e-08,
+      "loss": 0.3544,
+      "step": 6443
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.0265222549424315,
+      "learning_rate": 2.0625752823044686e-08,
+      "loss": 0.4098,
+      "step": 6444
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.533815621059066,
+      "learning_rate": 2.0404717288498577e-08,
+      "loss": 0.4227,
+      "step": 6445
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.676943038409958,
+      "learning_rate": 2.018487008313974e-08,
+      "loss": 0.381,
+      "step": 6446
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.973367476038849,
+      "learning_rate": 1.9966211259433433e-08,
+      "loss": 0.4027,
+      "step": 6447
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.031779613165817,
+      "learning_rate": 1.974874086955958e-08,
+      "loss": 0.4119,
+      "step": 6448
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.328670306126444,
+      "learning_rate": 1.953245896541611e-08,
+      "loss": 0.526,
+      "step": 6449
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.590351684617457,
+      "learning_rate": 1.9317365598615635e-08,
+      "loss": 0.4596,
+      "step": 6450
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.563792384853496,
+      "learning_rate": 1.9103460820488195e-08,
+      "loss": 0.4054,
+      "step": 6451
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.425370254450405,
+      "learning_rate": 1.8890744682080743e-08,
+      "loss": 0.4289,
+      "step": 6452
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.296060980983347,
+      "learning_rate": 1.8679217234154335e-08,
+      "loss": 0.3997,
+      "step": 6453
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.9644398418069136,
+      "learning_rate": 1.846887852718915e-08,
+      "loss": 0.4395,
+      "step": 6454
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.971959801762859,
+      "learning_rate": 1.825972861137948e-08,
+      "loss": 0.4254,
+      "step": 6455
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.982906798503222,
+      "learning_rate": 1.8051767536637066e-08,
+      "loss": 0.3822,
+      "step": 6456
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.471977421220158,
+      "learning_rate": 1.7844995352589988e-08,
+      "loss": 0.5026,
+      "step": 6457
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.7127969029382575,
+      "learning_rate": 1.763941210858211e-08,
+      "loss": 0.4338,
+      "step": 6458
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.80231238149647,
+      "learning_rate": 1.7435017853673632e-08,
+      "loss": 0.3877,
+      "step": 6459
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.180318034663948,
+      "learning_rate": 1.7231812636641644e-08,
+      "loss": 0.4328,
+      "step": 6460
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.83264658347131,
+      "learning_rate": 1.7029796505977912e-08,
+      "loss": 0.4705,
+      "step": 6461
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.284943448992248,
+      "learning_rate": 1.682896950989277e-08,
+      "loss": 0.4559,
+      "step": 6462
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.31368289069024,
+      "learning_rate": 1.6629331696310646e-08,
+      "loss": 0.4537,
+      "step": 6463
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.517304645963296,
+      "learning_rate": 1.6430883112872887e-08,
+      "loss": 0.4226,
+      "step": 6464
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.36662082306044,
+      "learning_rate": 1.623362380693827e-08,
+      "loss": 0.4989,
+      "step": 6465
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.240004835862762,
+      "learning_rate": 1.6037553825579145e-08,
+      "loss": 0.4172,
+      "step": 6466
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.60713598166183,
+      "learning_rate": 1.5842673215587524e-08,
+      "loss": 0.4081,
+      "step": 6467
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 16.947889877563853,
+      "learning_rate": 1.564898202346843e-08,
+      "loss": 0.4346,
+      "step": 6468
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 27.907913229975527,
+      "learning_rate": 1.5456480295443776e-08,
+      "loss": 0.4075,
+      "step": 6469
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 11.434089112970549,
+      "learning_rate": 1.5265168077453485e-08,
+      "loss": 0.3341,
+      "step": 6470
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.98913586305444,
+      "learning_rate": 1.507504541515159e-08,
+      "loss": 0.4662,
+      "step": 6471
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.395349280237933,
+      "learning_rate": 1.4886112353908467e-08,
+      "loss": 0.4337,
+      "step": 6472
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.922925738733252,
+      "learning_rate": 1.4698368938811935e-08,
+      "loss": 0.3577,
+      "step": 6473
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.582957335555795,
+      "learning_rate": 1.4511815214664493e-08,
+      "loss": 0.3827,
+      "step": 6474
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.362638682698845,
+      "learning_rate": 1.4326451225985527e-08,
+      "loss": 0.4074,
+      "step": 6475
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.472823235569069,
+      "learning_rate": 1.4142277017009653e-08,
+      "loss": 0.4173,
+      "step": 6476
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.531352106977007,
+      "learning_rate": 1.3959292631688937e-08,
+      "loss": 0.372,
+      "step": 6477
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.9699405184098495,
+      "learning_rate": 1.3777498113690114e-08,
+      "loss": 0.4276,
+      "step": 6478
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.988353369551623,
+      "learning_rate": 1.3596893506396813e-08,
+      "loss": 0.4201,
+      "step": 6479
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 11.52132734798953,
+      "learning_rate": 1.341747885290845e-08,
+      "loss": 0.4028,
+      "step": 6480
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.005767209905986,
+      "learning_rate": 1.3239254196040775e-08,
+      "loss": 0.3939,
+      "step": 6481
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 27.225755104526392,
+      "learning_rate": 1.3062219578324765e-08,
+      "loss": 0.4322,
+      "step": 6482
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.125091154746191,
+      "learning_rate": 1.2886375042008292e-08,
+      "loss": 0.4799,
+      "step": 6483
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 32.370190311616646,
+      "learning_rate": 1.2711720629055014e-08,
+      "loss": 0.5071,
+      "step": 6484
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.030081262550622,
+      "learning_rate": 1.253825638114381e-08,
+      "loss": 0.4175,
+      "step": 6485
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.618309708074467,
+      "learning_rate": 1.2365982339670456e-08,
+      "loss": 0.4048,
+      "step": 6486
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.404533079230864,
+      "learning_rate": 1.2194898545745958e-08,
+      "loss": 0.4034,
+      "step": 6487
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 11.198495791155826,
+      "learning_rate": 1.2025005040198768e-08,
+      "loss": 0.48,
+      "step": 6488
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.748227644663272,
+      "learning_rate": 1.1856301863570895e-08,
+      "loss": 0.4119,
+      "step": 6489
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.334942972879035,
+      "learning_rate": 1.1688789056122918e-08,
+      "loss": 0.3999,
+      "step": 6490
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.231604751158212,
+      "learning_rate": 1.1522466657828968e-08,
+      "loss": 0.4366,
+      "step": 6491
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.260880021913545,
+      "learning_rate": 1.135733470838063e-08,
+      "loss": 0.4054,
+      "step": 6492
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 23.534543982647225,
+      "learning_rate": 1.1193393247185268e-08,
+      "loss": 0.3619,
+      "step": 6493
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.631052704548437,
+      "learning_rate": 1.1030642313364926e-08,
+      "loss": 0.4331,
+      "step": 6494
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.3703449951391065,
+      "learning_rate": 1.0869081945759086e-08,
+      "loss": 0.4452,
+      "step": 6495
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.7736360582718165,
+      "learning_rate": 1.070871218292302e-08,
+      "loss": 0.3481,
+      "step": 6496
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 11.95300179592741,
+      "learning_rate": 1.0549533063126117e-08,
+      "loss": 0.5196,
+      "step": 6497
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.830245134714541,
+      "learning_rate": 1.0391544624355764e-08,
+      "loss": 0.4423,
+      "step": 6498
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.061023454513648,
+      "learning_rate": 1.0234746904313475e-08,
+      "loss": 0.4119,
+      "step": 6499
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.385398856404234,
+      "learning_rate": 1.0079139940418204e-08,
+      "loss": 0.4342,
+      "step": 6500
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.509958243540073,
+      "learning_rate": 9.924723769803025e-09,
+      "loss": 0.4132,
+      "step": 6501
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.835439704650566,
+      "learning_rate": 9.771498429319015e-09,
+      "loss": 0.4379,
+      "step": 6502
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.23291163722602,
+      "learning_rate": 9.619463955530817e-09,
+      "loss": 0.4798,
+      "step": 6503
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.479682096238707,
+      "learning_rate": 9.468620384720517e-09,
+      "loss": 0.4282,
+      "step": 6504
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.836813033634854,
+      "learning_rate": 9.318967752884878e-09,
+      "loss": 0.427,
+      "step": 6505
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.46167019601088,
+      "learning_rate": 9.170506095736997e-09,
+      "loss": 0.3369,
+      "step": 6506
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.130726300039751,
+      "learning_rate": 9.023235448706313e-09,
+      "loss": 0.534,
+      "step": 6507
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3674810779102473,
+      "learning_rate": 8.877155846936935e-09,
+      "loss": 0.5374,
+      "step": 6508
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.813987012740299,
+      "learning_rate": 8.732267325289313e-09,
+      "loss": 0.4373,
+      "step": 6509
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.064414994709687,
+      "learning_rate": 8.588569918339119e-09,
+      "loss": 0.4477,
+      "step": 6510
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.540154813893036,
+      "learning_rate": 8.446063660379478e-09,
+      "loss": 0.46,
+      "step": 6511
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.74478979995469,
+      "learning_rate": 8.304748585417077e-09,
+      "loss": 0.4481,
+      "step": 6512
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 11.72029709553076,
+      "learning_rate": 8.164624727175497e-09,
+      "loss": 0.4275,
+      "step": 6513
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.644498929976521,
+      "learning_rate": 8.025692119094098e-09,
+      "loss": 0.383,
+      "step": 6514
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 13.074901320376275,
+      "learning_rate": 7.88795079432747e-09,
+      "loss": 0.4662,
+      "step": 6515
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.666675314853856,
+      "learning_rate": 7.751400785746543e-09,
+      "loss": 0.4556,
+      "step": 6516
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.840353480782144,
+      "learning_rate": 7.616042125936918e-09,
+      "loss": 0.3813,
+      "step": 6517
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.339618661586898,
+      "learning_rate": 7.481874847201087e-09,
+      "loss": 0.471,
+      "step": 6518
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.546323179783108,
+      "learning_rate": 7.3488989815567735e-09,
+      "loss": 0.3861,
+      "step": 6519
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.962552033136589,
+      "learning_rate": 7.217114560736926e-09,
+      "loss": 0.4295,
+      "step": 6520
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.465100018527397,
+      "learning_rate": 7.0865216161902785e-09,
+      "loss": 0.4431,
+      "step": 6521
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.280162373325864,
+      "learning_rate": 6.9571201790818995e-09,
+      "loss": 0.4637,
+      "step": 6522
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.439019209192012,
+      "learning_rate": 6.828910280292644e-09,
+      "loss": 0.3516,
+      "step": 6523
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.612471242078843,
+      "learning_rate": 6.701891950417483e-09,
+      "loss": 0.4384,
+      "step": 6524
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 16.100636376468643,
+      "learning_rate": 6.57606521976828e-09,
+      "loss": 0.3679,
+      "step": 6525
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.748531413414321,
+      "learning_rate": 6.4514301183726814e-09,
+      "loss": 0.4333,
+      "step": 6526
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 12.888410363356453,
+      "learning_rate": 6.327986675973563e-09,
+      "loss": 0.4349,
+      "step": 6527
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.808863334609339,
+      "learning_rate": 6.2057349220295826e-09,
+      "loss": 0.3965,
+      "step": 6528
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.2689411029315725,
+      "learning_rate": 6.084674885714071e-09,
+      "loss": 0.4758,
+      "step": 6529
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.448588047583308,
+      "learning_rate": 5.9648065959172494e-09,
+      "loss": 0.4242,
+      "step": 6530
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.5780481416273915,
+      "learning_rate": 5.846130081244017e-09,
+      "loss": 0.3675,
+      "step": 6531
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.587618761015889,
+      "learning_rate": 5.728645370016162e-09,
+      "loss": 0.3922,
+      "step": 6532
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.455604512657786,
+      "learning_rate": 5.612352490270146e-09,
+      "loss": 0.4353,
+      "step": 6533
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.4866405693172,
+      "learning_rate": 5.497251469757103e-09,
+      "loss": 0.3575,
+      "step": 6534
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 11.315193835655208,
+      "learning_rate": 5.383342335945063e-09,
+      "loss": 0.3951,
+      "step": 6535
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.324848658753268,
+      "learning_rate": 5.270625116017836e-09,
+      "loss": 0.4027,
+      "step": 6536
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 13.213150254317688,
+      "learning_rate": 5.159099836873904e-09,
+      "loss": 0.4901,
+      "step": 6537
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 11.83011277915953,
+      "learning_rate": 5.048766525127535e-09,
+      "loss": 0.3583,
+      "step": 6538
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 11.540333420472034,
+      "learning_rate": 4.939625207108778e-09,
+      "loss": 0.4371,
+      "step": 6539
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.124167799218047,
+      "learning_rate": 4.831675908862909e-09,
+      "loss": 0.5524,
+      "step": 6540
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.156708144595947,
+      "learning_rate": 4.724918656150989e-09,
+      "loss": 0.4206,
+      "step": 6541
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.6059051023093405,
+      "learning_rate": 4.619353474449307e-09,
+      "loss": 0.3906,
+      "step": 6542
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.144270153131836,
+      "learning_rate": 4.514980388950485e-09,
+      "loss": 0.4635,
+      "step": 6543
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.639924959058924,
+      "learning_rate": 4.4117994245618245e-09,
+      "loss": 0.4325,
+      "step": 6544
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 38.88023063435594,
+      "learning_rate": 4.309810605906406e-09,
+      "loss": 0.372,
+      "step": 6545
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.249254992165918,
+      "learning_rate": 4.209013957322539e-09,
+      "loss": 0.3909,
+      "step": 6546
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 18.528362047653776,
+      "learning_rate": 4.109409502864314e-09,
+      "loss": 0.4541,
+      "step": 6547
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 13.014625042343413,
+      "learning_rate": 4.0109972663021635e-09,
+      "loss": 0.4147,
+      "step": 6548
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.254836553459079,
+      "learning_rate": 3.913777271120078e-09,
+      "loss": 0.4276,
+      "step": 6549
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.532058709482877,
+      "learning_rate": 3.8177495405189446e-09,
+      "loss": 0.3945,
+      "step": 6550
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.771941482763166,
+      "learning_rate": 3.7229140974148757e-09,
+      "loss": 0.3926,
+      "step": 6551
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 11.156158299037292,
+      "learning_rate": 3.629270964439213e-09,
+      "loss": 0.4161,
+      "step": 6552
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.222151091147984,
+      "learning_rate": 3.536820163939636e-09,
+      "loss": 0.44,
+      "step": 6553
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.58184317440906,
+      "learning_rate": 3.445561717977941e-09,
+      "loss": 0.4105,
+      "step": 6554
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2496824915060523,
+      "learning_rate": 3.355495648331708e-09,
+      "loss": 0.5197,
+      "step": 6555
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.906055352633679,
+      "learning_rate": 3.266621976494855e-09,
+      "loss": 0.4662,
+      "step": 6556
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.821292519469058,
+      "learning_rate": 3.1789407236759717e-09,
+      "loss": 0.4664,
+      "step": 6557
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.503331485089174,
+      "learning_rate": 3.092451910799432e-09,
+      "loss": 0.4512,
+      "step": 6558
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.748408859264843,
+      "learning_rate": 3.0071555585048374e-09,
+      "loss": 0.4857,
+      "step": 6559
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 11.606029133978572,
+      "learning_rate": 2.923051687147016e-09,
+      "loss": 0.4215,
+      "step": 6560
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.149931012582169,
+      "learning_rate": 2.8401403167971354e-09,
+      "loss": 0.4684,
+      "step": 6561
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.638336343633409,
+      "learning_rate": 2.7584214672404798e-09,
+      "loss": 0.4807,
+      "step": 6562
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.243235657287498,
+      "learning_rate": 2.6778951579786717e-09,
+      "loss": 0.3751,
+      "step": 6563
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.697764161737031,
+      "learning_rate": 2.5985614082280063e-09,
+      "loss": 0.4407,
+      "step": 6564
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.8456726226448765,
+      "learning_rate": 2.5204202369216723e-09,
+      "loss": 0.4288,
+      "step": 6565
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.727840986219048,
+      "learning_rate": 2.4434716627064204e-09,
+      "loss": 0.3953,
+      "step": 6566
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.195060705796738,
+      "learning_rate": 2.3677157039453393e-09,
+      "loss": 0.4806,
+      "step": 6567
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.34204319057369,
+      "learning_rate": 2.2931523787173005e-09,
+      "loss": 0.4438,
+      "step": 6568
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.539791407107073,
+      "learning_rate": 2.219781704815849e-09,
+      "loss": 0.415,
+      "step": 6569
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.58090501874538,
+      "learning_rate": 2.147603699749756e-09,
+      "loss": 0.442,
+      "step": 6570
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 38.58295245731931,
+      "learning_rate": 2.076618380744133e-09,
+      "loss": 0.3881,
+      "step": 6571
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9625456204477675,
+      "learning_rate": 2.006825764738762e-09,
+      "loss": 0.4412,
+      "step": 6572
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 39.65751533306592,
+      "learning_rate": 1.9382258683880994e-09,
+      "loss": 0.3999,
+      "step": 6573
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 14.39197389725519,
+      "learning_rate": 1.8708187080640485e-09,
+      "loss": 0.4232,
+      "step": 6574
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.179578887633754,
+      "learning_rate": 1.8046042998520752e-09,
+      "loss": 0.3887,
+      "step": 6575
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.451879978689478,
+      "learning_rate": 1.7395826595534293e-09,
+      "loss": 0.3454,
+      "step": 6576
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.372670869397812,
+      "learning_rate": 1.6757538026851427e-09,
+      "loss": 0.3853,
+      "step": 6577
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 13.200350153763086,
+      "learning_rate": 1.6131177444789204e-09,
+      "loss": 0.4,
+      "step": 6578
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.603486397424962,
+      "learning_rate": 1.551674499882805e-09,
+      "loss": 0.4786,
+      "step": 6579
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.893169855004569,
+      "learning_rate": 1.4914240835589566e-09,
+      "loss": 0.4106,
+      "step": 6580
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.533020205604693,
+      "learning_rate": 1.4323665098858742e-09,
+      "loss": 0.3893,
+      "step": 6581
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 11.986702272762939,
+      "learning_rate": 1.3745017929572835e-09,
+      "loss": 0.428,
+      "step": 6582
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.18985005789812,
+      "learning_rate": 1.3178299465810284e-09,
+      "loss": 0.415,
+      "step": 6583
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.2391142689865315,
+      "learning_rate": 1.2623509842824012e-09,
+      "loss": 0.3544,
+      "step": 6584
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.990832735871294,
+      "learning_rate": 1.208064919300811e-09,
+      "loss": 0.4187,
+      "step": 6585
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 25.063464070699546,
+      "learning_rate": 1.1549717645903401e-09,
+      "loss": 0.44,
+      "step": 6586
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 25.693869849188705,
+      "learning_rate": 1.1030715328208542e-09,
+      "loss": 0.5064,
+      "step": 6587
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 12.066246357515496,
+      "learning_rate": 1.052364236379111e-09,
+      "loss": 0.4221,
+      "step": 6588
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.175332876522486,
+      "learning_rate": 1.0028498873648762e-09,
+      "loss": 0.4247,
+      "step": 6589
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 11.305405270864528,
+      "learning_rate": 9.545284975948087e-10,
+      "loss": 0.3669,
+      "step": 6590
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.32658780767732,
+      "learning_rate": 9.074000786002401e-10,
+      "loss": 0.4048,
+      "step": 6591
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.559429392659263,
+      "learning_rate": 8.614646416271743e-10,
+      "loss": 0.4335,
+      "step": 6592
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.719723370195331,
+      "learning_rate": 8.167221976385087e-10,
+      "loss": 0.4058,
+      "step": 6593
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.581774500446054,
+      "learning_rate": 7.731727573112579e-10,
+      "loss": 0.4005,
+      "step": 6594
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.905245832103491,
+      "learning_rate": 7.308163310376648e-10,
+      "loss": 0.4101,
+      "step": 6595
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.825437010211374,
+      "learning_rate": 6.896529289263099e-10,
+      "loss": 0.4374,
+      "step": 6596
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 21.966968092284404,
+      "learning_rate": 6.496825607998913e-10,
+      "loss": 0.4726,
+      "step": 6597
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 10.83335582114488,
+      "learning_rate": 6.109052361974455e-10,
+      "loss": 0.412,
+      "step": 6598
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 11.801321703736502,
+      "learning_rate": 5.733209643721261e-10,
+      "loss": 0.4545,
+      "step": 6599
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 18.632137042897106,
+      "learning_rate": 5.369297542934249e-10,
+      "loss": 0.4216,
+      "step": 6600
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 17.15857827560764,
+      "learning_rate": 5.017316146460616e-10,
+      "loss": 0.415,
+      "step": 6601
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0137337311868537,
+      "learning_rate": 4.677265538288733e-10,
+      "loss": 0.5045,
+      "step": 6602
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 69.79505030196971,
+      "learning_rate": 4.3491457995759043e-10,
+      "loss": 0.4233,
+      "step": 6603
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.510704024596302,
+      "learning_rate": 4.032957008620608e-10,
+      "loss": 0.4161,
+      "step": 6604
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.869329637398582,
+      "learning_rate": 3.7286992408791525e-10,
+      "loss": 0.4433,
+      "step": 6605
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.833557035854727,
+      "learning_rate": 3.436372568954571e-10,
+      "loss": 0.428,
+      "step": 6606
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.40474502917948,
+      "learning_rate": 3.1559770626188314e-10,
+      "loss": 0.4433,
+      "step": 6607
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.947195508435455,
+      "learning_rate": 2.887512788773972e-10,
+      "loss": 0.4798,
+      "step": 6608
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 26.713151892568415,
+      "learning_rate": 2.6309798114965144e-10,
+      "loss": 0.3868,
+      "step": 6609
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 19.001760220794395,
+      "learning_rate": 2.386378191998606e-10,
+      "loss": 0.4535,
+      "step": 6610
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.791105649715406,
+      "learning_rate": 2.1537079886502222e-10,
+      "loss": 0.4605,
+      "step": 6611
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 19.269514717272305,
+      "learning_rate": 1.9329692569791668e-10,
+      "loss": 0.5126,
+      "step": 6612
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.205303360602723,
+      "learning_rate": 1.7241620496655232e-10,
+      "loss": 0.4116,
+      "step": 6613
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.960901078804037,
+      "learning_rate": 1.527286416530549e-10,
+      "loss": 0.4354,
+      "step": 6614
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.194733853937215,
+      "learning_rate": 1.342342404564434e-10,
+      "loss": 0.434,
+      "step": 6615
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.488440618503094,
+      "learning_rate": 1.1693300578985435e-10,
+      "loss": 0.3522,
+      "step": 6616
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 10.740956668666344,
+      "learning_rate": 1.0082494178220715e-10,
+      "loss": 0.4596,
+      "step": 6617
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 18.339751169153274,
+      "learning_rate": 8.591005227764904e-11,
+      "loss": 0.4394,
+      "step": 6618
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.073532136462251,
+      "learning_rate": 7.218834083499992e-11,
+      "loss": 0.4025,
+      "step": 6619
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 33.15054376801335,
+      "learning_rate": 5.965981072941774e-11,
+      "loss": 0.3995,
+      "step": 6620
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.6902728097162525,
+      "learning_rate": 4.832446494962284e-11,
+      "loss": 0.4442,
+      "step": 6621
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.682580276493064,
+      "learning_rate": 3.818230620233898e-11,
+      "loss": 0.4302,
+      "step": 6622
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.521696077434697,
+      "learning_rate": 2.9233336906187016e-11,
+      "loss": 0.4311,
+      "step": 6623
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.832883945114147,
+      "learning_rate": 2.1477559197791153e-11,
+      "loss": 0.4352,
+      "step": 6624
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.817620300061387,
+      "learning_rate": 1.491497492789318e-11,
+      "loss": 0.5094,
+      "step": 6625
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.369565709244112,
+      "learning_rate": 9.545585662462665e-12,
+      "loss": 0.454,
+      "step": 6626
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.242508760294443,
+      "learning_rate": 5.369392682696983e-12,
+      "loss": 0.3994,
+      "step": 6627
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.977820450694954,
+      "learning_rate": 2.3863969850213e-12,
+      "loss": 0.3659,
+      "step": 6628
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.541882542541116,
+      "learning_rate": 5.965992816436838e-13,
+      "loss": 0.4509,
+      "step": 6629
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.2167060447028115,
+      "learning_rate": 0.0,
+      "loss": 0.5483,
+      "step": 6630
+    },
+    {
+      "epoch": 1.0,
+      "step": 6630,
+      "total_flos": 3531656507129856.0,
+      "train_loss": 0.47476094288254217,
+      "train_runtime": 79275.3724,
+      "train_samples_per_second": 10.704,
+      "train_steps_per_second": 0.084
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 6630,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 3000,
+  "total_flos": 3531656507129856.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}