diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,39831 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.9792235231974646,
+  "eval_steps": 710,
+  "global_step": 5678,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.14370673894882202,
+      "learning_rate": 2e-05,
+      "loss": 1.4342,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "eval_loss": 1.4139364957809448,
+      "eval_runtime": 685.7235,
+      "eval_samples_per_second": 10.027,
+      "eval_steps_per_second": 5.014,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.14010629057884216,
+      "learning_rate": 4e-05,
+      "loss": 1.4461,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.14629808068275452,
+      "learning_rate": 6e-05,
+      "loss": 1.5264,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.14778390526771545,
+      "learning_rate": 8e-05,
+      "loss": 1.4242,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.13870327174663544,
+      "learning_rate": 0.0001,
+      "loss": 1.4421,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.13916651904582977,
+      "learning_rate": 0.00012,
+      "loss": 1.4097,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.14333531260490417,
+      "learning_rate": 0.00014,
+      "loss": 1.4631,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.12884660065174103,
+      "learning_rate": 0.00016,
+      "loss": 1.3788,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.10925424844026566,
+      "learning_rate": 0.00018,
+      "loss": 1.3871,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.1368756741285324,
+      "learning_rate": 0.0002,
+      "loss": 1.4336,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.14384619891643524,
+      "learning_rate": 0.00019999999616659956,
+      "loss": 1.3772,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.14267297089099884,
+      "learning_rate": 0.00019999998466639854,
+      "loss": 1.3686,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.1526922881603241,
+      "learning_rate": 0.00019999996549939782,
+      "loss": 1.3395,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.12112465500831604,
+      "learning_rate": 0.00019999993866559883,
+      "loss": 1.2573,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.1302354335784912,
+      "learning_rate": 0.00019999990416500366,
+      "loss": 1.3261,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.18590843677520752,
+      "learning_rate": 0.000199999861997615,
+      "loss": 1.3111,
+      "step": 16
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.12813763320446014,
+      "learning_rate": 0.000199999812163436,
+      "loss": 1.4044,
+      "step": 17
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.123746357858181,
+      "learning_rate": 0.00019999975466247053,
+      "loss": 1.3508,
+      "step": 18
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.13045178353786469,
+      "learning_rate": 0.00019999968949472297,
+      "loss": 1.3344,
+      "step": 19
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.1230260580778122,
+      "learning_rate": 0.00019999961666019838,
+      "loss": 1.348,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.10891468077898026,
+      "learning_rate": 0.00019999953615890226,
+      "loss": 1.3366,
+      "step": 21
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.10767161101102829,
+      "learning_rate": 0.00019999944799084082,
+      "loss": 1.2208,
+      "step": 22
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.11098026484251022,
+      "learning_rate": 0.00019999935215602084,
+      "loss": 1.3507,
+      "step": 23
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.10552258044481277,
+      "learning_rate": 0.00019999924865444962,
+      "loss": 1.285,
+      "step": 24
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.10019772499799728,
+      "learning_rate": 0.00019999913748613515,
+      "loss": 1.3506,
+      "step": 25
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.19789332151412964,
+      "learning_rate": 0.00019999901865108588,
+      "loss": 1.2869,
+      "step": 26
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.10755759477615356,
+      "learning_rate": 0.000199998892149311,
+      "loss": 1.3804,
+      "step": 27
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.1173456609249115,
+      "learning_rate": 0.00019999875798082018,
+      "loss": 1.3407,
+      "step": 28
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.10115078836679459,
+      "learning_rate": 0.00019999861614562368,
+      "loss": 1.3737,
+      "step": 29
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.10270245373249054,
+      "learning_rate": 0.0001999984666437324,
+      "loss": 1.3361,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.11115173250436783,
+      "learning_rate": 0.0001999983094751578,
+      "loss": 1.2959,
+      "step": 31
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.10555535554885864,
+      "learning_rate": 0.00019999814463991195,
+      "loss": 1.3213,
+      "step": 32
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.10262320190668106,
+      "learning_rate": 0.00019999797213800743,
+      "loss": 1.2701,
+      "step": 33
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.10323359072208405,
+      "learning_rate": 0.0001999977919694575,
+      "loss": 1.2588,
+      "step": 34
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.09919729828834534,
+      "learning_rate": 0.00019999760413427602,
+      "loss": 1.2837,
+      "step": 35
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.1005103588104248,
+      "learning_rate": 0.00019999740863247732,
+      "loss": 1.2368,
+      "step": 36
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.10752934962511063,
+      "learning_rate": 0.00019999720546407638,
+      "loss": 1.3274,
+      "step": 37
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.10152855515480042,
+      "learning_rate": 0.00019999699462908886,
+      "loss": 1.3079,
+      "step": 38
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.10640770941972733,
+      "learning_rate": 0.00019999677612753082,
+      "loss": 1.2958,
+      "step": 39
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.10559240728616714,
+      "learning_rate": 0.0001999965499594191,
+      "loss": 1.3084,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.105494424700737,
+      "learning_rate": 0.000199996316124771,
+      "loss": 1.298,
+      "step": 41
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.10321727395057678,
+      "learning_rate": 0.00019999607462360444,
+      "loss": 1.1781,
+      "step": 42
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.11195287108421326,
+      "learning_rate": 0.000199995825455938,
+      "loss": 1.2647,
+      "step": 43
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.10888040065765381,
+      "learning_rate": 0.00019999556862179064,
+      "loss": 1.2142,
+      "step": 44
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.11213697493076324,
+      "learning_rate": 0.00019999530412118222,
+      "loss": 1.2636,
+      "step": 45
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.11512047797441483,
+      "learning_rate": 0.00019999503195413293,
+      "loss": 1.3122,
+      "step": 46
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.11261572688817978,
+      "learning_rate": 0.0001999947521206636,
+      "loss": 1.3468,
+      "step": 47
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.11023643612861633,
+      "learning_rate": 0.00019999446462079577,
+      "loss": 1.2716,
+      "step": 48
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.11236844211816788,
+      "learning_rate": 0.00019999416945455146,
+      "loss": 1.1672,
+      "step": 49
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.10870081931352615,
+      "learning_rate": 0.00019999386662195324,
+      "loss": 1.2377,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.11656371504068375,
+      "learning_rate": 0.0001999935561230244,
+      "loss": 1.309,
+      "step": 51
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.11757932603359222,
+      "learning_rate": 0.00019999323795778874,
+      "loss": 1.3768,
+      "step": 52
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.11963064968585968,
+      "learning_rate": 0.00019999291212627058,
+      "loss": 1.2987,
+      "step": 53
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.12015822529792786,
+      "learning_rate": 0.000199992578628495,
+      "loss": 1.2777,
+      "step": 54
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.10828964412212372,
+      "learning_rate": 0.00019999223746448746,
+      "loss": 1.2406,
+      "step": 55
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.12686830759048462,
+      "learning_rate": 0.00019999188863427421,
+      "loss": 1.2505,
+      "step": 56
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.11787737905979156,
+      "learning_rate": 0.00019999153213788197,
+      "loss": 1.2192,
+      "step": 57
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.11279816180467606,
+      "learning_rate": 0.00019999116797533804,
+      "loss": 1.2675,
+      "step": 58
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.11661232262849808,
+      "learning_rate": 0.0001999907961466704,
+      "loss": 1.2613,
+      "step": 59
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.12005340307950974,
+      "learning_rate": 0.00019999041665190746,
+      "loss": 1.301,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.12159264087677002,
+      "learning_rate": 0.0001999900294910784,
+      "loss": 1.261,
+      "step": 61
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.12154638767242432,
+      "learning_rate": 0.0001999896346642129,
+      "loss": 1.1817,
+      "step": 62
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.1204255223274231,
+      "learning_rate": 0.0001999892321713412,
+      "loss": 1.2293,
+      "step": 63
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.11920502036809921,
+      "learning_rate": 0.00019998882201249413,
+      "loss": 1.2225,
+      "step": 64
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.12662023305892944,
+      "learning_rate": 0.00019998840418770323,
+      "loss": 1.2439,
+      "step": 65
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.12107300013303757,
+      "learning_rate": 0.00019998797869700044,
+      "loss": 1.263,
+      "step": 66
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.11776495724916458,
+      "learning_rate": 0.00019998754554041842,
+      "loss": 1.2241,
+      "step": 67
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.13094747066497803,
+      "learning_rate": 0.00019998710471799035,
+      "loss": 1.3078,
+      "step": 68
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.1258789449930191,
+      "learning_rate": 0.0001999866562297501,
+      "loss": 1.2448,
+      "step": 69
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.13428783416748047,
+      "learning_rate": 0.00019998620007573198,
+      "loss": 1.3025,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.12916766107082367,
+      "learning_rate": 0.000199985736255971,
+      "loss": 1.2096,
+      "step": 71
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.12558427453041077,
+      "learning_rate": 0.0001999852647705027,
+      "loss": 1.1837,
+      "step": 72
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.13561181724071503,
+      "learning_rate": 0.00019998478561936324,
+      "loss": 1.2618,
+      "step": 73
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.12791557610034943,
+      "learning_rate": 0.0001999842988025894,
+      "loss": 1.1922,
+      "step": 74
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.12207165360450745,
+      "learning_rate": 0.00019998380432021838,
+      "loss": 1.2192,
+      "step": 75
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.13057637214660645,
+      "learning_rate": 0.0001999833021722882,
+      "loss": 1.2109,
+      "step": 76
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1369403749704361,
+      "learning_rate": 0.00019998279235883734,
+      "loss": 1.2083,
+      "step": 77
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1295352429151535,
+      "learning_rate": 0.00019998227487990486,
+      "loss": 1.2487,
+      "step": 78
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.12950530648231506,
+      "learning_rate": 0.00019998174973553048,
+      "loss": 1.2241,
+      "step": 79
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1299756020307541,
+      "learning_rate": 0.0001999812169257544,
+      "loss": 1.198,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1389460414648056,
+      "learning_rate": 0.0001999806764506175,
+      "loss": 1.2143,
+      "step": 81
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.13119052350521088,
+      "learning_rate": 0.0001999801283101612,
+      "loss": 1.2537,
+      "step": 82
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.13153623044490814,
+      "learning_rate": 0.00019997957250442755,
+      "loss": 1.203,
+      "step": 83
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.13375413417816162,
+      "learning_rate": 0.00019997900903345918,
+      "loss": 1.1866,
+      "step": 84
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.13613435626029968,
+      "learning_rate": 0.00019997843789729925,
+      "loss": 1.2383,
+      "step": 85
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.13316839933395386,
+      "learning_rate": 0.00019997785909599153,
+      "loss": 1.1934,
+      "step": 86
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.13064943253993988,
+      "learning_rate": 0.00019997727262958045,
+      "loss": 1.1934,
+      "step": 87
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1421564519405365,
+      "learning_rate": 0.00019997667849811098,
+      "loss": 1.2651,
+      "step": 88
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1368361860513687,
+      "learning_rate": 0.00019997607670162862,
+      "loss": 1.2236,
+      "step": 89
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1279619336128235,
+      "learning_rate": 0.0001999754672401795,
+      "loss": 1.224,
+      "step": 90
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.13549309968948364,
+      "learning_rate": 0.0001999748501138104,
+      "loss": 1.2221,
+      "step": 91
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.131364107131958,
+      "learning_rate": 0.0001999742253225686,
+      "loss": 1.2194,
+      "step": 92
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.13993385434150696,
+      "learning_rate": 0.00019997359286650203,
+      "loss": 1.2473,
+      "step": 93
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.13425293564796448,
+      "learning_rate": 0.00019997295274565917,
+      "loss": 1.1737,
+      "step": 94
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1355249583721161,
+      "learning_rate": 0.00019997230496008906,
+      "loss": 1.2063,
+      "step": 95
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.14003869891166687,
+      "learning_rate": 0.00019997164950984138,
+      "loss": 1.203,
+      "step": 96
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.13123658299446106,
+      "learning_rate": 0.0001999709863949664,
+      "loss": 1.1163,
+      "step": 97
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1321428418159485,
+      "learning_rate": 0.00019997031561551498,
+      "loss": 1.17,
+      "step": 98
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.13351233303546906,
+      "learning_rate": 0.0001999696371715385,
+      "loss": 1.1809,
+      "step": 99
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.12996333837509155,
+      "learning_rate": 0.000199968951063089,
+      "loss": 1.1476,
+      "step": 100
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.13364213705062866,
+      "learning_rate": 0.0001999682572902191,
+      "loss": 1.1802,
+      "step": 101
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.13364386558532715,
+      "learning_rate": 0.00019996755585298192,
+      "loss": 1.2397,
+      "step": 102
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.137208953499794,
+      "learning_rate": 0.0001999668467514313,
+      "loss": 1.2739,
+      "step": 103
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1342509239912033,
+      "learning_rate": 0.00019996612998562162,
+      "loss": 1.2535,
+      "step": 104
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.15147949755191803,
+      "learning_rate": 0.0001999654055556078,
+      "loss": 1.1905,
+      "step": 105
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.12738168239593506,
+      "learning_rate": 0.00019996467346144537,
+      "loss": 1.1332,
+      "step": 106
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1382182538509369,
+      "learning_rate": 0.00019996393370319047,
+      "loss": 1.2299,
+      "step": 107
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.14213505387306213,
+      "learning_rate": 0.0001999631862808998,
+      "loss": 1.2613,
+      "step": 108
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.13719429075717926,
+      "learning_rate": 0.00019996243119463074,
+      "loss": 1.2,
+      "step": 109
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.13508781790733337,
+      "learning_rate": 0.00019996166844444107,
+      "loss": 1.1761,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1376897543668747,
+      "learning_rate": 0.00019996089803038937,
+      "loss": 1.2237,
+      "step": 111
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.14335481822490692,
+      "learning_rate": 0.00019996011995253462,
+      "loss": 1.1745,
+      "step": 112
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1393982619047165,
+      "learning_rate": 0.0001999593342109365,
+      "loss": 1.2872,
+      "step": 113
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.14151644706726074,
+      "learning_rate": 0.00019995854080565528,
+      "loss": 1.2365,
+      "step": 114
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.13480781018733978,
+      "learning_rate": 0.0001999577397367518,
+      "loss": 1.2138,
+      "step": 115
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1416129618883133,
+      "learning_rate": 0.0001999569310042874,
+      "loss": 1.2345,
+      "step": 116
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1319161355495453,
+      "learning_rate": 0.00019995611460832415,
+      "loss": 1.1598,
+      "step": 117
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.13647343218326569,
+      "learning_rate": 0.00019995529054892463,
+      "loss": 1.1749,
+      "step": 118
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1369137465953827,
+      "learning_rate": 0.00019995445882615203,
+      "loss": 1.2051,
+      "step": 119
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1341824233531952,
+      "learning_rate": 0.00019995361944007007,
+      "loss": 1.262,
+      "step": 120
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1396571397781372,
+      "learning_rate": 0.00019995277239074315,
+      "loss": 1.1545,
+      "step": 121
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.13041186332702637,
+      "learning_rate": 0.0001999519176782362,
+      "loss": 1.1699,
+      "step": 122
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.12751354277133942,
+      "learning_rate": 0.0001999510553026147,
+      "loss": 1.1315,
+      "step": 123
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.13470140099525452,
+      "learning_rate": 0.00019995018526394485,
+      "loss": 1.2279,
+      "step": 124
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1435287892818451,
+      "learning_rate": 0.0001999493075622933,
+      "loss": 1.224,
+      "step": 125
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.14406907558441162,
+      "learning_rate": 0.00019994842219772739,
+      "loss": 1.1869,
+      "step": 126
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.13931530714035034,
+      "learning_rate": 0.00019994752917031492,
+      "loss": 1.2554,
+      "step": 127
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.13751220703125,
+      "learning_rate": 0.0001999466284801244,
+      "loss": 1.2116,
+      "step": 128
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1349681317806244,
+      "learning_rate": 0.00019994572012722492,
+      "loss": 1.1736,
+      "step": 129
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1377020627260208,
+      "learning_rate": 0.00019994480411168609,
+      "loss": 1.1921,
+      "step": 130
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.13730505108833313,
+      "learning_rate": 0.00019994388043357807,
+      "loss": 1.2046,
+      "step": 131
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.14024712145328522,
+      "learning_rate": 0.0001999429490929718,
+      "loss": 1.1561,
+      "step": 132
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1297326534986496,
+      "learning_rate": 0.00019994201008993864,
+      "loss": 1.1856,
+      "step": 133
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.129353865981102,
+      "learning_rate": 0.00019994106342455053,
+      "loss": 1.1283,
+      "step": 134
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.14311642944812775,
+      "learning_rate": 0.00019994010909688012,
+      "loss": 1.2684,
+      "step": 135
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.13943718373775482,
+      "learning_rate": 0.00019993914710700052,
+      "loss": 1.2236,
+      "step": 136
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.14227281510829926,
+      "learning_rate": 0.0001999381774549855,
+      "loss": 1.287,
+      "step": 137
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.14150571823120117,
+      "learning_rate": 0.00019993720014090945,
+      "loss": 1.1848,
+      "step": 138
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1421256810426712,
+      "learning_rate": 0.0001999362151648472,
+      "loss": 1.1925,
+      "step": 139
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1376809924840927,
+      "learning_rate": 0.00019993522252687438,
+      "loss": 1.2419,
+      "step": 140
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.15068894624710083,
+      "learning_rate": 0.00019993422222706704,
+      "loss": 1.206,
+      "step": 141
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.15035270154476166,
+      "learning_rate": 0.0001999332142655018,
+      "loss": 1.2368,
+      "step": 142
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.136764794588089,
+      "learning_rate": 0.00019993219864225607,
+      "loss": 1.162,
+      "step": 143
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1427554190158844,
+      "learning_rate": 0.00019993117535740764,
+      "loss": 1.1967,
+      "step": 144
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1433485597372055,
+      "learning_rate": 0.00019993014441103496,
+      "loss": 1.1817,
+      "step": 145
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.13792641460895538,
+      "learning_rate": 0.0001999291058032171,
+      "loss": 1.1668,
+      "step": 146
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1383659839630127,
+      "learning_rate": 0.0001999280595340337,
+      "loss": 1.2255,
+      "step": 147
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.13606032729148865,
+      "learning_rate": 0.0001999270056035649,
+      "loss": 1.1995,
+      "step": 148
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.14569996297359467,
+      "learning_rate": 0.00019992594401189157,
+      "loss": 1.2617,
+      "step": 149
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.13744774460792542,
+      "learning_rate": 0.00019992487475909508,
+      "loss": 1.2096,
+      "step": 150
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.13937488198280334,
+      "learning_rate": 0.0001999237978452574,
+      "loss": 1.2524,
+      "step": 151
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.13918696343898773,
+      "learning_rate": 0.00019992271327046114,
+      "loss": 1.1093,
+      "step": 152
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.13565635681152344,
+      "learning_rate": 0.0001999216210347894,
+      "loss": 1.1467,
+      "step": 153
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.14065219461917877,
+      "learning_rate": 0.00019992052113832593,
+      "loss": 1.1114,
+      "step": 154
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.14353357255458832,
+      "learning_rate": 0.0001999194135811551,
+      "loss": 1.1444,
+      "step": 155
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.14801861345767975,
+      "learning_rate": 0.00019991829836336176,
+      "loss": 1.2434,
+      "step": 156
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.15463212132453918,
+      "learning_rate": 0.00019991717548503143,
+      "loss": 1.1974,
+      "step": 157
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.15134255588054657,
+      "learning_rate": 0.00019991604494625021,
+      "loss": 1.18,
+      "step": 158
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.15339814126491547,
+      "learning_rate": 0.0001999149067471048,
+      "loss": 1.1829,
+      "step": 159
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.14708739519119263,
+      "learning_rate": 0.0001999137608876824,
+      "loss": 1.1795,
+      "step": 160
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.15173062682151794,
+      "learning_rate": 0.0001999126073680709,
+      "loss": 1.2195,
+      "step": 161
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.16870474815368652,
+      "learning_rate": 0.00019991144618835877,
+      "loss": 1.2889,
+      "step": 162
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.14730185270309448,
+      "learning_rate": 0.000199910277348635,
+      "loss": 1.1631,
+      "step": 163
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.14397159218788147,
+      "learning_rate": 0.0001999091008489892,
+      "loss": 1.1906,
+      "step": 164
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1512182652950287,
+      "learning_rate": 0.00019990791668951155,
+      "loss": 1.1496,
+      "step": 165
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.15269646048545837,
+      "learning_rate": 0.00019990672487029288,
+      "loss": 1.1805,
+      "step": 166
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1465330868959427,
+      "learning_rate": 0.00019990552539142454,
+      "loss": 1.2078,
+      "step": 167
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.14436009526252747,
+      "learning_rate": 0.0001999043182529985,
+      "loss": 1.2336,
+      "step": 168
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1508115530014038,
+      "learning_rate": 0.00019990310345510733,
+      "loss": 1.304,
+      "step": 169
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.14511720836162567,
+      "learning_rate": 0.00019990188099784412,
+      "loss": 1.2004,
+      "step": 170
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.14771127700805664,
+      "learning_rate": 0.0001999006508813026,
+      "loss": 1.2467,
+      "step": 171
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.14662198722362518,
+      "learning_rate": 0.0001998994131055771,
+      "loss": 1.1317,
+      "step": 172
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1433134377002716,
+      "learning_rate": 0.00019989816767076253,
+      "loss": 1.1357,
+      "step": 173
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.15154090523719788,
+      "learning_rate": 0.0001998969145769544,
+      "loss": 1.2298,
+      "step": 174
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.13884200155735016,
+      "learning_rate": 0.00019989565382424869,
+      "loss": 1.2133,
+      "step": 175
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.14153452217578888,
+      "learning_rate": 0.0001998943854127421,
+      "loss": 1.1647,
+      "step": 176
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.14802685379981995,
+      "learning_rate": 0.00019989310934253191,
+      "loss": 1.1019,
+      "step": 177
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1560146063566208,
+      "learning_rate": 0.00019989182561371593,
+      "loss": 1.168,
+      "step": 178
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.14783112704753876,
+      "learning_rate": 0.0001998905342263926,
+      "loss": 1.2003,
+      "step": 179
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.14809173345565796,
+      "learning_rate": 0.00019988923518066089,
+      "loss": 1.2171,
+      "step": 180
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1465620994567871,
+      "learning_rate": 0.00019988792847662042,
+      "loss": 1.1109,
+      "step": 181
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.14426849782466888,
+      "learning_rate": 0.00019988661411437137,
+      "loss": 1.1649,
+      "step": 182
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.15189453959465027,
+      "learning_rate": 0.0001998852920940145,
+      "loss": 1.1772,
+      "step": 183
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.16095630824565887,
+      "learning_rate": 0.00019988396241565117,
+      "loss": 1.0906,
+      "step": 184
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.13917481899261475,
+      "learning_rate": 0.00019988262507938335,
+      "loss": 1.1591,
+      "step": 185
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.15862134099006653,
+      "learning_rate": 0.00019988128008531354,
+      "loss": 1.1531,
+      "step": 186
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.14392463862895966,
+      "learning_rate": 0.00019987992743354487,
+      "loss": 1.2193,
+      "step": 187
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.14952170848846436,
+      "learning_rate": 0.00019987856712418104,
+      "loss": 1.2214,
+      "step": 188
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1452377885580063,
+      "learning_rate": 0.00019987719915732634,
+      "loss": 1.1678,
+      "step": 189
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1571136713027954,
+      "learning_rate": 0.00019987582353308566,
+      "loss": 1.1832,
+      "step": 190
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.15406474471092224,
+      "learning_rate": 0.00019987444025156446,
+      "loss": 1.1609,
+      "step": 191
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.16054005920886993,
+      "learning_rate": 0.0001998730493128688,
+      "loss": 1.1787,
+      "step": 192
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.14553441107273102,
+      "learning_rate": 0.00019987165071710527,
+      "loss": 1.1636,
+      "step": 193
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.14451570808887482,
+      "learning_rate": 0.0001998702444643812,
+      "loss": 1.1658,
+      "step": 194
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.149046391248703,
+      "learning_rate": 0.0001998688305548043,
+      "loss": 1.1452,
+      "step": 195
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1507096290588379,
+      "learning_rate": 0.00019986740898848306,
+      "loss": 1.1181,
+      "step": 196
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.150601327419281,
+      "learning_rate": 0.0001998659797655264,
+      "loss": 1.1789,
+      "step": 197
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.14738313853740692,
+      "learning_rate": 0.00019986454288604394,
+      "loss": 1.1728,
+      "step": 198
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.15052588284015656,
+      "learning_rate": 0.0001998630983501458,
+      "loss": 1.2037,
+      "step": 199
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.15669110417366028,
+      "learning_rate": 0.0001998616461579428,
+      "loss": 1.1426,
+      "step": 200
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.155398428440094,
+      "learning_rate": 0.00019986018630954616,
+      "loss": 1.1367,
+      "step": 201
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.14136084914207458,
+      "learning_rate": 0.00019985871880506793,
+      "loss": 1.1654,
+      "step": 202
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.14925213158130646,
+      "learning_rate": 0.00019985724364462057,
+      "loss": 1.1997,
+      "step": 203
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.15234960615634918,
+      "learning_rate": 0.00019985576082831715,
+      "loss": 1.142,
+      "step": 204
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1498517394065857,
+      "learning_rate": 0.00019985427035627137,
+      "loss": 1.1695,
+      "step": 205
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.14819611608982086,
+      "learning_rate": 0.00019985277222859754,
+      "loss": 1.1948,
+      "step": 206
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.14833801984786987,
+      "learning_rate": 0.00019985126644541046,
+      "loss": 1.2263,
+      "step": 207
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.138643279671669,
+      "learning_rate": 0.0001998497530068256,
+      "loss": 1.1655,
+      "step": 208
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.15962731838226318,
+      "learning_rate": 0.000199848231912959,
+      "loss": 1.2159,
+      "step": 209
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.14703673124313354,
+      "learning_rate": 0.00019984670316392728,
+      "loss": 1.1169,
+      "step": 210
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.14648592472076416,
+      "learning_rate": 0.0001998451667598477,
+      "loss": 1.266,
+      "step": 211
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.15541663765907288,
+      "learning_rate": 0.0001998436227008379,
+      "loss": 1.1395,
+      "step": 212
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.14851492643356323,
+      "learning_rate": 0.00019984207098701638,
+      "loss": 1.1509,
+      "step": 213
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.15682092308998108,
+      "learning_rate": 0.0001998405116185021,
+      "loss": 1.2423,
+      "step": 214
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.16800767183303833,
+      "learning_rate": 0.0001998389445954146,
+      "loss": 1.1532,
+      "step": 215
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1449185460805893,
+      "learning_rate": 0.000199837369917874,
+      "loss": 1.1184,
+      "step": 216
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.15644285082817078,
+      "learning_rate": 0.00019983578758600104,
+      "loss": 1.2339,
+      "step": 217
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.14295175671577454,
+      "learning_rate": 0.00019983419759991703,
+      "loss": 1.1165,
+      "step": 218
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.15597833693027496,
+      "learning_rate": 0.0001998325999597439,
+      "loss": 1.2016,
+      "step": 219
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.14484375715255737,
+      "learning_rate": 0.0001998309946656041,
+      "loss": 1.1666,
+      "step": 220
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.16649185121059418,
+      "learning_rate": 0.0001998293817176207,
+      "loss": 1.1367,
+      "step": 221
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1692935675382614,
+      "learning_rate": 0.00019982776111591744,
+      "loss": 1.1454,
+      "step": 222
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1514350026845932,
+      "learning_rate": 0.00019982613286061846,
+      "loss": 1.1746,
+      "step": 223
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.15292806923389435,
+      "learning_rate": 0.00019982449695184865,
+      "loss": 1.1006,
+      "step": 224
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.16936799883842468,
+      "learning_rate": 0.00019982285338973344,
+      "loss": 1.1535,
+      "step": 225
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1575537770986557,
+      "learning_rate": 0.00019982120217439884,
+      "loss": 1.2035,
+      "step": 226
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.14317381381988525,
+      "learning_rate": 0.00019981954330597143,
+      "loss": 1.1423,
+      "step": 227
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.16129539906978607,
+      "learning_rate": 0.00019981787678457837,
+      "loss": 1.0953,
+      "step": 228
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1630077362060547,
+      "learning_rate": 0.00019981620261034743,
+      "loss": 1.0997,
+      "step": 229
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.14313393831253052,
+      "learning_rate": 0.00019981452078340702,
+      "loss": 1.1185,
+      "step": 230
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.15655392408370972,
+      "learning_rate": 0.00019981283130388606,
+      "loss": 1.0938,
+      "step": 231
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.14617329835891724,
+      "learning_rate": 0.00019981113417191407,
+      "loss": 1.0718,
+      "step": 232
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.15672458708286285,
+      "learning_rate": 0.00019980942938762115,
+      "loss": 1.1351,
+      "step": 233
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.15482361614704132,
+      "learning_rate": 0.000199807716951138,
+      "loss": 1.1938,
+      "step": 234
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.15706798434257507,
+      "learning_rate": 0.00019980599686259594,
+      "loss": 1.2067,
+      "step": 235
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1500590741634369,
+      "learning_rate": 0.00019980426912212686,
+      "loss": 1.1681,
+      "step": 236
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.16141961514949799,
+      "learning_rate": 0.00019980253372986316,
+      "loss": 1.1418,
+      "step": 237
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1429641842842102,
+      "learning_rate": 0.00019980079068593793,
+      "loss": 1.1352,
+      "step": 238
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.15207724273204803,
+      "learning_rate": 0.00019979903999048482,
+      "loss": 1.2145,
+      "step": 239
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.16128486394882202,
+      "learning_rate": 0.000199797281643638,
+      "loss": 1.14,
+      "step": 240
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.15816397964954376,
+      "learning_rate": 0.00019979551564553233,
+      "loss": 1.1786,
+      "step": 241
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.15580213069915771,
+      "learning_rate": 0.0001997937419963032,
+      "loss": 1.1646,
+      "step": 242
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.14587156474590302,
+      "learning_rate": 0.0001997919606960865,
+      "loss": 1.1389,
+      "step": 243
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.14832736551761627,
+      "learning_rate": 0.00019979017174501896,
+      "loss": 1.0983,
+      "step": 244
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1540011465549469,
+      "learning_rate": 0.00019978837514323763,
+      "loss": 1.122,
+      "step": 245
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1595630794763565,
+      "learning_rate": 0.00019978657089088028,
+      "loss": 1.1561,
+      "step": 246
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.14259830117225647,
+      "learning_rate": 0.00019978475898808524,
+      "loss": 1.0869,
+      "step": 247
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.14668312668800354,
+      "learning_rate": 0.0001997829394349914,
+      "loss": 1.1542,
+      "step": 248
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.14027823507785797,
+      "learning_rate": 0.00019978111223173828,
+      "loss": 1.1879,
+      "step": 249
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.15351636707782745,
+      "learning_rate": 0.00019977927737846595,
+      "loss": 1.1686,
+      "step": 250
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.15014703571796417,
+      "learning_rate": 0.00019977743487531514,
+      "loss": 1.1307,
+      "step": 251
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1641966849565506,
+      "learning_rate": 0.00019977558472242704,
+      "loss": 1.1601,
+      "step": 252
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.15307414531707764,
+      "learning_rate": 0.00019977372691994355,
+      "loss": 1.1866,
+      "step": 253
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.15638311207294464,
+      "learning_rate": 0.00019977186146800707,
+      "loss": 1.2126,
+      "step": 254
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.14870685338974,
+      "learning_rate": 0.00019976998836676067,
+      "loss": 1.1191,
+      "step": 255
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1501302719116211,
+      "learning_rate": 0.00019976810761634787,
+      "loss": 1.172,
+      "step": 256
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.15548266470432281,
+      "learning_rate": 0.00019976621921691297,
+      "loss": 1.139,
+      "step": 257
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.15052320063114166,
+      "learning_rate": 0.00019976432316860067,
+      "loss": 1.1248,
+      "step": 258
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.14842264354228973,
+      "learning_rate": 0.00019976241947155634,
+      "loss": 1.2016,
+      "step": 259
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.15101812779903412,
+      "learning_rate": 0.00019976050812592594,
+      "loss": 1.1772,
+      "step": 260
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.14940795302391052,
+      "learning_rate": 0.00019975858913185609,
+      "loss": 1.1316,
+      "step": 261
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.14583760499954224,
+      "learning_rate": 0.00019975666248949378,
+      "loss": 1.1271,
+      "step": 262
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.15480725467205048,
+      "learning_rate": 0.00019975472819898683,
+      "loss": 1.1886,
+      "step": 263
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.14589324593544006,
+      "learning_rate": 0.0001997527862604835,
+      "loss": 1.2016,
+      "step": 264
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.14960989356040955,
+      "learning_rate": 0.00019975083667413265,
+      "loss": 1.1269,
+      "step": 265
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1421205699443817,
+      "learning_rate": 0.00019974887944008377,
+      "loss": 1.052,
+      "step": 266
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.15329426527023315,
+      "learning_rate": 0.00019974691455848696,
+      "loss": 1.1607,
+      "step": 267
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.14127160608768463,
+      "learning_rate": 0.0001997449420294928,
+      "loss": 1.1725,
+      "step": 268
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.16101525723934174,
+      "learning_rate": 0.00019974296185325255,
+      "loss": 1.1535,
+      "step": 269
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.15944872796535492,
+      "learning_rate": 0.000199740974029918,
+      "loss": 1.1134,
+      "step": 270
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1493426114320755,
+      "learning_rate": 0.0001997389785596416,
+      "loss": 1.1452,
+      "step": 271
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.14942991733551025,
+      "learning_rate": 0.0001997369754425763,
+      "loss": 1.2205,
+      "step": 272
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.15282519161701202,
+      "learning_rate": 0.00019973496467887568,
+      "loss": 1.1538,
+      "step": 273
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1616443246603012,
+      "learning_rate": 0.00019973294626869393,
+      "loss": 1.2279,
+      "step": 274
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.15750998258590698,
+      "learning_rate": 0.00019973092021218576,
+      "loss": 1.1482,
+      "step": 275
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.14668568968772888,
+      "learning_rate": 0.00019972888650950652,
+      "loss": 1.1397,
+      "step": 276
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.14951111376285553,
+      "learning_rate": 0.00019972684516081214,
+      "loss": 1.1073,
+      "step": 277
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.15213429927825928,
+      "learning_rate": 0.0001997247961662591,
+      "loss": 1.1308,
+      "step": 278
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.18259520828723907,
+      "learning_rate": 0.00019972273952600448,
+      "loss": 1.1946,
+      "step": 279
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.15054333209991455,
+      "learning_rate": 0.00019972067524020605,
+      "loss": 1.1469,
+      "step": 280
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.15350620448589325,
+      "learning_rate": 0.00019971860330902198,
+      "loss": 1.164,
+      "step": 281
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.16368921101093292,
+      "learning_rate": 0.00019971652373261116,
+      "loss": 1.2219,
+      "step": 282
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1547555923461914,
+      "learning_rate": 0.000199714436511133,
+      "loss": 1.091,
+      "step": 283
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1510656177997589,
+      "learning_rate": 0.00019971234164474754,
+      "loss": 1.1535,
+      "step": 284
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.15672382712364197,
+      "learning_rate": 0.0001997102391336154,
+      "loss": 1.0743,
+      "step": 285
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.16902947425842285,
+      "learning_rate": 0.0001997081289778978,
+      "loss": 1.1292,
+      "step": 286
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.14788220822811127,
+      "learning_rate": 0.00019970601117775644,
+      "loss": 1.1239,
+      "step": 287
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1836184412240982,
+      "learning_rate": 0.00019970388573335374,
+      "loss": 1.1346,
+      "step": 288
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.14677155017852783,
+      "learning_rate": 0.00019970175264485266,
+      "loss": 1.1423,
+      "step": 289
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.15332669019699097,
+      "learning_rate": 0.00019969961191241675,
+      "loss": 1.1339,
+      "step": 290
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1484992951154709,
+      "learning_rate": 0.0001996974635362101,
+      "loss": 1.169,
+      "step": 291
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1394282877445221,
+      "learning_rate": 0.00019969530751639742,
+      "loss": 1.0936,
+      "step": 292
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.15574292838573456,
+      "learning_rate": 0.00019969314385314405,
+      "loss": 1.1508,
+      "step": 293
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.15783685445785522,
+      "learning_rate": 0.0001996909725466158,
+      "loss": 1.1752,
+      "step": 294
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1484525501728058,
+      "learning_rate": 0.00019968879359697926,
+      "loss": 1.1611,
+      "step": 295
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1568101942539215,
+      "learning_rate": 0.00019968660700440136,
+      "loss": 1.0999,
+      "step": 296
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1591518223285675,
+      "learning_rate": 0.0001996844127690498,
+      "loss": 1.0899,
+      "step": 297
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.14987587928771973,
+      "learning_rate": 0.00019968221089109283,
+      "loss": 1.1148,
+      "step": 298
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1563010960817337,
+      "learning_rate": 0.0001996800013706992,
+      "loss": 1.0537,
+      "step": 299
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.16013358533382416,
+      "learning_rate": 0.00019967778420803834,
+      "loss": 1.0721,
+      "step": 300
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.16744442284107208,
+      "learning_rate": 0.00019967555940328028,
+      "loss": 1.1195,
+      "step": 301
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.14828696846961975,
+      "learning_rate": 0.00019967332695659554,
+      "loss": 1.0544,
+      "step": 302
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.16506452858448029,
+      "learning_rate": 0.00019967108686815528,
+      "loss": 1.1455,
+      "step": 303
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.178831547498703,
+      "learning_rate": 0.00019966883913813125,
+      "loss": 1.0979,
+      "step": 304
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.15071475505828857,
+      "learning_rate": 0.00019966658376669577,
+      "loss": 1.1009,
+      "step": 305
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.17837685346603394,
+      "learning_rate": 0.00019966432075402182,
+      "loss": 1.1708,
+      "step": 306
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.18097737431526184,
+      "learning_rate": 0.0001996620501002828,
+      "loss": 1.1184,
+      "step": 307
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.14264380931854248,
+      "learning_rate": 0.00019965977180565284,
+      "loss": 1.1471,
+      "step": 308
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1658606231212616,
+      "learning_rate": 0.00019965748587030664,
+      "loss": 1.0892,
+      "step": 309
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.16822941601276398,
+      "learning_rate": 0.00019965519229441943,
+      "loss": 1.135,
+      "step": 310
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1563027799129486,
+      "learning_rate": 0.00019965289107816707,
+      "loss": 1.2013,
+      "step": 311
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.15105265378952026,
+      "learning_rate": 0.00019965058222172594,
+      "loss": 1.1629,
+      "step": 312
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.15457677841186523,
+      "learning_rate": 0.00019964826572527314,
+      "loss": 1.1253,
+      "step": 313
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.18937544524669647,
+      "learning_rate": 0.00019964594158898617,
+      "loss": 1.155,
+      "step": 314
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1488984227180481,
+      "learning_rate": 0.0001996436098130433,
+      "loss": 1.1303,
+      "step": 315
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1458614617586136,
+      "learning_rate": 0.0001996412703976233,
+      "loss": 1.1398,
+      "step": 316
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1635604351758957,
+      "learning_rate": 0.00019963892334290547,
+      "loss": 1.0879,
+      "step": 317
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1525888293981552,
+      "learning_rate": 0.00019963656864906978,
+      "loss": 1.1006,
+      "step": 318
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.15084509551525116,
+      "learning_rate": 0.00019963420631629678,
+      "loss": 1.1898,
+      "step": 319
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.16898000240325928,
+      "learning_rate": 0.00019963183634476756,
+      "loss": 1.1698,
+      "step": 320
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.15749044716358185,
+      "learning_rate": 0.00019962945873466383,
+      "loss": 1.212,
+      "step": 321
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.15450423955917358,
+      "learning_rate": 0.0001996270734861679,
+      "loss": 1.162,
+      "step": 322
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.15154510736465454,
+      "learning_rate": 0.0001996246805994626,
+      "loss": 1.1085,
+      "step": 323
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.15856508910655975,
+      "learning_rate": 0.00019962228007473144,
+      "loss": 1.0942,
+      "step": 324
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.17117822170257568,
+      "learning_rate": 0.00019961987191215841,
+      "loss": 1.1871,
+      "step": 325
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.15272203087806702,
+      "learning_rate": 0.00019961745611192817,
+      "loss": 1.0386,
+      "step": 326
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.15557795763015747,
+      "learning_rate": 0.0001996150326742259,
+      "loss": 1.121,
+      "step": 327
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.16436509788036346,
+      "learning_rate": 0.00019961260159923744,
+      "loss": 1.1567,
+      "step": 328
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.16334863007068634,
+      "learning_rate": 0.00019961016288714918,
+      "loss": 1.0672,
+      "step": 329
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.16035060584545135,
+      "learning_rate": 0.00019960771653814807,
+      "loss": 1.1199,
+      "step": 330
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.15284912288188934,
+      "learning_rate": 0.00019960526255242164,
+      "loss": 1.0992,
+      "step": 331
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.15060201287269592,
+      "learning_rate": 0.00019960280093015808,
+      "loss": 1.1156,
+      "step": 332
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.16274458169937134,
+      "learning_rate": 0.0001996003316715461,
+      "loss": 1.1305,
+      "step": 333
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1730499416589737,
+      "learning_rate": 0.00019959785477677503,
+      "loss": 1.1174,
+      "step": 334
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.15439949929714203,
+      "learning_rate": 0.0001995953702460347,
+      "loss": 1.0511,
+      "step": 335
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.15664365887641907,
+      "learning_rate": 0.0001995928780795157,
+      "loss": 1.141,
+      "step": 336
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.16615834832191467,
+      "learning_rate": 0.00019959037827740902,
+      "loss": 1.156,
+      "step": 337
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.15994668006896973,
+      "learning_rate": 0.00019958787083990633,
+      "loss": 1.1022,
+      "step": 338
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.15766587853431702,
+      "learning_rate": 0.00019958535576719991,
+      "loss": 1.1335,
+      "step": 339
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.15908043086528778,
+      "learning_rate": 0.00019958283305948252,
+      "loss": 1.172,
+      "step": 340
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.16546039283275604,
+      "learning_rate": 0.00019958030271694766,
+      "loss": 1.2044,
+      "step": 341
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.16330744326114655,
+      "learning_rate": 0.00019957776473978923,
+      "loss": 1.1237,
+      "step": 342
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1535853147506714,
+      "learning_rate": 0.00019957521912820187,
+      "loss": 1.114,
+      "step": 343
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1673428863286972,
+      "learning_rate": 0.0001995726658823807,
+      "loss": 1.0633,
+      "step": 344
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.15542234480381012,
+      "learning_rate": 0.00019957010500252153,
+      "loss": 1.1305,
+      "step": 345
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.15837039053440094,
+      "learning_rate": 0.00019956753648882068,
+      "loss": 1.1461,
+      "step": 346
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.16113828122615814,
+      "learning_rate": 0.00019956496034147505,
+      "loss": 1.1012,
+      "step": 347
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.150481179356575,
+      "learning_rate": 0.00019956237656068218,
+      "loss": 1.0548,
+      "step": 348
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.15013793110847473,
+      "learning_rate": 0.00019955978514664012,
+      "loss": 1.0639,
+      "step": 349
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.16781456768512726,
+      "learning_rate": 0.00019955718609954758,
+      "loss": 1.0529,
+      "step": 350
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.15649698674678802,
+      "learning_rate": 0.00019955457941960383,
+      "loss": 1.0471,
+      "step": 351
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.15318119525909424,
+      "learning_rate": 0.0001995519651070087,
+      "loss": 1.1111,
+      "step": 352
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1564999669790268,
+      "learning_rate": 0.00019954934316196261,
+      "loss": 1.1563,
+      "step": 353
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.15984880924224854,
+      "learning_rate": 0.00019954671358466663,
+      "loss": 1.1305,
+      "step": 354
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.17183969914913177,
+      "learning_rate": 0.00019954407637532234,
+      "loss": 1.1312,
+      "step": 355
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.15881489217281342,
+      "learning_rate": 0.0001995414315341319,
+      "loss": 1.1161,
+      "step": 356
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.15662528574466705,
+      "learning_rate": 0.0001995387790612981,
+      "loss": 1.1191,
+      "step": 357
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1573573797941208,
+      "learning_rate": 0.00019953611895702435,
+      "loss": 1.0853,
+      "step": 358
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.15552809834480286,
+      "learning_rate": 0.00019953345122151453,
+      "loss": 1.0935,
+      "step": 359
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.15939250588417053,
+      "learning_rate": 0.00019953077585497319,
+      "loss": 1.1027,
+      "step": 360
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1561977118253708,
+      "learning_rate": 0.00019952809285760545,
+      "loss": 1.1131,
+      "step": 361
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.15809029340744019,
+      "learning_rate": 0.000199525402229617,
+      "loss": 1.1493,
+      "step": 362
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1678066998720169,
+      "learning_rate": 0.00019952270397121416,
+      "loss": 1.1568,
+      "step": 363
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.16522270441055298,
+      "learning_rate": 0.00019951999808260376,
+      "loss": 1.1968,
+      "step": 364
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.15423065423965454,
+      "learning_rate": 0.00019951728456399327,
+      "loss": 1.1335,
+      "step": 365
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.16434919834136963,
+      "learning_rate": 0.0001995145634155907,
+      "loss": 1.1583,
+      "step": 366
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.15384919941425323,
+      "learning_rate": 0.00019951183463760475,
+      "loss": 1.1789,
+      "step": 367
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.15325228869915009,
+      "learning_rate": 0.00019950909823024455,
+      "loss": 1.1198,
+      "step": 368
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.15763159096240997,
+      "learning_rate": 0.00019950635419371996,
+      "loss": 1.184,
+      "step": 369
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.17088153958320618,
+      "learning_rate": 0.0001995036025282413,
+      "loss": 1.1653,
+      "step": 370
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.15483850240707397,
+      "learning_rate": 0.00019950084323401958,
+      "loss": 1.0983,
+      "step": 371
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.164934441447258,
+      "learning_rate": 0.00019949807631126633,
+      "loss": 1.0897,
+      "step": 372
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.16767363250255585,
+      "learning_rate": 0.00019949530176019368,
+      "loss": 1.1031,
+      "step": 373
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.16566498577594757,
+      "learning_rate": 0.00019949251958101437,
+      "loss": 1.176,
+      "step": 374
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.15819025039672852,
+      "learning_rate": 0.0001994897297739417,
+      "loss": 1.2291,
+      "step": 375
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.15242032706737518,
+      "learning_rate": 0.00019948693233918952,
+      "loss": 1.1527,
+      "step": 376
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.15688876807689667,
+      "learning_rate": 0.00019948412727697234,
+      "loss": 1.1566,
+      "step": 377
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.15263889729976654,
+      "learning_rate": 0.00019948131458750524,
+      "loss": 1.1477,
+      "step": 378
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1564534604549408,
+      "learning_rate": 0.00019947849427100382,
+      "loss": 1.1185,
+      "step": 379
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1585642248392105,
+      "learning_rate": 0.0001994756663276843,
+      "loss": 1.1919,
+      "step": 380
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.15505757927894592,
+      "learning_rate": 0.00019947283075776352,
+      "loss": 1.1588,
+      "step": 381
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.14966590702533722,
+      "learning_rate": 0.0001994699875614589,
+      "loss": 1.0691,
+      "step": 382
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.16230319440364838,
+      "learning_rate": 0.0001994671367389884,
+      "loss": 1.0874,
+      "step": 383
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1507527083158493,
+      "learning_rate": 0.00019946427829057057,
+      "loss": 1.1406,
+      "step": 384
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.15179996192455292,
+      "learning_rate": 0.00019946141221642458,
+      "loss": 1.0817,
+      "step": 385
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.15593743324279785,
+      "learning_rate": 0.00019945853851677014,
+      "loss": 1.0832,
+      "step": 386
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.15427644550800323,
+      "learning_rate": 0.00019945565719182762,
+      "loss": 1.0984,
+      "step": 387
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.14921724796295166,
+      "learning_rate": 0.00019945276824181787,
+      "loss": 1.1237,
+      "step": 388
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.15762227773666382,
+      "learning_rate": 0.00019944987166696244,
+      "loss": 1.1492,
+      "step": 389
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.17605511844158173,
+      "learning_rate": 0.00019944696746748334,
+      "loss": 1.077,
+      "step": 390
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.15459680557250977,
+      "learning_rate": 0.00019944405564360326,
+      "loss": 1.0774,
+      "step": 391
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.15118005871772766,
+      "learning_rate": 0.00019944113619554545,
+      "loss": 1.0614,
+      "step": 392
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.15591846406459808,
+      "learning_rate": 0.00019943820912353375,
+      "loss": 1.2164,
+      "step": 393
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1590687781572342,
+      "learning_rate": 0.00019943527442779256,
+      "loss": 1.1047,
+      "step": 394
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.15030692517757416,
+      "learning_rate": 0.00019943233210854686,
+      "loss": 1.0685,
+      "step": 395
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.16304552555084229,
+      "learning_rate": 0.00019942938216602223,
+      "loss": 1.1315,
+      "step": 396
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.16410882771015167,
+      "learning_rate": 0.00019942642460044486,
+      "loss": 1.0894,
+      "step": 397
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.15879030525684357,
+      "learning_rate": 0.00019942345941204148,
+      "loss": 1.1264,
+      "step": 398
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.17829300463199615,
+      "learning_rate": 0.00019942048660103948,
+      "loss": 1.1409,
+      "step": 399
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1717354953289032,
+      "learning_rate": 0.0001994175061676667,
+      "loss": 1.1616,
+      "step": 400
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.16701237857341766,
+      "learning_rate": 0.00019941451811215168,
+      "loss": 1.1964,
+      "step": 401
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.16376936435699463,
+      "learning_rate": 0.00019941152243472353,
+      "loss": 1.1349,
+      "step": 402
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1571936160326004,
+      "learning_rate": 0.00019940851913561187,
+      "loss": 1.1143,
+      "step": 403
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1656806915998459,
+      "learning_rate": 0.000199405508215047,
+      "loss": 1.1323,
+      "step": 404
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.16776010394096375,
+      "learning_rate": 0.00019940248967325976,
+      "loss": 1.1367,
+      "step": 405
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.14835897088050842,
+      "learning_rate": 0.00019939946351048158,
+      "loss": 1.1074,
+      "step": 406
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1668928563594818,
+      "learning_rate": 0.00019939642972694442,
+      "loss": 1.076,
+      "step": 407
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.16557957231998444,
+      "learning_rate": 0.0001993933883228809,
+      "loss": 1.1482,
+      "step": 408
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.15333621203899384,
+      "learning_rate": 0.00019939033929852425,
+      "loss": 1.1367,
+      "step": 409
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.15507064759731293,
+      "learning_rate": 0.00019938728265410816,
+      "loss": 1.1554,
+      "step": 410
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.15727224946022034,
+      "learning_rate": 0.000199384218389867,
+      "loss": 1.1084,
+      "step": 411
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1558888554573059,
+      "learning_rate": 0.00019938114650603573,
+      "loss": 1.1928,
+      "step": 412
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1571364551782608,
+      "learning_rate": 0.00019937806700284986,
+      "loss": 1.0771,
+      "step": 413
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.16041377186775208,
+      "learning_rate": 0.00019937497988054546,
+      "loss": 1.14,
+      "step": 414
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.15569083392620087,
+      "learning_rate": 0.00019937188513935921,
+      "loss": 1.1003,
+      "step": 415
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.15835022926330566,
+      "learning_rate": 0.0001993687827795284,
+      "loss": 1.0935,
+      "step": 416
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.16400137543678284,
+      "learning_rate": 0.00019936567280129085,
+      "loss": 1.1469,
+      "step": 417
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.17146562039852142,
+      "learning_rate": 0.00019936255520488504,
+      "loss": 1.1114,
+      "step": 418
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.16136683523654938,
+      "learning_rate": 0.00019935942999055,
+      "loss": 1.1084,
+      "step": 419
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.15654154121875763,
+      "learning_rate": 0.00019935629715852525,
+      "loss": 1.0972,
+      "step": 420
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1646607220172882,
+      "learning_rate": 0.00019935315670905105,
+      "loss": 1.1281,
+      "step": 421
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1610719859600067,
+      "learning_rate": 0.0001993500086423682,
+      "loss": 1.1131,
+      "step": 422
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.16374924778938293,
+      "learning_rate": 0.00019934685295871796,
+      "loss": 1.1501,
+      "step": 423
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.14906828105449677,
+      "learning_rate": 0.00019934368965834236,
+      "loss": 1.0911,
+      "step": 424
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.15239720046520233,
+      "learning_rate": 0.00019934051874148385,
+      "loss": 1.1018,
+      "step": 425
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.16222716867923737,
+      "learning_rate": 0.00019933734020838558,
+      "loss": 1.1582,
+      "step": 426
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.16466976702213287,
+      "learning_rate": 0.00019933415405929124,
+      "loss": 1.1741,
+      "step": 427
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.16280680894851685,
+      "learning_rate": 0.00019933096029444512,
+      "loss": 1.1858,
+      "step": 428
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1722201555967331,
+      "learning_rate": 0.00019932775891409204,
+      "loss": 1.1564,
+      "step": 429
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1672065407037735,
+      "learning_rate": 0.00019932454991847745,
+      "loss": 1.0803,
+      "step": 430
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.15295496582984924,
+      "learning_rate": 0.0001993213333078474,
+      "loss": 1.0711,
+      "step": 431
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1692061722278595,
+      "learning_rate": 0.0001993181090824485,
+      "loss": 1.1564,
+      "step": 432
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.17974984645843506,
+      "learning_rate": 0.00019931487724252796,
+      "loss": 1.1416,
+      "step": 433
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2724705934524536,
+      "learning_rate": 0.00019931163778833352,
+      "loss": 1.1635,
+      "step": 434
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.16062946617603302,
+      "learning_rate": 0.00019930839072011353,
+      "loss": 1.2007,
+      "step": 435
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1821368783712387,
+      "learning_rate": 0.00019930513603811699,
+      "loss": 1.0672,
+      "step": 436
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.17222100496292114,
+      "learning_rate": 0.00019930187374259337,
+      "loss": 1.1162,
+      "step": 437
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.16529572010040283,
+      "learning_rate": 0.00019929860383379285,
+      "loss": 1.1139,
+      "step": 438
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.18187189102172852,
+      "learning_rate": 0.00019929532631196608,
+      "loss": 1.1688,
+      "step": 439
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.18088270723819733,
+      "learning_rate": 0.00019929204117736436,
+      "loss": 1.1112,
+      "step": 440
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1561852991580963,
+      "learning_rate": 0.00019928874843023957,
+      "loss": 1.1113,
+      "step": 441
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1627861112356186,
+      "learning_rate": 0.0001992854480708441,
+      "loss": 1.0826,
+      "step": 442
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.16556496918201447,
+      "learning_rate": 0.00019928214009943106,
+      "loss": 1.1006,
+      "step": 443
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.18377332389354706,
+      "learning_rate": 0.00019927882451625402,
+      "loss": 1.1345,
+      "step": 444
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1585996001958847,
+      "learning_rate": 0.00019927550132156716,
+      "loss": 1.0774,
+      "step": 445
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1622677743434906,
+      "learning_rate": 0.00019927217051562531,
+      "loss": 1.0757,
+      "step": 446
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1750255525112152,
+      "learning_rate": 0.00019926883209868382,
+      "loss": 1.1532,
+      "step": 447
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1599978357553482,
+      "learning_rate": 0.00019926548607099865,
+      "loss": 1.0104,
+      "step": 448
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.15971815586090088,
+      "learning_rate": 0.00019926213243282628,
+      "loss": 1.0962,
+      "step": 449
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.16724000871181488,
+      "learning_rate": 0.00019925877118442386,
+      "loss": 1.1142,
+      "step": 450
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.17150260508060455,
+      "learning_rate": 0.00019925540232604915,
+      "loss": 1.1623,
+      "step": 451
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1774536669254303,
+      "learning_rate": 0.00019925202585796033,
+      "loss": 1.1171,
+      "step": 452
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1575271040201187,
+      "learning_rate": 0.00019924864178041634,
+      "loss": 1.095,
+      "step": 453
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3351128101348877,
+      "learning_rate": 0.00019924525009367658,
+      "loss": 1.2996,
+      "step": 454
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1803962141275406,
+      "learning_rate": 0.00019924185079800113,
+      "loss": 1.084,
+      "step": 455
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.17613469064235687,
+      "learning_rate": 0.0001992384438936506,
+      "loss": 1.1243,
+      "step": 456
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.17339295148849487,
+      "learning_rate": 0.00019923502938088615,
+      "loss": 1.0991,
+      "step": 457
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.16048653423786163,
+      "learning_rate": 0.0001992316072599696,
+      "loss": 1.0882,
+      "step": 458
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.15557259321212769,
+      "learning_rate": 0.0001992281775311633,
+      "loss": 1.0639,
+      "step": 459
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.17755229771137238,
+      "learning_rate": 0.00019922474019473023,
+      "loss": 1.1323,
+      "step": 460
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.16753968596458435,
+      "learning_rate": 0.00019922129525093389,
+      "loss": 1.0643,
+      "step": 461
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.15593862533569336,
+      "learning_rate": 0.00019921784270003843,
+      "loss": 1.1237,
+      "step": 462
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1679830104112625,
+      "learning_rate": 0.0001992143825423085,
+      "loss": 1.1163,
+      "step": 463
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.17055770754814148,
+      "learning_rate": 0.00019921091477800946,
+      "loss": 1.1181,
+      "step": 464
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1715300977230072,
+      "learning_rate": 0.0001992074394074071,
+      "loss": 1.1767,
+      "step": 465
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.15965238213539124,
+      "learning_rate": 0.0001992039564307679,
+      "loss": 1.1201,
+      "step": 466
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1599990427494049,
+      "learning_rate": 0.00019920046584835887,
+      "loss": 1.1146,
+      "step": 467
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.16567906737327576,
+      "learning_rate": 0.00019919696766044768,
+      "loss": 1.1322,
+      "step": 468
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.15802539885044098,
+      "learning_rate": 0.0001991934618673025,
+      "loss": 1.0508,
+      "step": 469
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.161991149187088,
+      "learning_rate": 0.0001991899484691921,
+      "loss": 1.1892,
+      "step": 470
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.16740918159484863,
+      "learning_rate": 0.00019918642746638584,
+      "loss": 1.1625,
+      "step": 471
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.15869614481925964,
+      "learning_rate": 0.0001991828988591537,
+      "loss": 1.126,
+      "step": 472
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.17097507417201996,
+      "learning_rate": 0.0001991793626477662,
+      "loss": 1.0856,
+      "step": 473
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.15990661084651947,
+      "learning_rate": 0.00019917581883249446,
+      "loss": 1.0745,
+      "step": 474
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1636432260274887,
+      "learning_rate": 0.00019917226741361015,
+      "loss": 1.192,
+      "step": 475
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.16329345107078552,
+      "learning_rate": 0.00019916870839138556,
+      "loss": 1.0846,
+      "step": 476
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.16847611963748932,
+      "learning_rate": 0.00019916514176609358,
+      "loss": 1.1549,
+      "step": 477
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.16813425719738007,
+      "learning_rate": 0.00019916156753800764,
+      "loss": 1.0764,
+      "step": 478
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1625032275915146,
+      "learning_rate": 0.00019915798570740173,
+      "loss": 1.1547,
+      "step": 479
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.170704185962677,
+      "learning_rate": 0.00019915439627455052,
+      "loss": 1.0889,
+      "step": 480
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.17293334007263184,
+      "learning_rate": 0.00019915079923972918,
+      "loss": 1.202,
+      "step": 481
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.15445169806480408,
+      "learning_rate": 0.0001991471946032135,
+      "loss": 1.0738,
+      "step": 482
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.16902180016040802,
+      "learning_rate": 0.00019914358236527982,
+      "loss": 1.0554,
+      "step": 483
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.15919239819049835,
+      "learning_rate": 0.0001991399625262051,
+      "loss": 1.1367,
+      "step": 484
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.16527967154979706,
+      "learning_rate": 0.00019913633508626685,
+      "loss": 1.1041,
+      "step": 485
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.16559232771396637,
+      "learning_rate": 0.0001991327000457432,
+      "loss": 1.0582,
+      "step": 486
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.16566972434520721,
+      "learning_rate": 0.0001991290574049128,
+      "loss": 1.0881,
+      "step": 487
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.15967406332492828,
+      "learning_rate": 0.00019912540716405497,
+      "loss": 1.0668,
+      "step": 488
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.16620907187461853,
+      "learning_rate": 0.00019912174932344953,
+      "loss": 1.0945,
+      "step": 489
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.16342945396900177,
+      "learning_rate": 0.00019911808388337697,
+      "loss": 1.0558,
+      "step": 490
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.14973656833171844,
+      "learning_rate": 0.00019911441084411827,
+      "loss": 1.042,
+      "step": 491
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1684689223766327,
+      "learning_rate": 0.00019911073020595504,
+      "loss": 1.1281,
+      "step": 492
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.15556031465530396,
+      "learning_rate": 0.00019910704196916948,
+      "loss": 1.0338,
+      "step": 493
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.16132386028766632,
+      "learning_rate": 0.00019910334613404434,
+      "loss": 1.0586,
+      "step": 494
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.16225095093250275,
+      "learning_rate": 0.000199099642700863,
+      "loss": 1.1608,
+      "step": 495
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.17955203354358673,
+      "learning_rate": 0.00019909593166990934,
+      "loss": 1.0858,
+      "step": 496
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.153148353099823,
+      "learning_rate": 0.00019909221304146795,
+      "loss": 1.1217,
+      "step": 497
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.16156357526779175,
+      "learning_rate": 0.00019908848681582391,
+      "loss": 1.08,
+      "step": 498
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.167068749666214,
+      "learning_rate": 0.00019908475299326286,
+      "loss": 1.1021,
+      "step": 499
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.17191778123378754,
+      "learning_rate": 0.00019908101157407112,
+      "loss": 1.085,
+      "step": 500
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.17865735292434692,
+      "learning_rate": 0.00019907726255853547,
+      "loss": 1.118,
+      "step": 501
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1564328372478485,
+      "learning_rate": 0.0001990735059469434,
+      "loss": 1.0634,
+      "step": 502
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1716470867395401,
+      "learning_rate": 0.0001990697417395829,
+      "loss": 1.0566,
+      "step": 503
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1743009388446808,
+      "learning_rate": 0.00019906596993674258,
+      "loss": 1.0836,
+      "step": 504
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1646343320608139,
+      "learning_rate": 0.0001990621905387116,
+      "loss": 1.0982,
+      "step": 505
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.18098647892475128,
+      "learning_rate": 0.00019905840354577972,
+      "loss": 1.1517,
+      "step": 506
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.16818532347679138,
+      "learning_rate": 0.00019905460895823727,
+      "loss": 1.13,
+      "step": 507
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.16616329550743103,
+      "learning_rate": 0.00019905080677637518,
+      "loss": 1.1478,
+      "step": 508
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.16015742719173431,
+      "learning_rate": 0.000199046997000485,
+      "loss": 1.0683,
+      "step": 509
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.16552738845348358,
+      "learning_rate": 0.00019904317963085876,
+      "loss": 1.0965,
+      "step": 510
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.15932437777519226,
+      "learning_rate": 0.00019903935466778915,
+      "loss": 1.0809,
+      "step": 511
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.157403826713562,
+      "learning_rate": 0.00019903552211156939,
+      "loss": 1.073,
+      "step": 512
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1659756451845169,
+      "learning_rate": 0.0001990316819624934,
+      "loss": 1.111,
+      "step": 513
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1708558350801468,
+      "learning_rate": 0.0001990278342208555,
+      "loss": 1.0917,
+      "step": 514
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1662391573190689,
+      "learning_rate": 0.00019902397888695077,
+      "loss": 1.1407,
+      "step": 515
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.16887429356575012,
+      "learning_rate": 0.00019902011596107473,
+      "loss": 1.025,
+      "step": 516
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.17284426093101501,
+      "learning_rate": 0.00019901624544352357,
+      "loss": 1.0788,
+      "step": 517
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1698935627937317,
+      "learning_rate": 0.00019901236733459404,
+      "loss": 1.1118,
+      "step": 518
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.15765266120433807,
+      "learning_rate": 0.00019900848163458346,
+      "loss": 1.0623,
+      "step": 519
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1714356690645218,
+      "learning_rate": 0.00019900458834378972,
+      "loss": 1.1496,
+      "step": 520
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.16561633348464966,
+      "learning_rate": 0.00019900068746251135,
+      "loss": 1.0381,
+      "step": 521
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.16743022203445435,
+      "learning_rate": 0.00019899677899104737,
+      "loss": 1.0783,
+      "step": 522
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.16489222645759583,
+      "learning_rate": 0.00019899286292969747,
+      "loss": 1.1166,
+      "step": 523
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1696144938468933,
+      "learning_rate": 0.0001989889392787619,
+      "loss": 1.1433,
+      "step": 524
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.16742746531963348,
+      "learning_rate": 0.00019898500803854145,
+      "loss": 1.0809,
+      "step": 525
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1641719490289688,
+      "learning_rate": 0.00019898106920933755,
+      "loss": 1.1071,
+      "step": 526
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.17350101470947266,
+      "learning_rate": 0.00019897712279145214,
+      "loss": 1.0923,
+      "step": 527
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.17832672595977783,
+      "learning_rate": 0.0001989731687851878,
+      "loss": 1.1332,
+      "step": 528
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.15399731695652008,
+      "learning_rate": 0.0001989692071908477,
+      "loss": 1.0644,
+      "step": 529
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.16712157428264618,
+      "learning_rate": 0.00019896523800873555,
+      "loss": 1.1236,
+      "step": 530
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1874651312828064,
+      "learning_rate": 0.00019896126123915563,
+      "loss": 1.1011,
+      "step": 531
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.17110246419906616,
+      "learning_rate": 0.00019895727688241287,
+      "loss": 1.0806,
+      "step": 532
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.17345298826694489,
+      "learning_rate": 0.00019895328493881276,
+      "loss": 1.1803,
+      "step": 533
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1709088385105133,
+      "learning_rate": 0.0001989492854086613,
+      "loss": 1.1504,
+      "step": 534
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.16704735159873962,
+      "learning_rate": 0.00019894527829226517,
+      "loss": 1.1459,
+      "step": 535
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1761869490146637,
+      "learning_rate": 0.00019894126358993157,
+      "loss": 1.0465,
+      "step": 536
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.16007749736309052,
+      "learning_rate": 0.00019893724130196828,
+      "loss": 1.0763,
+      "step": 537
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1607150137424469,
+      "learning_rate": 0.00019893321142868377,
+      "loss": 1.0575,
+      "step": 538
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.17952270805835724,
+      "learning_rate": 0.00019892917397038685,
+      "loss": 1.1514,
+      "step": 539
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.165918231010437,
+      "learning_rate": 0.00019892512892738718,
+      "loss": 1.1026,
+      "step": 540
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.17264124751091003,
+      "learning_rate": 0.00019892107629999485,
+      "loss": 1.0882,
+      "step": 541
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4207189977169037,
+      "learning_rate": 0.00019891701608852058,
+      "loss": 1.2084,
+      "step": 542
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.22649544477462769,
+      "learning_rate": 0.00019891294829327562,
+      "loss": 1.0122,
+      "step": 543
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.18632979691028595,
+      "learning_rate": 0.0001989088729145719,
+      "loss": 1.0788,
+      "step": 544
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.16893143951892853,
+      "learning_rate": 0.00019890478995272183,
+      "loss": 1.1284,
+      "step": 545
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.33213508129119873,
+      "learning_rate": 0.00019890069940803847,
+      "loss": 1.0338,
+      "step": 546
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.18078738451004028,
+      "learning_rate": 0.00019889660128083537,
+      "loss": 1.1017,
+      "step": 547
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1705341935157776,
+      "learning_rate": 0.00019889249557142677,
+      "loss": 1.1644,
+      "step": 548
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1637420952320099,
+      "learning_rate": 0.00019888838228012746,
+      "loss": 1.1165,
+      "step": 549
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.18197381496429443,
+      "learning_rate": 0.00019888426140725283,
+      "loss": 1.1492,
+      "step": 550
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1665923148393631,
+      "learning_rate": 0.00019888013295311875,
+      "loss": 1.0797,
+      "step": 551
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.17134587466716766,
+      "learning_rate": 0.00019887599691804174,
+      "loss": 1.1146,
+      "step": 552
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.16827426850795746,
+      "learning_rate": 0.00019887185330233893,
+      "loss": 1.0843,
+      "step": 553
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1690630316734314,
+      "learning_rate": 0.00019886770210632802,
+      "loss": 1.1285,
+      "step": 554
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1599999964237213,
+      "learning_rate": 0.00019886354333032723,
+      "loss": 1.0514,
+      "step": 555
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.16263416409492493,
+      "learning_rate": 0.00019885937697465545,
+      "loss": 1.0951,
+      "step": 556
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1642608791589737,
+      "learning_rate": 0.00019885520303963204,
+      "loss": 1.0927,
+      "step": 557
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.16561810672283173,
+      "learning_rate": 0.00019885102152557708,
+      "loss": 1.1361,
+      "step": 558
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.16234217584133148,
+      "learning_rate": 0.00019884683243281116,
+      "loss": 1.0727,
+      "step": 559
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.15540958940982819,
+      "learning_rate": 0.0001988426357616554,
+      "loss": 0.9542,
+      "step": 560
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.16096945106983185,
+      "learning_rate": 0.00019883843151243156,
+      "loss": 1.0627,
+      "step": 561
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1687619537115097,
+      "learning_rate": 0.00019883421968546196,
+      "loss": 1.1358,
+      "step": 562
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.16691409051418304,
+      "learning_rate": 0.00019883000028106954,
+      "loss": 1.0642,
+      "step": 563
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.16787758469581604,
+      "learning_rate": 0.00019882577329957782,
+      "loss": 1.1038,
+      "step": 564
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.16505680978298187,
+      "learning_rate": 0.00019882153874131083,
+      "loss": 1.0917,
+      "step": 565
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1684773862361908,
+      "learning_rate": 0.00019881729660659324,
+      "loss": 1.1094,
+      "step": 566
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.16181319952011108,
+      "learning_rate": 0.00019881304689575025,
+      "loss": 1.1195,
+      "step": 567
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.16680899262428284,
+      "learning_rate": 0.00019880878960910772,
+      "loss": 1.0675,
+      "step": 568
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.16721315681934357,
+      "learning_rate": 0.00019880452474699205,
+      "loss": 1.0641,
+      "step": 569
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.16467982530593872,
+      "learning_rate": 0.0001988002523097302,
+      "loss": 1.068,
+      "step": 570
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.16453714668750763,
+      "learning_rate": 0.00019879597229764974,
+      "loss": 1.1099,
+      "step": 571
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.17011506855487823,
+      "learning_rate": 0.0001987916847110788,
+      "loss": 1.0738,
+      "step": 572
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.16118289530277252,
+      "learning_rate": 0.00019878738955034607,
+      "loss": 1.1012,
+      "step": 573
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.16739653050899506,
+      "learning_rate": 0.00019878308681578095,
+      "loss": 1.0256,
+      "step": 574
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.16780000925064087,
+      "learning_rate": 0.00019877877650771323,
+      "loss": 1.1269,
+      "step": 575
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.16901536285877228,
+      "learning_rate": 0.0001987744586264734,
+      "loss": 1.0934,
+      "step": 576
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1620328426361084,
+      "learning_rate": 0.0001987701331723925,
+      "loss": 1.1137,
+      "step": 577
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1608792394399643,
+      "learning_rate": 0.00019876580014580215,
+      "loss": 1.0165,
+      "step": 578
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.17747630178928375,
+      "learning_rate": 0.00019876145954703458,
+      "loss": 1.0936,
+      "step": 579
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.16932570934295654,
+      "learning_rate": 0.00019875711137642258,
+      "loss": 1.1101,
+      "step": 580
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.15958163142204285,
+      "learning_rate": 0.00019875275563429945,
+      "loss": 1.0829,
+      "step": 581
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.16792955994606018,
+      "learning_rate": 0.00019874839232099918,
+      "loss": 1.1083,
+      "step": 582
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1686125099658966,
+      "learning_rate": 0.00019874402143685633,
+      "loss": 1.0504,
+      "step": 583
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.158395916223526,
+      "learning_rate": 0.00019873964298220597,
+      "loss": 1.0783,
+      "step": 584
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.15973375737667084,
+      "learning_rate": 0.00019873525695738374,
+      "loss": 1.093,
+      "step": 585
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.16699771583080292,
+      "learning_rate": 0.00019873086336272603,
+      "loss": 1.0864,
+      "step": 586
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.16269387304782867,
+      "learning_rate": 0.00019872646219856957,
+      "loss": 1.0638,
+      "step": 587
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.16882598400115967,
+      "learning_rate": 0.00019872205346525186,
+      "loss": 1.1471,
+      "step": 588
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.16338156163692474,
+      "learning_rate": 0.00019871763716311085,
+      "loss": 1.0887,
+      "step": 589
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1657753735780716,
+      "learning_rate": 0.0001987132132924852,
+      "loss": 1.1273,
+      "step": 590
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1670033186674118,
+      "learning_rate": 0.000198708781853714,
+      "loss": 1.1171,
+      "step": 591
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.16076108813285828,
+      "learning_rate": 0.0001987043428471371,
+      "loss": 1.0601,
+      "step": 592
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1607861965894699,
+      "learning_rate": 0.00019869989627309474,
+      "loss": 1.0151,
+      "step": 593
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.15327143669128418,
+      "learning_rate": 0.00019869544213192787,
+      "loss": 1.0779,
+      "step": 594
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.16481177508831024,
+      "learning_rate": 0.00019869098042397796,
+      "loss": 1.0822,
+      "step": 595
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.17603227496147156,
+      "learning_rate": 0.00019868651114958707,
+      "loss": 1.0934,
+      "step": 596
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.15565873682498932,
+      "learning_rate": 0.0001986820343090979,
+      "loss": 0.9236,
+      "step": 597
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.15814374387264252,
+      "learning_rate": 0.00019867754990285366,
+      "loss": 1.0454,
+      "step": 598
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.17093122005462646,
+      "learning_rate": 0.00019867305793119816,
+      "loss": 1.0845,
+      "step": 599
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1707032173871994,
+      "learning_rate": 0.00019866855839447575,
+      "loss": 1.0129,
+      "step": 600
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1676967591047287,
+      "learning_rate": 0.00019866405129303146,
+      "loss": 1.0709,
+      "step": 601
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.16182450950145721,
+      "learning_rate": 0.00019865953662721083,
+      "loss": 1.1096,
+      "step": 602
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.16513295471668243,
+      "learning_rate": 0.00019865501439735993,
+      "loss": 1.1376,
+      "step": 603
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1725083738565445,
+      "learning_rate": 0.00019865048460382555,
+      "loss": 1.1235,
+      "step": 604
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.174422025680542,
+      "learning_rate": 0.00019864594724695493,
+      "loss": 1.1195,
+      "step": 605
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.16189318895339966,
+      "learning_rate": 0.00019864140232709598,
+      "loss": 1.0649,
+      "step": 606
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.176430344581604,
+      "learning_rate": 0.0001986368498445971,
+      "loss": 1.1182,
+      "step": 607
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.16489064693450928,
+      "learning_rate": 0.00019863228979980735,
+      "loss": 1.0936,
+      "step": 608
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.16108497977256775,
+      "learning_rate": 0.00019862772219307636,
+      "loss": 1.1095,
+      "step": 609
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.163023442029953,
+      "learning_rate": 0.00019862314702475425,
+      "loss": 1.0862,
+      "step": 610
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.16625739634037018,
+      "learning_rate": 0.0001986185642951919,
+      "loss": 1.0695,
+      "step": 611
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.18615350127220154,
+      "learning_rate": 0.0001986139740047405,
+      "loss": 1.1538,
+      "step": 612
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.17098894715309143,
+      "learning_rate": 0.00019860937615375214,
+      "loss": 1.0959,
+      "step": 613
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1671309918165207,
+      "learning_rate": 0.00019860477074257923,
+      "loss": 1.0731,
+      "step": 614
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.18721477687358856,
+      "learning_rate": 0.00019860015777157491,
+      "loss": 1.1827,
+      "step": 615
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.16247648000717163,
+      "learning_rate": 0.00019859553724109282,
+      "loss": 1.0652,
+      "step": 616
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.16883952915668488,
+      "learning_rate": 0.0001985909091514872,
+      "loss": 1.0913,
+      "step": 617
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.16720356047153473,
+      "learning_rate": 0.0001985862735031129,
+      "loss": 1.0836,
+      "step": 618
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.16027386486530304,
+      "learning_rate": 0.00019858163029632528,
+      "loss": 1.0332,
+      "step": 619
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.15760108828544617,
+      "learning_rate": 0.00019857697953148037,
+      "loss": 1.0635,
+      "step": 620
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.16750258207321167,
+      "learning_rate": 0.00019857232120893477,
+      "loss": 1.1103,
+      "step": 621
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.16163256764411926,
+      "learning_rate": 0.0001985676553290455,
+      "loss": 1.0626,
+      "step": 622
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.16259679198265076,
+      "learning_rate": 0.0001985629818921704,
+      "loss": 1.0135,
+      "step": 623
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.16069941222667694,
+      "learning_rate": 0.0001985583008986678,
+      "loss": 1.0288,
+      "step": 624
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.16990730166435242,
+      "learning_rate": 0.0001985536123488964,
+      "loss": 1.14,
+      "step": 625
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1745760291814804,
+      "learning_rate": 0.00019854891624321587,
+      "loss": 1.0704,
+      "step": 626
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.16768278181552887,
+      "learning_rate": 0.00019854421258198613,
+      "loss": 1.0832,
+      "step": 627
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.16163139045238495,
+      "learning_rate": 0.00019853950136556781,
+      "loss": 1.0877,
+      "step": 628
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.16793616116046906,
+      "learning_rate": 0.00019853478259432213,
+      "loss": 1.1423,
+      "step": 629
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1579504758119583,
+      "learning_rate": 0.0001985300562686109,
+      "loss": 1.0608,
+      "step": 630
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.17058435082435608,
+      "learning_rate": 0.00019852532238879645,
+      "loss": 1.0879,
+      "step": 631
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.16249874234199524,
+      "learning_rate": 0.0001985205809552417,
+      "loss": 1.0558,
+      "step": 632
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.160230353474617,
+      "learning_rate": 0.00019851583196831015,
+      "loss": 0.9504,
+      "step": 633
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.163030743598938,
+      "learning_rate": 0.00019851107542836598,
+      "loss": 1.0638,
+      "step": 634
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.16014380753040314,
+      "learning_rate": 0.00019850631133577377,
+      "loss": 1.0828,
+      "step": 635
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1663341075181961,
+      "learning_rate": 0.0001985015396908988,
+      "loss": 1.0808,
+      "step": 636
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.16736695170402527,
+      "learning_rate": 0.00019849676049410694,
+      "loss": 1.0942,
+      "step": 637
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.16381196677684784,
+      "learning_rate": 0.00019849197374576457,
+      "loss": 1.1072,
+      "step": 638
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.16722634434700012,
+      "learning_rate": 0.00019848717944623872,
+      "loss": 1.0613,
+      "step": 639
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.15986140072345734,
+      "learning_rate": 0.00019848237759589692,
+      "loss": 1.0266,
+      "step": 640
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.17291495203971863,
+      "learning_rate": 0.0001984775681951073,
+      "loss": 1.0988,
+      "step": 641
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.15927840769290924,
+      "learning_rate": 0.00019847275124423863,
+      "loss": 1.0066,
+      "step": 642
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.16089984774589539,
+      "learning_rate": 0.00019846792674366016,
+      "loss": 1.0368,
+      "step": 643
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.16961584985256195,
+      "learning_rate": 0.00019846309469374185,
+      "loss": 1.1053,
+      "step": 644
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.16907954216003418,
+      "learning_rate": 0.0001984582550948541,
+      "loss": 1.1217,
+      "step": 645
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1658971607685089,
+      "learning_rate": 0.00019845340794736802,
+      "loss": 1.0813,
+      "step": 646
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.18253681063652039,
+      "learning_rate": 0.00019844855325165518,
+      "loss": 1.0798,
+      "step": 647
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.16371911764144897,
+      "learning_rate": 0.0001984436910080878,
+      "loss": 1.0753,
+      "step": 648
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1763499230146408,
+      "learning_rate": 0.00019843882121703863,
+      "loss": 1.1702,
+      "step": 649
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.17335723340511322,
+      "learning_rate": 0.00019843394387888104,
+      "loss": 1.1431,
+      "step": 650
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.16688063740730286,
+      "learning_rate": 0.00019842905899398897,
+      "loss": 1.0716,
+      "step": 651
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.16677774488925934,
+      "learning_rate": 0.00019842416656273696,
+      "loss": 1.0497,
+      "step": 652
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.16049520671367645,
+      "learning_rate": 0.00019841926658550007,
+      "loss": 1.0556,
+      "step": 653
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.16821910440921783,
+      "learning_rate": 0.000198414359062654,
+      "loss": 1.0611,
+      "step": 654
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.16483479738235474,
+      "learning_rate": 0.00019840944399457492,
+      "loss": 1.0531,
+      "step": 655
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.17143693566322327,
+      "learning_rate": 0.00019840452138163977,
+      "loss": 1.0541,
+      "step": 656
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.17825862765312195,
+      "learning_rate": 0.0001983995912242259,
+      "loss": 1.0862,
+      "step": 657
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.16992264986038208,
+      "learning_rate": 0.0001983946535227113,
+      "loss": 1.1236,
+      "step": 658
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.17565155029296875,
+      "learning_rate": 0.00019838970827747453,
+      "loss": 1.023,
+      "step": 659
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.16197721660137177,
+      "learning_rate": 0.00019838475548889471,
+      "loss": 1.0602,
+      "step": 660
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.157365620136261,
+      "learning_rate": 0.00019837979515735166,
+      "loss": 1.0197,
+      "step": 661
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.16862985491752625,
+      "learning_rate": 0.00019837482728322553,
+      "loss": 1.1263,
+      "step": 662
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.16315187513828278,
+      "learning_rate": 0.0001983698518668973,
+      "loss": 1.078,
+      "step": 663
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.16445574164390564,
+      "learning_rate": 0.00019836486890874845,
+      "loss": 1.1159,
+      "step": 664
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1527933031320572,
+      "learning_rate": 0.00019835987840916092,
+      "loss": 1.0328,
+      "step": 665
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.16762986779212952,
+      "learning_rate": 0.00019835488036851735,
+      "loss": 1.0463,
+      "step": 666
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.17638851702213287,
+      "learning_rate": 0.00019834987478720096,
+      "loss": 1.0314,
+      "step": 667
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1704005002975464,
+      "learning_rate": 0.0001983448616655955,
+      "loss": 1.1023,
+      "step": 668
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1858178824186325,
+      "learning_rate": 0.00019833984100408531,
+      "loss": 1.0918,
+      "step": 669
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.22606873512268066,
+      "learning_rate": 0.00019833481280305535,
+      "loss": 0.9917,
+      "step": 670
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.16366678476333618,
+      "learning_rate": 0.00019832977706289108,
+      "loss": 1.0597,
+      "step": 671
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1654898226261139,
+      "learning_rate": 0.0001983247337839786,
+      "loss": 1.062,
+      "step": 672
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.18031100928783417,
+      "learning_rate": 0.00019831968296670454,
+      "loss": 1.111,
+      "step": 673
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.17027471959590912,
+      "learning_rate": 0.00019831462461145617,
+      "loss": 1.1478,
+      "step": 674
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.16689349710941315,
+      "learning_rate": 0.0001983095587186213,
+      "loss": 1.0514,
+      "step": 675
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.17473068833351135,
+      "learning_rate": 0.00019830448528858832,
+      "loss": 1.1135,
+      "step": 676
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.15812206268310547,
+      "learning_rate": 0.00019829940432174617,
+      "loss": 1.06,
+      "step": 677
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.16691286861896515,
+      "learning_rate": 0.00019829431581848446,
+      "loss": 1.0982,
+      "step": 678
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1686815768480301,
+      "learning_rate": 0.00019828921977919323,
+      "loss": 1.0948,
+      "step": 679
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1836208552122116,
+      "learning_rate": 0.0001982841162042633,
+      "loss": 1.1331,
+      "step": 680
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1622176170349121,
+      "learning_rate": 0.00019827900509408581,
+      "loss": 1.0369,
+      "step": 681
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1718926578760147,
+      "learning_rate": 0.0001982738864490527,
+      "loss": 1.0552,
+      "step": 682
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1771085262298584,
+      "learning_rate": 0.00019826876026955644,
+      "loss": 1.0502,
+      "step": 683
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.16786454617977142,
+      "learning_rate": 0.00019826362655599,
+      "loss": 1.0486,
+      "step": 684
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.16586992144584656,
+      "learning_rate": 0.00019825848530874692,
+      "loss": 1.0809,
+      "step": 685
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1600944846868515,
+      "learning_rate": 0.00019825333652822146,
+      "loss": 1.0243,
+      "step": 686
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1946248859167099,
+      "learning_rate": 0.00019824818021480832,
+      "loss": 1.0905,
+      "step": 687
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.16281233727931976,
+      "learning_rate": 0.0001982430163689028,
+      "loss": 1.0344,
+      "step": 688
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.16939066350460052,
+      "learning_rate": 0.00019823784499090088,
+      "loss": 1.0619,
+      "step": 689
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.18849721550941467,
+      "learning_rate": 0.00019823266608119897,
+      "loss": 1.0294,
+      "step": 690
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.17455913126468658,
+      "learning_rate": 0.00019822747964019416,
+      "loss": 1.0463,
+      "step": 691
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.16168515384197235,
+      "learning_rate": 0.0001982222856682841,
+      "loss": 1.0988,
+      "step": 692
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.17095373570919037,
+      "learning_rate": 0.00019821708416586692,
+      "loss": 1.0377,
+      "step": 693
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.17848733067512512,
+      "learning_rate": 0.0001982118751333415,
+      "loss": 1.0138,
+      "step": 694
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1678147166967392,
+      "learning_rate": 0.00019820665857110716,
+      "loss": 1.0738,
+      "step": 695
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.16244111955165863,
+      "learning_rate": 0.00019820143447956388,
+      "loss": 1.0178,
+      "step": 696
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.17898660898208618,
+      "learning_rate": 0.00019819620285911212,
+      "loss": 1.0424,
+      "step": 697
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.18440993130207062,
+      "learning_rate": 0.00019819096371015302,
+      "loss": 1.087,
+      "step": 698
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1611272543668747,
+      "learning_rate": 0.00019818571703308826,
+      "loss": 1.0659,
+      "step": 699
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.17365390062332153,
+      "learning_rate": 0.00019818046282832005,
+      "loss": 1.0261,
+      "step": 700
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1745585799217224,
+      "learning_rate": 0.00019817520109625128,
+      "loss": 1.0753,
+      "step": 701
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.16264259815216064,
+      "learning_rate": 0.00019816993183728533,
+      "loss": 1.1131,
+      "step": 702
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.16973325610160828,
+      "learning_rate": 0.00019816465505182617,
+      "loss": 1.1309,
+      "step": 703
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.16402168571949005,
+      "learning_rate": 0.00019815937074027837,
+      "loss": 1.0483,
+      "step": 704
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.174351766705513,
+      "learning_rate": 0.00019815407890304707,
+      "loss": 1.1422,
+      "step": 705
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.16474570333957672,
+      "learning_rate": 0.00019814877954053798,
+      "loss": 1.0539,
+      "step": 706
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.16664201021194458,
+      "learning_rate": 0.00019814347265315738,
+      "loss": 1.0573,
+      "step": 707
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.17523398995399475,
+      "learning_rate": 0.00019813815824131216,
+      "loss": 1.1203,
+      "step": 708
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.16613712906837463,
+      "learning_rate": 0.0001981328363054098,
+      "loss": 1.0876,
+      "step": 709
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1733550727367401,
+      "learning_rate": 0.00019812750684585823,
+      "loss": 1.0538,
+      "step": 710
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 1.0733253955841064,
+      "eval_runtime": 679.1234,
+      "eval_samples_per_second": 10.125,
+      "eval_steps_per_second": 5.062,
+      "step": 710
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.16474869847297668,
+      "learning_rate": 0.00019812216986306612,
+      "loss": 1.1042,
+      "step": 711
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1642632931470871,
+      "learning_rate": 0.00019811682535744263,
+      "loss": 1.0718,
+      "step": 712
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.17970983684062958,
+      "learning_rate": 0.0001981114733293975,
+      "loss": 1.0629,
+      "step": 713
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.15759165585041046,
+      "learning_rate": 0.00019810611377934104,
+      "loss": 1.0588,
+      "step": 714
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.17333929240703583,
+      "learning_rate": 0.00019810074670768423,
+      "loss": 1.0582,
+      "step": 715
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1946728527545929,
+      "learning_rate": 0.00019809537211483847,
+      "loss": 1.1664,
+      "step": 716
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1639917492866516,
+      "learning_rate": 0.00019808999000121586,
+      "loss": 1.1074,
+      "step": 717
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.19192171096801758,
+      "learning_rate": 0.00019808460036722906,
+      "loss": 1.0592,
+      "step": 718
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1635810136795044,
+      "learning_rate": 0.00019807920321329123,
+      "loss": 1.038,
+      "step": 719
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.16381768882274628,
+      "learning_rate": 0.0001980737985398162,
+      "loss": 1.0321,
+      "step": 720
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1705102175474167,
+      "learning_rate": 0.00019806838634721832,
+      "loss": 1.1072,
+      "step": 721
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1895795315504074,
+      "learning_rate": 0.0001980629666359125,
+      "loss": 1.0437,
+      "step": 722
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1661144644021988,
+      "learning_rate": 0.0001980575394063143,
+      "loss": 1.0303,
+      "step": 723
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.18807817995548248,
+      "learning_rate": 0.0001980521046588398,
+      "loss": 1.1596,
+      "step": 724
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.16420172154903412,
+      "learning_rate": 0.00019804666239390568,
+      "loss": 0.9703,
+      "step": 725
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1850784718990326,
+      "learning_rate": 0.0001980412126119292,
+      "loss": 1.1502,
+      "step": 726
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.16995462775230408,
+      "learning_rate": 0.00019803575531332816,
+      "loss": 1.0608,
+      "step": 727
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.16244123876094818,
+      "learning_rate": 0.00019803029049852096,
+      "loss": 1.0032,
+      "step": 728
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.16180339455604553,
+      "learning_rate": 0.00019802481816792657,
+      "loss": 1.0867,
+      "step": 729
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.16730904579162598,
+      "learning_rate": 0.00019801933832196456,
+      "loss": 1.0569,
+      "step": 730
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.17155852913856506,
+      "learning_rate": 0.0001980138509610551,
+      "loss": 1.0967,
+      "step": 731
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.17094744741916656,
+      "learning_rate": 0.0001980083560856188,
+      "loss": 1.0308,
+      "step": 732
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.17406611144542694,
+      "learning_rate": 0.000198002853696077,
+      "loss": 1.0859,
+      "step": 733
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1751219928264618,
+      "learning_rate": 0.00019799734379285155,
+      "loss": 1.1167,
+      "step": 734
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.16861020028591156,
+      "learning_rate": 0.0001979918263763649,
+      "loss": 1.0223,
+      "step": 735
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.17369449138641357,
+      "learning_rate": 0.00019798630144704,
+      "loss": 1.0547,
+      "step": 736
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.17756544053554535,
+      "learning_rate": 0.00019798076900530052,
+      "loss": 1.105,
+      "step": 737
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.17094124853610992,
+      "learning_rate": 0.00019797522905157054,
+      "loss": 1.0508,
+      "step": 738
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.17031005024909973,
+      "learning_rate": 0.00019796968158627488,
+      "loss": 1.0735,
+      "step": 739
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.17707915604114532,
+      "learning_rate": 0.00019796412660983876,
+      "loss": 1.0652,
+      "step": 740
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1686302125453949,
+      "learning_rate": 0.00019795856412268813,
+      "loss": 1.0452,
+      "step": 741
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1696590632200241,
+      "learning_rate": 0.00019795299412524945,
+      "loss": 1.056,
+      "step": 742
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1714959293603897,
+      "learning_rate": 0.00019794741661794977,
+      "loss": 1.0193,
+      "step": 743
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.16320788860321045,
+      "learning_rate": 0.00019794183160121666,
+      "loss": 1.088,
+      "step": 744
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.15728294849395752,
+      "learning_rate": 0.00019793623907547834,
+      "loss": 1.0298,
+      "step": 745
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.17356523871421814,
+      "learning_rate": 0.0001979306390411636,
+      "loss": 1.1152,
+      "step": 746
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.17470648884773254,
+      "learning_rate": 0.00019792503149870173,
+      "loss": 1.0811,
+      "step": 747
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.17432372272014618,
+      "learning_rate": 0.00019791941644852273,
+      "loss": 1.1509,
+      "step": 748
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.16345620155334473,
+      "learning_rate": 0.000197913793891057,
+      "loss": 1.0062,
+      "step": 749
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.17330077290534973,
+      "learning_rate": 0.00019790816382673568,
+      "loss": 0.9442,
+      "step": 750
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1774456650018692,
+      "learning_rate": 0.0001979025262559904,
+      "loss": 1.1019,
+      "step": 751
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1685924530029297,
+      "learning_rate": 0.00019789688117925335,
+      "loss": 1.116,
+      "step": 752
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.17273302376270294,
+      "learning_rate": 0.00019789122859695737,
+      "loss": 1.0744,
+      "step": 753
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1677582710981369,
+      "learning_rate": 0.0001978855685095358,
+      "loss": 1.0655,
+      "step": 754
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.16952186822891235,
+      "learning_rate": 0.0001978799009174226,
+      "loss": 1.0505,
+      "step": 755
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1648993194103241,
+      "learning_rate": 0.00019787422582105227,
+      "loss": 1.1172,
+      "step": 756
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1641387939453125,
+      "learning_rate": 0.00019786854322085997,
+      "loss": 1.0762,
+      "step": 757
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.17605146765708923,
+      "learning_rate": 0.0001978628531172813,
+      "loss": 1.0586,
+      "step": 758
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.17868219316005707,
+      "learning_rate": 0.0001978571555107526,
+      "loss": 1.0759,
+      "step": 759
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1822500377893448,
+      "learning_rate": 0.0001978514504017106,
+      "loss": 1.116,
+      "step": 760
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.17165422439575195,
+      "learning_rate": 0.0001978457377905927,
+      "loss": 1.0491,
+      "step": 761
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1742304265499115,
+      "learning_rate": 0.000197840017677837,
+      "loss": 1.0565,
+      "step": 762
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.16227282583713531,
+      "learning_rate": 0.0001978342900638819,
+      "loss": 1.067,
+      "step": 763
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.17510762810707092,
+      "learning_rate": 0.0001978285549491666,
+      "loss": 1.0838,
+      "step": 764
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1696232557296753,
+      "learning_rate": 0.0001978228123341308,
+      "loss": 1.0111,
+      "step": 765
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1753813475370407,
+      "learning_rate": 0.00019781706221921473,
+      "loss": 1.0459,
+      "step": 766
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.16556529700756073,
+      "learning_rate": 0.0001978113046048593,
+      "loss": 1.0401,
+      "step": 767
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.16358666121959686,
+      "learning_rate": 0.0001978055394915059,
+      "loss": 1.0166,
+      "step": 768
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1697453260421753,
+      "learning_rate": 0.00019779976687959652,
+      "loss": 1.0732,
+      "step": 769
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.17842546105384827,
+      "learning_rate": 0.00019779398676957375,
+      "loss": 1.0354,
+      "step": 770
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.17023664712905884,
+      "learning_rate": 0.00019778819916188076,
+      "loss": 1.1196,
+      "step": 771
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.18845778703689575,
+      "learning_rate": 0.00019778240405696123,
+      "loss": 1.084,
+      "step": 772
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.17198894917964935,
+      "learning_rate": 0.0001977766014552595,
+      "loss": 1.0415,
+      "step": 773
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.17236502468585968,
+      "learning_rate": 0.00019777079135722042,
+      "loss": 1.0656,
+      "step": 774
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1829390674829483,
+      "learning_rate": 0.00019776497376328944,
+      "loss": 1.0939,
+      "step": 775
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.16156379878520966,
+      "learning_rate": 0.00019775914867391258,
+      "loss": 1.0654,
+      "step": 776
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.17911884188652039,
+      "learning_rate": 0.0001977533160895365,
+      "loss": 1.0854,
+      "step": 777
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.17780935764312744,
+      "learning_rate": 0.00019774747601060825,
+      "loss": 1.1216,
+      "step": 778
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.16712267696857452,
+      "learning_rate": 0.0001977416284375757,
+      "loss": 1.0472,
+      "step": 779
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.17137813568115234,
+      "learning_rate": 0.0001977357733708871,
+      "loss": 1.0768,
+      "step": 780
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.16970902681350708,
+      "learning_rate": 0.00019772991081099136,
+      "loss": 1.1034,
+      "step": 781
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.17829537391662598,
+      "learning_rate": 0.00019772404075833795,
+      "loss": 1.1081,
+      "step": 782
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1679857075214386,
+      "learning_rate": 0.00019771816321337692,
+      "loss": 1.0901,
+      "step": 783
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.17669101059436798,
+      "learning_rate": 0.00019771227817655892,
+      "loss": 1.008,
+      "step": 784
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.18435387313365936,
+      "learning_rate": 0.0001977063856483351,
+      "loss": 1.0802,
+      "step": 785
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.16921114921569824,
+      "learning_rate": 0.00019770048562915723,
+      "loss": 1.0708,
+      "step": 786
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.17006106674671173,
+      "learning_rate": 0.00019769457811947766,
+      "loss": 1.0267,
+      "step": 787
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.19145065546035767,
+      "learning_rate": 0.00019768866311974933,
+      "loss": 1.1084,
+      "step": 788
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.17966581881046295,
+      "learning_rate": 0.00019768274063042572,
+      "loss": 1.0757,
+      "step": 789
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.16242194175720215,
+      "learning_rate": 0.00019767681065196085,
+      "loss": 1.06,
+      "step": 790
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.17664171755313873,
+      "learning_rate": 0.00019767087318480945,
+      "loss": 1.0989,
+      "step": 791
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.18222616612911224,
+      "learning_rate": 0.00019766492822942668,
+      "loss": 1.1169,
+      "step": 792
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.17824456095695496,
+      "learning_rate": 0.0001976589757862683,
+      "loss": 1.0474,
+      "step": 793
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.17561721801757812,
+      "learning_rate": 0.00019765301585579072,
+      "loss": 1.0891,
+      "step": 794
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1854829043149948,
+      "learning_rate": 0.00019764704843845086,
+      "loss": 1.0897,
+      "step": 795
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.17442169785499573,
+      "learning_rate": 0.0001976410735347062,
+      "loss": 1.008,
+      "step": 796
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.16896742582321167,
+      "learning_rate": 0.0001976350911450149,
+      "loss": 1.1197,
+      "step": 797
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1652253419160843,
+      "learning_rate": 0.00019762910126983556,
+      "loss": 1.0353,
+      "step": 798
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.17013123631477356,
+      "learning_rate": 0.0001976231039096274,
+      "loss": 1.0953,
+      "step": 799
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.16675332188606262,
+      "learning_rate": 0.00019761709906485026,
+      "loss": 1.0707,
+      "step": 800
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.164289191365242,
+      "learning_rate": 0.00019761108673596448,
+      "loss": 1.035,
+      "step": 801
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1685546636581421,
+      "learning_rate": 0.00019760506692343107,
+      "loss": 1.0126,
+      "step": 802
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.159869983792305,
+      "learning_rate": 0.00019759903962771156,
+      "loss": 1.0443,
+      "step": 803
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.17488044500350952,
+      "learning_rate": 0.00019759300484926796,
+      "loss": 1.0821,
+      "step": 804
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.18100354075431824,
+      "learning_rate": 0.00019758696258856303,
+      "loss": 1.0498,
+      "step": 805
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.16927726566791534,
+      "learning_rate": 0.00019758091284606,
+      "loss": 1.0592,
+      "step": 806
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.17606563866138458,
+      "learning_rate": 0.00019757485562222265,
+      "loss": 1.0358,
+      "step": 807
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1672973483800888,
+      "learning_rate": 0.00019756879091751544,
+      "loss": 0.9804,
+      "step": 808
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1743594855070114,
+      "learning_rate": 0.0001975627187324033,
+      "loss": 1.0395,
+      "step": 809
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.17184039950370789,
+      "learning_rate": 0.0001975566390673518,
+      "loss": 1.0169,
+      "step": 810
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.17724105715751648,
+      "learning_rate": 0.000197550551922827,
+      "loss": 0.9956,
+      "step": 811
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.16138318181037903,
+      "learning_rate": 0.00019754445729929562,
+      "loss": 1.0382,
+      "step": 812
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.18038204312324524,
+      "learning_rate": 0.00019753835519722494,
+      "loss": 1.0985,
+      "step": 813
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.17275594174861908,
+      "learning_rate": 0.0001975322456170828,
+      "loss": 1.0409,
+      "step": 814
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.17210884392261505,
+      "learning_rate": 0.00019752612855933755,
+      "loss": 1.0578,
+      "step": 815
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.15895509719848633,
+      "learning_rate": 0.00019752000402445825,
+      "loss": 0.9716,
+      "step": 816
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1858808845281601,
+      "learning_rate": 0.00019751387201291442,
+      "loss": 1.0542,
+      "step": 817
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.16655674576759338,
+      "learning_rate": 0.00019750773252517618,
+      "loss": 1.0328,
+      "step": 818
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.17627952992916107,
+      "learning_rate": 0.00019750158556171426,
+      "loss": 1.0429,
+      "step": 819
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.16836844384670258,
+      "learning_rate": 0.0001974954311229999,
+      "loss": 1.0179,
+      "step": 820
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.17369742691516876,
+      "learning_rate": 0.000197489269209505,
+      "loss": 1.0074,
+      "step": 821
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.17023475468158722,
+      "learning_rate": 0.0001974830998217019,
+      "loss": 1.0429,
+      "step": 822
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.18552331626415253,
+      "learning_rate": 0.00019747692296006366,
+      "loss": 1.1223,
+      "step": 823
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.18040619790554047,
+      "learning_rate": 0.00019747073862506383,
+      "loss": 1.0913,
+      "step": 824
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1917116492986679,
+      "learning_rate": 0.00019746454681717656,
+      "loss": 1.0605,
+      "step": 825
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.18105146288871765,
+      "learning_rate": 0.00019745834753687652,
+      "loss": 1.0573,
+      "step": 826
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.16846537590026855,
+      "learning_rate": 0.00019745214078463908,
+      "loss": 1.0284,
+      "step": 827
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.18332545459270477,
+      "learning_rate": 0.00019744592656094004,
+      "loss": 1.0732,
+      "step": 828
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.18125556409358978,
+      "learning_rate": 0.00019743970486625582,
+      "loss": 0.9678,
+      "step": 829
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.16851511597633362,
+      "learning_rate": 0.00019743347570106348,
+      "loss": 1.0347,
+      "step": 830
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.18086205422878265,
+      "learning_rate": 0.00019742723906584052,
+      "loss": 1.1649,
+      "step": 831
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.176468163728714,
+      "learning_rate": 0.0001974209949610652,
+      "loss": 1.0727,
+      "step": 832
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.17329242825508118,
+      "learning_rate": 0.00019741474338721613,
+      "loss": 1.0446,
+      "step": 833
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.186629056930542,
+      "learning_rate": 0.00019740848434477267,
+      "loss": 1.028,
+      "step": 834
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.17307981848716736,
+      "learning_rate": 0.00019740221783421467,
+      "loss": 1.0449,
+      "step": 835
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.17103774845600128,
+      "learning_rate": 0.00019739594385602256,
+      "loss": 1.1006,
+      "step": 836
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.18144570291042328,
+      "learning_rate": 0.00019738966241067738,
+      "loss": 1.1385,
+      "step": 837
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.17040754854679108,
+      "learning_rate": 0.00019738337349866071,
+      "loss": 1.0435,
+      "step": 838
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.16767257452011108,
+      "learning_rate": 0.0001973770771204547,
+      "loss": 1.0754,
+      "step": 839
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1669158935546875,
+      "learning_rate": 0.00019737077327654204,
+      "loss": 1.0454,
+      "step": 840
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.18039458990097046,
+      "learning_rate": 0.00019736446196740612,
+      "loss": 1.083,
+      "step": 841
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.18282979726791382,
+      "learning_rate": 0.00019735814319353078,
+      "loss": 1.065,
+      "step": 842
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.18812443315982819,
+      "learning_rate": 0.00019735181695540043,
+      "loss": 1.0857,
+      "step": 843
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.17648079991340637,
+      "learning_rate": 0.00019734548325350013,
+      "loss": 1.0887,
+      "step": 844
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.18308205902576447,
+      "learning_rate": 0.00019733914208831547,
+      "loss": 1.0436,
+      "step": 845
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1741240918636322,
+      "learning_rate": 0.0001973327934603326,
+      "loss": 0.9797,
+      "step": 846
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.17795363068580627,
+      "learning_rate": 0.00019732643737003827,
+      "loss": 0.9717,
+      "step": 847
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.18939168751239777,
+      "learning_rate": 0.00019732007381791978,
+      "loss": 1.1139,
+      "step": 848
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.17296737432479858,
+      "learning_rate": 0.00019731370280446497,
+      "loss": 1.0673,
+      "step": 849
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.16962456703186035,
+      "learning_rate": 0.00019730732433016236,
+      "loss": 1.0461,
+      "step": 850
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.170401930809021,
+      "learning_rate": 0.00019730093839550098,
+      "loss": 1.0249,
+      "step": 851
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1670999675989151,
+      "learning_rate": 0.00019729454500097036,
+      "loss": 1.0277,
+      "step": 852
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.17348629236221313,
+      "learning_rate": 0.00019728814414706074,
+      "loss": 1.1394,
+      "step": 853
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1759617030620575,
+      "learning_rate": 0.0001972817358342628,
+      "loss": 1.0365,
+      "step": 854
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.16651113331317902,
+      "learning_rate": 0.00019727532006306788,
+      "loss": 1.0067,
+      "step": 855
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1713470071554184,
+      "learning_rate": 0.00019726889683396786,
+      "loss": 1.0365,
+      "step": 856
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1733425259590149,
+      "learning_rate": 0.0001972624661474552,
+      "loss": 1.1274,
+      "step": 857
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.16293053328990936,
+      "learning_rate": 0.00019725602800402294,
+      "loss": 1.0848,
+      "step": 858
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1699129045009613,
+      "learning_rate": 0.00019724958240416467,
+      "loss": 1.0418,
+      "step": 859
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1770995557308197,
+      "learning_rate": 0.00019724312934837454,
+      "loss": 1.1251,
+      "step": 860
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.17217662930488586,
+      "learning_rate": 0.00019723666883714733,
+      "loss": 1.075,
+      "step": 861
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.18040764331817627,
+      "learning_rate": 0.00019723020087097833,
+      "loss": 1.0622,
+      "step": 862
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.16920557618141174,
+      "learning_rate": 0.00019722372545036342,
+      "loss": 1.0461,
+      "step": 863
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.17057543992996216,
+      "learning_rate": 0.00019721724257579907,
+      "loss": 1.0893,
+      "step": 864
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.18235866725444794,
+      "learning_rate": 0.0001972107522477823,
+      "loss": 1.026,
+      "step": 865
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.18047012388706207,
+      "learning_rate": 0.00019720425446681077,
+      "loss": 1.0588,
+      "step": 866
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1593666523694992,
+      "learning_rate": 0.00019719774923338253,
+      "loss": 0.9946,
+      "step": 867
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.17716600000858307,
+      "learning_rate": 0.00019719123654799646,
+      "loss": 1.0427,
+      "step": 868
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.168783500790596,
+      "learning_rate": 0.00019718471641115177,
+      "loss": 1.0223,
+      "step": 869
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.18226021528244019,
+      "learning_rate": 0.00019717818882334837,
+      "loss": 1.0657,
+      "step": 870
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.17665214836597443,
+      "learning_rate": 0.00019717165378508678,
+      "loss": 1.0003,
+      "step": 871
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.17209143936634064,
+      "learning_rate": 0.00019716511129686795,
+      "loss": 1.0387,
+      "step": 872
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.17868731915950775,
+      "learning_rate": 0.00019715856135919352,
+      "loss": 1.0744,
+      "step": 873
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.16554374992847443,
+      "learning_rate": 0.00019715200397256563,
+      "loss": 1.049,
+      "step": 874
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.15895216166973114,
+      "learning_rate": 0.00019714543913748706,
+      "loss": 0.9866,
+      "step": 875
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1729590743780136,
+      "learning_rate": 0.0001971388668544611,
+      "loss": 1.0204,
+      "step": 876
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.17698253691196442,
+      "learning_rate": 0.00019713228712399167,
+      "loss": 1.0451,
+      "step": 877
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.18253202736377716,
+      "learning_rate": 0.00019712569994658315,
+      "loss": 1.1374,
+      "step": 878
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.17986972630023956,
+      "learning_rate": 0.00019711910532274062,
+      "loss": 1.0667,
+      "step": 879
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.18846715986728668,
+      "learning_rate": 0.00019711250325296967,
+      "loss": 1.1056,
+      "step": 880
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.17018909752368927,
+      "learning_rate": 0.00019710589373777646,
+      "loss": 0.9997,
+      "step": 881
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1827511042356491,
+      "learning_rate": 0.00019709927677766774,
+      "loss": 1.0255,
+      "step": 882
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.18419188261032104,
+      "learning_rate": 0.0001970926523731508,
+      "loss": 1.0579,
+      "step": 883
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.17267294228076935,
+      "learning_rate": 0.00019708602052473357,
+      "loss": 1.0677,
+      "step": 884
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1671350598335266,
+      "learning_rate": 0.00019707938123292442,
+      "loss": 1.0914,
+      "step": 885
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.16826584935188293,
+      "learning_rate": 0.00019707273449823243,
+      "loss": 1.0836,
+      "step": 886
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.16998471319675446,
+      "learning_rate": 0.00019706608032116714,
+      "loss": 1.0238,
+      "step": 887
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1634981334209442,
+      "learning_rate": 0.0001970594187022388,
+      "loss": 1.0445,
+      "step": 888
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.16182340681552887,
+      "learning_rate": 0.00019705274964195807,
+      "loss": 0.9969,
+      "step": 889
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.17049826681613922,
+      "learning_rate": 0.0001970460731408363,
+      "loss": 1.0772,
+      "step": 890
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1803371161222458,
+      "learning_rate": 0.0001970393891993853,
+      "loss": 1.07,
+      "step": 891
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.17657095193862915,
+      "learning_rate": 0.00019703269781811756,
+      "loss": 1.0898,
+      "step": 892
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1745598018169403,
+      "learning_rate": 0.00019702599899754607,
+      "loss": 1.0386,
+      "step": 893
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1789242923259735,
+      "learning_rate": 0.00019701929273818446,
+      "loss": 1.0609,
+      "step": 894
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.17516635358333588,
+      "learning_rate": 0.00019701257904054686,
+      "loss": 1.0211,
+      "step": 895
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.17327910661697388,
+      "learning_rate": 0.000197005857905148,
+      "loss": 1.0538,
+      "step": 896
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1742434799671173,
+      "learning_rate": 0.00019699912933250315,
+      "loss": 1.085,
+      "step": 897
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.17944319546222687,
+      "learning_rate": 0.0001969923933231282,
+      "loss": 0.9968,
+      "step": 898
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.17509445548057556,
+      "learning_rate": 0.00019698564987753958,
+      "loss": 1.0612,
+      "step": 899
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.18131905794143677,
+      "learning_rate": 0.0001969788989962543,
+      "loss": 1.0615,
+      "step": 900
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.17456522583961487,
+      "learning_rate": 0.00019697214067978999,
+      "loss": 1.0396,
+      "step": 901
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.18266192078590393,
+      "learning_rate": 0.00019696537492866468,
+      "loss": 1.0479,
+      "step": 902
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1651952564716339,
+      "learning_rate": 0.00019695860174339714,
+      "loss": 1.0023,
+      "step": 903
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1714158058166504,
+      "learning_rate": 0.0001969518211245067,
+      "loss": 1.0011,
+      "step": 904
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.18282106518745422,
+      "learning_rate": 0.00019694503307251317,
+      "loss": 1.1196,
+      "step": 905
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.16715645790100098,
+      "learning_rate": 0.000196938237587937,
+      "loss": 1.0366,
+      "step": 906
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.18058747053146362,
+      "learning_rate": 0.00019693143467129916,
+      "loss": 1.0491,
+      "step": 907
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.18357785046100616,
+      "learning_rate": 0.00019692462432312124,
+      "loss": 1.0959,
+      "step": 908
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.16413332521915436,
+      "learning_rate": 0.00019691780654392535,
+      "loss": 0.9923,
+      "step": 909
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.17373456060886383,
+      "learning_rate": 0.00019691098133423423,
+      "loss": 1.0187,
+      "step": 910
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.18135735392570496,
+      "learning_rate": 0.00019690414869457117,
+      "loss": 1.0138,
+      "step": 911
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.16327910125255585,
+      "learning_rate": 0.00019689730862545995,
+      "loss": 1.01,
+      "step": 912
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1778937131166458,
+      "learning_rate": 0.00019689046112742503,
+      "loss": 1.0789,
+      "step": 913
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.16726310551166534,
+      "learning_rate": 0.00019688360620099133,
+      "loss": 0.986,
+      "step": 914
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1793222278356552,
+      "learning_rate": 0.00019687674384668451,
+      "loss": 1.1003,
+      "step": 915
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.16796770691871643,
+      "learning_rate": 0.00019686987406503063,
+      "loss": 1.0466,
+      "step": 916
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.16727067530155182,
+      "learning_rate": 0.00019686299685655638,
+      "loss": 0.987,
+      "step": 917
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.17687837779521942,
+      "learning_rate": 0.00019685611222178905,
+      "loss": 1.0444,
+      "step": 918
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1811404675245285,
+      "learning_rate": 0.0001968492201612564,
+      "loss": 1.0468,
+      "step": 919
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1711091250181198,
+      "learning_rate": 0.00019684232067548695,
+      "loss": 1.0123,
+      "step": 920
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1783980429172516,
+      "learning_rate": 0.00019683541376500955,
+      "loss": 1.0674,
+      "step": 921
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.18069174885749817,
+      "learning_rate": 0.00019682849943035384,
+      "loss": 1.067,
+      "step": 922
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1692749261856079,
+      "learning_rate": 0.00019682157767204986,
+      "loss": 1.012,
+      "step": 923
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.17725811898708344,
+      "learning_rate": 0.00019681464849062832,
+      "loss": 0.9682,
+      "step": 924
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.17122116684913635,
+      "learning_rate": 0.00019680771188662044,
+      "loss": 1.0045,
+      "step": 925
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.17642873525619507,
+      "learning_rate": 0.00019680076786055806,
+      "loss": 1.0636,
+      "step": 926
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.19407136738300323,
+      "learning_rate": 0.00019679381641297353,
+      "loss": 1.1208,
+      "step": 927
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.17200174927711487,
+      "learning_rate": 0.00019678685754439987,
+      "loss": 1.0143,
+      "step": 928
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.17304272949695587,
+      "learning_rate": 0.00019677989125537054,
+      "loss": 0.9638,
+      "step": 929
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.18978925049304962,
+      "learning_rate": 0.00019677291754641965,
+      "loss": 1.007,
+      "step": 930
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.17057448625564575,
+      "learning_rate": 0.0001967659364180819,
+      "loss": 1.0672,
+      "step": 931
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.17902058362960815,
+      "learning_rate": 0.00019675894787089243,
+      "loss": 1.0503,
+      "step": 932
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.19893060624599457,
+      "learning_rate": 0.00019675195190538714,
+      "loss": 1.07,
+      "step": 933
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.16946718096733093,
+      "learning_rate": 0.0001967449485221023,
+      "loss": 1.0448,
+      "step": 934
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.18813353776931763,
+      "learning_rate": 0.00019673793772157492,
+      "loss": 1.1135,
+      "step": 935
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1719808280467987,
+      "learning_rate": 0.00019673091950434248,
+      "loss": 1.0385,
+      "step": 936
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.179408997297287,
+      "learning_rate": 0.00019672389387094302,
+      "loss": 1.0326,
+      "step": 937
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1784079670906067,
+      "learning_rate": 0.00019671686082191526,
+      "loss": 1.0154,
+      "step": 938
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.19028711318969727,
+      "learning_rate": 0.00019670982035779834,
+      "loss": 1.0589,
+      "step": 939
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1741345077753067,
+      "learning_rate": 0.00019670277247913205,
+      "loss": 1.0797,
+      "step": 940
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.18272146582603455,
+      "learning_rate": 0.00019669571718645675,
+      "loss": 1.0764,
+      "step": 941
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.17313426733016968,
+      "learning_rate": 0.00019668865448031335,
+      "loss": 1.0728,
+      "step": 942
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.17928192019462585,
+      "learning_rate": 0.00019668158436124332,
+      "loss": 1.0319,
+      "step": 943
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.16490381956100464,
+      "learning_rate": 0.00019667450682978876,
+      "loss": 1.0106,
+      "step": 944
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1769440919160843,
+      "learning_rate": 0.00019666742188649225,
+      "loss": 1.0644,
+      "step": 945
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1783764362335205,
+      "learning_rate": 0.000196660329531897,
+      "loss": 1.0492,
+      "step": 946
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.18292124569416046,
+      "learning_rate": 0.0001966532297665467,
+      "loss": 1.0893,
+      "step": 947
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.16674058139324188,
+      "learning_rate": 0.00019664612259098578,
+      "loss": 1.0172,
+      "step": 948
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.17283639311790466,
+      "learning_rate": 0.00019663900800575906,
+      "loss": 1.0527,
+      "step": 949
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1824640929698944,
+      "learning_rate": 0.00019663188601141203,
+      "loss": 1.1194,
+      "step": 950
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.17792586982250214,
+      "learning_rate": 0.00019662475660849066,
+      "loss": 1.0048,
+      "step": 951
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1702018827199936,
+      "learning_rate": 0.00019661761979754166,
+      "loss": 1.0611,
+      "step": 952
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.16764649748802185,
+      "learning_rate": 0.00019661047557911212,
+      "loss": 1.0499,
+      "step": 953
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1887722909450531,
+      "learning_rate": 0.00019660332395374979,
+      "loss": 1.0431,
+      "step": 954
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1827782541513443,
+      "learning_rate": 0.00019659616492200295,
+      "loss": 1.074,
+      "step": 955
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.17926305532455444,
+      "learning_rate": 0.00019658899848442048,
+      "loss": 1.0444,
+      "step": 956
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.16336272656917572,
+      "learning_rate": 0.00019658182464155183,
+      "loss": 0.9583,
+      "step": 957
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1775018870830536,
+      "learning_rate": 0.000196574643393947,
+      "loss": 0.9603,
+      "step": 958
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.18180976808071136,
+      "learning_rate": 0.00019656745474215656,
+      "loss": 1.0505,
+      "step": 959
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1839762032032013,
+      "learning_rate": 0.00019656025868673165,
+      "loss": 1.0521,
+      "step": 960
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.18608036637306213,
+      "learning_rate": 0.00019655305522822399,
+      "loss": 1.0153,
+      "step": 961
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.17250166833400726,
+      "learning_rate": 0.0001965458443671858,
+      "loss": 1.0238,
+      "step": 962
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.17518852651119232,
+      "learning_rate": 0.00019653862610417,
+      "loss": 0.9844,
+      "step": 963
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.18602406978607178,
+      "learning_rate": 0.00019653140043972994,
+      "loss": 1.0637,
+      "step": 964
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.16935952007770538,
+      "learning_rate": 0.0001965241673744196,
+      "loss": 1.0002,
+      "step": 965
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2699510157108307,
+      "learning_rate": 0.0001965169269087936,
+      "loss": 1.0215,
+      "step": 966
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.17568126320838928,
+      "learning_rate": 0.00019650967904340694,
+      "loss": 1.0628,
+      "step": 967
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.16942435503005981,
+      "learning_rate": 0.00019650242377881538,
+      "loss": 1.0706,
+      "step": 968
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.17300187051296234,
+      "learning_rate": 0.00019649516111557515,
+      "loss": 1.0659,
+      "step": 969
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1726924628019333,
+      "learning_rate": 0.00019648789105424305,
+      "loss": 1.102,
+      "step": 970
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1732690930366516,
+      "learning_rate": 0.00019648061359537646,
+      "loss": 1.1284,
+      "step": 971
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.17127656936645508,
+      "learning_rate": 0.00019647332873953336,
+      "loss": 1.0179,
+      "step": 972
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.17334236204624176,
+      "learning_rate": 0.00019646603648727223,
+      "loss": 1.0578,
+      "step": 973
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.16784228384494781,
+      "learning_rate": 0.00019645873683915213,
+      "loss": 1.0121,
+      "step": 974
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.167621910572052,
+      "learning_rate": 0.00019645142979573277,
+      "loss": 1.0196,
+      "step": 975
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1630851924419403,
+      "learning_rate": 0.00019644411535757437,
+      "loss": 1.0524,
+      "step": 976
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.16767187416553497,
+      "learning_rate": 0.00019643679352523763,
+      "loss": 1.04,
+      "step": 977
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.18620026111602783,
+      "learning_rate": 0.00019642946429928397,
+      "loss": 1.0708,
+      "step": 978
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.17906849086284637,
+      "learning_rate": 0.0001964221276802753,
+      "loss": 1.0709,
+      "step": 979
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1792149543762207,
+      "learning_rate": 0.0001964147836687741,
+      "loss": 1.0365,
+      "step": 980
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.17698103189468384,
+      "learning_rate": 0.0001964074322653434,
+      "loss": 1.0732,
+      "step": 981
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.17077916860580444,
+      "learning_rate": 0.00019640007347054684,
+      "loss": 1.0463,
+      "step": 982
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.17265799641609192,
+      "learning_rate": 0.0001963927072849486,
+      "loss": 1.0835,
+      "step": 983
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1759408712387085,
+      "learning_rate": 0.00019638533370911341,
+      "loss": 1.0469,
+      "step": 984
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.17300966382026672,
+      "learning_rate": 0.00019637795274360663,
+      "loss": 1.0623,
+      "step": 985
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1715632528066635,
+      "learning_rate": 0.0001963705643889941,
+      "loss": 1.0799,
+      "step": 986
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.17230769991874695,
+      "learning_rate": 0.0001963631686458423,
+      "loss": 1.0518,
+      "step": 987
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.17746640741825104,
+      "learning_rate": 0.0001963557655147182,
+      "loss": 1.0129,
+      "step": 988
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1871126890182495,
+      "learning_rate": 0.00019634835499618948,
+      "loss": 1.0845,
+      "step": 989
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.17565259337425232,
+      "learning_rate": 0.0001963409370908242,
+      "loss": 1.0699,
+      "step": 990
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1731378436088562,
+      "learning_rate": 0.00019633351179919109,
+      "loss": 0.9835,
+      "step": 991
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.18255259096622467,
+      "learning_rate": 0.00019632607912185949,
+      "loss": 0.964,
+      "step": 992
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1729213297367096,
+      "learning_rate": 0.00019631863905939916,
+      "loss": 1.0397,
+      "step": 993
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.17974966764450073,
+      "learning_rate": 0.0001963111916123806,
+      "loss": 1.0543,
+      "step": 994
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.17525100708007812,
+      "learning_rate": 0.0001963037367813747,
+      "loss": 1.0449,
+      "step": 995
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.18129664659500122,
+      "learning_rate": 0.0001962962745669531,
+      "loss": 1.0856,
+      "step": 996
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1881393939256668,
+      "learning_rate": 0.00019628880496968786,
+      "loss": 1.0776,
+      "step": 997
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.17546981573104858,
+      "learning_rate": 0.00019628132799015167,
+      "loss": 1.0396,
+      "step": 998
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1705467253923416,
+      "learning_rate": 0.00019627384362891776,
+      "loss": 1.0109,
+      "step": 999
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1715412139892578,
+      "learning_rate": 0.00019626635188656,
+      "loss": 1.0649,
+      "step": 1000
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.17314362525939941,
+      "learning_rate": 0.00019625885276365268,
+      "loss": 0.984,
+      "step": 1001
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.18583662807941437,
+      "learning_rate": 0.00019625134626077083,
+      "loss": 1.0614,
+      "step": 1002
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.18308642506599426,
+      "learning_rate": 0.00019624383237848986,
+      "loss": 1.0291,
+      "step": 1003
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.17275825142860413,
+      "learning_rate": 0.00019623631111738595,
+      "loss": 1.0661,
+      "step": 1004
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.17974300682544708,
+      "learning_rate": 0.00019622878247803569,
+      "loss": 1.0458,
+      "step": 1005
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.17539826035499573,
+      "learning_rate": 0.00019622124646101626,
+      "loss": 1.0174,
+      "step": 1006
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1848883181810379,
+      "learning_rate": 0.00019621370306690546,
+      "loss": 1.0746,
+      "step": 1007
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.16872359812259674,
+      "learning_rate": 0.00019620615229628164,
+      "loss": 1.0053,
+      "step": 1008
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.16738854348659515,
+      "learning_rate": 0.00019619859414972366,
+      "loss": 1.0096,
+      "step": 1009
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1777888685464859,
+      "learning_rate": 0.00019619102862781105,
+      "loss": 0.946,
+      "step": 1010
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.18343977630138397,
+      "learning_rate": 0.00019618345573112377,
+      "loss": 0.9957,
+      "step": 1011
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.20510901510715485,
+      "learning_rate": 0.0001961758754602425,
+      "loss": 1.0588,
+      "step": 1012
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.18343272805213928,
+      "learning_rate": 0.00019616828781574835,
+      "loss": 1.0221,
+      "step": 1013
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.17978180944919586,
+      "learning_rate": 0.00019616069279822305,
+      "loss": 1.0499,
+      "step": 1014
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1911647468805313,
+      "learning_rate": 0.00019615309040824888,
+      "loss": 1.075,
+      "step": 1015
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.17616963386535645,
+      "learning_rate": 0.00019614548064640877,
+      "loss": 1.0689,
+      "step": 1016
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.17211149632930756,
+      "learning_rate": 0.00019613786351328609,
+      "loss": 0.986,
+      "step": 1017
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1717488318681717,
+      "learning_rate": 0.00019613023900946484,
+      "loss": 1.0073,
+      "step": 1018
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.18268033862113953,
+      "learning_rate": 0.00019612260713552956,
+      "loss": 1.0969,
+      "step": 1019
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1773374378681183,
+      "learning_rate": 0.0001961149678920654,
+      "loss": 0.9878,
+      "step": 1020
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.17691907286643982,
+      "learning_rate": 0.00019610732127965804,
+      "loss": 1.1073,
+      "step": 1021
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.18169642984867096,
+      "learning_rate": 0.00019609966729889372,
+      "loss": 1.0644,
+      "step": 1022
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.17204183340072632,
+      "learning_rate": 0.00019609200595035927,
+      "loss": 0.9929,
+      "step": 1023
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1788758486509323,
+      "learning_rate": 0.00019608433723464206,
+      "loss": 1.0369,
+      "step": 1024
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.17738159000873566,
+      "learning_rate": 0.00019607666115233,
+      "loss": 1.0628,
+      "step": 1025
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1694379299879074,
+      "learning_rate": 0.00019606897770401167,
+      "loss": 1.0282,
+      "step": 1026
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.16664545238018036,
+      "learning_rate": 0.0001960612868902761,
+      "loss": 0.9741,
+      "step": 1027
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.16760802268981934,
+      "learning_rate": 0.00019605358871171294,
+      "loss": 0.9351,
+      "step": 1028
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.17086480557918549,
+      "learning_rate": 0.0001960458831689124,
+      "loss": 1.0735,
+      "step": 1029
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.17389248311519623,
+      "learning_rate": 0.00019603817026246523,
+      "loss": 0.9759,
+      "step": 1030
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1694263219833374,
+      "learning_rate": 0.0001960304499929628,
+      "loss": 1.0005,
+      "step": 1031
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.17080332338809967,
+      "learning_rate": 0.00019602272236099693,
+      "loss": 0.9775,
+      "step": 1032
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1651020646095276,
+      "learning_rate": 0.00019601498736716017,
+      "loss": 1.0308,
+      "step": 1033
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1663987785577774,
+      "learning_rate": 0.00019600724501204553,
+      "loss": 0.9909,
+      "step": 1034
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.17150457203388214,
+      "learning_rate": 0.00019599949529624654,
+      "loss": 1.0159,
+      "step": 1035
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.17601971328258514,
+      "learning_rate": 0.00019599173822035744,
+      "loss": 0.997,
+      "step": 1036
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.18561767041683197,
+      "learning_rate": 0.00019598397378497287,
+      "loss": 1.0094,
+      "step": 1037
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.17770105600357056,
+      "learning_rate": 0.00019597620199068817,
+      "loss": 1.0259,
+      "step": 1038
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.17730176448822021,
+      "learning_rate": 0.00019596842283809917,
+      "loss": 1.0212,
+      "step": 1039
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1914602667093277,
+      "learning_rate": 0.0001959606363278023,
+      "loss": 1.086,
+      "step": 1040
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.18201208114624023,
+      "learning_rate": 0.00019595284246039447,
+      "loss": 1.0525,
+      "step": 1041
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.16805338859558105,
+      "learning_rate": 0.0001959450412364733,
+      "loss": 1.0172,
+      "step": 1042
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.18455561995506287,
+      "learning_rate": 0.00019593723265663685,
+      "loss": 1.0622,
+      "step": 1043
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.18691962957382202,
+      "learning_rate": 0.0001959294167214838,
+      "loss": 1.0064,
+      "step": 1044
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.17232896387577057,
+      "learning_rate": 0.00019592159343161342,
+      "loss": 0.9567,
+      "step": 1045
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.17483456432819366,
+      "learning_rate": 0.00019591376278762541,
+      "loss": 1.0152,
+      "step": 1046
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.18186868727207184,
+      "learning_rate": 0.00019590592479012023,
+      "loss": 0.988,
+      "step": 1047
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.19091114401817322,
+      "learning_rate": 0.00019589807943969873,
+      "loss": 1.0016,
+      "step": 1048
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.17381857335567474,
+      "learning_rate": 0.00019589022673696245,
+      "loss": 1.0549,
+      "step": 1049
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1698593646287918,
+      "learning_rate": 0.0001958823666825134,
+      "loss": 1.0751,
+      "step": 1050
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1881944239139557,
+      "learning_rate": 0.00019587449927695424,
+      "loss": 1.0665,
+      "step": 1051
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.17412430047988892,
+      "learning_rate": 0.00019586662452088813,
+      "loss": 1.0164,
+      "step": 1052
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.17845436930656433,
+      "learning_rate": 0.0001958587424149188,
+      "loss": 1.0372,
+      "step": 1053
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.17655451595783234,
+      "learning_rate": 0.00019585085295965055,
+      "loss": 1.0524,
+      "step": 1054
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.17986845970153809,
+      "learning_rate": 0.00019584295615568826,
+      "loss": 1.0226,
+      "step": 1055
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.17774830758571625,
+      "learning_rate": 0.00019583505200363734,
+      "loss": 1.0196,
+      "step": 1056
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.18801482021808624,
+      "learning_rate": 0.00019582714050410386,
+      "loss": 1.0183,
+      "step": 1057
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.173845112323761,
+      "learning_rate": 0.00019581922165769428,
+      "loss": 1.0156,
+      "step": 1058
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.17635942995548248,
+      "learning_rate": 0.0001958112954650158,
+      "loss": 1.0471,
+      "step": 1059
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.18246427178382874,
+      "learning_rate": 0.00019580336192667606,
+      "loss": 1.0075,
+      "step": 1060
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.18411587178707123,
+      "learning_rate": 0.0001957954210432833,
+      "loss": 1.0057,
+      "step": 1061
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.16751395165920258,
+      "learning_rate": 0.00019578747281544638,
+      "loss": 0.9679,
+      "step": 1062
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1719505786895752,
+      "learning_rate": 0.00019577951724377462,
+      "loss": 1.0377,
+      "step": 1063
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.16536471247673035,
+      "learning_rate": 0.00019577155432887804,
+      "loss": 1.0532,
+      "step": 1064
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.19073115289211273,
+      "learning_rate": 0.00019576358407136707,
+      "loss": 1.0635,
+      "step": 1065
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.18181023001670837,
+      "learning_rate": 0.00019575560647185278,
+      "loss": 1.0941,
+      "step": 1066
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.16925771534442902,
+      "learning_rate": 0.0001957476215309468,
+      "loss": 1.0137,
+      "step": 1067
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.17827220261096954,
+      "learning_rate": 0.00019573962924926134,
+      "loss": 1.047,
+      "step": 1068
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1751449555158615,
+      "learning_rate": 0.00019573162962740914,
+      "loss": 1.0427,
+      "step": 1069
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.18006440997123718,
+      "learning_rate": 0.00019572362266600353,
+      "loss": 1.0312,
+      "step": 1070
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.17753449082374573,
+      "learning_rate": 0.00019571560836565834,
+      "loss": 1.0177,
+      "step": 1071
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.17010781168937683,
+      "learning_rate": 0.00019570758672698805,
+      "loss": 1.087,
+      "step": 1072
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.17299331724643707,
+      "learning_rate": 0.00019569955775060766,
+      "loss": 1.0537,
+      "step": 1073
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.17425057291984558,
+      "learning_rate": 0.00019569152143713274,
+      "loss": 0.9863,
+      "step": 1074
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.17348161339759827,
+      "learning_rate": 0.0001956834777871794,
+      "loss": 1.0099,
+      "step": 1075
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.17897789180278778,
+      "learning_rate": 0.00019567542680136438,
+      "loss": 1.1186,
+      "step": 1076
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.16750477254390717,
+      "learning_rate": 0.00019566736848030484,
+      "loss": 0.9746,
+      "step": 1077
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.18266913294792175,
+      "learning_rate": 0.00019565930282461867,
+      "loss": 1.0324,
+      "step": 1078
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.17874056100845337,
+      "learning_rate": 0.00019565122983492423,
+      "loss": 1.0307,
+      "step": 1079
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1887243390083313,
+      "learning_rate": 0.00019564314951184047,
+      "loss": 1.048,
+      "step": 1080
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.17233537137508392,
+      "learning_rate": 0.00019563506185598685,
+      "loss": 1.0384,
+      "step": 1081
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.18821567296981812,
+      "learning_rate": 0.0001956269668679835,
+      "loss": 1.0694,
+      "step": 1082
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.19365891814231873,
+      "learning_rate": 0.00019561886454845097,
+      "loss": 1.0784,
+      "step": 1083
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.18988920748233795,
+      "learning_rate": 0.00019561075489801054,
+      "loss": 1.0649,
+      "step": 1084
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.18308693170547485,
+      "learning_rate": 0.00019560263791728387,
+      "loss": 1.0837,
+      "step": 1085
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1714373528957367,
+      "learning_rate": 0.0001955945136068933,
+      "loss": 1.071,
+      "step": 1086
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.17758604884147644,
+      "learning_rate": 0.00019558638196746176,
+      "loss": 1.0841,
+      "step": 1087
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.17851248383522034,
+      "learning_rate": 0.0001955782429996126,
+      "loss": 1.0355,
+      "step": 1088
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.16533468663692474,
+      "learning_rate": 0.0001955700967039699,
+      "loss": 1.0278,
+      "step": 1089
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.17190004885196686,
+      "learning_rate": 0.00019556194308115816,
+      "loss": 1.0253,
+      "step": 1090
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.17942769825458527,
+      "learning_rate": 0.00019555378213180252,
+      "loss": 0.9406,
+      "step": 1091
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.17903082072734833,
+      "learning_rate": 0.0001955456138565287,
+      "loss": 1.037,
+      "step": 1092
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1841680407524109,
+      "learning_rate": 0.0001955374382559629,
+      "loss": 1.0799,
+      "step": 1093
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.18083688616752625,
+      "learning_rate": 0.00019552925533073193,
+      "loss": 1.0175,
+      "step": 1094
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1914413422346115,
+      "learning_rate": 0.00019552106508146318,
+      "loss": 1.1054,
+      "step": 1095
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.18662455677986145,
+      "learning_rate": 0.00019551286750878454,
+      "loss": 1.0377,
+      "step": 1096
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.17855742573738098,
+      "learning_rate": 0.00019550466261332455,
+      "loss": 1.0656,
+      "step": 1097
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.17760635912418365,
+      "learning_rate": 0.00019549645039571226,
+      "loss": 1.0806,
+      "step": 1098
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.17919620871543884,
+      "learning_rate": 0.00019548823085657727,
+      "loss": 0.9743,
+      "step": 1099
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.17144104838371277,
+      "learning_rate": 0.00019548000399654975,
+      "loss": 0.9721,
+      "step": 1100
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1793820559978485,
+      "learning_rate": 0.00019547176981626042,
+      "loss": 1.038,
+      "step": 1101
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.17242422699928284,
+      "learning_rate": 0.00019546352831634062,
+      "loss": 0.9687,
+      "step": 1102
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.17942006886005402,
+      "learning_rate": 0.0001954552794974222,
+      "loss": 1.0924,
+      "step": 1103
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.19152596592903137,
+      "learning_rate": 0.00019544702336013754,
+      "loss": 1.0564,
+      "step": 1104
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.17344453930854797,
+      "learning_rate": 0.00019543875990511969,
+      "loss": 1.0529,
+      "step": 1105
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.17758041620254517,
+      "learning_rate": 0.0001954304891330021,
+      "loss": 1.0792,
+      "step": 1106
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.17804868519306183,
+      "learning_rate": 0.00019542221104441895,
+      "loss": 1.0285,
+      "step": 1107
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.191632479429245,
+      "learning_rate": 0.00019541392564000488,
+      "loss": 0.9587,
+      "step": 1108
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.17585574090480804,
+      "learning_rate": 0.00019540563292039513,
+      "loss": 0.9943,
+      "step": 1109
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.16620998084545135,
+      "learning_rate": 0.0001953973328862255,
+      "loss": 1.012,
+      "step": 1110
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1953848898410797,
+      "learning_rate": 0.00019538902553813226,
+      "loss": 1.0499,
+      "step": 1111
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.18621788918972015,
+      "learning_rate": 0.0001953807108767524,
+      "loss": 1.0103,
+      "step": 1112
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1733444482088089,
+      "learning_rate": 0.00019537238890272334,
+      "loss": 1.1045,
+      "step": 1113
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1846238672733307,
+      "learning_rate": 0.00019536405961668313,
+      "loss": 1.0661,
+      "step": 1114
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.18024377524852753,
+      "learning_rate": 0.00019535572301927032,
+      "loss": 1.0236,
+      "step": 1115
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1756822168827057,
+      "learning_rate": 0.00019534737911112412,
+      "loss": 1.0008,
+      "step": 1116
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.17604592442512512,
+      "learning_rate": 0.00019533902789288426,
+      "loss": 1.0103,
+      "step": 1117
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.18618983030319214,
+      "learning_rate": 0.00019533066936519092,
+      "loss": 1.0273,
+      "step": 1118
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.17751984298229218,
+      "learning_rate": 0.00019532230352868499,
+      "loss": 1.0341,
+      "step": 1119
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2014542818069458,
+      "learning_rate": 0.00019531393038400786,
+      "loss": 1.0502,
+      "step": 1120
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.17488035559654236,
+      "learning_rate": 0.00019530554993180151,
+      "loss": 1.0167,
+      "step": 1121
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.18563339114189148,
+      "learning_rate": 0.00019529716217270835,
+      "loss": 1.0396,
+      "step": 1122
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.17215563356876373,
+      "learning_rate": 0.00019528876710737155,
+      "loss": 1.0216,
+      "step": 1123
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.18212297558784485,
+      "learning_rate": 0.00019528036473643473,
+      "loss": 0.9686,
+      "step": 1124
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1783788949251175,
+      "learning_rate": 0.00019527195506054203,
+      "loss": 1.0206,
+      "step": 1125
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1766308844089508,
+      "learning_rate": 0.00019526353808033825,
+      "loss": 1.0558,
+      "step": 1126
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.18010054528713226,
+      "learning_rate": 0.0001952551137964687,
+      "loss": 1.0818,
+      "step": 1127
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.19283832609653473,
+      "learning_rate": 0.00019524668220957927,
+      "loss": 1.1005,
+      "step": 1128
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1859970986843109,
+      "learning_rate": 0.00019523824332031632,
+      "loss": 1.0519,
+      "step": 1129
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.17316271364688873,
+      "learning_rate": 0.00019522979712932693,
+      "loss": 1.039,
+      "step": 1130
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1714569330215454,
+      "learning_rate": 0.0001952213436372586,
+      "loss": 1.0384,
+      "step": 1131
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.18875844776630402,
+      "learning_rate": 0.00019521288284475943,
+      "loss": 1.0167,
+      "step": 1132
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1846914291381836,
+      "learning_rate": 0.00019520441475247814,
+      "loss": 0.9885,
+      "step": 1133
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.18210475146770477,
+      "learning_rate": 0.00019519593936106395,
+      "loss": 0.9981,
+      "step": 1134
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1718529462814331,
+      "learning_rate": 0.00019518745667116663,
+      "loss": 0.9566,
+      "step": 1135
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.17763008177280426,
+      "learning_rate": 0.00019517896668343654,
+      "loss": 1.0626,
+      "step": 1136
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.17546570301055908,
+      "learning_rate": 0.00019517046939852462,
+      "loss": 1.026,
+      "step": 1137
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.174907848238945,
+      "learning_rate": 0.00019516196481708228,
+      "loss": 1.0155,
+      "step": 1138
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.18386012315750122,
+      "learning_rate": 0.00019515345293976156,
+      "loss": 1.0359,
+      "step": 1139
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.16808752715587616,
+      "learning_rate": 0.00019514493376721512,
+      "loss": 1.0262,
+      "step": 1140
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.17714083194732666,
+      "learning_rate": 0.00019513640730009604,
+      "loss": 1.0467,
+      "step": 1141
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.19117549061775208,
+      "learning_rate": 0.00019512787353905803,
+      "loss": 0.9753,
+      "step": 1142
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.17344416677951813,
+      "learning_rate": 0.0001951193324847554,
+      "loss": 1.078,
+      "step": 1143
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1775476038455963,
+      "learning_rate": 0.00019511078413784295,
+      "loss": 1.031,
+      "step": 1144
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.17347300052642822,
+      "learning_rate": 0.00019510222849897604,
+      "loss": 1.0538,
+      "step": 1145
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.17518822848796844,
+      "learning_rate": 0.00019509366556881065,
+      "loss": 1.0162,
+      "step": 1146
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.18036702275276184,
+      "learning_rate": 0.00019508509534800326,
+      "loss": 1.0921,
+      "step": 1147
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1831878125667572,
+      "learning_rate": 0.00019507651783721094,
+      "loss": 1.079,
+      "step": 1148
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1758822798728943,
+      "learning_rate": 0.00019506793303709132,
+      "loss": 1.0403,
+      "step": 1149
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.18290312588214874,
+      "learning_rate": 0.00019505934094830258,
+      "loss": 1.0331,
+      "step": 1150
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.190335214138031,
+      "learning_rate": 0.00019505074157150342,
+      "loss": 0.9999,
+      "step": 1151
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.16814689338207245,
+      "learning_rate": 0.0001950421349073532,
+      "loss": 1.0396,
+      "step": 1152
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.189680278301239,
+      "learning_rate": 0.00019503352095651174,
+      "loss": 1.022,
+      "step": 1153
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.18800580501556396,
+      "learning_rate": 0.00019502489971963942,
+      "loss": 1.0712,
+      "step": 1154
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.18286903202533722,
+      "learning_rate": 0.00019501627119739728,
+      "loss": 1.0245,
+      "step": 1155
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1748037189245224,
+      "learning_rate": 0.00019500763539044683,
+      "loss": 1.0192,
+      "step": 1156
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.19290375709533691,
+      "learning_rate": 0.00019499899229945012,
+      "loss": 1.0468,
+      "step": 1157
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1722959280014038,
+      "learning_rate": 0.00019499034192506983,
+      "loss": 1.0388,
+      "step": 1158
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.17889316380023956,
+      "learning_rate": 0.00019498168426796919,
+      "loss": 1.0789,
+      "step": 1159
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.17683634161949158,
+      "learning_rate": 0.00019497301932881194,
+      "loss": 1.0811,
+      "step": 1160
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.18203513324260712,
+      "learning_rate": 0.0001949643471082624,
+      "loss": 1.0446,
+      "step": 1161
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.18062527477741241,
+      "learning_rate": 0.00019495566760698547,
+      "loss": 1.0625,
+      "step": 1162
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.18026340007781982,
+      "learning_rate": 0.00019494698082564655,
+      "loss": 1.0631,
+      "step": 1163
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.17580631375312805,
+      "learning_rate": 0.0001949382867649117,
+      "loss": 0.9806,
+      "step": 1164
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1745951920747757,
+      "learning_rate": 0.00019492958542544742,
+      "loss": 1.0034,
+      "step": 1165
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.18265682458877563,
+      "learning_rate": 0.00019492087680792085,
+      "loss": 1.044,
+      "step": 1166
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.17897681891918182,
+      "learning_rate": 0.00019491216091299966,
+      "loss": 1.025,
+      "step": 1167
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.18096914887428284,
+      "learning_rate": 0.00019490343774135208,
+      "loss": 0.9971,
+      "step": 1168
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1791757494211197,
+      "learning_rate": 0.00019489470729364692,
+      "loss": 1.0091,
+      "step": 1169
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.18002015352249146,
+      "learning_rate": 0.00019488596957055348,
+      "loss": 1.0668,
+      "step": 1170
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.18384124338626862,
+      "learning_rate": 0.00019487722457274168,
+      "loss": 1.0324,
+      "step": 1171
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.17454668879508972,
+      "learning_rate": 0.000194868472300882,
+      "loss": 1.0107,
+      "step": 1172
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.17286750674247742,
+      "learning_rate": 0.00019485971275564546,
+      "loss": 1.0268,
+      "step": 1173
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.17558777332305908,
+      "learning_rate": 0.00019485094593770362,
+      "loss": 1.0354,
+      "step": 1174
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1863299012184143,
+      "learning_rate": 0.00019484217184772865,
+      "loss": 1.0204,
+      "step": 1175
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.18392202258110046,
+      "learning_rate": 0.00019483339048639316,
+      "loss": 1.0339,
+      "step": 1176
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.17207181453704834,
+      "learning_rate": 0.00019482460185437045,
+      "loss": 1.0036,
+      "step": 1177
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1741429716348648,
+      "learning_rate": 0.00019481580595233434,
+      "loss": 1.0321,
+      "step": 1178
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.18753582239151,
+      "learning_rate": 0.0001948070027809592,
+      "loss": 1.0876,
+      "step": 1179
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.16908766329288483,
+      "learning_rate": 0.00019479819234091993,
+      "loss": 0.9712,
+      "step": 1180
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.16926541924476624,
+      "learning_rate": 0.00019478937463289202,
+      "loss": 1.0192,
+      "step": 1181
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.18049052357673645,
+      "learning_rate": 0.0001947805496575515,
+      "loss": 0.9744,
+      "step": 1182
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.17877183854579926,
+      "learning_rate": 0.00019477171741557494,
+      "loss": 1.0187,
+      "step": 1183
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.17817731201648712,
+      "learning_rate": 0.00019476287790763953,
+      "loss": 1.0025,
+      "step": 1184
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.17894314229488373,
+      "learning_rate": 0.00019475403113442297,
+      "loss": 0.9888,
+      "step": 1185
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1737499237060547,
+      "learning_rate": 0.00019474517709660346,
+      "loss": 1.0435,
+      "step": 1186
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1924152672290802,
+      "learning_rate": 0.00019473631579485995,
+      "loss": 1.021,
+      "step": 1187
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.17013049125671387,
+      "learning_rate": 0.0001947274472298717,
+      "loss": 1.0198,
+      "step": 1188
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.17775459587574005,
+      "learning_rate": 0.00019471857140231872,
+      "loss": 1.0803,
+      "step": 1189
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.17278152704238892,
+      "learning_rate": 0.00019470968831288143,
+      "loss": 0.991,
+      "step": 1190
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.17254570126533508,
+      "learning_rate": 0.00019470079796224095,
+      "loss": 0.9979,
+      "step": 1191
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.16804873943328857,
+      "learning_rate": 0.00019469190035107886,
+      "loss": 1.0026,
+      "step": 1192
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1779593974351883,
+      "learning_rate": 0.00019468299548007728,
+      "loss": 1.0992,
+      "step": 1193
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.17888137698173523,
+      "learning_rate": 0.00019467408334991902,
+      "loss": 1.0435,
+      "step": 1194
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.17595483362674713,
+      "learning_rate": 0.00019466516396128726,
+      "loss": 0.9999,
+      "step": 1195
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.17406252026557922,
+      "learning_rate": 0.00019465623731486592,
+      "loss": 1.0375,
+      "step": 1196
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.18004122376441956,
+      "learning_rate": 0.0001946473034113393,
+      "loss": 1.0238,
+      "step": 1197
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.18256844580173492,
+      "learning_rate": 0.0001946383622513924,
+      "loss": 0.9845,
+      "step": 1198
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1867818981409073,
+      "learning_rate": 0.00019462941383571072,
+      "loss": 1.0887,
+      "step": 1199
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.17373467981815338,
+      "learning_rate": 0.0001946204581649803,
+      "loss": 0.9685,
+      "step": 1200
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.17235921323299408,
+      "learning_rate": 0.00019461149523988772,
+      "loss": 1.006,
+      "step": 1201
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.17519314587116241,
+      "learning_rate": 0.00019460252506112025,
+      "loss": 1.0133,
+      "step": 1202
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.17482595145702362,
+      "learning_rate": 0.00019459354762936553,
+      "loss": 1.0493,
+      "step": 1203
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.17520184814929962,
+      "learning_rate": 0.00019458456294531185,
+      "loss": 1.0206,
+      "step": 1204
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.17929650843143463,
+      "learning_rate": 0.0001945755710096481,
+      "loss": 1.0205,
+      "step": 1205
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3840513527393341,
+      "learning_rate": 0.00019456657182306358,
+      "loss": 0.9666,
+      "step": 1206
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.18337073922157288,
+      "learning_rate": 0.00019455756538624833,
+      "loss": 1.0845,
+      "step": 1207
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.17474819719791412,
+      "learning_rate": 0.00019454855169989284,
+      "loss": 1.048,
+      "step": 1208
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.17507345974445343,
+      "learning_rate": 0.00019453953076468814,
+      "loss": 1.0749,
+      "step": 1209
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.17816558480262756,
+      "learning_rate": 0.00019453050258132586,
+      "loss": 1.0486,
+      "step": 1210
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1680316925048828,
+      "learning_rate": 0.00019452146715049818,
+      "loss": 0.9808,
+      "step": 1211
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1678476631641388,
+      "learning_rate": 0.0001945124244728978,
+      "loss": 0.9723,
+      "step": 1212
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.17107614874839783,
+      "learning_rate": 0.00019450337454921806,
+      "loss": 0.9654,
+      "step": 1213
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.17758767306804657,
+      "learning_rate": 0.00019449431738015273,
+      "loss": 1.0573,
+      "step": 1214
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.176117405295372,
+      "learning_rate": 0.00019448525296639625,
+      "loss": 1.0669,
+      "step": 1215
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.17450761795043945,
+      "learning_rate": 0.00019447618130864357,
+      "loss": 0.9906,
+      "step": 1216
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1826401948928833,
+      "learning_rate": 0.0001944671024075902,
+      "loss": 1.0665,
+      "step": 1217
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.17291821539402008,
+      "learning_rate": 0.00019445801626393218,
+      "loss": 0.9476,
+      "step": 1218
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.16751059889793396,
+      "learning_rate": 0.00019444892287836613,
+      "loss": 1.016,
+      "step": 1219
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.17721164226531982,
+      "learning_rate": 0.00019443982225158924,
+      "loss": 1.0528,
+      "step": 1220
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.17266419529914856,
+      "learning_rate": 0.00019443071438429921,
+      "loss": 1.0224,
+      "step": 1221
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.17830513417720795,
+      "learning_rate": 0.00019442159927719434,
+      "loss": 1.0509,
+      "step": 1222
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.17815667390823364,
+      "learning_rate": 0.00019441247693097346,
+      "loss": 0.9926,
+      "step": 1223
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.17313189804553986,
+      "learning_rate": 0.000194403347346336,
+      "loss": 0.9687,
+      "step": 1224
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.18634116649627686,
+      "learning_rate": 0.00019439421052398182,
+      "loss": 1.0251,
+      "step": 1225
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1939675211906433,
+      "learning_rate": 0.0001943850664646115,
+      "loss": 1.047,
+      "step": 1226
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.17838935554027557,
+      "learning_rate": 0.00019437591516892607,
+      "loss": 1.0672,
+      "step": 1227
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.17744366824626923,
+      "learning_rate": 0.00019436675663762716,
+      "loss": 0.9857,
+      "step": 1228
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.16612915694713593,
+      "learning_rate": 0.0001943575908714169,
+      "loss": 0.9924,
+      "step": 1229
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1929885447025299,
+      "learning_rate": 0.00019434841787099803,
+      "loss": 1.0051,
+      "step": 1230
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.189718559384346,
+      "learning_rate": 0.00019433923763707383,
+      "loss": 1.1377,
+      "step": 1231
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.19058705866336823,
+      "learning_rate": 0.00019433005017034815,
+      "loss": 1.0702,
+      "step": 1232
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.17250509560108185,
+      "learning_rate": 0.00019432085547152532,
+      "loss": 1.0217,
+      "step": 1233
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.17606855928897858,
+      "learning_rate": 0.0001943116535413103,
+      "loss": 0.9705,
+      "step": 1234
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.17175276577472687,
+      "learning_rate": 0.00019430244438040862,
+      "loss": 1.0118,
+      "step": 1235
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.17503391206264496,
+      "learning_rate": 0.0001942932279895263,
+      "loss": 0.9583,
+      "step": 1236
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.17668573558330536,
+      "learning_rate": 0.00019428400436936994,
+      "loss": 1.0018,
+      "step": 1237
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.17263321578502655,
+      "learning_rate": 0.0001942747735206467,
+      "loss": 1.0072,
+      "step": 1238
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1817825585603714,
+      "learning_rate": 0.0001942655354440643,
+      "loss": 1.077,
+      "step": 1239
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.18519999086856842,
+      "learning_rate": 0.00019425629014033097,
+      "loss": 1.041,
+      "step": 1240
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1884918361902237,
+      "learning_rate": 0.0001942470376101556,
+      "loss": 0.971,
+      "step": 1241
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.17670756578445435,
+      "learning_rate": 0.00019423777785424748,
+      "loss": 1.0509,
+      "step": 1242
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.18296687304973602,
+      "learning_rate": 0.0001942285108733166,
+      "loss": 1.0273,
+      "step": 1243
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.17323486506938934,
+      "learning_rate": 0.0001942192366680734,
+      "loss": 1.0364,
+      "step": 1244
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.18001830577850342,
+      "learning_rate": 0.00019420995523922895,
+      "loss": 1.0587,
+      "step": 1245
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.18932732939720154,
+      "learning_rate": 0.00019420066658749481,
+      "loss": 1.0051,
+      "step": 1246
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.18397623300552368,
+      "learning_rate": 0.00019419137071358312,
+      "loss": 0.9917,
+      "step": 1247
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1875317096710205,
+      "learning_rate": 0.0001941820676182066,
+      "loss": 1.052,
+      "step": 1248
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.18426766991615295,
+      "learning_rate": 0.0001941727573020785,
+      "loss": 1.1202,
+      "step": 1249
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.181876078248024,
+      "learning_rate": 0.00019416343976591261,
+      "loss": 1.0529,
+      "step": 1250
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.17538008093833923,
+      "learning_rate": 0.00019415411501042326,
+      "loss": 0.9733,
+      "step": 1251
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.16881094872951508,
+      "learning_rate": 0.00019414478303632544,
+      "loss": 1.0014,
+      "step": 1252
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.17428913712501526,
+      "learning_rate": 0.00019413544384433453,
+      "loss": 1.0101,
+      "step": 1253
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.16802147030830383,
+      "learning_rate": 0.00019412609743516657,
+      "loss": 1.0394,
+      "step": 1254
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.18043246865272522,
+      "learning_rate": 0.00019411674380953817,
+      "loss": 0.997,
+      "step": 1255
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1805686354637146,
+      "learning_rate": 0.0001941073829681664,
+      "loss": 1.0549,
+      "step": 1256
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1995280236005783,
+      "learning_rate": 0.000194098014911769,
+      "loss": 1.0735,
+      "step": 1257
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.174880713224411,
+      "learning_rate": 0.00019408863964106413,
+      "loss": 0.9948,
+      "step": 1258
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.16793687641620636,
+      "learning_rate": 0.00019407925715677062,
+      "loss": 0.992,
+      "step": 1259
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.19079770147800446,
+      "learning_rate": 0.0001940698674596078,
+      "loss": 1.0717,
+      "step": 1260
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.18188640475273132,
+      "learning_rate": 0.00019406047055029553,
+      "loss": 1.0195,
+      "step": 1261
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.18434570729732513,
+      "learning_rate": 0.00019405106642955428,
+      "loss": 1.05,
+      "step": 1262
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.18857026100158691,
+      "learning_rate": 0.00019404165509810504,
+      "loss": 1.0704,
+      "step": 1263
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.17161917686462402,
+      "learning_rate": 0.00019403223655666937,
+      "loss": 1.0103,
+      "step": 1264
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.17019355297088623,
+      "learning_rate": 0.00019402281080596934,
+      "loss": 0.9766,
+      "step": 1265
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1757907271385193,
+      "learning_rate": 0.0001940133778467276,
+      "loss": 1.0475,
+      "step": 1266
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.183481827378273,
+      "learning_rate": 0.00019400393767966744,
+      "loss": 1.0011,
+      "step": 1267
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1849091649055481,
+      "learning_rate": 0.0001939944903055125,
+      "loss": 1.029,
+      "step": 1268
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.17956006526947021,
+      "learning_rate": 0.00019398503572498718,
+      "loss": 1.0034,
+      "step": 1269
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.172621950507164,
+      "learning_rate": 0.00019397557393881633,
+      "loss": 1.0086,
+      "step": 1270
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.18043522536754608,
+      "learning_rate": 0.00019396610494772532,
+      "loss": 1.025,
+      "step": 1271
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.17680267989635468,
+      "learning_rate": 0.00019395662875244015,
+      "loss": 1.042,
+      "step": 1272
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.17402620613574982,
+      "learning_rate": 0.00019394714535368732,
+      "loss": 0.9135,
+      "step": 1273
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1817319095134735,
+      "learning_rate": 0.00019393765475219393,
+      "loss": 1.0469,
+      "step": 1274
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.18868154287338257,
+      "learning_rate": 0.0001939281569486876,
+      "loss": 1.0571,
+      "step": 1275
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.16820600628852844,
+      "learning_rate": 0.0001939186519438965,
+      "loss": 1.0429,
+      "step": 1276
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1712152361869812,
+      "learning_rate": 0.00019390913973854938,
+      "loss": 0.9995,
+      "step": 1277
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1772497594356537,
+      "learning_rate": 0.00019389962033337552,
+      "loss": 0.9795,
+      "step": 1278
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.17727500200271606,
+      "learning_rate": 0.0001938900937291047,
+      "loss": 0.9998,
+      "step": 1279
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.17319205403327942,
+      "learning_rate": 0.00019388055992646737,
+      "loss": 0.9909,
+      "step": 1280
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1945336014032364,
+      "learning_rate": 0.00019387101892619443,
+      "loss": 1.0173,
+      "step": 1281
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.18670299649238586,
+      "learning_rate": 0.0001938614707290174,
+      "loss": 1.023,
+      "step": 1282
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.17243389785289764,
+      "learning_rate": 0.00019385191533566828,
+      "loss": 1.0632,
+      "step": 1283
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.19126373529434204,
+      "learning_rate": 0.0001938423527468797,
+      "loss": 0.9837,
+      "step": 1284
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.19175255298614502,
+      "learning_rate": 0.0001938327829633848,
+      "loss": 1.0314,
+      "step": 1285
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1695779263973236,
+      "learning_rate": 0.00019382320598591727,
+      "loss": 1.0078,
+      "step": 1286
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.17427422106266022,
+      "learning_rate": 0.00019381362181521134,
+      "loss": 0.9935,
+      "step": 1287
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.18275946378707886,
+      "learning_rate": 0.00019380403045200182,
+      "loss": 1.0181,
+      "step": 1288
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.18247005343437195,
+      "learning_rate": 0.0001937944318970241,
+      "loss": 1.0173,
+      "step": 1289
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1783648580312729,
+      "learning_rate": 0.00019378482615101401,
+      "loss": 1.027,
+      "step": 1290
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.18437276780605316,
+      "learning_rate": 0.00019377521321470805,
+      "loss": 1.0709,
+      "step": 1291
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1733713001012802,
+      "learning_rate": 0.00019376559308884324,
+      "loss": 1.0126,
+      "step": 1292
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.17454306781291962,
+      "learning_rate": 0.0001937559657741571,
+      "loss": 1.0644,
+      "step": 1293
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.18214353919029236,
+      "learning_rate": 0.00019374633127138772,
+      "loss": 0.9976,
+      "step": 1294
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1960800439119339,
+      "learning_rate": 0.00019373668958127383,
+      "loss": 1.0927,
+      "step": 1295
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.18307611346244812,
+      "learning_rate": 0.00019372704070455456,
+      "loss": 1.0643,
+      "step": 1296
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.17762793600559235,
+      "learning_rate": 0.0001937173846419697,
+      "loss": 0.9654,
+      "step": 1297
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1836734414100647,
+      "learning_rate": 0.00019370772139425962,
+      "loss": 0.9957,
+      "step": 1298
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.17774513363838196,
+      "learning_rate": 0.00019369805096216509,
+      "loss": 1.0072,
+      "step": 1299
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1814611703157425,
+      "learning_rate": 0.00019368837334642758,
+      "loss": 1.0244,
+      "step": 1300
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4324641227722168,
+      "learning_rate": 0.00019367868854778904,
+      "loss": 1.035,
+      "step": 1301
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1968534290790558,
+      "learning_rate": 0.00019366899656699196,
+      "loss": 0.9609,
+      "step": 1302
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.18885602056980133,
+      "learning_rate": 0.0001936592974047794,
+      "loss": 1.0719,
+      "step": 1303
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1806947886943817,
+      "learning_rate": 0.00019364959106189504,
+      "loss": 1.0308,
+      "step": 1304
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1710716187953949,
+      "learning_rate": 0.000193639877539083,
+      "loss": 0.999,
+      "step": 1305
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.17838624119758606,
+      "learning_rate": 0.000193630156837088,
+      "loss": 0.9507,
+      "step": 1306
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.17988012731075287,
+      "learning_rate": 0.0001936204289566553,
+      "loss": 1.0071,
+      "step": 1307
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1811203956604004,
+      "learning_rate": 0.00019361069389853074,
+      "loss": 1.0272,
+      "step": 1308
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1934630423784256,
+      "learning_rate": 0.00019360095166346067,
+      "loss": 1.0003,
+      "step": 1309
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1737378090620041,
+      "learning_rate": 0.00019359120225219203,
+      "loss": 1.0585,
+      "step": 1310
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2015053778886795,
+      "learning_rate": 0.00019358144566547226,
+      "loss": 1.0141,
+      "step": 1311
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1798975020647049,
+      "learning_rate": 0.00019357168190404936,
+      "loss": 1.0133,
+      "step": 1312
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.18216365575790405,
+      "learning_rate": 0.00019356191096867197,
+      "loss": 1.0467,
+      "step": 1313
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1877415031194687,
+      "learning_rate": 0.00019355213286008916,
+      "loss": 1.081,
+      "step": 1314
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.18899911642074585,
+      "learning_rate": 0.00019354234757905055,
+      "loss": 1.0421,
+      "step": 1315
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.18335410952568054,
+      "learning_rate": 0.00019353255512630647,
+      "loss": 0.9673,
+      "step": 1316
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1879153847694397,
+      "learning_rate": 0.00019352275550260759,
+      "loss": 1.028,
+      "step": 1317
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.18537075817584991,
+      "learning_rate": 0.00019351294870870528,
+      "loss": 0.984,
+      "step": 1318
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.18217268586158752,
+      "learning_rate": 0.0001935031347453514,
+      "loss": 1.0041,
+      "step": 1319
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1934812366962433,
+      "learning_rate": 0.00019349331361329835,
+      "loss": 1.0278,
+      "step": 1320
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.19314013421535492,
+      "learning_rate": 0.0001934834853132991,
+      "loss": 1.0594,
+      "step": 1321
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.18223635852336884,
+      "learning_rate": 0.00019347364984610718,
+      "loss": 1.0376,
+      "step": 1322
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1895359456539154,
+      "learning_rate": 0.00019346380721247669,
+      "loss": 1.0477,
+      "step": 1323
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.19515518844127655,
+      "learning_rate": 0.00019345395741316216,
+      "loss": 1.0081,
+      "step": 1324
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.17444954812526703,
+      "learning_rate": 0.0001934441004489188,
+      "loss": 0.9642,
+      "step": 1325
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1794186234474182,
+      "learning_rate": 0.00019343423632050232,
+      "loss": 0.952,
+      "step": 1326
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1817828118801117,
+      "learning_rate": 0.00019342436502866898,
+      "loss": 1.0201,
+      "step": 1327
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.18109366297721863,
+      "learning_rate": 0.0001934144865741756,
+      "loss": 1.0059,
+      "step": 1328
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.18579837679862976,
+      "learning_rate": 0.00019340460095777957,
+      "loss": 1.0449,
+      "step": 1329
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.17582041025161743,
+      "learning_rate": 0.00019339470818023876,
+      "loss": 0.9718,
+      "step": 1330
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1840338110923767,
+      "learning_rate": 0.00019338480824231162,
+      "loss": 1.0338,
+      "step": 1331
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.18756501376628876,
+      "learning_rate": 0.00019337490114475719,
+      "loss": 1.0245,
+      "step": 1332
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1740320324897766,
+      "learning_rate": 0.000193364986888335,
+      "loss": 0.9619,
+      "step": 1333
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.18331877887248993,
+      "learning_rate": 0.00019335506547380518,
+      "loss": 0.9684,
+      "step": 1334
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.18273407220840454,
+      "learning_rate": 0.0001933451369019284,
+      "loss": 1.0624,
+      "step": 1335
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.17652833461761475,
+      "learning_rate": 0.0001933352011734658,
+      "loss": 0.9889,
+      "step": 1336
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.18905626237392426,
+      "learning_rate": 0.0001933252582891792,
+      "loss": 1.0827,
+      "step": 1337
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.17634400725364685,
+      "learning_rate": 0.00019331530824983085,
+      "loss": 0.9897,
+      "step": 1338
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.18231314420700073,
+      "learning_rate": 0.00019330535105618367,
+      "loss": 1.0558,
+      "step": 1339
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.19120241701602936,
+      "learning_rate": 0.00019329538670900096,
+      "loss": 1.027,
+      "step": 1340
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1788601577281952,
+      "learning_rate": 0.00019328541520904677,
+      "loss": 1.0448,
+      "step": 1341
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1818162500858307,
+      "learning_rate": 0.0001932754365570855,
+      "loss": 1.0101,
+      "step": 1342
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.193072110414505,
+      "learning_rate": 0.00019326545075388225,
+      "loss": 1.1396,
+      "step": 1343
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1790308952331543,
+      "learning_rate": 0.0001932554578002026,
+      "loss": 0.9894,
+      "step": 1344
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.19239982962608337,
+      "learning_rate": 0.0001932454576968127,
+      "loss": 1.0331,
+      "step": 1345
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1802983582019806,
+      "learning_rate": 0.0001932354504444792,
+      "loss": 1.0169,
+      "step": 1346
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.18328198790550232,
+      "learning_rate": 0.0001932254360439694,
+      "loss": 1.0175,
+      "step": 1347
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.17947252094745636,
+      "learning_rate": 0.000193215414496051,
+      "loss": 1.0286,
+      "step": 1348
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.18461839854717255,
+      "learning_rate": 0.00019320538580149243,
+      "loss": 1.0737,
+      "step": 1349
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.18745096027851105,
+      "learning_rate": 0.0001931953499610625,
+      "loss": 1.011,
+      "step": 1350
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1850559115409851,
+      "learning_rate": 0.00019318530697553062,
+      "loss": 1.0351,
+      "step": 1351
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.17920838296413422,
+      "learning_rate": 0.00019317525684566685,
+      "loss": 0.9879,
+      "step": 1352
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.18742653727531433,
+      "learning_rate": 0.00019316519957224165,
+      "loss": 0.9737,
+      "step": 1353
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.18532788753509521,
+      "learning_rate": 0.00019315513515602613,
+      "loss": 0.9625,
+      "step": 1354
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.17812992632389069,
+      "learning_rate": 0.00019314506359779186,
+      "loss": 0.9914,
+      "step": 1355
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.17627066373825073,
+      "learning_rate": 0.00019313498489831102,
+      "loss": 1.0321,
+      "step": 1356
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.18436536192893982,
+      "learning_rate": 0.00019312489905835636,
+      "loss": 1.0064,
+      "step": 1357
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.18245936930179596,
+      "learning_rate": 0.00019311480607870108,
+      "loss": 1.0026,
+      "step": 1358
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1713896244764328,
+      "learning_rate": 0.00019310470596011907,
+      "loss": 0.942,
+      "step": 1359
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.17367751896381378,
+      "learning_rate": 0.00019309459870338463,
+      "loss": 0.9872,
+      "step": 1360
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1733555942773819,
+      "learning_rate": 0.0001930844843092727,
+      "loss": 1.0109,
+      "step": 1361
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.17760616540908813,
+      "learning_rate": 0.00019307436277855866,
+      "loss": 1.0196,
+      "step": 1362
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1863793581724167,
+      "learning_rate": 0.00019306423411201857,
+      "loss": 1.05,
+      "step": 1363
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1735723912715912,
+      "learning_rate": 0.00019305409831042898,
+      "loss": 1.0202,
+      "step": 1364
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.18256759643554688,
+      "learning_rate": 0.00019304395537456692,
+      "loss": 1.0575,
+      "step": 1365
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1789254993200302,
+      "learning_rate": 0.00019303380530521008,
+      "loss": 1.0937,
+      "step": 1366
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.17587143182754517,
+      "learning_rate": 0.00019302364810313668,
+      "loss": 1.0526,
+      "step": 1367
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.17091909050941467,
+      "learning_rate": 0.00019301348376912536,
+      "loss": 0.9902,
+      "step": 1368
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.18148952722549438,
+      "learning_rate": 0.00019300331230395545,
+      "loss": 1.029,
+      "step": 1369
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1739834100008011,
+      "learning_rate": 0.00019299313370840676,
+      "loss": 1.0009,
+      "step": 1370
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.18367813527584076,
+      "learning_rate": 0.0001929829479832597,
+      "loss": 1.0411,
+      "step": 1371
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.17575769126415253,
+      "learning_rate": 0.00019297275512929516,
+      "loss": 1.0155,
+      "step": 1372
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1805105209350586,
+      "learning_rate": 0.0001929625551472946,
+      "loss": 1.0153,
+      "step": 1373
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.17204175889492035,
+      "learning_rate": 0.00019295234803804004,
+      "loss": 0.9689,
+      "step": 1374
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1902618110179901,
+      "learning_rate": 0.00019294213380231402,
+      "loss": 0.9915,
+      "step": 1375
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.17897525429725647,
+      "learning_rate": 0.00019293191244089972,
+      "loss": 0.9946,
+      "step": 1376
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.17735904455184937,
+      "learning_rate": 0.0001929216839545807,
+      "loss": 0.963,
+      "step": 1377
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1746383160352707,
+      "learning_rate": 0.0001929114483441412,
+      "loss": 0.9751,
+      "step": 1378
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.17372174561023712,
+      "learning_rate": 0.000192901205610366,
+      "loss": 1.0304,
+      "step": 1379
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.17814914882183075,
+      "learning_rate": 0.0001928909557540403,
+      "loss": 0.9754,
+      "step": 1380
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.17424102127552032,
+      "learning_rate": 0.00019288069877594998,
+      "loss": 1.0329,
+      "step": 1381
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.17923283576965332,
+      "learning_rate": 0.00019287043467688145,
+      "loss": 1.0678,
+      "step": 1382
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.17741328477859497,
+      "learning_rate": 0.00019286016345762162,
+      "loss": 0.9972,
+      "step": 1383
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1734980046749115,
+      "learning_rate": 0.00019284988511895793,
+      "loss": 1.0207,
+      "step": 1384
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.17344945669174194,
+      "learning_rate": 0.00019283959966167843,
+      "loss": 0.98,
+      "step": 1385
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.18832659721374512,
+      "learning_rate": 0.00019282930708657169,
+      "loss": 1.0533,
+      "step": 1386
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.18710127472877502,
+      "learning_rate": 0.00019281900739442684,
+      "loss": 1.0172,
+      "step": 1387
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.18463793396949768,
+      "learning_rate": 0.0001928087005860335,
+      "loss": 1.0912,
+      "step": 1388
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1741846352815628,
+      "learning_rate": 0.00019279838666218187,
+      "loss": 0.9459,
+      "step": 1389
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.18016698956489563,
+      "learning_rate": 0.00019278806562366275,
+      "loss": 1.0024,
+      "step": 1390
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.18188920617103577,
+      "learning_rate": 0.00019277773747126736,
+      "loss": 1.0591,
+      "step": 1391
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1762164831161499,
+      "learning_rate": 0.00019276740220578758,
+      "loss": 1.0583,
+      "step": 1392
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.17220334708690643,
+      "learning_rate": 0.00019275705982801578,
+      "loss": 1.0282,
+      "step": 1393
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1795509308576584,
+      "learning_rate": 0.00019274671033874492,
+      "loss": 1.0277,
+      "step": 1394
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.17990824580192566,
+      "learning_rate": 0.00019273635373876844,
+      "loss": 1.0421,
+      "step": 1395
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.18492265045642853,
+      "learning_rate": 0.0001927259900288804,
+      "loss": 0.9711,
+      "step": 1396
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.18191231787204742,
+      "learning_rate": 0.0001927156192098753,
+      "loss": 1.0157,
+      "step": 1397
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1850530356168747,
+      "learning_rate": 0.0001927052412825483,
+      "loss": 1.0406,
+      "step": 1398
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.19039122760295868,
+      "learning_rate": 0.00019269485624769507,
+      "loss": 0.9688,
+      "step": 1399
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.17096638679504395,
+      "learning_rate": 0.00019268446410611176,
+      "loss": 1.0016,
+      "step": 1400
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.19254373013973236,
+      "learning_rate": 0.00019267406485859513,
+      "loss": 1.0246,
+      "step": 1401
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1767900288105011,
+      "learning_rate": 0.0001926636585059425,
+      "loss": 1.0083,
+      "step": 1402
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.18407025933265686,
+      "learning_rate": 0.00019265324504895168,
+      "loss": 1.012,
+      "step": 1403
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.18007557094097137,
+      "learning_rate": 0.00019264282448842103,
+      "loss": 1.0839,
+      "step": 1404
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.18103165924549103,
+      "learning_rate": 0.00019263239682514952,
+      "loss": 1.0383,
+      "step": 1405
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.18077121675014496,
+      "learning_rate": 0.00019262196205993657,
+      "loss": 1.0107,
+      "step": 1406
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.17406535148620605,
+      "learning_rate": 0.00019261152019358222,
+      "loss": 1.0433,
+      "step": 1407
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1833508461713791,
+      "learning_rate": 0.00019260107122688705,
+      "loss": 1.0356,
+      "step": 1408
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.17652858793735504,
+      "learning_rate": 0.0001925906151606521,
+      "loss": 0.9974,
+      "step": 1409
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.17765457928180695,
+      "learning_rate": 0.00019258015199567904,
+      "loss": 0.9817,
+      "step": 1410
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1752379983663559,
+      "learning_rate": 0.0001925696817327701,
+      "loss": 0.9987,
+      "step": 1411
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.17794077098369598,
+      "learning_rate": 0.000192559204372728,
+      "loss": 0.9804,
+      "step": 1412
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1826150268316269,
+      "learning_rate": 0.00019254871991635598,
+      "loss": 1.0078,
+      "step": 1413
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.18269848823547363,
+      "learning_rate": 0.00019253822836445787,
+      "loss": 0.99,
+      "step": 1414
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.19174599647521973,
+      "learning_rate": 0.00019252772971783806,
+      "loss": 1.0228,
+      "step": 1415
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1771102249622345,
+      "learning_rate": 0.00019251722397730148,
+      "loss": 0.9811,
+      "step": 1416
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.17833948135375977,
+      "learning_rate": 0.00019250671114365352,
+      "loss": 1.0279,
+      "step": 1417
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1791437864303589,
+      "learning_rate": 0.0001924961912177002,
+      "loss": 1.0045,
+      "step": 1418
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1782650500535965,
+      "learning_rate": 0.00019248566420024813,
+      "loss": 0.962,
+      "step": 1419
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.17893794178962708,
+      "learning_rate": 0.00019247513009210433,
+      "loss": 1.0527,
+      "step": 1420
+    },
+    {
+      "epoch": 0.5,
+      "eval_loss": 1.0163156986236572,
+      "eval_runtime": 680.0523,
+      "eval_samples_per_second": 10.111,
+      "eval_steps_per_second": 5.055,
+      "step": 1420
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1815989464521408,
+      "learning_rate": 0.00019246458889407643,
+      "loss": 1.0524,
+      "step": 1421
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.17999018728733063,
+      "learning_rate": 0.0001924540406069726,
+      "loss": 0.9832,
+      "step": 1422
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.18894697725772858,
+      "learning_rate": 0.00019244348523160155,
+      "loss": 1.0769,
+      "step": 1423
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1834978312253952,
+      "learning_rate": 0.00019243292276877258,
+      "loss": 1.0129,
+      "step": 1424
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1804187297821045,
+      "learning_rate": 0.00019242235321929548,
+      "loss": 1.0363,
+      "step": 1425
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.17544397711753845,
+      "learning_rate": 0.0001924117765839806,
+      "loss": 0.9883,
+      "step": 1426
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.18312135338783264,
+      "learning_rate": 0.0001924011928636388,
+      "loss": 0.9859,
+      "step": 1427
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.17146514356136322,
+      "learning_rate": 0.00019239060205908154,
+      "loss": 0.9585,
+      "step": 1428
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.18334969878196716,
+      "learning_rate": 0.00019238000417112078,
+      "loss": 1.0252,
+      "step": 1429
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.180964857339859,
+      "learning_rate": 0.00019236939920056902,
+      "loss": 1.0354,
+      "step": 1430
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.19952955842018127,
+      "learning_rate": 0.0001923587871482394,
+      "loss": 1.0261,
+      "step": 1431
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.18565645813941956,
+      "learning_rate": 0.00019234816801494542,
+      "loss": 0.9784,
+      "step": 1432
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.18717040121555328,
+      "learning_rate": 0.00019233754180150129,
+      "loss": 0.9832,
+      "step": 1433
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.18185356259346008,
+      "learning_rate": 0.00019232690850872172,
+      "loss": 1.021,
+      "step": 1434
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.17485785484313965,
+      "learning_rate": 0.0001923162681374219,
+      "loss": 0.9975,
+      "step": 1435
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.18476179242134094,
+      "learning_rate": 0.0001923056206884176,
+      "loss": 0.9592,
+      "step": 1436
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.18408514559268951,
+      "learning_rate": 0.0001922949661625252,
+      "loss": 1.0284,
+      "step": 1437
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1944282501935959,
+      "learning_rate": 0.00019228430456056148,
+      "loss": 1.0313,
+      "step": 1438
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1734023094177246,
+      "learning_rate": 0.0001922736358833439,
+      "loss": 1.0098,
+      "step": 1439
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.18978969752788544,
+      "learning_rate": 0.00019226296013169042,
+      "loss": 0.9933,
+      "step": 1440
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.18699029088020325,
+      "learning_rate": 0.00019225227730641943,
+      "loss": 1.0292,
+      "step": 1441
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.17445528507232666,
+      "learning_rate": 0.00019224158740835007,
+      "loss": 1.0196,
+      "step": 1442
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.17488212883472443,
+      "learning_rate": 0.0001922308904383019,
+      "loss": 1.0401,
+      "step": 1443
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1726742386817932,
+      "learning_rate": 0.00019222018639709497,
+      "loss": 0.9805,
+      "step": 1444
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.17496153712272644,
+      "learning_rate": 0.00019220947528554997,
+      "loss": 0.9819,
+      "step": 1445
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.18209873139858246,
+      "learning_rate": 0.00019219875710448813,
+      "loss": 1.0321,
+      "step": 1446
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.17869645357131958,
+      "learning_rate": 0.00019218803185473116,
+      "loss": 0.941,
+      "step": 1447
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.19009461998939514,
+      "learning_rate": 0.00019217729953710134,
+      "loss": 1.0568,
+      "step": 1448
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1776958853006363,
+      "learning_rate": 0.0001921665601524215,
+      "loss": 1.0039,
+      "step": 1449
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.18249039351940155,
+      "learning_rate": 0.00019215581370151505,
+      "loss": 0.9773,
+      "step": 1450
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.19207897782325745,
+      "learning_rate": 0.00019214506018520582,
+      "loss": 1.0134,
+      "step": 1451
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.18842819333076477,
+      "learning_rate": 0.00019213429960431832,
+      "loss": 1.0259,
+      "step": 1452
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.18662334978580475,
+      "learning_rate": 0.00019212353195967752,
+      "loss": 0.9814,
+      "step": 1453
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.18366405367851257,
+      "learning_rate": 0.00019211275725210898,
+      "loss": 1.0715,
+      "step": 1454
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.18426047265529633,
+      "learning_rate": 0.00019210197548243874,
+      "loss": 1.0182,
+      "step": 1455
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.17692065238952637,
+      "learning_rate": 0.00019209118665149344,
+      "loss": 0.9778,
+      "step": 1456
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1945948749780655,
+      "learning_rate": 0.0001920803907601002,
+      "loss": 1.0043,
+      "step": 1457
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.17809897661209106,
+      "learning_rate": 0.00019206958780908679,
+      "loss": 1.0542,
+      "step": 1458
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.18257014453411102,
+      "learning_rate": 0.00019205877779928136,
+      "loss": 1.0037,
+      "step": 1459
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1871815174818039,
+      "learning_rate": 0.00019204796073151278,
+      "loss": 1.0409,
+      "step": 1460
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.17258718609809875,
+      "learning_rate": 0.00019203713660661032,
+      "loss": 0.9901,
+      "step": 1461
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1797480434179306,
+      "learning_rate": 0.00019202630542540386,
+      "loss": 1.0053,
+      "step": 1462
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.18446534872055054,
+      "learning_rate": 0.00019201546718872382,
+      "loss": 1.0494,
+      "step": 1463
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1935523897409439,
+      "learning_rate": 0.0001920046218974011,
+      "loss": 1.0355,
+      "step": 1464
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1764344573020935,
+      "learning_rate": 0.00019199376955226724,
+      "loss": 0.9837,
+      "step": 1465
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.17583085596561432,
+      "learning_rate": 0.00019198291015415424,
+      "loss": 1.0092,
+      "step": 1466
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.18314938247203827,
+      "learning_rate": 0.00019197204370389467,
+      "loss": 1.0004,
+      "step": 1467
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1774096041917801,
+      "learning_rate": 0.00019196117020232165,
+      "loss": 0.9551,
+      "step": 1468
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.18138380348682404,
+      "learning_rate": 0.00019195028965026883,
+      "loss": 0.996,
+      "step": 1469
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1724422425031662,
+      "learning_rate": 0.00019193940204857035,
+      "loss": 0.9521,
+      "step": 1470
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.18512949347496033,
+      "learning_rate": 0.000191928507398061,
+      "loss": 1.0161,
+      "step": 1471
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.17651984095573425,
+      "learning_rate": 0.00019191760569957605,
+      "loss": 1.0599,
+      "step": 1472
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1797732263803482,
+      "learning_rate": 0.00019190669695395127,
+      "loss": 1.0212,
+      "step": 1473
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.18342822790145874,
+      "learning_rate": 0.00019189578116202307,
+      "loss": 1.0642,
+      "step": 1474
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.17123663425445557,
+      "learning_rate": 0.0001918848583246283,
+      "loss": 1.0349,
+      "step": 1475
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.17873786389827728,
+      "learning_rate": 0.00019187392844260438,
+      "loss": 0.994,
+      "step": 1476
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.18375077843666077,
+      "learning_rate": 0.00019186299151678933,
+      "loss": 0.9608,
+      "step": 1477
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1748666614294052,
+      "learning_rate": 0.00019185204754802162,
+      "loss": 0.9826,
+      "step": 1478
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.17497992515563965,
+      "learning_rate": 0.0001918410965371403,
+      "loss": 0.9611,
+      "step": 1479
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.17814147472381592,
+      "learning_rate": 0.00019183013848498505,
+      "loss": 0.9753,
+      "step": 1480
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.18198871612548828,
+      "learning_rate": 0.00019181917339239587,
+      "loss": 1.0253,
+      "step": 1481
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1947755515575409,
+      "learning_rate": 0.0001918082012602135,
+      "loss": 1.0288,
+      "step": 1482
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1716606318950653,
+      "learning_rate": 0.00019179722208927916,
+      "loss": 1.0063,
+      "step": 1483
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.17634867131710052,
+      "learning_rate": 0.0001917862358804346,
+      "loss": 0.9734,
+      "step": 1484
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.16377978026866913,
+      "learning_rate": 0.00019177524263452208,
+      "loss": 0.8907,
+      "step": 1485
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1761971414089203,
+      "learning_rate": 0.00019176424235238445,
+      "loss": 0.9301,
+      "step": 1486
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1791556477546692,
+      "learning_rate": 0.00019175323503486507,
+      "loss": 1.0068,
+      "step": 1487
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.18452045321464539,
+      "learning_rate": 0.00019174222068280784,
+      "loss": 1.0406,
+      "step": 1488
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.17810030281543732,
+      "learning_rate": 0.00019173119929705726,
+      "loss": 1.0032,
+      "step": 1489
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.18171797692775726,
+      "learning_rate": 0.00019172017087845826,
+      "loss": 1.0213,
+      "step": 1490
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.19251057505607605,
+      "learning_rate": 0.0001917091354278564,
+      "loss": 1.0263,
+      "step": 1491
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1829746812582016,
+      "learning_rate": 0.00019169809294609772,
+      "loss": 1.0505,
+      "step": 1492
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1797158122062683,
+      "learning_rate": 0.00019168704343402887,
+      "loss": 1.0091,
+      "step": 1493
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1812933087348938,
+      "learning_rate": 0.00019167598689249696,
+      "loss": 1.0804,
+      "step": 1494
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.17717424035072327,
+      "learning_rate": 0.00019166492332234968,
+      "loss": 1.017,
+      "step": 1495
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1779852956533432,
+      "learning_rate": 0.00019165385272443524,
+      "loss": 0.9392,
+      "step": 1496
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.18181325495243073,
+      "learning_rate": 0.0001916427750996024,
+      "loss": 0.9957,
+      "step": 1497
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.18540653586387634,
+      "learning_rate": 0.0001916316904487005,
+      "loss": 0.9936,
+      "step": 1498
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1858024299144745,
+      "learning_rate": 0.00019162059877257933,
+      "loss": 0.9896,
+      "step": 1499
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.187087744474411,
+      "learning_rate": 0.00019160950007208926,
+      "loss": 1.0209,
+      "step": 1500
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.18074601888656616,
+      "learning_rate": 0.00019159839434808125,
+      "loss": 0.9911,
+      "step": 1501
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.18445174396038055,
+      "learning_rate": 0.00019158728160140675,
+      "loss": 1.0246,
+      "step": 1502
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.18151141703128815,
+      "learning_rate": 0.00019157616183291772,
+      "loss": 1.0255,
+      "step": 1503
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1752597987651825,
+      "learning_rate": 0.00019156503504346673,
+      "loss": 0.9441,
+      "step": 1504
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1818474382162094,
+      "learning_rate": 0.0001915539012339068,
+      "loss": 0.9716,
+      "step": 1505
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.18579250574111938,
+      "learning_rate": 0.00019154276040509157,
+      "loss": 1.0682,
+      "step": 1506
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.18655996024608612,
+      "learning_rate": 0.00019153161255787518,
+      "loss": 0.9676,
+      "step": 1507
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1754051148891449,
+      "learning_rate": 0.00019152045769311232,
+      "loss": 0.9886,
+      "step": 1508
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1764054149389267,
+      "learning_rate": 0.00019150929581165818,
+      "loss": 0.983,
+      "step": 1509
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.17365694046020508,
+      "learning_rate": 0.00019149812691436854,
+      "loss": 0.9749,
+      "step": 1510
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.19486962258815765,
+      "learning_rate": 0.0001914869510020997,
+      "loss": 1.0016,
+      "step": 1511
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.17791058123111725,
+      "learning_rate": 0.00019147576807570854,
+      "loss": 1.0725,
+      "step": 1512
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.18877653777599335,
+      "learning_rate": 0.00019146457813605232,
+      "loss": 1.0221,
+      "step": 1513
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.18828843533992767,
+      "learning_rate": 0.00019145338118398906,
+      "loss": 1.0174,
+      "step": 1514
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.18070997297763824,
+      "learning_rate": 0.00019144217722037713,
+      "loss": 1.026,
+      "step": 1515
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.18031719326972961,
+      "learning_rate": 0.00019143096624607558,
+      "loss": 1.0322,
+      "step": 1516
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.18009379506111145,
+      "learning_rate": 0.00019141974826194388,
+      "loss": 1.0786,
+      "step": 1517
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1836472451686859,
+      "learning_rate": 0.00019140852326884216,
+      "loss": 1.0363,
+      "step": 1518
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1833399087190628,
+      "learning_rate": 0.00019139729126763092,
+      "loss": 0.9989,
+      "step": 1519
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.17795605957508087,
+      "learning_rate": 0.00019138605225917137,
+      "loss": 1.0266,
+      "step": 1520
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.18254658579826355,
+      "learning_rate": 0.00019137480624432517,
+      "loss": 1.0618,
+      "step": 1521
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.18196122348308563,
+      "learning_rate": 0.0001913635532239545,
+      "loss": 0.9893,
+      "step": 1522
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.18685205280780792,
+      "learning_rate": 0.0001913522931989221,
+      "loss": 1.0147,
+      "step": 1523
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.18355733156204224,
+      "learning_rate": 0.00019134102617009133,
+      "loss": 0.9925,
+      "step": 1524
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.17968907952308655,
+      "learning_rate": 0.00019132975213832594,
+      "loss": 1.0555,
+      "step": 1525
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.17847761511802673,
+      "learning_rate": 0.0001913184711044903,
+      "loss": 1.0557,
+      "step": 1526
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.17835336923599243,
+      "learning_rate": 0.00019130718306944932,
+      "loss": 0.9833,
+      "step": 1527
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1996411383152008,
+      "learning_rate": 0.00019129588803406842,
+      "loss": 1.0438,
+      "step": 1528
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.17643500864505768,
+      "learning_rate": 0.00019128458599921357,
+      "loss": 1.0064,
+      "step": 1529
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.18694715201854706,
+      "learning_rate": 0.00019127327696575125,
+      "loss": 1.001,
+      "step": 1530
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.18230758607387543,
+      "learning_rate": 0.0001912619609345486,
+      "loss": 1.062,
+      "step": 1531
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.17499038577079773,
+      "learning_rate": 0.00019125063790647305,
+      "loss": 0.9381,
+      "step": 1532
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.17816944420337677,
+      "learning_rate": 0.00019123930788239281,
+      "loss": 1.0399,
+      "step": 1533
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.19370074570178986,
+      "learning_rate": 0.00019122797086317653,
+      "loss": 0.9949,
+      "step": 1534
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1971571147441864,
+      "learning_rate": 0.00019121662684969335,
+      "loss": 0.9631,
+      "step": 1535
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.18476071953773499,
+      "learning_rate": 0.00019120527584281304,
+      "loss": 1.0312,
+      "step": 1536
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.18056322634220123,
+      "learning_rate": 0.0001911939178434058,
+      "loss": 0.9848,
+      "step": 1537
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.19656671583652496,
+      "learning_rate": 0.0001911825528523425,
+      "loss": 1.0728,
+      "step": 1538
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.19638994336128235,
+      "learning_rate": 0.00019117118087049438,
+      "loss": 0.9967,
+      "step": 1539
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.19183266162872314,
+      "learning_rate": 0.00019115980189873341,
+      "loss": 0.9916,
+      "step": 1540
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1767280101776123,
+      "learning_rate": 0.00019114841593793194,
+      "loss": 1.0588,
+      "step": 1541
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1942903995513916,
+      "learning_rate": 0.0001911370229889629,
+      "loss": 0.943,
+      "step": 1542
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1888478547334671,
+      "learning_rate": 0.0001911256230526998,
+      "loss": 1.0352,
+      "step": 1543
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.18362468481063843,
+      "learning_rate": 0.00019111421613001662,
+      "loss": 0.9774,
+      "step": 1544
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.17923595011234283,
+      "learning_rate": 0.0001911028022217879,
+      "loss": 0.9808,
+      "step": 1545
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.19007225334644318,
+      "learning_rate": 0.00019109138132888872,
+      "loss": 1.002,
+      "step": 1546
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.17761003971099854,
+      "learning_rate": 0.00019107995345219475,
+      "loss": 0.9513,
+      "step": 1547
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.18594999611377716,
+      "learning_rate": 0.0001910685185925821,
+      "loss": 1.0227,
+      "step": 1548
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1720389872789383,
+      "learning_rate": 0.00019105707675092743,
+      "loss": 0.9558,
+      "step": 1549
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.18154871463775635,
+      "learning_rate": 0.000191045627928108,
+      "loss": 0.995,
+      "step": 1550
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.19383765757083893,
+      "learning_rate": 0.00019103417212500162,
+      "loss": 1.0508,
+      "step": 1551
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.18130627274513245,
+      "learning_rate": 0.00019102270934248647,
+      "loss": 1.0096,
+      "step": 1552
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.18290060758590698,
+      "learning_rate": 0.00019101123958144146,
+      "loss": 1.0364,
+      "step": 1553
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.19369938969612122,
+      "learning_rate": 0.0001909997628427459,
+      "loss": 0.9589,
+      "step": 1554
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.19195522367954254,
+      "learning_rate": 0.00019098827912727976,
+      "loss": 1.0424,
+      "step": 1555
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.18206413090229034,
+      "learning_rate": 0.00019097678843592344,
+      "loss": 0.9539,
+      "step": 1556
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1879245638847351,
+      "learning_rate": 0.00019096529076955787,
+      "loss": 1.0335,
+      "step": 1557
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1857089251279831,
+      "learning_rate": 0.0001909537861290646,
+      "loss": 1.0136,
+      "step": 1558
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.20278511941432953,
+      "learning_rate": 0.00019094227451532563,
+      "loss": 1.002,
+      "step": 1559
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.197901651263237,
+      "learning_rate": 0.00019093075592922358,
+      "loss": 1.0214,
+      "step": 1560
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.18838657438755035,
+      "learning_rate": 0.00019091923037164156,
+      "loss": 1.0061,
+      "step": 1561
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.18443070352077484,
+      "learning_rate": 0.00019090769784346314,
+      "loss": 1.0103,
+      "step": 1562
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.18846985697746277,
+      "learning_rate": 0.00019089615834557252,
+      "loss": 0.9854,
+      "step": 1563
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1843843013048172,
+      "learning_rate": 0.0001908846118788545,
+      "loss": 1.0166,
+      "step": 1564
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.18994472920894623,
+      "learning_rate": 0.0001908730584441942,
+      "loss": 1.0502,
+      "step": 1565
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1749737709760666,
+      "learning_rate": 0.00019086149804247747,
+      "loss": 0.9723,
+      "step": 1566
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1871078461408615,
+      "learning_rate": 0.0001908499306745906,
+      "loss": 0.9993,
+      "step": 1567
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.18963484466075897,
+      "learning_rate": 0.00019083835634142043,
+      "loss": 1.0489,
+      "step": 1568
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1813851296901703,
+      "learning_rate": 0.00019082677504385438,
+      "loss": 0.9793,
+      "step": 1569
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.181089848279953,
+      "learning_rate": 0.00019081518678278035,
+      "loss": 0.9995,
+      "step": 1570
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.18877080082893372,
+      "learning_rate": 0.00019080359155908676,
+      "loss": 1.0485,
+      "step": 1571
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.17573611438274384,
+      "learning_rate": 0.0001907919893736626,
+      "loss": 0.9614,
+      "step": 1572
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.18343812227249146,
+      "learning_rate": 0.0001907803802273974,
+      "loss": 1.0384,
+      "step": 1573
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.16975995898246765,
+      "learning_rate": 0.00019076876412118122,
+      "loss": 0.9212,
+      "step": 1574
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.18836268782615662,
+      "learning_rate": 0.00019075714105590464,
+      "loss": 0.9728,
+      "step": 1575
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.18021149933338165,
+      "learning_rate": 0.00019074551103245876,
+      "loss": 1.0136,
+      "step": 1576
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.18562135100364685,
+      "learning_rate": 0.0001907338740517352,
+      "loss": 1.0479,
+      "step": 1577
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1791529655456543,
+      "learning_rate": 0.00019072223011462627,
+      "loss": 1.0383,
+      "step": 1578
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1899656504392624,
+      "learning_rate": 0.00019071057922202456,
+      "loss": 1.0454,
+      "step": 1579
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.17192921042442322,
+      "learning_rate": 0.00019069892137482333,
+      "loss": 0.9879,
+      "step": 1580
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.17805634438991547,
+      "learning_rate": 0.00019068725657391645,
+      "loss": 1.0285,
+      "step": 1581
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.17237581312656403,
+      "learning_rate": 0.00019067558482019811,
+      "loss": 0.9978,
+      "step": 1582
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.17143943905830383,
+      "learning_rate": 0.00019066390611456328,
+      "loss": 0.9697,
+      "step": 1583
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.16979925334453583,
+      "learning_rate": 0.0001906522204579073,
+      "loss": 0.9267,
+      "step": 1584
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.18201367557048798,
+      "learning_rate": 0.00019064052785112607,
+      "loss": 0.9633,
+      "step": 1585
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.17835623025894165,
+      "learning_rate": 0.00019062882829511607,
+      "loss": 0.9812,
+      "step": 1586
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.18148957192897797,
+      "learning_rate": 0.00019061712179077424,
+      "loss": 1.0285,
+      "step": 1587
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.18641547858715057,
+      "learning_rate": 0.00019060540833899814,
+      "loss": 1.0153,
+      "step": 1588
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1883457750082016,
+      "learning_rate": 0.00019059368794068578,
+      "loss": 0.9653,
+      "step": 1589
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.18426276743412018,
+      "learning_rate": 0.00019058196059673575,
+      "loss": 1.0388,
+      "step": 1590
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.19204431772232056,
+      "learning_rate": 0.00019057022630804716,
+      "loss": 1.0635,
+      "step": 1591
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1714024692773819,
+      "learning_rate": 0.00019055848507551967,
+      "loss": 0.9776,
+      "step": 1592
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1830907016992569,
+      "learning_rate": 0.00019054673690005344,
+      "loss": 0.9496,
+      "step": 1593
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.17566922307014465,
+      "learning_rate": 0.0001905349817825492,
+      "loss": 0.9616,
+      "step": 1594
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1792527139186859,
+      "learning_rate": 0.00019052321972390815,
+      "loss": 0.9606,
+      "step": 1595
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.18835650384426117,
+      "learning_rate": 0.00019051145072503215,
+      "loss": 1.0187,
+      "step": 1596
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.17348940670490265,
+      "learning_rate": 0.00019049967478682338,
+      "loss": 0.9624,
+      "step": 1597
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1860518902540207,
+      "learning_rate": 0.00019048789191018477,
+      "loss": 0.963,
+      "step": 1598
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1794481873512268,
+      "learning_rate": 0.0001904761020960197,
+      "loss": 0.9907,
+      "step": 1599
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.18411587178707123,
+      "learning_rate": 0.00019046430534523198,
+      "loss": 1.0253,
+      "step": 1600
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1995529979467392,
+      "learning_rate": 0.00019045250165872614,
+      "loss": 1.0077,
+      "step": 1601
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.17933034896850586,
+      "learning_rate": 0.00019044069103740707,
+      "loss": 0.9448,
+      "step": 1602
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.17415685951709747,
+      "learning_rate": 0.00019042887348218033,
+      "loss": 0.9509,
+      "step": 1603
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.17275942862033844,
+      "learning_rate": 0.00019041704899395192,
+      "loss": 0.975,
+      "step": 1604
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.17786523699760437,
+      "learning_rate": 0.0001904052175736284,
+      "loss": 1.0138,
+      "step": 1605
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.19810418784618378,
+      "learning_rate": 0.00019039337922211683,
+      "loss": 1.0133,
+      "step": 1606
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.17100881040096283,
+      "learning_rate": 0.0001903815339403249,
+      "loss": 0.9999,
+      "step": 1607
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1892804652452469,
+      "learning_rate": 0.0001903696817291607,
+      "loss": 0.9823,
+      "step": 1608
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.18206055462360382,
+      "learning_rate": 0.00019035782258953297,
+      "loss": 0.9637,
+      "step": 1609
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.18716344237327576,
+      "learning_rate": 0.0001903459565223509,
+      "loss": 1.0761,
+      "step": 1610
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.17864565551280975,
+      "learning_rate": 0.00019033408352852422,
+      "loss": 1.0491,
+      "step": 1611
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.18046505749225616,
+      "learning_rate": 0.00019032220360896325,
+      "loss": 1.0324,
+      "step": 1612
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.18205080926418304,
+      "learning_rate": 0.00019031031676457875,
+      "loss": 0.9451,
+      "step": 1613
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1753171980381012,
+      "learning_rate": 0.00019029842299628208,
+      "loss": 1.0274,
+      "step": 1614
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1871657818555832,
+      "learning_rate": 0.00019028652230498516,
+      "loss": 1.0019,
+      "step": 1615
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.17937354743480682,
+      "learning_rate": 0.00019027461469160032,
+      "loss": 0.9694,
+      "step": 1616
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.17640085518360138,
+      "learning_rate": 0.00019026270015704054,
+      "loss": 1.0143,
+      "step": 1617
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.18968771398067474,
+      "learning_rate": 0.00019025077870221929,
+      "loss": 0.9798,
+      "step": 1618
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.18246042728424072,
+      "learning_rate": 0.0001902388503280505,
+      "loss": 1.0376,
+      "step": 1619
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.18239466845989227,
+      "learning_rate": 0.00019022691503544877,
+      "loss": 1.0622,
+      "step": 1620
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.17998111248016357,
+      "learning_rate": 0.00019021497282532907,
+      "loss": 1.0192,
+      "step": 1621
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.18229350447654724,
+      "learning_rate": 0.00019020302369860708,
+      "loss": 1.0453,
+      "step": 1622
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.17646130919456482,
+      "learning_rate": 0.00019019106765619888,
+      "loss": 0.9841,
+      "step": 1623
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.18252936005592346,
+      "learning_rate": 0.00019017910469902108,
+      "loss": 1.0199,
+      "step": 1624
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.17558379471302032,
+      "learning_rate": 0.00019016713482799088,
+      "loss": 0.9245,
+      "step": 1625
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1946648508310318,
+      "learning_rate": 0.00019015515804402604,
+      "loss": 1.0411,
+      "step": 1626
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1894584596157074,
+      "learning_rate": 0.00019014317434804472,
+      "loss": 0.977,
+      "step": 1627
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1869843602180481,
+      "learning_rate": 0.0001901311837409657,
+      "loss": 1.0375,
+      "step": 1628
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1728430539369583,
+      "learning_rate": 0.0001901191862237083,
+      "loss": 0.9841,
+      "step": 1629
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.18686573207378387,
+      "learning_rate": 0.00019010718179719236,
+      "loss": 0.9953,
+      "step": 1630
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.19124893844127655,
+      "learning_rate": 0.00019009517046233816,
+      "loss": 1.0541,
+      "step": 1631
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1842876523733139,
+      "learning_rate": 0.00019008315222006667,
+      "loss": 0.9448,
+      "step": 1632
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.17653100192546844,
+      "learning_rate": 0.00019007112707129927,
+      "loss": 1.0036,
+      "step": 1633
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.17082680761814117,
+      "learning_rate": 0.0001900590950169579,
+      "loss": 0.9503,
+      "step": 1634
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.17388126254081726,
+      "learning_rate": 0.00019004705605796504,
+      "loss": 0.9682,
+      "step": 1635
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.19252508878707886,
+      "learning_rate": 0.00019003501019524368,
+      "loss": 0.9718,
+      "step": 1636
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.18236425518989563,
+      "learning_rate": 0.00019002295742971738,
+      "loss": 1.0233,
+      "step": 1637
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.18430474400520325,
+      "learning_rate": 0.0001900108977623102,
+      "loss": 1.0188,
+      "step": 1638
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1759885847568512,
+      "learning_rate": 0.00018999883119394668,
+      "loss": 0.9615,
+      "step": 1639
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1740919053554535,
+      "learning_rate": 0.000189986757725552,
+      "loss": 1.0351,
+      "step": 1640
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.188923642039299,
+      "learning_rate": 0.0001899746773580518,
+      "loss": 1.004,
+      "step": 1641
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.17621168494224548,
+      "learning_rate": 0.0001899625900923722,
+      "loss": 0.986,
+      "step": 1642
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1900838315486908,
+      "learning_rate": 0.00018995049592943997,
+      "loss": 1.0189,
+      "step": 1643
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.18396282196044922,
+      "learning_rate": 0.00018993839487018232,
+      "loss": 0.8813,
+      "step": 1644
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.18651731312274933,
+      "learning_rate": 0.00018992628691552702,
+      "loss": 1.0907,
+      "step": 1645
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.18417122960090637,
+      "learning_rate": 0.00018991417206640237,
+      "loss": 1.0003,
+      "step": 1646
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.19596096873283386,
+      "learning_rate": 0.00018990205032373716,
+      "loss": 1.0152,
+      "step": 1647
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.17158856987953186,
+      "learning_rate": 0.0001898899216884608,
+      "loss": 0.9455,
+      "step": 1648
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.17305323481559753,
+      "learning_rate": 0.00018987778616150308,
+      "loss": 0.9388,
+      "step": 1649
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1787583976984024,
+      "learning_rate": 0.0001898656437437945,
+      "loss": 1.0368,
+      "step": 1650
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.18098847568035126,
+      "learning_rate": 0.00018985349443626594,
+      "loss": 1.0413,
+      "step": 1651
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.18061774969100952,
+      "learning_rate": 0.00018984133823984886,
+      "loss": 1.0476,
+      "step": 1652
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.18374860286712646,
+      "learning_rate": 0.0001898291751554753,
+      "loss": 1.0636,
+      "step": 1653
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.17729640007019043,
+      "learning_rate": 0.00018981700518407775,
+      "loss": 0.9942,
+      "step": 1654
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.18511778116226196,
+      "learning_rate": 0.00018980482832658922,
+      "loss": 1.0035,
+      "step": 1655
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.18169361352920532,
+      "learning_rate": 0.00018979264458394334,
+      "loss": 0.998,
+      "step": 1656
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.18548671901226044,
+      "learning_rate": 0.00018978045395707418,
+      "loss": 1.0163,
+      "step": 1657
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1922914832830429,
+      "learning_rate": 0.00018976825644691637,
+      "loss": 1.065,
+      "step": 1658
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.17772276699543,
+      "learning_rate": 0.00018975605205440513,
+      "loss": 0.9657,
+      "step": 1659
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.17630743980407715,
+      "learning_rate": 0.00018974384078047606,
+      "loss": 0.9632,
+      "step": 1660
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.17994816601276398,
+      "learning_rate": 0.00018973162262606543,
+      "loss": 0.946,
+      "step": 1661
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.18123731017112732,
+      "learning_rate": 0.00018971939759210998,
+      "loss": 1.0613,
+      "step": 1662
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.18457649648189545,
+      "learning_rate": 0.0001897071656795469,
+      "loss": 0.984,
+      "step": 1663
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.18137028813362122,
+      "learning_rate": 0.00018969492688931412,
+      "loss": 0.999,
+      "step": 1664
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.18939583003520966,
+      "learning_rate": 0.00018968268122234987,
+      "loss": 0.9991,
+      "step": 1665
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.18159011006355286,
+      "learning_rate": 0.000189670428679593,
+      "loss": 0.9348,
+      "step": 1666
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1750485897064209,
+      "learning_rate": 0.00018965816926198295,
+      "loss": 1.0172,
+      "step": 1667
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.18323971331119537,
+      "learning_rate": 0.00018964590297045958,
+      "loss": 0.9855,
+      "step": 1668
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1904216706752777,
+      "learning_rate": 0.00018963362980596333,
+      "loss": 1.0082,
+      "step": 1669
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.17032112181186676,
+      "learning_rate": 0.00018962134976943516,
+      "loss": 0.9158,
+      "step": 1670
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.18121746182441711,
+      "learning_rate": 0.00018960906286181653,
+      "loss": 0.9998,
+      "step": 1671
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.17808766663074493,
+      "learning_rate": 0.00018959676908404948,
+      "loss": 0.9863,
+      "step": 1672
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.19097678363323212,
+      "learning_rate": 0.00018958446843707658,
+      "loss": 0.9903,
+      "step": 1673
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2249254286289215,
+      "learning_rate": 0.00018957216092184084,
+      "loss": 0.9,
+      "step": 1674
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.17637252807617188,
+      "learning_rate": 0.0001895598465392859,
+      "loss": 1.0117,
+      "step": 1675
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.17421402037143707,
+      "learning_rate": 0.00018954752529035584,
+      "loss": 0.9635,
+      "step": 1676
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.17446309328079224,
+      "learning_rate": 0.0001895351971759953,
+      "loss": 0.985,
+      "step": 1677
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.18266022205352783,
+      "learning_rate": 0.00018952286219714952,
+      "loss": 1.0179,
+      "step": 1678
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1776091754436493,
+      "learning_rate": 0.0001895105203547641,
+      "loss": 0.9669,
+      "step": 1679
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.18488481640815735,
+      "learning_rate": 0.00018949817164978536,
+      "loss": 1.0574,
+      "step": 1680
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.17778055369853973,
+      "learning_rate": 0.00018948581608315999,
+      "loss": 1.0059,
+      "step": 1681
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1818065345287323,
+      "learning_rate": 0.00018947345365583526,
+      "loss": 1.0158,
+      "step": 1682
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.18250834941864014,
+      "learning_rate": 0.00018946108436875903,
+      "loss": 0.9978,
+      "step": 1683
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1843356043100357,
+      "learning_rate": 0.00018944870822287956,
+      "loss": 1.066,
+      "step": 1684
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.18172258138656616,
+      "learning_rate": 0.00018943632521914577,
+      "loss": 0.9483,
+      "step": 1685
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.17615202069282532,
+      "learning_rate": 0.000189423935358507,
+      "loss": 0.9955,
+      "step": 1686
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.17779870331287384,
+      "learning_rate": 0.00018941153864191317,
+      "loss": 1.0287,
+      "step": 1687
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.18770545721054077,
+      "learning_rate": 0.0001893991350703147,
+      "loss": 1.0095,
+      "step": 1688
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.18199758231639862,
+      "learning_rate": 0.00018938672464466255,
+      "loss": 0.9966,
+      "step": 1689
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1823631227016449,
+      "learning_rate": 0.0001893743073659082,
+      "loss": 1.0017,
+      "step": 1690
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.18197377026081085,
+      "learning_rate": 0.00018936188323500367,
+      "loss": 0.9994,
+      "step": 1691
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.17887642979621887,
+      "learning_rate": 0.00018934945225290148,
+      "loss": 0.9761,
+      "step": 1692
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.18398606777191162,
+      "learning_rate": 0.0001893370144205547,
+      "loss": 0.9836,
+      "step": 1693
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.18904682993888855,
+      "learning_rate": 0.00018932456973891695,
+      "loss": 1.0479,
+      "step": 1694
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.18656067550182343,
+      "learning_rate": 0.00018931211820894225,
+      "loss": 0.987,
+      "step": 1695
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.17840254306793213,
+      "learning_rate": 0.00018929965983158532,
+      "loss": 0.9536,
+      "step": 1696
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.187407448887825,
+      "learning_rate": 0.00018928719460780123,
+      "loss": 1.0308,
+      "step": 1697
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.18530529737472534,
+      "learning_rate": 0.00018927472253854575,
+      "loss": 1.0369,
+      "step": 1698
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.17825321853160858,
+      "learning_rate": 0.00018926224362477508,
+      "loss": 1.013,
+      "step": 1699
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.17379194498062134,
+      "learning_rate": 0.0001892497578674459,
+      "loss": 1.0158,
+      "step": 1700
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.18412184715270996,
+      "learning_rate": 0.00018923726526751548,
+      "loss": 1.0062,
+      "step": 1701
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.18638308346271515,
+      "learning_rate": 0.00018922476582594163,
+      "loss": 1.0076,
+      "step": 1702
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.17837877571582794,
+      "learning_rate": 0.00018921225954368268,
+      "loss": 0.8609,
+      "step": 1703
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.184190034866333,
+      "learning_rate": 0.00018919974642169738,
+      "loss": 0.9835,
+      "step": 1704
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.19461223483085632,
+      "learning_rate": 0.00018918722646094516,
+      "loss": 1.0066,
+      "step": 1705
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1890239417552948,
+      "learning_rate": 0.00018917469966238585,
+      "loss": 0.9629,
+      "step": 1706
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.17515210807323456,
+      "learning_rate": 0.00018916216602697991,
+      "loss": 0.9093,
+      "step": 1707
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1932542473077774,
+      "learning_rate": 0.00018914962555568825,
+      "loss": 0.9947,
+      "step": 1708
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.17919421195983887,
+      "learning_rate": 0.00018913707824947228,
+      "loss": 0.971,
+      "step": 1709
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.18187792599201202,
+      "learning_rate": 0.000189124524109294,
+      "loss": 0.994,
+      "step": 1710
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1855144202709198,
+      "learning_rate": 0.00018911196313611597,
+      "loss": 0.9237,
+      "step": 1711
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.19063417613506317,
+      "learning_rate": 0.00018909939533090113,
+      "loss": 1.0746,
+      "step": 1712
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1843891590833664,
+      "learning_rate": 0.00018908682069461303,
+      "loss": 1.0307,
+      "step": 1713
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.19610640406608582,
+      "learning_rate": 0.00018907423922821584,
+      "loss": 1.0087,
+      "step": 1714
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.19017747044563293,
+      "learning_rate": 0.00018906165093267405,
+      "loss": 0.9879,
+      "step": 1715
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1834624856710434,
+      "learning_rate": 0.00018904905580895284,
+      "loss": 0.996,
+      "step": 1716
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.18889987468719482,
+      "learning_rate": 0.00018903645385801783,
+      "loss": 0.9959,
+      "step": 1717
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.18375161290168762,
+      "learning_rate": 0.00018902384508083517,
+      "loss": 1.0468,
+      "step": 1718
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.17865221202373505,
+      "learning_rate": 0.0001890112294783716,
+      "loss": 0.9869,
+      "step": 1719
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.18713891506195068,
+      "learning_rate": 0.00018899860705159428,
+      "loss": 0.984,
+      "step": 1720
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.18462030589580536,
+      "learning_rate": 0.00018898597780147097,
+      "loss": 0.9846,
+      "step": 1721
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.19466228783130646,
+      "learning_rate": 0.00018897334172896993,
+      "loss": 0.9924,
+      "step": 1722
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.17786429822444916,
+      "learning_rate": 0.00018896069883505992,
+      "loss": 1.0272,
+      "step": 1723
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1776842325925827,
+      "learning_rate": 0.00018894804912071033,
+      "loss": 0.9687,
+      "step": 1724
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.17448489367961884,
+      "learning_rate": 0.00018893539258689089,
+      "loss": 0.9592,
+      "step": 1725
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1864984929561615,
+      "learning_rate": 0.00018892272923457198,
+      "loss": 1.0414,
+      "step": 1726
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.17481493949890137,
+      "learning_rate": 0.0001889100590647245,
+      "loss": 1.0149,
+      "step": 1727
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.17752544581890106,
+      "learning_rate": 0.0001888973820783198,
+      "loss": 0.9654,
+      "step": 1728
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.17314109206199646,
+      "learning_rate": 0.00018888469827632985,
+      "loss": 1.0088,
+      "step": 1729
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.19234046339988708,
+      "learning_rate": 0.00018887200765972705,
+      "loss": 0.9895,
+      "step": 1730
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1833478957414627,
+      "learning_rate": 0.0001888593102294844,
+      "loss": 1.0427,
+      "step": 1731
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.18196554481983185,
+      "learning_rate": 0.00018884660598657535,
+      "loss": 1.0088,
+      "step": 1732
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.18826749920845032,
+      "learning_rate": 0.00018883389493197394,
+      "loss": 1.0019,
+      "step": 1733
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.18463777005672455,
+      "learning_rate": 0.00018882117706665467,
+      "loss": 0.9894,
+      "step": 1734
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.17686514556407928,
+      "learning_rate": 0.00018880845239159264,
+      "loss": 0.9487,
+      "step": 1735
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.18111956119537354,
+      "learning_rate": 0.0001887957209077634,
+      "loss": 1.0057,
+      "step": 1736
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.17422161996364594,
+      "learning_rate": 0.00018878298261614304,
+      "loss": 1.0053,
+      "step": 1737
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1680193990468979,
+      "learning_rate": 0.00018877023751770816,
+      "loss": 0.9142,
+      "step": 1738
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1755419671535492,
+      "learning_rate": 0.00018875748561343595,
+      "loss": 1.0097,
+      "step": 1739
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1781400442123413,
+      "learning_rate": 0.00018874472690430408,
+      "loss": 0.9885,
+      "step": 1740
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.18728172779083252,
+      "learning_rate": 0.00018873196139129067,
+      "loss": 1.0323,
+      "step": 1741
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.185468390583992,
+      "learning_rate": 0.00018871918907537445,
+      "loss": 0.9909,
+      "step": 1742
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.18841905891895294,
+      "learning_rate": 0.00018870640995753468,
+      "loss": 0.9763,
+      "step": 1743
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.18700815737247467,
+      "learning_rate": 0.00018869362403875107,
+      "loss": 1.0644,
+      "step": 1744
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.18525230884552002,
+      "learning_rate": 0.00018868083132000393,
+      "loss": 1.0113,
+      "step": 1745
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1759456992149353,
+      "learning_rate": 0.00018866803180227402,
+      "loss": 1.0158,
+      "step": 1746
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.18815183639526367,
+      "learning_rate": 0.00018865522548654266,
+      "loss": 1.0415,
+      "step": 1747
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.17770276963710785,
+      "learning_rate": 0.0001886424123737917,
+      "loss": 0.9943,
+      "step": 1748
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.17623820900917053,
+      "learning_rate": 0.0001886295924650035,
+      "loss": 1.0151,
+      "step": 1749
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1780458390712738,
+      "learning_rate": 0.0001886167657611609,
+      "loss": 1.0295,
+      "step": 1750
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.18125370144844055,
+      "learning_rate": 0.00018860393226324734,
+      "loss": 0.9904,
+      "step": 1751
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1796397864818573,
+      "learning_rate": 0.00018859109197224668,
+      "loss": 0.9625,
+      "step": 1752
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.18634067475795746,
+      "learning_rate": 0.00018857824488914345,
+      "loss": 0.9659,
+      "step": 1753
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1882651001214981,
+      "learning_rate": 0.00018856539101492256,
+      "loss": 1.0561,
+      "step": 1754
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.18537171185016632,
+      "learning_rate": 0.00018855253035056943,
+      "loss": 1.0249,
+      "step": 1755
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1841856837272644,
+      "learning_rate": 0.00018853966289707018,
+      "loss": 1.0308,
+      "step": 1756
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.18117250502109528,
+      "learning_rate": 0.00018852678865541122,
+      "loss": 1.0331,
+      "step": 1757
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.18570590019226074,
+      "learning_rate": 0.0001885139076265797,
+      "loss": 0.966,
+      "step": 1758
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1879170686006546,
+      "learning_rate": 0.0001885010198115631,
+      "loss": 0.9882,
+      "step": 1759
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.18629556894302368,
+      "learning_rate": 0.00018848812521134956,
+      "loss": 0.9312,
+      "step": 1760
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.17607717216014862,
+      "learning_rate": 0.00018847522382692763,
+      "loss": 0.9519,
+      "step": 1761
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1829354614019394,
+      "learning_rate": 0.00018846231565928644,
+      "loss": 0.9932,
+      "step": 1762
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.18066377937793732,
+      "learning_rate": 0.0001884494007094157,
+      "loss": 0.9585,
+      "step": 1763
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.17497728765010834,
+      "learning_rate": 0.0001884364789783055,
+      "loss": 0.9822,
+      "step": 1764
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.18487730622291565,
+      "learning_rate": 0.00018842355046694656,
+      "loss": 1.0206,
+      "step": 1765
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1813206523656845,
+      "learning_rate": 0.00018841061517633006,
+      "loss": 0.9873,
+      "step": 1766
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1820863038301468,
+      "learning_rate": 0.00018839767310744777,
+      "loss": 1.002,
+      "step": 1767
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.18336263298988342,
+      "learning_rate": 0.00018838472426129185,
+      "loss": 0.9764,
+      "step": 1768
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.18435858190059662,
+      "learning_rate": 0.00018837176863885512,
+      "loss": 1.0391,
+      "step": 1769
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.17950308322906494,
+      "learning_rate": 0.00018835880624113087,
+      "loss": 0.9695,
+      "step": 1770
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.18525420129299164,
+      "learning_rate": 0.00018834583706911286,
+      "loss": 1.0169,
+      "step": 1771
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.18522191047668457,
+      "learning_rate": 0.00018833286112379547,
+      "loss": 1.0383,
+      "step": 1772
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.17976364493370056,
+      "learning_rate": 0.00018831987840617348,
+      "loss": 0.9452,
+      "step": 1773
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.19244349002838135,
+      "learning_rate": 0.00018830688891724226,
+      "loss": 1.0472,
+      "step": 1774
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1925305873155594,
+      "learning_rate": 0.00018829389265799775,
+      "loss": 1.0674,
+      "step": 1775
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1825743019580841,
+      "learning_rate": 0.00018828088962943626,
+      "loss": 0.9906,
+      "step": 1776
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1989976316690445,
+      "learning_rate": 0.00018826787983255473,
+      "loss": 0.9963,
+      "step": 1777
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1910424530506134,
+      "learning_rate": 0.00018825486326835063,
+      "loss": 0.9898,
+      "step": 1778
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.17530961334705353,
+      "learning_rate": 0.00018824183993782192,
+      "loss": 1.0211,
+      "step": 1779
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1942378282546997,
+      "learning_rate": 0.00018822880984196703,
+      "loss": 1.0093,
+      "step": 1780
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1794646829366684,
+      "learning_rate": 0.00018821577298178496,
+      "loss": 0.9389,
+      "step": 1781
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.19524367153644562,
+      "learning_rate": 0.00018820272935827524,
+      "loss": 0.99,
+      "step": 1782
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1962813436985016,
+      "learning_rate": 0.00018818967897243785,
+      "loss": 0.9494,
+      "step": 1783
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.17543120682239532,
+      "learning_rate": 0.00018817662182527342,
+      "loss": 0.96,
+      "step": 1784
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.18655969202518463,
+      "learning_rate": 0.00018816355791778294,
+      "loss": 1.0292,
+      "step": 1785
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1929544359445572,
+      "learning_rate": 0.00018815048725096805,
+      "loss": 1.0016,
+      "step": 1786
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1946924775838852,
+      "learning_rate": 0.00018813740982583083,
+      "loss": 1.0131,
+      "step": 1787
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1820334494113922,
+      "learning_rate": 0.00018812432564337388,
+      "loss": 0.9705,
+      "step": 1788
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.18293553590774536,
+      "learning_rate": 0.00018811123470460035,
+      "loss": 0.9449,
+      "step": 1789
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.18539494276046753,
+      "learning_rate": 0.0001880981370105139,
+      "loss": 0.9884,
+      "step": 1790
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.18742531538009644,
+      "learning_rate": 0.00018808503256211872,
+      "loss": 1.0232,
+      "step": 1791
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1922466903924942,
+      "learning_rate": 0.00018807192136041943,
+      "loss": 0.9864,
+      "step": 1792
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.19350063800811768,
+      "learning_rate": 0.00018805880340642137,
+      "loss": 0.9306,
+      "step": 1793
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1808355748653412,
+      "learning_rate": 0.00018804567870113017,
+      "loss": 0.948,
+      "step": 1794
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1811530888080597,
+      "learning_rate": 0.0001880325472455521,
+      "loss": 1.0872,
+      "step": 1795
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.18100488185882568,
+      "learning_rate": 0.0001880194090406939,
+      "loss": 1.0122,
+      "step": 1796
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.20256204903125763,
+      "learning_rate": 0.00018800626408756288,
+      "loss": 1.0903,
+      "step": 1797
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.19154663383960724,
+      "learning_rate": 0.00018799311238716683,
+      "loss": 1.0409,
+      "step": 1798
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1828100085258484,
+      "learning_rate": 0.00018797995394051408,
+      "loss": 0.9465,
+      "step": 1799
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1766197383403778,
+      "learning_rate": 0.00018796678874861343,
+      "loss": 0.9912,
+      "step": 1800
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.19492526352405548,
+      "learning_rate": 0.00018795361681247426,
+      "loss": 0.9834,
+      "step": 1801
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2047867476940155,
+      "learning_rate": 0.00018794043813310637,
+      "loss": 1.0388,
+      "step": 1802
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.17892782390117645,
+      "learning_rate": 0.00018792725271152025,
+      "loss": 0.9646,
+      "step": 1803
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.17773431539535522,
+      "learning_rate": 0.00018791406054872672,
+      "loss": 0.9839,
+      "step": 1804
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1797155886888504,
+      "learning_rate": 0.00018790086164573722,
+      "loss": 1.0079,
+      "step": 1805
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.18631033599376678,
+      "learning_rate": 0.0001878876560035637,
+      "loss": 1.0001,
+      "step": 1806
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.18853716552257538,
+      "learning_rate": 0.00018787444362321857,
+      "loss": 0.9939,
+      "step": 1807
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1785057783126831,
+      "learning_rate": 0.00018786122450571485,
+      "loss": 0.9477,
+      "step": 1808
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.18913787603378296,
+      "learning_rate": 0.00018784799865206596,
+      "loss": 0.9952,
+      "step": 1809
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.18841691315174103,
+      "learning_rate": 0.00018783476606328594,
+      "loss": 1.0136,
+      "step": 1810
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.17880366742610931,
+      "learning_rate": 0.00018782152674038933,
+      "loss": 1.0012,
+      "step": 1811
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.18537718057632446,
+      "learning_rate": 0.00018780828068439112,
+      "loss": 1.0081,
+      "step": 1812
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.18226812779903412,
+      "learning_rate": 0.00018779502789630686,
+      "loss": 0.9489,
+      "step": 1813
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.18130582571029663,
+      "learning_rate": 0.00018778176837715263,
+      "loss": 0.9939,
+      "step": 1814
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1837967485189438,
+      "learning_rate": 0.000187768502127945,
+      "loss": 1.0102,
+      "step": 1815
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.18316712975502014,
+      "learning_rate": 0.0001877552291497011,
+      "loss": 0.9439,
+      "step": 1816
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1856684684753418,
+      "learning_rate": 0.0001877419494434385,
+      "loss": 0.9549,
+      "step": 1817
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.18236926198005676,
+      "learning_rate": 0.00018772866301017534,
+      "loss": 1.0313,
+      "step": 1818
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.18475475907325745,
+      "learning_rate": 0.00018771536985093026,
+      "loss": 1.026,
+      "step": 1819
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.18892961740493774,
+      "learning_rate": 0.00018770206996672245,
+      "loss": 0.968,
+      "step": 1820
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.18441183865070343,
+      "learning_rate": 0.0001876887633585716,
+      "loss": 1.0116,
+      "step": 1821
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.18181204795837402,
+      "learning_rate": 0.00018767545002749782,
+      "loss": 1.0226,
+      "step": 1822
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1826629638671875,
+      "learning_rate": 0.00018766212997452185,
+      "loss": 0.9366,
+      "step": 1823
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1744999885559082,
+      "learning_rate": 0.00018764880320066497,
+      "loss": 0.9736,
+      "step": 1824
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.18155504763126373,
+      "learning_rate": 0.00018763546970694886,
+      "loss": 0.9828,
+      "step": 1825
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.18461792171001434,
+      "learning_rate": 0.00018762212949439578,
+      "loss": 1.0278,
+      "step": 1826
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.18739303946495056,
+      "learning_rate": 0.00018760878256402852,
+      "loss": 0.9921,
+      "step": 1827
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1814883053302765,
+      "learning_rate": 0.0001875954289168703,
+      "loss": 1.0008,
+      "step": 1828
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.17120951414108276,
+      "learning_rate": 0.000187582068553945,
+      "loss": 0.9497,
+      "step": 1829
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.183866485953331,
+      "learning_rate": 0.0001875687014762769,
+      "loss": 1.0129,
+      "step": 1830
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.186480313539505,
+      "learning_rate": 0.00018755532768489078,
+      "loss": 0.9857,
+      "step": 1831
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.19442792236804962,
+      "learning_rate": 0.00018754194718081205,
+      "loss": 0.9795,
+      "step": 1832
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.18308177590370178,
+      "learning_rate": 0.00018752855996506654,
+      "loss": 1.0739,
+      "step": 1833
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1863255500793457,
+      "learning_rate": 0.00018751516603868066,
+      "loss": 0.9829,
+      "step": 1834
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.17856715619564056,
+      "learning_rate": 0.0001875017654026812,
+      "loss": 1.0195,
+      "step": 1835
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.18626144528388977,
+      "learning_rate": 0.00018748835805809568,
+      "loss": 0.9666,
+      "step": 1836
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.19257773458957672,
+      "learning_rate": 0.00018747494400595191,
+      "loss": 1.0068,
+      "step": 1837
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.18437279760837555,
+      "learning_rate": 0.00018746152324727836,
+      "loss": 1.0515,
+      "step": 1838
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.18242545425891876,
+      "learning_rate": 0.00018744809578310397,
+      "loss": 0.9792,
+      "step": 1839
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1814119666814804,
+      "learning_rate": 0.00018743466161445823,
+      "loss": 0.9656,
+      "step": 1840
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.18533895909786224,
+      "learning_rate": 0.00018742122074237106,
+      "loss": 0.9934,
+      "step": 1841
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.18429550528526306,
+      "learning_rate": 0.00018740777316787296,
+      "loss": 1.0336,
+      "step": 1842
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.19191229343414307,
+      "learning_rate": 0.00018739431889199497,
+      "loss": 1.0694,
+      "step": 1843
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.18915875256061554,
+      "learning_rate": 0.00018738085791576855,
+      "loss": 1.0051,
+      "step": 1844
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1901199370622635,
+      "learning_rate": 0.0001873673902402257,
+      "loss": 1.0248,
+      "step": 1845
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.18844999372959137,
+      "learning_rate": 0.00018735391586639904,
+      "loss": 1.06,
+      "step": 1846
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1894400715827942,
+      "learning_rate": 0.0001873404347953216,
+      "loss": 0.9807,
+      "step": 1847
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.18685835599899292,
+      "learning_rate": 0.00018732694702802694,
+      "loss": 0.9532,
+      "step": 1848
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.176396906375885,
+      "learning_rate": 0.0001873134525655491,
+      "loss": 1.0345,
+      "step": 1849
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.18638281524181366,
+      "learning_rate": 0.00018729995140892274,
+      "loss": 1.0103,
+      "step": 1850
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.17325080931186676,
+      "learning_rate": 0.00018728644355918292,
+      "loss": 1.0141,
+      "step": 1851
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1763128787279129,
+      "learning_rate": 0.00018727292901736528,
+      "loss": 0.9846,
+      "step": 1852
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.18143166601657867,
+      "learning_rate": 0.00018725940778450598,
+      "loss": 0.9628,
+      "step": 1853
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.18402259051799774,
+      "learning_rate": 0.00018724587986164158,
+      "loss": 0.9959,
+      "step": 1854
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.19441376626491547,
+      "learning_rate": 0.0001872323452498093,
+      "loss": 0.9977,
+      "step": 1855
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.18418952822685242,
+      "learning_rate": 0.00018721880395004682,
+      "loss": 0.849,
+      "step": 1856
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1789008378982544,
+      "learning_rate": 0.00018720525596339232,
+      "loss": 0.9796,
+      "step": 1857
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.18109720945358276,
+      "learning_rate": 0.00018719170129088449,
+      "loss": 0.996,
+      "step": 1858
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.18615257740020752,
+      "learning_rate": 0.0001871781399335625,
+      "loss": 1.0433,
+      "step": 1859
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.19213053584098816,
+      "learning_rate": 0.00018716457189246614,
+      "loss": 1.0045,
+      "step": 1860
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.18630923330783844,
+      "learning_rate": 0.0001871509971686356,
+      "loss": 0.9688,
+      "step": 1861
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.18649254739284515,
+      "learning_rate": 0.00018713741576311166,
+      "loss": 1.0137,
+      "step": 1862
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1861242651939392,
+      "learning_rate": 0.00018712382767693554,
+      "loss": 0.9949,
+      "step": 1863
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.18726636469364166,
+      "learning_rate": 0.00018711023291114902,
+      "loss": 0.9067,
+      "step": 1864
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1886870563030243,
+      "learning_rate": 0.00018709663146679442,
+      "loss": 0.9772,
+      "step": 1865
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.18354971706867218,
+      "learning_rate": 0.0001870830233449145,
+      "loss": 1.0024,
+      "step": 1866
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.17389263212680817,
+      "learning_rate": 0.00018706940854655257,
+      "loss": 0.9816,
+      "step": 1867
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.17908309400081635,
+      "learning_rate": 0.00018705578707275248,
+      "loss": 0.9976,
+      "step": 1868
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.21369712054729462,
+      "learning_rate": 0.0001870421589245585,
+      "loss": 1.0394,
+      "step": 1869
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.19233554601669312,
+      "learning_rate": 0.00018702852410301554,
+      "loss": 0.9666,
+      "step": 1870
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1837395280599594,
+      "learning_rate": 0.00018701488260916893,
+      "loss": 1.0495,
+      "step": 1871
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.18571002781391144,
+      "learning_rate": 0.0001870012344440645,
+      "loss": 1.0363,
+      "step": 1872
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.19713854789733887,
+      "learning_rate": 0.0001869875796087487,
+      "loss": 0.9658,
+      "step": 1873
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.19223444163799286,
+      "learning_rate": 0.00018697391810426836,
+      "loss": 1.0167,
+      "step": 1874
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.18585839867591858,
+      "learning_rate": 0.00018696024993167088,
+      "loss": 0.9921,
+      "step": 1875
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1860339492559433,
+      "learning_rate": 0.00018694657509200422,
+      "loss": 0.9184,
+      "step": 1876
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.18812976777553558,
+      "learning_rate": 0.0001869328935863168,
+      "loss": 1.0175,
+      "step": 1877
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.19325508177280426,
+      "learning_rate": 0.00018691920541565746,
+      "loss": 1.103,
+      "step": 1878
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.18449616432189941,
+      "learning_rate": 0.00018690551058107575,
+      "loss": 1.0278,
+      "step": 1879
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1745263785123825,
+      "learning_rate": 0.0001868918090836216,
+      "loss": 0.9653,
+      "step": 1880
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.18073584139347076,
+      "learning_rate": 0.00018687810092434543,
+      "loss": 1.0015,
+      "step": 1881
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.19327598810195923,
+      "learning_rate": 0.0001868643861042983,
+      "loss": 1.016,
+      "step": 1882
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1936059147119522,
+      "learning_rate": 0.0001868506646245316,
+      "loss": 0.9642,
+      "step": 1883
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.19047310948371887,
+      "learning_rate": 0.0001868369364860974,
+      "loss": 1.1297,
+      "step": 1884
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.17914171516895294,
+      "learning_rate": 0.00018682320169004818,
+      "loss": 0.9462,
+      "step": 1885
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.17842666804790497,
+      "learning_rate": 0.00018680946023743697,
+      "loss": 0.9738,
+      "step": 1886
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.18439552187919617,
+      "learning_rate": 0.0001867957121293173,
+      "loss": 0.9926,
+      "step": 1887
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1877158284187317,
+      "learning_rate": 0.00018678195736674319,
+      "loss": 1.0793,
+      "step": 1888
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1864771544933319,
+      "learning_rate": 0.00018676819595076925,
+      "loss": 0.9975,
+      "step": 1889
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.18416166305541992,
+      "learning_rate": 0.00018675442788245047,
+      "loss": 0.9312,
+      "step": 1890
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.17662164568901062,
+      "learning_rate": 0.00018674065316284243,
+      "loss": 0.935,
+      "step": 1891
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.17577767372131348,
+      "learning_rate": 0.00018672687179300125,
+      "loss": 0.9825,
+      "step": 1892
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.17717039585113525,
+      "learning_rate": 0.0001867130837739835,
+      "loss": 0.9599,
+      "step": 1893
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.17826318740844727,
+      "learning_rate": 0.00018669928910684626,
+      "loss": 1.055,
+      "step": 1894
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.18381334841251373,
+      "learning_rate": 0.00018668548779264717,
+      "loss": 1.0138,
+      "step": 1895
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.17650206387043,
+      "learning_rate": 0.00018667167983244434,
+      "loss": 1.0064,
+      "step": 1896
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.17828193306922913,
+      "learning_rate": 0.00018665786522729637,
+      "loss": 0.9707,
+      "step": 1897
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.18671469390392303,
+      "learning_rate": 0.00018664404397826246,
+      "loss": 0.9894,
+      "step": 1898
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.17767448723316193,
+      "learning_rate": 0.00018663021608640224,
+      "loss": 0.967,
+      "step": 1899
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.19075635075569153,
+      "learning_rate": 0.0001866163815527758,
+      "loss": 0.993,
+      "step": 1900
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2010372281074524,
+      "learning_rate": 0.00018660254037844388,
+      "loss": 0.959,
+      "step": 1901
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.17780248820781708,
+      "learning_rate": 0.00018658869256446762,
+      "loss": 0.9479,
+      "step": 1902
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.18399259448051453,
+      "learning_rate": 0.0001865748381119087,
+      "loss": 1.0196,
+      "step": 1903
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1945018470287323,
+      "learning_rate": 0.00018656097702182938,
+      "loss": 0.9681,
+      "step": 1904
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1856342852115631,
+      "learning_rate": 0.00018654710929529226,
+      "loss": 1.0241,
+      "step": 1905
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1952160745859146,
+      "learning_rate": 0.00018653323493336062,
+      "loss": 1.033,
+      "step": 1906
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.17957225441932678,
+      "learning_rate": 0.00018651935393709814,
+      "loss": 0.8973,
+      "step": 1907
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.19565623998641968,
+      "learning_rate": 0.00018650546630756911,
+      "loss": 1.0472,
+      "step": 1908
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.17848311364650726,
+      "learning_rate": 0.0001864915720458382,
+      "loss": 0.9535,
+      "step": 1909
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1798015534877777,
+      "learning_rate": 0.00018647767115297068,
+      "loss": 0.9587,
+      "step": 1910
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.18396656215190887,
+      "learning_rate": 0.0001864637636300323,
+      "loss": 1.0296,
+      "step": 1911
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1760600209236145,
+      "learning_rate": 0.00018644984947808934,
+      "loss": 0.8995,
+      "step": 1912
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.195132777094841,
+      "learning_rate": 0.00018643592869820857,
+      "loss": 1.0686,
+      "step": 1913
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.19066643714904785,
+      "learning_rate": 0.00018642200129145723,
+      "loss": 0.967,
+      "step": 1914
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.18508267402648926,
+      "learning_rate": 0.0001864080672589031,
+      "loss": 1.0049,
+      "step": 1915
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.18527060747146606,
+      "learning_rate": 0.00018639412660161457,
+      "loss": 0.977,
+      "step": 1916
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.18264782428741455,
+      "learning_rate": 0.00018638017932066036,
+      "loss": 0.9405,
+      "step": 1917
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.17923100292682648,
+      "learning_rate": 0.00018636622541710982,
+      "loss": 1.01,
+      "step": 1918
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1840544193983078,
+      "learning_rate": 0.00018635226489203271,
+      "loss": 1.0268,
+      "step": 1919
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1948777288198471,
+      "learning_rate": 0.0001863382977464994,
+      "loss": 0.9392,
+      "step": 1920
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.18678011000156403,
+      "learning_rate": 0.00018632432398158074,
+      "loss": 0.9766,
+      "step": 1921
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.17854583263397217,
+      "learning_rate": 0.000186310343598348,
+      "loss": 0.9723,
+      "step": 1922
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.18715204298496246,
+      "learning_rate": 0.00018629635659787316,
+      "loss": 0.9783,
+      "step": 1923
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.18582193553447723,
+      "learning_rate": 0.00018628236298122842,
+      "loss": 0.9439,
+      "step": 1924
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.18433134257793427,
+      "learning_rate": 0.00018626836274948674,
+      "loss": 0.9874,
+      "step": 1925
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.18748904764652252,
+      "learning_rate": 0.00018625435590372143,
+      "loss": 1.0405,
+      "step": 1926
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.17972299456596375,
+      "learning_rate": 0.00018624034244500647,
+      "loss": 0.9845,
+      "step": 1927
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.18048334121704102,
+      "learning_rate": 0.00018622632237441612,
+      "loss": 0.9354,
+      "step": 1928
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1844087690114975,
+      "learning_rate": 0.00018621229569302532,
+      "loss": 0.9674,
+      "step": 1929
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.18459495902061462,
+      "learning_rate": 0.0001861982624019095,
+      "loss": 1.0473,
+      "step": 1930
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.19419121742248535,
+      "learning_rate": 0.00018618422250214452,
+      "loss": 1.0508,
+      "step": 1931
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.17196646332740784,
+      "learning_rate": 0.00018617017599480682,
+      "loss": 0.9615,
+      "step": 1932
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.18627747893333435,
+      "learning_rate": 0.0001861561228809733,
+      "loss": 1.0027,
+      "step": 1933
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.19207796454429626,
+      "learning_rate": 0.0001861420631617214,
+      "loss": 1.0159,
+      "step": 1934
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.18138843774795532,
+      "learning_rate": 0.00018612799683812904,
+      "loss": 0.9165,
+      "step": 1935
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.18867118656635284,
+      "learning_rate": 0.00018611392391127467,
+      "loss": 0.9952,
+      "step": 1936
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.18637476861476898,
+      "learning_rate": 0.0001860998443822372,
+      "loss": 0.9637,
+      "step": 1937
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.19077537953853607,
+      "learning_rate": 0.0001860857582520961,
+      "loss": 1.0254,
+      "step": 1938
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1727682203054428,
+      "learning_rate": 0.00018607166552193135,
+      "loss": 0.9634,
+      "step": 1939
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.186919167637825,
+      "learning_rate": 0.00018605756619282336,
+      "loss": 1.0402,
+      "step": 1940
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.18637052178382874,
+      "learning_rate": 0.00018604346026585312,
+      "loss": 1.0357,
+      "step": 1941
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.18484027683734894,
+      "learning_rate": 0.00018602934774210214,
+      "loss": 0.9575,
+      "step": 1942
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1901281327009201,
+      "learning_rate": 0.00018601522862265237,
+      "loss": 0.9959,
+      "step": 1943
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1861106902360916,
+      "learning_rate": 0.00018600110290858629,
+      "loss": 1.0286,
+      "step": 1944
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.19661380350589752,
+      "learning_rate": 0.00018598697060098686,
+      "loss": 1.02,
+      "step": 1945
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1923096925020218,
+      "learning_rate": 0.00018597283170093763,
+      "loss": 1.0177,
+      "step": 1946
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.19194743037223816,
+      "learning_rate": 0.0001859586862095226,
+      "loss": 1.0526,
+      "step": 1947
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.18508084118366241,
+      "learning_rate": 0.00018594453412782623,
+      "loss": 1.0435,
+      "step": 1948
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.18176361918449402,
+      "learning_rate": 0.00018593037545693356,
+      "loss": 1.0053,
+      "step": 1949
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1897421032190323,
+      "learning_rate": 0.00018591621019793013,
+      "loss": 1.0259,
+      "step": 1950
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.18557140231132507,
+      "learning_rate": 0.00018590203835190192,
+      "loss": 0.9631,
+      "step": 1951
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.19048331677913666,
+      "learning_rate": 0.0001858878599199355,
+      "loss": 1.0471,
+      "step": 1952
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.17859672009944916,
+      "learning_rate": 0.00018587367490311785,
+      "loss": 1.0225,
+      "step": 1953
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.18589475750923157,
+      "learning_rate": 0.00018585948330253652,
+      "loss": 0.9908,
+      "step": 1954
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1961270272731781,
+      "learning_rate": 0.0001858452851192796,
+      "loss": 1.0266,
+      "step": 1955
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1803094893693924,
+      "learning_rate": 0.0001858310803544356,
+      "loss": 0.9862,
+      "step": 1956
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.18154191970825195,
+      "learning_rate": 0.00018581686900909355,
+      "loss": 0.9889,
+      "step": 1957
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1789197474718094,
+      "learning_rate": 0.00018580265108434306,
+      "loss": 0.9266,
+      "step": 1958
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.18821920454502106,
+      "learning_rate": 0.00018578842658127415,
+      "loss": 0.9923,
+      "step": 1959
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1863432079553604,
+      "learning_rate": 0.0001857741955009774,
+      "loss": 0.9733,
+      "step": 1960
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.18911732733249664,
+      "learning_rate": 0.00018575995784454386,
+      "loss": 0.9425,
+      "step": 1961
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.18338856101036072,
+      "learning_rate": 0.0001857457136130651,
+      "loss": 0.9391,
+      "step": 1962
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.18629801273345947,
+      "learning_rate": 0.00018573146280763324,
+      "loss": 1.0108,
+      "step": 1963
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.18729791045188904,
+      "learning_rate": 0.00018571720542934084,
+      "loss": 0.9739,
+      "step": 1964
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.17371436953544617,
+      "learning_rate": 0.00018570294147928092,
+      "loss": 0.9305,
+      "step": 1965
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1936265081167221,
+      "learning_rate": 0.0001856886709585472,
+      "loss": 1.0033,
+      "step": 1966
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.187389075756073,
+      "learning_rate": 0.00018567439386823367,
+      "loss": 0.9921,
+      "step": 1967
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.19305671751499176,
+      "learning_rate": 0.00018566011020943496,
+      "loss": 0.9698,
+      "step": 1968
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.18552327156066895,
+      "learning_rate": 0.00018564581998324614,
+      "loss": 1.0511,
+      "step": 1969
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.19322232902050018,
+      "learning_rate": 0.00018563152319076286,
+      "loss": 0.9734,
+      "step": 1970
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.18172651529312134,
+      "learning_rate": 0.0001856172198330812,
+      "loss": 0.9552,
+      "step": 1971
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1780477911233902,
+      "learning_rate": 0.00018560290991129777,
+      "loss": 0.9882,
+      "step": 1972
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.18396319448947906,
+      "learning_rate": 0.00018558859342650969,
+      "loss": 0.9748,
+      "step": 1973
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.18518681824207306,
+      "learning_rate": 0.00018557427037981458,
+      "loss": 0.9754,
+      "step": 1974
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.17571696639060974,
+      "learning_rate": 0.00018555994077231055,
+      "loss": 0.9538,
+      "step": 1975
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1853148639202118,
+      "learning_rate": 0.00018554560460509623,
+      "loss": 0.9983,
+      "step": 1976
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.18485432863235474,
+      "learning_rate": 0.00018553126187927072,
+      "loss": 0.9954,
+      "step": 1977
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.19659732282161713,
+      "learning_rate": 0.00018551691259593368,
+      "loss": 1.0263,
+      "step": 1978
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.17942313849925995,
+      "learning_rate": 0.00018550255675618528,
+      "loss": 0.9816,
+      "step": 1979
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.19176405668258667,
+      "learning_rate": 0.00018548819436112603,
+      "loss": 0.9783,
+      "step": 1980
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.20149610936641693,
+      "learning_rate": 0.00018547382541185718,
+      "loss": 0.9438,
+      "step": 1981
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.18515346944332123,
+      "learning_rate": 0.0001854594499094803,
+      "loss": 1.0068,
+      "step": 1982
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.19630250334739685,
+      "learning_rate": 0.00018544506785509758,
+      "loss": 0.992,
+      "step": 1983
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.18869252502918243,
+      "learning_rate": 0.00018543067924981166,
+      "loss": 0.9607,
+      "step": 1984
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2046448290348053,
+      "learning_rate": 0.00018541628409472565,
+      "loss": 1.0241,
+      "step": 1985
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.17943765223026276,
+      "learning_rate": 0.00018540188239094321,
+      "loss": 1.0121,
+      "step": 1986
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.18648383021354675,
+      "learning_rate": 0.00018538747413956853,
+      "loss": 0.9549,
+      "step": 1987
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.18941380083560944,
+      "learning_rate": 0.0001853730593417062,
+      "loss": 1.0078,
+      "step": 1988
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.18444235622882843,
+      "learning_rate": 0.0001853586379984614,
+      "loss": 1.0314,
+      "step": 1989
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.18902845680713654,
+      "learning_rate": 0.00018534421011093982,
+      "loss": 1.014,
+      "step": 1990
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.17776677012443542,
+      "learning_rate": 0.00018532977568024757,
+      "loss": 0.9594,
+      "step": 1991
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1941346377134323,
+      "learning_rate": 0.00018531533470749132,
+      "loss": 1.0107,
+      "step": 1992
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.19353926181793213,
+      "learning_rate": 0.00018530088719377825,
+      "loss": 1.0376,
+      "step": 1993
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.18152490258216858,
+      "learning_rate": 0.000185286433140216,
+      "loss": 0.9123,
+      "step": 1994
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.18493008613586426,
+      "learning_rate": 0.0001852719725479127,
+      "loss": 0.999,
+      "step": 1995
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.18543054163455963,
+      "learning_rate": 0.00018525750541797712,
+      "loss": 0.9819,
+      "step": 1996
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.18111182749271393,
+      "learning_rate": 0.00018524303175151833,
+      "loss": 0.9611,
+      "step": 1997
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.18060384690761566,
+      "learning_rate": 0.00018522855154964605,
+      "loss": 0.9838,
+      "step": 1998
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.18530574440956116,
+      "learning_rate": 0.0001852140648134704,
+      "loss": 1.0593,
+      "step": 1999
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.18004360795021057,
+      "learning_rate": 0.00018519957154410207,
+      "loss": 0.9767,
+      "step": 2000
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.18427857756614685,
+      "learning_rate": 0.0001851850717426523,
+      "loss": 1.0015,
+      "step": 2001
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.18990133702754974,
+      "learning_rate": 0.00018517056541023262,
+      "loss": 1.0502,
+      "step": 2002
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.19328436255455017,
+      "learning_rate": 0.00018515605254795534,
+      "loss": 1.0043,
+      "step": 2003
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1968836933374405,
+      "learning_rate": 0.00018514153315693307,
+      "loss": 1.0026,
+      "step": 2004
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.18505604565143585,
+      "learning_rate": 0.00018512700723827892,
+      "loss": 1.0009,
+      "step": 2005
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.17385464906692505,
+      "learning_rate": 0.00018511247479310669,
+      "loss": 0.9709,
+      "step": 2006
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.185564324259758,
+      "learning_rate": 0.00018509793582253048,
+      "loss": 0.9357,
+      "step": 2007
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.186592698097229,
+      "learning_rate": 0.00018508339032766494,
+      "loss": 0.9939,
+      "step": 2008
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1900765746831894,
+      "learning_rate": 0.00018506883830962534,
+      "loss": 0.9358,
+      "step": 2009
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.175924152135849,
+      "learning_rate": 0.00018505427976952724,
+      "loss": 0.9777,
+      "step": 2010
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.17754964530467987,
+      "learning_rate": 0.00018503971470848688,
+      "loss": 0.9231,
+      "step": 2011
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.18660208582878113,
+      "learning_rate": 0.00018502514312762096,
+      "loss": 0.9708,
+      "step": 2012
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.18159836530685425,
+      "learning_rate": 0.00018501056502804656,
+      "loss": 0.9688,
+      "step": 2013
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1811097264289856,
+      "learning_rate": 0.00018499598041088144,
+      "loss": 0.977,
+      "step": 2014
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.19245150685310364,
+      "learning_rate": 0.00018498138927724376,
+      "loss": 1.0467,
+      "step": 2015
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1905694156885147,
+      "learning_rate": 0.00018496679162825216,
+      "loss": 1.0223,
+      "step": 2016
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1842517852783203,
+      "learning_rate": 0.00018495218746502582,
+      "loss": 0.9497,
+      "step": 2017
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.18859322369098663,
+      "learning_rate": 0.00018493757678868445,
+      "loss": 1.0115,
+      "step": 2018
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.19249236583709717,
+      "learning_rate": 0.00018492295960034815,
+      "loss": 1.002,
+      "step": 2019
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.18580509722232819,
+      "learning_rate": 0.00018490833590113767,
+      "loss": 0.9569,
+      "step": 2020
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.19517533481121063,
+      "learning_rate": 0.00018489370569217415,
+      "loss": 0.9678,
+      "step": 2021
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.17847387492656708,
+      "learning_rate": 0.00018487906897457924,
+      "loss": 1.008,
+      "step": 2022
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.18228110671043396,
+      "learning_rate": 0.00018486442574947511,
+      "loss": 0.9509,
+      "step": 2023
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.18810895085334778,
+      "learning_rate": 0.00018484977601798444,
+      "loss": 0.965,
+      "step": 2024
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.18813060224056244,
+      "learning_rate": 0.0001848351197812304,
+      "loss": 1.026,
+      "step": 2025
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1792432814836502,
+      "learning_rate": 0.00018482045704033663,
+      "loss": 0.9821,
+      "step": 2026
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.19384756684303284,
+      "learning_rate": 0.00018480578779642734,
+      "loss": 0.9905,
+      "step": 2027
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1915934830904007,
+      "learning_rate": 0.00018479111205062715,
+      "loss": 1.0132,
+      "step": 2028
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.21746857464313507,
+      "learning_rate": 0.00018477642980406126,
+      "loss": 0.888,
+      "step": 2029
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1871354579925537,
+      "learning_rate": 0.00018476174105785527,
+      "loss": 1.1023,
+      "step": 2030
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.19254761934280396,
+      "learning_rate": 0.00018474704581313537,
+      "loss": 0.9672,
+      "step": 2031
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.18293356895446777,
+      "learning_rate": 0.0001847323440710282,
+      "loss": 1.0073,
+      "step": 2032
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1897682547569275,
+      "learning_rate": 0.00018471763583266095,
+      "loss": 0.9788,
+      "step": 2033
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.178378164768219,
+      "learning_rate": 0.00018470292109916127,
+      "loss": 0.9124,
+      "step": 2034
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1845291703939438,
+      "learning_rate": 0.00018468819987165725,
+      "loss": 1.0039,
+      "step": 2035
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.18279947340488434,
+      "learning_rate": 0.0001846734721512776,
+      "loss": 0.9861,
+      "step": 2036
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2001655399799347,
+      "learning_rate": 0.00018465873793915141,
+      "loss": 0.9734,
+      "step": 2037
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1830873042345047,
+      "learning_rate": 0.00018464399723640837,
+      "loss": 0.9534,
+      "step": 2038
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.18374669551849365,
+      "learning_rate": 0.00018462925004417862,
+      "loss": 1.0282,
+      "step": 2039
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.18520888686180115,
+      "learning_rate": 0.00018461449636359277,
+      "loss": 0.9944,
+      "step": 2040
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.18403057754039764,
+      "learning_rate": 0.00018459973619578193,
+      "loss": 0.946,
+      "step": 2041
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1820298135280609,
+      "learning_rate": 0.00018458496954187783,
+      "loss": 0.9202,
+      "step": 2042
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1788777858018875,
+      "learning_rate": 0.0001845701964030125,
+      "loss": 1.0047,
+      "step": 2043
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.18801230192184448,
+      "learning_rate": 0.0001845554167803186,
+      "loss": 0.9193,
+      "step": 2044
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.17807801067829132,
+      "learning_rate": 0.0001845406306749293,
+      "loss": 0.9653,
+      "step": 2045
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.19261790812015533,
+      "learning_rate": 0.00018452583808797814,
+      "loss": 0.9817,
+      "step": 2046
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.18605123460292816,
+      "learning_rate": 0.0001845110390205993,
+      "loss": 0.987,
+      "step": 2047
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.17385923862457275,
+      "learning_rate": 0.00018449623347392737,
+      "loss": 0.9361,
+      "step": 2048
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.188616544008255,
+      "learning_rate": 0.00018448142144909747,
+      "loss": 1.0278,
+      "step": 2049
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1917959302663803,
+      "learning_rate": 0.0001844666029472452,
+      "loss": 0.9451,
+      "step": 2050
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.18233361840248108,
+      "learning_rate": 0.00018445177796950669,
+      "loss": 0.9313,
+      "step": 2051
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.19012048840522766,
+      "learning_rate": 0.0001844369465170185,
+      "loss": 1.0405,
+      "step": 2052
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.18029513955116272,
+      "learning_rate": 0.00018442210859091776,
+      "loss": 0.9404,
+      "step": 2053
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.200601264834404,
+      "learning_rate": 0.00018440726419234203,
+      "loss": 1.0396,
+      "step": 2054
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.18682414293289185,
+      "learning_rate": 0.00018439241332242944,
+      "loss": 0.9931,
+      "step": 2055
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.17674268782138824,
+      "learning_rate": 0.00018437755598231856,
+      "loss": 0.9469,
+      "step": 2056
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1858448088169098,
+      "learning_rate": 0.0001843626921731485,
+      "loss": 0.9994,
+      "step": 2057
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.18077892065048218,
+      "learning_rate": 0.00018434782189605877,
+      "loss": 0.9673,
+      "step": 2058
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1898825615644455,
+      "learning_rate": 0.0001843329451521895,
+      "loss": 0.9681,
+      "step": 2059
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.17584487795829773,
+      "learning_rate": 0.00018431806194268126,
+      "loss": 0.9353,
+      "step": 2060
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2292255014181137,
+      "learning_rate": 0.0001843031722686751,
+      "loss": 0.896,
+      "step": 2061
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1748904138803482,
+      "learning_rate": 0.0001842882761313126,
+      "loss": 0.8746,
+      "step": 2062
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.17828615009784698,
+      "learning_rate": 0.00018427337353173578,
+      "loss": 0.9323,
+      "step": 2063
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.18547767400741577,
+      "learning_rate": 0.00018425846447108718,
+      "loss": 0.9273,
+      "step": 2064
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.18525430560112,
+      "learning_rate": 0.00018424354895050994,
+      "loss": 1.0056,
+      "step": 2065
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.17384150624275208,
+      "learning_rate": 0.00018422862697114754,
+      "loss": 0.9105,
+      "step": 2066
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.18717877566814423,
+      "learning_rate": 0.00018421369853414401,
+      "loss": 0.9566,
+      "step": 2067
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.19069983065128326,
+      "learning_rate": 0.00018419876364064388,
+      "loss": 1.0089,
+      "step": 2068
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.17962156236171722,
+      "learning_rate": 0.0001841838222917922,
+      "loss": 0.9166,
+      "step": 2069
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.18591444194316864,
+      "learning_rate": 0.00018416887448873453,
+      "loss": 0.9007,
+      "step": 2070
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1889745593070984,
+      "learning_rate": 0.00018415392023261683,
+      "loss": 1.0553,
+      "step": 2071
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.18329787254333496,
+      "learning_rate": 0.00018413895952458563,
+      "loss": 0.9331,
+      "step": 2072
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.18456338346004486,
+      "learning_rate": 0.0001841239923657879,
+      "loss": 0.9649,
+      "step": 2073
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2080213725566864,
+      "learning_rate": 0.0001841090187573712,
+      "loss": 0.973,
+      "step": 2074
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.17959704995155334,
+      "learning_rate": 0.00018409403870048356,
+      "loss": 0.9257,
+      "step": 2075
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.18018606305122375,
+      "learning_rate": 0.00018407905219627337,
+      "loss": 0.9597,
+      "step": 2076
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1891738921403885,
+      "learning_rate": 0.00018406405924588967,
+      "loss": 0.9433,
+      "step": 2077
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.18608370423316956,
+      "learning_rate": 0.00018404905985048192,
+      "loss": 0.9511,
+      "step": 2078
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.18085582554340363,
+      "learning_rate": 0.00018403405401120013,
+      "loss": 1.0328,
+      "step": 2079
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.19775338470935822,
+      "learning_rate": 0.00018401904172919475,
+      "loss": 0.8966,
+      "step": 2080
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1780966818332672,
+      "learning_rate": 0.00018400402300561673,
+      "loss": 0.9511,
+      "step": 2081
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.19541499018669128,
+      "learning_rate": 0.00018398899784161752,
+      "loss": 1.0246,
+      "step": 2082
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.18545083701610565,
+      "learning_rate": 0.0001839739662383491,
+      "loss": 1.0099,
+      "step": 2083
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.17320199310779572,
+      "learning_rate": 0.00018395892819696389,
+      "loss": 0.9432,
+      "step": 2084
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.18442998826503754,
+      "learning_rate": 0.00018394388371861483,
+      "loss": 1.0097,
+      "step": 2085
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.17621727287769318,
+      "learning_rate": 0.00018392883280445537,
+      "loss": 1.0116,
+      "step": 2086
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.18226811289787292,
+      "learning_rate": 0.00018391377545563938,
+      "loss": 1.0203,
+      "step": 2087
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.18577659130096436,
+      "learning_rate": 0.00018389871167332134,
+      "loss": 0.9453,
+      "step": 2088
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.18943215906620026,
+      "learning_rate": 0.00018388364145865613,
+      "loss": 0.9726,
+      "step": 2089
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.18668906390666962,
+      "learning_rate": 0.00018386856481279916,
+      "loss": 0.9741,
+      "step": 2090
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.17848534882068634,
+      "learning_rate": 0.00018385348173690632,
+      "loss": 0.9991,
+      "step": 2091
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.18247073888778687,
+      "learning_rate": 0.000183838392232134,
+      "loss": 0.9163,
+      "step": 2092
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1850193589925766,
+      "learning_rate": 0.00018382329629963907,
+      "loss": 0.9621,
+      "step": 2093
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.18309763073921204,
+      "learning_rate": 0.00018380819394057894,
+      "loss": 0.9825,
+      "step": 2094
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.19286571443080902,
+      "learning_rate": 0.00018379308515611146,
+      "loss": 0.994,
+      "step": 2095
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1862298846244812,
+      "learning_rate": 0.00018377796994739498,
+      "loss": 0.961,
+      "step": 2096
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.18532899022102356,
+      "learning_rate": 0.00018376284831558834,
+      "loss": 1.0029,
+      "step": 2097
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.18360939621925354,
+      "learning_rate": 0.00018374772026185095,
+      "loss": 0.9917,
+      "step": 2098
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.18725305795669556,
+      "learning_rate": 0.00018373258578734255,
+      "loss": 0.9478,
+      "step": 2099
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.18553733825683594,
+      "learning_rate": 0.00018371744489322357,
+      "loss": 0.9833,
+      "step": 2100
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.17913220822811127,
+      "learning_rate": 0.00018370229758065478,
+      "loss": 0.9987,
+      "step": 2101
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1810830533504486,
+      "learning_rate": 0.0001836871438507975,
+      "loss": 0.9334,
+      "step": 2102
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.19181925058364868,
+      "learning_rate": 0.00018367198370481354,
+      "loss": 0.9945,
+      "step": 2103
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1797618567943573,
+      "learning_rate": 0.00018365681714386516,
+      "loss": 0.9155,
+      "step": 2104
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.18634293973445892,
+      "learning_rate": 0.0001836416441691152,
+      "loss": 0.9568,
+      "step": 2105
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1783328354358673,
+      "learning_rate": 0.00018362646478172696,
+      "loss": 0.9257,
+      "step": 2106
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.19061227142810822,
+      "learning_rate": 0.00018361127898286416,
+      "loss": 1.0015,
+      "step": 2107
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.18546177446842194,
+      "learning_rate": 0.00018359608677369108,
+      "loss": 0.9977,
+      "step": 2108
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.18221919238567352,
+      "learning_rate": 0.00018358088815537246,
+      "loss": 0.9498,
+      "step": 2109
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.18326273560523987,
+      "learning_rate": 0.0001835656831290736,
+      "loss": 0.9702,
+      "step": 2110
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1829109638929367,
+      "learning_rate": 0.0001835504716959602,
+      "loss": 0.9142,
+      "step": 2111
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.18340815603733063,
+      "learning_rate": 0.0001835352538571985,
+      "loss": 0.9222,
+      "step": 2112
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.19066669046878815,
+      "learning_rate": 0.0001835200296139552,
+      "loss": 0.9965,
+      "step": 2113
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.18047630786895752,
+      "learning_rate": 0.00018350479896739756,
+      "loss": 0.9369,
+      "step": 2114
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1809929460287094,
+      "learning_rate": 0.00018348956191869328,
+      "loss": 0.9247,
+      "step": 2115
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.18403151631355286,
+      "learning_rate": 0.00018347431846901048,
+      "loss": 0.9744,
+      "step": 2116
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.19564421474933624,
+      "learning_rate": 0.0001834590686195179,
+      "loss": 0.9778,
+      "step": 2117
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1840590536594391,
+      "learning_rate": 0.00018344381237138472,
+      "loss": 0.9687,
+      "step": 2118
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.18494993448257446,
+      "learning_rate": 0.00018342854972578062,
+      "loss": 0.9746,
+      "step": 2119
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.18578451871871948,
+      "learning_rate": 0.00018341328068387572,
+      "loss": 1.0011,
+      "step": 2120
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.18412530422210693,
+      "learning_rate": 0.00018339800524684068,
+      "loss": 0.907,
+      "step": 2121
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.18739020824432373,
+      "learning_rate": 0.00018338272341584662,
+      "loss": 0.961,
+      "step": 2122
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1956738978624344,
+      "learning_rate": 0.0001833674351920652,
+      "loss": 0.9948,
+      "step": 2123
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.18302105367183685,
+      "learning_rate": 0.00018335214057666854,
+      "loss": 0.9506,
+      "step": 2124
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.18477575480937958,
+      "learning_rate": 0.0001833368395708292,
+      "loss": 0.9732,
+      "step": 2125
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.18320126831531525,
+      "learning_rate": 0.00018332153217572034,
+      "loss": 0.9378,
+      "step": 2126
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.18950019776821136,
+      "learning_rate": 0.00018330621839251547,
+      "loss": 0.9409,
+      "step": 2127
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1776246428489685,
+      "learning_rate": 0.00018329089822238874,
+      "loss": 0.9854,
+      "step": 2128
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.19237257540225983,
+      "learning_rate": 0.0001832755716665147,
+      "loss": 0.958,
+      "step": 2129
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.19101490080356598,
+      "learning_rate": 0.0001832602387260684,
+      "loss": 0.9659,
+      "step": 2130
+    },
+    {
+      "epoch": 0.75,
+      "eval_loss": 0.9828736782073975,
+      "eval_runtime": 680.0071,
+      "eval_samples_per_second": 10.112,
+      "eval_steps_per_second": 5.056,
+      "step": 2130
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.19474919140338898,
+      "learning_rate": 0.00018324489940222538,
+      "loss": 1.0003,
+      "step": 2131
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.19254593551158905,
+      "learning_rate": 0.00018322955369616165,
+      "loss": 0.9636,
+      "step": 2132
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.18535253405570984,
+      "learning_rate": 0.00018321420160905382,
+      "loss": 0.9502,
+      "step": 2133
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.18019430339336395,
+      "learning_rate": 0.00018319884314207878,
+      "loss": 1.0317,
+      "step": 2134
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.18415750563144684,
+      "learning_rate": 0.00018318347829641412,
+      "loss": 0.9957,
+      "step": 2135
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.18982534110546112,
+      "learning_rate": 0.00018316810707323783,
+      "loss": 0.9908,
+      "step": 2136
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.19444747269153595,
+      "learning_rate": 0.00018315272947372837,
+      "loss": 1.0274,
+      "step": 2137
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.19093841314315796,
+      "learning_rate": 0.0001831373454990647,
+      "loss": 1.0041,
+      "step": 2138
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.18283675611019135,
+      "learning_rate": 0.00018312195515042632,
+      "loss": 0.9485,
+      "step": 2139
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.17775124311447144,
+      "learning_rate": 0.00018310655842899312,
+      "loss": 0.9003,
+      "step": 2140
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.20296508073806763,
+      "learning_rate": 0.00018309115533594558,
+      "loss": 1.0346,
+      "step": 2141
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.191764697432518,
+      "learning_rate": 0.00018307574587246459,
+      "loss": 0.9972,
+      "step": 2142
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.19153252243995667,
+      "learning_rate": 0.0001830603300397316,
+      "loss": 0.9488,
+      "step": 2143
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1953708678483963,
+      "learning_rate": 0.00018304490783892848,
+      "loss": 1.1073,
+      "step": 2144
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.20149077475070953,
+      "learning_rate": 0.00018302947927123766,
+      "loss": 0.9729,
+      "step": 2145
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.19051818549633026,
+      "learning_rate": 0.00018301404433784194,
+      "loss": 1.0179,
+      "step": 2146
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.18015353381633759,
+      "learning_rate": 0.0001829986030399248,
+      "loss": 0.862,
+      "step": 2147
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.19477394223213196,
+      "learning_rate": 0.00018298315537866998,
+      "loss": 1.0168,
+      "step": 2148
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1973867565393448,
+      "learning_rate": 0.0001829677013552619,
+      "loss": 1.0129,
+      "step": 2149
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.18476396799087524,
+      "learning_rate": 0.00018295224097088534,
+      "loss": 0.9461,
+      "step": 2150
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.18751783668994904,
+      "learning_rate": 0.00018293677422672567,
+      "loss": 1.0069,
+      "step": 2151
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.19476616382598877,
+      "learning_rate": 0.00018292130112396865,
+      "loss": 0.9646,
+      "step": 2152
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1982491910457611,
+      "learning_rate": 0.00018290582166380058,
+      "loss": 1.0004,
+      "step": 2153
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.18301215767860413,
+      "learning_rate": 0.0001828903358474082,
+      "loss": 0.9201,
+      "step": 2154
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.19771361351013184,
+      "learning_rate": 0.00018287484367597888,
+      "loss": 0.9588,
+      "step": 2155
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.18548940122127533,
+      "learning_rate": 0.00018285934515070026,
+      "loss": 0.9673,
+      "step": 2156
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1889079064130783,
+      "learning_rate": 0.00018284384027276068,
+      "loss": 1.0334,
+      "step": 2157
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.20975716412067413,
+      "learning_rate": 0.00018282832904334883,
+      "loss": 1.0518,
+      "step": 2158
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.19074541330337524,
+      "learning_rate": 0.0001828128114636539,
+      "loss": 0.9866,
+      "step": 2159
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.20263442397117615,
+      "learning_rate": 0.00018279728753486558,
+      "loss": 1.0949,
+      "step": 2160
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.19469326734542847,
+      "learning_rate": 0.00018278175725817413,
+      "loss": 0.9735,
+      "step": 2161
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.19660715758800507,
+      "learning_rate": 0.00018276622063477017,
+      "loss": 0.9821,
+      "step": 2162
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.18767257034778595,
+      "learning_rate": 0.00018275067766584488,
+      "loss": 0.9505,
+      "step": 2163
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.17904675006866455,
+      "learning_rate": 0.0001827351283525899,
+      "loss": 0.965,
+      "step": 2164
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.18472503125667572,
+      "learning_rate": 0.00018271957269619737,
+      "loss": 0.9939,
+      "step": 2165
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.18497197329998016,
+      "learning_rate": 0.00018270401069785992,
+      "loss": 1.0184,
+      "step": 2166
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1877606064081192,
+      "learning_rate": 0.00018268844235877062,
+      "loss": 0.923,
+      "step": 2167
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.18791928887367249,
+      "learning_rate": 0.00018267286768012313,
+      "loss": 1.0198,
+      "step": 2168
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1900901347398758,
+      "learning_rate": 0.00018265728666311146,
+      "loss": 1.0139,
+      "step": 2169
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.18016625940799713,
+      "learning_rate": 0.00018264169930893022,
+      "loss": 0.9624,
+      "step": 2170
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1885920614004135,
+      "learning_rate": 0.00018262610561877444,
+      "loss": 0.9212,
+      "step": 2171
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.19056080281734467,
+      "learning_rate": 0.00018261050559383968,
+      "loss": 0.9597,
+      "step": 2172
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.19492998719215393,
+      "learning_rate": 0.00018259489923532195,
+      "loss": 1.0138,
+      "step": 2173
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.17653758823871613,
+      "learning_rate": 0.00018257928654441777,
+      "loss": 0.9123,
+      "step": 2174
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.17693361639976501,
+      "learning_rate": 0.00018256366752232407,
+      "loss": 0.9263,
+      "step": 2175
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.18627983331680298,
+      "learning_rate": 0.00018254804217023843,
+      "loss": 0.9921,
+      "step": 2176
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.18731026351451874,
+      "learning_rate": 0.00018253241048935874,
+      "loss": 0.9612,
+      "step": 2177
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.19288547337055206,
+      "learning_rate": 0.00018251677248088348,
+      "loss": 1.0085,
+      "step": 2178
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.18385879695415497,
+      "learning_rate": 0.00018250112814601158,
+      "loss": 0.9856,
+      "step": 2179
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1872677505016327,
+      "learning_rate": 0.00018248547748594244,
+      "loss": 1.0597,
+      "step": 2180
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.18108715116977692,
+      "learning_rate": 0.000182469820501876,
+      "loss": 0.9922,
+      "step": 2181
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.18408574163913727,
+      "learning_rate": 0.00018245415719501265,
+      "loss": 0.9816,
+      "step": 2182
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.19224940240383148,
+      "learning_rate": 0.00018243848756655323,
+      "loss": 1.0338,
+      "step": 2183
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.18560826778411865,
+      "learning_rate": 0.0001824228116176991,
+      "loss": 1.0483,
+      "step": 2184
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.18587149679660797,
+      "learning_rate": 0.00018240712934965218,
+      "loss": 0.9734,
+      "step": 2185
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.19045782089233398,
+      "learning_rate": 0.0001823914407636147,
+      "loss": 1.0319,
+      "step": 2186
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.17874370515346527,
+      "learning_rate": 0.0001823757458607895,
+      "loss": 0.9584,
+      "step": 2187
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.19332368671894073,
+      "learning_rate": 0.00018236004464237992,
+      "loss": 1.0145,
+      "step": 2188
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.18693716824054718,
+      "learning_rate": 0.0001823443371095897,
+      "loss": 1.0324,
+      "step": 2189
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1814510077238083,
+      "learning_rate": 0.0001823286232636231,
+      "loss": 1.0154,
+      "step": 2190
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1795709729194641,
+      "learning_rate": 0.0001823129031056849,
+      "loss": 0.9817,
+      "step": 2191
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.18347346782684326,
+      "learning_rate": 0.00018229717663698032,
+      "loss": 0.955,
+      "step": 2192
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.190670445561409,
+      "learning_rate": 0.0001822814438587151,
+      "loss": 0.9807,
+      "step": 2193
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.18033914268016815,
+      "learning_rate": 0.00018226570477209534,
+      "loss": 1.0029,
+      "step": 2194
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.18933604657649994,
+      "learning_rate": 0.00018224995937832786,
+      "loss": 0.9991,
+      "step": 2195
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1753634661436081,
+      "learning_rate": 0.00018223420767861978,
+      "loss": 0.9316,
+      "step": 2196
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.18977148830890656,
+      "learning_rate": 0.0001822184496741787,
+      "loss": 0.9702,
+      "step": 2197
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.18761374056339264,
+      "learning_rate": 0.00018220268536621283,
+      "loss": 1.0516,
+      "step": 2198
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.18400248885154724,
+      "learning_rate": 0.00018218691475593073,
+      "loss": 0.962,
+      "step": 2199
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1761545091867447,
+      "learning_rate": 0.00018217113784454151,
+      "loss": 0.9558,
+      "step": 2200
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.18742233514785767,
+      "learning_rate": 0.0001821553546332548,
+      "loss": 0.967,
+      "step": 2201
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.18793874979019165,
+      "learning_rate": 0.00018213956512328064,
+      "loss": 1.0253,
+      "step": 2202
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.18570399284362793,
+      "learning_rate": 0.00018212376931582956,
+      "loss": 0.9804,
+      "step": 2203
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.18871964514255524,
+      "learning_rate": 0.0001821079672121126,
+      "loss": 1.021,
+      "step": 2204
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.19510523974895477,
+      "learning_rate": 0.0001820921588133413,
+      "loss": 1.0137,
+      "step": 2205
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.18368487060070038,
+      "learning_rate": 0.00018207634412072764,
+      "loss": 0.9795,
+      "step": 2206
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1859757900238037,
+      "learning_rate": 0.00018206052313548413,
+      "loss": 0.9345,
+      "step": 2207
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2072616070508957,
+      "learning_rate": 0.00018204469585882365,
+      "loss": 1.0181,
+      "step": 2208
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1866096556186676,
+      "learning_rate": 0.00018202886229195975,
+      "loss": 0.9891,
+      "step": 2209
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.18876902759075165,
+      "learning_rate": 0.0001820130224361063,
+      "loss": 1.0037,
+      "step": 2210
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.18180185556411743,
+      "learning_rate": 0.00018199717629247773,
+      "loss": 1.0389,
+      "step": 2211
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.19424739480018616,
+      "learning_rate": 0.0001819813238622889,
+      "loss": 0.9684,
+      "step": 2212
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.188624769449234,
+      "learning_rate": 0.00018196546514675523,
+      "loss": 0.956,
+      "step": 2213
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1843746155500412,
+      "learning_rate": 0.00018194960014709257,
+      "loss": 0.9866,
+      "step": 2214
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.18615764379501343,
+      "learning_rate": 0.0001819337288645172,
+      "loss": 0.9164,
+      "step": 2215
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.17863336205482483,
+      "learning_rate": 0.000181917851300246,
+      "loss": 1.0111,
+      "step": 2216
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.20061299204826355,
+      "learning_rate": 0.00018190196745549627,
+      "loss": 0.9398,
+      "step": 2217
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.18024347722530365,
+      "learning_rate": 0.00018188607733148578,
+      "loss": 0.9883,
+      "step": 2218
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.17821189761161804,
+      "learning_rate": 0.00018187018092943278,
+      "loss": 0.9278,
+      "step": 2219
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1844683438539505,
+      "learning_rate": 0.00018185427825055602,
+      "loss": 1.044,
+      "step": 2220
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.19516977667808533,
+      "learning_rate": 0.00018183836929607473,
+      "loss": 1.0143,
+      "step": 2221
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.192205011844635,
+      "learning_rate": 0.0001818224540672086,
+      "loss": 1.0169,
+      "step": 2222
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.20235204696655273,
+      "learning_rate": 0.00018180653256517787,
+      "loss": 1.0375,
+      "step": 2223
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.18561050295829773,
+      "learning_rate": 0.0001817906047912032,
+      "loss": 0.923,
+      "step": 2224
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.19041159749031067,
+      "learning_rate": 0.0001817746707465057,
+      "loss": 1.0645,
+      "step": 2225
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.18981195986270905,
+      "learning_rate": 0.000181758730432307,
+      "loss": 0.935,
+      "step": 2226
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.187282532453537,
+      "learning_rate": 0.00018174278384982926,
+      "loss": 1.0661,
+      "step": 2227
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.18467959761619568,
+      "learning_rate": 0.00018172683100029503,
+      "loss": 0.9665,
+      "step": 2228
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.19500401616096497,
+      "learning_rate": 0.00018171087188492742,
+      "loss": 1.0129,
+      "step": 2229
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.19473527371883392,
+      "learning_rate": 0.00018169490650494995,
+      "loss": 1.0049,
+      "step": 2230
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.19505290687084198,
+      "learning_rate": 0.0001816789348615867,
+      "loss": 1.0335,
+      "step": 2231
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1958623081445694,
+      "learning_rate": 0.0001816629569560621,
+      "loss": 1.0676,
+      "step": 2232
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.18890038132667542,
+      "learning_rate": 0.00018164697278960122,
+      "loss": 0.9362,
+      "step": 2233
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.18806125223636627,
+      "learning_rate": 0.00018163098236342952,
+      "loss": 0.9476,
+      "step": 2234
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.19116836786270142,
+      "learning_rate": 0.0001816149856787729,
+      "loss": 0.9534,
+      "step": 2235
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1891220211982727,
+      "learning_rate": 0.00018159898273685787,
+      "loss": 0.9905,
+      "step": 2236
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.18333634734153748,
+      "learning_rate": 0.0001815829735389113,
+      "loss": 1.0209,
+      "step": 2237
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.18993903696537018,
+      "learning_rate": 0.00018156695808616062,
+      "loss": 0.9481,
+      "step": 2238
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1848689764738083,
+      "learning_rate": 0.00018155093637983367,
+      "loss": 0.9124,
+      "step": 2239
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.19026519358158112,
+      "learning_rate": 0.00018153490842115879,
+      "loss": 1.0066,
+      "step": 2240
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.18695740401744843,
+      "learning_rate": 0.00018151887421136484,
+      "loss": 0.991,
+      "step": 2241
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.18972289562225342,
+      "learning_rate": 0.00018150283375168114,
+      "loss": 0.9192,
+      "step": 2242
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.19145484268665314,
+      "learning_rate": 0.00018148678704333745,
+      "loss": 1.0402,
+      "step": 2243
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.19382283091545105,
+      "learning_rate": 0.00018147073408756405,
+      "loss": 0.9467,
+      "step": 2244
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.19206614792346954,
+      "learning_rate": 0.0001814546748855917,
+      "loss": 0.9458,
+      "step": 2245
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.18128708004951477,
+      "learning_rate": 0.0001814386094386516,
+      "loss": 0.93,
+      "step": 2246
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.17813989520072937,
+      "learning_rate": 0.0001814225377479755,
+      "loss": 0.9856,
+      "step": 2247
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.18716156482696533,
+      "learning_rate": 0.00018140645981479554,
+      "loss": 0.9642,
+      "step": 2248
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.20171204209327698,
+      "learning_rate": 0.0001813903756403444,
+      "loss": 0.9714,
+      "step": 2249
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.18735407292842865,
+      "learning_rate": 0.0001813742852258552,
+      "loss": 0.9901,
+      "step": 2250
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.18513000011444092,
+      "learning_rate": 0.00018135818857256163,
+      "loss": 0.926,
+      "step": 2251
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.19058287143707275,
+      "learning_rate": 0.00018134208568169774,
+      "loss": 0.9119,
+      "step": 2252
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1841873973608017,
+      "learning_rate": 0.00018132597655449805,
+      "loss": 0.972,
+      "step": 2253
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1925683617591858,
+      "learning_rate": 0.00018130986119219774,
+      "loss": 1.0006,
+      "step": 2254
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.18620803952217102,
+      "learning_rate": 0.00018129373959603222,
+      "loss": 1.022,
+      "step": 2255
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1849374920129776,
+      "learning_rate": 0.00018127761176723757,
+      "loss": 0.9537,
+      "step": 2256
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.19724594056606293,
+      "learning_rate": 0.00018126147770705027,
+      "loss": 0.9363,
+      "step": 2257
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1982530653476715,
+      "learning_rate": 0.00018124533741670725,
+      "loss": 1.0424,
+      "step": 2258
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.18769797682762146,
+      "learning_rate": 0.00018122919089744603,
+      "loss": 1.006,
+      "step": 2259
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.18896335363388062,
+      "learning_rate": 0.00018121303815050449,
+      "loss": 1.0598,
+      "step": 2260
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1882297247648239,
+      "learning_rate": 0.00018119687917712097,
+      "loss": 0.9607,
+      "step": 2261
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.18507346510887146,
+      "learning_rate": 0.00018118071397853447,
+      "loss": 1.0024,
+      "step": 2262
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.18987086415290833,
+      "learning_rate": 0.00018116454255598423,
+      "loss": 0.9818,
+      "step": 2263
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.18397410213947296,
+      "learning_rate": 0.00018114836491071014,
+      "loss": 1.0325,
+      "step": 2264
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.18793408572673798,
+      "learning_rate": 0.00018113218104395248,
+      "loss": 0.9989,
+      "step": 2265
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1876463145017624,
+      "learning_rate": 0.00018111599095695205,
+      "loss": 0.9804,
+      "step": 2266
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.17934471368789673,
+      "learning_rate": 0.00018109979465095013,
+      "loss": 0.9507,
+      "step": 2267
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.17715683579444885,
+      "learning_rate": 0.00018108359212718841,
+      "loss": 0.9766,
+      "step": 2268
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.17394061386585236,
+      "learning_rate": 0.00018106738338690917,
+      "loss": 0.9053,
+      "step": 2269
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.17950716614723206,
+      "learning_rate": 0.00018105116843135504,
+      "loss": 0.9292,
+      "step": 2270
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.18406367301940918,
+      "learning_rate": 0.00018103494726176922,
+      "loss": 0.9567,
+      "step": 2271
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.18885083496570587,
+      "learning_rate": 0.00018101871987939535,
+      "loss": 0.9399,
+      "step": 2272
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.19340074062347412,
+      "learning_rate": 0.0001810024862854775,
+      "loss": 1.0206,
+      "step": 2273
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.182267963886261,
+      "learning_rate": 0.0001809862464812604,
+      "loss": 0.9908,
+      "step": 2274
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1867702603340149,
+      "learning_rate": 0.00018097000046798898,
+      "loss": 0.9909,
+      "step": 2275
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.18310391902923584,
+      "learning_rate": 0.0001809537482469089,
+      "loss": 1.0097,
+      "step": 2276
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1698244959115982,
+      "learning_rate": 0.00018093748981926609,
+      "loss": 0.906,
+      "step": 2277
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1908068060874939,
+      "learning_rate": 0.00018092122518630709,
+      "loss": 0.9443,
+      "step": 2278
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.18088297545909882,
+      "learning_rate": 0.00018090495434927893,
+      "loss": 0.9008,
+      "step": 2279
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.18029263615608215,
+      "learning_rate": 0.00018088867730942898,
+      "loss": 1.0218,
+      "step": 2280
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.18683558702468872,
+      "learning_rate": 0.00018087239406800522,
+      "loss": 1.0452,
+      "step": 2281
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.18489494919776917,
+      "learning_rate": 0.00018085610462625603,
+      "loss": 0.9451,
+      "step": 2282
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1905013769865036,
+      "learning_rate": 0.00018083980898543032,
+      "loss": 1.015,
+      "step": 2283
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.18374451994895935,
+      "learning_rate": 0.00018082350714677739,
+      "loss": 0.9872,
+      "step": 2284
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.19135090708732605,
+      "learning_rate": 0.0001808071991115471,
+      "loss": 0.9376,
+      "step": 2285
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.19377972185611725,
+      "learning_rate": 0.00018079088488098978,
+      "loss": 1.1116,
+      "step": 2286
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.18835166096687317,
+      "learning_rate": 0.0001807745644563562,
+      "loss": 0.9739,
+      "step": 2287
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1849549114704132,
+      "learning_rate": 0.00018075823783889755,
+      "loss": 1.0043,
+      "step": 2288
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1912752389907837,
+      "learning_rate": 0.00018074190502986563,
+      "loss": 0.9567,
+      "step": 2289
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1859811693429947,
+      "learning_rate": 0.00018072556603051265,
+      "loss": 0.9896,
+      "step": 2290
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.17959462106227875,
+      "learning_rate": 0.00018070922084209123,
+      "loss": 0.9897,
+      "step": 2291
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.18935257196426392,
+      "learning_rate": 0.00018069286946585455,
+      "loss": 0.9496,
+      "step": 2292
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.19748300313949585,
+      "learning_rate": 0.00018067651190305627,
+      "loss": 0.9379,
+      "step": 2293
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1971241533756256,
+      "learning_rate": 0.00018066014815495046,
+      "loss": 1.0008,
+      "step": 2294
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.18202240765094757,
+      "learning_rate": 0.00018064377822279168,
+      "loss": 0.9954,
+      "step": 2295
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.19593803584575653,
+      "learning_rate": 0.000180627402107835,
+      "loss": 1.0572,
+      "step": 2296
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.18760298192501068,
+      "learning_rate": 0.00018061101981133596,
+      "loss": 0.9804,
+      "step": 2297
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1931912750005722,
+      "learning_rate": 0.00018059463133455051,
+      "loss": 0.9814,
+      "step": 2298
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.17908045649528503,
+      "learning_rate": 0.00018057823667873518,
+      "loss": 0.8676,
+      "step": 2299
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.18545086681842804,
+      "learning_rate": 0.0001805618358451469,
+      "loss": 1.0557,
+      "step": 2300
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.19220340251922607,
+      "learning_rate": 0.00018054542883504305,
+      "loss": 0.9436,
+      "step": 2301
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.18264839053153992,
+      "learning_rate": 0.00018052901564968158,
+      "loss": 1.0022,
+      "step": 2302
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.18435068428516388,
+      "learning_rate": 0.00018051259629032083,
+      "loss": 0.9704,
+      "step": 2303
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1799696683883667,
+      "learning_rate": 0.00018049617075821962,
+      "loss": 0.8971,
+      "step": 2304
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.18429692089557648,
+      "learning_rate": 0.00018047973905463728,
+      "loss": 0.9938,
+      "step": 2305
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.18325746059417725,
+      "learning_rate": 0.0001804633011808336,
+      "loss": 0.9852,
+      "step": 2306
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.18153943121433258,
+      "learning_rate": 0.00018044685713806881,
+      "loss": 0.9657,
+      "step": 2307
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.18153220415115356,
+      "learning_rate": 0.00018043040692760368,
+      "loss": 0.9414,
+      "step": 2308
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1895730197429657,
+      "learning_rate": 0.00018041395055069944,
+      "loss": 0.9586,
+      "step": 2309
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.19165296852588654,
+      "learning_rate": 0.00018039748800861768,
+      "loss": 1.0116,
+      "step": 2310
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1748407781124115,
+      "learning_rate": 0.00018038101930262064,
+      "loss": 0.9276,
+      "step": 2311
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.18681953847408295,
+      "learning_rate": 0.00018036454443397086,
+      "loss": 0.976,
+      "step": 2312
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.17880845069885254,
+      "learning_rate": 0.00018034806340393153,
+      "loss": 0.9367,
+      "step": 2313
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.19864773750305176,
+      "learning_rate": 0.00018033157621376612,
+      "loss": 0.9697,
+      "step": 2314
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.19031883776187897,
+      "learning_rate": 0.00018031508286473874,
+      "loss": 1.0077,
+      "step": 2315
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.18201903998851776,
+      "learning_rate": 0.0001802985833581139,
+      "loss": 0.9863,
+      "step": 2316
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.19363977015018463,
+      "learning_rate": 0.0001802820776951565,
+      "loss": 1.0851,
+      "step": 2317
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.19829228520393372,
+      "learning_rate": 0.0001802655658771321,
+      "loss": 0.9617,
+      "step": 2318
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.17512492835521698,
+      "learning_rate": 0.0001802490479053066,
+      "loss": 0.9007,
+      "step": 2319
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.18068177998065948,
+      "learning_rate": 0.00018023252378094635,
+      "loss": 0.9897,
+      "step": 2320
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.18359632790088654,
+      "learning_rate": 0.0001802159935053183,
+      "loss": 1.0025,
+      "step": 2321
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.18495793640613556,
+      "learning_rate": 0.00018019945707968972,
+      "loss": 0.9347,
+      "step": 2322
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.18118029832839966,
+      "learning_rate": 0.00018018291450532848,
+      "loss": 0.9876,
+      "step": 2323
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.193434476852417,
+      "learning_rate": 0.00018016636578350282,
+      "loss": 0.9605,
+      "step": 2324
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.19045786559581757,
+      "learning_rate": 0.00018014981091548152,
+      "loss": 0.9737,
+      "step": 2325
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.18402209877967834,
+      "learning_rate": 0.00018013324990253385,
+      "loss": 0.967,
+      "step": 2326
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.19088177382946014,
+      "learning_rate": 0.00018011668274592942,
+      "loss": 0.9235,
+      "step": 2327
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.19194436073303223,
+      "learning_rate": 0.00018010010944693848,
+      "loss": 0.9763,
+      "step": 2328
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1945270448923111,
+      "learning_rate": 0.00018008353000683163,
+      "loss": 0.932,
+      "step": 2329
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.38292306661605835,
+      "learning_rate": 0.00018006694442687998,
+      "loss": 0.9527,
+      "step": 2330
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.19509528577327728,
+      "learning_rate": 0.00018005035270835516,
+      "loss": 0.9853,
+      "step": 2331
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.19551971554756165,
+      "learning_rate": 0.00018003375485252919,
+      "loss": 0.9776,
+      "step": 2332
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.17869053781032562,
+      "learning_rate": 0.00018001715086067457,
+      "loss": 0.9142,
+      "step": 2333
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1891433149576187,
+      "learning_rate": 0.00018000054073406436,
+      "loss": 0.9876,
+      "step": 2334
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1875545084476471,
+      "learning_rate": 0.00017998392447397197,
+      "loss": 1.0268,
+      "step": 2335
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.19381293654441833,
+      "learning_rate": 0.00017996730208167132,
+      "loss": 0.9694,
+      "step": 2336
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.20120295882225037,
+      "learning_rate": 0.00017995067355843687,
+      "loss": 1.0433,
+      "step": 2337
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1825202852487564,
+      "learning_rate": 0.0001799340389055435,
+      "loss": 0.9054,
+      "step": 2338
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.18435628712177277,
+      "learning_rate": 0.00017991739812426652,
+      "loss": 0.9426,
+      "step": 2339
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.18861481547355652,
+      "learning_rate": 0.00017990075121588172,
+      "loss": 0.989,
+      "step": 2340
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1839429885149002,
+      "learning_rate": 0.00017988409818166546,
+      "loss": 0.9547,
+      "step": 2341
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.18472762405872345,
+      "learning_rate": 0.00017986743902289446,
+      "loss": 0.9353,
+      "step": 2342
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.18218994140625,
+      "learning_rate": 0.00017985077374084592,
+      "loss": 0.9985,
+      "step": 2343
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.18237660825252533,
+      "learning_rate": 0.00017983410233679754,
+      "loss": 0.9403,
+      "step": 2344
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.32768821716308594,
+      "learning_rate": 0.0001798174248120275,
+      "loss": 0.9062,
+      "step": 2345
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.19033341109752655,
+      "learning_rate": 0.00017980074116781447,
+      "loss": 0.9599,
+      "step": 2346
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1890161782503128,
+      "learning_rate": 0.00017978405140543747,
+      "loss": 1.0005,
+      "step": 2347
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.18877194821834564,
+      "learning_rate": 0.00017976735552617613,
+      "loss": 0.9992,
+      "step": 2348
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.18344587087631226,
+      "learning_rate": 0.00017975065353131048,
+      "loss": 0.9636,
+      "step": 2349
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.19172963500022888,
+      "learning_rate": 0.000179733945422121,
+      "loss": 0.9603,
+      "step": 2350
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1853189468383789,
+      "learning_rate": 0.00017971723119988872,
+      "loss": 0.9867,
+      "step": 2351
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.18967324495315552,
+      "learning_rate": 0.00017970051086589503,
+      "loss": 0.9997,
+      "step": 2352
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1847553700208664,
+      "learning_rate": 0.0001796837844214219,
+      "loss": 0.9851,
+      "step": 2353
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.19326816499233246,
+      "learning_rate": 0.00017966705186775166,
+      "loss": 0.9813,
+      "step": 2354
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.18615196645259857,
+      "learning_rate": 0.0001796503132061672,
+      "loss": 0.9783,
+      "step": 2355
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.17984063923358917,
+      "learning_rate": 0.00017963356843795177,
+      "loss": 0.9427,
+      "step": 2356
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.17674154043197632,
+      "learning_rate": 0.00017961681756438927,
+      "loss": 0.9691,
+      "step": 2357
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.18042780458927155,
+      "learning_rate": 0.00017960006058676388,
+      "loss": 0.9217,
+      "step": 2358
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1813279390335083,
+      "learning_rate": 0.00017958329750636037,
+      "loss": 1.0397,
+      "step": 2359
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.18235574662685394,
+      "learning_rate": 0.00017956652832446386,
+      "loss": 0.9653,
+      "step": 2360
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.19341222941875458,
+      "learning_rate": 0.00017954975304236006,
+      "loss": 0.9598,
+      "step": 2361
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.18789170682430267,
+      "learning_rate": 0.00017953297166133512,
+      "loss": 0.9359,
+      "step": 2362
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.18214201927185059,
+      "learning_rate": 0.00017951618418267557,
+      "loss": 0.9558,
+      "step": 2363
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1955145299434662,
+      "learning_rate": 0.00017949939060766856,
+      "loss": 1.019,
+      "step": 2364
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.18358461558818817,
+      "learning_rate": 0.00017948259093760155,
+      "loss": 0.9635,
+      "step": 2365
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.18531377613544464,
+      "learning_rate": 0.0001794657851737625,
+      "loss": 0.9763,
+      "step": 2366
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.18523001670837402,
+      "learning_rate": 0.00017944897331743999,
+      "loss": 0.9611,
+      "step": 2367
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.19314917922019958,
+      "learning_rate": 0.00017943215536992289,
+      "loss": 0.9916,
+      "step": 2368
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.19348804652690887,
+      "learning_rate": 0.0001794153313325006,
+      "loss": 0.9994,
+      "step": 2369
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.23391282558441162,
+      "learning_rate": 0.000179398501206463,
+      "loss": 0.915,
+      "step": 2370
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.18983489274978638,
+      "learning_rate": 0.00017938166499310035,
+      "loss": 0.9543,
+      "step": 2371
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1920963078737259,
+      "learning_rate": 0.00017936482269370355,
+      "loss": 1.0246,
+      "step": 2372
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.19405518472194672,
+      "learning_rate": 0.00017934797430956383,
+      "loss": 0.9415,
+      "step": 2373
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.24989286065101624,
+      "learning_rate": 0.0001793311198419729,
+      "loss": 0.8994,
+      "step": 2374
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.18366257846355438,
+      "learning_rate": 0.000179314259292223,
+      "loss": 0.9825,
+      "step": 2375
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.19646799564361572,
+      "learning_rate": 0.00017929739266160674,
+      "loss": 0.9353,
+      "step": 2376
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.19476506114006042,
+      "learning_rate": 0.0001792805199514173,
+      "loss": 1.0047,
+      "step": 2377
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1821003556251526,
+      "learning_rate": 0.00017926364116294824,
+      "loss": 1.0428,
+      "step": 2378
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.19362856447696686,
+      "learning_rate": 0.00017924675629749364,
+      "loss": 0.9599,
+      "step": 2379
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1914760172367096,
+      "learning_rate": 0.00017922986535634804,
+      "loss": 0.9749,
+      "step": 2380
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1982416957616806,
+      "learning_rate": 0.0001792129683408064,
+      "loss": 0.9649,
+      "step": 2381
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.18061783909797668,
+      "learning_rate": 0.00017919606525216422,
+      "loss": 0.9222,
+      "step": 2382
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.17703884840011597,
+      "learning_rate": 0.00017917915609171742,
+      "loss": 0.9463,
+      "step": 2383
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.19080610573291779,
+      "learning_rate": 0.00017916224086076237,
+      "loss": 1.0145,
+      "step": 2384
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.17901524901390076,
+      "learning_rate": 0.00017914531956059594,
+      "loss": 0.9843,
+      "step": 2385
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.19128692150115967,
+      "learning_rate": 0.00017912839219251547,
+      "loss": 1.02,
+      "step": 2386
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1813100278377533,
+      "learning_rate": 0.0001791114587578187,
+      "loss": 0.8977,
+      "step": 2387
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1859530359506607,
+      "learning_rate": 0.00017909451925780394,
+      "loss": 0.9613,
+      "step": 2388
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1826142519712448,
+      "learning_rate": 0.00017907757369376985,
+      "loss": 0.9584,
+      "step": 2389
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.19591540098190308,
+      "learning_rate": 0.0001790606220670157,
+      "loss": 1.0151,
+      "step": 2390
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1937379240989685,
+      "learning_rate": 0.00017904366437884102,
+      "loss": 0.9983,
+      "step": 2391
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.18422500789165497,
+      "learning_rate": 0.00017902670063054597,
+      "loss": 0.903,
+      "step": 2392
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.181779146194458,
+      "learning_rate": 0.0001790097308234312,
+      "loss": 0.951,
+      "step": 2393
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.18863825500011444,
+      "learning_rate": 0.00017899275495879765,
+      "loss": 1.015,
+      "step": 2394
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1767343282699585,
+      "learning_rate": 0.0001789757730379469,
+      "loss": 0.8843,
+      "step": 2395
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1767064779996872,
+      "learning_rate": 0.00017895878506218085,
+      "loss": 0.9389,
+      "step": 2396
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1905364841222763,
+      "learning_rate": 0.00017894179103280198,
+      "loss": 0.9241,
+      "step": 2397
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.184580460190773,
+      "learning_rate": 0.00017892479095111318,
+      "loss": 0.9488,
+      "step": 2398
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.18272309005260468,
+      "learning_rate": 0.00017890778481841781,
+      "loss": 0.9805,
+      "step": 2399
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1822749674320221,
+      "learning_rate": 0.0001788907726360197,
+      "loss": 0.9814,
+      "step": 2400
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.18006400763988495,
+      "learning_rate": 0.00017887375440522313,
+      "loss": 0.974,
+      "step": 2401
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.18517065048217773,
+      "learning_rate": 0.00017885673012733288,
+      "loss": 0.9311,
+      "step": 2402
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.18556898832321167,
+      "learning_rate": 0.00017883969980365413,
+      "loss": 0.9755,
+      "step": 2403
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.17660577595233917,
+      "learning_rate": 0.00017882266343549258,
+      "loss": 0.9523,
+      "step": 2404
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.18745553493499756,
+      "learning_rate": 0.0001788056210241544,
+      "loss": 1.0022,
+      "step": 2405
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1848209798336029,
+      "learning_rate": 0.00017878857257094615,
+      "loss": 0.9009,
+      "step": 2406
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1945207267999649,
+      "learning_rate": 0.00017877151807717492,
+      "loss": 0.9552,
+      "step": 2407
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.17822642624378204,
+      "learning_rate": 0.00017875445754414826,
+      "loss": 0.8894,
+      "step": 2408
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.17673255503177643,
+      "learning_rate": 0.00017873739097317416,
+      "loss": 0.9404,
+      "step": 2409
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.18701551854610443,
+      "learning_rate": 0.00017872031836556104,
+      "loss": 0.9207,
+      "step": 2410
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.19993789494037628,
+      "learning_rate": 0.0001787032397226179,
+      "loss": 0.9728,
+      "step": 2411
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.18984495103359222,
+      "learning_rate": 0.00017868615504565405,
+      "loss": 0.9743,
+      "step": 2412
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.19204598665237427,
+      "learning_rate": 0.0001786690643359794,
+      "loss": 0.9662,
+      "step": 2413
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.19300125539302826,
+      "learning_rate": 0.00017865196759490418,
+      "loss": 0.9866,
+      "step": 2414
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.19442309439182281,
+      "learning_rate": 0.00017863486482373926,
+      "loss": 1.0097,
+      "step": 2415
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.198700413107872,
+      "learning_rate": 0.00017861775602379581,
+      "loss": 1.0121,
+      "step": 2416
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1955936849117279,
+      "learning_rate": 0.00017860064119638557,
+      "loss": 0.9874,
+      "step": 2417
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1855369508266449,
+      "learning_rate": 0.00017858352034282068,
+      "loss": 0.9305,
+      "step": 2418
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.20869863033294678,
+      "learning_rate": 0.00017856639346441373,
+      "loss": 0.9681,
+      "step": 2419
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.18309414386749268,
+      "learning_rate": 0.00017854926056247786,
+      "loss": 0.9009,
+      "step": 2420
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.18978330492973328,
+      "learning_rate": 0.00017853212163832655,
+      "loss": 0.9706,
+      "step": 2421
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.18249556422233582,
+      "learning_rate": 0.00017851497669327387,
+      "loss": 0.9336,
+      "step": 2422
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1928057074546814,
+      "learning_rate": 0.00017849782572863427,
+      "loss": 0.9682,
+      "step": 2423
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2058742493391037,
+      "learning_rate": 0.00017848066874572266,
+      "loss": 1.0131,
+      "step": 2424
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.19305527210235596,
+      "learning_rate": 0.00017846350574585445,
+      "loss": 1.0091,
+      "step": 2425
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1902138888835907,
+      "learning_rate": 0.0001784463367303455,
+      "loss": 1.0441,
+      "step": 2426
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.19089341163635254,
+      "learning_rate": 0.0001784291617005121,
+      "loss": 0.9453,
+      "step": 2427
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.19015923142433167,
+      "learning_rate": 0.00017841198065767107,
+      "loss": 0.9733,
+      "step": 2428
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.20033025741577148,
+      "learning_rate": 0.00017839479360313957,
+      "loss": 0.972,
+      "step": 2429
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.20651842653751373,
+      "learning_rate": 0.00017837760053823537,
+      "loss": 0.9752,
+      "step": 2430
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.19228796660900116,
+      "learning_rate": 0.00017836040146427658,
+      "loss": 1.0119,
+      "step": 2431
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.19866019487380981,
+      "learning_rate": 0.00017834319638258185,
+      "loss": 0.9802,
+      "step": 2432
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.19379134476184845,
+      "learning_rate": 0.00017832598529447026,
+      "loss": 0.9727,
+      "step": 2433
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.20367862284183502,
+      "learning_rate": 0.00017830876820126133,
+      "loss": 0.9894,
+      "step": 2434
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.20174342393875122,
+      "learning_rate": 0.00017829154510427505,
+      "loss": 0.9858,
+      "step": 2435
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1984124779701233,
+      "learning_rate": 0.00017827431600483194,
+      "loss": 1.0181,
+      "step": 2436
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.19504335522651672,
+      "learning_rate": 0.00017825708090425283,
+      "loss": 1.0525,
+      "step": 2437
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.19187763333320618,
+      "learning_rate": 0.0001782398398038592,
+      "loss": 0.9748,
+      "step": 2438
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.19658133387565613,
+      "learning_rate": 0.00017822259270497282,
+      "loss": 0.9971,
+      "step": 2439
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2015499472618103,
+      "learning_rate": 0.00017820533960891601,
+      "loss": 0.9745,
+      "step": 2440
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.18588942289352417,
+      "learning_rate": 0.0001781880805170115,
+      "loss": 0.9684,
+      "step": 2441
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.19589237868785858,
+      "learning_rate": 0.0001781708154305826,
+      "loss": 1.0866,
+      "step": 2442
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.19100069999694824,
+      "learning_rate": 0.00017815354435095292,
+      "loss": 0.9363,
+      "step": 2443
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.20036999881267548,
+      "learning_rate": 0.0001781362672794466,
+      "loss": 0.9627,
+      "step": 2444
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2073875069618225,
+      "learning_rate": 0.00017811898421738826,
+      "loss": 0.9667,
+      "step": 2445
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.19809319078922272,
+      "learning_rate": 0.00017810169516610294,
+      "loss": 1.0038,
+      "step": 2446
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.18860866129398346,
+      "learning_rate": 0.00017808440012691618,
+      "loss": 1.0263,
+      "step": 2447
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1854882389307022,
+      "learning_rate": 0.00017806709910115392,
+      "loss": 0.9482,
+      "step": 2448
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.20285579562187195,
+      "learning_rate": 0.00017804979209014264,
+      "loss": 0.9279,
+      "step": 2449
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.21248310804367065,
+      "learning_rate": 0.0001780324790952092,
+      "loss": 0.9916,
+      "step": 2450
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1837083101272583,
+      "learning_rate": 0.00017801516011768095,
+      "loss": 0.9036,
+      "step": 2451
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.18199291825294495,
+      "learning_rate": 0.00017799783515888576,
+      "loss": 0.9792,
+      "step": 2452
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.19304387271404266,
+      "learning_rate": 0.0001779805042201518,
+      "loss": 0.9757,
+      "step": 2453
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.18951784074306488,
+      "learning_rate": 0.0001779631673028079,
+      "loss": 0.9124,
+      "step": 2454
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1998693346977234,
+      "learning_rate": 0.00017794582440818318,
+      "loss": 0.9957,
+      "step": 2455
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.24689649045467377,
+      "learning_rate": 0.0001779284755376073,
+      "loss": 0.8665,
+      "step": 2456
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.18024122714996338,
+      "learning_rate": 0.00017791112069241036,
+      "loss": 0.9108,
+      "step": 2457
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.19234396517276764,
+      "learning_rate": 0.00017789375987392294,
+      "loss": 0.9928,
+      "step": 2458
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.18758738040924072,
+      "learning_rate": 0.00017787639308347608,
+      "loss": 0.9688,
+      "step": 2459
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.18658681213855743,
+      "learning_rate": 0.0001778590203224012,
+      "loss": 1.0176,
+      "step": 2460
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.19142453372478485,
+      "learning_rate": 0.00017784164159203027,
+      "loss": 0.9591,
+      "step": 2461
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1846286803483963,
+      "learning_rate": 0.00017782425689369567,
+      "loss": 0.9502,
+      "step": 2462
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1844143271446228,
+      "learning_rate": 0.00017780686622873026,
+      "loss": 0.9868,
+      "step": 2463
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.18845881521701813,
+      "learning_rate": 0.00017778946959846737,
+      "loss": 0.9394,
+      "step": 2464
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.19077198207378387,
+      "learning_rate": 0.0001777720670042407,
+      "loss": 1.027,
+      "step": 2465
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.18606038391590118,
+      "learning_rate": 0.00017775465844738453,
+      "loss": 0.9552,
+      "step": 2466
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.18549682199954987,
+      "learning_rate": 0.00017773724392923352,
+      "loss": 0.9809,
+      "step": 2467
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.191581591963768,
+      "learning_rate": 0.00017771982345112278,
+      "loss": 0.9697,
+      "step": 2468
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.18413247168064117,
+      "learning_rate": 0.00017770239701438799,
+      "loss": 0.9741,
+      "step": 2469
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.19427776336669922,
+      "learning_rate": 0.00017768496462036509,
+      "loss": 1.0082,
+      "step": 2470
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.18671664595603943,
+      "learning_rate": 0.00017766752627039063,
+      "loss": 1.0208,
+      "step": 2471
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.18228982388973236,
+      "learning_rate": 0.0001776500819658016,
+      "loss": 0.9956,
+      "step": 2472
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.17968390882015228,
+      "learning_rate": 0.00017763263170793541,
+      "loss": 0.9631,
+      "step": 2473
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.18487539887428284,
+      "learning_rate": 0.00017761517549812992,
+      "loss": 0.9693,
+      "step": 2474
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.18275010585784912,
+      "learning_rate": 0.00017759771333772346,
+      "loss": 0.918,
+      "step": 2475
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1807790845632553,
+      "learning_rate": 0.00017758024522805484,
+      "loss": 0.9267,
+      "step": 2476
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.18188652396202087,
+      "learning_rate": 0.00017756277117046332,
+      "loss": 0.9219,
+      "step": 2477
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1865534782409668,
+      "learning_rate": 0.00017754529116628852,
+      "loss": 0.9727,
+      "step": 2478
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.17771273851394653,
+      "learning_rate": 0.0001775278052168707,
+      "loss": 0.9443,
+      "step": 2479
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.18896722793579102,
+      "learning_rate": 0.0001775103133235504,
+      "loss": 0.944,
+      "step": 2480
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.19240954518318176,
+      "learning_rate": 0.00017749281548766873,
+      "loss": 0.9398,
+      "step": 2481
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1850031018257141,
+      "learning_rate": 0.00017747531171056718,
+      "loss": 0.9118,
+      "step": 2482
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.19715161621570587,
+      "learning_rate": 0.00017745780199358776,
+      "loss": 1.0052,
+      "step": 2483
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1838032752275467,
+      "learning_rate": 0.00017744028633807286,
+      "loss": 0.9127,
+      "step": 2484
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.18692578375339508,
+      "learning_rate": 0.00017742276474536545,
+      "loss": 0.9777,
+      "step": 2485
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.18337734043598175,
+      "learning_rate": 0.0001774052372168088,
+      "loss": 0.9411,
+      "step": 2486
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.18327218294143677,
+      "learning_rate": 0.00017738770375374675,
+      "loss": 0.9248,
+      "step": 2487
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1908503621816635,
+      "learning_rate": 0.00017737016435752357,
+      "loss": 0.9677,
+      "step": 2488
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1790422648191452,
+      "learning_rate": 0.0001773526190294839,
+      "loss": 0.9293,
+      "step": 2489
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.17901550233364105,
+      "learning_rate": 0.000177335067770973,
+      "loss": 0.9304,
+      "step": 2490
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.18893276154994965,
+      "learning_rate": 0.0001773175105833364,
+      "loss": 0.9439,
+      "step": 2491
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.18400788307189941,
+      "learning_rate": 0.00017729994746792027,
+      "loss": 0.9323,
+      "step": 2492
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.18849042057991028,
+      "learning_rate": 0.00017728237842607104,
+      "loss": 1.0064,
+      "step": 2493
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.18234604597091675,
+      "learning_rate": 0.00017726480345913573,
+      "loss": 0.946,
+      "step": 2494
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.18465781211853027,
+      "learning_rate": 0.00017724722256846182,
+      "loss": 0.9963,
+      "step": 2495
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.17847271263599396,
+      "learning_rate": 0.00017722963575539716,
+      "loss": 0.9777,
+      "step": 2496
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.19351433217525482,
+      "learning_rate": 0.0001772120430212901,
+      "loss": 1.0891,
+      "step": 2497
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.19211028516292572,
+      "learning_rate": 0.00017719444436748943,
+      "loss": 0.9809,
+      "step": 2498
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.18651790916919708,
+      "learning_rate": 0.00017717683979534443,
+      "loss": 0.95,
+      "step": 2499
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.18656237423419952,
+      "learning_rate": 0.0001771592293062048,
+      "loss": 0.8969,
+      "step": 2500
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1903420239686966,
+      "learning_rate": 0.00017714161290142074,
+      "loss": 0.969,
+      "step": 2501
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.19045324623584747,
+      "learning_rate": 0.00017712399058234275,
+      "loss": 0.9857,
+      "step": 2502
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.20729127526283264,
+      "learning_rate": 0.000177106362350322,
+      "loss": 0.9877,
+      "step": 2503
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.18529750406742096,
+      "learning_rate": 0.00017708872820670996,
+      "loss": 0.9178,
+      "step": 2504
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.18597035109996796,
+      "learning_rate": 0.00017707108815285866,
+      "loss": 0.9055,
+      "step": 2505
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.19581322371959686,
+      "learning_rate": 0.00017705344219012047,
+      "loss": 1.0413,
+      "step": 2506
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.177476167678833,
+      "learning_rate": 0.0001770357903198483,
+      "loss": 0.9686,
+      "step": 2507
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.17651228606700897,
+      "learning_rate": 0.0001770181325433955,
+      "loss": 0.9821,
+      "step": 2508
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1783791333436966,
+      "learning_rate": 0.0001770004688621158,
+      "loss": 0.9437,
+      "step": 2509
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1889670491218567,
+      "learning_rate": 0.00017698279927736349,
+      "loss": 0.996,
+      "step": 2510
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.18272949755191803,
+      "learning_rate": 0.00017696512379049325,
+      "loss": 0.971,
+      "step": 2511
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.18914766609668732,
+      "learning_rate": 0.0001769474424028602,
+      "loss": 0.937,
+      "step": 2512
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1900867372751236,
+      "learning_rate": 0.00017692975511581998,
+      "loss": 1.0283,
+      "step": 2513
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1854158490896225,
+      "learning_rate": 0.0001769120619307286,
+      "loss": 0.9811,
+      "step": 2514
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.18695878982543945,
+      "learning_rate": 0.00017689436284894257,
+      "loss": 0.9677,
+      "step": 2515
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1806347817182541,
+      "learning_rate": 0.0001768766578718189,
+      "loss": 0.9546,
+      "step": 2516
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.18050779402256012,
+      "learning_rate": 0.00017685894700071489,
+      "loss": 0.9554,
+      "step": 2517
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.20120397210121155,
+      "learning_rate": 0.00017684123023698847,
+      "loss": 0.9677,
+      "step": 2518
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.19255037605762482,
+      "learning_rate": 0.00017682350758199794,
+      "loss": 0.9656,
+      "step": 2519
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.19251564145088196,
+      "learning_rate": 0.00017680577903710204,
+      "loss": 0.9783,
+      "step": 2520
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.19172608852386475,
+      "learning_rate": 0.00017678804460366,
+      "loss": 0.9441,
+      "step": 2521
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.17901365458965302,
+      "learning_rate": 0.0001767703042830315,
+      "loss": 0.8979,
+      "step": 2522
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1947840005159378,
+      "learning_rate": 0.00017675255807657661,
+      "loss": 0.982,
+      "step": 2523
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.192917600274086,
+      "learning_rate": 0.00017673480598565596,
+      "loss": 1.0014,
+      "step": 2524
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.18823352456092834,
+      "learning_rate": 0.00017671704801163046,
+      "loss": 0.9995,
+      "step": 2525
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.19002202153205872,
+      "learning_rate": 0.00017669928415586168,
+      "loss": 1.0014,
+      "step": 2526
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.18676120042800903,
+      "learning_rate": 0.0001766815144197115,
+      "loss": 0.9492,
+      "step": 2527
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.18423797190189362,
+      "learning_rate": 0.0001766637388045423,
+      "loss": 0.9578,
+      "step": 2528
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.18916915357112885,
+      "learning_rate": 0.00017664595731171692,
+      "loss": 0.9713,
+      "step": 2529
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.21028481423854828,
+      "learning_rate": 0.00017662816994259858,
+      "loss": 0.9628,
+      "step": 2530
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.17987662553787231,
+      "learning_rate": 0.00017661037669855105,
+      "loss": 0.8998,
+      "step": 2531
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.17913588881492615,
+      "learning_rate": 0.00017659257758093845,
+      "loss": 0.9486,
+      "step": 2532
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.20375816524028778,
+      "learning_rate": 0.0001765747725911255,
+      "loss": 1.0011,
+      "step": 2533
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.18085017800331116,
+      "learning_rate": 0.00017655696173047715,
+      "loss": 0.9419,
+      "step": 2534
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.20048849284648895,
+      "learning_rate": 0.000176539145000359,
+      "loss": 1.0196,
+      "step": 2535
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1985366940498352,
+      "learning_rate": 0.000176521322402137,
+      "loss": 1.0123,
+      "step": 2536
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.18767307698726654,
+      "learning_rate": 0.0001765034939371776,
+      "loss": 0.957,
+      "step": 2537
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.19573388993740082,
+      "learning_rate": 0.00017648565960684762,
+      "loss": 0.9504,
+      "step": 2538
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.19709470868110657,
+      "learning_rate": 0.00017646781941251443,
+      "loss": 1.0158,
+      "step": 2539
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.19420363008975983,
+      "learning_rate": 0.0001764499733555458,
+      "loss": 0.9909,
+      "step": 2540
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.19303253293037415,
+      "learning_rate": 0.0001764321214373099,
+      "loss": 1.0083,
+      "step": 2541
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.18523170053958893,
+      "learning_rate": 0.0001764142636591755,
+      "loss": 0.9079,
+      "step": 2542
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.18193301558494568,
+      "learning_rate": 0.0001763964000225116,
+      "loss": 0.9455,
+      "step": 2543
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.20065605640411377,
+      "learning_rate": 0.0001763785305286879,
+      "loss": 1.003,
+      "step": 2544
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.19543297588825226,
+      "learning_rate": 0.00017636065517907427,
+      "loss": 0.9628,
+      "step": 2545
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.19203916192054749,
+      "learning_rate": 0.0001763427739750413,
+      "loss": 0.9727,
+      "step": 2546
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1904589831829071,
+      "learning_rate": 0.00017632488691795987,
+      "loss": 0.9252,
+      "step": 2547
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.19423457980155945,
+      "learning_rate": 0.0001763069940092013,
+      "loss": 0.9845,
+      "step": 2548
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.18518464267253876,
+      "learning_rate": 0.00017628909525013744,
+      "loss": 0.9494,
+      "step": 2549
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.18881121277809143,
+      "learning_rate": 0.00017627119064214056,
+      "loss": 0.9423,
+      "step": 2550
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1885634958744049,
+      "learning_rate": 0.0001762532801865834,
+      "loss": 0.9789,
+      "step": 2551
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.18694055080413818,
+      "learning_rate": 0.00017623536388483905,
+      "loss": 0.9094,
+      "step": 2552
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.18675947189331055,
+      "learning_rate": 0.00017621744173828115,
+      "loss": 0.9737,
+      "step": 2553
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.18994884192943573,
+      "learning_rate": 0.00017619951374828373,
+      "loss": 0.9488,
+      "step": 2554
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1960517317056656,
+      "learning_rate": 0.00017618157991622138,
+      "loss": 0.9846,
+      "step": 2555
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1805785894393921,
+      "learning_rate": 0.00017616364024346893,
+      "loss": 0.9257,
+      "step": 2556
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.19077786803245544,
+      "learning_rate": 0.00017614569473140188,
+      "loss": 0.9756,
+      "step": 2557
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.19670559465885162,
+      "learning_rate": 0.000176127743381396,
+      "loss": 0.9607,
+      "step": 2558
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.19572590291500092,
+      "learning_rate": 0.00017610978619482764,
+      "loss": 0.9815,
+      "step": 2559
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.19013354182243347,
+      "learning_rate": 0.0001760918231730735,
+      "loss": 0.9206,
+      "step": 2560
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.17755639553070068,
+      "learning_rate": 0.0001760738543175108,
+      "loss": 0.9608,
+      "step": 2561
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.18971143662929535,
+      "learning_rate": 0.00017605587962951717,
+      "loss": 0.9529,
+      "step": 2562
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.20042835175991058,
+      "learning_rate": 0.0001760378991104707,
+      "loss": 1.0179,
+      "step": 2563
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.18564312160015106,
+      "learning_rate": 0.0001760199127617499,
+      "loss": 0.9651,
+      "step": 2564
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.19597946107387543,
+      "learning_rate": 0.00017600192058473378,
+      "loss": 0.9924,
+      "step": 2565
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.19503167271614075,
+      "learning_rate": 0.00017598392258080168,
+      "loss": 0.9508,
+      "step": 2566
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.18775935471057892,
+      "learning_rate": 0.0001759659187513336,
+      "loss": 0.9388,
+      "step": 2567
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.18213537335395813,
+      "learning_rate": 0.00017594790909770976,
+      "loss": 0.9227,
+      "step": 2568
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.20675437152385712,
+      "learning_rate": 0.00017592989362131097,
+      "loss": 1.002,
+      "step": 2569
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.17877671122550964,
+      "learning_rate": 0.0001759118723235184,
+      "loss": 0.9257,
+      "step": 2570
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.18973997235298157,
+      "learning_rate": 0.00017589384520571376,
+      "loss": 0.9784,
+      "step": 2571
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.19125111401081085,
+      "learning_rate": 0.0001758758122692791,
+      "loss": 0.9594,
+      "step": 2572
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.18701352179050446,
+      "learning_rate": 0.00017585777351559703,
+      "loss": 0.963,
+      "step": 2573
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.18558120727539062,
+      "learning_rate": 0.00017583972894605047,
+      "loss": 0.9009,
+      "step": 2574
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.18775434792041779,
+      "learning_rate": 0.00017582167856202295,
+      "loss": 1.0063,
+      "step": 2575
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1810017228126526,
+      "learning_rate": 0.00017580362236489823,
+      "loss": 0.8829,
+      "step": 2576
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.17568695545196533,
+      "learning_rate": 0.00017578556035606079,
+      "loss": 0.9182,
+      "step": 2577
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.18526588380336761,
+      "learning_rate": 0.00017576749253689532,
+      "loss": 0.9913,
+      "step": 2578
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.18252696096897125,
+      "learning_rate": 0.00017574941890878705,
+      "loss": 0.9919,
+      "step": 2579
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1952015608549118,
+      "learning_rate": 0.0001757313394731217,
+      "loss": 1.019,
+      "step": 2580
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.18415303528308868,
+      "learning_rate": 0.00017571325423128533,
+      "loss": 0.9776,
+      "step": 2581
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.18594758212566376,
+      "learning_rate": 0.00017569516318466453,
+      "loss": 0.9915,
+      "step": 2582
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.19213664531707764,
+      "learning_rate": 0.00017567706633464628,
+      "loss": 1.0335,
+      "step": 2583
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.18794137239456177,
+      "learning_rate": 0.00017565896368261803,
+      "loss": 0.9585,
+      "step": 2584
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.19076941907405853,
+      "learning_rate": 0.00017564085522996767,
+      "loss": 0.9427,
+      "step": 2585
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1841420829296112,
+      "learning_rate": 0.00017562274097808359,
+      "loss": 0.9604,
+      "step": 2586
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.17589803040027618,
+      "learning_rate": 0.00017560462092835452,
+      "loss": 0.9205,
+      "step": 2587
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.19224637746810913,
+      "learning_rate": 0.0001755864950821697,
+      "loss": 0.9698,
+      "step": 2588
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.18274915218353271,
+      "learning_rate": 0.00017556836344091882,
+      "loss": 0.9323,
+      "step": 2589
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.18005827069282532,
+      "learning_rate": 0.00017555022600599198,
+      "loss": 0.9531,
+      "step": 2590
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.18462547659873962,
+      "learning_rate": 0.00017553208277877973,
+      "loss": 1.0143,
+      "step": 2591
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.18477481603622437,
+      "learning_rate": 0.00017551393376067314,
+      "loss": 0.9429,
+      "step": 2592
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.18777184188365936,
+      "learning_rate": 0.00017549577895306358,
+      "loss": 0.9305,
+      "step": 2593
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.18475376069545746,
+      "learning_rate": 0.00017547761835734295,
+      "loss": 0.9892,
+      "step": 2594
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.176994189620018,
+      "learning_rate": 0.00017545945197490365,
+      "loss": 0.9627,
+      "step": 2595
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.180376335978508,
+      "learning_rate": 0.00017544127980713837,
+      "loss": 0.9661,
+      "step": 2596
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.18328218162059784,
+      "learning_rate": 0.00017542310185544041,
+      "loss": 0.9033,
+      "step": 2597
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.19031573832035065,
+      "learning_rate": 0.00017540491812120343,
+      "loss": 0.9594,
+      "step": 2598
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1834496259689331,
+      "learning_rate": 0.00017538672860582151,
+      "loss": 0.9981,
+      "step": 2599
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.19750292599201202,
+      "learning_rate": 0.00017536853331068922,
+      "loss": 1.1101,
+      "step": 2600
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.19078829884529114,
+      "learning_rate": 0.00017535033223720154,
+      "loss": 0.9598,
+      "step": 2601
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.18731549382209778,
+      "learning_rate": 0.00017533212538675395,
+      "loss": 0.9635,
+      "step": 2602
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.18684478104114532,
+      "learning_rate": 0.00017531391276074228,
+      "loss": 0.9185,
+      "step": 2603
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.18306666612625122,
+      "learning_rate": 0.00017529569436056287,
+      "loss": 0.9284,
+      "step": 2604
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.17791491746902466,
+      "learning_rate": 0.00017527747018761253,
+      "loss": 0.9262,
+      "step": 2605
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.18981628119945526,
+      "learning_rate": 0.00017525924024328844,
+      "loss": 1.0016,
+      "step": 2606
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1939721405506134,
+      "learning_rate": 0.00017524100452898824,
+      "loss": 0.9912,
+      "step": 2607
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.19684866070747375,
+      "learning_rate": 0.00017522276304611005,
+      "loss": 1.0439,
+      "step": 2608
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1944311410188675,
+      "learning_rate": 0.0001752045157960524,
+      "loss": 0.921,
+      "step": 2609
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.18503262102603912,
+      "learning_rate": 0.00017518626278021425,
+      "loss": 0.9166,
+      "step": 2610
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.17905835807323456,
+      "learning_rate": 0.00017516800399999507,
+      "loss": 0.9416,
+      "step": 2611
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.17581358551979065,
+      "learning_rate": 0.0001751497394567947,
+      "loss": 0.8916,
+      "step": 2612
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.18805000185966492,
+      "learning_rate": 0.0001751314691520134,
+      "loss": 0.9259,
+      "step": 2613
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.18567489087581635,
+      "learning_rate": 0.00017511319308705198,
+      "loss": 0.9211,
+      "step": 2614
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.19012784957885742,
+      "learning_rate": 0.00017509491126331162,
+      "loss": 0.9819,
+      "step": 2615
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.18892110884189606,
+      "learning_rate": 0.00017507662368219395,
+      "loss": 0.896,
+      "step": 2616
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.18657296895980835,
+      "learning_rate": 0.00017505833034510104,
+      "loss": 0.9289,
+      "step": 2617
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.19216986000537872,
+      "learning_rate": 0.00017504003125343536,
+      "loss": 0.9743,
+      "step": 2618
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.18053677678108215,
+      "learning_rate": 0.00017502172640859991,
+      "loss": 0.9065,
+      "step": 2619
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.19183310866355896,
+      "learning_rate": 0.0001750034158119981,
+      "loss": 1.0026,
+      "step": 2620
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.17989866435527802,
+      "learning_rate": 0.0001749850994650337,
+      "loss": 0.8568,
+      "step": 2621
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.19608454406261444,
+      "learning_rate": 0.0001749667773691111,
+      "loss": 0.9396,
+      "step": 2622
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1893497258424759,
+      "learning_rate": 0.00017494844952563492,
+      "loss": 1.0244,
+      "step": 2623
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1881159394979477,
+      "learning_rate": 0.00017493011593601034,
+      "loss": 0.9258,
+      "step": 2624
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.19302113354206085,
+      "learning_rate": 0.00017491177660164294,
+      "loss": 0.9426,
+      "step": 2625
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1864844262599945,
+      "learning_rate": 0.00017489343152393886,
+      "loss": 0.9482,
+      "step": 2626
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.19288282096385956,
+      "learning_rate": 0.00017487508070430447,
+      "loss": 1.0078,
+      "step": 2627
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.19508597254753113,
+      "learning_rate": 0.00017485672414414675,
+      "loss": 1.0081,
+      "step": 2628
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.19731415808200836,
+      "learning_rate": 0.00017483836184487304,
+      "loss": 1.0362,
+      "step": 2629
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.18698172271251678,
+      "learning_rate": 0.00017481999380789112,
+      "loss": 0.9922,
+      "step": 2630
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1864076554775238,
+      "learning_rate": 0.0001748016200346093,
+      "loss": 0.9954,
+      "step": 2631
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.19683367013931274,
+      "learning_rate": 0.00017478324052643617,
+      "loss": 0.9626,
+      "step": 2632
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1958743780851364,
+      "learning_rate": 0.00017476485528478093,
+      "loss": 0.9605,
+      "step": 2633
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.19047711789608002,
+      "learning_rate": 0.00017474646431105305,
+      "loss": 0.9122,
+      "step": 2634
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.18464651703834534,
+      "learning_rate": 0.00017472806760666262,
+      "loss": 0.9645,
+      "step": 2635
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.19960200786590576,
+      "learning_rate": 0.00017470966517302004,
+      "loss": 0.9316,
+      "step": 2636
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.18094268441200256,
+      "learning_rate": 0.00017469125701153622,
+      "loss": 0.9249,
+      "step": 2637
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.18596448004245758,
+      "learning_rate": 0.00017467284312362242,
+      "loss": 0.9562,
+      "step": 2638
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.19108310341835022,
+      "learning_rate": 0.00017465442351069044,
+      "loss": 0.9802,
+      "step": 2639
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.19825781881809235,
+      "learning_rate": 0.00017463599817415243,
+      "loss": 1.0161,
+      "step": 2640
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.18590597808361053,
+      "learning_rate": 0.0001746175671154211,
+      "loss": 0.8963,
+      "step": 2641
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.19133514165878296,
+      "learning_rate": 0.00017459913033590944,
+      "loss": 0.9536,
+      "step": 2642
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.18116304278373718,
+      "learning_rate": 0.000174580687837031,
+      "loss": 1.0119,
+      "step": 2643
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.19031855463981628,
+      "learning_rate": 0.00017456223962019975,
+      "loss": 0.9332,
+      "step": 2644
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.18714213371276855,
+      "learning_rate": 0.00017454378568683003,
+      "loss": 0.9149,
+      "step": 2645
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.20420034229755402,
+      "learning_rate": 0.00017452532603833668,
+      "loss": 0.9958,
+      "step": 2646
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1930292397737503,
+      "learning_rate": 0.00017450686067613502,
+      "loss": 0.9853,
+      "step": 2647
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.20497813820838928,
+      "learning_rate": 0.00017448838960164068,
+      "loss": 0.9942,
+      "step": 2648
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.19013455510139465,
+      "learning_rate": 0.00017446991281626982,
+      "loss": 0.999,
+      "step": 2649
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.19203388690948486,
+      "learning_rate": 0.00017445143032143905,
+      "loss": 0.9922,
+      "step": 2650
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1985304057598114,
+      "learning_rate": 0.00017443294211856534,
+      "loss": 0.9287,
+      "step": 2651
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.19734938442707062,
+      "learning_rate": 0.0001744144482090662,
+      "loss": 1.036,
+      "step": 2652
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.19779257476329803,
+      "learning_rate": 0.00017439594859435946,
+      "loss": 0.9858,
+      "step": 2653
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1939266473054886,
+      "learning_rate": 0.0001743774432758635,
+      "loss": 0.9208,
+      "step": 2654
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.19337120652198792,
+      "learning_rate": 0.00017435893225499705,
+      "loss": 0.9435,
+      "step": 2655
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.17661985754966736,
+      "learning_rate": 0.00017434041553317933,
+      "loss": 0.9302,
+      "step": 2656
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.19542555510997772,
+      "learning_rate": 0.00017432189311182995,
+      "loss": 0.956,
+      "step": 2657
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.19809268414974213,
+      "learning_rate": 0.00017430336499236904,
+      "loss": 1.0116,
+      "step": 2658
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.18813373148441315,
+      "learning_rate": 0.00017428483117621707,
+      "loss": 0.9388,
+      "step": 2659
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.18381525576114655,
+      "learning_rate": 0.00017426629166479503,
+      "loss": 1.002,
+      "step": 2660
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.18643952906131744,
+      "learning_rate": 0.00017424774645952425,
+      "loss": 0.9856,
+      "step": 2661
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.18275515735149384,
+      "learning_rate": 0.0001742291955618266,
+      "loss": 0.9922,
+      "step": 2662
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.18715785443782806,
+      "learning_rate": 0.00017421063897312439,
+      "loss": 0.9633,
+      "step": 2663
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.19189901649951935,
+      "learning_rate": 0.00017419207669484022,
+      "loss": 0.9482,
+      "step": 2664
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2036040723323822,
+      "learning_rate": 0.0001741735087283972,
+      "loss": 1.0095,
+      "step": 2665
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.19645290076732635,
+      "learning_rate": 0.00017415493507521904,
+      "loss": 0.9933,
+      "step": 2666
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1870705783367157,
+      "learning_rate": 0.00017413635573672963,
+      "loss": 0.9274,
+      "step": 2667
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.18717564642429352,
+      "learning_rate": 0.00017411777071435345,
+      "loss": 1.0066,
+      "step": 2668
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.18918143212795258,
+      "learning_rate": 0.00017409918000951534,
+      "loss": 0.9256,
+      "step": 2669
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.18597491085529327,
+      "learning_rate": 0.00017408058362364067,
+      "loss": 0.9461,
+      "step": 2670
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.18700912594795227,
+      "learning_rate": 0.0001740619815581552,
+      "loss": 0.9501,
+      "step": 2671
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.18128564953804016,
+      "learning_rate": 0.000174043373814485,
+      "loss": 0.9801,
+      "step": 2672
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.19424092769622803,
+      "learning_rate": 0.0001740247603940568,
+      "loss": 0.977,
+      "step": 2673
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.19749069213867188,
+      "learning_rate": 0.00017400614129829762,
+      "loss": 1.0002,
+      "step": 2674
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.19106240570545197,
+      "learning_rate": 0.00017398751652863495,
+      "loss": 0.9679,
+      "step": 2675
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.18876773118972778,
+      "learning_rate": 0.0001739688860864967,
+      "loss": 0.9475,
+      "step": 2676
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1802065521478653,
+      "learning_rate": 0.00017395024997331123,
+      "loss": 0.9605,
+      "step": 2677
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.18019360303878784,
+      "learning_rate": 0.00017393160819050737,
+      "loss": 0.9025,
+      "step": 2678
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.20628519356250763,
+      "learning_rate": 0.00017391296073951428,
+      "loss": 0.9899,
+      "step": 2679
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1838846504688263,
+      "learning_rate": 0.00017389430762176168,
+      "loss": 0.9286,
+      "step": 2680
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.196673184633255,
+      "learning_rate": 0.00017387564883867968,
+      "loss": 0.9841,
+      "step": 2681
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.19348804652690887,
+      "learning_rate": 0.00017385698439169875,
+      "loss": 0.981,
+      "step": 2682
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.19190922379493713,
+      "learning_rate": 0.0001738383142822499,
+      "loss": 0.9236,
+      "step": 2683
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1826351135969162,
+      "learning_rate": 0.00017381963851176454,
+      "loss": 0.9592,
+      "step": 2684
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.18728382885456085,
+      "learning_rate": 0.00017380095708167446,
+      "loss": 0.8939,
+      "step": 2685
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.19076916575431824,
+      "learning_rate": 0.00017378226999341197,
+      "loss": 0.9717,
+      "step": 2686
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.17857998609542847,
+      "learning_rate": 0.00017376357724840973,
+      "loss": 0.9,
+      "step": 2687
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1867329478263855,
+      "learning_rate": 0.00017374487884810088,
+      "loss": 0.957,
+      "step": 2688
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.18857437372207642,
+      "learning_rate": 0.00017372617479391903,
+      "loss": 0.9405,
+      "step": 2689
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.19370107352733612,
+      "learning_rate": 0.00017370746508729818,
+      "loss": 0.8735,
+      "step": 2690
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.18515117466449738,
+      "learning_rate": 0.00017368874972967273,
+      "loss": 0.8689,
+      "step": 2691
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.20236286520957947,
+      "learning_rate": 0.00017367002872247758,
+      "loss": 1.0547,
+      "step": 2692
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.19606740772724152,
+      "learning_rate": 0.000173651302067148,
+      "loss": 1.0677,
+      "step": 2693
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1996169537305832,
+      "learning_rate": 0.00017363256976511972,
+      "loss": 0.8994,
+      "step": 2694
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1887851357460022,
+      "learning_rate": 0.00017361383181782898,
+      "loss": 0.9454,
+      "step": 2695
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1962120532989502,
+      "learning_rate": 0.0001735950882267123,
+      "loss": 0.9883,
+      "step": 2696
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.184779554605484,
+      "learning_rate": 0.0001735763389932068,
+      "loss": 0.9725,
+      "step": 2697
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.19174376130104065,
+      "learning_rate": 0.00017355758411874983,
+      "loss": 0.9666,
+      "step": 2698
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.19610293209552765,
+      "learning_rate": 0.00017353882360477937,
+      "loss": 0.9812,
+      "step": 2699
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.19059418141841888,
+      "learning_rate": 0.0001735200574527337,
+      "loss": 0.9819,
+      "step": 2700
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.18484486639499664,
+      "learning_rate": 0.00017350128566405172,
+      "loss": 0.9654,
+      "step": 2701
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.19833725690841675,
+      "learning_rate": 0.00017348250824017242,
+      "loss": 0.9318,
+      "step": 2702
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1945193111896515,
+      "learning_rate": 0.00017346372518253558,
+      "loss": 0.978,
+      "step": 2703
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.18793296813964844,
+      "learning_rate": 0.00017344493649258118,
+      "loss": 0.9256,
+      "step": 2704
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1921757608652115,
+      "learning_rate": 0.00017342614217174976,
+      "loss": 0.981,
+      "step": 2705
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.19174674153327942,
+      "learning_rate": 0.00017340734222148223,
+      "loss": 0.981,
+      "step": 2706
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.19075125455856323,
+      "learning_rate": 0.00017338853664321992,
+      "loss": 0.985,
+      "step": 2707
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.18513436615467072,
+      "learning_rate": 0.00017336972543840467,
+      "loss": 1.0468,
+      "step": 2708
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1740717887878418,
+      "learning_rate": 0.00017335090860847862,
+      "loss": 0.911,
+      "step": 2709
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.20851874351501465,
+      "learning_rate": 0.00017333208615488447,
+      "loss": 1.1055,
+      "step": 2710
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1837608367204666,
+      "learning_rate": 0.0001733132580790653,
+      "loss": 0.9793,
+      "step": 2711
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.17361067235469818,
+      "learning_rate": 0.0001732944243824646,
+      "loss": 0.8594,
+      "step": 2712
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1763828694820404,
+      "learning_rate": 0.00017327558506652636,
+      "loss": 0.9235,
+      "step": 2713
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1873314380645752,
+      "learning_rate": 0.00017325674013269487,
+      "loss": 0.9479,
+      "step": 2714
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.19114772975444794,
+      "learning_rate": 0.00017323788958241502,
+      "loss": 0.9618,
+      "step": 2715
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.19344426691532135,
+      "learning_rate": 0.00017321903341713203,
+      "loss": 0.9868,
+      "step": 2716
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.19095127284526825,
+      "learning_rate": 0.00017320017163829148,
+      "loss": 0.9675,
+      "step": 2717
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.19331395626068115,
+      "learning_rate": 0.00017318130424733954,
+      "loss": 0.9681,
+      "step": 2718
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.19180914759635925,
+      "learning_rate": 0.00017316243124572273,
+      "loss": 0.9113,
+      "step": 2719
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1976597011089325,
+      "learning_rate": 0.000173143552634888,
+      "loss": 1.0001,
+      "step": 2720
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.19116324186325073,
+      "learning_rate": 0.00017312466841628272,
+      "loss": 0.9865,
+      "step": 2721
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1902334988117218,
+      "learning_rate": 0.0001731057785913547,
+      "loss": 0.9652,
+      "step": 2722
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1837366372346878,
+      "learning_rate": 0.00017308688316155225,
+      "loss": 0.8781,
+      "step": 2723
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.18876072764396667,
+      "learning_rate": 0.00017306798212832395,
+      "loss": 1.0168,
+      "step": 2724
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.18862731754779816,
+      "learning_rate": 0.000173049075493119,
+      "loss": 0.9721,
+      "step": 2725
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1875831037759781,
+      "learning_rate": 0.00017303016325738685,
+      "loss": 0.9598,
+      "step": 2726
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.18485279381275177,
+      "learning_rate": 0.00017301124542257748,
+      "loss": 0.9447,
+      "step": 2727
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1946990042924881,
+      "learning_rate": 0.00017299232199014135,
+      "loss": 0.9895,
+      "step": 2728
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.18176619708538055,
+      "learning_rate": 0.0001729733929615292,
+      "loss": 0.9482,
+      "step": 2729
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.19105994701385498,
+      "learning_rate": 0.00017295445833819236,
+      "loss": 0.9972,
+      "step": 2730
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.18423745036125183,
+      "learning_rate": 0.0001729355181215824,
+      "loss": 0.9693,
+      "step": 2731
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.18534033000469208,
+      "learning_rate": 0.00017291657231315152,
+      "loss": 0.8961,
+      "step": 2732
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1894759088754654,
+      "learning_rate": 0.00017289762091435225,
+      "loss": 0.96,
+      "step": 2733
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.18298478424549103,
+      "learning_rate": 0.00017287866392663752,
+      "loss": 0.8785,
+      "step": 2734
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.18672245740890503,
+      "learning_rate": 0.00017285970135146075,
+      "loss": 0.9642,
+      "step": 2735
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.18695160746574402,
+      "learning_rate": 0.00017284073319027575,
+      "loss": 0.9333,
+      "step": 2736
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1886570155620575,
+      "learning_rate": 0.00017282175944453675,
+      "loss": 0.9322,
+      "step": 2737
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1965176910161972,
+      "learning_rate": 0.00017280278011569847,
+      "loss": 0.9819,
+      "step": 2738
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.19003883004188538,
+      "learning_rate": 0.00017278379520521605,
+      "loss": 1.0588,
+      "step": 2739
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.19280295073986053,
+      "learning_rate": 0.00017276480471454492,
+      "loss": 0.9592,
+      "step": 2740
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1867469996213913,
+      "learning_rate": 0.00017274580864514108,
+      "loss": 0.9086,
+      "step": 2741
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1786963790655136,
+      "learning_rate": 0.000172726806998461,
+      "loss": 0.8676,
+      "step": 2742
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.18356633186340332,
+      "learning_rate": 0.0001727077997759614,
+      "loss": 0.9183,
+      "step": 2743
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.19396550953388214,
+      "learning_rate": 0.00017268878697909961,
+      "loss": 0.9955,
+      "step": 2744
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.17852063477039337,
+      "learning_rate": 0.0001726697686093332,
+      "loss": 0.9336,
+      "step": 2745
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.18407288193702698,
+      "learning_rate": 0.0001726507446681204,
+      "loss": 0.9827,
+      "step": 2746
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1954531967639923,
+      "learning_rate": 0.0001726317151569196,
+      "loss": 0.9903,
+      "step": 2747
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.18365801870822906,
+      "learning_rate": 0.00017261268007718984,
+      "loss": 0.9263,
+      "step": 2748
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.19449764490127563,
+      "learning_rate": 0.00017259363943039052,
+      "loss": 0.9948,
+      "step": 2749
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.18458783626556396,
+      "learning_rate": 0.00017257459321798137,
+      "loss": 0.9354,
+      "step": 2750
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.18690603971481323,
+      "learning_rate": 0.00017255554144142267,
+      "loss": 0.894,
+      "step": 2751
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.18686780333518982,
+      "learning_rate": 0.00017253648410217508,
+      "loss": 1.0131,
+      "step": 2752
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.182310089468956,
+      "learning_rate": 0.0001725174212016997,
+      "loss": 1.0015,
+      "step": 2753
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.18795253336429596,
+      "learning_rate": 0.00017249835274145805,
+      "loss": 0.9597,
+      "step": 2754
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.19204358756542206,
+      "learning_rate": 0.000172479278722912,
+      "loss": 0.9657,
+      "step": 2755
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.19570910930633545,
+      "learning_rate": 0.000172460199147524,
+      "loss": 0.9586,
+      "step": 2756
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.19368427991867065,
+      "learning_rate": 0.00017244111401675682,
+      "loss": 0.9487,
+      "step": 2757
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.19332663714885712,
+      "learning_rate": 0.00017242202333207364,
+      "loss": 0.9792,
+      "step": 2758
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.18540112674236298,
+      "learning_rate": 0.0001724029270949382,
+      "loss": 0.9517,
+      "step": 2759
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1930074840784073,
+      "learning_rate": 0.00017238382530681445,
+      "loss": 1.0426,
+      "step": 2760
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.18122000992298126,
+      "learning_rate": 0.00017236471796916694,
+      "loss": 0.9315,
+      "step": 2761
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.18953172862529755,
+      "learning_rate": 0.0001723456050834606,
+      "loss": 0.9644,
+      "step": 2762
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.18717069923877716,
+      "learning_rate": 0.0001723264866511608,
+      "loss": 1.0098,
+      "step": 2763
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.17663700878620148,
+      "learning_rate": 0.00017230736267373326,
+      "loss": 0.9677,
+      "step": 2764
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.18183469772338867,
+      "learning_rate": 0.0001722882331526442,
+      "loss": 0.9359,
+      "step": 2765
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.18459314107894897,
+      "learning_rate": 0.00017226909808936027,
+      "loss": 0.9705,
+      "step": 2766
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1858403980731964,
+      "learning_rate": 0.00017224995748534844,
+      "loss": 0.9647,
+      "step": 2767
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1881045550107956,
+      "learning_rate": 0.00017223081134207628,
+      "loss": 1.0115,
+      "step": 2768
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.18951913714408875,
+      "learning_rate": 0.00017221165966101163,
+      "loss": 0.9398,
+      "step": 2769
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.19097717106342316,
+      "learning_rate": 0.0001721925024436228,
+      "loss": 0.9834,
+      "step": 2770
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.18614451587200165,
+      "learning_rate": 0.0001721733396913786,
+      "loss": 0.9537,
+      "step": 2771
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1865241825580597,
+      "learning_rate": 0.0001721541714057481,
+      "loss": 0.9338,
+      "step": 2772
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1903114914894104,
+      "learning_rate": 0.00017213499758820096,
+      "loss": 0.9616,
+      "step": 2773
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1903352588415146,
+      "learning_rate": 0.00017211581824020723,
+      "loss": 0.9934,
+      "step": 2774
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.18006877601146698,
+      "learning_rate": 0.00017209663336323726,
+      "loss": 0.9624,
+      "step": 2775
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.19256149232387543,
+      "learning_rate": 0.00017207744295876198,
+      "loss": 0.9431,
+      "step": 2776
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.19010770320892334,
+      "learning_rate": 0.0001720582470282527,
+      "loss": 0.9376,
+      "step": 2777
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1907973736524582,
+      "learning_rate": 0.00017203904557318108,
+      "loss": 0.9791,
+      "step": 2778
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.19442807137966156,
+      "learning_rate": 0.00017201983859501928,
+      "loss": 1.0337,
+      "step": 2779
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.18805912137031555,
+      "learning_rate": 0.00017200062609523984,
+      "loss": 1.0056,
+      "step": 2780
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.19650529325008392,
+      "learning_rate": 0.00017198140807531576,
+      "loss": 1.0526,
+      "step": 2781
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.19417373836040497,
+      "learning_rate": 0.0001719621845367205,
+      "loss": 0.9836,
+      "step": 2782
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.17788995802402496,
+      "learning_rate": 0.0001719429554809278,
+      "loss": 0.9189,
+      "step": 2783
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.19120745360851288,
+      "learning_rate": 0.00017192372090941196,
+      "loss": 0.965,
+      "step": 2784
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.18333250284194946,
+      "learning_rate": 0.00017190448082364765,
+      "loss": 0.9157,
+      "step": 2785
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.19432075321674347,
+      "learning_rate": 0.00017188523522510996,
+      "loss": 0.9863,
+      "step": 2786
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1891552060842514,
+      "learning_rate": 0.00017186598411527446,
+      "loss": 0.9274,
+      "step": 2787
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.19520063698291779,
+      "learning_rate": 0.00017184672749561705,
+      "loss": 0.9257,
+      "step": 2788
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.18742509186267853,
+      "learning_rate": 0.00017182746536761404,
+      "loss": 0.9583,
+      "step": 2789
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.19111135601997375,
+      "learning_rate": 0.00017180819773274234,
+      "loss": 0.9878,
+      "step": 2790
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.18058103322982788,
+      "learning_rate": 0.00017178892459247908,
+      "loss": 0.962,
+      "step": 2791
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.18281134963035583,
+      "learning_rate": 0.00017176964594830193,
+      "loss": 0.9301,
+      "step": 2792
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.17827610671520233,
+      "learning_rate": 0.00017175036180168892,
+      "loss": 0.8986,
+      "step": 2793
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.19877582788467407,
+      "learning_rate": 0.00017173107215411854,
+      "loss": 0.9709,
+      "step": 2794
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.19208525121212006,
+      "learning_rate": 0.0001717117770070697,
+      "loss": 0.9325,
+      "step": 2795
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.19865301251411438,
+      "learning_rate": 0.0001716924763620217,
+      "loss": 0.9956,
+      "step": 2796
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.19380724430084229,
+      "learning_rate": 0.00017167317022045428,
+      "loss": 0.9418,
+      "step": 2797
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.19513899087905884,
+      "learning_rate": 0.00017165385858384765,
+      "loss": 1.0074,
+      "step": 2798
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1949981302022934,
+      "learning_rate": 0.0001716345414536823,
+      "loss": 0.999,
+      "step": 2799
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.18708331882953644,
+      "learning_rate": 0.00017161521883143934,
+      "loss": 1.0341,
+      "step": 2800
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.19197827577590942,
+      "learning_rate": 0.00017159589071860014,
+      "loss": 0.9257,
+      "step": 2801
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1827811300754547,
+      "learning_rate": 0.00017157655711664654,
+      "loss": 0.9209,
+      "step": 2802
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1887734979391098,
+      "learning_rate": 0.00017155721802706087,
+      "loss": 0.9731,
+      "step": 2803
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.19555433094501495,
+      "learning_rate": 0.00017153787345132573,
+      "loss": 0.9979,
+      "step": 2804
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.20035997033119202,
+      "learning_rate": 0.0001715185233909243,
+      "loss": 1.017,
+      "step": 2805
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.19120369851589203,
+      "learning_rate": 0.0001714991678473401,
+      "loss": 0.969,
+      "step": 2806
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1866525262594223,
+      "learning_rate": 0.00017147980682205702,
+      "loss": 0.9909,
+      "step": 2807
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1893429458141327,
+      "learning_rate": 0.00017146044031655951,
+      "loss": 0.9889,
+      "step": 2808
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.19175229966640472,
+      "learning_rate": 0.00017144106833233232,
+      "loss": 0.9641,
+      "step": 2809
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.18215136229991913,
+      "learning_rate": 0.0001714216908708607,
+      "loss": 0.9432,
+      "step": 2810
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.18707439303398132,
+      "learning_rate": 0.00017140230793363025,
+      "loss": 0.9354,
+      "step": 2811
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1920577585697174,
+      "learning_rate": 0.000171382919522127,
+      "loss": 0.9781,
+      "step": 2812
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.17737874388694763,
+      "learning_rate": 0.0001713635256378375,
+      "loss": 1.0113,
+      "step": 2813
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1806550770998001,
+      "learning_rate": 0.00017134412628224852,
+      "loss": 0.9249,
+      "step": 2814
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.18417756259441376,
+      "learning_rate": 0.00017132472145684746,
+      "loss": 0.9205,
+      "step": 2815
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.18996597826480865,
+      "learning_rate": 0.00017130531116312203,
+      "loss": 0.9541,
+      "step": 2816
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.19089314341545105,
+      "learning_rate": 0.0001712858954025604,
+      "loss": 0.9308,
+      "step": 2817
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.17796491086483002,
+      "learning_rate": 0.00017126647417665107,
+      "loss": 1.0109,
+      "step": 2818
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.19152474403381348,
+      "learning_rate": 0.0001712470474868831,
+      "loss": 0.9048,
+      "step": 2819
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.22874225676059723,
+      "learning_rate": 0.00017122761533474586,
+      "loss": 0.9911,
+      "step": 2820
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1857227236032486,
+      "learning_rate": 0.00017120817772172916,
+      "loss": 0.9237,
+      "step": 2821
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.18604129552841187,
+      "learning_rate": 0.00017118873464932327,
+      "loss": 0.9447,
+      "step": 2822
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.20199713110923767,
+      "learning_rate": 0.00017116928611901885,
+      "loss": 0.9754,
+      "step": 2823
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.20207203924655914,
+      "learning_rate": 0.000171149832132307,
+      "loss": 1.015,
+      "step": 2824
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.18461887538433075,
+      "learning_rate": 0.00017113037269067912,
+      "loss": 0.9923,
+      "step": 2825
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.187242791056633,
+      "learning_rate": 0.00017111090779562725,
+      "loss": 0.927,
+      "step": 2826
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.18785372376441956,
+      "learning_rate": 0.0001710914374486437,
+      "loss": 0.9408,
+      "step": 2827
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.18827159702777863,
+      "learning_rate": 0.00017107196165122113,
+      "loss": 0.9365,
+      "step": 2828
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1842368096113205,
+      "learning_rate": 0.00017105248040485282,
+      "loss": 0.943,
+      "step": 2829
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.18567433953285217,
+      "learning_rate": 0.0001710329937110323,
+      "loss": 0.9705,
+      "step": 2830
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.18705464899539948,
+      "learning_rate": 0.0001710135015712536,
+      "loss": 1.0149,
+      "step": 2831
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.19440047442913055,
+      "learning_rate": 0.00017099400398701113,
+      "loss": 0.9546,
+      "step": 2832
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.19255554676055908,
+      "learning_rate": 0.00017097450095979975,
+      "loss": 0.9735,
+      "step": 2833
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1825937181711197,
+      "learning_rate": 0.00017095499249111467,
+      "loss": 0.9555,
+      "step": 2834
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.19005203247070312,
+      "learning_rate": 0.00017093547858245162,
+      "loss": 0.9334,
+      "step": 2835
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.18393410742282867,
+      "learning_rate": 0.0001709159592353067,
+      "loss": 0.9209,
+      "step": 2836
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.18357278406620026,
+      "learning_rate": 0.00017089643445117637,
+      "loss": 0.9059,
+      "step": 2837
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1842675656080246,
+      "learning_rate": 0.00017087690423155756,
+      "loss": 0.9593,
+      "step": 2838
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.19786691665649414,
+      "learning_rate": 0.00017085736857794763,
+      "loss": 1.015,
+      "step": 2839
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.18587841093540192,
+      "learning_rate": 0.00017083782749184435,
+      "loss": 0.9738,
+      "step": 2840
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 0.9591795802116394,
+      "eval_runtime": 679.171,
+      "eval_samples_per_second": 10.124,
+      "eval_steps_per_second": 5.062,
+      "step": 2840
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1888989806175232,
+      "learning_rate": 0.0001708182809747459,
+      "loss": 0.954,
+      "step": 2841
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1919054090976715,
+      "learning_rate": 0.00017079872902815084,
+      "loss": 0.9808,
+      "step": 2842
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.20058830082416534,
+      "learning_rate": 0.00017077917165355823,
+      "loss": 1.0177,
+      "step": 2843
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.17403985559940338,
+      "learning_rate": 0.00017075960885246745,
+      "loss": 0.8599,
+      "step": 2844
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.19590140879154205,
+      "learning_rate": 0.00017074004062637833,
+      "loss": 0.9614,
+      "step": 2845
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.20473170280456543,
+      "learning_rate": 0.00017072046697679118,
+      "loss": 0.9607,
+      "step": 2846
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1959337592124939,
+      "learning_rate": 0.00017070088790520664,
+      "loss": 0.9394,
+      "step": 2847
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.17946115136146545,
+      "learning_rate": 0.0001706813034131258,
+      "loss": 0.9491,
+      "step": 2848
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1879536658525467,
+      "learning_rate": 0.00017066171350205012,
+      "loss": 0.931,
+      "step": 2849
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.18565255403518677,
+      "learning_rate": 0.00017064211817348164,
+      "loss": 0.8425,
+      "step": 2850
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.18562796711921692,
+      "learning_rate": 0.00017062251742892256,
+      "loss": 0.9054,
+      "step": 2851
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1931876242160797,
+      "learning_rate": 0.00017060291126987572,
+      "loss": 0.9934,
+      "step": 2852
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.19624903798103333,
+      "learning_rate": 0.00017058329969784426,
+      "loss": 0.9625,
+      "step": 2853
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.19228866696357727,
+      "learning_rate": 0.00017056368271433175,
+      "loss": 0.9353,
+      "step": 2854
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.19154101610183716,
+      "learning_rate": 0.0001705440603208422,
+      "loss": 0.9374,
+      "step": 2855
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.18350714445114136,
+      "learning_rate": 0.00017052443251888,
+      "loss": 0.9005,
+      "step": 2856
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.18323072791099548,
+      "learning_rate": 0.00017050479930995,
+      "loss": 0.9087,
+      "step": 2857
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.18845704197883606,
+      "learning_rate": 0.0001704851606955574,
+      "loss": 0.9322,
+      "step": 2858
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.19670087099075317,
+      "learning_rate": 0.00017046551667720787,
+      "loss": 0.9987,
+      "step": 2859
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.18626806139945984,
+      "learning_rate": 0.00017044586725640756,
+      "loss": 0.9809,
+      "step": 2860
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.19881881773471832,
+      "learning_rate": 0.00017042621243466277,
+      "loss": 0.9545,
+      "step": 2861
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.19658763706684113,
+      "learning_rate": 0.00017040655221348057,
+      "loss": 0.9976,
+      "step": 2862
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.1838339865207672,
+      "learning_rate": 0.0001703868865943682,
+      "loss": 0.9416,
+      "step": 2863
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.18808238208293915,
+      "learning_rate": 0.00017036721557883334,
+      "loss": 0.9292,
+      "step": 2864
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.19379892945289612,
+      "learning_rate": 0.00017034753916838423,
+      "loss": 1.0161,
+      "step": 2865
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.1916080266237259,
+      "learning_rate": 0.00017032785736452937,
+      "loss": 0.9748,
+      "step": 2866
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.19743172824382782,
+      "learning_rate": 0.00017030817016877768,
+      "loss": 1.0049,
+      "step": 2867
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.18762719631195068,
+      "learning_rate": 0.0001702884775826386,
+      "loss": 0.915,
+      "step": 2868
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.17494773864746094,
+      "learning_rate": 0.0001702687796076219,
+      "loss": 0.9419,
+      "step": 2869
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.18630997836589813,
+      "learning_rate": 0.0001702490762452378,
+      "loss": 0.9856,
+      "step": 2870
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.1823118031024933,
+      "learning_rate": 0.0001702293674969969,
+      "loss": 0.9582,
+      "step": 2871
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.201270192861557,
+      "learning_rate": 0.0001702096533644102,
+      "loss": 1.0076,
+      "step": 2872
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.17998550832271576,
+      "learning_rate": 0.00017018993384898922,
+      "loss": 0.8525,
+      "step": 2873
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.19493627548217773,
+      "learning_rate": 0.00017017020895224573,
+      "loss": 0.9494,
+      "step": 2874
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.18817171454429626,
+      "learning_rate": 0.0001701504786756921,
+      "loss": 0.9834,
+      "step": 2875
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.1890181452035904,
+      "learning_rate": 0.0001701307430208409,
+      "loss": 0.939,
+      "step": 2876
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.186468243598938,
+      "learning_rate": 0.0001701110019892053,
+      "loss": 0.8638,
+      "step": 2877
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.192847341299057,
+      "learning_rate": 0.00017009125558229874,
+      "loss": 0.9671,
+      "step": 2878
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.18834351003170013,
+      "learning_rate": 0.00017007150380163522,
+      "loss": 0.9366,
+      "step": 2879
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.1983274221420288,
+      "learning_rate": 0.00017005174664872904,
+      "loss": 0.9598,
+      "step": 2880
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.18392324447631836,
+      "learning_rate": 0.00017003198412509488,
+      "loss": 0.9406,
+      "step": 2881
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.19125406444072723,
+      "learning_rate": 0.00017001221623224797,
+      "loss": 0.9212,
+      "step": 2882
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.19399990141391754,
+      "learning_rate": 0.00016999244297170383,
+      "loss": 0.9889,
+      "step": 2883
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.18320515751838684,
+      "learning_rate": 0.00016997266434497848,
+      "loss": 0.9134,
+      "step": 2884
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.18350949883460999,
+      "learning_rate": 0.00016995288035358828,
+      "loss": 0.9717,
+      "step": 2885
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.18024645745754242,
+      "learning_rate": 0.00016993309099905003,
+      "loss": 0.9331,
+      "step": 2886
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.18857190012931824,
+      "learning_rate": 0.00016991329628288093,
+      "loss": 0.9852,
+      "step": 2887
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.18607757985591888,
+      "learning_rate": 0.00016989349620659866,
+      "loss": 0.9123,
+      "step": 2888
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.18504849076271057,
+      "learning_rate": 0.00016987369077172115,
+      "loss": 0.9228,
+      "step": 2889
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.18795835971832275,
+      "learning_rate": 0.00016985387997976693,
+      "loss": 0.889,
+      "step": 2890
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.19008661806583405,
+      "learning_rate": 0.0001698340638322548,
+      "loss": 0.9829,
+      "step": 2891
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.19802813231945038,
+      "learning_rate": 0.00016981424233070404,
+      "loss": 0.9692,
+      "step": 2892
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.20054937899112701,
+      "learning_rate": 0.00016979441547663435,
+      "loss": 0.939,
+      "step": 2893
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.18934161961078644,
+      "learning_rate": 0.00016977458327156582,
+      "loss": 0.9009,
+      "step": 2894
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.1872309148311615,
+      "learning_rate": 0.0001697547457170189,
+      "loss": 0.9415,
+      "step": 2895
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.1852417290210724,
+      "learning_rate": 0.00016973490281451455,
+      "loss": 0.9377,
+      "step": 2896
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.1943930685520172,
+      "learning_rate": 0.000169715054565574,
+      "loss": 0.9183,
+      "step": 2897
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.23046432435512543,
+      "learning_rate": 0.00016969520097171906,
+      "loss": 0.9881,
+      "step": 2898
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1862451732158661,
+      "learning_rate": 0.00016967534203447184,
+      "loss": 0.8953,
+      "step": 2899
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.18676815927028656,
+      "learning_rate": 0.00016965547775535492,
+      "loss": 0.9184,
+      "step": 2900
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.18597565591335297,
+      "learning_rate": 0.00016963560813589116,
+      "loss": 0.9172,
+      "step": 2901
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1836593896150589,
+      "learning_rate": 0.000169615733177604,
+      "loss": 0.9615,
+      "step": 2902
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2000277191400528,
+      "learning_rate": 0.00016959585288201719,
+      "loss": 0.9409,
+      "step": 2903
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1949191689491272,
+      "learning_rate": 0.0001695759672506549,
+      "loss": 0.8877,
+      "step": 2904
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.19184264540672302,
+      "learning_rate": 0.0001695560762850418,
+      "loss": 0.9198,
+      "step": 2905
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.18635979294776917,
+      "learning_rate": 0.00016953617998670279,
+      "loss": 0.9238,
+      "step": 2906
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.18243515491485596,
+      "learning_rate": 0.0001695162783571633,
+      "loss": 0.9363,
+      "step": 2907
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1891496181488037,
+      "learning_rate": 0.0001694963713979492,
+      "loss": 0.9342,
+      "step": 2908
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.19546544551849365,
+      "learning_rate": 0.00016947645911058668,
+      "loss": 0.944,
+      "step": 2909
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.18900978565216064,
+      "learning_rate": 0.00016945654149660237,
+      "loss": 0.8945,
+      "step": 2910
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1874096542596817,
+      "learning_rate": 0.0001694366185575233,
+      "loss": 0.9204,
+      "step": 2911
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.18490877747535706,
+      "learning_rate": 0.00016941669029487698,
+      "loss": 0.9533,
+      "step": 2912
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.18125925958156586,
+      "learning_rate": 0.00016939675671019124,
+      "loss": 0.8697,
+      "step": 2913
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.19065441191196442,
+      "learning_rate": 0.00016937681780499432,
+      "loss": 0.887,
+      "step": 2914
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.18740889430046082,
+      "learning_rate": 0.00016935687358081493,
+      "loss": 0.9241,
+      "step": 2915
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.1898462325334549,
+      "learning_rate": 0.00016933692403918219,
+      "loss": 0.9273,
+      "step": 2916
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.19207645952701569,
+      "learning_rate": 0.00016931696918162547,
+      "loss": 0.9765,
+      "step": 2917
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.19262659549713135,
+      "learning_rate": 0.0001692970090096748,
+      "loss": 0.9454,
+      "step": 2918
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.19295157492160797,
+      "learning_rate": 0.00016927704352486038,
+      "loss": 0.9456,
+      "step": 2919
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.20015589892864227,
+      "learning_rate": 0.000169257072728713,
+      "loss": 1.0075,
+      "step": 2920
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.177590474486351,
+      "learning_rate": 0.00016923709662276378,
+      "loss": 0.8629,
+      "step": 2921
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.19473347067832947,
+      "learning_rate": 0.00016921711520854422,
+      "loss": 0.9014,
+      "step": 2922
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.19092902541160583,
+      "learning_rate": 0.0001691971284875862,
+      "loss": 0.9194,
+      "step": 2923
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.18545444309711456,
+      "learning_rate": 0.00016917713646142222,
+      "loss": 0.9279,
+      "step": 2924
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.19252456724643707,
+      "learning_rate": 0.00016915713913158485,
+      "loss": 0.9169,
+      "step": 2925
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.18508480489253998,
+      "learning_rate": 0.00016913713649960738,
+      "loss": 0.8767,
+      "step": 2926
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.190322607755661,
+      "learning_rate": 0.0001691171285670233,
+      "loss": 0.8883,
+      "step": 2927
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.19125689566135406,
+      "learning_rate": 0.0001690971153353666,
+      "loss": 0.986,
+      "step": 2928
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.19923120737075806,
+      "learning_rate": 0.00016907709680617162,
+      "loss": 0.9492,
+      "step": 2929
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.18744708597660065,
+      "learning_rate": 0.00016905707298097323,
+      "loss": 0.8807,
+      "step": 2930
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.19923172891139984,
+      "learning_rate": 0.00016903704386130652,
+      "loss": 0.9619,
+      "step": 2931
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.17846401035785675,
+      "learning_rate": 0.00016901700944870713,
+      "loss": 0.8608,
+      "step": 2932
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.18846310675144196,
+      "learning_rate": 0.00016899696974471105,
+      "loss": 0.9406,
+      "step": 2933
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.1996917575597763,
+      "learning_rate": 0.00016897692475085468,
+      "loss": 0.9684,
+      "step": 2934
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.18243834376335144,
+      "learning_rate": 0.00016895687446867485,
+      "loss": 0.8787,
+      "step": 2935
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.18757185339927673,
+      "learning_rate": 0.00016893681889970874,
+      "loss": 0.9692,
+      "step": 2936
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.1936056762933731,
+      "learning_rate": 0.00016891675804549402,
+      "loss": 1.0149,
+      "step": 2937
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.18731190264225006,
+      "learning_rate": 0.00016889669190756868,
+      "loss": 0.9039,
+      "step": 2938
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.19040818512439728,
+      "learning_rate": 0.00016887662048747112,
+      "loss": 0.9222,
+      "step": 2939
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.19378627836704254,
+      "learning_rate": 0.00016885654378674024,
+      "loss": 0.9511,
+      "step": 2940
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.1890939325094223,
+      "learning_rate": 0.00016883646180691525,
+      "loss": 0.9387,
+      "step": 2941
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.19378086924552917,
+      "learning_rate": 0.00016881637454953583,
+      "loss": 0.8926,
+      "step": 2942
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.18797039985656738,
+      "learning_rate": 0.00016879628201614194,
+      "loss": 0.9298,
+      "step": 2943
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.21416975557804108,
+      "learning_rate": 0.00016877618420827413,
+      "loss": 0.9618,
+      "step": 2944
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.18781521916389465,
+      "learning_rate": 0.00016875608112747322,
+      "loss": 0.898,
+      "step": 2945
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.18484070897102356,
+      "learning_rate": 0.00016873597277528046,
+      "loss": 0.9284,
+      "step": 2946
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.20203427970409393,
+      "learning_rate": 0.00016871585915323757,
+      "loss": 0.9545,
+      "step": 2947
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.19452163577079773,
+      "learning_rate": 0.00016869574026288654,
+      "loss": 0.9575,
+      "step": 2948
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.1887548416852951,
+      "learning_rate": 0.0001686756161057699,
+      "loss": 0.9401,
+      "step": 2949
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.19443704187870026,
+      "learning_rate": 0.00016865548668343054,
+      "loss": 0.9254,
+      "step": 2950
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.19337326288223267,
+      "learning_rate": 0.0001686353519974117,
+      "loss": 0.9227,
+      "step": 2951
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.2151598483324051,
+      "learning_rate": 0.0001686152120492571,
+      "loss": 0.8004,
+      "step": 2952
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.19231447577476501,
+      "learning_rate": 0.0001685950668405108,
+      "loss": 0.9479,
+      "step": 2953
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.1957997977733612,
+      "learning_rate": 0.00016857491637271732,
+      "loss": 0.9058,
+      "step": 2954
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.19956451654434204,
+      "learning_rate": 0.00016855476064742155,
+      "loss": 0.9873,
+      "step": 2955
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.18961627781391144,
+      "learning_rate": 0.00016853459966616877,
+      "loss": 0.8979,
+      "step": 2956
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.19322636723518372,
+      "learning_rate": 0.00016851443343050474,
+      "loss": 0.9682,
+      "step": 2957
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.20402783155441284,
+      "learning_rate": 0.0001684942619419755,
+      "loss": 0.8716,
+      "step": 2958
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.19145797193050385,
+      "learning_rate": 0.00016847408520212758,
+      "loss": 0.9521,
+      "step": 2959
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.18774887919425964,
+      "learning_rate": 0.00016845390321250787,
+      "loss": 0.8892,
+      "step": 2960
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.19743181765079498,
+      "learning_rate": 0.00016843371597466373,
+      "loss": 0.96,
+      "step": 2961
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.2033877819776535,
+      "learning_rate": 0.00016841352349014285,
+      "loss": 0.9533,
+      "step": 2962
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.19369393587112427,
+      "learning_rate": 0.00016839332576049334,
+      "loss": 0.9296,
+      "step": 2963
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.1883126199245453,
+      "learning_rate": 0.0001683731227872637,
+      "loss": 0.9543,
+      "step": 2964
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.1869036704301834,
+      "learning_rate": 0.00016835291457200293,
+      "loss": 0.9109,
+      "step": 2965
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.18386952579021454,
+      "learning_rate": 0.00016833270111626028,
+      "loss": 0.9149,
+      "step": 2966
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.18549568951129913,
+      "learning_rate": 0.0001683124824215855,
+      "loss": 0.9119,
+      "step": 2967
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.18437707424163818,
+      "learning_rate": 0.0001682922584895287,
+      "loss": 0.9272,
+      "step": 2968
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.1883966028690338,
+      "learning_rate": 0.00016827202932164046,
+      "loss": 0.9637,
+      "step": 2969
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.1802394688129425,
+      "learning_rate": 0.00016825179491947165,
+      "loss": 0.9544,
+      "step": 2970
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.1978726089000702,
+      "learning_rate": 0.0001682315552845736,
+      "loss": 0.933,
+      "step": 2971
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.19730928540229797,
+      "learning_rate": 0.00016821131041849812,
+      "loss": 0.8994,
+      "step": 2972
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.18910084664821625,
+      "learning_rate": 0.00016819106032279723,
+      "loss": 0.9299,
+      "step": 2973
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.19059893488883972,
+      "learning_rate": 0.0001681708049990236,
+      "loss": 0.9231,
+      "step": 2974
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.19314156472682953,
+      "learning_rate": 0.00016815054444873002,
+      "loss": 0.9978,
+      "step": 2975
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.1947154849767685,
+      "learning_rate": 0.00016813027867346994,
+      "loss": 0.9429,
+      "step": 2976
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.19466917216777802,
+      "learning_rate": 0.000168110007674797,
+      "loss": 0.9488,
+      "step": 2977
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.19159191846847534,
+      "learning_rate": 0.00016808973145426547,
+      "loss": 0.9295,
+      "step": 2978
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.19307000935077667,
+      "learning_rate": 0.00016806945001342975,
+      "loss": 0.9588,
+      "step": 2979
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.19548317790031433,
+      "learning_rate": 0.00016804916335384486,
+      "loss": 0.9053,
+      "step": 2980
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.19848956167697906,
+      "learning_rate": 0.0001680288714770661,
+      "loss": 0.9296,
+      "step": 2981
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.1890234500169754,
+      "learning_rate": 0.00016800857438464923,
+      "loss": 0.9,
+      "step": 2982
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.1944311261177063,
+      "learning_rate": 0.0001679882720781504,
+      "loss": 0.8962,
+      "step": 2983
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.19622524082660675,
+      "learning_rate": 0.0001679679645591261,
+      "loss": 0.9182,
+      "step": 2984
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.18798577785491943,
+      "learning_rate": 0.00016794765182913328,
+      "loss": 0.9858,
+      "step": 2985
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.19635619223117828,
+      "learning_rate": 0.00016792733388972932,
+      "loss": 0.904,
+      "step": 2986
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.19752800464630127,
+      "learning_rate": 0.0001679070107424719,
+      "loss": 0.9645,
+      "step": 2987
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.19207364320755005,
+      "learning_rate": 0.00016788668238891923,
+      "loss": 0.8897,
+      "step": 2988
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.1890457272529602,
+      "learning_rate": 0.00016786634883062974,
+      "loss": 0.9724,
+      "step": 2989
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.1933799535036087,
+      "learning_rate": 0.00016784601006916244,
+      "loss": 0.9378,
+      "step": 2990
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.18908622860908508,
+      "learning_rate": 0.00016782566610607663,
+      "loss": 0.9372,
+      "step": 2991
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.2023576945066452,
+      "learning_rate": 0.00016780531694293206,
+      "loss": 0.994,
+      "step": 2992
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.18966807425022125,
+      "learning_rate": 0.00016778496258128888,
+      "loss": 0.8883,
+      "step": 2993
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.20282705128192902,
+      "learning_rate": 0.00016776460302270758,
+      "loss": 0.9586,
+      "step": 2994
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.19105058908462524,
+      "learning_rate": 0.00016774423826874908,
+      "loss": 0.972,
+      "step": 2995
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.19293305277824402,
+      "learning_rate": 0.0001677238683209747,
+      "loss": 0.9297,
+      "step": 2996
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.19731226563453674,
+      "learning_rate": 0.00016770349318094623,
+      "loss": 0.8922,
+      "step": 2997
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.1933281421661377,
+      "learning_rate": 0.00016768311285022573,
+      "loss": 0.9455,
+      "step": 2998
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.19647730886936188,
+      "learning_rate": 0.00016766272733037576,
+      "loss": 0.9428,
+      "step": 2999
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.1985509991645813,
+      "learning_rate": 0.00016764233662295918,
+      "loss": 0.9212,
+      "step": 3000
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.18913675844669342,
+      "learning_rate": 0.00016762194072953937,
+      "loss": 0.8769,
+      "step": 3001
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.19756121933460236,
+      "learning_rate": 0.00016760153965167995,
+      "loss": 0.8908,
+      "step": 3002
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.19995391368865967,
+      "learning_rate": 0.00016758113339094516,
+      "loss": 0.9668,
+      "step": 3003
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.19399406015872955,
+      "learning_rate": 0.00016756072194889943,
+      "loss": 0.9075,
+      "step": 3004
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.19179430603981018,
+      "learning_rate": 0.00016754030532710763,
+      "loss": 0.8711,
+      "step": 3005
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.19073912501335144,
+      "learning_rate": 0.00016751988352713511,
+      "loss": 0.9328,
+      "step": 3006
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.18982064723968506,
+      "learning_rate": 0.00016749945655054763,
+      "loss": 0.9255,
+      "step": 3007
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.1822153478860855,
+      "learning_rate": 0.00016747902439891117,
+      "loss": 0.9393,
+      "step": 3008
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.18544809520244598,
+      "learning_rate": 0.00016745858707379227,
+      "loss": 0.8614,
+      "step": 3009
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.19058829545974731,
+      "learning_rate": 0.00016743814457675785,
+      "loss": 0.9478,
+      "step": 3010
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.19758155941963196,
+      "learning_rate": 0.00016741769690937512,
+      "loss": 0.9714,
+      "step": 3011
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.1888519525527954,
+      "learning_rate": 0.00016739724407321185,
+      "loss": 0.8456,
+      "step": 3012
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.19396807253360748,
+      "learning_rate": 0.00016737678606983607,
+      "loss": 0.8806,
+      "step": 3013
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.20305144786834717,
+      "learning_rate": 0.00016735632290081624,
+      "loss": 0.9567,
+      "step": 3014
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.1895797699689865,
+      "learning_rate": 0.00016733585456772128,
+      "loss": 0.9095,
+      "step": 3015
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.1943138986825943,
+      "learning_rate": 0.00016731538107212039,
+      "loss": 0.9681,
+      "step": 3016
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.18373315036296844,
+      "learning_rate": 0.0001672949024155833,
+      "loss": 0.9072,
+      "step": 3017
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.19009727239608765,
+      "learning_rate": 0.00016727441859968004,
+      "loss": 0.9642,
+      "step": 3018
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.19613289833068848,
+      "learning_rate": 0.00016725392962598106,
+      "loss": 0.9783,
+      "step": 3019
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.18772883713245392,
+      "learning_rate": 0.0001672334354960572,
+      "loss": 0.8989,
+      "step": 3020
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.19232602417469025,
+      "learning_rate": 0.00016721293621147968,
+      "loss": 0.9428,
+      "step": 3021
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.19024088978767395,
+      "learning_rate": 0.00016719243177382024,
+      "loss": 0.8916,
+      "step": 3022
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.19054242968559265,
+      "learning_rate": 0.00016717192218465077,
+      "loss": 0.9239,
+      "step": 3023
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.2045336663722992,
+      "learning_rate": 0.00016715140744554384,
+      "loss": 0.9251,
+      "step": 3024
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.19345614314079285,
+      "learning_rate": 0.0001671308875580722,
+      "loss": 0.9296,
+      "step": 3025
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.20464305579662323,
+      "learning_rate": 0.00016711036252380909,
+      "loss": 0.9632,
+      "step": 3026
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.1874096393585205,
+      "learning_rate": 0.00016708983234432808,
+      "loss": 0.8621,
+      "step": 3027
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.19258873164653778,
+      "learning_rate": 0.00016706929702120323,
+      "loss": 0.9658,
+      "step": 3028
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.19458803534507751,
+      "learning_rate": 0.00016704875655600897,
+      "loss": 0.9901,
+      "step": 3029
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.19785954058170319,
+      "learning_rate": 0.00016702821095031998,
+      "loss": 0.9192,
+      "step": 3030
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.20203198492527008,
+      "learning_rate": 0.00016700766020571157,
+      "loss": 0.9488,
+      "step": 3031
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.18694931268692017,
+      "learning_rate": 0.00016698710432375925,
+      "loss": 0.9295,
+      "step": 3032
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.2125498652458191,
+      "learning_rate": 0.000166966543306039,
+      "loss": 0.9405,
+      "step": 3033
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.1959962248802185,
+      "learning_rate": 0.00016694597715412727,
+      "loss": 0.9279,
+      "step": 3034
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.19424889981746674,
+      "learning_rate": 0.00016692540586960074,
+      "loss": 0.9329,
+      "step": 3035
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.1940089464187622,
+      "learning_rate": 0.0001669048294540366,
+      "loss": 0.9566,
+      "step": 3036
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.1906086653470993,
+      "learning_rate": 0.00016688424790901242,
+      "loss": 0.9052,
+      "step": 3037
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.19040682911872864,
+      "learning_rate": 0.00016686366123610613,
+      "loss": 0.921,
+      "step": 3038
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.20232956111431122,
+      "learning_rate": 0.00016684306943689608,
+      "loss": 0.9927,
+      "step": 3039
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.20033952593803406,
+      "learning_rate": 0.000166822472512961,
+      "loss": 0.925,
+      "step": 3040
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.1834961622953415,
+      "learning_rate": 0.00016680187046588,
+      "loss": 0.8689,
+      "step": 3041
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.19811643660068512,
+      "learning_rate": 0.00016678126329723258,
+      "loss": 0.9067,
+      "step": 3042
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.1838274598121643,
+      "learning_rate": 0.00016676065100859868,
+      "loss": 0.9123,
+      "step": 3043
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.19470977783203125,
+      "learning_rate": 0.00016674003360155862,
+      "loss": 0.9938,
+      "step": 3044
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.1993788778781891,
+      "learning_rate": 0.00016671941107769306,
+      "loss": 0.945,
+      "step": 3045
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.19459454715251923,
+      "learning_rate": 0.00016669878343858309,
+      "loss": 0.8592,
+      "step": 3046
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.1934826523065567,
+      "learning_rate": 0.00016667815068581026,
+      "loss": 0.8991,
+      "step": 3047
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.19478553533554077,
+      "learning_rate": 0.00016665751282095634,
+      "loss": 0.933,
+      "step": 3048
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.18333174288272858,
+      "learning_rate": 0.00016663686984560365,
+      "loss": 0.8764,
+      "step": 3049
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.18906213343143463,
+      "learning_rate": 0.00016661622176133482,
+      "loss": 0.8985,
+      "step": 3050
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.19240978360176086,
+      "learning_rate": 0.0001665955685697329,
+      "loss": 0.9382,
+      "step": 3051
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.18707260489463806,
+      "learning_rate": 0.00016657491027238138,
+      "loss": 0.9029,
+      "step": 3052
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.19875676929950714,
+      "learning_rate": 0.00016655424687086402,
+      "loss": 0.9134,
+      "step": 3053
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.20501351356506348,
+      "learning_rate": 0.0001665335783667651,
+      "loss": 0.9141,
+      "step": 3054
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.2028193324804306,
+      "learning_rate": 0.0001665129047616692,
+      "loss": 0.9328,
+      "step": 3055
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.1942424327135086,
+      "learning_rate": 0.00016649222605716132,
+      "loss": 0.9344,
+      "step": 3056
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.19648103415966034,
+      "learning_rate": 0.00016647154225482688,
+      "loss": 0.9037,
+      "step": 3057
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.19552026689052582,
+      "learning_rate": 0.00016645085335625166,
+      "loss": 0.9409,
+      "step": 3058
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.20716822147369385,
+      "learning_rate": 0.0001664301593630218,
+      "loss": 0.9503,
+      "step": 3059
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.19085535407066345,
+      "learning_rate": 0.00016640946027672392,
+      "loss": 0.9805,
+      "step": 3060
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.19183199107646942,
+      "learning_rate": 0.00016638875609894494,
+      "loss": 0.9525,
+      "step": 3061
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.20048820972442627,
+      "learning_rate": 0.00016636804683127222,
+      "loss": 0.9452,
+      "step": 3062
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.19393523037433624,
+      "learning_rate": 0.0001663473324752935,
+      "loss": 0.9248,
+      "step": 3063
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.19443608820438385,
+      "learning_rate": 0.00016632661303259691,
+      "loss": 0.9816,
+      "step": 3064
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.20129789412021637,
+      "learning_rate": 0.00016630588850477098,
+      "loss": 0.9673,
+      "step": 3065
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.19859552383422852,
+      "learning_rate": 0.0001662851588934046,
+      "loss": 0.9159,
+      "step": 3066
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.20427550375461578,
+      "learning_rate": 0.00016626442420008706,
+      "loss": 0.9684,
+      "step": 3067
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.2060311883687973,
+      "learning_rate": 0.00016624368442640807,
+      "loss": 0.9626,
+      "step": 3068
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.20656219124794006,
+      "learning_rate": 0.0001662229395739577,
+      "loss": 0.9244,
+      "step": 3069
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.19876663386821747,
+      "learning_rate": 0.00016620218964432643,
+      "loss": 0.9035,
+      "step": 3070
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.1983165442943573,
+      "learning_rate": 0.00016618143463910506,
+      "loss": 0.9005,
+      "step": 3071
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.19417667388916016,
+      "learning_rate": 0.00016616067455988491,
+      "loss": 0.9707,
+      "step": 3072
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.19066518545150757,
+      "learning_rate": 0.00016613990940825754,
+      "loss": 0.941,
+      "step": 3073
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.19258198142051697,
+      "learning_rate": 0.00016611913918581505,
+      "loss": 0.9481,
+      "step": 3074
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.2004285454750061,
+      "learning_rate": 0.0001660983638941498,
+      "loss": 0.9491,
+      "step": 3075
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.19285230338573456,
+      "learning_rate": 0.00016607758353485462,
+      "loss": 0.9509,
+      "step": 3076
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.18702608346939087,
+      "learning_rate": 0.00016605679810952268,
+      "loss": 0.8851,
+      "step": 3077
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.1984139382839203,
+      "learning_rate": 0.00016603600761974755,
+      "loss": 0.9796,
+      "step": 3078
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.20252801477909088,
+      "learning_rate": 0.00016601521206712318,
+      "loss": 0.9725,
+      "step": 3079
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.20916445553302765,
+      "learning_rate": 0.000165994411453244,
+      "loss": 0.9426,
+      "step": 3080
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.19025819003582,
+      "learning_rate": 0.00016597360577970464,
+      "loss": 0.9129,
+      "step": 3081
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.19675807654857635,
+      "learning_rate": 0.00016595279504810035,
+      "loss": 0.9671,
+      "step": 3082
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.19846715033054352,
+      "learning_rate": 0.00016593197926002655,
+      "loss": 0.9286,
+      "step": 3083
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.19845189154148102,
+      "learning_rate": 0.00016591115841707918,
+      "loss": 0.9887,
+      "step": 3084
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.20557260513305664,
+      "learning_rate": 0.00016589033252085455,
+      "loss": 0.9767,
+      "step": 3085
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.19639475643634796,
+      "learning_rate": 0.00016586950157294932,
+      "loss": 0.9543,
+      "step": 3086
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.19399593770503998,
+      "learning_rate": 0.00016584866557496056,
+      "loss": 0.8986,
+      "step": 3087
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.1935609132051468,
+      "learning_rate": 0.00016582782452848575,
+      "loss": 0.9144,
+      "step": 3088
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.2042001187801361,
+      "learning_rate": 0.00016580697843512267,
+      "loss": 0.984,
+      "step": 3089
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.19425135850906372,
+      "learning_rate": 0.00016578612729646963,
+      "loss": 0.9508,
+      "step": 3090
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.19415859878063202,
+      "learning_rate": 0.00016576527111412516,
+      "loss": 0.899,
+      "step": 3091
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.18906068801879883,
+      "learning_rate": 0.00016574440988968833,
+      "loss": 0.9146,
+      "step": 3092
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.19608232378959656,
+      "learning_rate": 0.0001657235436247585,
+      "loss": 0.9263,
+      "step": 3093
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.19680467247962952,
+      "learning_rate": 0.00016570267232093544,
+      "loss": 0.8997,
+      "step": 3094
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.20155908167362213,
+      "learning_rate": 0.00016568179597981932,
+      "loss": 0.9227,
+      "step": 3095
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.19624723494052887,
+      "learning_rate": 0.0001656609146030107,
+      "loss": 0.8891,
+      "step": 3096
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.19039687514305115,
+      "learning_rate": 0.00016564002819211048,
+      "loss": 0.8529,
+      "step": 3097
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.19904066622257233,
+      "learning_rate": 0.00016561913674872,
+      "loss": 0.9535,
+      "step": 3098
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.1853899508714676,
+      "learning_rate": 0.00016559824027444098,
+      "loss": 0.8376,
+      "step": 3099
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.19816075265407562,
+      "learning_rate": 0.00016557733877087548,
+      "loss": 0.9628,
+      "step": 3100
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.20754671096801758,
+      "learning_rate": 0.000165556432239626,
+      "loss": 0.9895,
+      "step": 3101
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.1962323784828186,
+      "learning_rate": 0.0001655355206822954,
+      "loss": 0.948,
+      "step": 3102
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.1916942596435547,
+      "learning_rate": 0.0001655146041004869,
+      "loss": 0.924,
+      "step": 3103
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.1882028877735138,
+      "learning_rate": 0.00016549368249580416,
+      "loss": 0.905,
+      "step": 3104
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.1991576850414276,
+      "learning_rate": 0.00016547275586985122,
+      "loss": 0.9417,
+      "step": 3105
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.20659486949443817,
+      "learning_rate": 0.00016545182422423246,
+      "loss": 0.999,
+      "step": 3106
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.19963429868221283,
+      "learning_rate": 0.00016543088756055263,
+      "loss": 0.9741,
+      "step": 3107
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.2001790702342987,
+      "learning_rate": 0.00016540994588041695,
+      "loss": 0.8866,
+      "step": 3108
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.18648461997509003,
+      "learning_rate": 0.00016538899918543094,
+      "loss": 0.9174,
+      "step": 3109
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.1969536989927292,
+      "learning_rate": 0.0001653680474772006,
+      "loss": 0.9187,
+      "step": 3110
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.19499890506267548,
+      "learning_rate": 0.0001653470907573322,
+      "loss": 0.8722,
+      "step": 3111
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.20478104054927826,
+      "learning_rate": 0.00016532612902743246,
+      "loss": 0.8899,
+      "step": 3112
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.19237670302391052,
+      "learning_rate": 0.0001653051622891085,
+      "loss": 0.9049,
+      "step": 3113
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.20637308061122894,
+      "learning_rate": 0.0001652841905439678,
+      "loss": 0.9914,
+      "step": 3114
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.2024182230234146,
+      "learning_rate": 0.00016526321379361816,
+      "loss": 0.8817,
+      "step": 3115
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.19620735943317413,
+      "learning_rate": 0.0001652422320396679,
+      "loss": 0.9344,
+      "step": 3116
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.1945084184408188,
+      "learning_rate": 0.00016522124528372563,
+      "loss": 1.0151,
+      "step": 3117
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.20316572487354279,
+      "learning_rate": 0.0001652002535274003,
+      "loss": 1.0103,
+      "step": 3118
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.197440043091774,
+      "learning_rate": 0.0001651792567723014,
+      "loss": 0.9926,
+      "step": 3119
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.1884462982416153,
+      "learning_rate": 0.00016515825502003866,
+      "loss": 0.9498,
+      "step": 3120
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.19690294563770294,
+      "learning_rate": 0.00016513724827222227,
+      "loss": 0.9016,
+      "step": 3121
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.1945369392633438,
+      "learning_rate": 0.00016511623653046268,
+      "loss": 0.9326,
+      "step": 3122
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.20239035785198212,
+      "learning_rate": 0.00016509521979637094,
+      "loss": 0.9521,
+      "step": 3123
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.19082894921302795,
+      "learning_rate": 0.0001650741980715583,
+      "loss": 0.9664,
+      "step": 3124
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.19143463671207428,
+      "learning_rate": 0.00016505317135763652,
+      "loss": 0.8994,
+      "step": 3125
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.1881125271320343,
+      "learning_rate": 0.0001650321396562176,
+      "loss": 0.9712,
+      "step": 3126
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.19766877591609955,
+      "learning_rate": 0.000165011102968914,
+      "loss": 0.9933,
+      "step": 3127
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.19713425636291504,
+      "learning_rate": 0.00016499006129733857,
+      "loss": 0.9075,
+      "step": 3128
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.1889505535364151,
+      "learning_rate": 0.00016496901464310457,
+      "loss": 0.9361,
+      "step": 3129
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.19792018830776215,
+      "learning_rate": 0.0001649479630078256,
+      "loss": 0.9492,
+      "step": 3130
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.19634611904621124,
+      "learning_rate": 0.00016492690639311562,
+      "loss": 0.9335,
+      "step": 3131
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.19777911901474,
+      "learning_rate": 0.000164905844800589,
+      "loss": 0.9047,
+      "step": 3132
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.1970728635787964,
+      "learning_rate": 0.00016488477823186048,
+      "loss": 0.9206,
+      "step": 3133
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.20032721757888794,
+      "learning_rate": 0.00016486370668854523,
+      "loss": 0.9584,
+      "step": 3134
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.19533273577690125,
+      "learning_rate": 0.00016484263017225875,
+      "loss": 0.9534,
+      "step": 3135
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.1874852180480957,
+      "learning_rate": 0.00016482154868461692,
+      "loss": 0.8876,
+      "step": 3136
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.207426518201828,
+      "learning_rate": 0.00016480046222723603,
+      "loss": 0.9597,
+      "step": 3137
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.1896134465932846,
+      "learning_rate": 0.0001647793708017327,
+      "loss": 0.9288,
+      "step": 3138
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.19841088354587555,
+      "learning_rate": 0.000164758274409724,
+      "loss": 0.9661,
+      "step": 3139
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.1989322006702423,
+      "learning_rate": 0.00016473717305282742,
+      "loss": 0.9517,
+      "step": 3140
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.19085904955863953,
+      "learning_rate": 0.00016471606673266066,
+      "loss": 0.9135,
+      "step": 3141
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.1852741837501526,
+      "learning_rate": 0.00016469495545084185,
+      "loss": 0.8794,
+      "step": 3142
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.19000276923179626,
+      "learning_rate": 0.00016467383920898965,
+      "loss": 0.8926,
+      "step": 3143
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.18369747698307037,
+      "learning_rate": 0.00016465271800872303,
+      "loss": 0.8703,
+      "step": 3144
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.19372214376926422,
+      "learning_rate": 0.0001646315918516612,
+      "loss": 0.9348,
+      "step": 3145
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.1902085244655609,
+      "learning_rate": 0.00016461046073942397,
+      "loss": 0.8901,
+      "step": 3146
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.19396044313907623,
+      "learning_rate": 0.00016458932467363133,
+      "loss": 0.9754,
+      "step": 3147
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.19111724197864532,
+      "learning_rate": 0.00016456818365590377,
+      "loss": 0.9584,
+      "step": 3148
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.18766222894191742,
+      "learning_rate": 0.00016454703768786216,
+      "loss": 0.8653,
+      "step": 3149
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.1874326914548874,
+      "learning_rate": 0.00016452588677112767,
+      "loss": 0.8799,
+      "step": 3150
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.19378015398979187,
+      "learning_rate": 0.00016450473090732194,
+      "loss": 0.945,
+      "step": 3151
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.18832823634147644,
+      "learning_rate": 0.00016448357009806692,
+      "loss": 0.8718,
+      "step": 3152
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.1912168264389038,
+      "learning_rate": 0.000164462404344985,
+      "loss": 0.8851,
+      "step": 3153
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.1954006403684616,
+      "learning_rate": 0.00016444123364969885,
+      "loss": 0.9769,
+      "step": 3154
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.19893434643745422,
+      "learning_rate": 0.00016442005801383167,
+      "loss": 0.9268,
+      "step": 3155
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.19670341908931732,
+      "learning_rate": 0.00016439887743900693,
+      "loss": 0.9658,
+      "step": 3156
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.19159068167209625,
+      "learning_rate": 0.00016437769192684845,
+      "loss": 0.8905,
+      "step": 3157
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.18277385830879211,
+      "learning_rate": 0.00016435650147898052,
+      "loss": 0.832,
+      "step": 3158
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.19444088637828827,
+      "learning_rate": 0.00016433530609702776,
+      "loss": 0.9398,
+      "step": 3159
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.1958342045545578,
+      "learning_rate": 0.00016431410578261524,
+      "loss": 0.8834,
+      "step": 3160
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.19928790628910065,
+      "learning_rate": 0.00016429290053736822,
+      "loss": 0.9126,
+      "step": 3161
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.1952957957983017,
+      "learning_rate": 0.0001642716903629126,
+      "loss": 0.9472,
+      "step": 3162
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.18636207282543182,
+      "learning_rate": 0.0001642504752608744,
+      "loss": 0.9279,
+      "step": 3163
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.18926826119422913,
+      "learning_rate": 0.00016422925523288022,
+      "loss": 0.9201,
+      "step": 3164
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.1961440145969391,
+      "learning_rate": 0.00016420803028055692,
+      "loss": 0.998,
+      "step": 3165
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.19415853917598724,
+      "learning_rate": 0.00016418680040553181,
+      "loss": 0.9176,
+      "step": 3166
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.1978716403245926,
+      "learning_rate": 0.0001641655656094325,
+      "loss": 0.9559,
+      "step": 3167
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.19432327151298523,
+      "learning_rate": 0.00016414432589388706,
+      "loss": 0.8507,
+      "step": 3168
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.19534040987491608,
+      "learning_rate": 0.00016412308126052385,
+      "loss": 0.927,
+      "step": 3169
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.1897934377193451,
+      "learning_rate": 0.0001641018317109717,
+      "loss": 0.8587,
+      "step": 3170
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.20288267731666565,
+      "learning_rate": 0.00016408057724685976,
+      "loss": 0.9847,
+      "step": 3171
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.18903978168964386,
+      "learning_rate": 0.00016405931786981755,
+      "loss": 0.9283,
+      "step": 3172
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.1943960338830948,
+      "learning_rate": 0.00016403805358147495,
+      "loss": 0.936,
+      "step": 3173
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.18546950817108154,
+      "learning_rate": 0.00016401678438346235,
+      "loss": 0.967,
+      "step": 3174
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.19150441884994507,
+      "learning_rate": 0.0001639955102774103,
+      "loss": 0.9381,
+      "step": 3175
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.18767720460891724,
+      "learning_rate": 0.00016397423126494998,
+      "loss": 0.9682,
+      "step": 3176
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.19705381989479065,
+      "learning_rate": 0.0001639529473477127,
+      "loss": 0.8661,
+      "step": 3177
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.20014797151088715,
+      "learning_rate": 0.00016393165852733027,
+      "loss": 0.9955,
+      "step": 3178
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.1958370953798294,
+      "learning_rate": 0.00016391036480543488,
+      "loss": 0.9264,
+      "step": 3179
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.18295347690582275,
+      "learning_rate": 0.0001638890661836591,
+      "loss": 0.91,
+      "step": 3180
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.20895381271839142,
+      "learning_rate": 0.00016386776266363583,
+      "loss": 0.9342,
+      "step": 3181
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.1930595338344574,
+      "learning_rate": 0.00016384645424699835,
+      "loss": 0.9263,
+      "step": 3182
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.1938629299402237,
+      "learning_rate": 0.00016382514093538037,
+      "loss": 0.9294,
+      "step": 3183
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.1964857280254364,
+      "learning_rate": 0.00016380382273041593,
+      "loss": 0.9872,
+      "step": 3184
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.1964603066444397,
+      "learning_rate": 0.00016378249963373942,
+      "loss": 0.9155,
+      "step": 3185
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.20166955888271332,
+      "learning_rate": 0.00016376117164698567,
+      "loss": 0.9221,
+      "step": 3186
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.19599172472953796,
+      "learning_rate": 0.00016373983877178986,
+      "loss": 0.9071,
+      "step": 3187
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.196363627910614,
+      "learning_rate": 0.0001637185010097875,
+      "loss": 1.0266,
+      "step": 3188
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.1886577606201172,
+      "learning_rate": 0.00016369715836261458,
+      "loss": 0.9177,
+      "step": 3189
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.19875650107860565,
+      "learning_rate": 0.00016367581083190735,
+      "loss": 0.9816,
+      "step": 3190
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.19966274499893188,
+      "learning_rate": 0.0001636544584193025,
+      "loss": 0.8889,
+      "step": 3191
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.18415604531764984,
+      "learning_rate": 0.00016363310112643703,
+      "loss": 0.9384,
+      "step": 3192
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.17933693528175354,
+      "learning_rate": 0.00016361173895494845,
+      "loss": 0.875,
+      "step": 3193
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.19441203773021698,
+      "learning_rate": 0.0001635903719064745,
+      "loss": 0.9285,
+      "step": 3194
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.19981196522712708,
+      "learning_rate": 0.00016356899998265333,
+      "loss": 0.9445,
+      "step": 3195
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.18804876506328583,
+      "learning_rate": 0.00016354762318512354,
+      "loss": 0.9127,
+      "step": 3196
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.1986115276813507,
+      "learning_rate": 0.000163526241515524,
+      "loss": 0.9932,
+      "step": 3197
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.19924885034561157,
+      "learning_rate": 0.00016350485497549402,
+      "loss": 0.9489,
+      "step": 3198
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.1975104957818985,
+      "learning_rate": 0.00016348346356667324,
+      "loss": 1.0024,
+      "step": 3199
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.1849527508020401,
+      "learning_rate": 0.00016346206729070173,
+      "loss": 0.8958,
+      "step": 3200
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.1946130096912384,
+      "learning_rate": 0.00016344066614921987,
+      "loss": 0.8933,
+      "step": 3201
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.19796276092529297,
+      "learning_rate": 0.00016341926014386846,
+      "loss": 0.8946,
+      "step": 3202
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.19626247882843018,
+      "learning_rate": 0.00016339784927628867,
+      "loss": 0.9928,
+      "step": 3203
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.1890694499015808,
+      "learning_rate": 0.00016337643354812203,
+      "loss": 0.9041,
+      "step": 3204
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.19057592749595642,
+      "learning_rate": 0.00016335501296101037,
+      "loss": 0.9625,
+      "step": 3205
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.18700197339057922,
+      "learning_rate": 0.00016333358751659605,
+      "loss": 0.8718,
+      "step": 3206
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.19421708583831787,
+      "learning_rate": 0.00016331215721652166,
+      "loss": 0.9084,
+      "step": 3207
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.19596804678440094,
+      "learning_rate": 0.0001632907220624303,
+      "loss": 0.8881,
+      "step": 3208
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.20006729662418365,
+      "learning_rate": 0.00016326928205596527,
+      "loss": 0.9956,
+      "step": 3209
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.1877383142709732,
+      "learning_rate": 0.00016324783719877032,
+      "loss": 0.8683,
+      "step": 3210
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.20360927283763885,
+      "learning_rate": 0.00016322638749248969,
+      "loss": 0.9104,
+      "step": 3211
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.18599390983581543,
+      "learning_rate": 0.0001632049329387678,
+      "loss": 0.9254,
+      "step": 3212
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.20069356262683868,
+      "learning_rate": 0.00016318347353924957,
+      "loss": 0.9163,
+      "step": 3213
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.19795280694961548,
+      "learning_rate": 0.0001631620092955802,
+      "loss": 0.872,
+      "step": 3214
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.21090590953826904,
+      "learning_rate": 0.00016314054020940538,
+      "loss": 0.9532,
+      "step": 3215
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.19131018221378326,
+      "learning_rate": 0.00016311906628237104,
+      "loss": 0.8848,
+      "step": 3216
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.20162901282310486,
+      "learning_rate": 0.00016309758751612357,
+      "loss": 0.9615,
+      "step": 3217
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.20793616771697998,
+      "learning_rate": 0.00016307610391230973,
+      "loss": 0.9463,
+      "step": 3218
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.19533415138721466,
+      "learning_rate": 0.0001630546154725766,
+      "loss": 0.9327,
+      "step": 3219
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.200450137257576,
+      "learning_rate": 0.00016303312219857166,
+      "loss": 0.8954,
+      "step": 3220
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.1979542076587677,
+      "learning_rate": 0.00016301162409194272,
+      "loss": 0.9546,
+      "step": 3221
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.1875590831041336,
+      "learning_rate": 0.000162990121154338,
+      "loss": 0.9371,
+      "step": 3222
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.19330134987831116,
+      "learning_rate": 0.00016296861338740618,
+      "loss": 0.8646,
+      "step": 3223
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.18598540127277374,
+      "learning_rate": 0.00016294710079279613,
+      "loss": 0.9595,
+      "step": 3224
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.18598809838294983,
+      "learning_rate": 0.0001629255833721572,
+      "loss": 0.879,
+      "step": 3225
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.1924973428249359,
+      "learning_rate": 0.00016290406112713906,
+      "loss": 0.8916,
+      "step": 3226
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.1871059238910675,
+      "learning_rate": 0.00016288253405939183,
+      "loss": 0.9009,
+      "step": 3227
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.1845734864473343,
+      "learning_rate": 0.00016286100217056596,
+      "loss": 0.9263,
+      "step": 3228
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.19569912552833557,
+      "learning_rate": 0.00016283946546231218,
+      "loss": 0.9312,
+      "step": 3229
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.19267801940441132,
+      "learning_rate": 0.0001628179239362817,
+      "loss": 0.873,
+      "step": 3230
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.2015204131603241,
+      "learning_rate": 0.0001627963775941261,
+      "loss": 0.9102,
+      "step": 3231
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.19582143425941467,
+      "learning_rate": 0.00016277482643749726,
+      "loss": 0.9614,
+      "step": 3232
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.20045465230941772,
+      "learning_rate": 0.00016275327046804747,
+      "loss": 0.9631,
+      "step": 3233
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.18617044389247894,
+      "learning_rate": 0.0001627317096874294,
+      "loss": 0.8873,
+      "step": 3234
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.20299577713012695,
+      "learning_rate": 0.00016271014409729605,
+      "loss": 0.922,
+      "step": 3235
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.20413140952587128,
+      "learning_rate": 0.0001626885736993008,
+      "loss": 0.9555,
+      "step": 3236
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.18887461721897125,
+      "learning_rate": 0.00016266699849509743,
+      "loss": 0.9311,
+      "step": 3237
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.19803249835968018,
+      "learning_rate": 0.0001626454184863401,
+      "loss": 0.9104,
+      "step": 3238
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.19364400207996368,
+      "learning_rate": 0.00016262383367468325,
+      "loss": 0.9282,
+      "step": 3239
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.20137445628643036,
+      "learning_rate": 0.00016260224406178177,
+      "loss": 0.9215,
+      "step": 3240
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.19514375925064087,
+      "learning_rate": 0.00016258064964929088,
+      "loss": 0.9409,
+      "step": 3241
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.19378265738487244,
+      "learning_rate": 0.0001625590504388662,
+      "loss": 0.8933,
+      "step": 3242
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.18738335371017456,
+      "learning_rate": 0.00016253744643216368,
+      "loss": 0.9545,
+      "step": 3243
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.1987677812576294,
+      "learning_rate": 0.00016251583763083967,
+      "loss": 0.9697,
+      "step": 3244
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.20496296882629395,
+      "learning_rate": 0.00016249422403655087,
+      "loss": 0.96,
+      "step": 3245
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.20727890729904175,
+      "learning_rate": 0.00016247260565095437,
+      "loss": 1.0182,
+      "step": 3246
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.19182537496089935,
+      "learning_rate": 0.00016245098247570758,
+      "loss": 0.9521,
+      "step": 3247
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.2026839703321457,
+      "learning_rate": 0.0001624293545124683,
+      "loss": 0.9507,
+      "step": 3248
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.20389671623706818,
+      "learning_rate": 0.00016240772176289475,
+      "loss": 0.9635,
+      "step": 3249
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.1950456202030182,
+      "learning_rate": 0.00016238608422864542,
+      "loss": 0.8601,
+      "step": 3250
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.1940208077430725,
+      "learning_rate": 0.00016236444191137917,
+      "loss": 0.9209,
+      "step": 3251
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.19530853629112244,
+      "learning_rate": 0.00016234279481275543,
+      "loss": 0.9396,
+      "step": 3252
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.19468019902706146,
+      "learning_rate": 0.0001623211429344337,
+      "loss": 0.8943,
+      "step": 3253
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.1874203383922577,
+      "learning_rate": 0.00016229948627807404,
+      "loss": 0.9383,
+      "step": 3254
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.198331817984581,
+      "learning_rate": 0.00016227782484533681,
+      "loss": 0.9178,
+      "step": 3255
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.19345207512378693,
+      "learning_rate": 0.00016225615863788278,
+      "loss": 0.9393,
+      "step": 3256
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.1910363733768463,
+      "learning_rate": 0.000162234487657373,
+      "loss": 0.9635,
+      "step": 3257
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.18190106749534607,
+      "learning_rate": 0.000162212811905469,
+      "loss": 0.8974,
+      "step": 3258
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.19228288531303406,
+      "learning_rate": 0.00016219113138383258,
+      "loss": 0.9087,
+      "step": 3259
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.1946840137243271,
+      "learning_rate": 0.00016216944609412595,
+      "loss": 0.9161,
+      "step": 3260
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.1919838786125183,
+      "learning_rate": 0.00016214775603801167,
+      "loss": 0.9231,
+      "step": 3261
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.203338623046875,
+      "learning_rate": 0.00016212606121715267,
+      "loss": 0.9122,
+      "step": 3262
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.18521350622177124,
+      "learning_rate": 0.0001621043616332123,
+      "loss": 0.8616,
+      "step": 3263
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.19194412231445312,
+      "learning_rate": 0.00016208265728785416,
+      "loss": 0.9158,
+      "step": 3264
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.19357708096504211,
+      "learning_rate": 0.00016206094818274229,
+      "loss": 0.9746,
+      "step": 3265
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.1967921108007431,
+      "learning_rate": 0.00016203923431954112,
+      "loss": 0.9261,
+      "step": 3266
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.20249836146831512,
+      "learning_rate": 0.00016201751569991537,
+      "loss": 0.9597,
+      "step": 3267
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.19432705640792847,
+      "learning_rate": 0.0001619957923255302,
+      "loss": 0.9656,
+      "step": 3268
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.18914005160331726,
+      "learning_rate": 0.00016197406419805107,
+      "loss": 0.8933,
+      "step": 3269
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.20102287828922272,
+      "learning_rate": 0.00016195233131914382,
+      "loss": 0.9124,
+      "step": 3270
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.18623220920562744,
+      "learning_rate": 0.0001619305936904747,
+      "loss": 0.8506,
+      "step": 3271
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.19582891464233398,
+      "learning_rate": 0.00016190885131371028,
+      "loss": 0.9541,
+      "step": 3272
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.19561882317066193,
+      "learning_rate": 0.00016188710419051748,
+      "loss": 0.9532,
+      "step": 3273
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.19685816764831543,
+      "learning_rate": 0.00016186535232256364,
+      "loss": 0.959,
+      "step": 3274
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.20488882064819336,
+      "learning_rate": 0.00016184359571151644,
+      "loss": 0.953,
+      "step": 3275
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.1881592869758606,
+      "learning_rate": 0.00016182183435904389,
+      "loss": 0.9008,
+      "step": 3276
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.19552184641361237,
+      "learning_rate": 0.00016180006826681438,
+      "loss": 0.9552,
+      "step": 3277
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.1943063884973526,
+      "learning_rate": 0.00016177829743649672,
+      "loss": 0.8943,
+      "step": 3278
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.18469476699829102,
+      "learning_rate": 0.00016175652186975996,
+      "loss": 0.8334,
+      "step": 3279
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.1960335671901703,
+      "learning_rate": 0.0001617347415682737,
+      "loss": 0.9298,
+      "step": 3280
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.18966439366340637,
+      "learning_rate": 0.00016171295653370768,
+      "loss": 0.8931,
+      "step": 3281
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.18533951044082642,
+      "learning_rate": 0.00016169116676773219,
+      "loss": 0.873,
+      "step": 3282
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.19226530194282532,
+      "learning_rate": 0.00016166937227201776,
+      "loss": 0.9701,
+      "step": 3283
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.20334738492965698,
+      "learning_rate": 0.00016164757304823536,
+      "loss": 0.9888,
+      "step": 3284
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.19117876887321472,
+      "learning_rate": 0.0001616257690980563,
+      "loss": 0.8898,
+      "step": 3285
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.19045065343379974,
+      "learning_rate": 0.00016160396042315224,
+      "loss": 0.9351,
+      "step": 3286
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.19016839563846588,
+      "learning_rate": 0.00016158214702519517,
+      "loss": 0.9358,
+      "step": 3287
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.1997360736131668,
+      "learning_rate": 0.0001615603289058575,
+      "loss": 0.9525,
+      "step": 3288
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.1873968541622162,
+      "learning_rate": 0.00016153850606681202,
+      "loss": 0.8936,
+      "step": 3289
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.201723113656044,
+      "learning_rate": 0.00016151667850973183,
+      "loss": 0.9871,
+      "step": 3290
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.19081851840019226,
+      "learning_rate": 0.00016149484623629036,
+      "loss": 0.8942,
+      "step": 3291
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.18933135271072388,
+      "learning_rate": 0.00016147300924816146,
+      "loss": 0.9196,
+      "step": 3292
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.20124155282974243,
+      "learning_rate": 0.0001614511675470194,
+      "loss": 0.9442,
+      "step": 3293
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.18622492253780365,
+      "learning_rate": 0.0001614293211345386,
+      "loss": 0.9153,
+      "step": 3294
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.1968419998884201,
+      "learning_rate": 0.00016140747001239415,
+      "loss": 0.9315,
+      "step": 3295
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.19139516353607178,
+      "learning_rate": 0.0001613856141822612,
+      "loss": 0.8875,
+      "step": 3296
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.20368632674217224,
+      "learning_rate": 0.00016136375364581547,
+      "loss": 1.0369,
+      "step": 3297
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.1934148222208023,
+      "learning_rate": 0.0001613418884047329,
+      "loss": 0.8596,
+      "step": 3298
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.1904744654893875,
+      "learning_rate": 0.00016132001846068988,
+      "loss": 0.9054,
+      "step": 3299
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.18952694535255432,
+      "learning_rate": 0.00016129814381536322,
+      "loss": 0.9598,
+      "step": 3300
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.18965736031532288,
+      "learning_rate": 0.00016127626447042984,
+      "loss": 0.9516,
+      "step": 3301
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.19727297127246857,
+      "learning_rate": 0.00016125438042756732,
+      "loss": 0.8983,
+      "step": 3302
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.2037915289402008,
+      "learning_rate": 0.0001612324916884534,
+      "loss": 0.9696,
+      "step": 3303
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.18593308329582214,
+      "learning_rate": 0.0001612105982547663,
+      "loss": 0.883,
+      "step": 3304
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.1988077610731125,
+      "learning_rate": 0.00016118870012818447,
+      "loss": 0.9199,
+      "step": 3305
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.20564799010753632,
+      "learning_rate": 0.0001611667973103869,
+      "loss": 0.9839,
+      "step": 3306
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.19398616254329681,
+      "learning_rate": 0.00016114488980305272,
+      "loss": 0.9529,
+      "step": 3307
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.19221334159374237,
+      "learning_rate": 0.00016112297760786159,
+      "loss": 0.9208,
+      "step": 3308
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.19437314569950104,
+      "learning_rate": 0.0001611010607264935,
+      "loss": 0.9303,
+      "step": 3309
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.1968432515859604,
+      "learning_rate": 0.00016107913916062875,
+      "loss": 0.9664,
+      "step": 3310
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.19320005178451538,
+      "learning_rate": 0.000161057212911948,
+      "loss": 0.9252,
+      "step": 3311
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.20163682103157043,
+      "learning_rate": 0.00016103528198213232,
+      "loss": 0.9298,
+      "step": 3312
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.20061266422271729,
+      "learning_rate": 0.0001610133463728631,
+      "loss": 0.9236,
+      "step": 3313
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.20181746780872345,
+      "learning_rate": 0.0001609914060858221,
+      "loss": 0.9528,
+      "step": 3314
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.19283704459667206,
+      "learning_rate": 0.00016096946112269145,
+      "loss": 0.9264,
+      "step": 3315
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.1948464959859848,
+      "learning_rate": 0.0001609475114851536,
+      "loss": 0.9575,
+      "step": 3316
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.2059197872877121,
+      "learning_rate": 0.00016092555717489142,
+      "loss": 0.9476,
+      "step": 3317
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.19160623848438263,
+      "learning_rate": 0.00016090359819358806,
+      "loss": 0.9308,
+      "step": 3318
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.19544410705566406,
+      "learning_rate": 0.0001608816345429271,
+      "loss": 0.9077,
+      "step": 3319
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.19720809161663055,
+      "learning_rate": 0.0001608596662245925,
+      "loss": 0.9383,
+      "step": 3320
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.2031698375940323,
+      "learning_rate": 0.0001608376932402684,
+      "loss": 0.8995,
+      "step": 3321
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.21251250803470612,
+      "learning_rate": 0.00016081571559163952,
+      "loss": 0.9421,
+      "step": 3322
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.1931510716676712,
+      "learning_rate": 0.0001607937332803908,
+      "loss": 0.9147,
+      "step": 3323
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.19057153165340424,
+      "learning_rate": 0.00016077174630820765,
+      "loss": 0.8873,
+      "step": 3324
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.19663560390472412,
+      "learning_rate": 0.0001607497546767757,
+      "loss": 0.8729,
+      "step": 3325
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.197501540184021,
+      "learning_rate": 0.00016072775838778104,
+      "loss": 0.9721,
+      "step": 3326
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.201323002576828,
+      "learning_rate": 0.00016070575744291004,
+      "loss": 0.9891,
+      "step": 3327
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.19409815967082977,
+      "learning_rate": 0.0001606837518438495,
+      "loss": 0.9291,
+      "step": 3328
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.18671323359012604,
+      "learning_rate": 0.00016066174159228653,
+      "loss": 0.8786,
+      "step": 3329
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.20137618482112885,
+      "learning_rate": 0.00016063972668990862,
+      "loss": 0.8937,
+      "step": 3330
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.2120594084262848,
+      "learning_rate": 0.00016061770713840363,
+      "loss": 0.9768,
+      "step": 3331
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.20866088569164276,
+      "learning_rate": 0.0001605956829394597,
+      "loss": 0.911,
+      "step": 3332
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.1928453892469406,
+      "learning_rate": 0.00016057365409476545,
+      "loss": 0.9378,
+      "step": 3333
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.21047063171863556,
+      "learning_rate": 0.0001605516206060097,
+      "loss": 0.937,
+      "step": 3334
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.20211422443389893,
+      "learning_rate": 0.00016052958247488182,
+      "loss": 0.9226,
+      "step": 3335
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.21109655499458313,
+      "learning_rate": 0.0001605075397030714,
+      "loss": 0.9584,
+      "step": 3336
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.19364911317825317,
+      "learning_rate": 0.00016048549229226832,
+      "loss": 0.9552,
+      "step": 3337
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.2017887681722641,
+      "learning_rate": 0.00016046344024416302,
+      "loss": 0.9571,
+      "step": 3338
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.19468630850315094,
+      "learning_rate": 0.00016044138356044614,
+      "loss": 0.9033,
+      "step": 3339
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.1936691552400589,
+      "learning_rate": 0.00016041932224280877,
+      "loss": 0.9071,
+      "step": 3340
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.20116499066352844,
+      "learning_rate": 0.00016039725629294225,
+      "loss": 0.9469,
+      "step": 3341
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.19864040613174438,
+      "learning_rate": 0.00016037518571253832,
+      "loss": 0.9157,
+      "step": 3342
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.19681987166404724,
+      "learning_rate": 0.00016035311050328915,
+      "loss": 0.8943,
+      "step": 3343
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.19773036241531372,
+      "learning_rate": 0.00016033103066688715,
+      "loss": 0.9386,
+      "step": 3344
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.20723670721054077,
+      "learning_rate": 0.00016030894620502518,
+      "loss": 0.8724,
+      "step": 3345
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.1986563503742218,
+      "learning_rate": 0.00016028685711939636,
+      "loss": 0.918,
+      "step": 3346
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.20282043516635895,
+      "learning_rate": 0.00016026476341169426,
+      "loss": 0.9582,
+      "step": 3347
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.19087593257427216,
+      "learning_rate": 0.00016024266508361275,
+      "loss": 0.9584,
+      "step": 3348
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.18812496960163116,
+      "learning_rate": 0.00016022056213684609,
+      "loss": 0.9218,
+      "step": 3349
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.19818438589572906,
+      "learning_rate": 0.0001601984545730888,
+      "loss": 0.9222,
+      "step": 3350
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.19090117514133453,
+      "learning_rate": 0.00016017634239403589,
+      "loss": 0.9023,
+      "step": 3351
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.19678524136543274,
+      "learning_rate": 0.00016015422560138263,
+      "loss": 0.9563,
+      "step": 3352
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.18356016278266907,
+      "learning_rate": 0.00016013210419682465,
+      "loss": 0.8765,
+      "step": 3353
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.21214888989925385,
+      "learning_rate": 0.000160109978182058,
+      "loss": 0.9886,
+      "step": 3354
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.19988428056240082,
+      "learning_rate": 0.000160087847558779,
+      "loss": 0.9294,
+      "step": 3355
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.19415299594402313,
+      "learning_rate": 0.00016006571232868438,
+      "loss": 0.9029,
+      "step": 3356
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.1970900595188141,
+      "learning_rate": 0.00016004357249347118,
+      "loss": 0.9029,
+      "step": 3357
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.20032072067260742,
+      "learning_rate": 0.00016002142805483685,
+      "loss": 0.9369,
+      "step": 3358
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.19752494990825653,
+      "learning_rate": 0.00015999927901447913,
+      "loss": 0.934,
+      "step": 3359
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.1939263790845871,
+      "learning_rate": 0.00015997712537409615,
+      "loss": 0.9206,
+      "step": 3360
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.19927872717380524,
+      "learning_rate": 0.0001599549671353864,
+      "loss": 0.9109,
+      "step": 3361
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.19712553918361664,
+      "learning_rate": 0.0001599328043000487,
+      "loss": 0.955,
+      "step": 3362
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.18878357112407684,
+      "learning_rate": 0.00015991063686978226,
+      "loss": 0.9895,
+      "step": 3363
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.19527225196361542,
+      "learning_rate": 0.00015988846484628656,
+      "loss": 0.9226,
+      "step": 3364
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.1946769803762436,
+      "learning_rate": 0.0001598662882312615,
+      "loss": 0.9261,
+      "step": 3365
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.20022673904895782,
+      "learning_rate": 0.00015984410702640735,
+      "loss": 0.9437,
+      "step": 3366
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.19915050268173218,
+      "learning_rate": 0.00015982192123342468,
+      "loss": 0.9444,
+      "step": 3367
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.19298063218593597,
+      "learning_rate": 0.0001597997308540144,
+      "loss": 0.927,
+      "step": 3368
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.19722694158554077,
+      "learning_rate": 0.00015977753588987783,
+      "loss": 0.9354,
+      "step": 3369
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.1972188502550125,
+      "learning_rate": 0.00015975533634271662,
+      "loss": 0.972,
+      "step": 3370
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.19931070506572723,
+      "learning_rate": 0.00015973313221423274,
+      "loss": 0.9313,
+      "step": 3371
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.20220732688903809,
+      "learning_rate": 0.00015971092350612859,
+      "loss": 0.9237,
+      "step": 3372
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.1999865621328354,
+      "learning_rate": 0.00015968871022010677,
+      "loss": 0.988,
+      "step": 3373
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.19390293955802917,
+      "learning_rate": 0.00015966649235787044,
+      "loss": 0.9628,
+      "step": 3374
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.19825689494609833,
+      "learning_rate": 0.0001596442699211229,
+      "loss": 0.9282,
+      "step": 3375
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.19494299590587616,
+      "learning_rate": 0.000159622042911568,
+      "loss": 0.886,
+      "step": 3376
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.1895485520362854,
+      "learning_rate": 0.00015959981133090978,
+      "loss": 0.8627,
+      "step": 3377
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.1998288780450821,
+      "learning_rate": 0.00015957757518085267,
+      "loss": 0.9129,
+      "step": 3378
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.1855890303850174,
+      "learning_rate": 0.0001595553344631015,
+      "loss": 0.893,
+      "step": 3379
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.1998797506093979,
+      "learning_rate": 0.00015953308917936144,
+      "loss": 0.9685,
+      "step": 3380
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.20662716031074524,
+      "learning_rate": 0.000159510839331338,
+      "loss": 0.9312,
+      "step": 3381
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.20258738100528717,
+      "learning_rate": 0.00015948858492073696,
+      "loss": 0.9709,
+      "step": 3382
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.18903274834156036,
+      "learning_rate": 0.00015946632594926458,
+      "loss": 0.8984,
+      "step": 3383
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.2018447369337082,
+      "learning_rate": 0.00015944406241862738,
+      "loss": 0.9672,
+      "step": 3384
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.19322530925273895,
+      "learning_rate": 0.00015942179433053232,
+      "loss": 0.9422,
+      "step": 3385
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.19214372336864471,
+      "learning_rate": 0.00015939952168668656,
+      "loss": 0.9142,
+      "step": 3386
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.19462235271930695,
+      "learning_rate": 0.0001593772444887978,
+      "loss": 0.9098,
+      "step": 3387
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.19193169474601746,
+      "learning_rate": 0.0001593549627385739,
+      "loss": 0.9119,
+      "step": 3388
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.1959066390991211,
+      "learning_rate": 0.0001593326764377232,
+      "loss": 0.9462,
+      "step": 3389
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.19710206985473633,
+      "learning_rate": 0.00015931038558795435,
+      "loss": 0.9513,
+      "step": 3390
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.19761282205581665,
+      "learning_rate": 0.0001592880901909763,
+      "loss": 0.9108,
+      "step": 3391
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.19094908237457275,
+      "learning_rate": 0.0001592657902484985,
+      "loss": 0.9442,
+      "step": 3392
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.19495327770709991,
+      "learning_rate": 0.00015924348576223048,
+      "loss": 0.9221,
+      "step": 3393
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.19475141167640686,
+      "learning_rate": 0.00015922117673388244,
+      "loss": 0.9524,
+      "step": 3394
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.2048145830631256,
+      "learning_rate": 0.00015919886316516469,
+      "loss": 1.0231,
+      "step": 3395
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.19695360958576202,
+      "learning_rate": 0.0001591765450577879,
+      "loss": 0.9444,
+      "step": 3396
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.18348188698291779,
+      "learning_rate": 0.00015915422241346333,
+      "loss": 0.8791,
+      "step": 3397
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.18603946268558502,
+      "learning_rate": 0.00015913189523390227,
+      "loss": 0.892,
+      "step": 3398
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.19879482686519623,
+      "learning_rate": 0.00015910956352081657,
+      "loss": 0.9514,
+      "step": 3399
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.19644108414649963,
+      "learning_rate": 0.0001590872272759183,
+      "loss": 0.9085,
+      "step": 3400
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.1905001401901245,
+      "learning_rate": 0.00015906488650091997,
+      "loss": 0.9118,
+      "step": 3401
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.193280428647995,
+      "learning_rate": 0.0001590425411975344,
+      "loss": 0.9526,
+      "step": 3402
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.19824816286563873,
+      "learning_rate": 0.0001590201913674748,
+      "loss": 0.9328,
+      "step": 3403
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.267400860786438,
+      "learning_rate": 0.00015899783701245462,
+      "loss": 0.8283,
+      "step": 3404
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.19649048149585724,
+      "learning_rate": 0.00015897547813418776,
+      "loss": 0.9114,
+      "step": 3405
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.19132551550865173,
+      "learning_rate": 0.0001589531147343884,
+      "loss": 0.8749,
+      "step": 3406
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.19222450256347656,
+      "learning_rate": 0.0001589307468147711,
+      "loss": 0.9477,
+      "step": 3407
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.19050081074237823,
+      "learning_rate": 0.00015890837437705085,
+      "loss": 0.8783,
+      "step": 3408
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.18469731509685516,
+      "learning_rate": 0.00015888599742294282,
+      "loss": 0.9796,
+      "step": 3409
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.19211378693580627,
+      "learning_rate": 0.0001588636159541626,
+      "loss": 0.924,
+      "step": 3410
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.19958160817623138,
+      "learning_rate": 0.00015884122997242615,
+      "loss": 0.9866,
+      "step": 3411
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.19366028904914856,
+      "learning_rate": 0.00015881883947944974,
+      "loss": 0.8821,
+      "step": 3412
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.19694273173809052,
+      "learning_rate": 0.00015879644447695007,
+      "loss": 0.8971,
+      "step": 3413
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.18563781678676605,
+      "learning_rate": 0.00015877404496664402,
+      "loss": 0.9001,
+      "step": 3414
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.20295719802379608,
+      "learning_rate": 0.000158751640950249,
+      "loss": 0.8894,
+      "step": 3415
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.20849061012268066,
+      "learning_rate": 0.00015872923242948267,
+      "loss": 0.9321,
+      "step": 3416
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.19227440655231476,
+      "learning_rate": 0.00015870681940606302,
+      "loss": 0.8957,
+      "step": 3417
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.1898086965084076,
+      "learning_rate": 0.00015868440188170847,
+      "loss": 0.9095,
+      "step": 3418
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.19240839779376984,
+      "learning_rate": 0.0001586619798581376,
+      "loss": 0.9037,
+      "step": 3419
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.19419671595096588,
+      "learning_rate": 0.00015863955333706957,
+      "loss": 0.8636,
+      "step": 3420
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.20581892132759094,
+      "learning_rate": 0.00015861712232022375,
+      "loss": 0.9749,
+      "step": 3421
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.19966059923171997,
+      "learning_rate": 0.00015859468680931989,
+      "loss": 0.9905,
+      "step": 3422
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.18994350731372833,
+      "learning_rate": 0.00015857224680607804,
+      "loss": 0.9269,
+      "step": 3423
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.19134818017482758,
+      "learning_rate": 0.00015854980231221869,
+      "loss": 0.8993,
+      "step": 3424
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.1913415640592575,
+      "learning_rate": 0.00015852735332946253,
+      "loss": 0.9024,
+      "step": 3425
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.18681485950946808,
+      "learning_rate": 0.00015850489985953076,
+      "loss": 0.9176,
+      "step": 3426
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.2123211771249771,
+      "learning_rate": 0.0001584824419041448,
+      "loss": 0.9553,
+      "step": 3427
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.19513416290283203,
+      "learning_rate": 0.0001584599794650265,
+      "loss": 0.9052,
+      "step": 3428
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.20175153017044067,
+      "learning_rate": 0.00015843751254389796,
+      "loss": 0.9644,
+      "step": 3429
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.1934724897146225,
+      "learning_rate": 0.00015841504114248167,
+      "loss": 0.9069,
+      "step": 3430
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.19547207653522491,
+      "learning_rate": 0.0001583925652625005,
+      "loss": 0.8992,
+      "step": 3431
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.19625899195671082,
+      "learning_rate": 0.00015837008490567764,
+      "loss": 0.8821,
+      "step": 3432
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.19435423612594604,
+      "learning_rate": 0.00015834760007373658,
+      "loss": 0.9296,
+      "step": 3433
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.1921653300523758,
+      "learning_rate": 0.00015832511076840122,
+      "loss": 0.9497,
+      "step": 3434
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.20929893851280212,
+      "learning_rate": 0.00015830261699139575,
+      "loss": 0.9234,
+      "step": 3435
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.2004985213279724,
+      "learning_rate": 0.0001582801187444447,
+      "loss": 0.9163,
+      "step": 3436
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.20779582858085632,
+      "learning_rate": 0.000158257616029273,
+      "loss": 0.8724,
+      "step": 3437
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.2043759822845459,
+      "learning_rate": 0.0001582351088476059,
+      "loss": 0.8911,
+      "step": 3438
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.19317206740379333,
+      "learning_rate": 0.00015821259720116897,
+      "loss": 0.8972,
+      "step": 3439
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.20157548785209656,
+      "learning_rate": 0.00015819008109168808,
+      "loss": 0.9373,
+      "step": 3440
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.19609443843364716,
+      "learning_rate": 0.00015816756052088957,
+      "loss": 0.9662,
+      "step": 3441
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.19324661791324615,
+      "learning_rate": 0.00015814503549050002,
+      "loss": 0.8593,
+      "step": 3442
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.19976924359798431,
+      "learning_rate": 0.00015812250600224636,
+      "loss": 0.9051,
+      "step": 3443
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.19995374977588654,
+      "learning_rate": 0.0001580999720578559,
+      "loss": 0.9245,
+      "step": 3444
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.1970292627811432,
+      "learning_rate": 0.00015807743365905627,
+      "loss": 0.8853,
+      "step": 3445
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.19631227850914001,
+      "learning_rate": 0.00015805489080757544,
+      "loss": 0.8863,
+      "step": 3446
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.20259711146354675,
+      "learning_rate": 0.00015803234350514176,
+      "loss": 0.9394,
+      "step": 3447
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.21781596541404724,
+      "learning_rate": 0.00015800979175348382,
+      "loss": 0.9558,
+      "step": 3448
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.20329150557518005,
+      "learning_rate": 0.00015798723555433068,
+      "loss": 0.9256,
+      "step": 3449
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.18962758779525757,
+      "learning_rate": 0.00015796467490941164,
+      "loss": 0.856,
+      "step": 3450
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.21290762722492218,
+      "learning_rate": 0.00015794210982045636,
+      "loss": 0.9407,
+      "step": 3451
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.19620513916015625,
+      "learning_rate": 0.00015791954028919497,
+      "loss": 0.9601,
+      "step": 3452
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.20747888088226318,
+      "learning_rate": 0.00015789696631735768,
+      "loss": 0.9534,
+      "step": 3453
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.19664233922958374,
+      "learning_rate": 0.0001578743879066753,
+      "loss": 0.9279,
+      "step": 3454
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.19570444524288177,
+      "learning_rate": 0.00015785180505887883,
+      "loss": 0.9025,
+      "step": 3455
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.18608535826206207,
+      "learning_rate": 0.00015782921777569966,
+      "loss": 0.8819,
+      "step": 3456
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.20291505753993988,
+      "learning_rate": 0.00015780662605886951,
+      "loss": 0.9383,
+      "step": 3457
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.2000468522310257,
+      "learning_rate": 0.00015778402991012045,
+      "loss": 0.9386,
+      "step": 3458
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.19044055044651031,
+      "learning_rate": 0.00015776142933118488,
+      "loss": 0.9092,
+      "step": 3459
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.2033713459968567,
+      "learning_rate": 0.00015773882432379552,
+      "loss": 0.9389,
+      "step": 3460
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.19457867741584778,
+      "learning_rate": 0.0001577162148896855,
+      "loss": 0.8789,
+      "step": 3461
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.20580138266086578,
+      "learning_rate": 0.00015769360103058816,
+      "loss": 0.904,
+      "step": 3462
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.19230179488658905,
+      "learning_rate": 0.00015767098274823733,
+      "loss": 0.9108,
+      "step": 3463
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.19959823787212372,
+      "learning_rate": 0.0001576483600443671,
+      "loss": 0.9708,
+      "step": 3464
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.19323502480983734,
+      "learning_rate": 0.00015762573292071186,
+      "loss": 0.9289,
+      "step": 3465
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.2083415389060974,
+      "learning_rate": 0.00015760310137900643,
+      "loss": 1.0455,
+      "step": 3466
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.198960542678833,
+      "learning_rate": 0.00015758046542098595,
+      "loss": 0.888,
+      "step": 3467
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.1912854015827179,
+      "learning_rate": 0.0001575578250483858,
+      "loss": 0.9166,
+      "step": 3468
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.2046162486076355,
+      "learning_rate": 0.00015753518026294185,
+      "loss": 0.9227,
+      "step": 3469
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.19570709764957428,
+      "learning_rate": 0.00015751253106639014,
+      "loss": 0.8762,
+      "step": 3470
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.18846337497234344,
+      "learning_rate": 0.0001574898774604672,
+      "loss": 0.9041,
+      "step": 3471
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.18992602825164795,
+      "learning_rate": 0.00015746721944690986,
+      "loss": 0.9162,
+      "step": 3472
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.19614478945732117,
+      "learning_rate": 0.00015744455702745521,
+      "loss": 0.8904,
+      "step": 3473
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.19449754059314728,
+      "learning_rate": 0.00015742189020384077,
+      "loss": 0.9354,
+      "step": 3474
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.20135246217250824,
+      "learning_rate": 0.00015739921897780436,
+      "loss": 0.9433,
+      "step": 3475
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.19418463110923767,
+      "learning_rate": 0.0001573765433510841,
+      "loss": 0.8763,
+      "step": 3476
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.19569334387779236,
+      "learning_rate": 0.00015735386332541847,
+      "loss": 0.8775,
+      "step": 3477
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.19247597455978394,
+      "learning_rate": 0.0001573311789025464,
+      "loss": 0.8953,
+      "step": 3478
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.21302643418312073,
+      "learning_rate": 0.00015730849008420701,
+      "loss": 0.9473,
+      "step": 3479
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.19832172989845276,
+      "learning_rate": 0.00015728579687213974,
+      "loss": 0.9332,
+      "step": 3480
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.19264574348926544,
+      "learning_rate": 0.00015726309926808452,
+      "loss": 0.9148,
+      "step": 3481
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.1975184977054596,
+      "learning_rate": 0.00015724039727378148,
+      "loss": 0.8925,
+      "step": 3482
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.19813388586044312,
+      "learning_rate": 0.0001572176908909712,
+      "loss": 0.9127,
+      "step": 3483
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.20585894584655762,
+      "learning_rate": 0.00015719498012139446,
+      "loss": 0.9244,
+      "step": 3484
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.18802528083324432,
+      "learning_rate": 0.00015717226496679248,
+      "loss": 0.8992,
+      "step": 3485
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.18946048617362976,
+      "learning_rate": 0.00015714954542890677,
+      "loss": 0.9641,
+      "step": 3486
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.19644182920455933,
+      "learning_rate": 0.00015712682150947923,
+      "loss": 0.9015,
+      "step": 3487
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.19160401821136475,
+      "learning_rate": 0.00015710409321025202,
+      "loss": 0.9143,
+      "step": 3488
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.18982695043087006,
+      "learning_rate": 0.00015708136053296768,
+      "loss": 0.8535,
+      "step": 3489
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.19489288330078125,
+      "learning_rate": 0.00015705862347936914,
+      "loss": 0.9514,
+      "step": 3490
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.19520345330238342,
+      "learning_rate": 0.0001570358820511995,
+      "loss": 0.9161,
+      "step": 3491
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.19223810732364655,
+      "learning_rate": 0.0001570131362502024,
+      "loss": 0.8904,
+      "step": 3492
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.1916969120502472,
+      "learning_rate": 0.00015699038607812162,
+      "loss": 0.8793,
+      "step": 3493
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.19017493724822998,
+      "learning_rate": 0.00015696763153670142,
+      "loss": 0.8336,
+      "step": 3494
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.18902774155139923,
+      "learning_rate": 0.00015694487262768634,
+      "loss": 0.8857,
+      "step": 3495
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.19431182742118835,
+      "learning_rate": 0.0001569221093528213,
+      "loss": 0.87,
+      "step": 3496
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.18795667588710785,
+      "learning_rate": 0.00015689934171385147,
+      "loss": 0.882,
+      "step": 3497
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.1952623426914215,
+      "learning_rate": 0.00015687656971252238,
+      "loss": 0.928,
+      "step": 3498
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.18606235086917877,
+      "learning_rate": 0.00015685379335057996,
+      "loss": 0.8403,
+      "step": 3499
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.18356823921203613,
+      "learning_rate": 0.0001568310126297704,
+      "loss": 0.8661,
+      "step": 3500
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.2000182569026947,
+      "learning_rate": 0.00015680822755184028,
+      "loss": 0.9734,
+      "step": 3501
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.19715382158756256,
+      "learning_rate": 0.00015678543811853643,
+      "loss": 0.8895,
+      "step": 3502
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.1973632425069809,
+      "learning_rate": 0.00015676264433160615,
+      "loss": 0.9437,
+      "step": 3503
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.20098167657852173,
+      "learning_rate": 0.0001567398461927969,
+      "loss": 0.8708,
+      "step": 3504
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.2057405710220337,
+      "learning_rate": 0.0001567170437038567,
+      "loss": 0.9534,
+      "step": 3505
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.189925417304039,
+      "learning_rate": 0.00015669423686653366,
+      "loss": 0.8764,
+      "step": 3506
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.18974903225898743,
+      "learning_rate": 0.00015667142568257636,
+      "loss": 0.8952,
+      "step": 3507
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.19761645793914795,
+      "learning_rate": 0.0001566486101537337,
+      "loss": 0.9287,
+      "step": 3508
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.19054748117923737,
+      "learning_rate": 0.00015662579028175486,
+      "loss": 0.9112,
+      "step": 3509
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.19649845361709595,
+      "learning_rate": 0.0001566029660683895,
+      "loss": 0.9309,
+      "step": 3510
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.18728889524936676,
+      "learning_rate": 0.0001565801375153874,
+      "loss": 0.8955,
+      "step": 3511
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.18988218903541565,
+      "learning_rate": 0.00015655730462449882,
+      "loss": 0.9138,
+      "step": 3512
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.20179392397403717,
+      "learning_rate": 0.00015653446739747427,
+      "loss": 0.9328,
+      "step": 3513
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.19956812262535095,
+      "learning_rate": 0.00015651162583606474,
+      "loss": 0.9143,
+      "step": 3514
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.18776121735572815,
+      "learning_rate": 0.00015648877994202138,
+      "loss": 0.9249,
+      "step": 3515
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.20067352056503296,
+      "learning_rate": 0.0001564659297170957,
+      "loss": 0.943,
+      "step": 3516
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.1972014605998993,
+      "learning_rate": 0.00015644307516303964,
+      "loss": 0.9838,
+      "step": 3517
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.1812957376241684,
+      "learning_rate": 0.0001564202162816054,
+      "loss": 0.9204,
+      "step": 3518
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.20647543668746948,
+      "learning_rate": 0.0001563973530745455,
+      "loss": 0.9156,
+      "step": 3519
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.20621806383132935,
+      "learning_rate": 0.00015637448554361286,
+      "loss": 0.9443,
+      "step": 3520
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.19071152806282043,
+      "learning_rate": 0.00015635161369056066,
+      "loss": 0.9304,
+      "step": 3521
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.18997131288051605,
+      "learning_rate": 0.00015632873751714245,
+      "loss": 0.9167,
+      "step": 3522
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.1913936585187912,
+      "learning_rate": 0.0001563058570251121,
+      "loss": 0.8775,
+      "step": 3523
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.20011787116527557,
+      "learning_rate": 0.00015628297221622378,
+      "loss": 0.9236,
+      "step": 3524
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.20807957649230957,
+      "learning_rate": 0.00015626008309223207,
+      "loss": 0.9442,
+      "step": 3525
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.19250045716762543,
+      "learning_rate": 0.0001562371896548918,
+      "loss": 0.9152,
+      "step": 3526
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.19840964674949646,
+      "learning_rate": 0.0001562142919059582,
+      "loss": 0.924,
+      "step": 3527
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.20426321029663086,
+      "learning_rate": 0.00015619138984718673,
+      "loss": 0.9256,
+      "step": 3528
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.2006935477256775,
+      "learning_rate": 0.00015616848348033334,
+      "loss": 0.9052,
+      "step": 3529
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.20013563334941864,
+      "learning_rate": 0.00015614557280715412,
+      "loss": 0.9853,
+      "step": 3530
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.19713100790977478,
+      "learning_rate": 0.00015612265782940566,
+      "loss": 0.8925,
+      "step": 3531
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.19918465614318848,
+      "learning_rate": 0.00015609973854884473,
+      "loss": 0.9249,
+      "step": 3532
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.1967756599187851,
+      "learning_rate": 0.0001560768149672286,
+      "loss": 0.9007,
+      "step": 3533
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.18599602580070496,
+      "learning_rate": 0.0001560538870863147,
+      "loss": 0.8715,
+      "step": 3534
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.19670726358890533,
+      "learning_rate": 0.0001560309549078609,
+      "loss": 0.9482,
+      "step": 3535
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.1904277354478836,
+      "learning_rate": 0.00015600801843362535,
+      "loss": 0.9127,
+      "step": 3536
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.1943923681974411,
+      "learning_rate": 0.00015598507766536655,
+      "loss": 0.9423,
+      "step": 3537
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.20839069783687592,
+      "learning_rate": 0.0001559621326048433,
+      "loss": 0.9427,
+      "step": 3538
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.19504216313362122,
+      "learning_rate": 0.0001559391832538148,
+      "loss": 0.9384,
+      "step": 3539
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.19287019968032837,
+      "learning_rate": 0.0001559162296140405,
+      "loss": 0.9277,
+      "step": 3540
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.19911225140094757,
+      "learning_rate": 0.00015589327168728022,
+      "loss": 0.8505,
+      "step": 3541
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.19432951509952545,
+      "learning_rate": 0.00015587030947529412,
+      "loss": 0.9193,
+      "step": 3542
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.20329613983631134,
+      "learning_rate": 0.00015584734297984258,
+      "loss": 0.9654,
+      "step": 3543
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.19105875492095947,
+      "learning_rate": 0.00015582437220268647,
+      "loss": 0.9219,
+      "step": 3544
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.1942165344953537,
+      "learning_rate": 0.00015580139714558693,
+      "loss": 0.9053,
+      "step": 3545
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.1945033073425293,
+      "learning_rate": 0.00015577841781030537,
+      "loss": 0.9851,
+      "step": 3546
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.19224001467227936,
+      "learning_rate": 0.00015575543419860357,
+      "loss": 0.9162,
+      "step": 3547
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.1970493197441101,
+      "learning_rate": 0.00015573244631224365,
+      "loss": 0.9324,
+      "step": 3548
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.2003912329673767,
+      "learning_rate": 0.00015570945415298803,
+      "loss": 0.8751,
+      "step": 3549
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.21274451911449432,
+      "learning_rate": 0.0001556864577225995,
+      "loss": 1.0118,
+      "step": 3550
+    },
+    {
+      "epoch": 1.23,
+      "eval_loss": 0.9446007609367371,
+      "eval_runtime": 679.1052,
+      "eval_samples_per_second": 10.125,
+      "eval_steps_per_second": 5.063,
+      "step": 3550
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.18957918882369995,
+      "learning_rate": 0.0001556634570228412,
+      "loss": 0.9222,
+      "step": 3551
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.1931392103433609,
+      "learning_rate": 0.0001556404520554764,
+      "loss": 0.9261,
+      "step": 3552
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.19815470278263092,
+      "learning_rate": 0.00015561744282226896,
+      "loss": 0.9156,
+      "step": 3553
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.21058355271816254,
+      "learning_rate": 0.00015559442932498293,
+      "loss": 0.9234,
+      "step": 3554
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.20303329825401306,
+      "learning_rate": 0.00015557141156538268,
+      "loss": 0.9156,
+      "step": 3555
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.1917152851819992,
+      "learning_rate": 0.000155548389545233,
+      "loss": 0.88,
+      "step": 3556
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.19306671619415283,
+      "learning_rate": 0.00015552536326629888,
+      "loss": 0.8316,
+      "step": 3557
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.19891008734703064,
+      "learning_rate": 0.00015550233273034568,
+      "loss": 0.9051,
+      "step": 3558
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.21156953275203705,
+      "learning_rate": 0.0001554792979391392,
+      "loss": 0.9795,
+      "step": 3559
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.19550539553165436,
+      "learning_rate": 0.0001554562588944454,
+      "loss": 0.9283,
+      "step": 3560
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.19516271352767944,
+      "learning_rate": 0.00015543321559803065,
+      "loss": 0.9426,
+      "step": 3561
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.20502054691314697,
+      "learning_rate": 0.00015541016805166162,
+      "loss": 0.9175,
+      "step": 3562
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.18738922476768494,
+      "learning_rate": 0.00015538711625710536,
+      "loss": 0.8853,
+      "step": 3563
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.19391462206840515,
+      "learning_rate": 0.00015536406021612916,
+      "loss": 0.9396,
+      "step": 3564
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.19468341767787933,
+      "learning_rate": 0.00015534099993050072,
+      "loss": 0.9577,
+      "step": 3565
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.19213582575321198,
+      "learning_rate": 0.00015531793540198802,
+      "loss": 0.8845,
+      "step": 3566
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.19879239797592163,
+      "learning_rate": 0.00015529486663235935,
+      "loss": 0.9431,
+      "step": 3567
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.19197018444538116,
+      "learning_rate": 0.00015527179362338333,
+      "loss": 0.9384,
+      "step": 3568
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.19179058074951172,
+      "learning_rate": 0.00015524871637682898,
+      "loss": 0.8971,
+      "step": 3569
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.1983899474143982,
+      "learning_rate": 0.00015522563489446552,
+      "loss": 0.8992,
+      "step": 3570
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.18999382853507996,
+      "learning_rate": 0.00015520254917806265,
+      "loss": 0.9508,
+      "step": 3571
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.20845505595207214,
+      "learning_rate": 0.00015517945922939025,
+      "loss": 0.9412,
+      "step": 3572
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.1960909217596054,
+      "learning_rate": 0.00015515636505021853,
+      "loss": 0.9537,
+      "step": 3573
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.2100881189107895,
+      "learning_rate": 0.00015513326664231815,
+      "loss": 0.962,
+      "step": 3574
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.19650724530220032,
+      "learning_rate": 0.00015511016400746,
+      "loss": 0.9683,
+      "step": 3575
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.19711832702159882,
+      "learning_rate": 0.00015508705714741531,
+      "loss": 0.8819,
+      "step": 3576
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.1942337453365326,
+      "learning_rate": 0.00015506394606395568,
+      "loss": 0.9437,
+      "step": 3577
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.21626056730747223,
+      "learning_rate": 0.00015504083075885288,
+      "loss": 0.9474,
+      "step": 3578
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.20113731920719147,
+      "learning_rate": 0.00015501771123387922,
+      "loss": 0.9172,
+      "step": 3579
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.1971191167831421,
+      "learning_rate": 0.0001549945874908072,
+      "loss": 0.8949,
+      "step": 3580
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.20610153675079346,
+      "learning_rate": 0.00015497145953140962,
+      "loss": 0.9802,
+      "step": 3581
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.19442158937454224,
+      "learning_rate": 0.0001549483273574597,
+      "loss": 0.9386,
+      "step": 3582
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.19450068473815918,
+      "learning_rate": 0.00015492519097073097,
+      "loss": 1.0089,
+      "step": 3583
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.19892315566539764,
+      "learning_rate": 0.0001549020503729972,
+      "loss": 0.9811,
+      "step": 3584
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.19469180703163147,
+      "learning_rate": 0.00015487890556603254,
+      "loss": 0.9294,
+      "step": 3585
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.2058008313179016,
+      "learning_rate": 0.00015485575655161147,
+      "loss": 0.9309,
+      "step": 3586
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.19313830137252808,
+      "learning_rate": 0.0001548326033315088,
+      "loss": 0.9438,
+      "step": 3587
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.19705910980701447,
+      "learning_rate": 0.00015480944590749954,
+      "loss": 0.9313,
+      "step": 3588
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.19770359992980957,
+      "learning_rate": 0.00015478628428135924,
+      "loss": 0.927,
+      "step": 3589
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.20620097219944,
+      "learning_rate": 0.00015476311845486363,
+      "loss": 0.962,
+      "step": 3590
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.2014746069908142,
+      "learning_rate": 0.00015473994842978874,
+      "loss": 0.8598,
+      "step": 3591
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.18861185014247894,
+      "learning_rate": 0.00015471677420791102,
+      "loss": 0.9241,
+      "step": 3592
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.19452114403247833,
+      "learning_rate": 0.00015469359579100718,
+      "loss": 0.9356,
+      "step": 3593
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.19815759360790253,
+      "learning_rate": 0.00015467041318085423,
+      "loss": 0.9815,
+      "step": 3594
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.19805648922920227,
+      "learning_rate": 0.00015464722637922957,
+      "loss": 0.9215,
+      "step": 3595
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.1998823881149292,
+      "learning_rate": 0.00015462403538791088,
+      "loss": 0.9155,
+      "step": 3596
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.197417750954628,
+      "learning_rate": 0.00015460084020867617,
+      "loss": 0.8845,
+      "step": 3597
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.20399829745292664,
+      "learning_rate": 0.00015457764084330375,
+      "loss": 0.9541,
+      "step": 3598
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.20391467213630676,
+      "learning_rate": 0.00015455443729357228,
+      "loss": 0.8704,
+      "step": 3599
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.20385457575321198,
+      "learning_rate": 0.00015453122956126075,
+      "loss": 0.9182,
+      "step": 3600
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.19763916730880737,
+      "learning_rate": 0.00015450801764814838,
+      "loss": 0.9316,
+      "step": 3601
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.19183963537216187,
+      "learning_rate": 0.00015448480155601487,
+      "loss": 0.9035,
+      "step": 3602
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.20232851803302765,
+      "learning_rate": 0.00015446158128664013,
+      "loss": 0.9119,
+      "step": 3603
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.19677840173244476,
+      "learning_rate": 0.00015443835684180436,
+      "loss": 0.9383,
+      "step": 3604
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.1975477635860443,
+      "learning_rate": 0.0001544151282232882,
+      "loss": 0.8994,
+      "step": 3605
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.19218041002750397,
+      "learning_rate": 0.00015439189543287247,
+      "loss": 0.904,
+      "step": 3606
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.18120858073234558,
+      "learning_rate": 0.00015436865847233845,
+      "loss": 0.8842,
+      "step": 3607
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.19843339920043945,
+      "learning_rate": 0.00015434541734346766,
+      "loss": 0.9324,
+      "step": 3608
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.1892959624528885,
+      "learning_rate": 0.0001543221720480419,
+      "loss": 0.9368,
+      "step": 3609
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.1868380457162857,
+      "learning_rate": 0.00015429892258784337,
+      "loss": 0.9239,
+      "step": 3610
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.19214345514774323,
+      "learning_rate": 0.00015427566896465458,
+      "loss": 0.8936,
+      "step": 3611
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.20086589455604553,
+      "learning_rate": 0.00015425241118025834,
+      "loss": 0.9443,
+      "step": 3612
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.19924329221248627,
+      "learning_rate": 0.00015422914923643772,
+      "loss": 0.8946,
+      "step": 3613
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.18544204533100128,
+      "learning_rate": 0.00015420588313497623,
+      "loss": 0.8469,
+      "step": 3614
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.20435726642608643,
+      "learning_rate": 0.0001541826128776576,
+      "loss": 1.0018,
+      "step": 3615
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.1996217519044876,
+      "learning_rate": 0.00015415933846626593,
+      "loss": 0.9117,
+      "step": 3616
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.1999591886997223,
+      "learning_rate": 0.00015413605990258566,
+      "loss": 0.9222,
+      "step": 3617
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.19460801780223846,
+      "learning_rate": 0.00015411277718840145,
+      "loss": 0.9324,
+      "step": 3618
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.20450779795646667,
+      "learning_rate": 0.00015408949032549835,
+      "loss": 0.9441,
+      "step": 3619
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.20513708889484406,
+      "learning_rate": 0.00015406619931566172,
+      "loss": 0.9168,
+      "step": 3620
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.19295939803123474,
+      "learning_rate": 0.00015404290416067725,
+      "loss": 0.9253,
+      "step": 3621
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.1939520537853241,
+      "learning_rate": 0.00015401960486233093,
+      "loss": 0.886,
+      "step": 3622
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.20039387047290802,
+      "learning_rate": 0.0001539963014224091,
+      "loss": 0.9371,
+      "step": 3623
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.19236010313034058,
+      "learning_rate": 0.00015397299384269832,
+      "loss": 0.9098,
+      "step": 3624
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.20198063552379608,
+      "learning_rate": 0.00015394968212498555,
+      "loss": 0.9768,
+      "step": 3625
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.1999257355928421,
+      "learning_rate": 0.0001539263662710581,
+      "loss": 0.8571,
+      "step": 3626
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.1928747296333313,
+      "learning_rate": 0.00015390304628270355,
+      "loss": 0.9619,
+      "step": 3627
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.2093496322631836,
+      "learning_rate": 0.00015387972216170977,
+      "loss": 0.9477,
+      "step": 3628
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.2011002153158188,
+      "learning_rate": 0.00015385639390986494,
+      "loss": 0.907,
+      "step": 3629
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.19310890138149261,
+      "learning_rate": 0.00015383306152895766,
+      "loss": 0.9346,
+      "step": 3630
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.20581993460655212,
+      "learning_rate": 0.0001538097250207767,
+      "loss": 0.9707,
+      "step": 3631
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.1893724799156189,
+      "learning_rate": 0.00015378638438711133,
+      "loss": 0.9351,
+      "step": 3632
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.18906715512275696,
+      "learning_rate": 0.0001537630396297509,
+      "loss": 0.9621,
+      "step": 3633
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.1960383504629135,
+      "learning_rate": 0.00015373969075048533,
+      "loss": 0.934,
+      "step": 3634
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.197191521525383,
+      "learning_rate": 0.00015371633775110462,
+      "loss": 0.945,
+      "step": 3635
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.18677780032157898,
+      "learning_rate": 0.00015369298063339933,
+      "loss": 0.8289,
+      "step": 3636
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.20285886526107788,
+      "learning_rate": 0.00015366961939916008,
+      "loss": 0.9173,
+      "step": 3637
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.1963968425989151,
+      "learning_rate": 0.00015364625405017796,
+      "loss": 0.9221,
+      "step": 3638
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.19961443543434143,
+      "learning_rate": 0.00015362288458824438,
+      "loss": 0.9478,
+      "step": 3639
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.2008897364139557,
+      "learning_rate": 0.000153599511015151,
+      "loss": 0.882,
+      "step": 3640
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.20071753859519958,
+      "learning_rate": 0.00015357613333268985,
+      "loss": 0.9574,
+      "step": 3641
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.19547533988952637,
+      "learning_rate": 0.00015355275154265322,
+      "loss": 0.9415,
+      "step": 3642
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.20526045560836792,
+      "learning_rate": 0.0001535293656468338,
+      "loss": 0.9546,
+      "step": 3643
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.19212369620800018,
+      "learning_rate": 0.00015350597564702448,
+      "loss": 0.9417,
+      "step": 3644
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.20225495100021362,
+      "learning_rate": 0.00015348258154501853,
+      "loss": 0.9573,
+      "step": 3645
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.19001320004463196,
+      "learning_rate": 0.00015345918334260956,
+      "loss": 0.917,
+      "step": 3646
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.20381398499011993,
+      "learning_rate": 0.00015343578104159145,
+      "loss": 0.9687,
+      "step": 3647
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.19656528532505035,
+      "learning_rate": 0.00015341237464375844,
+      "loss": 0.9432,
+      "step": 3648
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.18959979712963104,
+      "learning_rate": 0.000153388964150905,
+      "loss": 0.9259,
+      "step": 3649
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.20691464841365814,
+      "learning_rate": 0.00015336554956482594,
+      "loss": 0.9746,
+      "step": 3650
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.1913762390613556,
+      "learning_rate": 0.0001533421308873165,
+      "loss": 0.9936,
+      "step": 3651
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.20307478308677673,
+      "learning_rate": 0.0001533187081201721,
+      "loss": 0.9708,
+      "step": 3652
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.19943755865097046,
+      "learning_rate": 0.00015329528126518848,
+      "loss": 0.8927,
+      "step": 3653
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.19460420310497284,
+      "learning_rate": 0.0001532718503241618,
+      "loss": 0.8652,
+      "step": 3654
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.20158496499061584,
+      "learning_rate": 0.0001532484152988884,
+      "loss": 0.9135,
+      "step": 3655
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.20068541169166565,
+      "learning_rate": 0.00015322497619116505,
+      "loss": 0.8876,
+      "step": 3656
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.3100888729095459,
+      "learning_rate": 0.00015320153300278877,
+      "loss": 0.8805,
+      "step": 3657
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.19288921356201172,
+      "learning_rate": 0.00015317808573555683,
+      "loss": 1.0011,
+      "step": 3658
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.19910819828510284,
+      "learning_rate": 0.000153154634391267,
+      "loss": 0.9088,
+      "step": 3659
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.1915523111820221,
+      "learning_rate": 0.00015313117897171717,
+      "loss": 0.9092,
+      "step": 3660
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.19844427704811096,
+      "learning_rate": 0.00015310771947870565,
+      "loss": 0.9244,
+      "step": 3661
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.19637028872966766,
+      "learning_rate": 0.00015308425591403104,
+      "loss": 0.9569,
+      "step": 3662
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.1928602159023285,
+      "learning_rate": 0.0001530607882794922,
+      "loss": 0.9344,
+      "step": 3663
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.18807289004325867,
+      "learning_rate": 0.0001530373165768884,
+      "loss": 0.922,
+      "step": 3664
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.21626749634742737,
+      "learning_rate": 0.0001530138408080191,
+      "loss": 1.0031,
+      "step": 3665
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.20268388092517853,
+      "learning_rate": 0.0001529903609746842,
+      "loss": 0.9478,
+      "step": 3666
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.20564964413642883,
+      "learning_rate": 0.0001529668770786839,
+      "loss": 0.8934,
+      "step": 3667
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.20551858842372894,
+      "learning_rate": 0.0001529433891218185,
+      "loss": 0.9278,
+      "step": 3668
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.1947879195213318,
+      "learning_rate": 0.00015291989710588898,
+      "loss": 0.7883,
+      "step": 3669
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.20598158240318298,
+      "learning_rate": 0.00015289640103269625,
+      "loss": 0.9217,
+      "step": 3670
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.1998508870601654,
+      "learning_rate": 0.0001528729009040418,
+      "loss": 0.9446,
+      "step": 3671
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.1991170197725296,
+      "learning_rate": 0.00015284939672172731,
+      "loss": 0.8942,
+      "step": 3672
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.19294770061969757,
+      "learning_rate": 0.0001528258884875548,
+      "loss": 0.9282,
+      "step": 3673
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.1996181309223175,
+      "learning_rate": 0.00015280237620332663,
+      "loss": 0.8971,
+      "step": 3674
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.19610020518302917,
+      "learning_rate": 0.0001527788598708454,
+      "loss": 0.9225,
+      "step": 3675
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.19014745950698853,
+      "learning_rate": 0.0001527553394919141,
+      "loss": 0.8461,
+      "step": 3676
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.19598208367824554,
+      "learning_rate": 0.00015273181506833593,
+      "loss": 0.9787,
+      "step": 3677
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.2032850831747055,
+      "learning_rate": 0.00015270828660191453,
+      "loss": 0.9879,
+      "step": 3678
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.1870907098054886,
+      "learning_rate": 0.00015268475409445373,
+      "loss": 0.8367,
+      "step": 3679
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.19532081484794617,
+      "learning_rate": 0.00015266121754775773,
+      "loss": 0.8911,
+      "step": 3680
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.194292813539505,
+      "learning_rate": 0.00015263767696363105,
+      "loss": 0.9161,
+      "step": 3681
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.19262300431728363,
+      "learning_rate": 0.0001526141323438785,
+      "loss": 0.9105,
+      "step": 3682
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.2018946409225464,
+      "learning_rate": 0.00015259058369030516,
+      "loss": 0.938,
+      "step": 3683
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.1982242912054062,
+      "learning_rate": 0.00015256703100471653,
+      "loss": 0.9816,
+      "step": 3684
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.19996307790279388,
+      "learning_rate": 0.00015254347428891825,
+      "loss": 0.9292,
+      "step": 3685
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.18849660456180573,
+      "learning_rate": 0.00015251991354471643,
+      "loss": 0.8891,
+      "step": 3686
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.1935540735721588,
+      "learning_rate": 0.00015249634877391742,
+      "loss": 0.9042,
+      "step": 3687
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.20464152097702026,
+      "learning_rate": 0.00015247277997832787,
+      "loss": 0.9383,
+      "step": 3688
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.19373281300067902,
+      "learning_rate": 0.00015244920715975482,
+      "loss": 0.8985,
+      "step": 3689
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.2008526623249054,
+      "learning_rate": 0.00015242563032000545,
+      "loss": 0.8887,
+      "step": 3690
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.21022352576255798,
+      "learning_rate": 0.00015240204946088737,
+      "loss": 0.9097,
+      "step": 3691
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.19161288440227509,
+      "learning_rate": 0.00015237846458420853,
+      "loss": 0.9256,
+      "step": 3692
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.2000376433134079,
+      "learning_rate": 0.00015235487569177707,
+      "loss": 0.8774,
+      "step": 3693
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.19469022750854492,
+      "learning_rate": 0.00015233128278540158,
+      "loss": 0.8713,
+      "step": 3694
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.19838149845600128,
+      "learning_rate": 0.00015230768586689086,
+      "loss": 0.9343,
+      "step": 3695
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.193618044257164,
+      "learning_rate": 0.00015228408493805397,
+      "loss": 0.8804,
+      "step": 3696
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.20271454751491547,
+      "learning_rate": 0.0001522604800007004,
+      "loss": 0.9614,
+      "step": 3697
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.1923941969871521,
+      "learning_rate": 0.0001522368710566399,
+      "loss": 0.9059,
+      "step": 3698
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.20286774635314941,
+      "learning_rate": 0.0001522132581076825,
+      "loss": 0.8956,
+      "step": 3699
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.19995754957199097,
+      "learning_rate": 0.0001521896411556386,
+      "loss": 0.9006,
+      "step": 3700
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.19672347605228424,
+      "learning_rate": 0.0001521660202023188,
+      "loss": 0.9199,
+      "step": 3701
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.19233226776123047,
+      "learning_rate": 0.00015214239524953408,
+      "loss": 0.8852,
+      "step": 3702
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.2116343080997467,
+      "learning_rate": 0.00015211876629909576,
+      "loss": 0.9474,
+      "step": 3703
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.18998068571090698,
+      "learning_rate": 0.00015209513335281543,
+      "loss": 0.8806,
+      "step": 3704
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.20243072509765625,
+      "learning_rate": 0.00015207149641250495,
+      "loss": 0.9505,
+      "step": 3705
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.19964581727981567,
+      "learning_rate": 0.00015204785547997648,
+      "loss": 0.9438,
+      "step": 3706
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.19266873598098755,
+      "learning_rate": 0.00015202421055704262,
+      "loss": 0.8944,
+      "step": 3707
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.20089896023273468,
+      "learning_rate": 0.00015200056164551607,
+      "loss": 0.9743,
+      "step": 3708
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.19461996853351593,
+      "learning_rate": 0.00015197690874721003,
+      "loss": 0.9625,
+      "step": 3709
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.22794750332832336,
+      "learning_rate": 0.00015195325186393794,
+      "loss": 1.0161,
+      "step": 3710
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.20262105762958527,
+      "learning_rate": 0.00015192959099751343,
+      "loss": 0.9498,
+      "step": 3711
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.19459131360054016,
+      "learning_rate": 0.0001519059261497506,
+      "loss": 0.8988,
+      "step": 3712
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.2000139057636261,
+      "learning_rate": 0.00015188225732246373,
+      "loss": 0.9255,
+      "step": 3713
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.19961777329444885,
+      "learning_rate": 0.00015185858451746752,
+      "loss": 0.9512,
+      "step": 3714
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.2007269263267517,
+      "learning_rate": 0.0001518349077365769,
+      "loss": 1.0035,
+      "step": 3715
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.18371418118476868,
+      "learning_rate": 0.00015181122698160714,
+      "loss": 0.9236,
+      "step": 3716
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.199965700507164,
+      "learning_rate": 0.00015178754225437372,
+      "loss": 0.911,
+      "step": 3717
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.19855192303657532,
+      "learning_rate": 0.00015176385355669255,
+      "loss": 0.9889,
+      "step": 3718
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.1886325627565384,
+      "learning_rate": 0.00015174016089037983,
+      "loss": 0.9105,
+      "step": 3719
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.1962376832962036,
+      "learning_rate": 0.00015171646425725204,
+      "loss": 0.9684,
+      "step": 3720
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.20075201988220215,
+      "learning_rate": 0.00015169276365912583,
+      "loss": 0.8843,
+      "step": 3721
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.2069513201713562,
+      "learning_rate": 0.0001516690590978184,
+      "loss": 0.9751,
+      "step": 3722
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.20042923092842102,
+      "learning_rate": 0.0001516453505751471,
+      "loss": 0.9644,
+      "step": 3723
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.19810305535793304,
+      "learning_rate": 0.00015162163809292957,
+      "loss": 0.9473,
+      "step": 3724
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.20136240124702454,
+      "learning_rate": 0.00015159792165298386,
+      "loss": 0.9029,
+      "step": 3725
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.19834771752357483,
+      "learning_rate": 0.00015157420125712825,
+      "loss": 0.9343,
+      "step": 3726
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.19727087020874023,
+      "learning_rate": 0.0001515504769071813,
+      "loss": 0.901,
+      "step": 3727
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.19946539402008057,
+      "learning_rate": 0.00015152674860496195,
+      "loss": 0.9179,
+      "step": 3728
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.19287483394145966,
+      "learning_rate": 0.00015150301635228936,
+      "loss": 0.9113,
+      "step": 3729
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.19016975164413452,
+      "learning_rate": 0.0001514792801509831,
+      "loss": 0.9222,
+      "step": 3730
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.19780956208705902,
+      "learning_rate": 0.0001514555400028629,
+      "loss": 0.9618,
+      "step": 3731
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.19044287502765656,
+      "learning_rate": 0.0001514317959097489,
+      "loss": 0.858,
+      "step": 3732
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.1943211555480957,
+      "learning_rate": 0.00015140804787346153,
+      "loss": 0.8965,
+      "step": 3733
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.2054436206817627,
+      "learning_rate": 0.00015138429589582148,
+      "loss": 0.9122,
+      "step": 3734
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.2009715586900711,
+      "learning_rate": 0.00015136053997864983,
+      "loss": 0.8797,
+      "step": 3735
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.18632814288139343,
+      "learning_rate": 0.00015133678012376777,
+      "loss": 0.8947,
+      "step": 3736
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.2046545445919037,
+      "learning_rate": 0.00015131301633299705,
+      "loss": 0.9694,
+      "step": 3737
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.18827247619628906,
+      "learning_rate": 0.00015128924860815954,
+      "loss": 0.9236,
+      "step": 3738
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.19354717433452606,
+      "learning_rate": 0.0001512654769510774,
+      "loss": 0.9056,
+      "step": 3739
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.20237867534160614,
+      "learning_rate": 0.00015124170136357332,
+      "loss": 0.8638,
+      "step": 3740
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.19454175233840942,
+      "learning_rate": 0.00015121792184746998,
+      "loss": 0.9331,
+      "step": 3741
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.19441847503185272,
+      "learning_rate": 0.00015119413840459055,
+      "loss": 0.9276,
+      "step": 3742
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.20244662463665009,
+      "learning_rate": 0.0001511703510367585,
+      "loss": 0.938,
+      "step": 3743
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.2046027034521103,
+      "learning_rate": 0.00015114655974579746,
+      "loss": 0.9821,
+      "step": 3744
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.19821032881736755,
+      "learning_rate": 0.00015112276453353157,
+      "loss": 0.9499,
+      "step": 3745
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.1993158459663391,
+      "learning_rate": 0.0001510989654017851,
+      "loss": 0.9264,
+      "step": 3746
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.19219717383384705,
+      "learning_rate": 0.00015107516235238273,
+      "loss": 0.9408,
+      "step": 3747
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.20064875483512878,
+      "learning_rate": 0.00015105135538714937,
+      "loss": 0.9396,
+      "step": 3748
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.19371774792671204,
+      "learning_rate": 0.0001510275445079102,
+      "loss": 0.8951,
+      "step": 3749
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.19230003654956818,
+      "learning_rate": 0.00015100372971649082,
+      "loss": 0.9125,
+      "step": 3750
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.1980389803647995,
+      "learning_rate": 0.000150979911014717,
+      "loss": 0.8656,
+      "step": 3751
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.19611652195453644,
+      "learning_rate": 0.000150956088404415,
+      "loss": 0.9375,
+      "step": 3752
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.19706304371356964,
+      "learning_rate": 0.0001509322618874111,
+      "loss": 0.8721,
+      "step": 3753
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.19378282129764557,
+      "learning_rate": 0.0001509084314655321,
+      "loss": 0.8858,
+      "step": 3754
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.20083080232143402,
+      "learning_rate": 0.00015088459714060503,
+      "loss": 0.9249,
+      "step": 3755
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.1986744999885559,
+      "learning_rate": 0.00015086075891445724,
+      "loss": 0.9235,
+      "step": 3756
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.19933581352233887,
+      "learning_rate": 0.00015083691678891633,
+      "loss": 0.9516,
+      "step": 3757
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.20001207292079926,
+      "learning_rate": 0.00015081307076581023,
+      "loss": 0.9141,
+      "step": 3758
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.1966947466135025,
+      "learning_rate": 0.00015078922084696715,
+      "loss": 0.8709,
+      "step": 3759
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.1952902227640152,
+      "learning_rate": 0.00015076536703421565,
+      "loss": 0.9097,
+      "step": 3760
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.19831472635269165,
+      "learning_rate": 0.00015074150932938455,
+      "loss": 0.8747,
+      "step": 3761
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.20838016271591187,
+      "learning_rate": 0.000150717647734303,
+      "loss": 0.8938,
+      "step": 3762
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.193468376994133,
+      "learning_rate": 0.00015069378225080032,
+      "loss": 0.9141,
+      "step": 3763
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.19065572321414948,
+      "learning_rate": 0.0001506699128807063,
+      "loss": 0.917,
+      "step": 3764
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.19850143790245056,
+      "learning_rate": 0.000150646039625851,
+      "loss": 0.9113,
+      "step": 3765
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.19818361103534698,
+      "learning_rate": 0.0001506221624880647,
+      "loss": 0.9928,
+      "step": 3766
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.20216867327690125,
+      "learning_rate": 0.00015059828146917793,
+      "loss": 0.9086,
+      "step": 3767
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.19267527759075165,
+      "learning_rate": 0.0001505743965710217,
+      "loss": 0.8939,
+      "step": 3768
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.19998551905155182,
+      "learning_rate": 0.00015055050779542718,
+      "loss": 0.8814,
+      "step": 3769
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.19110707938671112,
+      "learning_rate": 0.0001505266151442259,
+      "loss": 0.8582,
+      "step": 3770
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.19865083694458008,
+      "learning_rate": 0.00015050271861924967,
+      "loss": 0.9322,
+      "step": 3771
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.19565609097480774,
+      "learning_rate": 0.00015047881822233053,
+      "loss": 0.8699,
+      "step": 3772
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.1937209516763687,
+      "learning_rate": 0.0001504549139553009,
+      "loss": 0.8931,
+      "step": 3773
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.1972823441028595,
+      "learning_rate": 0.0001504310058199935,
+      "loss": 0.9337,
+      "step": 3774
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.20414002239704132,
+      "learning_rate": 0.0001504070938182413,
+      "loss": 0.9194,
+      "step": 3775
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.1957191377878189,
+      "learning_rate": 0.00015038317795187757,
+      "loss": 0.8957,
+      "step": 3776
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.2019643932580948,
+      "learning_rate": 0.00015035925822273592,
+      "loss": 0.9323,
+      "step": 3777
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.19081658124923706,
+      "learning_rate": 0.0001503353346326502,
+      "loss": 0.979,
+      "step": 3778
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.20455485582351685,
+      "learning_rate": 0.0001503114071834546,
+      "loss": 0.9567,
+      "step": 3779
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.19730593264102936,
+      "learning_rate": 0.00015028747587698363,
+      "loss": 0.8996,
+      "step": 3780
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.2002158761024475,
+      "learning_rate": 0.000150263540715072,
+      "loss": 0.9626,
+      "step": 3781
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.20014959573745728,
+      "learning_rate": 0.00015023960169955477,
+      "loss": 0.8964,
+      "step": 3782
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.19767984747886658,
+      "learning_rate": 0.00015021565883226732,
+      "loss": 0.9118,
+      "step": 3783
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.20115704834461212,
+      "learning_rate": 0.0001501917121150453,
+      "loss": 0.9544,
+      "step": 3784
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.20704735815525055,
+      "learning_rate": 0.00015016776154972466,
+      "loss": 0.9204,
+      "step": 3785
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.19559411704540253,
+      "learning_rate": 0.00015014380713814158,
+      "loss": 0.9321,
+      "step": 3786
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.198293074965477,
+      "learning_rate": 0.00015011984888213274,
+      "loss": 0.8662,
+      "step": 3787
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.20051126182079315,
+      "learning_rate": 0.00015009588678353483,
+      "loss": 0.8989,
+      "step": 3788
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.20624862611293793,
+      "learning_rate": 0.00015007192084418502,
+      "loss": 0.9028,
+      "step": 3789
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.20225314795970917,
+      "learning_rate": 0.00015004795106592078,
+      "loss": 0.9514,
+      "step": 3790
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.2037535309791565,
+      "learning_rate": 0.00015002397745057975,
+      "loss": 0.9128,
+      "step": 3791
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.20016394555568695,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 0.9348,
+      "step": 3792
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.19721950590610504,
+      "learning_rate": 0.00014997601871601982,
+      "loss": 0.9399,
+      "step": 3793
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.20106512308120728,
+      "learning_rate": 0.00014995203360047775,
+      "loss": 0.9711,
+      "step": 3794
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.1967417299747467,
+      "learning_rate": 0.00014992804465521276,
+      "loss": 0.95,
+      "step": 3795
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.19685690104961395,
+      "learning_rate": 0.00014990405188206401,
+      "loss": 0.9278,
+      "step": 3796
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.20026066899299622,
+      "learning_rate": 0.00014988005528287097,
+      "loss": 0.8825,
+      "step": 3797
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.1951385736465454,
+      "learning_rate": 0.0001498560548594734,
+      "loss": 0.8832,
+      "step": 3798
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.19592271745204926,
+      "learning_rate": 0.00014983205061371138,
+      "loss": 0.8754,
+      "step": 3799
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.18903060257434845,
+      "learning_rate": 0.00014980804254742525,
+      "loss": 0.9081,
+      "step": 3800
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.19677716493606567,
+      "learning_rate": 0.00014978403066245568,
+      "loss": 0.9202,
+      "step": 3801
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.19893287122249603,
+      "learning_rate": 0.0001497600149606436,
+      "loss": 0.9271,
+      "step": 3802
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.20432735979557037,
+      "learning_rate": 0.00014973599544383027,
+      "loss": 0.9466,
+      "step": 3803
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.19333679974079132,
+      "learning_rate": 0.0001497119721138572,
+      "loss": 0.9862,
+      "step": 3804
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.20335853099822998,
+      "learning_rate": 0.00014968794497256623,
+      "loss": 0.9579,
+      "step": 3805
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.19586090743541718,
+      "learning_rate": 0.00014966391402179943,
+      "loss": 0.8735,
+      "step": 3806
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.19172383844852448,
+      "learning_rate": 0.00014963987926339925,
+      "loss": 0.9268,
+      "step": 3807
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.20108972489833832,
+      "learning_rate": 0.00014961584069920835,
+      "loss": 0.894,
+      "step": 3808
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.19839361310005188,
+      "learning_rate": 0.00014959179833106973,
+      "loss": 0.902,
+      "step": 3809
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.18988490104675293,
+      "learning_rate": 0.0001495677521608267,
+      "loss": 0.919,
+      "step": 3810
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.201633483171463,
+      "learning_rate": 0.00014954370219032282,
+      "loss": 0.9612,
+      "step": 3811
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.19168926775455475,
+      "learning_rate": 0.00014951964842140192,
+      "loss": 0.8556,
+      "step": 3812
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.19750337302684784,
+      "learning_rate": 0.00014949559085590818,
+      "loss": 0.9275,
+      "step": 3813
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.20000135898590088,
+      "learning_rate": 0.00014947152949568605,
+      "loss": 0.942,
+      "step": 3814
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.19141016900539398,
+      "learning_rate": 0.00014944746434258026,
+      "loss": 0.9408,
+      "step": 3815
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.19368094205856323,
+      "learning_rate": 0.00014942339539843584,
+      "loss": 0.9202,
+      "step": 3816
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.19549012184143066,
+      "learning_rate": 0.00014939932266509807,
+      "loss": 0.9736,
+      "step": 3817
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.19119693338871002,
+      "learning_rate": 0.00014937524614441264,
+      "loss": 0.8828,
+      "step": 3818
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.1875646561384201,
+      "learning_rate": 0.00014935116583822538,
+      "loss": 0.8702,
+      "step": 3819
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.1974024921655655,
+      "learning_rate": 0.00014932708174838252,
+      "loss": 0.9605,
+      "step": 3820
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.2048822045326233,
+      "learning_rate": 0.00014930299387673054,
+      "loss": 0.9164,
+      "step": 3821
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.18495222926139832,
+      "learning_rate": 0.00014927890222511615,
+      "loss": 0.8123,
+      "step": 3822
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.19920560717582703,
+      "learning_rate": 0.00014925480679538647,
+      "loss": 0.9169,
+      "step": 3823
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.1953980028629303,
+      "learning_rate": 0.0001492307075893888,
+      "loss": 0.8839,
+      "step": 3824
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.20536701381206512,
+      "learning_rate": 0.00014920660460897083,
+      "loss": 0.9569,
+      "step": 3825
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.19224631786346436,
+      "learning_rate": 0.00014918249785598046,
+      "loss": 0.8981,
+      "step": 3826
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.2037082314491272,
+      "learning_rate": 0.00014915838733226594,
+      "loss": 1.0014,
+      "step": 3827
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.19932124018669128,
+      "learning_rate": 0.00014913427303967568,
+      "loss": 0.9632,
+      "step": 3828
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.20192798972129822,
+      "learning_rate": 0.00014911015498005859,
+      "loss": 0.9635,
+      "step": 3829
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.19329585134983063,
+      "learning_rate": 0.00014908603315526368,
+      "loss": 0.8993,
+      "step": 3830
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.19806140661239624,
+      "learning_rate": 0.0001490619075671404,
+      "loss": 0.9436,
+      "step": 3831
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.20333150029182434,
+      "learning_rate": 0.0001490377782175383,
+      "loss": 0.9349,
+      "step": 3832
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.194121852517128,
+      "learning_rate": 0.0001490136451083074,
+      "loss": 0.9253,
+      "step": 3833
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.19719244539737701,
+      "learning_rate": 0.0001489895082412979,
+      "loss": 0.992,
+      "step": 3834
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.1979801207780838,
+      "learning_rate": 0.0001489653676183604,
+      "loss": 0.9117,
+      "step": 3835
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.1975490152835846,
+      "learning_rate": 0.00014894122324134564,
+      "loss": 0.8944,
+      "step": 3836
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.2017369270324707,
+      "learning_rate": 0.00014891707511210473,
+      "loss": 0.9631,
+      "step": 3837
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.19351305067539215,
+      "learning_rate": 0.0001488929232324891,
+      "loss": 0.9337,
+      "step": 3838
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.18836168944835663,
+      "learning_rate": 0.0001488687676043504,
+      "loss": 0.862,
+      "step": 3839
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.19105665385723114,
+      "learning_rate": 0.00014884460822954057,
+      "loss": 0.9026,
+      "step": 3840
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.19071929156780243,
+      "learning_rate": 0.0001488204451099119,
+      "loss": 0.9249,
+      "step": 3841
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.2096261978149414,
+      "learning_rate": 0.00014879627824731693,
+      "loss": 0.9825,
+      "step": 3842
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.19523192942142487,
+      "learning_rate": 0.00014877210764360847,
+      "loss": 0.8733,
+      "step": 3843
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.19203850626945496,
+      "learning_rate": 0.0001487479333006396,
+      "loss": 0.8879,
+      "step": 3844
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.19315475225448608,
+      "learning_rate": 0.00014872375522026376,
+      "loss": 0.9211,
+      "step": 3845
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.1942511945962906,
+      "learning_rate": 0.00014869957340433461,
+      "loss": 0.9567,
+      "step": 3846
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.199487566947937,
+      "learning_rate": 0.00014867538785470617,
+      "loss": 0.8816,
+      "step": 3847
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.20415233075618744,
+      "learning_rate": 0.00014865119857323267,
+      "loss": 0.8951,
+      "step": 3848
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.19348014891147614,
+      "learning_rate": 0.0001486270055617686,
+      "loss": 0.9019,
+      "step": 3849
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.19001100957393646,
+      "learning_rate": 0.0001486028088221689,
+      "loss": 0.9418,
+      "step": 3850
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.18715202808380127,
+      "learning_rate": 0.00014857860835628856,
+      "loss": 0.8782,
+      "step": 3851
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.19490717351436615,
+      "learning_rate": 0.00014855440416598309,
+      "loss": 0.9294,
+      "step": 3852
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.21514825522899628,
+      "learning_rate": 0.00014853019625310813,
+      "loss": 0.9061,
+      "step": 3853
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.20610563457012177,
+      "learning_rate": 0.00014850598461951963,
+      "loss": 0.8979,
+      "step": 3854
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.1885242909193039,
+      "learning_rate": 0.0001484817692670739,
+      "loss": 0.915,
+      "step": 3855
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.20315773785114288,
+      "learning_rate": 0.00014845755019762743,
+      "loss": 0.9954,
+      "step": 3856
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.2000637799501419,
+      "learning_rate": 0.0001484333274130371,
+      "loss": 0.8913,
+      "step": 3857
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.20190002024173737,
+      "learning_rate": 0.00014840910091515997,
+      "loss": 0.8499,
+      "step": 3858
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.20083366334438324,
+      "learning_rate": 0.0001483848707058535,
+      "loss": 0.8497,
+      "step": 3859
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.20369665324687958,
+      "learning_rate": 0.0001483606367869753,
+      "loss": 0.9244,
+      "step": 3860
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.19197633862495422,
+      "learning_rate": 0.0001483363991603834,
+      "loss": 0.8958,
+      "step": 3861
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.19108688831329346,
+      "learning_rate": 0.000148312157827936,
+      "loss": 0.8965,
+      "step": 3862
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.19200675189495087,
+      "learning_rate": 0.00014828791279149168,
+      "loss": 0.9681,
+      "step": 3863
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.18933221697807312,
+      "learning_rate": 0.00014826366405290922,
+      "loss": 0.916,
+      "step": 3864
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.19670377671718597,
+      "learning_rate": 0.00014823941161404774,
+      "loss": 0.943,
+      "step": 3865
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.18979068100452423,
+      "learning_rate": 0.00014821515547676662,
+      "loss": 0.8877,
+      "step": 3866
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.18644832074642181,
+      "learning_rate": 0.00014819089564292552,
+      "loss": 0.8796,
+      "step": 3867
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.19427362084388733,
+      "learning_rate": 0.00014816663211438447,
+      "loss": 0.9048,
+      "step": 3868
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.20413526892662048,
+      "learning_rate": 0.0001481423648930036,
+      "loss": 0.9031,
+      "step": 3869
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.20067287981510162,
+      "learning_rate": 0.00014811809398064348,
+      "loss": 0.9111,
+      "step": 3870
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.1960747241973877,
+      "learning_rate": 0.00014809381937916488,
+      "loss": 0.8996,
+      "step": 3871
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.1922447681427002,
+      "learning_rate": 0.0001480695410904289,
+      "loss": 0.9269,
+      "step": 3872
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.19527967274188995,
+      "learning_rate": 0.000148045259116297,
+      "loss": 0.9108,
+      "step": 3873
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.18520699441432953,
+      "learning_rate": 0.0001480209734586307,
+      "loss": 0.8605,
+      "step": 3874
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.21494515240192413,
+      "learning_rate": 0.000147996684119292,
+      "loss": 0.9773,
+      "step": 3875
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.1982784867286682,
+      "learning_rate": 0.00014797239110014306,
+      "loss": 0.9689,
+      "step": 3876
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.1936819702386856,
+      "learning_rate": 0.00014794809440304646,
+      "loss": 0.9486,
+      "step": 3877
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.20160476863384247,
+      "learning_rate": 0.00014792379402986492,
+      "loss": 0.9043,
+      "step": 3878
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.20444336533546448,
+      "learning_rate": 0.0001478994899824615,
+      "loss": 0.9223,
+      "step": 3879
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.19447480142116547,
+      "learning_rate": 0.0001478751822626996,
+      "loss": 0.9149,
+      "step": 3880
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.19876286387443542,
+      "learning_rate": 0.00014785087087244275,
+      "loss": 0.9407,
+      "step": 3881
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.19465559720993042,
+      "learning_rate": 0.00014782655581355495,
+      "loss": 0.9301,
+      "step": 3882
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.19302017986774445,
+      "learning_rate": 0.0001478022370879003,
+      "loss": 0.8842,
+      "step": 3883
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.1959199607372284,
+      "learning_rate": 0.00014777791469734337,
+      "loss": 0.8692,
+      "step": 3884
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.18946966528892517,
+      "learning_rate": 0.00014775358864374885,
+      "loss": 0.8763,
+      "step": 3885
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.1975627839565277,
+      "learning_rate": 0.00014772925892898175,
+      "loss": 0.9775,
+      "step": 3886
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.2010650932788849,
+      "learning_rate": 0.0001477049255549074,
+      "loss": 0.9001,
+      "step": 3887
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.1925669014453888,
+      "learning_rate": 0.00014768058852339142,
+      "loss": 0.8867,
+      "step": 3888
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.19858621060848236,
+      "learning_rate": 0.00014765624783629963,
+      "loss": 0.9029,
+      "step": 3889
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.2079782485961914,
+      "learning_rate": 0.00014763190349549823,
+      "loss": 0.9251,
+      "step": 3890
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.1923612803220749,
+      "learning_rate": 0.00014760755550285366,
+      "loss": 0.8288,
+      "step": 3891
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.19295886158943176,
+      "learning_rate": 0.00014758320386023255,
+      "loss": 0.8466,
+      "step": 3892
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.19733382761478424,
+      "learning_rate": 0.00014755884856950196,
+      "loss": 0.8735,
+      "step": 3893
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.19566164910793304,
+      "learning_rate": 0.00014753448963252918,
+      "loss": 0.8891,
+      "step": 3894
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.1920028030872345,
+      "learning_rate": 0.00014751012705118172,
+      "loss": 0.9059,
+      "step": 3895
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.2053966224193573,
+      "learning_rate": 0.00014748576082732742,
+      "loss": 0.8818,
+      "step": 3896
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.20320646464824677,
+      "learning_rate": 0.00014746139096283436,
+      "loss": 0.8939,
+      "step": 3897
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.20649878680706024,
+      "learning_rate": 0.000147437017459571,
+      "loss": 0.9206,
+      "step": 3898
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.19128867983818054,
+      "learning_rate": 0.00014741264031940593,
+      "loss": 0.8968,
+      "step": 3899
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.18966726958751678,
+      "learning_rate": 0.00014738825954420815,
+      "loss": 0.8833,
+      "step": 3900
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.1958254873752594,
+      "learning_rate": 0.00014736387513584686,
+      "loss": 0.9248,
+      "step": 3901
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.19757814705371857,
+      "learning_rate": 0.00014733948709619158,
+      "loss": 0.9552,
+      "step": 3902
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.20112890005111694,
+      "learning_rate": 0.0001473150954271121,
+      "loss": 0.9392,
+      "step": 3903
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.2000960111618042,
+      "learning_rate": 0.00014729070013047845,
+      "loss": 0.9083,
+      "step": 3904
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.1944068968296051,
+      "learning_rate": 0.000147266301208161,
+      "loss": 0.9276,
+      "step": 3905
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.2013338953256607,
+      "learning_rate": 0.00014724189866203037,
+      "loss": 0.9418,
+      "step": 3906
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.2008303552865982,
+      "learning_rate": 0.00014721749249395737,
+      "loss": 0.9855,
+      "step": 3907
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.20033757388591766,
+      "learning_rate": 0.00014719308270581328,
+      "loss": 0.9395,
+      "step": 3908
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.21251152455806732,
+      "learning_rate": 0.0001471686692994695,
+      "loss": 0.9209,
+      "step": 3909
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.19628505408763885,
+      "learning_rate": 0.00014714425227679777,
+      "loss": 0.9154,
+      "step": 3910
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.1894054412841797,
+      "learning_rate": 0.0001471198316396701,
+      "loss": 0.9032,
+      "step": 3911
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.19757375121116638,
+      "learning_rate": 0.00014709540738995876,
+      "loss": 0.8706,
+      "step": 3912
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.20138540863990784,
+      "learning_rate": 0.0001470709795295363,
+      "loss": 0.9553,
+      "step": 3913
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.19919735193252563,
+      "learning_rate": 0.0001470465480602756,
+      "loss": 0.9502,
+      "step": 3914
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.19295333325862885,
+      "learning_rate": 0.00014702211298404968,
+      "loss": 0.9364,
+      "step": 3915
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.1922289878129959,
+      "learning_rate": 0.000146997674302732,
+      "loss": 0.9687,
+      "step": 3916
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.20261609554290771,
+      "learning_rate": 0.00014697323201819624,
+      "loss": 0.9599,
+      "step": 3917
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.1992306411266327,
+      "learning_rate": 0.0001469487861323163,
+      "loss": 0.8967,
+      "step": 3918
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.2102137804031372,
+      "learning_rate": 0.0001469243366469664,
+      "loss": 1.0319,
+      "step": 3919
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.1930590718984604,
+      "learning_rate": 0.00014689988356402108,
+      "loss": 0.9564,
+      "step": 3920
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.19371606409549713,
+      "learning_rate": 0.00014687542688535506,
+      "loss": 0.9345,
+      "step": 3921
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.20073796808719635,
+      "learning_rate": 0.00014685096661284336,
+      "loss": 0.8684,
+      "step": 3922
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.21918940544128418,
+      "learning_rate": 0.00014682650274836138,
+      "loss": 0.9269,
+      "step": 3923
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.21027129888534546,
+      "learning_rate": 0.00014680203529378468,
+      "loss": 0.9555,
+      "step": 3924
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.19333447515964508,
+      "learning_rate": 0.00014677756425098907,
+      "loss": 0.9541,
+      "step": 3925
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.20419958233833313,
+      "learning_rate": 0.0001467530896218508,
+      "loss": 0.9048,
+      "step": 3926
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.19317831099033356,
+      "learning_rate": 0.00014672861140824624,
+      "loss": 0.9029,
+      "step": 3927
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.20992499589920044,
+      "learning_rate": 0.00014670412961205207,
+      "loss": 0.9877,
+      "step": 3928
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.20966270565986633,
+      "learning_rate": 0.00014667964423514528,
+      "loss": 0.9048,
+      "step": 3929
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.20526836812496185,
+      "learning_rate": 0.00014665515527940312,
+      "loss": 0.8693,
+      "step": 3930
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.20329995453357697,
+      "learning_rate": 0.0001466306627467031,
+      "loss": 1.0054,
+      "step": 3931
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.20244084298610687,
+      "learning_rate": 0.000146606166638923,
+      "loss": 0.9703,
+      "step": 3932
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.192376971244812,
+      "learning_rate": 0.00014658166695794095,
+      "loss": 0.8323,
+      "step": 3933
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.18748922646045685,
+      "learning_rate": 0.00014655716370563522,
+      "loss": 0.8728,
+      "step": 3934
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.2070818841457367,
+      "learning_rate": 0.00014653265688388447,
+      "loss": 0.9134,
+      "step": 3935
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.20446988940238953,
+      "learning_rate": 0.00014650814649456754,
+      "loss": 0.9876,
+      "step": 3936
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.19711360335350037,
+      "learning_rate": 0.00014648363253956363,
+      "loss": 0.916,
+      "step": 3937
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.19562388956546783,
+      "learning_rate": 0.0001464591150207522,
+      "loss": 0.8866,
+      "step": 3938
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.1958800107240677,
+      "learning_rate": 0.00014643459394001292,
+      "loss": 0.8752,
+      "step": 3939
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.2000405341386795,
+      "learning_rate": 0.00014641006929922576,
+      "loss": 0.9439,
+      "step": 3940
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.2032128870487213,
+      "learning_rate": 0.000146385541100271,
+      "loss": 0.9456,
+      "step": 3941
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.20214006304740906,
+      "learning_rate": 0.00014636100934502916,
+      "loss": 0.8781,
+      "step": 3942
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.21211445331573486,
+      "learning_rate": 0.00014633647403538107,
+      "loss": 0.9876,
+      "step": 3943
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.20307664573192596,
+      "learning_rate": 0.00014631193517320777,
+      "loss": 0.9049,
+      "step": 3944
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.1985590010881424,
+      "learning_rate": 0.00014628739276039063,
+      "loss": 0.8968,
+      "step": 3945
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.20269496738910675,
+      "learning_rate": 0.00014626284679881125,
+      "loss": 0.9093,
+      "step": 3946
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.19618482887744904,
+      "learning_rate": 0.0001462382972903515,
+      "loss": 0.9047,
+      "step": 3947
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.18889334797859192,
+      "learning_rate": 0.00014621374423689357,
+      "loss": 0.8841,
+      "step": 3948
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.19615551829338074,
+      "learning_rate": 0.00014618918764031991,
+      "loss": 0.8833,
+      "step": 3949
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.20412616431713104,
+      "learning_rate": 0.0001461646275025132,
+      "loss": 0.9088,
+      "step": 3950
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.19538843631744385,
+      "learning_rate": 0.0001461400638253564,
+      "loss": 0.9257,
+      "step": 3951
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.1954936534166336,
+      "learning_rate": 0.0001461154966107328,
+      "loss": 0.874,
+      "step": 3952
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.20418086647987366,
+      "learning_rate": 0.0001460909258605259,
+      "loss": 0.9677,
+      "step": 3953
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.1930086761713028,
+      "learning_rate": 0.00014606635157661947,
+      "loss": 0.9172,
+      "step": 3954
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.18511292338371277,
+      "learning_rate": 0.00014604177376089763,
+      "loss": 0.8675,
+      "step": 3955
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.19437043368816376,
+      "learning_rate": 0.00014601719241524462,
+      "loss": 0.881,
+      "step": 3956
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.19371411204338074,
+      "learning_rate": 0.0001459926075415451,
+      "loss": 0.915,
+      "step": 3957
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.19958390295505524,
+      "learning_rate": 0.000145968019141684,
+      "loss": 0.9678,
+      "step": 3958
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.20946894586086273,
+      "learning_rate": 0.00014594342721754635,
+      "loss": 0.9311,
+      "step": 3959
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.19142213463783264,
+      "learning_rate": 0.00014591883177101763,
+      "loss": 0.8736,
+      "step": 3960
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.1921462118625641,
+      "learning_rate": 0.00014589423280398354,
+      "loss": 0.8867,
+      "step": 3961
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.19228462874889374,
+      "learning_rate": 0.00014586963031832994,
+      "loss": 0.9924,
+      "step": 3962
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.19412390887737274,
+      "learning_rate": 0.00014584502431594316,
+      "loss": 0.8861,
+      "step": 3963
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.19502438604831696,
+      "learning_rate": 0.00014582041479870966,
+      "loss": 0.8712,
+      "step": 3964
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.20914730429649353,
+      "learning_rate": 0.00014579580176851616,
+      "loss": 1.0042,
+      "step": 3965
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.2027062028646469,
+      "learning_rate": 0.00014577118522724977,
+      "loss": 0.9674,
+      "step": 3966
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.19774848222732544,
+      "learning_rate": 0.0001457465651767977,
+      "loss": 0.9081,
+      "step": 3967
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.19559279084205627,
+      "learning_rate": 0.0001457219416190476,
+      "loss": 0.8935,
+      "step": 3968
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.21025317907333374,
+      "learning_rate": 0.00014569731455588728,
+      "loss": 0.9731,
+      "step": 3969
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.1929987221956253,
+      "learning_rate": 0.0001456726839892048,
+      "loss": 0.8873,
+      "step": 3970
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.18292102217674255,
+      "learning_rate": 0.00014564804992088864,
+      "loss": 0.8333,
+      "step": 3971
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.20394857227802277,
+      "learning_rate": 0.00014562341235282734,
+      "loss": 1.0219,
+      "step": 3972
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.20659831166267395,
+      "learning_rate": 0.0001455987712869099,
+      "loss": 1.0135,
+      "step": 3973
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.1952691227197647,
+      "learning_rate": 0.00014557412672502542,
+      "loss": 0.8934,
+      "step": 3974
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.18991760909557343,
+      "learning_rate": 0.0001455494786690634,
+      "loss": 0.8838,
+      "step": 3975
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.19869369268417358,
+      "learning_rate": 0.00014552482712091357,
+      "loss": 0.9144,
+      "step": 3976
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.20484110713005066,
+      "learning_rate": 0.00014550017208246586,
+      "loss": 0.9326,
+      "step": 3977
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.19844062626361847,
+      "learning_rate": 0.0001454755135556106,
+      "loss": 0.9201,
+      "step": 3978
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.1981542855501175,
+      "learning_rate": 0.00014545085154223824,
+      "loss": 0.9259,
+      "step": 3979
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.19899936020374298,
+      "learning_rate": 0.00014542618604423957,
+      "loss": 0.8957,
+      "step": 3980
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.197379469871521,
+      "learning_rate": 0.0001454015170635057,
+      "loss": 0.9057,
+      "step": 3981
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.20835913717746735,
+      "learning_rate": 0.0001453768446019279,
+      "loss": 0.933,
+      "step": 3982
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.21079415082931519,
+      "learning_rate": 0.0001453521686613978,
+      "loss": 0.9159,
+      "step": 3983
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.1958971619606018,
+      "learning_rate": 0.00014532748924380722,
+      "loss": 0.9019,
+      "step": 3984
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.19664226472377777,
+      "learning_rate": 0.00014530280635104828,
+      "loss": 0.9261,
+      "step": 3985
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.1906144917011261,
+      "learning_rate": 0.00014527811998501342,
+      "loss": 0.8938,
+      "step": 3986
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.20075096189975739,
+      "learning_rate": 0.00014525343014759522,
+      "loss": 0.9138,
+      "step": 3987
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.20587298274040222,
+      "learning_rate": 0.00014522873684068665,
+      "loss": 0.9163,
+      "step": 3988
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.19886673986911774,
+      "learning_rate": 0.00014520404006618088,
+      "loss": 0.9025,
+      "step": 3989
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.19743849337100983,
+      "learning_rate": 0.00014517933982597137,
+      "loss": 0.9412,
+      "step": 3990
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.1906091868877411,
+      "learning_rate": 0.0001451546361219519,
+      "loss": 0.8986,
+      "step": 3991
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.1904173046350479,
+      "learning_rate": 0.0001451299289560163,
+      "loss": 0.8772,
+      "step": 3992
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.19968535006046295,
+      "learning_rate": 0.00014510521833005892,
+      "loss": 0.9069,
+      "step": 3993
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.20211651921272278,
+      "learning_rate": 0.0001450805042459743,
+      "loss": 0.9065,
+      "step": 3994
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.19885098934173584,
+      "learning_rate": 0.00014505578670565713,
+      "loss": 0.9555,
+      "step": 3995
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.19500750303268433,
+      "learning_rate": 0.00014503106571100259,
+      "loss": 0.8783,
+      "step": 3996
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.2058175504207611,
+      "learning_rate": 0.00014500634126390585,
+      "loss": 0.9622,
+      "step": 3997
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.19853660464286804,
+      "learning_rate": 0.00014498161336626254,
+      "loss": 0.9553,
+      "step": 3998
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.19556358456611633,
+      "learning_rate": 0.00014495688201996852,
+      "loss": 0.93,
+      "step": 3999
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.21239247918128967,
+      "learning_rate": 0.00014493214722691982,
+      "loss": 0.9389,
+      "step": 4000
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.20066651701927185,
+      "learning_rate": 0.0001449074089890129,
+      "loss": 0.9596,
+      "step": 4001
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.1968492865562439,
+      "learning_rate": 0.00014488266730814438,
+      "loss": 0.941,
+      "step": 4002
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.18987908959388733,
+      "learning_rate": 0.0001448579221862111,
+      "loss": 0.9014,
+      "step": 4003
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.18961165845394135,
+      "learning_rate": 0.00014483317362511027,
+      "loss": 0.8637,
+      "step": 4004
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.1986960470676422,
+      "learning_rate": 0.00014480842162673923,
+      "loss": 0.9423,
+      "step": 4005
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.2111463099718094,
+      "learning_rate": 0.00014478366619299576,
+      "loss": 0.8806,
+      "step": 4006
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.18590988218784332,
+      "learning_rate": 0.00014475890732577778,
+      "loss": 0.8824,
+      "step": 4007
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.19554437696933746,
+      "learning_rate": 0.0001447341450269835,
+      "loss": 0.9777,
+      "step": 4008
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.19814185798168182,
+      "learning_rate": 0.0001447093792985114,
+      "loss": 0.8583,
+      "step": 4009
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.2064286172389984,
+      "learning_rate": 0.0001446846101422602,
+      "loss": 0.9261,
+      "step": 4010
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.208113431930542,
+      "learning_rate": 0.00014465983756012892,
+      "loss": 0.9472,
+      "step": 4011
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.20458854734897614,
+      "learning_rate": 0.0001446350615540168,
+      "loss": 0.936,
+      "step": 4012
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.19451703131198883,
+      "learning_rate": 0.00014461028212582342,
+      "loss": 0.9528,
+      "step": 4013
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.1916043907403946,
+      "learning_rate": 0.00014458549927744854,
+      "loss": 0.8705,
+      "step": 4014
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.19787919521331787,
+      "learning_rate": 0.00014456071301079216,
+      "loss": 0.9302,
+      "step": 4015
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.20245125889778137,
+      "learning_rate": 0.00014453592332775466,
+      "loss": 0.9772,
+      "step": 4016
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.19883689284324646,
+      "learning_rate": 0.00014451113023023664,
+      "loss": 0.8607,
+      "step": 4017
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.20537172257900238,
+      "learning_rate": 0.00014448633372013886,
+      "loss": 0.8589,
+      "step": 4018
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.1990588903427124,
+      "learning_rate": 0.00014446153379936246,
+      "loss": 0.935,
+      "step": 4019
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.20056746900081635,
+      "learning_rate": 0.00014443673046980876,
+      "loss": 0.8643,
+      "step": 4020
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.19613921642303467,
+      "learning_rate": 0.00014441192373337947,
+      "loss": 0.9284,
+      "step": 4021
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.19420263171195984,
+      "learning_rate": 0.0001443871135919764,
+      "loss": 0.8983,
+      "step": 4022
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.19666177034378052,
+      "learning_rate": 0.00014436230004750172,
+      "loss": 0.8861,
+      "step": 4023
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.20106558501720428,
+      "learning_rate": 0.0001443374831018578,
+      "loss": 0.9547,
+      "step": 4024
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.1971885859966278,
+      "learning_rate": 0.00014431266275694734,
+      "loss": 0.9432,
+      "step": 4025
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.19694119691848755,
+      "learning_rate": 0.0001442878390146733,
+      "loss": 0.9428,
+      "step": 4026
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.2107475996017456,
+      "learning_rate": 0.0001442630118769388,
+      "loss": 1.0014,
+      "step": 4027
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.19050538539886475,
+      "learning_rate": 0.0001442381813456473,
+      "loss": 0.9064,
+      "step": 4028
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.2018805891275406,
+      "learning_rate": 0.00014421334742270256,
+      "loss": 0.9149,
+      "step": 4029
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.19574452936649323,
+      "learning_rate": 0.00014418851011000847,
+      "loss": 0.9126,
+      "step": 4030
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.19954833388328552,
+      "learning_rate": 0.00014416366940946934,
+      "loss": 0.8749,
+      "step": 4031
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.19605934619903564,
+      "learning_rate": 0.00014413882532298956,
+      "loss": 0.8863,
+      "step": 4032
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.2000497281551361,
+      "learning_rate": 0.00014411397785247398,
+      "loss": 0.9146,
+      "step": 4033
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.19151616096496582,
+      "learning_rate": 0.00014408912699982756,
+      "loss": 0.9435,
+      "step": 4034
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.1961325705051422,
+      "learning_rate": 0.00014406427276695552,
+      "loss": 0.8814,
+      "step": 4035
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.1881711333990097,
+      "learning_rate": 0.00014403941515576344,
+      "loss": 0.8064,
+      "step": 4036
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.19434835016727448,
+      "learning_rate": 0.0001440145541681571,
+      "loss": 0.888,
+      "step": 4037
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.19846580922603607,
+      "learning_rate": 0.00014398968980604252,
+      "loss": 0.942,
+      "step": 4038
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.21146689355373383,
+      "learning_rate": 0.00014396482207132606,
+      "loss": 0.9229,
+      "step": 4039
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.19547779858112335,
+      "learning_rate": 0.00014393995096591416,
+      "loss": 0.8803,
+      "step": 4040
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.19985531270503998,
+      "learning_rate": 0.00014391507649171375,
+      "loss": 0.9313,
+      "step": 4041
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.2019869089126587,
+      "learning_rate": 0.00014389019865063187,
+      "loss": 0.9602,
+      "step": 4042
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.18351396918296814,
+      "learning_rate": 0.00014386531744457585,
+      "loss": 0.8261,
+      "step": 4043
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.19523383677005768,
+      "learning_rate": 0.0001438404328754533,
+      "loss": 0.9308,
+      "step": 4044
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.19494594633579254,
+      "learning_rate": 0.000143815544945172,
+      "loss": 0.9157,
+      "step": 4045
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.19079479575157166,
+      "learning_rate": 0.00014379065365564017,
+      "loss": 0.8429,
+      "step": 4046
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.19934016466140747,
+      "learning_rate": 0.00014376575900876612,
+      "loss": 0.9156,
+      "step": 4047
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.20024937391281128,
+      "learning_rate": 0.00014374086100645846,
+      "loss": 0.9277,
+      "step": 4048
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.19894444942474365,
+      "learning_rate": 0.00014371595965062607,
+      "loss": 0.9388,
+      "step": 4049
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.2016160637140274,
+      "learning_rate": 0.0001436910549431781,
+      "loss": 0.8564,
+      "step": 4050
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.2054978460073471,
+      "learning_rate": 0.00014366614688602396,
+      "loss": 1.0209,
+      "step": 4051
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.18863257765769958,
+      "learning_rate": 0.00014364123548107327,
+      "loss": 0.8814,
+      "step": 4052
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.19060060381889343,
+      "learning_rate": 0.00014361632073023596,
+      "loss": 0.9216,
+      "step": 4053
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.2042149156332016,
+      "learning_rate": 0.0001435914026354222,
+      "loss": 0.9728,
+      "step": 4054
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.20097431540489197,
+      "learning_rate": 0.00014356648119854236,
+      "loss": 0.9218,
+      "step": 4055
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.20287711918354034,
+      "learning_rate": 0.0001435415564215072,
+      "loss": 0.9008,
+      "step": 4056
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.1915043294429779,
+      "learning_rate": 0.00014351662830622757,
+      "loss": 0.8405,
+      "step": 4057
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.20655031502246857,
+      "learning_rate": 0.00014349169685461471,
+      "loss": 1.0139,
+      "step": 4058
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.19209517538547516,
+      "learning_rate": 0.00014346676206858009,
+      "loss": 0.9055,
+      "step": 4059
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.20339743793010712,
+      "learning_rate": 0.00014344182395003533,
+      "loss": 0.9622,
+      "step": 4060
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.19336020946502686,
+      "learning_rate": 0.00014341688250089243,
+      "loss": 0.8918,
+      "step": 4061
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.19621950387954712,
+      "learning_rate": 0.0001433919377230636,
+      "loss": 0.8601,
+      "step": 4062
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.21147488057613373,
+      "learning_rate": 0.00014336698961846134,
+      "loss": 0.9483,
+      "step": 4063
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.2010038048028946,
+      "learning_rate": 0.0001433420381889983,
+      "loss": 0.9204,
+      "step": 4064
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.1948138177394867,
+      "learning_rate": 0.00014331708343658748,
+      "loss": 0.9032,
+      "step": 4065
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.20164598524570465,
+      "learning_rate": 0.00014329212536314217,
+      "loss": 0.8759,
+      "step": 4066
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.20524714887142181,
+      "learning_rate": 0.0001432671639705758,
+      "loss": 0.9107,
+      "step": 4067
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.2029474675655365,
+      "learning_rate": 0.00014324219926080208,
+      "loss": 0.8739,
+      "step": 4068
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.20881551504135132,
+      "learning_rate": 0.0001432172312357351,
+      "loss": 0.9433,
+      "step": 4069
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.198815256357193,
+      "learning_rate": 0.00014319225989728901,
+      "loss": 0.9242,
+      "step": 4070
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.19560791552066803,
+      "learning_rate": 0.0001431672852473784,
+      "loss": 0.9165,
+      "step": 4071
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.2154237926006317,
+      "learning_rate": 0.00014314230728791795,
+      "loss": 0.8763,
+      "step": 4072
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.20212805271148682,
+      "learning_rate": 0.00014311732602082273,
+      "loss": 0.9695,
+      "step": 4073
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.20487231016159058,
+      "learning_rate": 0.00014309234144800796,
+      "loss": 0.9338,
+      "step": 4074
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.19709345698356628,
+      "learning_rate": 0.00014306735357138916,
+      "loss": 0.9005,
+      "step": 4075
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.20154863595962524,
+      "learning_rate": 0.00014304236239288213,
+      "loss": 0.8645,
+      "step": 4076
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.20887543261051178,
+      "learning_rate": 0.0001430173679144029,
+      "loss": 0.9507,
+      "step": 4077
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.2201569527387619,
+      "learning_rate": 0.00014299237013786772,
+      "loss": 0.8901,
+      "step": 4078
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.20078633725643158,
+      "learning_rate": 0.00014296736906519312,
+      "loss": 0.8637,
+      "step": 4079
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.20457907021045685,
+      "learning_rate": 0.00014294236469829588,
+      "loss": 0.947,
+      "step": 4080
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.20842711627483368,
+      "learning_rate": 0.00014291735703909306,
+      "loss": 0.942,
+      "step": 4081
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.21422892808914185,
+      "learning_rate": 0.00014289234608950193,
+      "loss": 0.9287,
+      "step": 4082
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.20222635567188263,
+      "learning_rate": 0.00014286733185144004,
+      "loss": 0.9141,
+      "step": 4083
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.20117203891277313,
+      "learning_rate": 0.00014284231432682516,
+      "loss": 0.8867,
+      "step": 4084
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.20013990998268127,
+      "learning_rate": 0.00014281729351757534,
+      "loss": 0.9631,
+      "step": 4085
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.1925857663154602,
+      "learning_rate": 0.00014279226942560887,
+      "loss": 0.8808,
+      "step": 4086
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.19970326125621796,
+      "learning_rate": 0.00014276724205284434,
+      "loss": 0.9158,
+      "step": 4087
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.2049684375524521,
+      "learning_rate": 0.0001427422114012005,
+      "loss": 0.8965,
+      "step": 4088
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.21222761273384094,
+      "learning_rate": 0.00014271717747259646,
+      "loss": 0.9446,
+      "step": 4089
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.19456645846366882,
+      "learning_rate": 0.00014269214026895144,
+      "loss": 0.855,
+      "step": 4090
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.20363333821296692,
+      "learning_rate": 0.00014266709979218504,
+      "loss": 0.9686,
+      "step": 4091
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.20178015530109406,
+      "learning_rate": 0.00014264205604421704,
+      "loss": 0.9871,
+      "step": 4092
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.19955269992351532,
+      "learning_rate": 0.00014261700902696753,
+      "loss": 0.9779,
+      "step": 4093
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.2012970894575119,
+      "learning_rate": 0.00014259195874235682,
+      "loss": 0.891,
+      "step": 4094
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.19689302146434784,
+      "learning_rate": 0.0001425669051923054,
+      "loss": 0.8973,
+      "step": 4095
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.18733972311019897,
+      "learning_rate": 0.00014254184837873414,
+      "loss": 0.8695,
+      "step": 4096
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.19153453409671783,
+      "learning_rate": 0.00014251678830356408,
+      "loss": 0.9009,
+      "step": 4097
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.1985730081796646,
+      "learning_rate": 0.0001424917249687165,
+      "loss": 0.9046,
+      "step": 4098
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.19961155951023102,
+      "learning_rate": 0.000142466658376113,
+      "loss": 0.9044,
+      "step": 4099
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.1989118456840515,
+      "learning_rate": 0.00014244158852767534,
+      "loss": 0.9417,
+      "step": 4100
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.19771447777748108,
+      "learning_rate": 0.0001424165154253256,
+      "loss": 0.9204,
+      "step": 4101
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.18623396754264832,
+      "learning_rate": 0.0001423914390709861,
+      "loss": 0.8632,
+      "step": 4102
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.19544756412506104,
+      "learning_rate": 0.00014236635946657933,
+      "loss": 0.9045,
+      "step": 4103
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.20649179816246033,
+      "learning_rate": 0.0001423412766140282,
+      "loss": 0.9267,
+      "step": 4104
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.20009225606918335,
+      "learning_rate": 0.00014231619051525564,
+      "loss": 0.8781,
+      "step": 4105
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.20141996443271637,
+      "learning_rate": 0.00014229110117218503,
+      "loss": 0.8861,
+      "step": 4106
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.1917063295841217,
+      "learning_rate": 0.0001422660085867399,
+      "loss": 0.9038,
+      "step": 4107
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.19084444642066956,
+      "learning_rate": 0.00014224091276084406,
+      "loss": 0.8411,
+      "step": 4108
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.19777989387512207,
+      "learning_rate": 0.00014221581369642158,
+      "loss": 0.889,
+      "step": 4109
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.19790798425674438,
+      "learning_rate": 0.00014219071139539668,
+      "loss": 0.9385,
+      "step": 4110
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.19884786009788513,
+      "learning_rate": 0.00014216560585969396,
+      "loss": 0.9052,
+      "step": 4111
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.1907127946615219,
+      "learning_rate": 0.0001421404970912382,
+      "loss": 0.8382,
+      "step": 4112
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.19692125916481018,
+      "learning_rate": 0.00014211538509195442,
+      "loss": 0.8997,
+      "step": 4113
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.2003505676984787,
+      "learning_rate": 0.00014209026986376796,
+      "loss": 0.8873,
+      "step": 4114
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.20393992960453033,
+      "learning_rate": 0.00014206515140860427,
+      "loss": 0.9339,
+      "step": 4115
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.1977144181728363,
+      "learning_rate": 0.0001420400297283892,
+      "loss": 0.8991,
+      "step": 4116
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.1982010304927826,
+      "learning_rate": 0.0001420149048250488,
+      "loss": 0.9067,
+      "step": 4117
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.20188704133033752,
+      "learning_rate": 0.00014198977670050925,
+      "loss": 0.9074,
+      "step": 4118
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.19379226863384247,
+      "learning_rate": 0.00014196464535669718,
+      "loss": 0.9018,
+      "step": 4119
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.19640572369098663,
+      "learning_rate": 0.0001419395107955393,
+      "loss": 0.9235,
+      "step": 4120
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.19844560325145721,
+      "learning_rate": 0.0001419143730189626,
+      "loss": 0.9753,
+      "step": 4121
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.20532171428203583,
+      "learning_rate": 0.00014188923202889439,
+      "loss": 0.8703,
+      "step": 4122
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.2062155157327652,
+      "learning_rate": 0.00014186408782726218,
+      "loss": 0.8938,
+      "step": 4123
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.1967145800590515,
+      "learning_rate": 0.00014183894041599373,
+      "loss": 0.9188,
+      "step": 4124
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.19157996773719788,
+      "learning_rate": 0.000141813789797017,
+      "loss": 0.9115,
+      "step": 4125
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.1903439611196518,
+      "learning_rate": 0.00014178863597226027,
+      "loss": 0.8493,
+      "step": 4126
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.19151780009269714,
+      "learning_rate": 0.00014176347894365204,
+      "loss": 0.9128,
+      "step": 4127
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.20262552797794342,
+      "learning_rate": 0.00014173831871312103,
+      "loss": 0.923,
+      "step": 4128
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.20540249347686768,
+      "learning_rate": 0.00014171315528259623,
+      "loss": 0.9111,
+      "step": 4129
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.2055845409631729,
+      "learning_rate": 0.0001416879886540069,
+      "loss": 0.9231,
+      "step": 4130
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.19773642718791962,
+      "learning_rate": 0.00014166281882928245,
+      "loss": 0.9057,
+      "step": 4131
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.19996729493141174,
+      "learning_rate": 0.00014163764581035266,
+      "loss": 0.9112,
+      "step": 4132
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.21551403403282166,
+      "learning_rate": 0.00014161246959914744,
+      "loss": 0.9638,
+      "step": 4133
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.1965765506029129,
+      "learning_rate": 0.00014158729019759706,
+      "loss": 0.8685,
+      "step": 4134
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.20234429836273193,
+      "learning_rate": 0.00014156210760763198,
+      "loss": 0.94,
+      "step": 4135
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.19687169790267944,
+      "learning_rate": 0.00014153692183118283,
+      "loss": 0.9069,
+      "step": 4136
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.20247773826122284,
+      "learning_rate": 0.0001415117328701806,
+      "loss": 0.9672,
+      "step": 4137
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.19329603016376495,
+      "learning_rate": 0.00014148654072655646,
+      "loss": 0.8837,
+      "step": 4138
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.19726818799972534,
+      "learning_rate": 0.00014146134540224186,
+      "loss": 0.8691,
+      "step": 4139
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.19611580669879913,
+      "learning_rate": 0.00014143614689916847,
+      "loss": 0.9759,
+      "step": 4140
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.19384996592998505,
+      "learning_rate": 0.00014141094521926816,
+      "loss": 0.8821,
+      "step": 4141
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.20498308539390564,
+      "learning_rate": 0.00014138574036447318,
+      "loss": 0.933,
+      "step": 4142
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.20109273493289948,
+      "learning_rate": 0.00014136053233671587,
+      "loss": 0.9647,
+      "step": 4143
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.20127202570438385,
+      "learning_rate": 0.0001413353211379289,
+      "loss": 0.9003,
+      "step": 4144
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.1983189433813095,
+      "learning_rate": 0.0001413101067700452,
+      "loss": 0.8785,
+      "step": 4145
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.2038581520318985,
+      "learning_rate": 0.00014128488923499785,
+      "loss": 0.9107,
+      "step": 4146
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.1996109038591385,
+      "learning_rate": 0.00014125966853472026,
+      "loss": 0.8666,
+      "step": 4147
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.19804580509662628,
+      "learning_rate": 0.00014123444467114603,
+      "loss": 0.9559,
+      "step": 4148
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.1972654163837433,
+      "learning_rate": 0.000141209217646209,
+      "loss": 0.9336,
+      "step": 4149
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.1962662637233734,
+      "learning_rate": 0.00014118398746184338,
+      "loss": 0.8972,
+      "step": 4150
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.1934090405702591,
+      "learning_rate": 0.00014115875411998337,
+      "loss": 0.9185,
+      "step": 4151
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.19299843907356262,
+      "learning_rate": 0.00014113351762256367,
+      "loss": 0.8529,
+      "step": 4152
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.20182262361049652,
+      "learning_rate": 0.00014110827797151907,
+      "loss": 0.918,
+      "step": 4153
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.19320084154605865,
+      "learning_rate": 0.00014108303516878465,
+      "loss": 0.8898,
+      "step": 4154
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.20861463248729706,
+      "learning_rate": 0.00014105778921629576,
+      "loss": 0.9821,
+      "step": 4155
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.1959957629442215,
+      "learning_rate": 0.0001410325401159879,
+      "loss": 0.9332,
+      "step": 4156
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.20823851227760315,
+      "learning_rate": 0.0001410072878697969,
+      "loss": 0.9911,
+      "step": 4157
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.19281212985515594,
+      "learning_rate": 0.00014098203247965875,
+      "loss": 0.8473,
+      "step": 4158
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.20280300080776215,
+      "learning_rate": 0.0001409567739475098,
+      "loss": 0.9104,
+      "step": 4159
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.20401504635810852,
+      "learning_rate": 0.00014093151227528657,
+      "loss": 0.9004,
+      "step": 4160
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.19846060872077942,
+      "learning_rate": 0.00014090624746492577,
+      "loss": 0.9688,
+      "step": 4161
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.20426008105278015,
+      "learning_rate": 0.00014088097951836444,
+      "loss": 0.9435,
+      "step": 4162
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.201716348528862,
+      "learning_rate": 0.00014085570843753978,
+      "loss": 0.9067,
+      "step": 4163
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.19857504963874817,
+      "learning_rate": 0.00014083043422438935,
+      "loss": 0.8696,
+      "step": 4164
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.19757845997810364,
+      "learning_rate": 0.00014080515688085082,
+      "loss": 0.9303,
+      "step": 4165
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.19037018716335297,
+      "learning_rate": 0.00014077987640886215,
+      "loss": 0.8765,
+      "step": 4166
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.20469224452972412,
+      "learning_rate": 0.00014075459281036155,
+      "loss": 0.945,
+      "step": 4167
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.20371098816394806,
+      "learning_rate": 0.00014072930608728746,
+      "loss": 0.8623,
+      "step": 4168
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.19951358437538147,
+      "learning_rate": 0.0001407040162415786,
+      "loss": 0.9219,
+      "step": 4169
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.19785675406455994,
+      "learning_rate": 0.00014067872327517384,
+      "loss": 0.9279,
+      "step": 4170
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.19662170112133026,
+      "learning_rate": 0.00014065342719001236,
+      "loss": 0.8534,
+      "step": 4171
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.1945483237504959,
+      "learning_rate": 0.0001406281279880336,
+      "loss": 0.9185,
+      "step": 4172
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.2053336203098297,
+      "learning_rate": 0.00014060282567117716,
+      "loss": 0.9257,
+      "step": 4173
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.19595634937286377,
+      "learning_rate": 0.0001405775202413829,
+      "loss": 0.9356,
+      "step": 4174
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.19171977043151855,
+      "learning_rate": 0.00014055221170059096,
+      "loss": 0.8608,
+      "step": 4175
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.20737864077091217,
+      "learning_rate": 0.0001405269000507417,
+      "loss": 0.9602,
+      "step": 4176
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.19404630362987518,
+      "learning_rate": 0.00014050158529377573,
+      "loss": 0.812,
+      "step": 4177
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.20775002241134644,
+      "learning_rate": 0.00014047626743163383,
+      "loss": 0.9876,
+      "step": 4178
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.1968124359846115,
+      "learning_rate": 0.00014045094646625712,
+      "loss": 0.9084,
+      "step": 4179
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.20270562171936035,
+      "learning_rate": 0.00014042562239958688,
+      "loss": 0.9759,
+      "step": 4180
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.2031795233488083,
+      "learning_rate": 0.00014040029523356467,
+      "loss": 0.9155,
+      "step": 4181
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.20602475106716156,
+      "learning_rate": 0.0001403749649701323,
+      "loss": 0.9024,
+      "step": 4182
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.19497054815292358,
+      "learning_rate": 0.00014034963161123175,
+      "loss": 0.882,
+      "step": 4183
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.18592894077301025,
+      "learning_rate": 0.00014032429515880524,
+      "loss": 0.853,
+      "step": 4184
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.20189309120178223,
+      "learning_rate": 0.00014029895561479535,
+      "loss": 0.9096,
+      "step": 4185
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.2048352062702179,
+      "learning_rate": 0.00014027361298114474,
+      "loss": 0.8815,
+      "step": 4186
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.1962141990661621,
+      "learning_rate": 0.00014024826725979645,
+      "loss": 0.9115,
+      "step": 4187
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.1993790864944458,
+      "learning_rate": 0.00014022291845269363,
+      "loss": 0.9071,
+      "step": 4188
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.2057470977306366,
+      "learning_rate": 0.00014019756656177975,
+      "loss": 0.8741,
+      "step": 4189
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.19505277276039124,
+      "learning_rate": 0.0001401722115889985,
+      "loss": 0.8974,
+      "step": 4190
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.1985812783241272,
+      "learning_rate": 0.00014014685353629376,
+      "loss": 1.0093,
+      "step": 4191
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.19670027494430542,
+      "learning_rate": 0.0001401214924056097,
+      "loss": 0.9443,
+      "step": 4192
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.2052966207265854,
+      "learning_rate": 0.00014009612819889072,
+      "loss": 0.9262,
+      "step": 4193
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.2016931027173996,
+      "learning_rate": 0.0001400707609180814,
+      "loss": 0.9776,
+      "step": 4194
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.20231129229068756,
+      "learning_rate": 0.00014004539056512667,
+      "loss": 0.9499,
+      "step": 4195
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.1933603733778,
+      "learning_rate": 0.00014002001714197155,
+      "loss": 0.9471,
+      "step": 4196
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.1898379921913147,
+      "learning_rate": 0.0001399946406505614,
+      "loss": 0.8167,
+      "step": 4197
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.19617202877998352,
+      "learning_rate": 0.00013996926109284183,
+      "loss": 0.9083,
+      "step": 4198
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.1954466700553894,
+      "learning_rate": 0.00013994387847075855,
+      "loss": 0.8999,
+      "step": 4199
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.19861939549446106,
+      "learning_rate": 0.00013991849278625765,
+      "loss": 0.9274,
+      "step": 4200
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.1931011974811554,
+      "learning_rate": 0.00013989310404128542,
+      "loss": 0.9264,
+      "step": 4201
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.19421404600143433,
+      "learning_rate": 0.0001398677122377883,
+      "loss": 0.9099,
+      "step": 4202
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.2066340446472168,
+      "learning_rate": 0.0001398423173777131,
+      "loss": 0.9021,
+      "step": 4203
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.20380087196826935,
+      "learning_rate": 0.00013981691946300675,
+      "loss": 0.9444,
+      "step": 4204
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.2037912905216217,
+      "learning_rate": 0.00013979151849561648,
+      "loss": 0.9184,
+      "step": 4205
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.2029280811548233,
+      "learning_rate": 0.00013976611447748968,
+      "loss": 0.8991,
+      "step": 4206
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.18916313350200653,
+      "learning_rate": 0.0001397407074105741,
+      "loss": 0.9124,
+      "step": 4207
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.20596222579479218,
+      "learning_rate": 0.00013971529729681762,
+      "loss": 1.0275,
+      "step": 4208
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.22113633155822754,
+      "learning_rate": 0.00013968988413816833,
+      "loss": 1.0119,
+      "step": 4209
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.19565562903881073,
+      "learning_rate": 0.00013966446793657467,
+      "loss": 0.8888,
+      "step": 4210
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.20347142219543457,
+      "learning_rate": 0.00013963904869398528,
+      "loss": 0.9253,
+      "step": 4211
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.19310647249221802,
+      "learning_rate": 0.00013961362641234893,
+      "loss": 0.8565,
+      "step": 4212
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.20248988270759583,
+      "learning_rate": 0.0001395882010936147,
+      "loss": 0.9677,
+      "step": 4213
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.1965925544500351,
+      "learning_rate": 0.0001395627727397319,
+      "loss": 0.9319,
+      "step": 4214
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.21116456389427185,
+      "learning_rate": 0.00013953734135265013,
+      "loss": 0.9578,
+      "step": 4215
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.20068742334842682,
+      "learning_rate": 0.0001395119069343191,
+      "loss": 0.899,
+      "step": 4216
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.19809217751026154,
+      "learning_rate": 0.00013948646948668886,
+      "loss": 0.9072,
+      "step": 4217
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.20866192877292633,
+      "learning_rate": 0.00013946102901170963,
+      "loss": 0.8836,
+      "step": 4218
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.18966230750083923,
+      "learning_rate": 0.00013943558551133186,
+      "loss": 0.8439,
+      "step": 4219
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.20649796724319458,
+      "learning_rate": 0.00013941013898750631,
+      "loss": 0.9673,
+      "step": 4220
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.20510034263134003,
+      "learning_rate": 0.0001393846894421838,
+      "loss": 0.9216,
+      "step": 4221
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.2012736052274704,
+      "learning_rate": 0.00013935923687731567,
+      "loss": 0.8898,
+      "step": 4222
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.18844866752624512,
+      "learning_rate": 0.00013933378129485317,
+      "loss": 0.8726,
+      "step": 4223
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.19127140939235687,
+      "learning_rate": 0.00013930832269674797,
+      "loss": 0.8688,
+      "step": 4224
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.20248930156230927,
+      "learning_rate": 0.00013928286108495197,
+      "loss": 0.938,
+      "step": 4225
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.20166027545928955,
+      "learning_rate": 0.0001392573964614172,
+      "loss": 0.958,
+      "step": 4226
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.19940727949142456,
+      "learning_rate": 0.000139231928828096,
+      "loss": 0.8704,
+      "step": 4227
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.2015244960784912,
+      "learning_rate": 0.00013920645818694098,
+      "loss": 0.9304,
+      "step": 4228
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.1990206241607666,
+      "learning_rate": 0.00013918098453990485,
+      "loss": 0.9443,
+      "step": 4229
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.20932473242282867,
+      "learning_rate": 0.00013915550788894064,
+      "loss": 0.9485,
+      "step": 4230
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.19755159318447113,
+      "learning_rate": 0.00013913002823600162,
+      "loss": 0.9026,
+      "step": 4231
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.20039217174053192,
+      "learning_rate": 0.00013910454558304126,
+      "loss": 0.9063,
+      "step": 4232
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.2012191265821457,
+      "learning_rate": 0.00013907905993201328,
+      "loss": 0.9489,
+      "step": 4233
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.19405993819236755,
+      "learning_rate": 0.0001390535712848715,
+      "loss": 0.94,
+      "step": 4234
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.2001253366470337,
+      "learning_rate": 0.00013902807964357025,
+      "loss": 0.955,
+      "step": 4235
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.19616849720478058,
+      "learning_rate": 0.00013900258501006382,
+      "loss": 0.9086,
+      "step": 4236
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.18834726512432098,
+      "learning_rate": 0.00013897708738630686,
+      "loss": 0.8754,
+      "step": 4237
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.18973401188850403,
+      "learning_rate": 0.00013895158677425424,
+      "loss": 0.9142,
+      "step": 4238
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.19476298987865448,
+      "learning_rate": 0.00013892608317586097,
+      "loss": 0.829,
+      "step": 4239
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.1959441602230072,
+      "learning_rate": 0.00013890057659308246,
+      "loss": 0.9672,
+      "step": 4240
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.20653578639030457,
+      "learning_rate": 0.0001388750670278742,
+      "loss": 0.9431,
+      "step": 4241
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.20152495801448822,
+      "learning_rate": 0.00013884955448219196,
+      "loss": 0.9173,
+      "step": 4242
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.20955972373485565,
+      "learning_rate": 0.0001388240389579917,
+      "loss": 0.9552,
+      "step": 4243
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.1913870871067047,
+      "learning_rate": 0.00013879852045722968,
+      "loss": 0.9056,
+      "step": 4244
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.2016601711511612,
+      "learning_rate": 0.00013877299898186235,
+      "loss": 0.9518,
+      "step": 4245
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.20458324253559113,
+      "learning_rate": 0.00013874747453384643,
+      "loss": 0.933,
+      "step": 4246
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.20345601439476013,
+      "learning_rate": 0.00013872194711513877,
+      "loss": 0.937,
+      "step": 4247
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.19450397789478302,
+      "learning_rate": 0.00013869641672769654,
+      "loss": 0.9003,
+      "step": 4248
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.19730433821678162,
+      "learning_rate": 0.00013867088337347704,
+      "loss": 0.8861,
+      "step": 4249
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.21060039103031158,
+      "learning_rate": 0.00013864534705443796,
+      "loss": 1.0102,
+      "step": 4250
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.20124833285808563,
+      "learning_rate": 0.00013861980777253707,
+      "loss": 0.9119,
+      "step": 4251
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.1974477618932724,
+      "learning_rate": 0.0001385942655297324,
+      "loss": 0.8834,
+      "step": 4252
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.201642706990242,
+      "learning_rate": 0.00013856872032798226,
+      "loss": 0.9881,
+      "step": 4253
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.19658470153808594,
+      "learning_rate": 0.0001385431721692451,
+      "loss": 0.8564,
+      "step": 4254
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.2029627114534378,
+      "learning_rate": 0.0001385176210554797,
+      "loss": 0.9421,
+      "step": 4255
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.19619666039943695,
+      "learning_rate": 0.00013849206698864498,
+      "loss": 0.9308,
+      "step": 4256
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.21590463817119598,
+      "learning_rate": 0.00013846650997070012,
+      "loss": 0.9667,
+      "step": 4257
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.20509636402130127,
+      "learning_rate": 0.00013844095000360457,
+      "loss": 0.9068,
+      "step": 4258
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.2073296159505844,
+      "learning_rate": 0.0001384153870893179,
+      "loss": 0.9285,
+      "step": 4259
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.20427079498767853,
+      "learning_rate": 0.00013838982122979998,
+      "loss": 0.955,
+      "step": 4260
+    },
+    {
+      "epoch": 1.48,
+      "eval_loss": 0.9315346479415894,
+      "eval_runtime": 792.6138,
+      "eval_samples_per_second": 8.675,
+      "eval_steps_per_second": 4.338,
+      "step": 4260
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.2080182284116745,
+      "learning_rate": 0.0001383642524270109,
+      "loss": 0.9158,
+      "step": 4261
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.20025072991847992,
+      "learning_rate": 0.000138338680682911,
+      "loss": 0.9673,
+      "step": 4262
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.19327238202095032,
+      "learning_rate": 0.00013831310599946082,
+      "loss": 0.8705,
+      "step": 4263
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.20488376915454865,
+      "learning_rate": 0.00013828752837862103,
+      "loss": 0.9752,
+      "step": 4264
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.1918758749961853,
+      "learning_rate": 0.00013826194782235273,
+      "loss": 0.8837,
+      "step": 4265
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.19506444036960602,
+      "learning_rate": 0.00013823636433261705,
+      "loss": 0.9593,
+      "step": 4266
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.19961698353290558,
+      "learning_rate": 0.00013821077791137545,
+      "loss": 0.9217,
+      "step": 4267
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.1855212301015854,
+      "learning_rate": 0.00013818518856058962,
+      "loss": 0.9365,
+      "step": 4268
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.1939588487148285,
+      "learning_rate": 0.00013815959628222138,
+      "loss": 0.8809,
+      "step": 4269
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.19562089443206787,
+      "learning_rate": 0.0001381340010782329,
+      "loss": 0.9127,
+      "step": 4270
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.1901262402534485,
+      "learning_rate": 0.00013810840295058646,
+      "loss": 0.9086,
+      "step": 4271
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.1936071664094925,
+      "learning_rate": 0.00013808280190124466,
+      "loss": 0.9219,
+      "step": 4272
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.20040670037269592,
+      "learning_rate": 0.00013805719793217032,
+      "loss": 0.939,
+      "step": 4273
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.2018422931432724,
+      "learning_rate": 0.0001380315910453263,
+      "loss": 0.9056,
+      "step": 4274
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.18731442093849182,
+      "learning_rate": 0.000138005981242676,
+      "loss": 0.9318,
+      "step": 4275
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.19478319585323334,
+      "learning_rate": 0.00013798036852618275,
+      "loss": 0.8266,
+      "step": 4276
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.2014399617910385,
+      "learning_rate": 0.0001379547528978103,
+      "loss": 0.9502,
+      "step": 4277
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.19708126783370972,
+      "learning_rate": 0.00013792913435952252,
+      "loss": 0.8693,
+      "step": 4278
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.20559723675251007,
+      "learning_rate": 0.00013790351291328347,
+      "loss": 0.9499,
+      "step": 4279
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.19979485869407654,
+      "learning_rate": 0.0001378778885610576,
+      "loss": 0.8887,
+      "step": 4280
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.20127691328525543,
+      "learning_rate": 0.00013785226130480946,
+      "loss": 0.9415,
+      "step": 4281
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.20772793889045715,
+      "learning_rate": 0.00013782663114650379,
+      "loss": 0.9173,
+      "step": 4282
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.19286410510540009,
+      "learning_rate": 0.00013780099808810564,
+      "loss": 0.8666,
+      "step": 4283
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.2035040259361267,
+      "learning_rate": 0.00013777536213158023,
+      "loss": 0.9432,
+      "step": 4284
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.19898177683353424,
+      "learning_rate": 0.00013774972327889302,
+      "loss": 0.8985,
+      "step": 4285
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.1931130290031433,
+      "learning_rate": 0.0001377240815320097,
+      "loss": 0.9683,
+      "step": 4286
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.20073358714580536,
+      "learning_rate": 0.00013769843689289614,
+      "loss": 0.8938,
+      "step": 4287
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.20493336021900177,
+      "learning_rate": 0.00013767278936351854,
+      "loss": 0.9391,
+      "step": 4288
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.20389360189437866,
+      "learning_rate": 0.00013764713894584313,
+      "loss": 0.8701,
+      "step": 4289
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.20712493360042572,
+      "learning_rate": 0.0001376214856418366,
+      "loss": 0.9492,
+      "step": 4290
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.2005758136510849,
+      "learning_rate": 0.00013759582945346564,
+      "loss": 0.877,
+      "step": 4291
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.19810102880001068,
+      "learning_rate": 0.0001375701703826973,
+      "loss": 0.8873,
+      "step": 4292
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.19251015782356262,
+      "learning_rate": 0.00013754450843149883,
+      "loss": 0.8936,
+      "step": 4293
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.19246292114257812,
+      "learning_rate": 0.0001375188436018376,
+      "loss": 0.8776,
+      "step": 4294
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.20692507922649384,
+      "learning_rate": 0.00013749317589568138,
+      "loss": 0.9213,
+      "step": 4295
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.19895991683006287,
+      "learning_rate": 0.00013746750531499802,
+      "loss": 0.9143,
+      "step": 4296
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.19519945979118347,
+      "learning_rate": 0.0001374418318617556,
+      "loss": 0.9254,
+      "step": 4297
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.19164012372493744,
+      "learning_rate": 0.0001374161555379225,
+      "loss": 0.8961,
+      "step": 4298
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.1917676478624344,
+      "learning_rate": 0.00013739047634546729,
+      "loss": 0.8847,
+      "step": 4299
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.18594776093959808,
+      "learning_rate": 0.0001373647942863587,
+      "loss": 0.8765,
+      "step": 4300
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.1943291574716568,
+      "learning_rate": 0.00013733910936256568,
+      "loss": 0.8541,
+      "step": 4301
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.2048158198595047,
+      "learning_rate": 0.00013731342157605753,
+      "loss": 0.9455,
+      "step": 4302
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.19819948077201843,
+      "learning_rate": 0.00013728773092880364,
+      "loss": 0.8636,
+      "step": 4303
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.2028474062681198,
+      "learning_rate": 0.00013726203742277368,
+      "loss": 0.9195,
+      "step": 4304
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.20039261877536774,
+      "learning_rate": 0.0001372363410599375,
+      "loss": 0.8995,
+      "step": 4305
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.2012881338596344,
+      "learning_rate": 0.00013721064184226523,
+      "loss": 0.9096,
+      "step": 4306
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.2081228643655777,
+      "learning_rate": 0.0001371849397717271,
+      "loss": 0.958,
+      "step": 4307
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.19576063752174377,
+      "learning_rate": 0.00013715923485029365,
+      "loss": 0.9018,
+      "step": 4308
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.2039271742105484,
+      "learning_rate": 0.0001371335270799357,
+      "loss": 0.9569,
+      "step": 4309
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.1974891871213913,
+      "learning_rate": 0.00013710781646262418,
+      "loss": 0.929,
+      "step": 4310
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.19602631032466888,
+      "learning_rate": 0.00013708210300033023,
+      "loss": 0.9001,
+      "step": 4311
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.1993727684020996,
+      "learning_rate": 0.00013705638669502527,
+      "loss": 0.8957,
+      "step": 4312
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.20080606639385223,
+      "learning_rate": 0.00013703066754868095,
+      "loss": 0.896,
+      "step": 4313
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.20421402156352997,
+      "learning_rate": 0.0001370049455632691,
+      "loss": 0.9239,
+      "step": 4314
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.19175568222999573,
+      "learning_rate": 0.00013697922074076173,
+      "loss": 0.9108,
+      "step": 4315
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.2039186805486679,
+      "learning_rate": 0.00013695349308313114,
+      "loss": 0.8913,
+      "step": 4316
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.20334771275520325,
+      "learning_rate": 0.00013692776259234982,
+      "loss": 0.8841,
+      "step": 4317
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.20497700572013855,
+      "learning_rate": 0.00013690202927039047,
+      "loss": 0.8602,
+      "step": 4318
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.2132885605096817,
+      "learning_rate": 0.00013687629311922602,
+      "loss": 0.9771,
+      "step": 4319
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.20935708284378052,
+      "learning_rate": 0.00013685055414082962,
+      "loss": 0.9509,
+      "step": 4320
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.21456383168697357,
+      "learning_rate": 0.0001368248123371746,
+      "loss": 0.896,
+      "step": 4321
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.18961356580257416,
+      "learning_rate": 0.00013679906771023453,
+      "loss": 0.8618,
+      "step": 4322
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.20388853549957275,
+      "learning_rate": 0.0001367733202619832,
+      "loss": 0.9323,
+      "step": 4323
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.20104189217090607,
+      "learning_rate": 0.00013674756999439464,
+      "loss": 0.8609,
+      "step": 4324
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.20272791385650635,
+      "learning_rate": 0.00013672181690944308,
+      "loss": 0.9397,
+      "step": 4325
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.19907397031784058,
+      "learning_rate": 0.00013669606100910291,
+      "loss": 0.8918,
+      "step": 4326
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.2002435177564621,
+      "learning_rate": 0.00013667030229534883,
+      "loss": 0.9188,
+      "step": 4327
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.20999270677566528,
+      "learning_rate": 0.00013664454077015568,
+      "loss": 0.9373,
+      "step": 4328
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.19210919737815857,
+      "learning_rate": 0.00013661877643549858,
+      "loss": 0.9088,
+      "step": 4329
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.2012004852294922,
+      "learning_rate": 0.00013659300929335277,
+      "loss": 0.9251,
+      "step": 4330
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.1998959332704544,
+      "learning_rate": 0.00013656723934569383,
+      "loss": 0.8971,
+      "step": 4331
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.19553914666175842,
+      "learning_rate": 0.0001365414665944974,
+      "loss": 0.9171,
+      "step": 4332
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.19870030879974365,
+      "learning_rate": 0.00013651569104173954,
+      "loss": 0.9363,
+      "step": 4333
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.1972079575061798,
+      "learning_rate": 0.00013648991268939634,
+      "loss": 0.9027,
+      "step": 4334
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.20828427374362946,
+      "learning_rate": 0.00013646413153944423,
+      "loss": 0.9268,
+      "step": 4335
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.1969297230243683,
+      "learning_rate": 0.0001364383475938597,
+      "loss": 0.9146,
+      "step": 4336
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.20469069480895996,
+      "learning_rate": 0.00013641256085461967,
+      "loss": 0.9479,
+      "step": 4337
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.2146853655576706,
+      "learning_rate": 0.00013638677132370107,
+      "loss": 0.853,
+      "step": 4338
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.19474592804908752,
+      "learning_rate": 0.0001363609790030812,
+      "loss": 0.9035,
+      "step": 4339
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.19707852602005005,
+      "learning_rate": 0.00013633518389473742,
+      "loss": 0.9282,
+      "step": 4340
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.19231432676315308,
+      "learning_rate": 0.00013630938600064747,
+      "loss": 0.8703,
+      "step": 4341
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.21518941223621368,
+      "learning_rate": 0.0001362835853227892,
+      "loss": 0.9615,
+      "step": 4342
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.19725742936134338,
+      "learning_rate": 0.00013625778186314067,
+      "loss": 0.8598,
+      "step": 4343
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.2083866149187088,
+      "learning_rate": 0.0001362319756236802,
+      "loss": 0.9858,
+      "step": 4344
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.2003520131111145,
+      "learning_rate": 0.0001362061666063863,
+      "loss": 0.8966,
+      "step": 4345
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.20607808232307434,
+      "learning_rate": 0.00013618035481323775,
+      "loss": 0.9609,
+      "step": 4346
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.19025003910064697,
+      "learning_rate": 0.0001361545402462134,
+      "loss": 0.8294,
+      "step": 4347
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.1967662274837494,
+      "learning_rate": 0.0001361287229072924,
+      "loss": 0.938,
+      "step": 4348
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.20086213946342468,
+      "learning_rate": 0.00013610290279845422,
+      "loss": 0.9094,
+      "step": 4349
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.2035747617483139,
+      "learning_rate": 0.00013607707992167834,
+      "loss": 0.98,
+      "step": 4350
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.1947256624698639,
+      "learning_rate": 0.0001360512542789446,
+      "loss": 0.9259,
+      "step": 4351
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.19627153873443604,
+      "learning_rate": 0.00013602542587223295,
+      "loss": 0.8606,
+      "step": 4352
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.20166419446468353,
+      "learning_rate": 0.00013599959470352363,
+      "loss": 0.9869,
+      "step": 4353
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.2084067314863205,
+      "learning_rate": 0.0001359737607747971,
+      "loss": 0.89,
+      "step": 4354
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.20896011590957642,
+      "learning_rate": 0.0001359479240880339,
+      "loss": 0.8938,
+      "step": 4355
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.20715130865573883,
+      "learning_rate": 0.000135922084645215,
+      "loss": 0.9206,
+      "step": 4356
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.19523970782756805,
+      "learning_rate": 0.00013589624244832136,
+      "loss": 0.8097,
+      "step": 4357
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.19200685620307922,
+      "learning_rate": 0.00013587039749933432,
+      "loss": 0.8675,
+      "step": 4358
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.19039379060268402,
+      "learning_rate": 0.00013584454980023532,
+      "loss": 0.8412,
+      "step": 4359
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.19700483977794647,
+      "learning_rate": 0.00013581869935300604,
+      "loss": 0.8602,
+      "step": 4360
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.20600780844688416,
+      "learning_rate": 0.00013579284615962843,
+      "loss": 0.8716,
+      "step": 4361
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.19441702961921692,
+      "learning_rate": 0.00013576699022208453,
+      "loss": 0.859,
+      "step": 4362
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.20182082056999207,
+      "learning_rate": 0.00013574113154235675,
+      "loss": 0.92,
+      "step": 4363
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.20214153826236725,
+      "learning_rate": 0.00013571527012242758,
+      "loss": 0.8722,
+      "step": 4364
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.2132052481174469,
+      "learning_rate": 0.00013568940596427975,
+      "loss": 0.956,
+      "step": 4365
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.20397233963012695,
+      "learning_rate": 0.00013566353906989623,
+      "loss": 0.9429,
+      "step": 4366
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.19374212622642517,
+      "learning_rate": 0.0001356376694412602,
+      "loss": 0.9663,
+      "step": 4367
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.19922977685928345,
+      "learning_rate": 0.00013561179708035495,
+      "loss": 0.9123,
+      "step": 4368
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.20432357490062714,
+      "learning_rate": 0.0001355859219891642,
+      "loss": 0.9537,
+      "step": 4369
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.19571687281131744,
+      "learning_rate": 0.00013556004416967165,
+      "loss": 0.9415,
+      "step": 4370
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.20116746425628662,
+      "learning_rate": 0.00013553416362386132,
+      "loss": 0.9905,
+      "step": 4371
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.19328705966472626,
+      "learning_rate": 0.00013550828035371738,
+      "loss": 0.8699,
+      "step": 4372
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.20481045544147491,
+      "learning_rate": 0.00013548239436122434,
+      "loss": 0.9877,
+      "step": 4373
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.19554881751537323,
+      "learning_rate": 0.00013545650564836676,
+      "loss": 0.9079,
+      "step": 4374
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.19037027657032013,
+      "learning_rate": 0.0001354306142171295,
+      "loss": 0.8858,
+      "step": 4375
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.20761582255363464,
+      "learning_rate": 0.0001354047200694976,
+      "loss": 0.9034,
+      "step": 4376
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.2009163498878479,
+      "learning_rate": 0.00013537882320745627,
+      "loss": 0.9119,
+      "step": 4377
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.2040199190378189,
+      "learning_rate": 0.00013535292363299104,
+      "loss": 0.9102,
+      "step": 4378
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.1997002214193344,
+      "learning_rate": 0.00013532702134808755,
+      "loss": 0.9247,
+      "step": 4379
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.1887330412864685,
+      "learning_rate": 0.00013530111635473167,
+      "loss": 0.8357,
+      "step": 4380
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.20603513717651367,
+      "learning_rate": 0.0001352752086549095,
+      "loss": 0.8948,
+      "step": 4381
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.19191201031208038,
+      "learning_rate": 0.0001352492982506073,
+      "loss": 0.8744,
+      "step": 4382
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.20775219798088074,
+      "learning_rate": 0.0001352233851438116,
+      "loss": 0.962,
+      "step": 4383
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.187792107462883,
+      "learning_rate": 0.0001351974693365091,
+      "loss": 0.819,
+      "step": 4384
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.1982596069574356,
+      "learning_rate": 0.0001351715508306867,
+      "loss": 0.8621,
+      "step": 4385
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.20273452997207642,
+      "learning_rate": 0.00013514562962833158,
+      "loss": 0.8676,
+      "step": 4386
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.21657277643680573,
+      "learning_rate": 0.00013511970573143095,
+      "loss": 0.9464,
+      "step": 4387
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.2003757208585739,
+      "learning_rate": 0.00013509377914197246,
+      "loss": 0.9091,
+      "step": 4388
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.2024131715297699,
+      "learning_rate": 0.00013506784986194376,
+      "loss": 0.8762,
+      "step": 4389
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.19812408089637756,
+      "learning_rate": 0.00013504191789333284,
+      "loss": 0.9416,
+      "step": 4390
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.204255148768425,
+      "learning_rate": 0.00013501598323812793,
+      "loss": 0.9498,
+      "step": 4391
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.20179559290409088,
+      "learning_rate": 0.00013499004589831722,
+      "loss": 0.9185,
+      "step": 4392
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.1984333097934723,
+      "learning_rate": 0.00013496410587588939,
+      "loss": 0.9283,
+      "step": 4393
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.20066393911838531,
+      "learning_rate": 0.00013493816317283316,
+      "loss": 0.9169,
+      "step": 4394
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.20280592143535614,
+      "learning_rate": 0.00013491221779113755,
+      "loss": 0.9415,
+      "step": 4395
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.1986709088087082,
+      "learning_rate": 0.00013488626973279173,
+      "loss": 0.8959,
+      "step": 4396
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.19992804527282715,
+      "learning_rate": 0.00013486031899978505,
+      "loss": 0.9565,
+      "step": 4397
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.18855875730514526,
+      "learning_rate": 0.00013483436559410717,
+      "loss": 0.8672,
+      "step": 4398
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.19654472172260284,
+      "learning_rate": 0.0001348084095177478,
+      "loss": 0.9141,
+      "step": 4399
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.20703695714473724,
+      "learning_rate": 0.00013478245077269696,
+      "loss": 0.952,
+      "step": 4400
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.19787438213825226,
+      "learning_rate": 0.00013475648936094495,
+      "loss": 0.8901,
+      "step": 4401
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.19545020163059235,
+      "learning_rate": 0.00013473052528448201,
+      "loss": 0.8773,
+      "step": 4402
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.20343650877475739,
+      "learning_rate": 0.00013470455854529894,
+      "loss": 0.9165,
+      "step": 4403
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.20238184928894043,
+      "learning_rate": 0.00013467858914538642,
+      "loss": 0.9452,
+      "step": 4404
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.19688518345355988,
+      "learning_rate": 0.00013465261708673552,
+      "loss": 0.884,
+      "step": 4405
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.2027846872806549,
+      "learning_rate": 0.00013462664237133754,
+      "loss": 0.9246,
+      "step": 4406
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.19596917927265167,
+      "learning_rate": 0.00013460066500118372,
+      "loss": 0.9475,
+      "step": 4407
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.19403593242168427,
+      "learning_rate": 0.0001345746849782659,
+      "loss": 0.8755,
+      "step": 4408
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.20170363783836365,
+      "learning_rate": 0.0001345487023045758,
+      "loss": 0.8977,
+      "step": 4409
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.19403304159641266,
+      "learning_rate": 0.00013452271698210548,
+      "loss": 0.9049,
+      "step": 4410
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.2069365233182907,
+      "learning_rate": 0.00013449672901284722,
+      "loss": 0.9529,
+      "step": 4411
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.20516468584537506,
+      "learning_rate": 0.0001344707383987934,
+      "loss": 0.9236,
+      "step": 4412
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.19744600355625153,
+      "learning_rate": 0.0001344447451419367,
+      "loss": 0.9136,
+      "step": 4413
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.20329774916172028,
+      "learning_rate": 0.00013441874924427,
+      "loss": 0.8956,
+      "step": 4414
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.19856001436710358,
+      "learning_rate": 0.00013439275070778632,
+      "loss": 0.8937,
+      "step": 4415
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.1949041783809662,
+      "learning_rate": 0.00013436674953447894,
+      "loss": 0.9168,
+      "step": 4416
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.20832853019237518,
+      "learning_rate": 0.00013434074572634126,
+      "loss": 0.9482,
+      "step": 4417
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.21745702624320984,
+      "learning_rate": 0.00013431473928536702,
+      "loss": 0.9225,
+      "step": 4418
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.20856374502182007,
+      "learning_rate": 0.00013428873021355004,
+      "loss": 0.9394,
+      "step": 4419
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.19907131791114807,
+      "learning_rate": 0.0001342627185128844,
+      "loss": 0.9404,
+      "step": 4420
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.20088614523410797,
+      "learning_rate": 0.00013423670418536435,
+      "loss": 0.8747,
+      "step": 4421
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.19531351327896118,
+      "learning_rate": 0.00013421068723298434,
+      "loss": 0.9087,
+      "step": 4422
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.19716942310333252,
+      "learning_rate": 0.00013418466765773907,
+      "loss": 0.9053,
+      "step": 4423
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.2023046761751175,
+      "learning_rate": 0.0001341586454616234,
+      "loss": 0.917,
+      "step": 4424
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.2044750452041626,
+      "learning_rate": 0.0001341326206466324,
+      "loss": 0.9218,
+      "step": 4425
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.2067694514989853,
+      "learning_rate": 0.00013410659321476132,
+      "loss": 0.8862,
+      "step": 4426
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.1974387913942337,
+      "learning_rate": 0.00013408056316800568,
+      "loss": 0.8885,
+      "step": 4427
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.2019626647233963,
+      "learning_rate": 0.00013405453050836113,
+      "loss": 0.9814,
+      "step": 4428
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.20270082354545593,
+      "learning_rate": 0.00013402849523782353,
+      "loss": 0.8564,
+      "step": 4429
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.19102539122104645,
+      "learning_rate": 0.0001340024573583889,
+      "loss": 0.861,
+      "step": 4430
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.20177873969078064,
+      "learning_rate": 0.00013397641687205365,
+      "loss": 0.8803,
+      "step": 4431
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.2063576877117157,
+      "learning_rate": 0.0001339503737808141,
+      "loss": 0.9562,
+      "step": 4432
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.20372329652309418,
+      "learning_rate": 0.00013392432808666707,
+      "loss": 0.9515,
+      "step": 4433
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.2125784009695053,
+      "learning_rate": 0.00013389827979160929,
+      "loss": 0.9238,
+      "step": 4434
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.21021676063537598,
+      "learning_rate": 0.0001338722288976379,
+      "loss": 0.9715,
+      "step": 4435
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.2150159329175949,
+      "learning_rate": 0.0001338461754067502,
+      "loss": 0.9791,
+      "step": 4436
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.19582140445709229,
+      "learning_rate": 0.00013382011932094363,
+      "loss": 0.92,
+      "step": 4437
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.20285850763320923,
+      "learning_rate": 0.00013379406064221582,
+      "loss": 0.8859,
+      "step": 4438
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.1983582079410553,
+      "learning_rate": 0.00013376799937256468,
+      "loss": 0.9032,
+      "step": 4439
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.20245066285133362,
+      "learning_rate": 0.00013374193551398826,
+      "loss": 0.9366,
+      "step": 4440
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.20687803626060486,
+      "learning_rate": 0.00013371586906848486,
+      "loss": 0.9917,
+      "step": 4441
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.20650020241737366,
+      "learning_rate": 0.00013368980003805288,
+      "loss": 0.9583,
+      "step": 4442
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.20843304693698883,
+      "learning_rate": 0.00013366372842469105,
+      "loss": 0.9294,
+      "step": 4443
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.19253431260585785,
+      "learning_rate": 0.00013363765423039816,
+      "loss": 0.901,
+      "step": 4444
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.1946306675672531,
+      "learning_rate": 0.0001336115774571733,
+      "loss": 0.8535,
+      "step": 4445
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.2142277956008911,
+      "learning_rate": 0.00013358549810701573,
+      "loss": 0.8874,
+      "step": 4446
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.20530451834201813,
+      "learning_rate": 0.0001335594161819249,
+      "loss": 0.8352,
+      "step": 4447
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.1976638287305832,
+      "learning_rate": 0.00013353333168390045,
+      "loss": 0.9163,
+      "step": 4448
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.204092338681221,
+      "learning_rate": 0.00013350724461494223,
+      "loss": 0.9034,
+      "step": 4449
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.1945483237504959,
+      "learning_rate": 0.00013348115497705028,
+      "loss": 0.8528,
+      "step": 4450
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.19972464442253113,
+      "learning_rate": 0.00013345506277222485,
+      "loss": 0.852,
+      "step": 4451
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.20141078531742096,
+      "learning_rate": 0.00013342896800246638,
+      "loss": 0.9889,
+      "step": 4452
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.20517943799495697,
+      "learning_rate": 0.00013340287066977547,
+      "loss": 0.9725,
+      "step": 4453
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.2040693759918213,
+      "learning_rate": 0.00013337677077615302,
+      "loss": 0.9314,
+      "step": 4454
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.20373493432998657,
+      "learning_rate": 0.00013335066832359998,
+      "loss": 0.9459,
+      "step": 4455
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.19994668662548065,
+      "learning_rate": 0.00013332456331411761,
+      "loss": 0.8899,
+      "step": 4456
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.2057134211063385,
+      "learning_rate": 0.00013329845574970732,
+      "loss": 0.8571,
+      "step": 4457
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.20253203809261322,
+      "learning_rate": 0.00013327234563237074,
+      "loss": 0.8994,
+      "step": 4458
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.20478413999080658,
+      "learning_rate": 0.0001332462329641097,
+      "loss": 0.9676,
+      "step": 4459
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.20461991429328918,
+      "learning_rate": 0.00013322011774692614,
+      "loss": 0.9049,
+      "step": 4460
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.20310884714126587,
+      "learning_rate": 0.0001331939999828223,
+      "loss": 0.8697,
+      "step": 4461
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.20332156121730804,
+      "learning_rate": 0.0001331678796738006,
+      "loss": 0.9044,
+      "step": 4462
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.19390572607517242,
+      "learning_rate": 0.0001331417568218636,
+      "loss": 0.9324,
+      "step": 4463
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.19995063543319702,
+      "learning_rate": 0.00013311563142901408,
+      "loss": 0.8561,
+      "step": 4464
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.1999330073595047,
+      "learning_rate": 0.00013308950349725504,
+      "loss": 0.9186,
+      "step": 4465
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.19406409561634064,
+      "learning_rate": 0.00013306337302858968,
+      "loss": 0.8674,
+      "step": 4466
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.2086877077817917,
+      "learning_rate": 0.0001330372400250213,
+      "loss": 0.9126,
+      "step": 4467
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.2143045961856842,
+      "learning_rate": 0.00013301110448855357,
+      "loss": 0.9841,
+      "step": 4468
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.19664418697357178,
+      "learning_rate": 0.00013298496642119014,
+      "loss": 0.9627,
+      "step": 4469
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.19305755198001862,
+      "learning_rate": 0.00013295882582493502,
+      "loss": 0.9068,
+      "step": 4470
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.20256315171718597,
+      "learning_rate": 0.0001329326827017924,
+      "loss": 0.8876,
+      "step": 4471
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.2027614861726761,
+      "learning_rate": 0.0001329065370537665,
+      "loss": 0.8925,
+      "step": 4472
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.1951225847005844,
+      "learning_rate": 0.000132880388882862,
+      "loss": 0.9562,
+      "step": 4473
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.19473542273044586,
+      "learning_rate": 0.0001328542381910835,
+      "loss": 0.9228,
+      "step": 4474
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.1945401281118393,
+      "learning_rate": 0.00013282808498043597,
+      "loss": 0.9131,
+      "step": 4475
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.18757790327072144,
+      "learning_rate": 0.00013280192925292457,
+      "loss": 0.8966,
+      "step": 4476
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.19668155908584595,
+      "learning_rate": 0.00013277577101055454,
+      "loss": 0.8772,
+      "step": 4477
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.1970047950744629,
+      "learning_rate": 0.00013274961025533144,
+      "loss": 0.8969,
+      "step": 4478
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.19303329288959503,
+      "learning_rate": 0.0001327234469892609,
+      "loss": 0.863,
+      "step": 4479
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.1973162442445755,
+      "learning_rate": 0.00013269728121434882,
+      "loss": 0.961,
+      "step": 4480
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.20247642695903778,
+      "learning_rate": 0.00013267111293260134,
+      "loss": 0.8567,
+      "step": 4481
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.1964680701494217,
+      "learning_rate": 0.00013264494214602467,
+      "loss": 0.8572,
+      "step": 4482
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.20020325481891632,
+      "learning_rate": 0.0001326187688566253,
+      "loss": 0.9614,
+      "step": 4483
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.20274128019809723,
+      "learning_rate": 0.00013259259306640987,
+      "loss": 0.8822,
+      "step": 4484
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.1965228021144867,
+      "learning_rate": 0.0001325664147773852,
+      "loss": 0.9174,
+      "step": 4485
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.20517116785049438,
+      "learning_rate": 0.00013254023399155836,
+      "loss": 0.9295,
+      "step": 4486
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.20203058421611786,
+      "learning_rate": 0.00013251405071093662,
+      "loss": 0.8963,
+      "step": 4487
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.202811598777771,
+      "learning_rate": 0.00013248786493752732,
+      "loss": 0.9262,
+      "step": 4488
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.19520260393619537,
+      "learning_rate": 0.00013246167667333808,
+      "loss": 0.9205,
+      "step": 4489
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.2029237449169159,
+      "learning_rate": 0.00013243548592037675,
+      "loss": 0.9363,
+      "step": 4490
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.2026432752609253,
+      "learning_rate": 0.0001324092926806513,
+      "loss": 0.8972,
+      "step": 4491
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.2118445485830307,
+      "learning_rate": 0.0001323830969561699,
+      "loss": 0.9456,
+      "step": 4492
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.1965012401342392,
+      "learning_rate": 0.00013235689874894097,
+      "loss": 0.8599,
+      "step": 4493
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.20173734426498413,
+      "learning_rate": 0.00013233069806097301,
+      "loss": 0.8679,
+      "step": 4494
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.1969406008720398,
+      "learning_rate": 0.0001323044948942748,
+      "loss": 0.8983,
+      "step": 4495
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.2201402634382248,
+      "learning_rate": 0.00013227828925085533,
+      "loss": 0.8905,
+      "step": 4496
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.19682130217552185,
+      "learning_rate": 0.00013225208113272366,
+      "loss": 0.8263,
+      "step": 4497
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.20680737495422363,
+      "learning_rate": 0.00013222587054188917,
+      "loss": 0.9453,
+      "step": 4498
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.2154068499803543,
+      "learning_rate": 0.00013219965748036133,
+      "loss": 0.915,
+      "step": 4499
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.20928502082824707,
+      "learning_rate": 0.00013217344195014987,
+      "loss": 0.9181,
+      "step": 4500
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.20148850977420807,
+      "learning_rate": 0.0001321472239532647,
+      "loss": 0.9762,
+      "step": 4501
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.21044816076755524,
+      "learning_rate": 0.00013212100349171589,
+      "loss": 0.9026,
+      "step": 4502
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.19358424842357635,
+      "learning_rate": 0.00013209478056751367,
+      "loss": 0.8614,
+      "step": 4503
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.204145610332489,
+      "learning_rate": 0.00013206855518266855,
+      "loss": 0.8295,
+      "step": 4504
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.21280977129936218,
+      "learning_rate": 0.00013204232733919112,
+      "loss": 0.9205,
+      "step": 4505
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.2101668119430542,
+      "learning_rate": 0.00013201609703909227,
+      "loss": 0.9607,
+      "step": 4506
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.2028149664402008,
+      "learning_rate": 0.00013198986428438305,
+      "loss": 0.8992,
+      "step": 4507
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.20130696892738342,
+      "learning_rate": 0.0001319636290770746,
+      "loss": 0.8599,
+      "step": 4508
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.20510746538639069,
+      "learning_rate": 0.00013193739141917837,
+      "loss": 0.949,
+      "step": 4509
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.19777636229991913,
+      "learning_rate": 0.0001319111513127059,
+      "loss": 0.9049,
+      "step": 4510
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.2046719193458557,
+      "learning_rate": 0.00013188490875966902,
+      "loss": 0.9144,
+      "step": 4511
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.2182987481355667,
+      "learning_rate": 0.0001318586637620797,
+      "loss": 0.8728,
+      "step": 4512
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.21205998957157135,
+      "learning_rate": 0.00013183241632195,
+      "loss": 0.9209,
+      "step": 4513
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.1928020566701889,
+      "learning_rate": 0.0001318061664412924,
+      "loss": 0.8787,
+      "step": 4514
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.20081821084022522,
+      "learning_rate": 0.0001317799141221193,
+      "loss": 0.9282,
+      "step": 4515
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.1871049404144287,
+      "learning_rate": 0.0001317536593664435,
+      "loss": 0.8818,
+      "step": 4516
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.2123117744922638,
+      "learning_rate": 0.00013172740217627784,
+      "loss": 0.9548,
+      "step": 4517
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.19874443113803864,
+      "learning_rate": 0.00013170114255363542,
+      "loss": 0.8925,
+      "step": 4518
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.20230533182621002,
+      "learning_rate": 0.00013167488050052953,
+      "loss": 0.8894,
+      "step": 4519
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.2105279266834259,
+      "learning_rate": 0.00013164861601897362,
+      "loss": 0.9251,
+      "step": 4520
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.20260430872440338,
+      "learning_rate": 0.00013162234911098136,
+      "loss": 1.0058,
+      "step": 4521
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.1892397105693817,
+      "learning_rate": 0.00013159607977856656,
+      "loss": 0.8249,
+      "step": 4522
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.19726528227329254,
+      "learning_rate": 0.0001315698080237432,
+      "loss": 0.8644,
+      "step": 4523
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.21568602323532104,
+      "learning_rate": 0.00013154353384852558,
+      "loss": 0.9634,
+      "step": 4524
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.19832541048526764,
+      "learning_rate": 0.00013151725725492797,
+      "loss": 0.9232,
+      "step": 4525
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.194269597530365,
+      "learning_rate": 0.00013149097824496502,
+      "loss": 0.8723,
+      "step": 4526
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.19430376589298248,
+      "learning_rate": 0.00013146469682065147,
+      "loss": 0.9318,
+      "step": 4527
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.2051064819097519,
+      "learning_rate": 0.00013143841298400227,
+      "loss": 0.897,
+      "step": 4528
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.19880448281764984,
+      "learning_rate": 0.00013141212673703257,
+      "loss": 0.9256,
+      "step": 4529
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.22948618233203888,
+      "learning_rate": 0.00013138583808175761,
+      "loss": 0.8805,
+      "step": 4530
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.2091957926750183,
+      "learning_rate": 0.00013135954702019298,
+      "loss": 0.9641,
+      "step": 4531
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.1984574943780899,
+      "learning_rate": 0.00013133325355435432,
+      "loss": 0.8683,
+      "step": 4532
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.19852031767368317,
+      "learning_rate": 0.00013130695768625748,
+      "loss": 0.835,
+      "step": 4533
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.19653603434562683,
+      "learning_rate": 0.00013128065941791857,
+      "loss": 0.9282,
+      "step": 4534
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.19558171927928925,
+      "learning_rate": 0.0001312543587513537,
+      "loss": 0.8546,
+      "step": 4535
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.1992642730474472,
+      "learning_rate": 0.00013122805568857948,
+      "loss": 0.9177,
+      "step": 4536
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.2031051516532898,
+      "learning_rate": 0.00013120175023161235,
+      "loss": 0.9602,
+      "step": 4537
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.21117272973060608,
+      "learning_rate": 0.00013117544238246917,
+      "loss": 0.9495,
+      "step": 4538
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.189724400639534,
+      "learning_rate": 0.00013114913214316693,
+      "loss": 0.847,
+      "step": 4539
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.1954822987318039,
+      "learning_rate": 0.0001311228195157227,
+      "loss": 0.8963,
+      "step": 4540
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.18917202949523926,
+      "learning_rate": 0.00013109650450215393,
+      "loss": 0.8647,
+      "step": 4541
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.20269404351711273,
+      "learning_rate": 0.00013107018710447802,
+      "loss": 0.9066,
+      "step": 4542
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.20098453760147095,
+      "learning_rate": 0.00013104386732471274,
+      "loss": 0.9232,
+      "step": 4543
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.18985962867736816,
+      "learning_rate": 0.00013101754516487603,
+      "loss": 0.8402,
+      "step": 4544
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.20924219489097595,
+      "learning_rate": 0.00013099122062698582,
+      "loss": 0.9598,
+      "step": 4545
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.19651223719120026,
+      "learning_rate": 0.00013096489371306047,
+      "loss": 0.832,
+      "step": 4546
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.20140282809734344,
+      "learning_rate": 0.00013093856442511836,
+      "loss": 0.9027,
+      "step": 4547
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.1940830647945404,
+      "learning_rate": 0.0001309122327651781,
+      "loss": 0.8725,
+      "step": 4548
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.19329622387886047,
+      "learning_rate": 0.00013088589873525857,
+      "loss": 0.9091,
+      "step": 4549
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.19192615151405334,
+      "learning_rate": 0.00013085956233737864,
+      "loss": 0.882,
+      "step": 4550
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.20170480012893677,
+      "learning_rate": 0.0001308332235735575,
+      "loss": 0.9323,
+      "step": 4551
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.201574444770813,
+      "learning_rate": 0.00013080688244581452,
+      "loss": 0.95,
+      "step": 4552
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.21443568170070648,
+      "learning_rate": 0.0001307805389561692,
+      "loss": 0.8848,
+      "step": 4553
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.19022703170776367,
+      "learning_rate": 0.00013075419310664123,
+      "loss": 0.8998,
+      "step": 4554
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.1944078505039215,
+      "learning_rate": 0.00013072784489925055,
+      "loss": 0.9451,
+      "step": 4555
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.1990111768245697,
+      "learning_rate": 0.00013070149433601718,
+      "loss": 0.9322,
+      "step": 4556
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.20842678844928741,
+      "learning_rate": 0.00013067514141896136,
+      "loss": 0.9692,
+      "step": 4557
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.20555783808231354,
+      "learning_rate": 0.00013064878615010348,
+      "loss": 0.9217,
+      "step": 4558
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.19351287186145782,
+      "learning_rate": 0.00013062242853146427,
+      "loss": 0.9378,
+      "step": 4559
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.20051833987236023,
+      "learning_rate": 0.00013059606856506443,
+      "loss": 0.9638,
+      "step": 4560
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.19287043809890747,
+      "learning_rate": 0.00013056970625292488,
+      "loss": 0.8593,
+      "step": 4561
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.1942070871591568,
+      "learning_rate": 0.00013054334159706688,
+      "loss": 0.8507,
+      "step": 4562
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.1981026977300644,
+      "learning_rate": 0.00013051697459951167,
+      "loss": 0.8782,
+      "step": 4563
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.19574685394763947,
+      "learning_rate": 0.00013049060526228078,
+      "loss": 0.9031,
+      "step": 4564
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.20009981095790863,
+      "learning_rate": 0.00013046423358739593,
+      "loss": 0.9434,
+      "step": 4565
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.19632475078105927,
+      "learning_rate": 0.00013043785957687887,
+      "loss": 0.8625,
+      "step": 4566
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.20003314316272736,
+      "learning_rate": 0.0001304114832327518,
+      "loss": 0.9462,
+      "step": 4567
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.19301815330982208,
+      "learning_rate": 0.00013038510455703684,
+      "loss": 0.8618,
+      "step": 4568
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.22005648910999298,
+      "learning_rate": 0.0001303587235517564,
+      "loss": 1.0248,
+      "step": 4569
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.20169512927532196,
+      "learning_rate": 0.00013033234021893312,
+      "loss": 0.8955,
+      "step": 4570
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.21043767035007477,
+      "learning_rate": 0.00013030595456058966,
+      "loss": 0.9648,
+      "step": 4571
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.21289391815662384,
+      "learning_rate": 0.00013027956657874905,
+      "loss": 0.918,
+      "step": 4572
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.21469812095165253,
+      "learning_rate": 0.00013025317627543431,
+      "loss": 0.9514,
+      "step": 4573
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.19922086596488953,
+      "learning_rate": 0.00013022678365266885,
+      "loss": 0.997,
+      "step": 4574
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.1971648931503296,
+      "learning_rate": 0.00013020038871247607,
+      "loss": 0.9485,
+      "step": 4575
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.20105613768100739,
+      "learning_rate": 0.00013017399145687957,
+      "loss": 0.9337,
+      "step": 4576
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.1998489499092102,
+      "learning_rate": 0.00013014759188790325,
+      "loss": 0.9294,
+      "step": 4577
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.20058594644069672,
+      "learning_rate": 0.0001301211900075711,
+      "loss": 0.8733,
+      "step": 4578
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.19787292182445526,
+      "learning_rate": 0.00013009478581790725,
+      "loss": 0.8558,
+      "step": 4579
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.1924135535955429,
+      "learning_rate": 0.00013006837932093615,
+      "loss": 0.9001,
+      "step": 4580
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.19671164453029633,
+      "learning_rate": 0.00013004197051868225,
+      "loss": 0.9114,
+      "step": 4581
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.2027590572834015,
+      "learning_rate": 0.0001300155594131703,
+      "loss": 0.9188,
+      "step": 4582
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.20843097567558289,
+      "learning_rate": 0.00012998914600642516,
+      "loss": 0.9254,
+      "step": 4583
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.2043987512588501,
+      "learning_rate": 0.0001299627303004719,
+      "loss": 0.8734,
+      "step": 4584
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.21097159385681152,
+      "learning_rate": 0.00012993631229733582,
+      "loss": 0.9372,
+      "step": 4585
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.20122608542442322,
+      "learning_rate": 0.00012990989199904225,
+      "loss": 0.8941,
+      "step": 4586
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.19901883602142334,
+      "learning_rate": 0.00012988346940761682,
+      "loss": 0.925,
+      "step": 4587
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.2017354816198349,
+      "learning_rate": 0.00012985704452508527,
+      "loss": 0.8339,
+      "step": 4588
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.1998375803232193,
+      "learning_rate": 0.0001298306173534736,
+      "loss": 0.9294,
+      "step": 4589
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.21041107177734375,
+      "learning_rate": 0.00012980418789480791,
+      "loss": 0.9718,
+      "step": 4590
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.19458279013633728,
+      "learning_rate": 0.00012977775615111443,
+      "loss": 0.8662,
+      "step": 4591
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.20359832048416138,
+      "learning_rate": 0.00012975132212441972,
+      "loss": 0.8277,
+      "step": 4592
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.20850984752178192,
+      "learning_rate": 0.00012972488581675033,
+      "loss": 0.9371,
+      "step": 4593
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.20215162634849548,
+      "learning_rate": 0.00012969844723013317,
+      "loss": 0.874,
+      "step": 4594
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.20475077629089355,
+      "learning_rate": 0.0001296720063665952,
+      "loss": 0.921,
+      "step": 4595
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.20587964355945587,
+      "learning_rate": 0.00012964556322816358,
+      "loss": 0.9558,
+      "step": 4596
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.21276992559432983,
+      "learning_rate": 0.00012961911781686567,
+      "loss": 0.9088,
+      "step": 4597
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.19974417984485626,
+      "learning_rate": 0.00012959267013472892,
+      "loss": 0.869,
+      "step": 4598
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.20184177160263062,
+      "learning_rate": 0.00012956622018378116,
+      "loss": 0.8675,
+      "step": 4599
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.21369704604148865,
+      "learning_rate": 0.00012953976796605012,
+      "loss": 0.9403,
+      "step": 4600
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.1977105587720871,
+      "learning_rate": 0.0001295133134835639,
+      "loss": 0.9233,
+      "step": 4601
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.20390766859054565,
+      "learning_rate": 0.00012948685673835067,
+      "loss": 0.9024,
+      "step": 4602
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.19591738283634186,
+      "learning_rate": 0.00012946039773243889,
+      "loss": 0.8169,
+      "step": 4603
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.19412824511528015,
+      "learning_rate": 0.00012943393646785704,
+      "loss": 0.8583,
+      "step": 4604
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.20760369300842285,
+      "learning_rate": 0.0001294074729466339,
+      "loss": 0.895,
+      "step": 4605
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.20479688048362732,
+      "learning_rate": 0.00012938100717079838,
+      "loss": 0.8874,
+      "step": 4606
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.2013799101114273,
+      "learning_rate": 0.00012935453914237954,
+      "loss": 0.9256,
+      "step": 4607
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.21338863670825958,
+      "learning_rate": 0.00012932806886340666,
+      "loss": 0.9692,
+      "step": 4608
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.19070522487163544,
+      "learning_rate": 0.00012930159633590908,
+      "loss": 0.8283,
+      "step": 4609
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.20960764586925507,
+      "learning_rate": 0.00012927512156191647,
+      "loss": 0.9784,
+      "step": 4610
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.20272405445575714,
+      "learning_rate": 0.00012924864454345858,
+      "loss": 0.8618,
+      "step": 4611
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.2035459280014038,
+      "learning_rate": 0.00012922216528256538,
+      "loss": 0.916,
+      "step": 4612
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.20177391171455383,
+      "learning_rate": 0.00012919568378126693,
+      "loss": 0.9787,
+      "step": 4613
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.19269515573978424,
+      "learning_rate": 0.00012916920004159356,
+      "loss": 0.8297,
+      "step": 4614
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.19745002686977386,
+      "learning_rate": 0.0001291427140655757,
+      "loss": 0.9053,
+      "step": 4615
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.19630128145217896,
+      "learning_rate": 0.00012911622585524396,
+      "loss": 0.8849,
+      "step": 4616
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.19954022765159607,
+      "learning_rate": 0.0001290897354126292,
+      "loss": 0.8558,
+      "step": 4617
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.20994316041469574,
+      "learning_rate": 0.00012906324273976233,
+      "loss": 0.9551,
+      "step": 4618
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.19849319756031036,
+      "learning_rate": 0.00012903674783867447,
+      "loss": 0.8924,
+      "step": 4619
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.21105317771434784,
+      "learning_rate": 0.000129010250711397,
+      "loss": 0.9563,
+      "step": 4620
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.20199191570281982,
+      "learning_rate": 0.00012898375135996136,
+      "loss": 0.8566,
+      "step": 4621
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.20917585492134094,
+      "learning_rate": 0.00012895724978639924,
+      "loss": 0.8869,
+      "step": 4622
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.22021600604057312,
+      "learning_rate": 0.0001289307459927424,
+      "loss": 0.9356,
+      "step": 4623
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.20073702931404114,
+      "learning_rate": 0.0001289042399810229,
+      "loss": 0.9461,
+      "step": 4624
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.20655374228954315,
+      "learning_rate": 0.00012887773175327286,
+      "loss": 0.8657,
+      "step": 4625
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.21302717924118042,
+      "learning_rate": 0.00012885122131152467,
+      "loss": 0.9399,
+      "step": 4626
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.20100539922714233,
+      "learning_rate": 0.00012882470865781075,
+      "loss": 0.8738,
+      "step": 4627
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.19878286123275757,
+      "learning_rate": 0.00012879819379416383,
+      "loss": 0.8997,
+      "step": 4628
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.21100857853889465,
+      "learning_rate": 0.0001287716767226167,
+      "loss": 0.9444,
+      "step": 4629
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.18998339772224426,
+      "learning_rate": 0.00012874515744520243,
+      "loss": 0.8804,
+      "step": 4630
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.2040288895368576,
+      "learning_rate": 0.00012871863596395418,
+      "loss": 0.9086,
+      "step": 4631
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.20322178304195404,
+      "learning_rate": 0.0001286921122809053,
+      "loss": 0.9374,
+      "step": 4632
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.19871999323368073,
+      "learning_rate": 0.0001286655863980893,
+      "loss": 0.9024,
+      "step": 4633
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.2054329365491867,
+      "learning_rate": 0.00012863905831753984,
+      "loss": 0.9025,
+      "step": 4634
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.19915160536766052,
+      "learning_rate": 0.00012861252804129082,
+      "loss": 0.8937,
+      "step": 4635
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.19782203435897827,
+      "learning_rate": 0.00012858599557137628,
+      "loss": 0.8921,
+      "step": 4636
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.1984853893518448,
+      "learning_rate": 0.00012855946090983034,
+      "loss": 0.8927,
+      "step": 4637
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.2001384049654007,
+      "learning_rate": 0.00012853292405868742,
+      "loss": 0.9125,
+      "step": 4638
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.19436684250831604,
+      "learning_rate": 0.000128506385019982,
+      "loss": 0.8602,
+      "step": 4639
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.2030458152294159,
+      "learning_rate": 0.00012847984379574882,
+      "loss": 0.9206,
+      "step": 4640
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.1936401128768921,
+      "learning_rate": 0.0001284533003880227,
+      "loss": 0.8341,
+      "step": 4641
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.18999110162258148,
+      "learning_rate": 0.00012842675479883875,
+      "loss": 0.8246,
+      "step": 4642
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.20278599858283997,
+      "learning_rate": 0.00012840020703023207,
+      "loss": 0.8715,
+      "step": 4643
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.19653545320034027,
+      "learning_rate": 0.00012837365708423808,
+      "loss": 0.8786,
+      "step": 4644
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.2093815803527832,
+      "learning_rate": 0.00012834710496289228,
+      "loss": 0.9366,
+      "step": 4645
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.20715850591659546,
+      "learning_rate": 0.00012832055066823038,
+      "loss": 0.9033,
+      "step": 4646
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.1966526210308075,
+      "learning_rate": 0.0001282939942022883,
+      "loss": 0.9521,
+      "step": 4647
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.19843120872974396,
+      "learning_rate": 0.00012826743556710202,
+      "loss": 0.9032,
+      "step": 4648
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.20358197391033173,
+      "learning_rate": 0.0001282408747647077,
+      "loss": 0.9334,
+      "step": 4649
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.21139803528785706,
+      "learning_rate": 0.00012821431179714176,
+      "loss": 0.9502,
+      "step": 4650
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.1972806453704834,
+      "learning_rate": 0.0001281877466664407,
+      "loss": 0.9055,
+      "step": 4651
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.19850580394268036,
+      "learning_rate": 0.00012816117937464126,
+      "loss": 0.907,
+      "step": 4652
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.20600315928459167,
+      "learning_rate": 0.00012813460992378028,
+      "loss": 0.8872,
+      "step": 4653
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.20602796971797943,
+      "learning_rate": 0.00012810803831589474,
+      "loss": 0.872,
+      "step": 4654
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.20422983169555664,
+      "learning_rate": 0.0001280814645530219,
+      "loss": 0.9397,
+      "step": 4655
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.2100948840379715,
+      "learning_rate": 0.00012805488863719907,
+      "loss": 0.9563,
+      "step": 4656
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.19479598104953766,
+      "learning_rate": 0.00012802831057046377,
+      "loss": 0.855,
+      "step": 4657
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.20453958213329315,
+      "learning_rate": 0.00012800173035485375,
+      "loss": 0.9296,
+      "step": 4658
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.21561020612716675,
+      "learning_rate": 0.0001279751479924068,
+      "loss": 0.9152,
+      "step": 4659
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.21022316813468933,
+      "learning_rate": 0.00012794856348516095,
+      "loss": 0.8859,
+      "step": 4660
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.1952332854270935,
+      "learning_rate": 0.0001279219768351544,
+      "loss": 0.9103,
+      "step": 4661
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.19976872205734253,
+      "learning_rate": 0.0001278953880444255,
+      "loss": 0.7599,
+      "step": 4662
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.20693926513195038,
+      "learning_rate": 0.00012786879711501273,
+      "loss": 0.9499,
+      "step": 4663
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.21582749485969543,
+      "learning_rate": 0.00012784220404895476,
+      "loss": 0.9489,
+      "step": 4664
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.18999259173870087,
+      "learning_rate": 0.00012781560884829045,
+      "loss": 0.9087,
+      "step": 4665
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.2033642679452896,
+      "learning_rate": 0.0001277890115150588,
+      "loss": 0.8906,
+      "step": 4666
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.2041936218738556,
+      "learning_rate": 0.00012776241205129897,
+      "loss": 0.8471,
+      "step": 4667
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.2005212903022766,
+      "learning_rate": 0.0001277358104590503,
+      "loss": 0.912,
+      "step": 4668
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.2013813853263855,
+      "learning_rate": 0.00012770920674035226,
+      "loss": 0.9598,
+      "step": 4669
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.19489338994026184,
+      "learning_rate": 0.00012768260089724448,
+      "loss": 0.8349,
+      "step": 4670
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.20248544216156006,
+      "learning_rate": 0.00012765599293176687,
+      "loss": 0.9462,
+      "step": 4671
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.20362122356891632,
+      "learning_rate": 0.00012762938284595931,
+      "loss": 0.9654,
+      "step": 4672
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.19850721955299377,
+      "learning_rate": 0.000127602770641862,
+      "loss": 0.9181,
+      "step": 4673
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.21032880246639252,
+      "learning_rate": 0.0001275761563215152,
+      "loss": 0.9282,
+      "step": 4674
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.20525328814983368,
+      "learning_rate": 0.00012754953988695943,
+      "loss": 0.9266,
+      "step": 4675
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.21305762231349945,
+      "learning_rate": 0.00012752292134023528,
+      "loss": 0.9671,
+      "step": 4676
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.19181224703788757,
+      "learning_rate": 0.00012749630068338357,
+      "loss": 0.8691,
+      "step": 4677
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.19330871105194092,
+      "learning_rate": 0.00012746967791844524,
+      "loss": 0.8276,
+      "step": 4678
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.20834501087665558,
+      "learning_rate": 0.00012744305304746137,
+      "loss": 0.9028,
+      "step": 4679
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.19693732261657715,
+      "learning_rate": 0.00012741642607247327,
+      "loss": 0.8798,
+      "step": 4680
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.1976938545703888,
+      "learning_rate": 0.0001273897969955224,
+      "loss": 0.9087,
+      "step": 4681
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.19912458956241608,
+      "learning_rate": 0.0001273631658186503,
+      "loss": 0.9223,
+      "step": 4682
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.20554779469966888,
+      "learning_rate": 0.00012733653254389885,
+      "loss": 0.9097,
+      "step": 4683
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.20539045333862305,
+      "learning_rate": 0.00012730989717330978,
+      "loss": 0.9172,
+      "step": 4684
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.20110385119915009,
+      "learning_rate": 0.0001272832597089253,
+      "loss": 0.9048,
+      "step": 4685
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.20027123391628265,
+      "learning_rate": 0.00012725662015278765,
+      "loss": 0.9018,
+      "step": 4686
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.19668515026569366,
+      "learning_rate": 0.00012722997850693916,
+      "loss": 0.912,
+      "step": 4687
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.19159843027591705,
+      "learning_rate": 0.0001272033347734225,
+      "loss": 0.8726,
+      "step": 4688
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.19470220804214478,
+      "learning_rate": 0.00012717668895428027,
+      "loss": 0.8884,
+      "step": 4689
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.20279020071029663,
+      "learning_rate": 0.00012715004105155544,
+      "loss": 0.857,
+      "step": 4690
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.20406362414360046,
+      "learning_rate": 0.000127123391067291,
+      "loss": 0.8551,
+      "step": 4691
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.20731474459171295,
+      "learning_rate": 0.0001270967390035302,
+      "loss": 0.8923,
+      "step": 4692
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.2141023874282837,
+      "learning_rate": 0.0001270700848623164,
+      "loss": 0.9492,
+      "step": 4693
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.19725339114665985,
+      "learning_rate": 0.00012704342864569302,
+      "loss": 0.8739,
+      "step": 4694
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.2122267633676529,
+      "learning_rate": 0.00012701677035570387,
+      "loss": 0.976,
+      "step": 4695
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.20237216353416443,
+      "learning_rate": 0.00012699010999439275,
+      "loss": 0.9657,
+      "step": 4696
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.19359779357910156,
+      "learning_rate": 0.00012696344756380357,
+      "loss": 0.8157,
+      "step": 4697
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.2037573903799057,
+      "learning_rate": 0.00012693678306598062,
+      "loss": 0.9134,
+      "step": 4698
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.20230227708816528,
+      "learning_rate": 0.00012691011650296812,
+      "loss": 0.9193,
+      "step": 4699
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.1983824223279953,
+      "learning_rate": 0.00012688344787681056,
+      "loss": 0.8738,
+      "step": 4700
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.2047380656003952,
+      "learning_rate": 0.0001268567771895526,
+      "loss": 0.9393,
+      "step": 4701
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.20342564582824707,
+      "learning_rate": 0.00012683010444323897,
+      "loss": 0.9246,
+      "step": 4702
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.19755631685256958,
+      "learning_rate": 0.00012680342963991472,
+      "loss": 0.9372,
+      "step": 4703
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.19999569654464722,
+      "learning_rate": 0.00012677675278162484,
+      "loss": 0.9368,
+      "step": 4704
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.2041405439376831,
+      "learning_rate": 0.00012675007387041466,
+      "loss": 0.934,
+      "step": 4705
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.1956358402967453,
+      "learning_rate": 0.00012672339290832957,
+      "loss": 0.8872,
+      "step": 4706
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.2013956606388092,
+      "learning_rate": 0.00012669670989741517,
+      "loss": 0.9227,
+      "step": 4707
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.2008161097764969,
+      "learning_rate": 0.0001266700248397172,
+      "loss": 0.9199,
+      "step": 4708
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.2024879902601242,
+      "learning_rate": 0.0001266433377372815,
+      "loss": 0.894,
+      "step": 4709
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.19811367988586426,
+      "learning_rate": 0.00012661664859215413,
+      "loss": 0.9339,
+      "step": 4710
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.2045515477657318,
+      "learning_rate": 0.00012658995740638136,
+      "loss": 0.9124,
+      "step": 4711
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.2044452279806137,
+      "learning_rate": 0.00012656326418200947,
+      "loss": 0.9922,
+      "step": 4712
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.2022848129272461,
+      "learning_rate": 0.00012653656892108503,
+      "loss": 0.947,
+      "step": 4713
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.1936623454093933,
+      "learning_rate": 0.00012650987162565465,
+      "loss": 0.8892,
+      "step": 4714
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.2036278247833252,
+      "learning_rate": 0.00012648317229776524,
+      "loss": 0.922,
+      "step": 4715
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.20058897137641907,
+      "learning_rate": 0.00012645647093946372,
+      "loss": 0.8689,
+      "step": 4716
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.20195390284061432,
+      "learning_rate": 0.00012642976755279723,
+      "loss": 0.9431,
+      "step": 4717
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.20222632586956024,
+      "learning_rate": 0.00012640306213981315,
+      "loss": 0.9048,
+      "step": 4718
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.204377681016922,
+      "learning_rate": 0.00012637635470255879,
+      "loss": 0.9154,
+      "step": 4719
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.19898901879787445,
+      "learning_rate": 0.0001263496452430819,
+      "loss": 0.8935,
+      "step": 4720
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.19688498973846436,
+      "learning_rate": 0.00012632293376343016,
+      "loss": 0.8986,
+      "step": 4721
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.19992433488368988,
+      "learning_rate": 0.00012629622026565147,
+      "loss": 0.8598,
+      "step": 4722
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.2034882754087448,
+      "learning_rate": 0.00012626950475179397,
+      "loss": 0.8845,
+      "step": 4723
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.2034643441438675,
+      "learning_rate": 0.00012624278722390584,
+      "loss": 0.9494,
+      "step": 4724
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.20310690999031067,
+      "learning_rate": 0.00012621606768403547,
+      "loss": 0.9539,
+      "step": 4725
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.20767931640148163,
+      "learning_rate": 0.00012618934613423138,
+      "loss": 0.9233,
+      "step": 4726
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.19426491856575012,
+      "learning_rate": 0.00012616262257654224,
+      "loss": 0.9371,
+      "step": 4727
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.19253315031528473,
+      "learning_rate": 0.000126135897013017,
+      "loss": 0.8632,
+      "step": 4728
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.195959210395813,
+      "learning_rate": 0.00012610916944570452,
+      "loss": 0.8928,
+      "step": 4729
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.2083369642496109,
+      "learning_rate": 0.00012608243987665403,
+      "loss": 0.9695,
+      "step": 4730
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.1992987096309662,
+      "learning_rate": 0.0001260557083079148,
+      "loss": 0.9216,
+      "step": 4731
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.2010459452867508,
+      "learning_rate": 0.00012602897474153627,
+      "loss": 0.8297,
+      "step": 4732
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.20180568099021912,
+      "learning_rate": 0.0001260022391795681,
+      "loss": 0.9472,
+      "step": 4733
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.1999882608652115,
+      "learning_rate": 0.00012597550162406004,
+      "loss": 0.898,
+      "step": 4734
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.2005031853914261,
+      "learning_rate": 0.00012594876207706199,
+      "loss": 0.9134,
+      "step": 4735
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.194036066532135,
+      "learning_rate": 0.00012592202054062402,
+      "loss": 0.9065,
+      "step": 4736
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.21351303160190582,
+      "learning_rate": 0.0001258952770167963,
+      "loss": 0.8464,
+      "step": 4737
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.19651855528354645,
+      "learning_rate": 0.0001258685315076293,
+      "loss": 0.907,
+      "step": 4738
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.20761534571647644,
+      "learning_rate": 0.0001258417840151735,
+      "loss": 0.9521,
+      "step": 4739
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.21457886695861816,
+      "learning_rate": 0.00012581503454147958,
+      "loss": 0.9144,
+      "step": 4740
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.21407045423984528,
+      "learning_rate": 0.00012578828308859835,
+      "loss": 0.9467,
+      "step": 4741
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.1977054923772812,
+      "learning_rate": 0.00012576152965858078,
+      "loss": 0.9275,
+      "step": 4742
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.2006564736366272,
+      "learning_rate": 0.00012573477425347808,
+      "loss": 0.9641,
+      "step": 4743
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.19844360649585724,
+      "learning_rate": 0.00012570801687534147,
+      "loss": 0.9349,
+      "step": 4744
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.214335635304451,
+      "learning_rate": 0.00012568125752622237,
+      "loss": 0.9298,
+      "step": 4745
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.19964855909347534,
+      "learning_rate": 0.0001256544962081724,
+      "loss": 0.8417,
+      "step": 4746
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.1935940831899643,
+      "learning_rate": 0.0001256277329232433,
+      "loss": 0.8531,
+      "step": 4747
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.2027629017829895,
+      "learning_rate": 0.0001256009676734869,
+      "loss": 0.9267,
+      "step": 4748
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.1888791173696518,
+      "learning_rate": 0.00012557420046095533,
+      "loss": 0.8679,
+      "step": 4749
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.20005565881729126,
+      "learning_rate": 0.00012554743128770073,
+      "loss": 0.9021,
+      "step": 4750
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.19851699471473694,
+      "learning_rate": 0.00012552066015577543,
+      "loss": 0.8643,
+      "step": 4751
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.19369323551654816,
+      "learning_rate": 0.00012549388706723194,
+      "loss": 0.855,
+      "step": 4752
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.19861680269241333,
+      "learning_rate": 0.00012546711202412287,
+      "loss": 0.8797,
+      "step": 4753
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.19953753054141998,
+      "learning_rate": 0.00012544033502850108,
+      "loss": 0.9289,
+      "step": 4754
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.20581848919391632,
+      "learning_rate": 0.00012541355608241942,
+      "loss": 0.945,
+      "step": 4755
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.19809095561504364,
+      "learning_rate": 0.00012538677518793103,
+      "loss": 0.8942,
+      "step": 4756
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.2005758285522461,
+      "learning_rate": 0.00012535999234708912,
+      "loss": 0.9201,
+      "step": 4757
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.2029479295015335,
+      "learning_rate": 0.0001253332075619471,
+      "loss": 0.8957,
+      "step": 4758
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.199596107006073,
+      "learning_rate": 0.0001253064208345585,
+      "loss": 0.9275,
+      "step": 4759
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.19497540593147278,
+      "learning_rate": 0.00012527963216697703,
+      "loss": 0.8849,
+      "step": 4760
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.1980920284986496,
+      "learning_rate": 0.00012525284156125649,
+      "loss": 0.9224,
+      "step": 4761
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.20496365427970886,
+      "learning_rate": 0.00012522604901945084,
+      "loss": 0.9739,
+      "step": 4762
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.20327137410640717,
+      "learning_rate": 0.00012519925454361426,
+      "loss": 0.8848,
+      "step": 4763
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.20324772596359253,
+      "learning_rate": 0.000125172458135801,
+      "loss": 0.8823,
+      "step": 4764
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.2028888463973999,
+      "learning_rate": 0.00012514565979806553,
+      "loss": 0.9404,
+      "step": 4765
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.20583532750606537,
+      "learning_rate": 0.00012511885953246237,
+      "loss": 0.9499,
+      "step": 4766
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.18950171768665314,
+      "learning_rate": 0.00012509205734104623,
+      "loss": 0.8812,
+      "step": 4767
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.20389115810394287,
+      "learning_rate": 0.00012506525322587207,
+      "loss": 0.9047,
+      "step": 4768
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.20218288898468018,
+      "learning_rate": 0.00012503844718899485,
+      "loss": 0.8811,
+      "step": 4769
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.19585788249969482,
+      "learning_rate": 0.0001250116392324697,
+      "loss": 0.8725,
+      "step": 4770
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.20046821236610413,
+      "learning_rate": 0.000124984829358352,
+      "loss": 0.9135,
+      "step": 4771
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.20927603542804718,
+      "learning_rate": 0.0001249580175686972,
+      "loss": 0.9032,
+      "step": 4772
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.19529473781585693,
+      "learning_rate": 0.00012493120386556085,
+      "loss": 0.8774,
+      "step": 4773
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.20796741545200348,
+      "learning_rate": 0.00012490438825099877,
+      "loss": 0.8518,
+      "step": 4774
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.20350944995880127,
+      "learning_rate": 0.00012487757072706683,
+      "loss": 0.9255,
+      "step": 4775
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.21421313285827637,
+      "learning_rate": 0.00012485075129582107,
+      "loss": 1.0059,
+      "step": 4776
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.2068627029657364,
+      "learning_rate": 0.00012482392995931768,
+      "loss": 0.889,
+      "step": 4777
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.20098544657230377,
+      "learning_rate": 0.00012479710671961302,
+      "loss": 0.9105,
+      "step": 4778
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.2001083940267563,
+      "learning_rate": 0.00012477028157876355,
+      "loss": 0.9113,
+      "step": 4779
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.20307984948158264,
+      "learning_rate": 0.00012474345453882595,
+      "loss": 0.8946,
+      "step": 4780
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.20198144018650055,
+      "learning_rate": 0.00012471662560185694,
+      "loss": 0.9234,
+      "step": 4781
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.20412342250347137,
+      "learning_rate": 0.00012468979476991344,
+      "loss": 0.9353,
+      "step": 4782
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.21251653134822845,
+      "learning_rate": 0.00012466296204505256,
+      "loss": 0.8688,
+      "step": 4783
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.20212705433368683,
+      "learning_rate": 0.00012463612742933148,
+      "loss": 0.9185,
+      "step": 4784
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.19823376834392548,
+      "learning_rate": 0.00012460929092480757,
+      "loss": 0.8651,
+      "step": 4785
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.1942797750234604,
+      "learning_rate": 0.00012458245253353832,
+      "loss": 0.8739,
+      "step": 4786
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.2100387066602707,
+      "learning_rate": 0.0001245556122575814,
+      "loss": 0.8959,
+      "step": 4787
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.21781910955905914,
+      "learning_rate": 0.00012452877009899455,
+      "loss": 0.9808,
+      "step": 4788
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.2030804306268692,
+      "learning_rate": 0.00012450192605983578,
+      "loss": 0.909,
+      "step": 4789
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.20184268057346344,
+      "learning_rate": 0.0001244750801421631,
+      "loss": 0.9259,
+      "step": 4790
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.21130618453025818,
+      "learning_rate": 0.00012444823234803476,
+      "loss": 0.8928,
+      "step": 4791
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.19274461269378662,
+      "learning_rate": 0.00012442138267950911,
+      "loss": 0.887,
+      "step": 4792
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.21150325238704681,
+      "learning_rate": 0.0001243945311386447,
+      "loss": 0.966,
+      "step": 4793
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.19882428646087646,
+      "learning_rate": 0.00012436767772750016,
+      "loss": 0.8655,
+      "step": 4794
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.2035803347826004,
+      "learning_rate": 0.0001243408224481343,
+      "loss": 0.8424,
+      "step": 4795
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.20144344866275787,
+      "learning_rate": 0.00012431396530260605,
+      "loss": 0.8622,
+      "step": 4796
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.21372032165527344,
+      "learning_rate": 0.00012428710629297445,
+      "loss": 0.8885,
+      "step": 4797
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.21561861038208008,
+      "learning_rate": 0.0001242602454212988,
+      "loss": 0.9356,
+      "step": 4798
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.20541751384735107,
+      "learning_rate": 0.00012423338268963846,
+      "loss": 0.9053,
+      "step": 4799
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.1994425356388092,
+      "learning_rate": 0.0001242065181000529,
+      "loss": 0.9568,
+      "step": 4800
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.20360726118087769,
+      "learning_rate": 0.00012417965165460179,
+      "loss": 0.8055,
+      "step": 4801
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.21201220154762268,
+      "learning_rate": 0.00012415278335534496,
+      "loss": 0.9122,
+      "step": 4802
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.19801384210586548,
+      "learning_rate": 0.00012412591320434227,
+      "loss": 0.8405,
+      "step": 4803
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.21161265671253204,
+      "learning_rate": 0.00012409904120365392,
+      "loss": 0.9093,
+      "step": 4804
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.1941586285829544,
+      "learning_rate": 0.00012407216735534004,
+      "loss": 0.8723,
+      "step": 4805
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.1975100338459015,
+      "learning_rate": 0.00012404529166146102,
+      "loss": 0.9366,
+      "step": 4806
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.20188577473163605,
+      "learning_rate": 0.00012401841412407733,
+      "loss": 0.8629,
+      "step": 4807
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.1966887265443802,
+      "learning_rate": 0.00012399153474524968,
+      "loss": 0.8388,
+      "step": 4808
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.20200707018375397,
+      "learning_rate": 0.00012396465352703886,
+      "loss": 0.8711,
+      "step": 4809
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.20064467191696167,
+      "learning_rate": 0.00012393777047150573,
+      "loss": 0.9107,
+      "step": 4810
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.21187013387680054,
+      "learning_rate": 0.00012391088558071146,
+      "loss": 0.922,
+      "step": 4811
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.20777884125709534,
+      "learning_rate": 0.00012388399885671715,
+      "loss": 0.9048,
+      "step": 4812
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.20996583998203278,
+      "learning_rate": 0.00012385711030158422,
+      "loss": 0.9906,
+      "step": 4813
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.20055098831653595,
+      "learning_rate": 0.00012383021991737415,
+      "loss": 0.8662,
+      "step": 4814
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.21267199516296387,
+      "learning_rate": 0.00012380332770614856,
+      "loss": 0.894,
+      "step": 4815
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.20670823752880096,
+      "learning_rate": 0.00012377643366996927,
+      "loss": 0.9398,
+      "step": 4816
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.20194852352142334,
+      "learning_rate": 0.00012374953781089812,
+      "loss": 0.9119,
+      "step": 4817
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.19411972165107727,
+      "learning_rate": 0.0001237226401309972,
+      "loss": 0.8739,
+      "step": 4818
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.2021835595369339,
+      "learning_rate": 0.0001236957406323287,
+      "loss": 0.8944,
+      "step": 4819
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.20480941236019135,
+      "learning_rate": 0.00012366883931695492,
+      "loss": 0.9131,
+      "step": 4820
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.19868041574954987,
+      "learning_rate": 0.0001236419361869384,
+      "loss": 0.9164,
+      "step": 4821
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.19789153337478638,
+      "learning_rate": 0.00012361503124434168,
+      "loss": 0.8702,
+      "step": 4822
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.18909890949726105,
+      "learning_rate": 0.00012358812449122754,
+      "loss": 0.8115,
+      "step": 4823
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.19557319581508636,
+      "learning_rate": 0.00012356121592965887,
+      "loss": 0.9356,
+      "step": 4824
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.20103345811367035,
+      "learning_rate": 0.0001235343055616987,
+      "loss": 0.9251,
+      "step": 4825
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.1910592019557953,
+      "learning_rate": 0.00012350739338941018,
+      "loss": 0.8771,
+      "step": 4826
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.1996619999408722,
+      "learning_rate": 0.00012348047941485659,
+      "loss": 0.8778,
+      "step": 4827
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.20036165416240692,
+      "learning_rate": 0.00012345356364010143,
+      "loss": 0.9669,
+      "step": 4828
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.19795265793800354,
+      "learning_rate": 0.00012342664606720822,
+      "loss": 0.8803,
+      "step": 4829
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.2031397819519043,
+      "learning_rate": 0.0001233997266982407,
+      "loss": 0.919,
+      "step": 4830
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.1937347948551178,
+      "learning_rate": 0.00012337280553526277,
+      "loss": 0.8887,
+      "step": 4831
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.20833584666252136,
+      "learning_rate": 0.00012334588258033834,
+      "loss": 0.9119,
+      "step": 4832
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.1977413147687912,
+      "learning_rate": 0.0001233189578355316,
+      "loss": 0.7891,
+      "step": 4833
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.19594435393810272,
+      "learning_rate": 0.0001232920313029068,
+      "loss": 0.9201,
+      "step": 4834
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.20837093889713287,
+      "learning_rate": 0.0001232651029845283,
+      "loss": 0.9205,
+      "step": 4835
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.2012907862663269,
+      "learning_rate": 0.00012323817288246072,
+      "loss": 0.8571,
+      "step": 4836
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.20674099028110504,
+      "learning_rate": 0.00012321124099876865,
+      "loss": 0.9526,
+      "step": 4837
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.19660276174545288,
+      "learning_rate": 0.00012318430733551699,
+      "loss": 0.8947,
+      "step": 4838
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.20618146657943726,
+      "learning_rate": 0.00012315737189477063,
+      "loss": 0.9915,
+      "step": 4839
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.20119090378284454,
+      "learning_rate": 0.00012313043467859468,
+      "loss": 0.9783,
+      "step": 4840
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.19183722138404846,
+      "learning_rate": 0.0001231034956890544,
+      "loss": 0.8596,
+      "step": 4841
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.19631201028823853,
+      "learning_rate": 0.00012307655492821507,
+      "loss": 0.8865,
+      "step": 4842
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.2002319097518921,
+      "learning_rate": 0.0001230496123981422,
+      "loss": 0.8704,
+      "step": 4843
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.2014852911233902,
+      "learning_rate": 0.00012302266810090148,
+      "loss": 0.9073,
+      "step": 4844
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.20717839896678925,
+      "learning_rate": 0.00012299572203855862,
+      "loss": 0.885,
+      "step": 4845
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.1985516995191574,
+      "learning_rate": 0.0001229687742131796,
+      "loss": 0.9247,
+      "step": 4846
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.20868374407291412,
+      "learning_rate": 0.0001229418246268303,
+      "loss": 0.9499,
+      "step": 4847
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.20093078911304474,
+      "learning_rate": 0.00012291487328157705,
+      "loss": 0.8834,
+      "step": 4848
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.2085324078798294,
+      "learning_rate": 0.00012288792017948607,
+      "loss": 1.0308,
+      "step": 4849
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.20289358496665955,
+      "learning_rate": 0.0001228609653226238,
+      "loss": 0.855,
+      "step": 4850
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.205068439245224,
+      "learning_rate": 0.00012283400871305694,
+      "loss": 1.0064,
+      "step": 4851
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.20595556497573853,
+      "learning_rate": 0.000122807050352852,
+      "loss": 0.9158,
+      "step": 4852
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.20259448885917664,
+      "learning_rate": 0.00012278009024407592,
+      "loss": 0.8504,
+      "step": 4853
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.20223982632160187,
+      "learning_rate": 0.00012275312838879574,
+      "loss": 0.942,
+      "step": 4854
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.20645855367183685,
+      "learning_rate": 0.00012272616478907846,
+      "loss": 0.917,
+      "step": 4855
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.20709437131881714,
+      "learning_rate": 0.0001226991994469914,
+      "loss": 0.8653,
+      "step": 4856
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.19745272397994995,
+      "learning_rate": 0.00012267223236460192,
+      "loss": 0.8475,
+      "step": 4857
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.21651792526245117,
+      "learning_rate": 0.00012264526354397755,
+      "loss": 0.9116,
+      "step": 4858
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.20984449982643127,
+      "learning_rate": 0.00012261829298718587,
+      "loss": 0.9508,
+      "step": 4859
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.22017496824264526,
+      "learning_rate": 0.0001225913206962947,
+      "loss": 1.0125,
+      "step": 4860
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.19838780164718628,
+      "learning_rate": 0.000122564346673372,
+      "loss": 0.8886,
+      "step": 4861
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.2012847363948822,
+      "learning_rate": 0.00012253737092048577,
+      "loss": 0.9365,
+      "step": 4862
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.20742902159690857,
+      "learning_rate": 0.00012251039343970417,
+      "loss": 0.9472,
+      "step": 4863
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.204187273979187,
+      "learning_rate": 0.0001224834142330955,
+      "loss": 0.8915,
+      "step": 4864
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.20868054032325745,
+      "learning_rate": 0.00012245643330272826,
+      "loss": 0.8737,
+      "step": 4865
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.20117759704589844,
+      "learning_rate": 0.00012242945065067098,
+      "loss": 0.8567,
+      "step": 4866
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.19919933378696442,
+      "learning_rate": 0.00012240246627899238,
+      "loss": 0.9152,
+      "step": 4867
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.2044895738363266,
+      "learning_rate": 0.00012237548018976133,
+      "loss": 0.9284,
+      "step": 4868
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.21247272193431854,
+      "learning_rate": 0.00012234849238504673,
+      "loss": 0.9213,
+      "step": 4869
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.2041437327861786,
+      "learning_rate": 0.00012232150286691773,
+      "loss": 0.9402,
+      "step": 4870
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.206780806183815,
+      "learning_rate": 0.00012229451163744354,
+      "loss": 0.8597,
+      "step": 4871
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.20794974267482758,
+      "learning_rate": 0.00012226751869869355,
+      "loss": 0.9023,
+      "step": 4872
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.20286689698696136,
+      "learning_rate": 0.00012224052405273724,
+      "loss": 0.8945,
+      "step": 4873
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.19486016035079956,
+      "learning_rate": 0.00012221352770164424,
+      "loss": 0.8933,
+      "step": 4874
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.19991089403629303,
+      "learning_rate": 0.00012218652964748428,
+      "loss": 0.9064,
+      "step": 4875
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.20874367654323578,
+      "learning_rate": 0.00012215952989232728,
+      "loss": 0.9028,
+      "step": 4876
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.21600410342216492,
+      "learning_rate": 0.00012213252843824325,
+      "loss": 0.9277,
+      "step": 4877
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.20367151498794556,
+      "learning_rate": 0.00012210552528730235,
+      "loss": 0.8838,
+      "step": 4878
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.18601082265377045,
+      "learning_rate": 0.00012207852044157484,
+      "loss": 0.8004,
+      "step": 4879
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.2015962302684784,
+      "learning_rate": 0.0001220515139031311,
+      "loss": 0.9,
+      "step": 4880
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.20441995561122894,
+      "learning_rate": 0.00012202450567404171,
+      "loss": 0.8845,
+      "step": 4881
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.20740507543087006,
+      "learning_rate": 0.00012199749575637733,
+      "loss": 0.8861,
+      "step": 4882
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.210323765873909,
+      "learning_rate": 0.00012197048415220875,
+      "loss": 0.9043,
+      "step": 4883
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.20224831998348236,
+      "learning_rate": 0.00012194347086360692,
+      "loss": 0.8717,
+      "step": 4884
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.20253652334213257,
+      "learning_rate": 0.00012191645589264283,
+      "loss": 0.9113,
+      "step": 4885
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.19361186027526855,
+      "learning_rate": 0.00012188943924138771,
+      "loss": 0.8605,
+      "step": 4886
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.1945461481809616,
+      "learning_rate": 0.00012186242091191292,
+      "loss": 0.8891,
+      "step": 4887
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.2082328498363495,
+      "learning_rate": 0.0001218354009062898,
+      "loss": 0.9408,
+      "step": 4888
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.20235101878643036,
+      "learning_rate": 0.00012180837922659001,
+      "loss": 0.8895,
+      "step": 4889
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.20408551394939423,
+      "learning_rate": 0.00012178135587488515,
+      "loss": 0.8531,
+      "step": 4890
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.20237864553928375,
+      "learning_rate": 0.00012175433085324714,
+      "loss": 0.9257,
+      "step": 4891
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.2030421644449234,
+      "learning_rate": 0.00012172730416374791,
+      "loss": 0.9628,
+      "step": 4892
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.19940567016601562,
+      "learning_rate": 0.00012170027580845953,
+      "loss": 0.8989,
+      "step": 4893
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.19683948159217834,
+      "learning_rate": 0.00012167324578945423,
+      "loss": 0.9434,
+      "step": 4894
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.21224354207515717,
+      "learning_rate": 0.00012164621410880427,
+      "loss": 0.9947,
+      "step": 4895
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.20414133369922638,
+      "learning_rate": 0.00012161918076858223,
+      "loss": 0.9006,
+      "step": 4896
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.20747117698192596,
+      "learning_rate": 0.00012159214577086062,
+      "loss": 0.908,
+      "step": 4897
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.19901777803897858,
+      "learning_rate": 0.00012156510911771221,
+      "loss": 0.9242,
+      "step": 4898
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.21684972941875458,
+      "learning_rate": 0.00012153807081120983,
+      "loss": 1.0294,
+      "step": 4899
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.20291352272033691,
+      "learning_rate": 0.0001215110308534264,
+      "loss": 0.8793,
+      "step": 4900
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.1894405335187912,
+      "learning_rate": 0.0001214839892464351,
+      "loss": 0.8511,
+      "step": 4901
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.2027515321969986,
+      "learning_rate": 0.00012145694599230913,
+      "loss": 0.9226,
+      "step": 4902
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.20232322812080383,
+      "learning_rate": 0.00012142990109312182,
+      "loss": 0.8925,
+      "step": 4903
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.19063344597816467,
+      "learning_rate": 0.00012140285455094669,
+      "loss": 0.8289,
+      "step": 4904
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.20592787861824036,
+      "learning_rate": 0.00012137580636785726,
+      "loss": 0.9255,
+      "step": 4905
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.1977776288986206,
+      "learning_rate": 0.00012134875654592736,
+      "loss": 0.9156,
+      "step": 4906
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.20104831457138062,
+      "learning_rate": 0.00012132170508723082,
+      "loss": 0.9236,
+      "step": 4907
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.2730604410171509,
+      "learning_rate": 0.00012129465199384157,
+      "loss": 0.8175,
+      "step": 4908
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.20264138281345367,
+      "learning_rate": 0.00012126759726783379,
+      "loss": 0.9311,
+      "step": 4909
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.21460282802581787,
+      "learning_rate": 0.00012124054091128165,
+      "loss": 0.9825,
+      "step": 4910
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.20077942311763763,
+      "learning_rate": 0.00012121348292625954,
+      "loss": 0.8766,
+      "step": 4911
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.19806306064128876,
+      "learning_rate": 0.00012118642331484194,
+      "loss": 0.8929,
+      "step": 4912
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.19765718281269073,
+      "learning_rate": 0.00012115936207910343,
+      "loss": 0.9899,
+      "step": 4913
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.20788128674030304,
+      "learning_rate": 0.00012113229922111877,
+      "loss": 0.9416,
+      "step": 4914
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.20780085027217865,
+      "learning_rate": 0.00012110523474296281,
+      "loss": 0.9248,
+      "step": 4915
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.21029643714427948,
+      "learning_rate": 0.00012107816864671054,
+      "loss": 0.9705,
+      "step": 4916
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.2099684178829193,
+      "learning_rate": 0.00012105110093443703,
+      "loss": 0.9051,
+      "step": 4917
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.20328673720359802,
+      "learning_rate": 0.00012102403160821753,
+      "loss": 0.9653,
+      "step": 4918
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.20479902625083923,
+      "learning_rate": 0.0001209969606701274,
+      "loss": 0.9216,
+      "step": 4919
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.1987340748310089,
+      "learning_rate": 0.00012096988812224208,
+      "loss": 0.8951,
+      "step": 4920
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.1998843103647232,
+      "learning_rate": 0.00012094281396663722,
+      "loss": 0.8768,
+      "step": 4921
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.20857340097427368,
+      "learning_rate": 0.0001209157382053885,
+      "loss": 0.9482,
+      "step": 4922
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.21716679632663727,
+      "learning_rate": 0.0001208886608405718,
+      "loss": 0.8987,
+      "step": 4923
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.20676922798156738,
+      "learning_rate": 0.00012086158187426304,
+      "loss": 0.8846,
+      "step": 4924
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.20093367993831635,
+      "learning_rate": 0.00012083450130853833,
+      "loss": 0.8664,
+      "step": 4925
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.20721964538097382,
+      "learning_rate": 0.00012080741914547391,
+      "loss": 0.8971,
+      "step": 4926
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.2019132673740387,
+      "learning_rate": 0.00012078033538714611,
+      "loss": 0.8943,
+      "step": 4927
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.18864411115646362,
+      "learning_rate": 0.00012075325003563133,
+      "loss": 0.8439,
+      "step": 4928
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.2038918435573578,
+      "learning_rate": 0.00012072616309300625,
+      "loss": 0.9259,
+      "step": 4929
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.19137632846832275,
+      "learning_rate": 0.00012069907456134746,
+      "loss": 0.8448,
+      "step": 4930
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.20036281645298004,
+      "learning_rate": 0.00012067198444273187,
+      "loss": 0.923,
+      "step": 4931
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.2106323540210724,
+      "learning_rate": 0.00012064489273923638,
+      "loss": 0.9007,
+      "step": 4932
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.20353010296821594,
+      "learning_rate": 0.0001206177994529381,
+      "loss": 0.9127,
+      "step": 4933
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.1995949000120163,
+      "learning_rate": 0.00012059070458591414,
+      "loss": 0.878,
+      "step": 4934
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.19901487231254578,
+      "learning_rate": 0.00012056360814024188,
+      "loss": 0.9539,
+      "step": 4935
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.2054724097251892,
+      "learning_rate": 0.00012053651011799876,
+      "loss": 0.9506,
+      "step": 4936
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.19489338994026184,
+      "learning_rate": 0.00012050941052126226,
+      "loss": 0.8724,
+      "step": 4937
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.20409442484378815,
+      "learning_rate": 0.0001204823093521101,
+      "loss": 0.873,
+      "step": 4938
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.21491308510303497,
+      "learning_rate": 0.0001204552066126201,
+      "loss": 0.9649,
+      "step": 4939
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.20102648437023163,
+      "learning_rate": 0.0001204281023048701,
+      "loss": 0.8908,
+      "step": 4940
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.1951928734779358,
+      "learning_rate": 0.00012040099643093818,
+      "loss": 0.8707,
+      "step": 4941
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.19312453269958496,
+      "learning_rate": 0.00012037388899290252,
+      "loss": 0.9023,
+      "step": 4942
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.19747990369796753,
+      "learning_rate": 0.0001203467799928413,
+      "loss": 0.8916,
+      "step": 4943
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.19958235323429108,
+      "learning_rate": 0.00012031966943283303,
+      "loss": 0.9143,
+      "step": 4944
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.2075386494398117,
+      "learning_rate": 0.00012029255731495613,
+      "loss": 0.8906,
+      "step": 4945
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.1987982839345932,
+      "learning_rate": 0.00012026544364128926,
+      "loss": 0.8988,
+      "step": 4946
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.21121381223201752,
+      "learning_rate": 0.0001202383284139112,
+      "loss": 0.8728,
+      "step": 4947
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.20049889385700226,
+      "learning_rate": 0.00012021121163490078,
+      "loss": 0.8491,
+      "step": 4948
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.20627477765083313,
+      "learning_rate": 0.00012018409330633704,
+      "loss": 0.931,
+      "step": 4949
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.2119373083114624,
+      "learning_rate": 0.00012015697343029903,
+      "loss": 0.9066,
+      "step": 4950
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.20603740215301514,
+      "learning_rate": 0.00012012985200886602,
+      "loss": 0.911,
+      "step": 4951
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.19633600115776062,
+      "learning_rate": 0.00012010272904411732,
+      "loss": 0.9493,
+      "step": 4952
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.20193667709827423,
+      "learning_rate": 0.00012007560453813241,
+      "loss": 0.9339,
+      "step": 4953
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.1957622915506363,
+      "learning_rate": 0.0001200484784929909,
+      "loss": 0.8399,
+      "step": 4954
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.1923246532678604,
+      "learning_rate": 0.00012002135091077244,
+      "loss": 0.8511,
+      "step": 4955
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.2052861452102661,
+      "learning_rate": 0.00011999422179355689,
+      "loss": 0.9062,
+      "step": 4956
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.20123454928398132,
+      "learning_rate": 0.00011996709114342417,
+      "loss": 0.8783,
+      "step": 4957
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.19618423283100128,
+      "learning_rate": 0.00011993995896245429,
+      "loss": 0.9468,
+      "step": 4958
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.20387451350688934,
+      "learning_rate": 0.00011991282525272751,
+      "loss": 0.8704,
+      "step": 4959
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.1999095231294632,
+      "learning_rate": 0.00011988569001632403,
+      "loss": 0.9186,
+      "step": 4960
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.2033764272928238,
+      "learning_rate": 0.00011985855325532432,
+      "loss": 0.9135,
+      "step": 4961
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.19539391994476318,
+      "learning_rate": 0.00011983141497180885,
+      "loss": 0.8872,
+      "step": 4962
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.19846411049365997,
+      "learning_rate": 0.00011980427516785829,
+      "loss": 0.8634,
+      "step": 4963
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.19619372487068176,
+      "learning_rate": 0.00011977713384555343,
+      "loss": 0.8708,
+      "step": 4964
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.21282833814620972,
+      "learning_rate": 0.00011974999100697503,
+      "loss": 0.9144,
+      "step": 4965
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.21747606992721558,
+      "learning_rate": 0.00011972284665420417,
+      "loss": 0.9693,
+      "step": 4966
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.206832617521286,
+      "learning_rate": 0.00011969570078932193,
+      "loss": 0.8923,
+      "step": 4967
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.19994622468948364,
+      "learning_rate": 0.00011966855341440952,
+      "loss": 0.9185,
+      "step": 4968
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.19521817564964294,
+      "learning_rate": 0.00011964140453154833,
+      "loss": 0.9909,
+      "step": 4969
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.20193180441856384,
+      "learning_rate": 0.0001196142541428197,
+      "loss": 0.9187,
+      "step": 4970
+    },
+    {
+      "epoch": 1.73,
+      "eval_loss": 0.9203041195869446,
+      "eval_runtime": 757.3258,
+      "eval_samples_per_second": 9.079,
+      "eval_steps_per_second": 4.54,
+      "step": 4970
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.20618854463100433,
+      "learning_rate": 0.00011958710225030529,
+      "loss": 0.8969,
+      "step": 4971
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.20384158194065094,
+      "learning_rate": 0.00011955994885608677,
+      "loss": 0.9259,
+      "step": 4972
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.19744277000427246,
+      "learning_rate": 0.00011953279396224588,
+      "loss": 0.8548,
+      "step": 4973
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.19920489192008972,
+      "learning_rate": 0.00011950563757086461,
+      "loss": 0.941,
+      "step": 4974
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.18816323578357697,
+      "learning_rate": 0.0001194784796840249,
+      "loss": 0.842,
+      "step": 4975
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.20093213021755219,
+      "learning_rate": 0.00011945132030380897,
+      "loss": 0.9264,
+      "step": 4976
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.21550847589969635,
+      "learning_rate": 0.00011942415943229903,
+      "loss": 0.9098,
+      "step": 4977
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.19980955123901367,
+      "learning_rate": 0.00011939699707157748,
+      "loss": 0.8741,
+      "step": 4978
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.19830453395843506,
+      "learning_rate": 0.00011936983322372679,
+      "loss": 0.8678,
+      "step": 4979
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.21155405044555664,
+      "learning_rate": 0.00011934266789082953,
+      "loss": 0.9155,
+      "step": 4980
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.20855197310447693,
+      "learning_rate": 0.00011931550107496846,
+      "loss": 0.866,
+      "step": 4981
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.200531005859375,
+      "learning_rate": 0.00011928833277822638,
+      "loss": 0.8902,
+      "step": 4982
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.2080719918012619,
+      "learning_rate": 0.0001192611630026862,
+      "loss": 0.9563,
+      "step": 4983
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.20154033601284027,
+      "learning_rate": 0.00011923399175043103,
+      "loss": 0.8818,
+      "step": 4984
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.19243967533111572,
+      "learning_rate": 0.00011920681902354405,
+      "loss": 0.8583,
+      "step": 4985
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.20547130703926086,
+      "learning_rate": 0.00011917964482410843,
+      "loss": 0.8994,
+      "step": 4986
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.2043677717447281,
+      "learning_rate": 0.00011915246915420768,
+      "loss": 0.8972,
+      "step": 4987
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.19929523766040802,
+      "learning_rate": 0.00011912529201592522,
+      "loss": 0.9099,
+      "step": 4988
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.19746260344982147,
+      "learning_rate": 0.00011909811341134472,
+      "loss": 0.871,
+      "step": 4989
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.20300501585006714,
+      "learning_rate": 0.00011907093334254993,
+      "loss": 0.8044,
+      "step": 4990
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.19856730103492737,
+      "learning_rate": 0.0001190437518116246,
+      "loss": 0.8879,
+      "step": 4991
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.20286604762077332,
+      "learning_rate": 0.00011901656882065276,
+      "loss": 0.9892,
+      "step": 4992
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.20987409353256226,
+      "learning_rate": 0.00011898938437171842,
+      "loss": 0.8848,
+      "step": 4993
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.19633051753044128,
+      "learning_rate": 0.00011896219846690582,
+      "loss": 0.8424,
+      "step": 4994
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.19561024010181427,
+      "learning_rate": 0.00011893501110829926,
+      "loss": 0.8976,
+      "step": 4995
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.19931741058826447,
+      "learning_rate": 0.00011890782229798305,
+      "loss": 0.9,
+      "step": 4996
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.2051405906677246,
+      "learning_rate": 0.00011888063203804176,
+      "loss": 0.8666,
+      "step": 4997
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.20421981811523438,
+      "learning_rate": 0.00011885344033056,
+      "loss": 0.9034,
+      "step": 4998
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.2062828689813614,
+      "learning_rate": 0.00011882624717762252,
+      "loss": 0.867,
+      "step": 4999
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.2086305171251297,
+      "learning_rate": 0.00011879905258131418,
+      "loss": 0.9182,
+      "step": 5000
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.19741983711719513,
+      "learning_rate": 0.00011877185654371987,
+      "loss": 0.8988,
+      "step": 5001
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.20895710587501526,
+      "learning_rate": 0.00011874465906692473,
+      "loss": 0.9702,
+      "step": 5002
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.20342940092086792,
+      "learning_rate": 0.00011871746015301389,
+      "loss": 0.9123,
+      "step": 5003
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.202351376414299,
+      "learning_rate": 0.00011869025980407264,
+      "loss": 0.9364,
+      "step": 5004
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.20114704966545105,
+      "learning_rate": 0.00011866305802218642,
+      "loss": 0.8252,
+      "step": 5005
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.21236039698123932,
+      "learning_rate": 0.00011863585480944066,
+      "loss": 0.9436,
+      "step": 5006
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.21095523238182068,
+      "learning_rate": 0.00011860865016792105,
+      "loss": 0.9577,
+      "step": 5007
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.20168866217136383,
+      "learning_rate": 0.00011858144409971326,
+      "loss": 0.9228,
+      "step": 5008
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.19422025978565216,
+      "learning_rate": 0.00011855423660690317,
+      "loss": 0.8558,
+      "step": 5009
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.20206353068351746,
+      "learning_rate": 0.0001185270276915767,
+      "loss": 0.8905,
+      "step": 5010
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.20214833319187164,
+      "learning_rate": 0.0001184998173558199,
+      "loss": 0.9154,
+      "step": 5011
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.20583628118038177,
+      "learning_rate": 0.00011847260560171896,
+      "loss": 0.9023,
+      "step": 5012
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.20077918469905853,
+      "learning_rate": 0.00011844539243136008,
+      "loss": 0.9255,
+      "step": 5013
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.20009995996952057,
+      "learning_rate": 0.00011841817784682974,
+      "loss": 0.9335,
+      "step": 5014
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.20166489481925964,
+      "learning_rate": 0.00011839096185021438,
+      "loss": 0.9916,
+      "step": 5015
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.1968214064836502,
+      "learning_rate": 0.00011836374444360056,
+      "loss": 0.8258,
+      "step": 5016
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.2118564248085022,
+      "learning_rate": 0.00011833652562907504,
+      "loss": 0.916,
+      "step": 5017
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.20984144508838654,
+      "learning_rate": 0.0001183093054087246,
+      "loss": 0.8969,
+      "step": 5018
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.20295755565166473,
+      "learning_rate": 0.00011828208378463619,
+      "loss": 0.8394,
+      "step": 5019
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.2015915960073471,
+      "learning_rate": 0.00011825486075889683,
+      "loss": 0.9105,
+      "step": 5020
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.20518988370895386,
+      "learning_rate": 0.0001182276363335936,
+      "loss": 0.8709,
+      "step": 5021
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.1976424902677536,
+      "learning_rate": 0.0001182004105108138,
+      "loss": 0.8797,
+      "step": 5022
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.2075275331735611,
+      "learning_rate": 0.00011817318329264477,
+      "loss": 0.9243,
+      "step": 5023
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.20736710727214813,
+      "learning_rate": 0.00011814595468117397,
+      "loss": 0.9199,
+      "step": 5024
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.2057676911354065,
+      "learning_rate": 0.00011811872467848897,
+      "loss": 0.9225,
+      "step": 5025
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.206811785697937,
+      "learning_rate": 0.0001180914932866774,
+      "loss": 0.9511,
+      "step": 5026
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.1979030966758728,
+      "learning_rate": 0.00011806426050782709,
+      "loss": 0.8755,
+      "step": 5027
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.21560432016849518,
+      "learning_rate": 0.00011803702634402587,
+      "loss": 0.9617,
+      "step": 5028
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.22279982268810272,
+      "learning_rate": 0.00011800979079736176,
+      "loss": 0.9276,
+      "step": 5029
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.1999060958623886,
+      "learning_rate": 0.00011798255386992288,
+      "loss": 0.8727,
+      "step": 5030
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.2031540870666504,
+      "learning_rate": 0.00011795531556379737,
+      "loss": 0.9602,
+      "step": 5031
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.20556539297103882,
+      "learning_rate": 0.00011792807588107357,
+      "loss": 0.9124,
+      "step": 5032
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.20039956271648407,
+      "learning_rate": 0.00011790083482383989,
+      "loss": 0.8745,
+      "step": 5033
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.20217512547969818,
+      "learning_rate": 0.00011787359239418485,
+      "loss": 0.9396,
+      "step": 5034
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.20315247774124146,
+      "learning_rate": 0.0001178463485941971,
+      "loss": 0.9068,
+      "step": 5035
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.20155490934848785,
+      "learning_rate": 0.00011781910342596529,
+      "loss": 0.9777,
+      "step": 5036
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.20154711604118347,
+      "learning_rate": 0.00011779185689157835,
+      "loss": 0.9135,
+      "step": 5037
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.20197591185569763,
+      "learning_rate": 0.00011776460899312514,
+      "loss": 0.8422,
+      "step": 5038
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.19243617355823517,
+      "learning_rate": 0.00011773735973269472,
+      "loss": 0.886,
+      "step": 5039
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.21045725047588348,
+      "learning_rate": 0.00011771010911237628,
+      "loss": 0.9763,
+      "step": 5040
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.2029496282339096,
+      "learning_rate": 0.00011768285713425901,
+      "loss": 0.9022,
+      "step": 5041
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.20475132763385773,
+      "learning_rate": 0.0001176556038004323,
+      "loss": 0.8977,
+      "step": 5042
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.1879488229751587,
+      "learning_rate": 0.00011762834911298561,
+      "loss": 0.8861,
+      "step": 5043
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.1994534581899643,
+      "learning_rate": 0.00011760109307400849,
+      "loss": 0.8782,
+      "step": 5044
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.20935015380382538,
+      "learning_rate": 0.00011757383568559061,
+      "loss": 0.9391,
+      "step": 5045
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.20415537059307098,
+      "learning_rate": 0.00011754657694982174,
+      "loss": 0.8946,
+      "step": 5046
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.19910888373851776,
+      "learning_rate": 0.00011751931686879177,
+      "loss": 0.857,
+      "step": 5047
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.18923191726207733,
+      "learning_rate": 0.00011749205544459063,
+      "loss": 0.8723,
+      "step": 5048
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.2021467089653015,
+      "learning_rate": 0.00011746479267930844,
+      "loss": 0.8768,
+      "step": 5049
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.19616064429283142,
+      "learning_rate": 0.00011743752857503538,
+      "loss": 0.8571,
+      "step": 5050
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.19986507296562195,
+      "learning_rate": 0.00011741026313386172,
+      "loss": 0.885,
+      "step": 5051
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.20162515342235565,
+      "learning_rate": 0.00011738299635787786,
+      "loss": 0.9214,
+      "step": 5052
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.2050357162952423,
+      "learning_rate": 0.00011735572824917427,
+      "loss": 0.8715,
+      "step": 5053
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.1958906352519989,
+      "learning_rate": 0.00011732845880984153,
+      "loss": 0.8739,
+      "step": 5054
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.2050018459558487,
+      "learning_rate": 0.00011730118804197042,
+      "loss": 0.851,
+      "step": 5055
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.19149170815944672,
+      "learning_rate": 0.00011727391594765163,
+      "loss": 0.9196,
+      "step": 5056
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.20418493449687958,
+      "learning_rate": 0.0001172466425289761,
+      "loss": 0.8943,
+      "step": 5057
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.19700072705745697,
+      "learning_rate": 0.00011721936778803484,
+      "loss": 0.8402,
+      "step": 5058
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.19113709032535553,
+      "learning_rate": 0.00011719209172691892,
+      "loss": 0.8822,
+      "step": 5059
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.19682195782661438,
+      "learning_rate": 0.00011716481434771958,
+      "loss": 0.9305,
+      "step": 5060
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.19861525297164917,
+      "learning_rate": 0.0001171375356525281,
+      "loss": 0.8543,
+      "step": 5061
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.20508785545825958,
+      "learning_rate": 0.0001171102556434359,
+      "loss": 0.8968,
+      "step": 5062
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.19325780868530273,
+      "learning_rate": 0.00011708297432253444,
+      "loss": 0.8555,
+      "step": 5063
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.202621191740036,
+      "learning_rate": 0.00011705569169191534,
+      "loss": 0.8708,
+      "step": 5064
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.20007860660552979,
+      "learning_rate": 0.00011702840775367034,
+      "loss": 0.9425,
+      "step": 5065
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.20317086577415466,
+      "learning_rate": 0.00011700112250989122,
+      "loss": 0.9159,
+      "step": 5066
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.21285471320152283,
+      "learning_rate": 0.00011697383596266993,
+      "loss": 0.9658,
+      "step": 5067
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.20765405893325806,
+      "learning_rate": 0.0001169465481140984,
+      "loss": 0.8651,
+      "step": 5068
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.20606768131256104,
+      "learning_rate": 0.00011691925896626877,
+      "loss": 0.8446,
+      "step": 5069
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.193104088306427,
+      "learning_rate": 0.00011689196852127325,
+      "loss": 0.8825,
+      "step": 5070
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.19602930545806885,
+      "learning_rate": 0.0001168646767812041,
+      "loss": 0.8814,
+      "step": 5071
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.19966290891170502,
+      "learning_rate": 0.00011683738374815382,
+      "loss": 0.8934,
+      "step": 5072
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.21192531287670135,
+      "learning_rate": 0.00011681008942421483,
+      "loss": 0.9528,
+      "step": 5073
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.2027505785226822,
+      "learning_rate": 0.00011678279381147975,
+      "loss": 0.8551,
+      "step": 5074
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.2016507089138031,
+      "learning_rate": 0.00011675549691204129,
+      "loss": 0.925,
+      "step": 5075
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.21192465722560883,
+      "learning_rate": 0.00011672819872799223,
+      "loss": 0.8924,
+      "step": 5076
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.2008252888917923,
+      "learning_rate": 0.00011670089926142553,
+      "loss": 0.8713,
+      "step": 5077
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.2006734311580658,
+      "learning_rate": 0.00011667359851443411,
+      "loss": 0.9093,
+      "step": 5078
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.20377005636692047,
+      "learning_rate": 0.00011664629648911108,
+      "loss": 0.8406,
+      "step": 5079
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.20884154736995697,
+      "learning_rate": 0.00011661899318754965,
+      "loss": 0.9309,
+      "step": 5080
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.20641496777534485,
+      "learning_rate": 0.00011659168861184308,
+      "loss": 0.8986,
+      "step": 5081
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.20506221055984497,
+      "learning_rate": 0.00011656438276408484,
+      "loss": 0.9009,
+      "step": 5082
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.19785983860492706,
+      "learning_rate": 0.00011653707564636833,
+      "loss": 0.9092,
+      "step": 5083
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.19785176217556,
+      "learning_rate": 0.00011650976726078712,
+      "loss": 0.8682,
+      "step": 5084
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.2017369270324707,
+      "learning_rate": 0.00011648245760943499,
+      "loss": 0.9217,
+      "step": 5085
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.2052314579486847,
+      "learning_rate": 0.0001164551466944056,
+      "loss": 0.8482,
+      "step": 5086
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.19984367489814758,
+      "learning_rate": 0.00011642783451779294,
+      "loss": 0.8843,
+      "step": 5087
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.21052081882953644,
+      "learning_rate": 0.00011640052108169088,
+      "loss": 0.9245,
+      "step": 5088
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.19494381546974182,
+      "learning_rate": 0.00011637320638819349,
+      "loss": 0.8456,
+      "step": 5089
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.2073839157819748,
+      "learning_rate": 0.00011634589043939501,
+      "loss": 0.8419,
+      "step": 5090
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.19387920200824738,
+      "learning_rate": 0.00011631857323738966,
+      "loss": 0.8577,
+      "step": 5091
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.20236201584339142,
+      "learning_rate": 0.00011629125478427178,
+      "loss": 0.9158,
+      "step": 5092
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.20327600836753845,
+      "learning_rate": 0.00011626393508213586,
+      "loss": 0.9471,
+      "step": 5093
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.19547243416309357,
+      "learning_rate": 0.00011623661413307639,
+      "loss": 0.853,
+      "step": 5094
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.20263858139514923,
+      "learning_rate": 0.00011620929193918806,
+      "loss": 0.9924,
+      "step": 5095
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.2074957937002182,
+      "learning_rate": 0.00011618196850256557,
+      "loss": 0.9559,
+      "step": 5096
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.19843412935733795,
+      "learning_rate": 0.0001161546438253038,
+      "loss": 0.8685,
+      "step": 5097
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.20189149677753448,
+      "learning_rate": 0.00011612731790949767,
+      "loss": 0.8867,
+      "step": 5098
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.19635272026062012,
+      "learning_rate": 0.00011609999075724214,
+      "loss": 0.8595,
+      "step": 5099
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.20424078404903412,
+      "learning_rate": 0.00011607266237063242,
+      "loss": 0.9016,
+      "step": 5100
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.19970911741256714,
+      "learning_rate": 0.00011604533275176366,
+      "loss": 0.9319,
+      "step": 5101
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.2071448713541031,
+      "learning_rate": 0.0001160180019027312,
+      "loss": 0.8947,
+      "step": 5102
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.20625151693820953,
+      "learning_rate": 0.00011599066982563043,
+      "loss": 0.8917,
+      "step": 5103
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.20384685695171356,
+      "learning_rate": 0.0001159633365225568,
+      "loss": 0.8873,
+      "step": 5104
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.19870717823505402,
+      "learning_rate": 0.00011593600199560599,
+      "loss": 0.8186,
+      "step": 5105
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.19644193351268768,
+      "learning_rate": 0.00011590866624687362,
+      "loss": 0.8968,
+      "step": 5106
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.2026110291481018,
+      "learning_rate": 0.00011588132927845552,
+      "loss": 0.9062,
+      "step": 5107
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.20113103091716766,
+      "learning_rate": 0.0001158539910924475,
+      "loss": 0.8294,
+      "step": 5108
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.20522204041481018,
+      "learning_rate": 0.00011582665169094554,
+      "loss": 0.9486,
+      "step": 5109
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.1949976086616516,
+      "learning_rate": 0.00011579931107604571,
+      "loss": 0.8544,
+      "step": 5110
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.20692569017410278,
+      "learning_rate": 0.00011577196924984418,
+      "loss": 0.8792,
+      "step": 5111
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.19932514429092407,
+      "learning_rate": 0.00011574462621443714,
+      "loss": 0.9022,
+      "step": 5112
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.20348139107227325,
+      "learning_rate": 0.000115717281971921,
+      "loss": 0.8854,
+      "step": 5113
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.1993340402841568,
+      "learning_rate": 0.0001156899365243921,
+      "loss": 0.9246,
+      "step": 5114
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.2046261727809906,
+      "learning_rate": 0.00011566258987394704,
+      "loss": 0.8866,
+      "step": 5115
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.21360766887664795,
+      "learning_rate": 0.00011563524202268241,
+      "loss": 0.9124,
+      "step": 5116
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.2009619027376175,
+      "learning_rate": 0.00011560789297269487,
+      "loss": 0.9078,
+      "step": 5117
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.20256301760673523,
+      "learning_rate": 0.0001155805427260813,
+      "loss": 0.9127,
+      "step": 5118
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.19327925145626068,
+      "learning_rate": 0.00011555319128493851,
+      "loss": 0.8556,
+      "step": 5119
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.19143442809581757,
+      "learning_rate": 0.00011552583865136353,
+      "loss": 0.8526,
+      "step": 5120
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.20160861313343048,
+      "learning_rate": 0.0001154984848274534,
+      "loss": 0.8435,
+      "step": 5121
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.19750486314296722,
+      "learning_rate": 0.00011547112981530531,
+      "loss": 0.8792,
+      "step": 5122
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.2013956904411316,
+      "learning_rate": 0.0001154437736170165,
+      "loss": 0.8681,
+      "step": 5123
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.20969411730766296,
+      "learning_rate": 0.00011541641623468431,
+      "loss": 0.9353,
+      "step": 5124
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.2132592648267746,
+      "learning_rate": 0.0001153890576704062,
+      "loss": 0.9047,
+      "step": 5125
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.2049732804298401,
+      "learning_rate": 0.00011536169792627967,
+      "loss": 0.901,
+      "step": 5126
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.20980370044708252,
+      "learning_rate": 0.00011533433700440232,
+      "loss": 0.9547,
+      "step": 5127
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.19606991112232208,
+      "learning_rate": 0.00011530697490687194,
+      "loss": 0.9157,
+      "step": 5128
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.19489556550979614,
+      "learning_rate": 0.00011527961163578624,
+      "loss": 0.8315,
+      "step": 5129
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.20620688796043396,
+      "learning_rate": 0.00011525224719324314,
+      "loss": 0.9093,
+      "step": 5130
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.21620605885982513,
+      "learning_rate": 0.00011522488158134064,
+      "loss": 0.948,
+      "step": 5131
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.20374895632266998,
+      "learning_rate": 0.00011519751480217676,
+      "loss": 0.8583,
+      "step": 5132
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.19723199307918549,
+      "learning_rate": 0.0001151701468578497,
+      "loss": 0.9123,
+      "step": 5133
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.20205992460250854,
+      "learning_rate": 0.00011514277775045768,
+      "loss": 0.9027,
+      "step": 5134
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.19809818267822266,
+      "learning_rate": 0.00011511540748209903,
+      "loss": 0.9026,
+      "step": 5135
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.20418870449066162,
+      "learning_rate": 0.00011508803605487222,
+      "loss": 0.848,
+      "step": 5136
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.19335588812828064,
+      "learning_rate": 0.00011506066347087568,
+      "loss": 0.8672,
+      "step": 5137
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.20655111968517303,
+      "learning_rate": 0.00011503328973220811,
+      "loss": 0.8827,
+      "step": 5138
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.19734971225261688,
+      "learning_rate": 0.0001150059148409681,
+      "loss": 0.8798,
+      "step": 5139
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.2011449933052063,
+      "learning_rate": 0.0001149785387992545,
+      "loss": 0.9196,
+      "step": 5140
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.19869692623615265,
+      "learning_rate": 0.00011495116160916617,
+      "loss": 0.8314,
+      "step": 5141
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.20650891959667206,
+      "learning_rate": 0.00011492378327280202,
+      "loss": 0.8761,
+      "step": 5142
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.2007884383201599,
+      "learning_rate": 0.00011489640379226116,
+      "loss": 0.8532,
+      "step": 5143
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.19970685243606567,
+      "learning_rate": 0.00011486902316964265,
+      "loss": 0.9252,
+      "step": 5144
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.20636236667633057,
+      "learning_rate": 0.00011484164140704577,
+      "loss": 0.8707,
+      "step": 5145
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.2008487433195114,
+      "learning_rate": 0.00011481425850656978,
+      "loss": 0.899,
+      "step": 5146
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.2033643275499344,
+      "learning_rate": 0.00011478687447031407,
+      "loss": 0.9197,
+      "step": 5147
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.20582693815231323,
+      "learning_rate": 0.00011475948930037819,
+      "loss": 1.0164,
+      "step": 5148
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.2104448676109314,
+      "learning_rate": 0.0001147321029988616,
+      "loss": 0.9111,
+      "step": 5149
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.19654923677444458,
+      "learning_rate": 0.00011470471556786401,
+      "loss": 0.9337,
+      "step": 5150
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.19730478525161743,
+      "learning_rate": 0.00011467732700948519,
+      "loss": 0.955,
+      "step": 5151
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.19658218324184418,
+      "learning_rate": 0.0001146499373258249,
+      "loss": 0.893,
+      "step": 5152
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.2017667293548584,
+      "learning_rate": 0.00011462254651898312,
+      "loss": 0.933,
+      "step": 5153
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.19858188927173615,
+      "learning_rate": 0.00011459515459105978,
+      "loss": 0.8692,
+      "step": 5154
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.19772326946258545,
+      "learning_rate": 0.00011456776154415502,
+      "loss": 0.8988,
+      "step": 5155
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.19625569880008698,
+      "learning_rate": 0.00011454036738036899,
+      "loss": 0.8658,
+      "step": 5156
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.20135082304477692,
+      "learning_rate": 0.00011451297210180195,
+      "loss": 0.8202,
+      "step": 5157
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.20419570803642273,
+      "learning_rate": 0.00011448557571055423,
+      "loss": 0.9184,
+      "step": 5158
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.2074098289012909,
+      "learning_rate": 0.00011445817820872627,
+      "loss": 0.9901,
+      "step": 5159
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.20851200819015503,
+      "learning_rate": 0.00011443077959841856,
+      "loss": 0.9857,
+      "step": 5160
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.20411436259746552,
+      "learning_rate": 0.00011440337988173173,
+      "loss": 0.9306,
+      "step": 5161
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.2083197683095932,
+      "learning_rate": 0.00011437597906076641,
+      "loss": 0.8815,
+      "step": 5162
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.20529308915138245,
+      "learning_rate": 0.00011434857713762344,
+      "loss": 0.9152,
+      "step": 5163
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.20252062380313873,
+      "learning_rate": 0.00011432117411440359,
+      "loss": 0.9166,
+      "step": 5164
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.1970435082912445,
+      "learning_rate": 0.00011429376999320789,
+      "loss": 0.8705,
+      "step": 5165
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.2025022804737091,
+      "learning_rate": 0.00011426636477613728,
+      "loss": 0.8794,
+      "step": 5166
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.20577043294906616,
+      "learning_rate": 0.00011423895846529286,
+      "loss": 0.9026,
+      "step": 5167
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.2037774920463562,
+      "learning_rate": 0.00011421155106277588,
+      "loss": 0.9078,
+      "step": 5168
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.20879001915454865,
+      "learning_rate": 0.00011418414257068759,
+      "loss": 0.9129,
+      "step": 5169
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.1920623779296875,
+      "learning_rate": 0.00011415673299112932,
+      "loss": 0.8451,
+      "step": 5170
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.19346140325069427,
+      "learning_rate": 0.0001141293223262025,
+      "loss": 0.8318,
+      "step": 5171
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.19851449131965637,
+      "learning_rate": 0.00011410191057800868,
+      "loss": 0.9736,
+      "step": 5172
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.19635413587093353,
+      "learning_rate": 0.00011407449774864947,
+      "loss": 0.8718,
+      "step": 5173
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.2130098044872284,
+      "learning_rate": 0.00011404708384022654,
+      "loss": 0.8962,
+      "step": 5174
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.18981891870498657,
+      "learning_rate": 0.00011401966885484165,
+      "loss": 0.8019,
+      "step": 5175
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.2004854828119278,
+      "learning_rate": 0.00011399225279459669,
+      "loss": 0.8952,
+      "step": 5176
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.20224079489707947,
+      "learning_rate": 0.00011396483566159352,
+      "loss": 0.9395,
+      "step": 5177
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.2006518691778183,
+      "learning_rate": 0.00011393741745793425,
+      "loss": 0.9353,
+      "step": 5178
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.197438046336174,
+      "learning_rate": 0.00011390999818572093,
+      "loss": 0.8748,
+      "step": 5179
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.2060369998216629,
+      "learning_rate": 0.00011388257784705573,
+      "loss": 0.8877,
+      "step": 5180
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.20259609818458557,
+      "learning_rate": 0.00011385515644404095,
+      "loss": 0.8416,
+      "step": 5181
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.2101147472858429,
+      "learning_rate": 0.00011382773397877888,
+      "loss": 0.9739,
+      "step": 5182
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.19835442304611206,
+      "learning_rate": 0.00011380031045337201,
+      "loss": 0.8296,
+      "step": 5183
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.20819783210754395,
+      "learning_rate": 0.00011377288586992281,
+      "loss": 0.9094,
+      "step": 5184
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.20873680710792542,
+      "learning_rate": 0.00011374546023053388,
+      "loss": 0.9217,
+      "step": 5185
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.20260164141654968,
+      "learning_rate": 0.00011371803353730786,
+      "loss": 0.933,
+      "step": 5186
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.20431824028491974,
+      "learning_rate": 0.00011369060579234754,
+      "loss": 0.9032,
+      "step": 5187
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.21404831111431122,
+      "learning_rate": 0.00011366317699775573,
+      "loss": 0.9451,
+      "step": 5188
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.1977844089269638,
+      "learning_rate": 0.00011363574715563534,
+      "loss": 0.8453,
+      "step": 5189
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.21856170892715454,
+      "learning_rate": 0.00011360831626808938,
+      "loss": 0.9768,
+      "step": 5190
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.20709745585918427,
+      "learning_rate": 0.00011358088433722092,
+      "loss": 0.8808,
+      "step": 5191
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.20399156212806702,
+      "learning_rate": 0.00011355345136513307,
+      "loss": 0.9244,
+      "step": 5192
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.20622244477272034,
+      "learning_rate": 0.0001135260173539291,
+      "loss": 0.9885,
+      "step": 5193
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.19313405454158783,
+      "learning_rate": 0.00011349858230571233,
+      "loss": 0.8616,
+      "step": 5194
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.19614113867282867,
+      "learning_rate": 0.00011347114622258612,
+      "loss": 0.9269,
+      "step": 5195
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.1973341405391693,
+      "learning_rate": 0.00011344370910665393,
+      "loss": 0.925,
+      "step": 5196
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.2052834928035736,
+      "learning_rate": 0.00011341627096001937,
+      "loss": 0.9339,
+      "step": 5197
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.20807234942913055,
+      "learning_rate": 0.000113388831784786,
+      "loss": 0.9083,
+      "step": 5198
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.2042306512594223,
+      "learning_rate": 0.00011336139158305758,
+      "loss": 0.9299,
+      "step": 5199
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.20556193590164185,
+      "learning_rate": 0.00011333395035693786,
+      "loss": 0.9511,
+      "step": 5200
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.19857318699359894,
+      "learning_rate": 0.00011330650810853072,
+      "loss": 0.8736,
+      "step": 5201
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.1991368979215622,
+      "learning_rate": 0.00011327906483994008,
+      "loss": 0.8783,
+      "step": 5202
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.20878826081752777,
+      "learning_rate": 0.00011325162055327001,
+      "loss": 0.8912,
+      "step": 5203
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.2100679725408554,
+      "learning_rate": 0.00011322417525062456,
+      "loss": 0.9602,
+      "step": 5204
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.19470907747745514,
+      "learning_rate": 0.00011319672893410796,
+      "loss": 0.9297,
+      "step": 5205
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.20928390324115753,
+      "learning_rate": 0.00011316928160582438,
+      "loss": 0.9481,
+      "step": 5206
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.194465771317482,
+      "learning_rate": 0.00011314183326787822,
+      "loss": 0.9818,
+      "step": 5207
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.23972180485725403,
+      "learning_rate": 0.00011311438392237387,
+      "loss": 1.1,
+      "step": 5208
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.1954929530620575,
+      "learning_rate": 0.00011308693357141582,
+      "loss": 0.8943,
+      "step": 5209
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.203359916806221,
+      "learning_rate": 0.00011305948221710865,
+      "loss": 0.8708,
+      "step": 5210
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.20809759199619293,
+      "learning_rate": 0.00011303202986155696,
+      "loss": 0.8858,
+      "step": 5211
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.2132718712091446,
+      "learning_rate": 0.00011300457650686547,
+      "loss": 0.9174,
+      "step": 5212
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.20165328681468964,
+      "learning_rate": 0.00011297712215513903,
+      "loss": 0.8968,
+      "step": 5213
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.19978751242160797,
+      "learning_rate": 0.00011294966680848247,
+      "loss": 0.9146,
+      "step": 5214
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.19985590875148773,
+      "learning_rate": 0.00011292221046900074,
+      "loss": 0.9535,
+      "step": 5215
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.201857790350914,
+      "learning_rate": 0.00011289475313879885,
+      "loss": 0.8486,
+      "step": 5216
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.20473961532115936,
+      "learning_rate": 0.0001128672948199819,
+      "loss": 0.8757,
+      "step": 5217
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.2123999297618866,
+      "learning_rate": 0.00011283983551465511,
+      "loss": 0.8902,
+      "step": 5218
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.19770240783691406,
+      "learning_rate": 0.00011281237522492367,
+      "loss": 0.9091,
+      "step": 5219
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.20275746285915375,
+      "learning_rate": 0.00011278491395289296,
+      "loss": 0.9041,
+      "step": 5220
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.19588282704353333,
+      "learning_rate": 0.00011275745170066834,
+      "loss": 0.8687,
+      "step": 5221
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.20717737078666687,
+      "learning_rate": 0.0001127299884703553,
+      "loss": 0.9517,
+      "step": 5222
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.2018229365348816,
+      "learning_rate": 0.00011270252426405939,
+      "loss": 0.8683,
+      "step": 5223
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.20680849254131317,
+      "learning_rate": 0.00011267505908388625,
+      "loss": 0.8913,
+      "step": 5224
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.2059405893087387,
+      "learning_rate": 0.00011264759293194156,
+      "loss": 0.8993,
+      "step": 5225
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.20858971774578094,
+      "learning_rate": 0.0001126201258103311,
+      "loss": 0.9484,
+      "step": 5226
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.20857854187488556,
+      "learning_rate": 0.00011259265772116072,
+      "loss": 0.8994,
+      "step": 5227
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.1971004158258438,
+      "learning_rate": 0.00011256518866653636,
+      "loss": 0.867,
+      "step": 5228
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.20058520138263702,
+      "learning_rate": 0.000112537718648564,
+      "loss": 0.9123,
+      "step": 5229
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.19929702579975128,
+      "learning_rate": 0.00011251024766934973,
+      "loss": 0.9114,
+      "step": 5230
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.20408372581005096,
+      "learning_rate": 0.00011248277573099966,
+      "loss": 0.9088,
+      "step": 5231
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.21157945692539215,
+      "learning_rate": 0.00011245530283562,
+      "loss": 0.8944,
+      "step": 5232
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.1991204172372818,
+      "learning_rate": 0.00011242782898531712,
+      "loss": 0.9014,
+      "step": 5233
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.19715304672718048,
+      "learning_rate": 0.00011240035418219732,
+      "loss": 0.8788,
+      "step": 5234
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.19844309985637665,
+      "learning_rate": 0.00011237287842836707,
+      "loss": 0.8362,
+      "step": 5235
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.20528234541416168,
+      "learning_rate": 0.00011234540172593287,
+      "loss": 0.8645,
+      "step": 5236
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.21071383357048035,
+      "learning_rate": 0.00011231792407700125,
+      "loss": 0.8801,
+      "step": 5237
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.20560497045516968,
+      "learning_rate": 0.00011229044548367897,
+      "loss": 0.889,
+      "step": 5238
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.19932425022125244,
+      "learning_rate": 0.00011226296594807271,
+      "loss": 0.9381,
+      "step": 5239
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.21976758539676666,
+      "learning_rate": 0.00011223548547228924,
+      "loss": 0.8839,
+      "step": 5240
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.20521189272403717,
+      "learning_rate": 0.0001122080040584355,
+      "loss": 0.9186,
+      "step": 5241
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.2094995379447937,
+      "learning_rate": 0.00011218052170861835,
+      "loss": 0.956,
+      "step": 5242
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.2032909095287323,
+      "learning_rate": 0.00011215303842494489,
+      "loss": 0.9085,
+      "step": 5243
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.1991930603981018,
+      "learning_rate": 0.00011212555420952215,
+      "loss": 0.9086,
+      "step": 5244
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.20247501134872437,
+      "learning_rate": 0.0001120980690644573,
+      "loss": 0.8857,
+      "step": 5245
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.19612811505794525,
+      "learning_rate": 0.00011207058299185765,
+      "loss": 0.8563,
+      "step": 5246
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.2090568244457245,
+      "learning_rate": 0.00011204309599383035,
+      "loss": 0.9022,
+      "step": 5247
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.1988435536623001,
+      "learning_rate": 0.00011201560807248293,
+      "loss": 0.8883,
+      "step": 5248
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.2111375778913498,
+      "learning_rate": 0.00011198811922992274,
+      "loss": 0.9505,
+      "step": 5249
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.20622937381267548,
+      "learning_rate": 0.0001119606294682573,
+      "loss": 0.9294,
+      "step": 5250
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.21491630375385284,
+      "learning_rate": 0.00011193313878959427,
+      "loss": 0.9827,
+      "step": 5251
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.20232580602169037,
+      "learning_rate": 0.00011190564719604118,
+      "loss": 0.9016,
+      "step": 5252
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.2048744410276413,
+      "learning_rate": 0.00011187815468970586,
+      "loss": 0.8494,
+      "step": 5253
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.2030027210712433,
+      "learning_rate": 0.00011185066127269607,
+      "loss": 0.9294,
+      "step": 5254
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.21077583730220795,
+      "learning_rate": 0.00011182316694711965,
+      "loss": 0.8665,
+      "step": 5255
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.201682448387146,
+      "learning_rate": 0.00011179567171508463,
+      "loss": 0.8872,
+      "step": 5256
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.20252567529678345,
+      "learning_rate": 0.00011176817557869887,
+      "loss": 0.872,
+      "step": 5257
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.1869186908006668,
+      "learning_rate": 0.00011174067854007058,
+      "loss": 0.8627,
+      "step": 5258
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.20626534521579742,
+      "learning_rate": 0.00011171318060130783,
+      "loss": 0.959,
+      "step": 5259
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.21301251649856567,
+      "learning_rate": 0.00011168568176451881,
+      "loss": 0.8972,
+      "step": 5260
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.21265006065368652,
+      "learning_rate": 0.00011165818203181189,
+      "loss": 0.9791,
+      "step": 5261
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.21141482889652252,
+      "learning_rate": 0.00011163068140529531,
+      "loss": 0.9556,
+      "step": 5262
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.1883353292942047,
+      "learning_rate": 0.00011160317988707759,
+      "loss": 0.8423,
+      "step": 5263
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.19268368184566498,
+      "learning_rate": 0.00011157567747926716,
+      "loss": 0.89,
+      "step": 5264
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.1984165757894516,
+      "learning_rate": 0.00011154817418397257,
+      "loss": 0.9212,
+      "step": 5265
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.2028784602880478,
+      "learning_rate": 0.0001115206700033025,
+      "loss": 0.8673,
+      "step": 5266
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.19984976947307587,
+      "learning_rate": 0.00011149316493936555,
+      "loss": 0.9039,
+      "step": 5267
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.20026452839374542,
+      "learning_rate": 0.00011146565899427057,
+      "loss": 0.9054,
+      "step": 5268
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.20769165456295013,
+      "learning_rate": 0.00011143815217012633,
+      "loss": 0.9541,
+      "step": 5269
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.2046111524105072,
+      "learning_rate": 0.00011141064446904172,
+      "loss": 0.9185,
+      "step": 5270
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.2113402932882309,
+      "learning_rate": 0.00011138313589312577,
+      "loss": 0.9294,
+      "step": 5271
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.20995759963989258,
+      "learning_rate": 0.00011135562644448741,
+      "loss": 0.8825,
+      "step": 5272
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.20764729380607605,
+      "learning_rate": 0.0001113281161252358,
+      "loss": 0.8985,
+      "step": 5273
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.21376466751098633,
+      "learning_rate": 0.00011130060493748007,
+      "loss": 0.9318,
+      "step": 5274
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.19655349850654602,
+      "learning_rate": 0.00011127309288332945,
+      "loss": 0.8771,
+      "step": 5275
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.19957830011844635,
+      "learning_rate": 0.00011124557996489329,
+      "loss": 0.89,
+      "step": 5276
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.19681084156036377,
+      "learning_rate": 0.00011121806618428085,
+      "loss": 0.8815,
+      "step": 5277
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.21147172152996063,
+      "learning_rate": 0.00011119055154360165,
+      "loss": 0.8854,
+      "step": 5278
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.19879549741744995,
+      "learning_rate": 0.00011116303604496517,
+      "loss": 0.8512,
+      "step": 5279
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.1971660852432251,
+      "learning_rate": 0.00011113551969048089,
+      "loss": 0.8922,
+      "step": 5280
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.20926682651042938,
+      "learning_rate": 0.0001111080024822585,
+      "loss": 0.9082,
+      "step": 5281
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.20934419333934784,
+      "learning_rate": 0.0001110804844224077,
+      "loss": 0.948,
+      "step": 5282
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.2003125250339508,
+      "learning_rate": 0.0001110529655130382,
+      "loss": 0.8796,
+      "step": 5283
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.208330899477005,
+      "learning_rate": 0.00011102544575625988,
+      "loss": 0.92,
+      "step": 5284
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.20303812623023987,
+      "learning_rate": 0.00011099792515418252,
+      "loss": 0.9019,
+      "step": 5285
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.19247807562351227,
+      "learning_rate": 0.00011097040370891621,
+      "loss": 0.8892,
+      "step": 5286
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.20970343053340912,
+      "learning_rate": 0.00011094288142257086,
+      "loss": 0.9262,
+      "step": 5287
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.20390614867210388,
+      "learning_rate": 0.00011091535829725659,
+      "loss": 0.9264,
+      "step": 5288
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.19981978833675385,
+      "learning_rate": 0.00011088783433508352,
+      "loss": 0.8799,
+      "step": 5289
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.20050524175167084,
+      "learning_rate": 0.00011086030953816187,
+      "loss": 0.8639,
+      "step": 5290
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.2076718658208847,
+      "learning_rate": 0.0001108327839086019,
+      "loss": 0.8632,
+      "step": 5291
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.21506258845329285,
+      "learning_rate": 0.000110805257448514,
+      "loss": 0.912,
+      "step": 5292
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.21158722043037415,
+      "learning_rate": 0.00011077773016000849,
+      "loss": 0.936,
+      "step": 5293
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.21115998923778534,
+      "learning_rate": 0.00011075020204519588,
+      "loss": 0.9155,
+      "step": 5294
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.20540079474449158,
+      "learning_rate": 0.00011072267310618669,
+      "loss": 0.9158,
+      "step": 5295
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.19588345289230347,
+      "learning_rate": 0.0001106951433450915,
+      "loss": 0.846,
+      "step": 5296
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.20122110843658447,
+      "learning_rate": 0.00011066761276402095,
+      "loss": 0.846,
+      "step": 5297
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.19394049048423767,
+      "learning_rate": 0.00011064008136508581,
+      "loss": 0.8491,
+      "step": 5298
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.20480981469154358,
+      "learning_rate": 0.0001106125491503968,
+      "loss": 0.8792,
+      "step": 5299
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.205979123711586,
+      "learning_rate": 0.00011058501612206476,
+      "loss": 0.907,
+      "step": 5300
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.20306456089019775,
+      "learning_rate": 0.00011055748228220062,
+      "loss": 0.8966,
+      "step": 5301
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.20474489033222198,
+      "learning_rate": 0.00011052994763291533,
+      "loss": 0.8474,
+      "step": 5302
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.19772367179393768,
+      "learning_rate": 0.00011050241217631994,
+      "loss": 0.914,
+      "step": 5303
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.2080567330121994,
+      "learning_rate": 0.00011047487591452552,
+      "loss": 0.8831,
+      "step": 5304
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.20726226270198822,
+      "learning_rate": 0.0001104473388496432,
+      "loss": 0.9431,
+      "step": 5305
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.20143887400627136,
+      "learning_rate": 0.00011041980098378421,
+      "loss": 0.9187,
+      "step": 5306
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.20895779132843018,
+      "learning_rate": 0.00011039226231905986,
+      "loss": 0.8455,
+      "step": 5307
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.20928290486335754,
+      "learning_rate": 0.00011036472285758144,
+      "loss": 0.9203,
+      "step": 5308
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.20436181128025055,
+      "learning_rate": 0.00011033718260146036,
+      "loss": 0.9224,
+      "step": 5309
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.19997340440750122,
+      "learning_rate": 0.00011030964155280803,
+      "loss": 0.9114,
+      "step": 5310
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.19829915463924408,
+      "learning_rate": 0.00011028209971373605,
+      "loss": 0.8384,
+      "step": 5311
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.19507120549678802,
+      "learning_rate": 0.00011025455708635595,
+      "loss": 0.8925,
+      "step": 5312
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.1935853362083435,
+      "learning_rate": 0.0001102270136727794,
+      "loss": 0.8382,
+      "step": 5313
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.2118246704339981,
+      "learning_rate": 0.00011019946947511804,
+      "loss": 0.9267,
+      "step": 5314
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.20547886192798615,
+      "learning_rate": 0.00011017192449548365,
+      "loss": 0.9465,
+      "step": 5315
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.20173491537570953,
+      "learning_rate": 0.00011014437873598811,
+      "loss": 0.9013,
+      "step": 5316
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.19979619979858398,
+      "learning_rate": 0.00011011683219874323,
+      "loss": 0.9181,
+      "step": 5317
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.2074444591999054,
+      "learning_rate": 0.00011008928488586096,
+      "loss": 0.9071,
+      "step": 5318
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.2015903890132904,
+      "learning_rate": 0.00011006173679945333,
+      "loss": 0.8847,
+      "step": 5319
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.20304925739765167,
+      "learning_rate": 0.00011003418794163236,
+      "loss": 0.8981,
+      "step": 5320
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.209546759724617,
+      "learning_rate": 0.00011000663831451017,
+      "loss": 0.9768,
+      "step": 5321
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.20055563747882843,
+      "learning_rate": 0.00010997908792019898,
+      "loss": 0.9023,
+      "step": 5322
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.20044977962970734,
+      "learning_rate": 0.00010995153676081096,
+      "loss": 0.9173,
+      "step": 5323
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.20793165266513824,
+      "learning_rate": 0.00010992398483845845,
+      "loss": 0.9098,
+      "step": 5324
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.2018725723028183,
+      "learning_rate": 0.00010989643215525375,
+      "loss": 0.8702,
+      "step": 5325
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.21068057417869568,
+      "learning_rate": 0.00010986887871330934,
+      "loss": 0.9116,
+      "step": 5326
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.20038604736328125,
+      "learning_rate": 0.00010984132451473763,
+      "loss": 0.9127,
+      "step": 5327
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.21165314316749573,
+      "learning_rate": 0.00010981376956165117,
+      "loss": 0.939,
+      "step": 5328
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.2063116729259491,
+      "learning_rate": 0.00010978621385616255,
+      "loss": 0.8957,
+      "step": 5329
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.20496952533721924,
+      "learning_rate": 0.00010975865740038438,
+      "loss": 0.9017,
+      "step": 5330
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.21283504366874695,
+      "learning_rate": 0.00010973110019642938,
+      "loss": 0.9653,
+      "step": 5331
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.2020772397518158,
+      "learning_rate": 0.0001097035422464103,
+      "loss": 0.9211,
+      "step": 5332
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.20327343046665192,
+      "learning_rate": 0.00010967598355244,
+      "loss": 0.8538,
+      "step": 5333
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.19840683043003082,
+      "learning_rate": 0.00010964842411663128,
+      "loss": 0.8665,
+      "step": 5334
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.20020124316215515,
+      "learning_rate": 0.00010962086394109705,
+      "loss": 0.9099,
+      "step": 5335
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.2016928642988205,
+      "learning_rate": 0.0001095933030279504,
+      "loss": 0.862,
+      "step": 5336
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.21030563116073608,
+      "learning_rate": 0.00010956574137930429,
+      "loss": 0.9831,
+      "step": 5337
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.20427197217941284,
+      "learning_rate": 0.00010953817899727183,
+      "loss": 0.9434,
+      "step": 5338
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.1986006200313568,
+      "learning_rate": 0.00010951061588396618,
+      "loss": 0.8907,
+      "step": 5339
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.20330186188220978,
+      "learning_rate": 0.00010948305204150054,
+      "loss": 0.9612,
+      "step": 5340
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.20180857181549072,
+      "learning_rate": 0.00010945548747198819,
+      "loss": 0.9131,
+      "step": 5341
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.19510334730148315,
+      "learning_rate": 0.00010942792217754245,
+      "loss": 0.8674,
+      "step": 5342
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.20021146535873413,
+      "learning_rate": 0.00010940035616027669,
+      "loss": 0.9286,
+      "step": 5343
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.202460378408432,
+      "learning_rate": 0.00010937278942230435,
+      "loss": 0.9394,
+      "step": 5344
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.2063404768705368,
+      "learning_rate": 0.00010934522196573887,
+      "loss": 0.9157,
+      "step": 5345
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.2022324502468109,
+      "learning_rate": 0.00010931765379269385,
+      "loss": 0.9371,
+      "step": 5346
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.2042325735092163,
+      "learning_rate": 0.00010929008490528287,
+      "loss": 0.8429,
+      "step": 5347
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.20193937420845032,
+      "learning_rate": 0.0001092625153056196,
+      "loss": 0.8439,
+      "step": 5348
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.19789718091487885,
+      "learning_rate": 0.0001092349449958177,
+      "loss": 0.8861,
+      "step": 5349
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.20109151303768158,
+      "learning_rate": 0.00010920737397799095,
+      "loss": 0.8545,
+      "step": 5350
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.21629342436790466,
+      "learning_rate": 0.00010917980225425317,
+      "loss": 0.9521,
+      "step": 5351
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.21499894559383392,
+      "learning_rate": 0.00010915222982671825,
+      "loss": 0.9516,
+      "step": 5352
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.2034975290298462,
+      "learning_rate": 0.0001091246566975001,
+      "loss": 0.9194,
+      "step": 5353
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.19859300553798676,
+      "learning_rate": 0.00010909708286871265,
+      "loss": 0.8968,
+      "step": 5354
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.20240305364131927,
+      "learning_rate": 0.00010906950834246999,
+      "loss": 0.8883,
+      "step": 5355
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.21117694675922394,
+      "learning_rate": 0.0001090419331208862,
+      "loss": 0.9253,
+      "step": 5356
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.20424620807170868,
+      "learning_rate": 0.00010901435720607538,
+      "loss": 0.9334,
+      "step": 5357
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.19281210005283356,
+      "learning_rate": 0.00010898678060015175,
+      "loss": 0.8969,
+      "step": 5358
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.20645880699157715,
+      "learning_rate": 0.00010895920330522956,
+      "loss": 0.8592,
+      "step": 5359
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.18876774609088898,
+      "learning_rate": 0.00010893162532342305,
+      "loss": 0.8815,
+      "step": 5360
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.20588569343090057,
+      "learning_rate": 0.00010890404665684665,
+      "loss": 0.8889,
+      "step": 5361
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.20744764804840088,
+      "learning_rate": 0.00010887646730761472,
+      "loss": 0.8864,
+      "step": 5362
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.2112501859664917,
+      "learning_rate": 0.0001088488872778417,
+      "loss": 0.9765,
+      "step": 5363
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.1971355676651001,
+      "learning_rate": 0.00010882130656964213,
+      "loss": 0.9024,
+      "step": 5364
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.19987548887729645,
+      "learning_rate": 0.00010879372518513053,
+      "loss": 0.8502,
+      "step": 5365
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.21339647471904755,
+      "learning_rate": 0.00010876614312642153,
+      "loss": 0.9464,
+      "step": 5366
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.20979535579681396,
+      "learning_rate": 0.00010873856039562982,
+      "loss": 0.9526,
+      "step": 5367
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.20245172083377838,
+      "learning_rate": 0.00010871097699487005,
+      "loss": 0.9115,
+      "step": 5368
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.2003164291381836,
+      "learning_rate": 0.00010868339292625705,
+      "loss": 0.9226,
+      "step": 5369
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.2036878913640976,
+      "learning_rate": 0.00010865580819190557,
+      "loss": 0.8324,
+      "step": 5370
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.20090122520923615,
+      "learning_rate": 0.00010862822279393054,
+      "loss": 0.9138,
+      "step": 5371
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.20926058292388916,
+      "learning_rate": 0.00010860063673444685,
+      "loss": 0.9332,
+      "step": 5372
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.20082472264766693,
+      "learning_rate": 0.00010857305001556944,
+      "loss": 0.8328,
+      "step": 5373
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.21082501113414764,
+      "learning_rate": 0.00010854546263941339,
+      "loss": 0.9512,
+      "step": 5374
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.19284385442733765,
+      "learning_rate": 0.0001085178746080937,
+      "loss": 0.9419,
+      "step": 5375
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.19388435781002045,
+      "learning_rate": 0.00010849028592372555,
+      "loss": 0.9202,
+      "step": 5376
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.2086414396762848,
+      "learning_rate": 0.00010846269658842407,
+      "loss": 0.8658,
+      "step": 5377
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.20063738524913788,
+      "learning_rate": 0.00010843510660430447,
+      "loss": 0.9396,
+      "step": 5378
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.19891178607940674,
+      "learning_rate": 0.00010840751597348211,
+      "loss": 0.8995,
+      "step": 5379
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.20803341269493103,
+      "learning_rate": 0.00010837992469807218,
+      "loss": 0.9674,
+      "step": 5380
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.20508162677288055,
+      "learning_rate": 0.00010835233278019015,
+      "loss": 0.8785,
+      "step": 5381
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.19641000032424927,
+      "learning_rate": 0.00010832474022195138,
+      "loss": 0.8814,
+      "step": 5382
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.1972811073064804,
+      "learning_rate": 0.00010829714702547132,
+      "loss": 0.8922,
+      "step": 5383
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.2041017860174179,
+      "learning_rate": 0.00010826955319286558,
+      "loss": 0.9163,
+      "step": 5384
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.21064582467079163,
+      "learning_rate": 0.00010824195872624963,
+      "loss": 0.8693,
+      "step": 5385
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.20119690895080566,
+      "learning_rate": 0.00010821436362773912,
+      "loss": 1.0056,
+      "step": 5386
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.20159223675727844,
+      "learning_rate": 0.0001081867678994497,
+      "loss": 0.922,
+      "step": 5387
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.2144201546907425,
+      "learning_rate": 0.00010815917154349706,
+      "loss": 0.9242,
+      "step": 5388
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.20381437242031097,
+      "learning_rate": 0.00010813157456199705,
+      "loss": 0.9295,
+      "step": 5389
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.19609691202640533,
+      "learning_rate": 0.00010810397695706535,
+      "loss": 0.8845,
+      "step": 5390
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.20411543548107147,
+      "learning_rate": 0.0001080763787308179,
+      "loss": 0.9235,
+      "step": 5391
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.20571023225784302,
+      "learning_rate": 0.00010804877988537059,
+      "loss": 0.852,
+      "step": 5392
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.2107377052307129,
+      "learning_rate": 0.00010802118042283931,
+      "loss": 0.9347,
+      "step": 5393
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.19296841323375702,
+      "learning_rate": 0.00010799358034534015,
+      "loss": 0.7903,
+      "step": 5394
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.19987763464450836,
+      "learning_rate": 0.00010796597965498908,
+      "loss": 0.8674,
+      "step": 5395
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.19931264221668243,
+      "learning_rate": 0.00010793837835390219,
+      "loss": 0.914,
+      "step": 5396
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.20454499125480652,
+      "learning_rate": 0.00010791077644419567,
+      "loss": 0.8788,
+      "step": 5397
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.20866768062114716,
+      "learning_rate": 0.00010788317392798563,
+      "loss": 0.9846,
+      "step": 5398
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.20036864280700684,
+      "learning_rate": 0.00010785557080738839,
+      "loss": 0.8952,
+      "step": 5399
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.19880080223083496,
+      "learning_rate": 0.00010782796708452014,
+      "loss": 0.919,
+      "step": 5400
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.19589628279209137,
+      "learning_rate": 0.00010780036276149728,
+      "loss": 0.8344,
+      "step": 5401
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.20191796123981476,
+      "learning_rate": 0.00010777275784043613,
+      "loss": 0.8675,
+      "step": 5402
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.20033222436904907,
+      "learning_rate": 0.00010774515232345305,
+      "loss": 0.9171,
+      "step": 5403
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.20329776406288147,
+      "learning_rate": 0.00010771754621266466,
+      "loss": 0.8448,
+      "step": 5404
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.20034590363502502,
+      "learning_rate": 0.0001076899395101873,
+      "loss": 0.9414,
+      "step": 5405
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.20257985591888428,
+      "learning_rate": 0.00010766233221813762,
+      "loss": 0.8854,
+      "step": 5406
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.20048066973686218,
+      "learning_rate": 0.00010763472433863217,
+      "loss": 0.8535,
+      "step": 5407
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.20843522250652313,
+      "learning_rate": 0.0001076071158737876,
+      "loss": 0.9159,
+      "step": 5408
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.19864386320114136,
+      "learning_rate": 0.00010757950682572063,
+      "loss": 0.857,
+      "step": 5409
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.193053737282753,
+      "learning_rate": 0.00010755189719654795,
+      "loss": 0.8336,
+      "step": 5410
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.19992056488990784,
+      "learning_rate": 0.00010752428698838637,
+      "loss": 0.8091,
+      "step": 5411
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.2062869817018509,
+      "learning_rate": 0.00010749667620335266,
+      "loss": 0.8749,
+      "step": 5412
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.203649640083313,
+      "learning_rate": 0.00010746906484356372,
+      "loss": 0.8286,
+      "step": 5413
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.2094353884458542,
+      "learning_rate": 0.00010744145291113645,
+      "loss": 0.9006,
+      "step": 5414
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.2003667950630188,
+      "learning_rate": 0.00010741384040818782,
+      "loss": 0.8428,
+      "step": 5415
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.20398718118667603,
+      "learning_rate": 0.00010738622733683479,
+      "loss": 0.9237,
+      "step": 5416
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.19468474388122559,
+      "learning_rate": 0.00010735861369919442,
+      "loss": 0.8502,
+      "step": 5417
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.2077466994524002,
+      "learning_rate": 0.00010733099949738376,
+      "loss": 0.9258,
+      "step": 5418
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.19644780457019806,
+      "learning_rate": 0.00010730338473352001,
+      "loss": 0.8911,
+      "step": 5419
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.19907157123088837,
+      "learning_rate": 0.00010727576940972031,
+      "loss": 0.9216,
+      "step": 5420
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.21069294214248657,
+      "learning_rate": 0.00010724815352810183,
+      "loss": 0.9574,
+      "step": 5421
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.20110972225666046,
+      "learning_rate": 0.00010722053709078184,
+      "loss": 0.8654,
+      "step": 5422
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.2043222188949585,
+      "learning_rate": 0.00010719292009987765,
+      "loss": 0.8942,
+      "step": 5423
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.19561101496219635,
+      "learning_rate": 0.0001071653025575066,
+      "loss": 0.8844,
+      "step": 5424
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.21482384204864502,
+      "learning_rate": 0.00010713768446578612,
+      "loss": 0.868,
+      "step": 5425
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.1975470781326294,
+      "learning_rate": 0.00010711006582683352,
+      "loss": 0.8686,
+      "step": 5426
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.18916769325733185,
+      "learning_rate": 0.00010708244664276634,
+      "loss": 0.8201,
+      "step": 5427
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.20201200246810913,
+      "learning_rate": 0.00010705482691570207,
+      "loss": 0.9239,
+      "step": 5428
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.19807839393615723,
+      "learning_rate": 0.00010702720664775829,
+      "loss": 0.8829,
+      "step": 5429
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.1970471292734146,
+      "learning_rate": 0.00010699958584105261,
+      "loss": 0.9509,
+      "step": 5430
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.20313721895217896,
+      "learning_rate": 0.00010697196449770255,
+      "loss": 0.8714,
+      "step": 5431
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.19797174632549286,
+      "learning_rate": 0.00010694434261982589,
+      "loss": 0.9216,
+      "step": 5432
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.20096881687641144,
+      "learning_rate": 0.00010691672020954028,
+      "loss": 0.906,
+      "step": 5433
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.2070157676935196,
+      "learning_rate": 0.00010688909726896353,
+      "loss": 0.9213,
+      "step": 5434
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.20100905001163483,
+      "learning_rate": 0.00010686147380021342,
+      "loss": 0.8678,
+      "step": 5435
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.2077058106660843,
+      "learning_rate": 0.00010683384980540776,
+      "loss": 0.9615,
+      "step": 5436
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.21475407481193542,
+      "learning_rate": 0.00010680622528666444,
+      "loss": 0.967,
+      "step": 5437
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.2054007202386856,
+      "learning_rate": 0.00010677860024610139,
+      "loss": 0.8804,
+      "step": 5438
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.20920714735984802,
+      "learning_rate": 0.00010675097468583652,
+      "loss": 0.9073,
+      "step": 5439
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.19480057060718536,
+      "learning_rate": 0.00010672334860798794,
+      "loss": 0.8315,
+      "step": 5440
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.20092402398586273,
+      "learning_rate": 0.00010669572201467355,
+      "loss": 0.8936,
+      "step": 5441
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.20771557092666626,
+      "learning_rate": 0.00010666809490801148,
+      "loss": 0.8878,
+      "step": 5442
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.21026811003684998,
+      "learning_rate": 0.00010664046729011987,
+      "loss": 0.9235,
+      "step": 5443
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.21163447201251984,
+      "learning_rate": 0.00010661283916311684,
+      "loss": 0.9107,
+      "step": 5444
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.20571337640285492,
+      "learning_rate": 0.00010658521052912065,
+      "loss": 0.9288,
+      "step": 5445
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.20502333343029022,
+      "learning_rate": 0.00010655758139024942,
+      "loss": 0.8934,
+      "step": 5446
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.19831927120685577,
+      "learning_rate": 0.00010652995174862152,
+      "loss": 0.8669,
+      "step": 5447
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.21284537017345428,
+      "learning_rate": 0.00010650232160635519,
+      "loss": 0.8566,
+      "step": 5448
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.2107332944869995,
+      "learning_rate": 0.00010647469096556883,
+      "loss": 0.9208,
+      "step": 5449
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.19804082810878754,
+      "learning_rate": 0.00010644705982838081,
+      "loss": 0.9367,
+      "step": 5450
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.20850442349910736,
+      "learning_rate": 0.00010641942819690953,
+      "loss": 0.9054,
+      "step": 5451
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.20339353382587433,
+      "learning_rate": 0.0001063917960732735,
+      "loss": 0.9204,
+      "step": 5452
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.19770273566246033,
+      "learning_rate": 0.00010636416345959117,
+      "loss": 0.8832,
+      "step": 5453
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.21037791669368744,
+      "learning_rate": 0.0001063365303579811,
+      "loss": 0.9366,
+      "step": 5454
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.20975439250469208,
+      "learning_rate": 0.00010630889677056189,
+      "loss": 0.8848,
+      "step": 5455
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.21157856285572052,
+      "learning_rate": 0.00010628126269945211,
+      "loss": 0.9573,
+      "step": 5456
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.18961650133132935,
+      "learning_rate": 0.00010625362814677043,
+      "loss": 0.7911,
+      "step": 5457
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.2068662941455841,
+      "learning_rate": 0.0001062259931146355,
+      "loss": 0.9004,
+      "step": 5458
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.2038230299949646,
+      "learning_rate": 0.00010619835760516612,
+      "loss": 0.9452,
+      "step": 5459
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.20304733514785767,
+      "learning_rate": 0.00010617072162048099,
+      "loss": 0.8527,
+      "step": 5460
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.20015372335910797,
+      "learning_rate": 0.0001061430851626989,
+      "loss": 0.8984,
+      "step": 5461
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.2066597193479538,
+      "learning_rate": 0.00010611544823393873,
+      "loss": 0.9431,
+      "step": 5462
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.20975813269615173,
+      "learning_rate": 0.00010608781083631931,
+      "loss": 0.9342,
+      "step": 5463
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.1995689421892166,
+      "learning_rate": 0.00010606017297195956,
+      "loss": 0.9373,
+      "step": 5464
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.19717445969581604,
+      "learning_rate": 0.0001060325346429784,
+      "loss": 0.9466,
+      "step": 5465
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.2298421710729599,
+      "learning_rate": 0.00010600489585149484,
+      "loss": 0.9011,
+      "step": 5466
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.22285929322242737,
+      "learning_rate": 0.00010597725659962788,
+      "loss": 0.9999,
+      "step": 5467
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.2079205960035324,
+      "learning_rate": 0.00010594961688949654,
+      "loss": 0.9186,
+      "step": 5468
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.2025509625673294,
+      "learning_rate": 0.00010592197672321991,
+      "loss": 0.8173,
+      "step": 5469
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.1945144236087799,
+      "learning_rate": 0.00010589433610291713,
+      "loss": 0.8464,
+      "step": 5470
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.20176583528518677,
+      "learning_rate": 0.00010586669503070734,
+      "loss": 0.8429,
+      "step": 5471
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.1998482644557953,
+      "learning_rate": 0.00010583905350870971,
+      "loss": 0.8806,
+      "step": 5472
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.18746793270111084,
+      "learning_rate": 0.00010581141153904348,
+      "loss": 0.8106,
+      "step": 5473
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.2047126293182373,
+      "learning_rate": 0.0001057837691238279,
+      "loss": 0.9248,
+      "step": 5474
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.2051793783903122,
+      "learning_rate": 0.00010575612626518225,
+      "loss": 0.891,
+      "step": 5475
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.2004009336233139,
+      "learning_rate": 0.00010572848296522588,
+      "loss": 0.8177,
+      "step": 5476
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.20141753554344177,
+      "learning_rate": 0.00010570083922607812,
+      "loss": 0.8132,
+      "step": 5477
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.20456604659557343,
+      "learning_rate": 0.00010567319504985837,
+      "loss": 0.9394,
+      "step": 5478
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.2142971158027649,
+      "learning_rate": 0.00010564555043868604,
+      "loss": 0.9591,
+      "step": 5479
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.20223455131053925,
+      "learning_rate": 0.00010561790539468062,
+      "loss": 0.9081,
+      "step": 5480
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.20663003623485565,
+      "learning_rate": 0.00010559025991996157,
+      "loss": 0.8495,
+      "step": 5481
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.2013295292854309,
+      "learning_rate": 0.00010556261401664842,
+      "loss": 0.9142,
+      "step": 5482
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.20907741785049438,
+      "learning_rate": 0.00010553496768686073,
+      "loss": 0.958,
+      "step": 5483
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.21043746173381805,
+      "learning_rate": 0.00010550732093271807,
+      "loss": 0.9505,
+      "step": 5484
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.19993579387664795,
+      "learning_rate": 0.00010547967375634009,
+      "loss": 0.8621,
+      "step": 5485
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.21031905710697174,
+      "learning_rate": 0.00010545202615984646,
+      "loss": 0.9439,
+      "step": 5486
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.20142893493175507,
+      "learning_rate": 0.00010542437814535682,
+      "loss": 0.9008,
+      "step": 5487
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.20271191000938416,
+      "learning_rate": 0.00010539672971499093,
+      "loss": 0.9142,
+      "step": 5488
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.20329239964485168,
+      "learning_rate": 0.00010536908087086847,
+      "loss": 0.9426,
+      "step": 5489
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.20079709589481354,
+      "learning_rate": 0.0001053414316151093,
+      "loss": 0.9146,
+      "step": 5490
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.20103353261947632,
+      "learning_rate": 0.00010531378194983321,
+      "loss": 0.9292,
+      "step": 5491
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.20483773946762085,
+      "learning_rate": 0.00010528613187716004,
+      "loss": 0.9024,
+      "step": 5492
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.20483364164829254,
+      "learning_rate": 0.00010525848139920967,
+      "loss": 0.8938,
+      "step": 5493
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.19378747045993805,
+      "learning_rate": 0.00010523083051810197,
+      "loss": 0.884,
+      "step": 5494
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.20826518535614014,
+      "learning_rate": 0.00010520317923595693,
+      "loss": 0.9423,
+      "step": 5495
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.2042119950056076,
+      "learning_rate": 0.00010517552755489449,
+      "loss": 0.9201,
+      "step": 5496
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.2060500979423523,
+      "learning_rate": 0.00010514787547703466,
+      "loss": 0.8883,
+      "step": 5497
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.20850086212158203,
+      "learning_rate": 0.00010512022300449749,
+      "loss": 0.9031,
+      "step": 5498
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.21010854840278625,
+      "learning_rate": 0.00010509257013940299,
+      "loss": 0.9278,
+      "step": 5499
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.20167683064937592,
+      "learning_rate": 0.00010506491688387127,
+      "loss": 0.8522,
+      "step": 5500
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.20341619849205017,
+      "learning_rate": 0.00010503726324002248,
+      "loss": 0.8767,
+      "step": 5501
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.2075185924768448,
+      "learning_rate": 0.00010500960920997676,
+      "loss": 0.8936,
+      "step": 5502
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.20581576228141785,
+      "learning_rate": 0.00010498195479585427,
+      "loss": 0.8568,
+      "step": 5503
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.21758943796157837,
+      "learning_rate": 0.0001049542999997752,
+      "loss": 0.9707,
+      "step": 5504
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.19635866582393646,
+      "learning_rate": 0.00010492664482385981,
+      "loss": 0.8547,
+      "step": 5505
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.21904851496219635,
+      "learning_rate": 0.00010489898927022838,
+      "loss": 0.7793,
+      "step": 5506
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.2064557820558548,
+      "learning_rate": 0.00010487133334100124,
+      "loss": 0.9223,
+      "step": 5507
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.20519934594631195,
+      "learning_rate": 0.00010484367703829865,
+      "loss": 0.8152,
+      "step": 5508
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.20812737941741943,
+      "learning_rate": 0.00010481602036424094,
+      "loss": 0.8843,
+      "step": 5509
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.20696045458316803,
+      "learning_rate": 0.00010478836332094859,
+      "loss": 0.8851,
+      "step": 5510
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.20457100868225098,
+      "learning_rate": 0.00010476070591054193,
+      "loss": 0.913,
+      "step": 5511
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.2012094110250473,
+      "learning_rate": 0.00010473304813514146,
+      "loss": 0.8772,
+      "step": 5512
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.20374369621276855,
+      "learning_rate": 0.00010470538999686759,
+      "loss": 0.8856,
+      "step": 5513
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.20570087432861328,
+      "learning_rate": 0.00010467773149784083,
+      "loss": 0.9468,
+      "step": 5514
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.21451853215694427,
+      "learning_rate": 0.00010465007264018172,
+      "loss": 0.9318,
+      "step": 5515
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.20936594903469086,
+      "learning_rate": 0.00010462241342601081,
+      "loss": 0.8838,
+      "step": 5516
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.20236121118068695,
+      "learning_rate": 0.00010459475385744866,
+      "loss": 0.8472,
+      "step": 5517
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.20448645949363708,
+      "learning_rate": 0.0001045670939366159,
+      "loss": 0.8967,
+      "step": 5518
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.20693820714950562,
+      "learning_rate": 0.0001045394336656331,
+      "loss": 0.9088,
+      "step": 5519
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.20853427052497864,
+      "learning_rate": 0.00010451177304662098,
+      "loss": 0.9635,
+      "step": 5520
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.2139703333377838,
+      "learning_rate": 0.00010448411208170021,
+      "loss": 0.8683,
+      "step": 5521
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.196266308426857,
+      "learning_rate": 0.00010445645077299149,
+      "loss": 0.8463,
+      "step": 5522
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.20476581156253815,
+      "learning_rate": 0.00010442878912261558,
+      "loss": 0.936,
+      "step": 5523
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.20537997782230377,
+      "learning_rate": 0.00010440112713269319,
+      "loss": 0.9305,
+      "step": 5524
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.21176749467849731,
+      "learning_rate": 0.00010437346480534516,
+      "loss": 0.9282,
+      "step": 5525
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.19759652018547058,
+      "learning_rate": 0.0001043458021426923,
+      "loss": 0.9027,
+      "step": 5526
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.19534112513065338,
+      "learning_rate": 0.00010431813914685544,
+      "loss": 0.8246,
+      "step": 5527
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.2069014310836792,
+      "learning_rate": 0.00010429047581995546,
+      "loss": 0.9569,
+      "step": 5528
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.2005814164876938,
+      "learning_rate": 0.00010426281216411322,
+      "loss": 0.8368,
+      "step": 5529
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.20376819372177124,
+      "learning_rate": 0.00010423514818144968,
+      "loss": 0.9077,
+      "step": 5530
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.20055246353149414,
+      "learning_rate": 0.00010420748387408574,
+      "loss": 0.8813,
+      "step": 5531
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.20620344579219818,
+      "learning_rate": 0.00010417981924414242,
+      "loss": 0.8969,
+      "step": 5532
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.20845210552215576,
+      "learning_rate": 0.00010415215429374068,
+      "loss": 0.8923,
+      "step": 5533
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.20220451056957245,
+      "learning_rate": 0.00010412448902500149,
+      "loss": 0.8537,
+      "step": 5534
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.21402427554130554,
+      "learning_rate": 0.00010409682344004598,
+      "loss": 0.8981,
+      "step": 5535
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.20166133344173431,
+      "learning_rate": 0.00010406915754099518,
+      "loss": 0.8941,
+      "step": 5536
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.21470798552036285,
+      "learning_rate": 0.00010404149132997012,
+      "loss": 0.9071,
+      "step": 5537
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.2156185358762741,
+      "learning_rate": 0.00010401382480909205,
+      "loss": 0.9286,
+      "step": 5538
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.2048928141593933,
+      "learning_rate": 0.00010398615798048194,
+      "loss": 0.9145,
+      "step": 5539
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.21606525778770447,
+      "learning_rate": 0.00010395849084626111,
+      "loss": 0.9124,
+      "step": 5540
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.20806032419204712,
+      "learning_rate": 0.00010393082340855063,
+      "loss": 0.8577,
+      "step": 5541
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.2055501490831375,
+      "learning_rate": 0.00010390315566947174,
+      "loss": 0.905,
+      "step": 5542
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.22344693541526794,
+      "learning_rate": 0.0001038754876311457,
+      "loss": 0.8798,
+      "step": 5543
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.2084258794784546,
+      "learning_rate": 0.00010384781929569372,
+      "loss": 0.8989,
+      "step": 5544
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.20271384716033936,
+      "learning_rate": 0.00010382015066523709,
+      "loss": 0.8499,
+      "step": 5545
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.1985735446214676,
+      "learning_rate": 0.00010379248174189715,
+      "loss": 0.8993,
+      "step": 5546
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.2049076408147812,
+      "learning_rate": 0.00010376481252779513,
+      "loss": 0.8773,
+      "step": 5547
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.2088305801153183,
+      "learning_rate": 0.00010373714302505251,
+      "loss": 0.9891,
+      "step": 5548
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.2174626737833023,
+      "learning_rate": 0.0001037094732357905,
+      "loss": 0.9326,
+      "step": 5549
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.21114954352378845,
+      "learning_rate": 0.0001036818031621306,
+      "loss": 0.8854,
+      "step": 5550
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.2056587040424347,
+      "learning_rate": 0.00010365413280619418,
+      "loss": 0.9057,
+      "step": 5551
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.21228408813476562,
+      "learning_rate": 0.00010362646217010266,
+      "loss": 0.9768,
+      "step": 5552
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.2071719616651535,
+      "learning_rate": 0.00010359879125597753,
+      "loss": 0.9171,
+      "step": 5553
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.2048521190881729,
+      "learning_rate": 0.00010357112006594021,
+      "loss": 0.95,
+      "step": 5554
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.20829981565475464,
+      "learning_rate": 0.00010354344860211225,
+      "loss": 0.9078,
+      "step": 5555
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.21465882658958435,
+      "learning_rate": 0.00010351577686661513,
+      "loss": 0.8951,
+      "step": 5556
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.21603715419769287,
+      "learning_rate": 0.00010348810486157039,
+      "loss": 0.9098,
+      "step": 5557
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.20947815477848053,
+      "learning_rate": 0.00010346043258909963,
+      "loss": 0.897,
+      "step": 5558
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.19364146888256073,
+      "learning_rate": 0.00010343276005132436,
+      "loss": 0.8358,
+      "step": 5559
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.20746031403541565,
+      "learning_rate": 0.00010340508725036622,
+      "loss": 0.8454,
+      "step": 5560
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.22006148099899292,
+      "learning_rate": 0.00010337741418834684,
+      "loss": 0.8814,
+      "step": 5561
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.20638620853424072,
+      "learning_rate": 0.0001033497408673878,
+      "loss": 0.8643,
+      "step": 5562
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.20269517600536346,
+      "learning_rate": 0.00010332206728961085,
+      "loss": 0.8548,
+      "step": 5563
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.20134706795215607,
+      "learning_rate": 0.00010329439345713756,
+      "loss": 0.9191,
+      "step": 5564
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.19523276388645172,
+      "learning_rate": 0.00010326671937208971,
+      "loss": 0.8223,
+      "step": 5565
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.20336094498634338,
+      "learning_rate": 0.000103239045036589,
+      "loss": 0.9309,
+      "step": 5566
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.19728174805641174,
+      "learning_rate": 0.00010321137045275713,
+      "loss": 0.8479,
+      "step": 5567
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.20552708208560944,
+      "learning_rate": 0.00010318369562271593,
+      "loss": 0.8819,
+      "step": 5568
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.20273354649543762,
+      "learning_rate": 0.00010315602054858706,
+      "loss": 0.8672,
+      "step": 5569
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.2043936848640442,
+      "learning_rate": 0.00010312834523249243,
+      "loss": 0.8995,
+      "step": 5570
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.20490287244319916,
+      "learning_rate": 0.00010310066967655378,
+      "loss": 0.8514,
+      "step": 5571
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.19761638343334198,
+      "learning_rate": 0.00010307299388289296,
+      "loss": 0.8673,
+      "step": 5572
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.1997511088848114,
+      "learning_rate": 0.00010304531785363183,
+      "loss": 0.8748,
+      "step": 5573
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.2093546837568283,
+      "learning_rate": 0.00010301764159089222,
+      "loss": 0.9083,
+      "step": 5574
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.20290273427963257,
+      "learning_rate": 0.00010298996509679606,
+      "loss": 0.8587,
+      "step": 5575
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.20612122118473053,
+      "learning_rate": 0.00010296228837346522,
+      "loss": 0.9561,
+      "step": 5576
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.2027217149734497,
+      "learning_rate": 0.00010293461142302163,
+      "loss": 0.893,
+      "step": 5577
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.20365840196609497,
+      "learning_rate": 0.00010290693424758722,
+      "loss": 0.8572,
+      "step": 5578
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.20379266142845154,
+      "learning_rate": 0.00010287925684928396,
+      "loss": 0.9155,
+      "step": 5579
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.20149919390678406,
+      "learning_rate": 0.00010285157923023381,
+      "loss": 0.9137,
+      "step": 5580
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.20707061886787415,
+      "learning_rate": 0.00010282390139255878,
+      "loss": 0.9025,
+      "step": 5581
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.20673343539237976,
+      "learning_rate": 0.0001027962233383808,
+      "loss": 0.9554,
+      "step": 5582
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.21456193923950195,
+      "learning_rate": 0.00010276854506982196,
+      "loss": 0.9115,
+      "step": 5583
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.2044607400894165,
+      "learning_rate": 0.0001027408665890043,
+      "loss": 0.8923,
+      "step": 5584
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.19878527522087097,
+      "learning_rate": 0.00010271318789804987,
+      "loss": 0.9447,
+      "step": 5585
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.19507496058940887,
+      "learning_rate": 0.00010268550899908071,
+      "loss": 0.8844,
+      "step": 5586
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.1981414258480072,
+      "learning_rate": 0.0001026578298942189,
+      "loss": 0.8851,
+      "step": 5587
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.20765522122383118,
+      "learning_rate": 0.00010263015058558659,
+      "loss": 0.9509,
+      "step": 5588
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.20648297667503357,
+      "learning_rate": 0.00010260247107530587,
+      "loss": 0.8989,
+      "step": 5589
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.19959251582622528,
+      "learning_rate": 0.00010257479136549889,
+      "loss": 0.864,
+      "step": 5590
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.19457371532917023,
+      "learning_rate": 0.00010254711145828774,
+      "loss": 0.8499,
+      "step": 5591
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.21515204012393951,
+      "learning_rate": 0.00010251943135579464,
+      "loss": 0.9099,
+      "step": 5592
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.20317937433719635,
+      "learning_rate": 0.00010249175106014178,
+      "loss": 0.9047,
+      "step": 5593
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.19348680973052979,
+      "learning_rate": 0.00010246407057345131,
+      "loss": 0.8087,
+      "step": 5594
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.19511255621910095,
+      "learning_rate": 0.00010243638989784545,
+      "loss": 0.8805,
+      "step": 5595
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.20101837813854218,
+      "learning_rate": 0.00010240870903544644,
+      "loss": 0.8821,
+      "step": 5596
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.19923505187034607,
+      "learning_rate": 0.00010238102798837648,
+      "loss": 0.8281,
+      "step": 5597
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.20761752128601074,
+      "learning_rate": 0.00010235334675875787,
+      "loss": 0.9093,
+      "step": 5598
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.2042294442653656,
+      "learning_rate": 0.00010232566534871286,
+      "loss": 0.8987,
+      "step": 5599
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.20492446422576904,
+      "learning_rate": 0.0001022979837603637,
+      "loss": 0.8391,
+      "step": 5600
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.2098143994808197,
+      "learning_rate": 0.00010227030199583271,
+      "loss": 0.9079,
+      "step": 5601
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.20019160211086273,
+      "learning_rate": 0.00010224262005724217,
+      "loss": 0.8631,
+      "step": 5602
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.21009786427021027,
+      "learning_rate": 0.00010221493794671442,
+      "loss": 0.9236,
+      "step": 5603
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.20378471910953522,
+      "learning_rate": 0.0001021872556663718,
+      "loss": 0.9148,
+      "step": 5604
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.20996125042438507,
+      "learning_rate": 0.00010215957321833664,
+      "loss": 0.9775,
+      "step": 5605
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.19774721562862396,
+      "learning_rate": 0.0001021318906047313,
+      "loss": 0.9339,
+      "step": 5606
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.2106465995311737,
+      "learning_rate": 0.00010210420782767813,
+      "loss": 0.8776,
+      "step": 5607
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.20622365176677704,
+      "learning_rate": 0.00010207652488929958,
+      "loss": 0.9095,
+      "step": 5608
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.20106709003448486,
+      "learning_rate": 0.00010204884179171797,
+      "loss": 0.9572,
+      "step": 5609
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.20053446292877197,
+      "learning_rate": 0.00010202115853705576,
+      "loss": 0.8891,
+      "step": 5610
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.21232885122299194,
+      "learning_rate": 0.00010199347512743533,
+      "loss": 0.8949,
+      "step": 5611
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.2083313763141632,
+      "learning_rate": 0.00010196579156497912,
+      "loss": 0.8339,
+      "step": 5612
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.20660704374313354,
+      "learning_rate": 0.00010193810785180962,
+      "loss": 0.8816,
+      "step": 5613
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.19979746639728546,
+      "learning_rate": 0.00010191042399004922,
+      "loss": 0.894,
+      "step": 5614
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.20635730028152466,
+      "learning_rate": 0.00010188273998182045,
+      "loss": 0.922,
+      "step": 5615
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.21255262196063995,
+      "learning_rate": 0.00010185505582924573,
+      "loss": 0.9356,
+      "step": 5616
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.19595664739608765,
+      "learning_rate": 0.00010182737153444758,
+      "loss": 0.8222,
+      "step": 5617
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.2002105861902237,
+      "learning_rate": 0.00010179968709954847,
+      "loss": 0.9032,
+      "step": 5618
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.2023312747478485,
+      "learning_rate": 0.00010177200252667096,
+      "loss": 0.8727,
+      "step": 5619
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.2006431221961975,
+      "learning_rate": 0.00010174431781793754,
+      "loss": 0.8597,
+      "step": 5620
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.20587624609470367,
+      "learning_rate": 0.00010171663297547076,
+      "loss": 0.8933,
+      "step": 5621
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.2047545611858368,
+      "learning_rate": 0.0001016889480013931,
+      "loss": 0.9077,
+      "step": 5622
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.2119145542383194,
+      "learning_rate": 0.00010166126289782721,
+      "loss": 0.8601,
+      "step": 5623
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.20192745327949524,
+      "learning_rate": 0.00010163357766689558,
+      "loss": 0.8753,
+      "step": 5624
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.20670832693576813,
+      "learning_rate": 0.00010160589231072082,
+      "loss": 0.9099,
+      "step": 5625
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.21294564008712769,
+      "learning_rate": 0.00010157820683142549,
+      "loss": 0.8567,
+      "step": 5626
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.194480761885643,
+      "learning_rate": 0.00010155052123113215,
+      "loss": 0.8606,
+      "step": 5627
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.20041505992412567,
+      "learning_rate": 0.00010152283551196347,
+      "loss": 0.925,
+      "step": 5628
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.20152615010738373,
+      "learning_rate": 0.00010149514967604202,
+      "loss": 0.8863,
+      "step": 5629
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.20309147238731384,
+      "learning_rate": 0.00010146746372549043,
+      "loss": 0.8359,
+      "step": 5630
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.21429483592510223,
+      "learning_rate": 0.00010143977766243129,
+      "loss": 1.0148,
+      "step": 5631
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.19496549665927887,
+      "learning_rate": 0.00010141209148898727,
+      "loss": 0.9032,
+      "step": 5632
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.2035844326019287,
+      "learning_rate": 0.00010138440520728101,
+      "loss": 0.8656,
+      "step": 5633
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.19976894557476044,
+      "learning_rate": 0.00010135671881943518,
+      "loss": 0.8728,
+      "step": 5634
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.20249591767787933,
+      "learning_rate": 0.0001013290323275724,
+      "loss": 0.8846,
+      "step": 5635
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.1962704211473465,
+      "learning_rate": 0.00010130134573381538,
+      "loss": 0.8569,
+      "step": 5636
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.2182995229959488,
+      "learning_rate": 0.00010127365904028672,
+      "loss": 0.9651,
+      "step": 5637
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.18703502416610718,
+      "learning_rate": 0.00010124597224910921,
+      "loss": 0.7896,
+      "step": 5638
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.2021150439977646,
+      "learning_rate": 0.00010121828536240548,
+      "loss": 0.8708,
+      "step": 5639
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.21872109174728394,
+      "learning_rate": 0.00010119059838229823,
+      "loss": 0.886,
+      "step": 5640
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.20219261944293976,
+      "learning_rate": 0.00010116291131091016,
+      "loss": 0.9473,
+      "step": 5641
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.20678988099098206,
+      "learning_rate": 0.000101135224150364,
+      "loss": 0.8906,
+      "step": 5642
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.1961716264486313,
+      "learning_rate": 0.00010110753690278248,
+      "loss": 0.8878,
+      "step": 5643
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.2135319858789444,
+      "learning_rate": 0.00010107984957028829,
+      "loss": 0.956,
+      "step": 5644
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.20089684426784515,
+      "learning_rate": 0.00010105216215500422,
+      "loss": 0.8844,
+      "step": 5645
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.19634021818637848,
+      "learning_rate": 0.00010102447465905293,
+      "loss": 0.872,
+      "step": 5646
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.20131458342075348,
+      "learning_rate": 0.00010099678708455722,
+      "loss": 0.8583,
+      "step": 5647
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.20227965712547302,
+      "learning_rate": 0.00010096909943363983,
+      "loss": 0.9438,
+      "step": 5648
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.2081882357597351,
+      "learning_rate": 0.00010094141170842353,
+      "loss": 0.9539,
+      "step": 5649
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.20764583349227905,
+      "learning_rate": 0.00010091372391103107,
+      "loss": 0.8397,
+      "step": 5650
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.19473712146282196,
+      "learning_rate": 0.00010088603604358521,
+      "loss": 0.8871,
+      "step": 5651
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.2055901437997818,
+      "learning_rate": 0.00010085834810820871,
+      "loss": 0.9054,
+      "step": 5652
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.20546817779541016,
+      "learning_rate": 0.0001008306601070244,
+      "loss": 0.9194,
+      "step": 5653
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.2075883150100708,
+      "learning_rate": 0.00010080297204215501,
+      "loss": 0.9154,
+      "step": 5654
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.1996559500694275,
+      "learning_rate": 0.00010077528391572338,
+      "loss": 0.8629,
+      "step": 5655
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.20948490500450134,
+      "learning_rate": 0.00010074759572985227,
+      "loss": 0.9011,
+      "step": 5656
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.19752012193202972,
+      "learning_rate": 0.00010071990748666445,
+      "loss": 0.9041,
+      "step": 5657
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.2031921148300171,
+      "learning_rate": 0.00010069221918828278,
+      "loss": 0.8492,
+      "step": 5658
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.20917463302612305,
+      "learning_rate": 0.00010066453083683006,
+      "loss": 0.9072,
+      "step": 5659
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.21156203746795654,
+      "learning_rate": 0.00010063684243442906,
+      "loss": 0.8781,
+      "step": 5660
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.2049502581357956,
+      "learning_rate": 0.00010060915398320262,
+      "loss": 0.9036,
+      "step": 5661
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.2090534120798111,
+      "learning_rate": 0.00010058146548527357,
+      "loss": 0.8892,
+      "step": 5662
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.20506758987903595,
+      "learning_rate": 0.0001005537769427647,
+      "loss": 0.8151,
+      "step": 5663
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.20771467685699463,
+      "learning_rate": 0.00010052608835779887,
+      "loss": 0.9167,
+      "step": 5664
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.21362152695655823,
+      "learning_rate": 0.00010049839973249888,
+      "loss": 0.9169,
+      "step": 5665
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.19874230027198792,
+      "learning_rate": 0.0001004707110689876,
+      "loss": 0.8439,
+      "step": 5666
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.19645732641220093,
+      "learning_rate": 0.00010044302236938781,
+      "loss": 0.8103,
+      "step": 5667
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.2064276784658432,
+      "learning_rate": 0.0001004153336358224,
+      "loss": 0.8839,
+      "step": 5668
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.1985475718975067,
+      "learning_rate": 0.00010038764487041419,
+      "loss": 0.8588,
+      "step": 5669
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.19935201108455658,
+      "learning_rate": 0.00010035995607528599,
+      "loss": 0.8933,
+      "step": 5670
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.20382753014564514,
+      "learning_rate": 0.00010033226725256071,
+      "loss": 0.9026,
+      "step": 5671
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.2041386365890503,
+      "learning_rate": 0.00010030457840436112,
+      "loss": 0.9249,
+      "step": 5672
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.20588405430316925,
+      "learning_rate": 0.00010027688953281014,
+      "loss": 0.9808,
+      "step": 5673
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.20263512432575226,
+      "learning_rate": 0.00010024920064003059,
+      "loss": 0.8611,
+      "step": 5674
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.20413020253181458,
+      "learning_rate": 0.0001002215117281453,
+      "loss": 0.8908,
+      "step": 5675
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.20494332909584045,
+      "learning_rate": 0.00010019382279927718,
+      "loss": 0.9052,
+      "step": 5676
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.2025218904018402,
+      "learning_rate": 0.00010016613385554903,
+      "loss": 0.891,
+      "step": 5677
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.20399563014507294,
+      "learning_rate": 0.00010013844489908376,
+      "loss": 0.9221,
+      "step": 5678
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 11356,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 4,
+  "save_steps": 2839,
+  "total_flos": 1.183008524297306e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}