diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,17536 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9970005998800238,
+  "eval_steps": 500,
+  "global_step": 2499,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0011997600479904018,
+      "grad_norm": 36.17856378068024,
+      "learning_rate": 0.0,
+      "loss": 6.5806,
+      "step": 1
+    },
+    {
+      "epoch": 0.0023995200959808036,
+      "grad_norm": 34.358282645308925,
+      "learning_rate": 4e-08,
+      "loss": 5.5196,
+      "step": 2
+    },
+    {
+      "epoch": 0.003599280143971206,
+      "grad_norm": 34.63248495550367,
+      "learning_rate": 8e-08,
+      "loss": 5.6467,
+      "step": 3
+    },
+    {
+      "epoch": 0.004799040191961607,
+      "grad_norm": 37.406685178178456,
+      "learning_rate": 1.2000000000000002e-07,
+      "loss": 7.1574,
+      "step": 4
+    },
+    {
+      "epoch": 0.00599880023995201,
+      "grad_norm": 38.40041370330666,
+      "learning_rate": 1.6e-07,
+      "loss": 5.0629,
+      "step": 5
+    },
+    {
+      "epoch": 0.007198560287942412,
+      "grad_norm": 32.30499073167661,
+      "learning_rate": 2.0000000000000002e-07,
+      "loss": 5.4575,
+      "step": 6
+    },
+    {
+      "epoch": 0.008398320335932814,
+      "grad_norm": 37.67738186414492,
+      "learning_rate": 2.4000000000000003e-07,
+      "loss": 6.8774,
+      "step": 7
+    },
+    {
+      "epoch": 0.009598080383923215,
+      "grad_norm": 29.847807359572833,
+      "learning_rate": 2.8e-07,
+      "loss": 5.0336,
+      "step": 8
+    },
+    {
+      "epoch": 0.010797840431913617,
+      "grad_norm": 33.87761978237807,
+      "learning_rate": 3.2e-07,
+      "loss": 5.3535,
+      "step": 9
+    },
+    {
+      "epoch": 0.01199760047990402,
+      "grad_norm": 36.02370970827611,
+      "learning_rate": 3.6e-07,
+      "loss": 6.5137,
+      "step": 10
+    },
+    {
+      "epoch": 0.013197360527894421,
+      "grad_norm": 30.794586404273485,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 4.6923,
+      "step": 11
+    },
+    {
+      "epoch": 0.014397120575884824,
+      "grad_norm": 36.095798724282986,
+      "learning_rate": 4.4e-07,
+      "loss": 5.5734,
+      "step": 12
+    },
+    {
+      "epoch": 0.015596880623875225,
+      "grad_norm": 36.625714284092496,
+      "learning_rate": 4.800000000000001e-07,
+      "loss": 6.2458,
+      "step": 13
+    },
+    {
+      "epoch": 0.016796640671865627,
+      "grad_norm": 38.38403373240824,
+      "learning_rate": 5.2e-07,
+      "loss": 6.8462,
+      "step": 14
+    },
+    {
+      "epoch": 0.017996400719856028,
+      "grad_norm": 36.381283291167584,
+      "learning_rate": 5.6e-07,
+      "loss": 6.8765,
+      "step": 15
+    },
+    {
+      "epoch": 0.01919616076784643,
+      "grad_norm": 32.176585157359426,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 5.0196,
+      "step": 16
+    },
+    {
+      "epoch": 0.020395920815836834,
+      "grad_norm": 36.517388905365706,
+      "learning_rate": 6.4e-07,
+      "loss": 7.2797,
+      "step": 17
+    },
+    {
+      "epoch": 0.021595680863827234,
+      "grad_norm": 33.5867632244418,
+      "learning_rate": 6.800000000000001e-07,
+      "loss": 4.8825,
+      "step": 18
+    },
+    {
+      "epoch": 0.022795440911817635,
+      "grad_norm": 37.36761591767671,
+      "learning_rate": 7.2e-07,
+      "loss": 6.9421,
+      "step": 19
+    },
+    {
+      "epoch": 0.02399520095980804,
+      "grad_norm": 34.62445788638932,
+      "learning_rate": 7.6e-07,
+      "loss": 5.6175,
+      "step": 20
+    },
+    {
+      "epoch": 0.02519496100779844,
+      "grad_norm": 35.15779448273933,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 5.8947,
+      "step": 21
+    },
+    {
+      "epoch": 0.026394721055788842,
+      "grad_norm": 29.819457245845634,
+      "learning_rate": 8.400000000000001e-07,
+      "loss": 4.8043,
+      "step": 22
+    },
+    {
+      "epoch": 0.027594481103779243,
+      "grad_norm": 33.64012081915802,
+      "learning_rate": 8.8e-07,
+      "loss": 5.4804,
+      "step": 23
+    },
+    {
+      "epoch": 0.028794241151769647,
+      "grad_norm": 35.618352356220015,
+      "learning_rate": 9.200000000000001e-07,
+      "loss": 6.6431,
+      "step": 24
+    },
+    {
+      "epoch": 0.029994001199760048,
+      "grad_norm": 34.112079688785066,
+      "learning_rate": 9.600000000000001e-07,
+      "loss": 6.0226,
+      "step": 25
+    },
+    {
+      "epoch": 0.03119376124775045,
+      "grad_norm": 30.405498175713877,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 5.2086,
+      "step": 26
+    },
+    {
+      "epoch": 0.03239352129574085,
+      "grad_norm": 26.22721194386834,
+      "learning_rate": 1.04e-06,
+      "loss": 3.9648,
+      "step": 27
+    },
+    {
+      "epoch": 0.033593281343731254,
+      "grad_norm": 32.474716650442616,
+      "learning_rate": 1.08e-06,
+      "loss": 5.42,
+      "step": 28
+    },
+    {
+      "epoch": 0.03479304139172166,
+      "grad_norm": 29.509812337295603,
+      "learning_rate": 1.12e-06,
+      "loss": 4.4588,
+      "step": 29
+    },
+    {
+      "epoch": 0.035992801439712056,
+      "grad_norm": 25.26807361414514,
+      "learning_rate": 1.1600000000000001e-06,
+      "loss": 4.4965,
+      "step": 30
+    },
+    {
+      "epoch": 0.03719256148770246,
+      "grad_norm": 17.211406917325288,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 3.1409,
+      "step": 31
+    },
+    {
+      "epoch": 0.03839232153569286,
+      "grad_norm": 16.58496319508288,
+      "learning_rate": 1.2400000000000002e-06,
+      "loss": 3.2392,
+      "step": 32
+    },
+    {
+      "epoch": 0.03959208158368326,
+      "grad_norm": 20.256709229042606,
+      "learning_rate": 1.28e-06,
+      "loss": 3.9943,
+      "step": 33
+    },
+    {
+      "epoch": 0.04079184163167367,
+      "grad_norm": 19.386587675168716,
+      "learning_rate": 1.32e-06,
+      "loss": 4.0614,
+      "step": 34
+    },
+    {
+      "epoch": 0.041991601679664065,
+      "grad_norm": 21.31388017416675,
+      "learning_rate": 1.3600000000000001e-06,
+      "loss": 4.8848,
+      "step": 35
+    },
+    {
+      "epoch": 0.04319136172765447,
+      "grad_norm": 16.440197605883533,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 4.1091,
+      "step": 36
+    },
+    {
+      "epoch": 0.04439112177564487,
+      "grad_norm": 14.127689181062076,
+      "learning_rate": 1.44e-06,
+      "loss": 3.9764,
+      "step": 37
+    },
+    {
+      "epoch": 0.04559088182363527,
+      "grad_norm": 11.382233478930736,
+      "learning_rate": 1.48e-06,
+      "loss": 3.1144,
+      "step": 38
+    },
+    {
+      "epoch": 0.046790641871625675,
+      "grad_norm": 10.94943616284639,
+      "learning_rate": 1.52e-06,
+      "loss": 3.4877,
+      "step": 39
+    },
+    {
+      "epoch": 0.04799040191961608,
+      "grad_norm": 8.0400059342567,
+      "learning_rate": 1.56e-06,
+      "loss": 3.048,
+      "step": 40
+    },
+    {
+      "epoch": 0.04919016196760648,
+      "grad_norm": 13.04940759359585,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 3.2501,
+      "step": 41
+    },
+    {
+      "epoch": 0.05038992201559688,
+      "grad_norm": 11.668121415877163,
+      "learning_rate": 1.6400000000000002e-06,
+      "loss": 2.6705,
+      "step": 42
+    },
+    {
+      "epoch": 0.05158968206358728,
+      "grad_norm": 15.351939200735895,
+      "learning_rate": 1.6800000000000002e-06,
+      "loss": 3.3811,
+      "step": 43
+    },
+    {
+      "epoch": 0.052789442111577684,
+      "grad_norm": 9.517039626972272,
+      "learning_rate": 1.72e-06,
+      "loss": 3.0159,
+      "step": 44
+    },
+    {
+      "epoch": 0.05398920215956809,
+      "grad_norm": 9.580365005252844,
+      "learning_rate": 1.76e-06,
+      "loss": 3.2343,
+      "step": 45
+    },
+    {
+      "epoch": 0.055188962207558485,
+      "grad_norm": 5.1751309603875235,
+      "learning_rate": 1.8000000000000001e-06,
+      "loss": 2.7038,
+      "step": 46
+    },
+    {
+      "epoch": 0.05638872225554889,
+      "grad_norm": 5.6080543773161775,
+      "learning_rate": 1.8400000000000002e-06,
+      "loss": 3.1211,
+      "step": 47
+    },
+    {
+      "epoch": 0.057588482303539294,
+      "grad_norm": 4.310981133231568,
+      "learning_rate": 1.8800000000000002e-06,
+      "loss": 2.6889,
+      "step": 48
+    },
+    {
+      "epoch": 0.05878824235152969,
+      "grad_norm": 4.3584143180433745,
+      "learning_rate": 1.9200000000000003e-06,
+      "loss": 2.5729,
+      "step": 49
+    },
+    {
+      "epoch": 0.059988002399520096,
+      "grad_norm": 5.498808904661779,
+      "learning_rate": 1.9600000000000003e-06,
+      "loss": 2.8867,
+      "step": 50
+    },
+    {
+      "epoch": 0.0611877624475105,
+      "grad_norm": 2.6927761049069723,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.6145,
+      "step": 51
+    },
+    {
+      "epoch": 0.0623875224955009,
+      "grad_norm": 5.4214774551834735,
+      "learning_rate": 2.04e-06,
+      "loss": 3.1939,
+      "step": 52
+    },
+    {
+      "epoch": 0.0635872825434913,
+      "grad_norm": 5.406801526942302,
+      "learning_rate": 2.08e-06,
+      "loss": 2.6354,
+      "step": 53
+    },
+    {
+      "epoch": 0.0647870425914817,
+      "grad_norm": 3.714353613351617,
+      "learning_rate": 2.12e-06,
+      "loss": 2.5379,
+      "step": 54
+    },
+    {
+      "epoch": 0.06598680263947211,
+      "grad_norm": 2.9173663728629085,
+      "learning_rate": 2.16e-06,
+      "loss": 2.1469,
+      "step": 55
+    },
+    {
+      "epoch": 0.06718656268746251,
+      "grad_norm": 3.608786726950732,
+      "learning_rate": 2.2e-06,
+      "loss": 2.6828,
+      "step": 56
+    },
+    {
+      "epoch": 0.0683863227354529,
+      "grad_norm": 3.763597670515027,
+      "learning_rate": 2.24e-06,
+      "loss": 2.6978,
+      "step": 57
+    },
+    {
+      "epoch": 0.06958608278344332,
+      "grad_norm": 3.4041860826087795,
+      "learning_rate": 2.28e-06,
+      "loss": 2.3874,
+      "step": 58
+    },
+    {
+      "epoch": 0.07078584283143372,
+      "grad_norm": 3.4740369457844755,
+      "learning_rate": 2.3200000000000002e-06,
+      "loss": 2.9403,
+      "step": 59
+    },
+    {
+      "epoch": 0.07198560287942411,
+      "grad_norm": 3.307937151104892,
+      "learning_rate": 2.3600000000000003e-06,
+      "loss": 2.6198,
+      "step": 60
+    },
+    {
+      "epoch": 0.07318536292741452,
+      "grad_norm": 2.7933319240440566,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 2.2052,
+      "step": 61
+    },
+    {
+      "epoch": 0.07438512297540492,
+      "grad_norm": 2.5093361221749064,
+      "learning_rate": 2.4400000000000004e-06,
+      "loss": 2.0,
+      "step": 62
+    },
+    {
+      "epoch": 0.07558488302339532,
+      "grad_norm": 3.7731916081151016,
+      "learning_rate": 2.4800000000000004e-06,
+      "loss": 2.5814,
+      "step": 63
+    },
+    {
+      "epoch": 0.07678464307138572,
+      "grad_norm": 2.7883126052257388,
+      "learning_rate": 2.52e-06,
+      "loss": 2.2478,
+      "step": 64
+    },
+    {
+      "epoch": 0.07798440311937613,
+      "grad_norm": 2.973289118741531,
+      "learning_rate": 2.56e-06,
+      "loss": 2.6007,
+      "step": 65
+    },
+    {
+      "epoch": 0.07918416316736653,
+      "grad_norm": 2.6177891365858152,
+      "learning_rate": 2.6e-06,
+      "loss": 2.3398,
+      "step": 66
+    },
+    {
+      "epoch": 0.08038392321535692,
+      "grad_norm": 3.5881470293988573,
+      "learning_rate": 2.64e-06,
+      "loss": 2.3946,
+      "step": 67
+    },
+    {
+      "epoch": 0.08158368326334733,
+      "grad_norm": 2.3121432357226976,
+      "learning_rate": 2.68e-06,
+      "loss": 1.8217,
+      "step": 68
+    },
+    {
+      "epoch": 0.08278344331133773,
+      "grad_norm": 2.9061157483024544,
+      "learning_rate": 2.7200000000000002e-06,
+      "loss": 2.1346,
+      "step": 69
+    },
+    {
+      "epoch": 0.08398320335932813,
+      "grad_norm": 3.0322906921976975,
+      "learning_rate": 2.7600000000000003e-06,
+      "loss": 2.3916,
+      "step": 70
+    },
+    {
+      "epoch": 0.08518296340731854,
+      "grad_norm": 3.0090713796463433,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 2.0897,
+      "step": 71
+    },
+    {
+      "epoch": 0.08638272345530894,
+      "grad_norm": 2.5732336424427413,
+      "learning_rate": 2.84e-06,
+      "loss": 1.8334,
+      "step": 72
+    },
+    {
+      "epoch": 0.08758248350329934,
+      "grad_norm": 2.511452530855231,
+      "learning_rate": 2.88e-06,
+      "loss": 1.8359,
+      "step": 73
+    },
+    {
+      "epoch": 0.08878224355128975,
+      "grad_norm": 2.1722861397430075,
+      "learning_rate": 2.92e-06,
+      "loss": 1.7097,
+      "step": 74
+    },
+    {
+      "epoch": 0.08998200359928014,
+      "grad_norm": 2.2110913265788446,
+      "learning_rate": 2.96e-06,
+      "loss": 1.8823,
+      "step": 75
+    },
+    {
+      "epoch": 0.09118176364727054,
+      "grad_norm": 3.1903636836969853,
+      "learning_rate": 3e-06,
+      "loss": 2.1091,
+      "step": 76
+    },
+    {
+      "epoch": 0.09238152369526095,
+      "grad_norm": 2.139050909268807,
+      "learning_rate": 3.04e-06,
+      "loss": 1.7386,
+      "step": 77
+    },
+    {
+      "epoch": 0.09358128374325135,
+      "grad_norm": 2.781821724211591,
+      "learning_rate": 3.08e-06,
+      "loss": 1.7271,
+      "step": 78
+    },
+    {
+      "epoch": 0.09478104379124175,
+      "grad_norm": 2.1741472134292894,
+      "learning_rate": 3.12e-06,
+      "loss": 1.7486,
+      "step": 79
+    },
+    {
+      "epoch": 0.09598080383923216,
+      "grad_norm": 2.373728486452905,
+      "learning_rate": 3.1600000000000002e-06,
+      "loss": 1.9319,
+      "step": 80
+    },
+    {
+      "epoch": 0.09718056388722256,
+      "grad_norm": 1.7379021917646496,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 1.4739,
+      "step": 81
+    },
+    {
+      "epoch": 0.09838032393521295,
+      "grad_norm": 1.7320610918875294,
+      "learning_rate": 3.2400000000000003e-06,
+      "loss": 1.5421,
+      "step": 82
+    },
+    {
+      "epoch": 0.09958008398320337,
+      "grad_norm": 2.144805174214645,
+      "learning_rate": 3.2800000000000004e-06,
+      "loss": 1.2531,
+      "step": 83
+    },
+    {
+      "epoch": 0.10077984403119376,
+      "grad_norm": 1.4760948538526937,
+      "learning_rate": 3.3200000000000004e-06,
+      "loss": 1.3801,
+      "step": 84
+    },
+    {
+      "epoch": 0.10197960407918416,
+      "grad_norm": 1.6876558775793726,
+      "learning_rate": 3.3600000000000004e-06,
+      "loss": 1.5244,
+      "step": 85
+    },
+    {
+      "epoch": 0.10317936412717456,
+      "grad_norm": 2.6308562265498656,
+      "learning_rate": 3.4000000000000005e-06,
+      "loss": 1.6274,
+      "step": 86
+    },
+    {
+      "epoch": 0.10437912417516497,
+      "grad_norm": 2.079207911605781,
+      "learning_rate": 3.44e-06,
+      "loss": 1.4456,
+      "step": 87
+    },
+    {
+      "epoch": 0.10557888422315537,
+      "grad_norm": 1.949221335018367,
+      "learning_rate": 3.48e-06,
+      "loss": 1.5871,
+      "step": 88
+    },
+    {
+      "epoch": 0.10677864427114576,
+      "grad_norm": 1.8368415904194118,
+      "learning_rate": 3.52e-06,
+      "loss": 1.3269,
+      "step": 89
+    },
+    {
+      "epoch": 0.10797840431913618,
+      "grad_norm": 1.7700541461140875,
+      "learning_rate": 3.5600000000000002e-06,
+      "loss": 1.5883,
+      "step": 90
+    },
+    {
+      "epoch": 0.10917816436712657,
+      "grad_norm": 1.1320104168531329,
+      "learning_rate": 3.6000000000000003e-06,
+      "loss": 1.0229,
+      "step": 91
+    },
+    {
+      "epoch": 0.11037792441511697,
+      "grad_norm": 1.9859979163482226,
+      "learning_rate": 3.6400000000000003e-06,
+      "loss": 1.4403,
+      "step": 92
+    },
+    {
+      "epoch": 0.11157768446310738,
+      "grad_norm": 1.6894108669888084,
+      "learning_rate": 3.6800000000000003e-06,
+      "loss": 1.2201,
+      "step": 93
+    },
+    {
+      "epoch": 0.11277744451109778,
+      "grad_norm": 1.5071412717650037,
+      "learning_rate": 3.7200000000000004e-06,
+      "loss": 1.4269,
+      "step": 94
+    },
+    {
+      "epoch": 0.11397720455908818,
+      "grad_norm": 1.6485135378859557,
+      "learning_rate": 3.7600000000000004e-06,
+      "loss": 1.4154,
+      "step": 95
+    },
+    {
+      "epoch": 0.11517696460707859,
+      "grad_norm": 1.6754056219688747,
+      "learning_rate": 3.8000000000000005e-06,
+      "loss": 1.3267,
+      "step": 96
+    },
+    {
+      "epoch": 0.11637672465506899,
+      "grad_norm": 1.2972761759551974,
+      "learning_rate": 3.8400000000000005e-06,
+      "loss": 1.2842,
+      "step": 97
+    },
+    {
+      "epoch": 0.11757648470305938,
+      "grad_norm": 1.2720323324361733,
+      "learning_rate": 3.88e-06,
+      "loss": 1.131,
+      "step": 98
+    },
+    {
+      "epoch": 0.1187762447510498,
+      "grad_norm": 1.2532439771283779,
+      "learning_rate": 3.920000000000001e-06,
+      "loss": 1.2646,
+      "step": 99
+    },
+    {
+      "epoch": 0.11997600479904019,
+      "grad_norm": 1.7000249287596594,
+      "learning_rate": 3.96e-06,
+      "loss": 1.3158,
+      "step": 100
+    },
+    {
+      "epoch": 0.12117576484703059,
+      "grad_norm": 1.5235594363476401,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.3772,
+      "step": 101
+    },
+    {
+      "epoch": 0.122375524895021,
+      "grad_norm": 1.301839610617614,
+      "learning_rate": 4.04e-06,
+      "loss": 1.2182,
+      "step": 102
+    },
+    {
+      "epoch": 0.1235752849430114,
+      "grad_norm": 1.210313514607747,
+      "learning_rate": 4.08e-06,
+      "loss": 1.1669,
+      "step": 103
+    },
+    {
+      "epoch": 0.1247750449910018,
+      "grad_norm": 1.30142947639613,
+      "learning_rate": 4.12e-06,
+      "loss": 1.0672,
+      "step": 104
+    },
+    {
+      "epoch": 0.1259748050389922,
+      "grad_norm": 3.128370236473383,
+      "learning_rate": 4.16e-06,
+      "loss": 1.2039,
+      "step": 105
+    },
+    {
+      "epoch": 0.1271745650869826,
+      "grad_norm": 1.4045390458763363,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 1.3575,
+      "step": 106
+    },
+    {
+      "epoch": 0.12837432513497302,
+      "grad_norm": 1.3667067390135492,
+      "learning_rate": 4.24e-06,
+      "loss": 1.2138,
+      "step": 107
+    },
+    {
+      "epoch": 0.1295740851829634,
+      "grad_norm": 1.2285297203972647,
+      "learning_rate": 4.2800000000000005e-06,
+      "loss": 1.2163,
+      "step": 108
+    },
+    {
+      "epoch": 0.1307738452309538,
+      "grad_norm": 1.3090900238777463,
+      "learning_rate": 4.32e-06,
+      "loss": 1.2605,
+      "step": 109
+    },
+    {
+      "epoch": 0.13197360527894422,
+      "grad_norm": 2.052730934540717,
+      "learning_rate": 4.360000000000001e-06,
+      "loss": 1.2653,
+      "step": 110
+    },
+    {
+      "epoch": 0.1331733653269346,
+      "grad_norm": 1.144841186760904,
+      "learning_rate": 4.4e-06,
+      "loss": 1.0678,
+      "step": 111
+    },
+    {
+      "epoch": 0.13437312537492502,
+      "grad_norm": 1.246084710168215,
+      "learning_rate": 4.440000000000001e-06,
+      "loss": 1.2,
+      "step": 112
+    },
+    {
+      "epoch": 0.13557288542291543,
+      "grad_norm": 1.6933198526774484,
+      "learning_rate": 4.48e-06,
+      "loss": 1.3187,
+      "step": 113
+    },
+    {
+      "epoch": 0.1367726454709058,
+      "grad_norm": 1.1999792730510719,
+      "learning_rate": 4.520000000000001e-06,
+      "loss": 1.1219,
+      "step": 114
+    },
+    {
+      "epoch": 0.13797240551889622,
+      "grad_norm": 1.0620219067763197,
+      "learning_rate": 4.56e-06,
+      "loss": 1.1237,
+      "step": 115
+    },
+    {
+      "epoch": 0.13917216556688664,
+      "grad_norm": 1.2599012736624202,
+      "learning_rate": 4.600000000000001e-06,
+      "loss": 1.1174,
+      "step": 116
+    },
+    {
+      "epoch": 0.14037192561487702,
+      "grad_norm": 1.0542956424602081,
+      "learning_rate": 4.6400000000000005e-06,
+      "loss": 1.0556,
+      "step": 117
+    },
+    {
+      "epoch": 0.14157168566286743,
+      "grad_norm": 1.4137377397969535,
+      "learning_rate": 4.680000000000001e-06,
+      "loss": 1.2119,
+      "step": 118
+    },
+    {
+      "epoch": 0.14277144571085784,
+      "grad_norm": 1.3907909495429018,
+      "learning_rate": 4.7200000000000005e-06,
+      "loss": 1.1808,
+      "step": 119
+    },
+    {
+      "epoch": 0.14397120575884823,
+      "grad_norm": 1.1932500680807583,
+      "learning_rate": 4.76e-06,
+      "loss": 1.1249,
+      "step": 120
+    },
+    {
+      "epoch": 0.14517096580683864,
+      "grad_norm": 1.0846222377259351,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 1.0507,
+      "step": 121
+    },
+    {
+      "epoch": 0.14637072585482905,
+      "grad_norm": 1.0610851208774656,
+      "learning_rate": 4.84e-06,
+      "loss": 1.1116,
+      "step": 122
+    },
+    {
+      "epoch": 0.14757048590281943,
+      "grad_norm": 0.886202868439996,
+      "learning_rate": 4.880000000000001e-06,
+      "loss": 0.9611,
+      "step": 123
+    },
+    {
+      "epoch": 0.14877024595080984,
+      "grad_norm": 1.6115026528324286,
+      "learning_rate": 4.92e-06,
+      "loss": 1.115,
+      "step": 124
+    },
+    {
+      "epoch": 0.14997000599880023,
+      "grad_norm": 1.0806217358873074,
+      "learning_rate": 4.960000000000001e-06,
+      "loss": 1.0021,
+      "step": 125
+    },
+    {
+      "epoch": 0.15116976604679064,
+      "grad_norm": 0.9920194026980443,
+      "learning_rate": 5e-06,
+      "loss": 1.062,
+      "step": 126
+    },
+    {
+      "epoch": 0.15236952609478105,
+      "grad_norm": 0.9327889709338532,
+      "learning_rate": 5.04e-06,
+      "loss": 1.0048,
+      "step": 127
+    },
+    {
+      "epoch": 0.15356928614277143,
+      "grad_norm": 1.0056043354123612,
+      "learning_rate": 5.0800000000000005e-06,
+      "loss": 0.9148,
+      "step": 128
+    },
+    {
+      "epoch": 0.15476904619076184,
+      "grad_norm": 1.0190425193471366,
+      "learning_rate": 5.12e-06,
+      "loss": 1.0364,
+      "step": 129
+    },
+    {
+      "epoch": 0.15596880623875226,
+      "grad_norm": 0.8777918587504914,
+      "learning_rate": 5.1600000000000006e-06,
+      "loss": 1.014,
+      "step": 130
+    },
+    {
+      "epoch": 0.15716856628674264,
+      "grad_norm": 1.1201006530409465,
+      "learning_rate": 5.2e-06,
+      "loss": 1.0061,
+      "step": 131
+    },
+    {
+      "epoch": 0.15836832633473305,
+      "grad_norm": 1.0665861855030594,
+      "learning_rate": 5.240000000000001e-06,
+      "loss": 1.0281,
+      "step": 132
+    },
+    {
+      "epoch": 0.15956808638272346,
+      "grad_norm": 1.2894515134551434,
+      "learning_rate": 5.28e-06,
+      "loss": 0.9923,
+      "step": 133
+    },
+    {
+      "epoch": 0.16076784643071385,
+      "grad_norm": 1.0389990888777316,
+      "learning_rate": 5.320000000000001e-06,
+      "loss": 1.1106,
+      "step": 134
+    },
+    {
+      "epoch": 0.16196760647870426,
+      "grad_norm": 1.0917817440032695,
+      "learning_rate": 5.36e-06,
+      "loss": 1.0392,
+      "step": 135
+    },
+    {
+      "epoch": 0.16316736652669467,
+      "grad_norm": 1.0587762299260288,
+      "learning_rate": 5.400000000000001e-06,
+      "loss": 1.0769,
+      "step": 136
+    },
+    {
+      "epoch": 0.16436712657468505,
+      "grad_norm": 0.8716176938258128,
+      "learning_rate": 5.4400000000000004e-06,
+      "loss": 1.0114,
+      "step": 137
+    },
+    {
+      "epoch": 0.16556688662267546,
+      "grad_norm": 1.1810297418673554,
+      "learning_rate": 5.480000000000001e-06,
+      "loss": 1.0641,
+      "step": 138
+    },
+    {
+      "epoch": 0.16676664667066587,
+      "grad_norm": 0.97161665734718,
+      "learning_rate": 5.5200000000000005e-06,
+      "loss": 1.0058,
+      "step": 139
+    },
+    {
+      "epoch": 0.16796640671865626,
+      "grad_norm": 1.0199814307298578,
+      "learning_rate": 5.560000000000001e-06,
+      "loss": 0.9774,
+      "step": 140
+    },
+    {
+      "epoch": 0.16916616676664667,
+      "grad_norm": 1.2324988972290545,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 1.0081,
+      "step": 141
+    },
+    {
+      "epoch": 0.17036592681463708,
+      "grad_norm": 0.9165804188035397,
+      "learning_rate": 5.64e-06,
+      "loss": 0.9421,
+      "step": 142
+    },
+    {
+      "epoch": 0.17156568686262746,
+      "grad_norm": 1.0948757075942301,
+      "learning_rate": 5.68e-06,
+      "loss": 0.9672,
+      "step": 143
+    },
+    {
+      "epoch": 0.17276544691061788,
+      "grad_norm": 1.3547639743881061,
+      "learning_rate": 5.72e-06,
+      "loss": 1.0737,
+      "step": 144
+    },
+    {
+      "epoch": 0.1739652069586083,
+      "grad_norm": 0.8432488035480182,
+      "learning_rate": 5.76e-06,
+      "loss": 0.9153,
+      "step": 145
+    },
+    {
+      "epoch": 0.17516496700659867,
+      "grad_norm": 0.9801210666456533,
+      "learning_rate": 5.8e-06,
+      "loss": 0.9635,
+      "step": 146
+    },
+    {
+      "epoch": 0.17636472705458908,
+      "grad_norm": 0.9391319754858495,
+      "learning_rate": 5.84e-06,
+      "loss": 0.9016,
+      "step": 147
+    },
+    {
+      "epoch": 0.1775644871025795,
+      "grad_norm": 1.0482432220624154,
+      "learning_rate": 5.8800000000000005e-06,
+      "loss": 0.9298,
+      "step": 148
+    },
+    {
+      "epoch": 0.17876424715056988,
+      "grad_norm": 0.8909811362011276,
+      "learning_rate": 5.92e-06,
+      "loss": 0.8352,
+      "step": 149
+    },
+    {
+      "epoch": 0.1799640071985603,
+      "grad_norm": 0.9822132770617572,
+      "learning_rate": 5.9600000000000005e-06,
+      "loss": 0.9224,
+      "step": 150
+    },
+    {
+      "epoch": 0.1811637672465507,
+      "grad_norm": 1.01230157575695,
+      "learning_rate": 6e-06,
+      "loss": 1.1265,
+      "step": 151
+    },
+    {
+      "epoch": 0.18236352729454108,
+      "grad_norm": 0.9437580880985108,
+      "learning_rate": 6.040000000000001e-06,
+      "loss": 0.958,
+      "step": 152
+    },
+    {
+      "epoch": 0.1835632873425315,
+      "grad_norm": 0.972791293358992,
+      "learning_rate": 6.08e-06,
+      "loss": 0.9232,
+      "step": 153
+    },
+    {
+      "epoch": 0.1847630473905219,
+      "grad_norm": 0.8858717763933377,
+      "learning_rate": 6.120000000000001e-06,
+      "loss": 0.9367,
+      "step": 154
+    },
+    {
+      "epoch": 0.1859628074385123,
+      "grad_norm": 0.9341273071118985,
+      "learning_rate": 6.16e-06,
+      "loss": 0.9434,
+      "step": 155
+    },
+    {
+      "epoch": 0.1871625674865027,
+      "grad_norm": 0.9816475511556988,
+      "learning_rate": 6.200000000000001e-06,
+      "loss": 1.0585,
+      "step": 156
+    },
+    {
+      "epoch": 0.1883623275344931,
+      "grad_norm": 0.8700040253263074,
+      "learning_rate": 6.24e-06,
+      "loss": 0.8949,
+      "step": 157
+    },
+    {
+      "epoch": 0.1895620875824835,
+      "grad_norm": 0.8317898088035741,
+      "learning_rate": 6.280000000000001e-06,
+      "loss": 0.9559,
+      "step": 158
+    },
+    {
+      "epoch": 0.1907618476304739,
+      "grad_norm": 1.1531747673452089,
+      "learning_rate": 6.3200000000000005e-06,
+      "loss": 1.0648,
+      "step": 159
+    },
+    {
+      "epoch": 0.19196160767846432,
+      "grad_norm": 0.8882450955578516,
+      "learning_rate": 6.360000000000001e-06,
+      "loss": 0.911,
+      "step": 160
+    },
+    {
+      "epoch": 0.1931613677264547,
+      "grad_norm": 0.8890920488919913,
+      "learning_rate": 6.4000000000000006e-06,
+      "loss": 0.8869,
+      "step": 161
+    },
+    {
+      "epoch": 0.1943611277744451,
+      "grad_norm": 0.9109956956678691,
+      "learning_rate": 6.440000000000001e-06,
+      "loss": 0.9941,
+      "step": 162
+    },
+    {
+      "epoch": 0.19556088782243553,
+      "grad_norm": 0.831996363776618,
+      "learning_rate": 6.480000000000001e-06,
+      "loss": 0.8425,
+      "step": 163
+    },
+    {
+      "epoch": 0.1967606478704259,
+      "grad_norm": 0.8344823541062902,
+      "learning_rate": 6.520000000000001e-06,
+      "loss": 0.8864,
+      "step": 164
+    },
+    {
+      "epoch": 0.19796040791841632,
+      "grad_norm": 1.093300270347132,
+      "learning_rate": 6.560000000000001e-06,
+      "loss": 0.9468,
+      "step": 165
+    },
+    {
+      "epoch": 0.19916016796640673,
+      "grad_norm": 0.9247298608370305,
+      "learning_rate": 6.600000000000001e-06,
+      "loss": 0.9053,
+      "step": 166
+    },
+    {
+      "epoch": 0.20035992801439712,
+      "grad_norm": 0.8701588667382361,
+      "learning_rate": 6.640000000000001e-06,
+      "loss": 0.9582,
+      "step": 167
+    },
+    {
+      "epoch": 0.20155968806238753,
+      "grad_norm": 0.9402890562149786,
+      "learning_rate": 6.680000000000001e-06,
+      "loss": 0.9583,
+      "step": 168
+    },
+    {
+      "epoch": 0.20275944811037794,
+      "grad_norm": 0.9482865539009554,
+      "learning_rate": 6.720000000000001e-06,
+      "loss": 0.9701,
+      "step": 169
+    },
+    {
+      "epoch": 0.20395920815836832,
+      "grad_norm": 0.9602660237737364,
+      "learning_rate": 6.760000000000001e-06,
+      "loss": 0.8874,
+      "step": 170
+    },
+    {
+      "epoch": 0.20515896820635873,
+      "grad_norm": 0.9616673957223468,
+      "learning_rate": 6.800000000000001e-06,
+      "loss": 0.9417,
+      "step": 171
+    },
+    {
+      "epoch": 0.20635872825434912,
+      "grad_norm": 0.9294318758101818,
+      "learning_rate": 6.8400000000000014e-06,
+      "loss": 1.0462,
+      "step": 172
+    },
+    {
+      "epoch": 0.20755848830233953,
+      "grad_norm": 1.025443081693529,
+      "learning_rate": 6.88e-06,
+      "loss": 1.0368,
+      "step": 173
+    },
+    {
+      "epoch": 0.20875824835032994,
+      "grad_norm": 0.9243795087667857,
+      "learning_rate": 6.92e-06,
+      "loss": 0.9266,
+      "step": 174
+    },
+    {
+      "epoch": 0.20995800839832032,
+      "grad_norm": 0.8913356558613602,
+      "learning_rate": 6.96e-06,
+      "loss": 0.8184,
+      "step": 175
+    },
+    {
+      "epoch": 0.21115776844631073,
+      "grad_norm": 0.8235026040452708,
+      "learning_rate": 7e-06,
+      "loss": 0.8625,
+      "step": 176
+    },
+    {
+      "epoch": 0.21235752849430115,
+      "grad_norm": 0.8610225245675168,
+      "learning_rate": 7.04e-06,
+      "loss": 0.8752,
+      "step": 177
+    },
+    {
+      "epoch": 0.21355728854229153,
+      "grad_norm": 0.9258435972085604,
+      "learning_rate": 7.08e-06,
+      "loss": 0.8718,
+      "step": 178
+    },
+    {
+      "epoch": 0.21475704859028194,
+      "grad_norm": 0.9663599667481519,
+      "learning_rate": 7.1200000000000004e-06,
+      "loss": 0.9455,
+      "step": 179
+    },
+    {
+      "epoch": 0.21595680863827235,
+      "grad_norm": 0.8891513081143456,
+      "learning_rate": 7.16e-06,
+      "loss": 0.8913,
+      "step": 180
+    },
+    {
+      "epoch": 0.21715656868626274,
+      "grad_norm": 0.8901809205969442,
+      "learning_rate": 7.2000000000000005e-06,
+      "loss": 0.806,
+      "step": 181
+    },
+    {
+      "epoch": 0.21835632873425315,
+      "grad_norm": 0.9125751194829773,
+      "learning_rate": 7.24e-06,
+      "loss": 0.91,
+      "step": 182
+    },
+    {
+      "epoch": 0.21955608878224356,
+      "grad_norm": 0.8419253308107584,
+      "learning_rate": 7.280000000000001e-06,
+      "loss": 0.899,
+      "step": 183
+    },
+    {
+      "epoch": 0.22075584883023394,
+      "grad_norm": 0.9065419690594954,
+      "learning_rate": 7.32e-06,
+      "loss": 0.8456,
+      "step": 184
+    },
+    {
+      "epoch": 0.22195560887822435,
+      "grad_norm": 0.8982973381361369,
+      "learning_rate": 7.360000000000001e-06,
+      "loss": 0.8708,
+      "step": 185
+    },
+    {
+      "epoch": 0.22315536892621476,
+      "grad_norm": 0.8089289556734092,
+      "learning_rate": 7.4e-06,
+      "loss": 0.8851,
+      "step": 186
+    },
+    {
+      "epoch": 0.22435512897420515,
+      "grad_norm": 0.9659392279213572,
+      "learning_rate": 7.440000000000001e-06,
+      "loss": 0.9239,
+      "step": 187
+    },
+    {
+      "epoch": 0.22555488902219556,
+      "grad_norm": 0.9918304286118667,
+      "learning_rate": 7.48e-06,
+      "loss": 0.8922,
+      "step": 188
+    },
+    {
+      "epoch": 0.22675464907018597,
+      "grad_norm": 0.8870035129018767,
+      "learning_rate": 7.520000000000001e-06,
+      "loss": 0.7712,
+      "step": 189
+    },
+    {
+      "epoch": 0.22795440911817635,
+      "grad_norm": 0.8304274635703632,
+      "learning_rate": 7.5600000000000005e-06,
+      "loss": 0.874,
+      "step": 190
+    },
+    {
+      "epoch": 0.22915416916616677,
+      "grad_norm": 0.9364956937637816,
+      "learning_rate": 7.600000000000001e-06,
+      "loss": 0.9543,
+      "step": 191
+    },
+    {
+      "epoch": 0.23035392921415718,
+      "grad_norm": 0.8742969550728668,
+      "learning_rate": 7.640000000000001e-06,
+      "loss": 0.9191,
+      "step": 192
+    },
+    {
+      "epoch": 0.23155368926214756,
+      "grad_norm": 0.9691851065618883,
+      "learning_rate": 7.680000000000001e-06,
+      "loss": 0.8632,
+      "step": 193
+    },
+    {
+      "epoch": 0.23275344931013797,
+      "grad_norm": 0.8438033215721245,
+      "learning_rate": 7.72e-06,
+      "loss": 0.9172,
+      "step": 194
+    },
+    {
+      "epoch": 0.23395320935812838,
+      "grad_norm": 0.9049771481466592,
+      "learning_rate": 7.76e-06,
+      "loss": 0.9471,
+      "step": 195
+    },
+    {
+      "epoch": 0.23515296940611877,
+      "grad_norm": 0.989123646719365,
+      "learning_rate": 7.800000000000002e-06,
+      "loss": 0.8708,
+      "step": 196
+    },
+    {
+      "epoch": 0.23635272945410918,
+      "grad_norm": 0.9706825120362642,
+      "learning_rate": 7.840000000000001e-06,
+      "loss": 0.9137,
+      "step": 197
+    },
+    {
+      "epoch": 0.2375524895020996,
+      "grad_norm": 0.8874052204518389,
+      "learning_rate": 7.88e-06,
+      "loss": 0.9451,
+      "step": 198
+    },
+    {
+      "epoch": 0.23875224955008997,
+      "grad_norm": 0.822402836098356,
+      "learning_rate": 7.92e-06,
+      "loss": 0.8645,
+      "step": 199
+    },
+    {
+      "epoch": 0.23995200959808038,
+      "grad_norm": 0.8547125621693455,
+      "learning_rate": 7.960000000000002e-06,
+      "loss": 0.846,
+      "step": 200
+    },
+    {
+      "epoch": 0.2411517696460708,
+      "grad_norm": 1.1387174253435128,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.952,
+      "step": 201
+    },
+    {
+      "epoch": 0.24235152969406118,
+      "grad_norm": 0.9321636106022483,
+      "learning_rate": 8.040000000000001e-06,
+      "loss": 0.7929,
+      "step": 202
+    },
+    {
+      "epoch": 0.2435512897420516,
+      "grad_norm": 0.8972716580989439,
+      "learning_rate": 8.08e-06,
+      "loss": 0.9002,
+      "step": 203
+    },
+    {
+      "epoch": 0.244751049790042,
+      "grad_norm": 0.8627127688612859,
+      "learning_rate": 8.120000000000002e-06,
+      "loss": 0.8346,
+      "step": 204
+    },
+    {
+      "epoch": 0.24595080983803239,
+      "grad_norm": 0.9804340797810391,
+      "learning_rate": 8.16e-06,
+      "loss": 0.8275,
+      "step": 205
+    },
+    {
+      "epoch": 0.2471505698860228,
+      "grad_norm": 0.9665883639615581,
+      "learning_rate": 8.2e-06,
+      "loss": 0.8732,
+      "step": 206
+    },
+    {
+      "epoch": 0.2483503299340132,
+      "grad_norm": 0.9978598494460422,
+      "learning_rate": 8.24e-06,
+      "loss": 0.8873,
+      "step": 207
+    },
+    {
+      "epoch": 0.2495500899820036,
+      "grad_norm": 0.835752279975912,
+      "learning_rate": 8.28e-06,
+      "loss": 0.7762,
+      "step": 208
+    },
+    {
+      "epoch": 0.250749850029994,
+      "grad_norm": 0.8056768392805612,
+      "learning_rate": 8.32e-06,
+      "loss": 0.8065,
+      "step": 209
+    },
+    {
+      "epoch": 0.2519496100779844,
+      "grad_norm": 1.0454200086020167,
+      "learning_rate": 8.36e-06,
+      "loss": 0.9384,
+      "step": 210
+    },
+    {
+      "epoch": 0.2531493701259748,
+      "grad_norm": 0.8460330286684768,
+      "learning_rate": 8.400000000000001e-06,
+      "loss": 0.8941,
+      "step": 211
+    },
+    {
+      "epoch": 0.2543491301739652,
+      "grad_norm": 0.8447583533784957,
+      "learning_rate": 8.44e-06,
+      "loss": 0.8714,
+      "step": 212
+    },
+    {
+      "epoch": 0.2555488902219556,
+      "grad_norm": 0.8376650804593687,
+      "learning_rate": 8.48e-06,
+      "loss": 0.8926,
+      "step": 213
+    },
+    {
+      "epoch": 0.25674865026994603,
+      "grad_norm": 0.8489358021207093,
+      "learning_rate": 8.52e-06,
+      "loss": 0.8818,
+      "step": 214
+    },
+    {
+      "epoch": 0.2579484103179364,
+      "grad_norm": 0.9236326997232677,
+      "learning_rate": 8.560000000000001e-06,
+      "loss": 0.9013,
+      "step": 215
+    },
+    {
+      "epoch": 0.2591481703659268,
+      "grad_norm": 0.8525808314403832,
+      "learning_rate": 8.6e-06,
+      "loss": 0.8728,
+      "step": 216
+    },
+    {
+      "epoch": 0.26034793041391724,
+      "grad_norm": 0.79761436147496,
+      "learning_rate": 8.64e-06,
+      "loss": 0.8644,
+      "step": 217
+    },
+    {
+      "epoch": 0.2615476904619076,
+      "grad_norm": 0.8960550194483666,
+      "learning_rate": 8.68e-06,
+      "loss": 0.8378,
+      "step": 218
+    },
+    {
+      "epoch": 0.262747450509898,
+      "grad_norm": 0.8592564745427348,
+      "learning_rate": 8.720000000000001e-06,
+      "loss": 0.7725,
+      "step": 219
+    },
+    {
+      "epoch": 0.26394721055788845,
+      "grad_norm": 0.9498899750514396,
+      "learning_rate": 8.76e-06,
+      "loss": 0.9275,
+      "step": 220
+    },
+    {
+      "epoch": 0.26514697060587883,
+      "grad_norm": 0.9793881370045231,
+      "learning_rate": 8.8e-06,
+      "loss": 0.9023,
+      "step": 221
+    },
+    {
+      "epoch": 0.2663467306538692,
+      "grad_norm": 0.7546525454223989,
+      "learning_rate": 8.84e-06,
+      "loss": 0.8087,
+      "step": 222
+    },
+    {
+      "epoch": 0.26754649070185965,
+      "grad_norm": 0.895743262472989,
+      "learning_rate": 8.880000000000001e-06,
+      "loss": 0.8462,
+      "step": 223
+    },
+    {
+      "epoch": 0.26874625074985004,
+      "grad_norm": 0.8508391890954724,
+      "learning_rate": 8.920000000000001e-06,
+      "loss": 0.9041,
+      "step": 224
+    },
+    {
+      "epoch": 0.2699460107978404,
+      "grad_norm": 0.8441234806191734,
+      "learning_rate": 8.96e-06,
+      "loss": 0.8307,
+      "step": 225
+    },
+    {
+      "epoch": 0.27114577084583086,
+      "grad_norm": 0.8792012530549744,
+      "learning_rate": 9e-06,
+      "loss": 0.8054,
+      "step": 226
+    },
+    {
+      "epoch": 0.27234553089382124,
+      "grad_norm": 0.8862693866179844,
+      "learning_rate": 9.040000000000002e-06,
+      "loss": 0.8082,
+      "step": 227
+    },
+    {
+      "epoch": 0.2735452909418116,
+      "grad_norm": 0.936765475346329,
+      "learning_rate": 9.080000000000001e-06,
+      "loss": 0.8213,
+      "step": 228
+    },
+    {
+      "epoch": 0.27474505098980206,
+      "grad_norm": 0.8845364940014634,
+      "learning_rate": 9.12e-06,
+      "loss": 0.8575,
+      "step": 229
+    },
+    {
+      "epoch": 0.27594481103779245,
+      "grad_norm": 0.8943979321402359,
+      "learning_rate": 9.16e-06,
+      "loss": 0.8593,
+      "step": 230
+    },
+    {
+      "epoch": 0.27714457108578283,
+      "grad_norm": 0.8746312602873112,
+      "learning_rate": 9.200000000000002e-06,
+      "loss": 0.8003,
+      "step": 231
+    },
+    {
+      "epoch": 0.27834433113377327,
+      "grad_norm": 0.8863435690762195,
+      "learning_rate": 9.240000000000001e-06,
+      "loss": 0.8141,
+      "step": 232
+    },
+    {
+      "epoch": 0.27954409118176365,
+      "grad_norm": 0.8886892400062681,
+      "learning_rate": 9.280000000000001e-06,
+      "loss": 0.8593,
+      "step": 233
+    },
+    {
+      "epoch": 0.28074385122975404,
+      "grad_norm": 0.8938765308959884,
+      "learning_rate": 9.32e-06,
+      "loss": 0.7926,
+      "step": 234
+    },
+    {
+      "epoch": 0.2819436112777445,
+      "grad_norm": 0.8905592256018503,
+      "learning_rate": 9.360000000000002e-06,
+      "loss": 0.7802,
+      "step": 235
+    },
+    {
+      "epoch": 0.28314337132573486,
+      "grad_norm": 0.944437847777848,
+      "learning_rate": 9.4e-06,
+      "loss": 0.8699,
+      "step": 236
+    },
+    {
+      "epoch": 0.28434313137372524,
+      "grad_norm": 0.9264151484947536,
+      "learning_rate": 9.440000000000001e-06,
+      "loss": 0.767,
+      "step": 237
+    },
+    {
+      "epoch": 0.2855428914217157,
+      "grad_norm": 0.7480605034270361,
+      "learning_rate": 9.48e-06,
+      "loss": 0.7868,
+      "step": 238
+    },
+    {
+      "epoch": 0.28674265146970607,
+      "grad_norm": 1.1013740308750852,
+      "learning_rate": 9.52e-06,
+      "loss": 0.8251,
+      "step": 239
+    },
+    {
+      "epoch": 0.28794241151769645,
+      "grad_norm": 0.817306124632072,
+      "learning_rate": 9.56e-06,
+      "loss": 0.7849,
+      "step": 240
+    },
+    {
+      "epoch": 0.2891421715656869,
+      "grad_norm": 0.8677167723424389,
+      "learning_rate": 9.600000000000001e-06,
+      "loss": 0.8002,
+      "step": 241
+    },
+    {
+      "epoch": 0.2903419316136773,
+      "grad_norm": 0.9154139015776457,
+      "learning_rate": 9.640000000000001e-06,
+      "loss": 0.8361,
+      "step": 242
+    },
+    {
+      "epoch": 0.29154169166166766,
+      "grad_norm": 0.9346294929702866,
+      "learning_rate": 9.68e-06,
+      "loss": 0.8009,
+      "step": 243
+    },
+    {
+      "epoch": 0.2927414517096581,
+      "grad_norm": 0.7915758604161942,
+      "learning_rate": 9.72e-06,
+      "loss": 0.828,
+      "step": 244
+    },
+    {
+      "epoch": 0.2939412117576485,
+      "grad_norm": 0.9537088344426334,
+      "learning_rate": 9.760000000000001e-06,
+      "loss": 0.8671,
+      "step": 245
+    },
+    {
+      "epoch": 0.29514097180563886,
+      "grad_norm": 1.0483821102743403,
+      "learning_rate": 9.800000000000001e-06,
+      "loss": 0.885,
+      "step": 246
+    },
+    {
+      "epoch": 0.29634073185362925,
+      "grad_norm": 0.9102427378891443,
+      "learning_rate": 9.84e-06,
+      "loss": 0.7805,
+      "step": 247
+    },
+    {
+      "epoch": 0.2975404919016197,
+      "grad_norm": 0.9252259031766393,
+      "learning_rate": 9.88e-06,
+      "loss": 0.8074,
+      "step": 248
+    },
+    {
+      "epoch": 0.29874025194961007,
+      "grad_norm": 0.8499107060128288,
+      "learning_rate": 9.920000000000002e-06,
+      "loss": 0.776,
+      "step": 249
+    },
+    {
+      "epoch": 0.29994001199760045,
+      "grad_norm": 1.281569882769472,
+      "learning_rate": 9.960000000000001e-06,
+      "loss": 0.8668,
+      "step": 250
+    },
+    {
+      "epoch": 0.3011397720455909,
+      "grad_norm": 0.8619792688405079,
+      "learning_rate": 1e-05,
+      "loss": 0.8864,
+      "step": 251
+    },
+    {
+      "epoch": 0.3023395320935813,
+      "grad_norm": 0.8951400482973322,
+      "learning_rate": 9.99999512178685e-06,
+      "loss": 0.8235,
+      "step": 252
+    },
+    {
+      "epoch": 0.30353929214157166,
+      "grad_norm": 0.967921131549879,
+      "learning_rate": 9.99998048715692e-06,
+      "loss": 0.7992,
+      "step": 253
+    },
+    {
+      "epoch": 0.3047390521895621,
+      "grad_norm": 0.840452525603444,
+      "learning_rate": 9.999956096138765e-06,
+      "loss": 0.7427,
+      "step": 254
+    },
+    {
+      "epoch": 0.3059388122375525,
+      "grad_norm": 1.0000789029765795,
+      "learning_rate": 9.999921948779978e-06,
+      "loss": 0.873,
+      "step": 255
+    },
+    {
+      "epoch": 0.30713857228554287,
+      "grad_norm": 0.8363738524608543,
+      "learning_rate": 9.99987804514719e-06,
+      "loss": 0.8984,
+      "step": 256
+    },
+    {
+      "epoch": 0.3083383323335333,
+      "grad_norm": 0.903315872911508,
+      "learning_rate": 9.999824385326073e-06,
+      "loss": 0.8936,
+      "step": 257
+    },
+    {
+      "epoch": 0.3095380923815237,
+      "grad_norm": 0.8112526213382579,
+      "learning_rate": 9.99976096942133e-06,
+      "loss": 0.8415,
+      "step": 258
+    },
+    {
+      "epoch": 0.31073785242951407,
+      "grad_norm": 0.8540518199641305,
+      "learning_rate": 9.999687797556705e-06,
+      "loss": 0.8041,
+      "step": 259
+    },
+    {
+      "epoch": 0.3119376124775045,
+      "grad_norm": 0.7614123112280738,
+      "learning_rate": 9.999604869874975e-06,
+      "loss": 0.9015,
+      "step": 260
+    },
+    {
+      "epoch": 0.3131373725254949,
+      "grad_norm": 0.8245082227173874,
+      "learning_rate": 9.999512186537957e-06,
+      "loss": 0.8514,
+      "step": 261
+    },
+    {
+      "epoch": 0.3143371325734853,
+      "grad_norm": 0.8054911223663823,
+      "learning_rate": 9.999409747726502e-06,
+      "loss": 0.8167,
+      "step": 262
+    },
+    {
+      "epoch": 0.3155368926214757,
+      "grad_norm": 0.8903884772939344,
+      "learning_rate": 9.999297553640498e-06,
+      "loss": 0.8892,
+      "step": 263
+    },
+    {
+      "epoch": 0.3167366526694661,
+      "grad_norm": 0.8458806392274165,
+      "learning_rate": 9.999175604498867e-06,
+      "loss": 0.8686,
+      "step": 264
+    },
+    {
+      "epoch": 0.3179364127174565,
+      "grad_norm": 0.8306255702145542,
+      "learning_rate": 9.999043900539568e-06,
+      "loss": 0.8528,
+      "step": 265
+    },
+    {
+      "epoch": 0.3191361727654469,
+      "grad_norm": 0.942771419645441,
+      "learning_rate": 9.99890244201959e-06,
+      "loss": 0.7698,
+      "step": 266
+    },
+    {
+      "epoch": 0.3203359328134373,
+      "grad_norm": 0.9116916527123582,
+      "learning_rate": 9.998751229214962e-06,
+      "loss": 0.7459,
+      "step": 267
+    },
+    {
+      "epoch": 0.3215356928614277,
+      "grad_norm": 0.8315868072518188,
+      "learning_rate": 9.998590262420743e-06,
+      "loss": 0.7962,
+      "step": 268
+    },
+    {
+      "epoch": 0.32273545290941813,
+      "grad_norm": 0.8560623972580833,
+      "learning_rate": 9.998419541951024e-06,
+      "loss": 0.8476,
+      "step": 269
+    },
+    {
+      "epoch": 0.3239352129574085,
+      "grad_norm": 0.9529061379410192,
+      "learning_rate": 9.998239068138929e-06,
+      "loss": 0.8171,
+      "step": 270
+    },
+    {
+      "epoch": 0.3251349730053989,
+      "grad_norm": 0.8807774080549793,
+      "learning_rate": 9.998048841336615e-06,
+      "loss": 0.8538,
+      "step": 271
+    },
+    {
+      "epoch": 0.32633473305338934,
+      "grad_norm": 0.85447879331899,
+      "learning_rate": 9.997848861915268e-06,
+      "loss": 0.8696,
+      "step": 272
+    },
+    {
+      "epoch": 0.3275344931013797,
+      "grad_norm": 1.018623091493027,
+      "learning_rate": 9.997639130265104e-06,
+      "loss": 0.7725,
+      "step": 273
+    },
+    {
+      "epoch": 0.3287342531493701,
+      "grad_norm": 0.8746987739592668,
+      "learning_rate": 9.997419646795372e-06,
+      "loss": 0.86,
+      "step": 274
+    },
+    {
+      "epoch": 0.32993401319736054,
+      "grad_norm": 0.921512734396811,
+      "learning_rate": 9.997190411934346e-06,
+      "loss": 0.7502,
+      "step": 275
+    },
+    {
+      "epoch": 0.3311337732453509,
+      "grad_norm": 0.9161536633828824,
+      "learning_rate": 9.996951426129327e-06,
+      "loss": 0.8332,
+      "step": 276
+    },
+    {
+      "epoch": 0.3323335332933413,
+      "grad_norm": 0.8829663327810717,
+      "learning_rate": 9.996702689846645e-06,
+      "loss": 0.7447,
+      "step": 277
+    },
+    {
+      "epoch": 0.33353329334133175,
+      "grad_norm": 0.8870347194650321,
+      "learning_rate": 9.996444203571655e-06,
+      "loss": 0.7957,
+      "step": 278
+    },
+    {
+      "epoch": 0.33473305338932213,
+      "grad_norm": 1.0771228842939597,
+      "learning_rate": 9.996175967808738e-06,
+      "loss": 0.8305,
+      "step": 279
+    },
+    {
+      "epoch": 0.3359328134373125,
+      "grad_norm": 1.0172416792396413,
+      "learning_rate": 9.995897983081301e-06,
+      "loss": 0.8696,
+      "step": 280
+    },
+    {
+      "epoch": 0.33713257348530296,
+      "grad_norm": 0.9606670435787978,
+      "learning_rate": 9.995610249931769e-06,
+      "loss": 0.8481,
+      "step": 281
+    },
+    {
+      "epoch": 0.33833233353329334,
+      "grad_norm": 0.9026045069176893,
+      "learning_rate": 9.995312768921592e-06,
+      "loss": 0.8514,
+      "step": 282
+    },
+    {
+      "epoch": 0.3395320935812837,
+      "grad_norm": 0.9391209813217448,
+      "learning_rate": 9.995005540631239e-06,
+      "loss": 0.7811,
+      "step": 283
+    },
+    {
+      "epoch": 0.34073185362927416,
+      "grad_norm": 0.8262010905204108,
+      "learning_rate": 9.994688565660201e-06,
+      "loss": 0.7676,
+      "step": 284
+    },
+    {
+      "epoch": 0.34193161367726455,
+      "grad_norm": 0.8447905564304076,
+      "learning_rate": 9.994361844626988e-06,
+      "loss": 0.7969,
+      "step": 285
+    },
+    {
+      "epoch": 0.34313137372525493,
+      "grad_norm": 0.8682002293146721,
+      "learning_rate": 9.994025378169123e-06,
+      "loss": 0.7306,
+      "step": 286
+    },
+    {
+      "epoch": 0.34433113377324537,
+      "grad_norm": 1.0013520695774265,
+      "learning_rate": 9.99367916694315e-06,
+      "loss": 0.8426,
+      "step": 287
+    },
+    {
+      "epoch": 0.34553089382123575,
+      "grad_norm": 0.8399202189103822,
+      "learning_rate": 9.993323211624625e-06,
+      "loss": 0.7884,
+      "step": 288
+    },
+    {
+      "epoch": 0.34673065386922614,
+      "grad_norm": 0.7564780955692382,
+      "learning_rate": 9.992957512908121e-06,
+      "loss": 0.7966,
+      "step": 289
+    },
+    {
+      "epoch": 0.3479304139172166,
+      "grad_norm": 0.8786188259825324,
+      "learning_rate": 9.992582071507217e-06,
+      "loss": 0.83,
+      "step": 290
+    },
+    {
+      "epoch": 0.34913017396520696,
+      "grad_norm": 0.8263855819597197,
+      "learning_rate": 9.992196888154508e-06,
+      "loss": 0.7298,
+      "step": 291
+    },
+    {
+      "epoch": 0.35032993401319734,
+      "grad_norm": 0.8335583134477816,
+      "learning_rate": 9.991801963601595e-06,
+      "loss": 0.7603,
+      "step": 292
+    },
+    {
+      "epoch": 0.3515296940611878,
+      "grad_norm": 0.8523253593159443,
+      "learning_rate": 9.991397298619091e-06,
+      "loss": 0.7732,
+      "step": 293
+    },
+    {
+      "epoch": 0.35272945410917816,
+      "grad_norm": 0.9147643300032103,
+      "learning_rate": 9.990982893996612e-06,
+      "loss": 0.8953,
+      "step": 294
+    },
+    {
+      "epoch": 0.35392921415716855,
+      "grad_norm": 0.8053352375638536,
+      "learning_rate": 9.990558750542778e-06,
+      "loss": 0.8504,
+      "step": 295
+    },
+    {
+      "epoch": 0.355128974205159,
+      "grad_norm": 0.8642158057007054,
+      "learning_rate": 9.990124869085216e-06,
+      "loss": 0.8018,
+      "step": 296
+    },
+    {
+      "epoch": 0.35632873425314937,
+      "grad_norm": 0.8944881258422162,
+      "learning_rate": 9.989681250470551e-06,
+      "loss": 0.7362,
+      "step": 297
+    },
+    {
+      "epoch": 0.35752849430113975,
+      "grad_norm": 0.8445661968651367,
+      "learning_rate": 9.98922789556441e-06,
+      "loss": 0.7657,
+      "step": 298
+    },
+    {
+      "epoch": 0.3587282543491302,
+      "grad_norm": 0.7985049700643867,
+      "learning_rate": 9.988764805251417e-06,
+      "loss": 0.7836,
+      "step": 299
+    },
+    {
+      "epoch": 0.3599280143971206,
+      "grad_norm": 0.814789600523998,
+      "learning_rate": 9.988291980435194e-06,
+      "loss": 0.8507,
+      "step": 300
+    },
+    {
+      "epoch": 0.36112777444511096,
+      "grad_norm": 0.877151130445411,
+      "learning_rate": 9.98780942203836e-06,
+      "loss": 0.8123,
+      "step": 301
+    },
+    {
+      "epoch": 0.3623275344931014,
+      "grad_norm": 0.8341864995834861,
+      "learning_rate": 9.987317131002519e-06,
+      "loss": 0.795,
+      "step": 302
+    },
+    {
+      "epoch": 0.3635272945410918,
+      "grad_norm": 0.9050260956788025,
+      "learning_rate": 9.986815108288273e-06,
+      "loss": 0.7917,
+      "step": 303
+    },
+    {
+      "epoch": 0.36472705458908217,
+      "grad_norm": 0.8567073056297471,
+      "learning_rate": 9.986303354875213e-06,
+      "loss": 0.8435,
+      "step": 304
+    },
+    {
+      "epoch": 0.3659268146370726,
+      "grad_norm": 0.8858524400221668,
+      "learning_rate": 9.985781871761915e-06,
+      "loss": 0.7367,
+      "step": 305
+    },
+    {
+      "epoch": 0.367126574685063,
+      "grad_norm": 0.8350475034028912,
+      "learning_rate": 9.98525065996594e-06,
+      "loss": 0.7849,
+      "step": 306
+    },
+    {
+      "epoch": 0.3683263347330534,
+      "grad_norm": 1.0440799632179578,
+      "learning_rate": 9.984709720523836e-06,
+      "loss": 0.8932,
+      "step": 307
+    },
+    {
+      "epoch": 0.3695260947810438,
+      "grad_norm": 0.724952917346773,
+      "learning_rate": 9.98415905449113e-06,
+      "loss": 0.7721,
+      "step": 308
+    },
+    {
+      "epoch": 0.3707258548290342,
+      "grad_norm": 0.8473537207653666,
+      "learning_rate": 9.983598662942325e-06,
+      "loss": 0.8477,
+      "step": 309
+    },
+    {
+      "epoch": 0.3719256148770246,
+      "grad_norm": 1.1641508886486074,
+      "learning_rate": 9.983028546970908e-06,
+      "loss": 0.817,
+      "step": 310
+    },
+    {
+      "epoch": 0.373125374925015,
+      "grad_norm": 0.8475121849998504,
+      "learning_rate": 9.982448707689338e-06,
+      "loss": 0.7466,
+      "step": 311
+    },
+    {
+      "epoch": 0.3743251349730054,
+      "grad_norm": 0.9417474319435428,
+      "learning_rate": 9.981859146229047e-06,
+      "loss": 0.7882,
+      "step": 312
+    },
+    {
+      "epoch": 0.3755248950209958,
+      "grad_norm": 0.8832965885957832,
+      "learning_rate": 9.981259863740435e-06,
+      "loss": 0.8116,
+      "step": 313
+    },
+    {
+      "epoch": 0.3767246550689862,
+      "grad_norm": 0.81201642834222,
+      "learning_rate": 9.980650861392877e-06,
+      "loss": 0.8432,
+      "step": 314
+    },
+    {
+      "epoch": 0.3779244151169766,
+      "grad_norm": 0.84708845432052,
+      "learning_rate": 9.980032140374706e-06,
+      "loss": 0.8342,
+      "step": 315
+    },
+    {
+      "epoch": 0.379124175164967,
+      "grad_norm": 0.9425876218381913,
+      "learning_rate": 9.979403701893226e-06,
+      "loss": 0.8462,
+      "step": 316
+    },
+    {
+      "epoch": 0.38032393521295743,
+      "grad_norm": 0.8137886430648797,
+      "learning_rate": 9.9787655471747e-06,
+      "loss": 0.7919,
+      "step": 317
+    },
+    {
+      "epoch": 0.3815236952609478,
+      "grad_norm": 0.7908051128093959,
+      "learning_rate": 9.97811767746435e-06,
+      "loss": 0.8084,
+      "step": 318
+    },
+    {
+      "epoch": 0.3827234553089382,
+      "grad_norm": 0.7982496312542597,
+      "learning_rate": 9.97746009402635e-06,
+      "loss": 0.8085,
+      "step": 319
+    },
+    {
+      "epoch": 0.38392321535692864,
+      "grad_norm": 0.776071189899833,
+      "learning_rate": 9.976792798143837e-06,
+      "loss": 0.7764,
+      "step": 320
+    },
+    {
+      "epoch": 0.385122975404919,
+      "grad_norm": 1.4618534145096531,
+      "learning_rate": 9.976115791118897e-06,
+      "loss": 0.8238,
+      "step": 321
+    },
+    {
+      "epoch": 0.3863227354529094,
+      "grad_norm": 0.7150889271178531,
+      "learning_rate": 9.97542907427256e-06,
+      "loss": 0.7196,
+      "step": 322
+    },
+    {
+      "epoch": 0.38752249550089984,
+      "grad_norm": 0.8050788339997501,
+      "learning_rate": 9.97473264894481e-06,
+      "loss": 0.7565,
+      "step": 323
+    },
+    {
+      "epoch": 0.3887222555488902,
+      "grad_norm": 0.9983272035179562,
+      "learning_rate": 9.97402651649457e-06,
+      "loss": 0.8429,
+      "step": 324
+    },
+    {
+      "epoch": 0.3899220155968806,
+      "grad_norm": 0.8623842545763719,
+      "learning_rate": 9.973310678299704e-06,
+      "loss": 0.7245,
+      "step": 325
+    },
+    {
+      "epoch": 0.39112177564487105,
+      "grad_norm": 0.8591131060288989,
+      "learning_rate": 9.97258513575702e-06,
+      "loss": 0.7125,
+      "step": 326
+    },
+    {
+      "epoch": 0.39232153569286143,
+      "grad_norm": 0.6711400146323111,
+      "learning_rate": 9.971849890282255e-06,
+      "loss": 0.7545,
+      "step": 327
+    },
+    {
+      "epoch": 0.3935212957408518,
+      "grad_norm": 0.768957585508672,
+      "learning_rate": 9.971104943310086e-06,
+      "loss": 0.7789,
+      "step": 328
+    },
+    {
+      "epoch": 0.39472105578884226,
+      "grad_norm": 1.1384395543413788,
+      "learning_rate": 9.970350296294114e-06,
+      "loss": 0.8106,
+      "step": 329
+    },
+    {
+      "epoch": 0.39592081583683264,
+      "grad_norm": 0.877220895635206,
+      "learning_rate": 9.969585950706872e-06,
+      "loss": 0.7892,
+      "step": 330
+    },
+    {
+      "epoch": 0.397120575884823,
+      "grad_norm": 0.8104918613582576,
+      "learning_rate": 9.968811908039818e-06,
+      "loss": 0.7972,
+      "step": 331
+    },
+    {
+      "epoch": 0.39832033593281346,
+      "grad_norm": 0.8234689194129275,
+      "learning_rate": 9.968028169803326e-06,
+      "loss": 0.809,
+      "step": 332
+    },
+    {
+      "epoch": 0.39952009598080385,
+      "grad_norm": 0.8441849338357255,
+      "learning_rate": 9.967234737526695e-06,
+      "loss": 0.775,
+      "step": 333
+    },
+    {
+      "epoch": 0.40071985602879423,
+      "grad_norm": 0.842204245498853,
+      "learning_rate": 9.966431612758138e-06,
+      "loss": 0.6613,
+      "step": 334
+    },
+    {
+      "epoch": 0.40191961607678467,
+      "grad_norm": 0.8800128959101974,
+      "learning_rate": 9.965618797064782e-06,
+      "loss": 0.747,
+      "step": 335
+    },
+    {
+      "epoch": 0.40311937612477505,
+      "grad_norm": 0.9511458805031532,
+      "learning_rate": 9.96479629203266e-06,
+      "loss": 0.7791,
+      "step": 336
+    },
+    {
+      "epoch": 0.40431913617276544,
+      "grad_norm": 0.9726969708248384,
+      "learning_rate": 9.963964099266715e-06,
+      "loss": 0.7684,
+      "step": 337
+    },
+    {
+      "epoch": 0.4055188962207559,
+      "grad_norm": 1.0222745919628025,
+      "learning_rate": 9.963122220390791e-06,
+      "loss": 0.7814,
+      "step": 338
+    },
+    {
+      "epoch": 0.40671865626874626,
+      "grad_norm": 1.0179699232834631,
+      "learning_rate": 9.962270657047634e-06,
+      "loss": 0.7701,
+      "step": 339
+    },
+    {
+      "epoch": 0.40791841631673664,
+      "grad_norm": 0.8559764361792539,
+      "learning_rate": 9.96140941089889e-06,
+      "loss": 0.8081,
+      "step": 340
+    },
+    {
+      "epoch": 0.4091181763647271,
+      "grad_norm": 0.8008244277348966,
+      "learning_rate": 9.960538483625092e-06,
+      "loss": 0.7081,
+      "step": 341
+    },
+    {
+      "epoch": 0.41031793641271747,
+      "grad_norm": 0.802937705951165,
+      "learning_rate": 9.959657876925671e-06,
+      "loss": 0.8017,
+      "step": 342
+    },
+    {
+      "epoch": 0.41151769646070785,
+      "grad_norm": 0.8658546979578301,
+      "learning_rate": 9.95876759251894e-06,
+      "loss": 0.7642,
+      "step": 343
+    },
+    {
+      "epoch": 0.41271745650869823,
+      "grad_norm": 0.8358056079222644,
+      "learning_rate": 9.957867632142098e-06,
+      "loss": 0.7632,
+      "step": 344
+    },
+    {
+      "epoch": 0.41391721655668867,
+      "grad_norm": 3.1390710745913935,
+      "learning_rate": 9.956957997551224e-06,
+      "loss": 0.7971,
+      "step": 345
+    },
+    {
+      "epoch": 0.41511697660467906,
+      "grad_norm": 1.0243036157293175,
+      "learning_rate": 9.956038690521276e-06,
+      "loss": 0.8726,
+      "step": 346
+    },
+    {
+      "epoch": 0.41631673665266944,
+      "grad_norm": 0.8289269012612019,
+      "learning_rate": 9.955109712846083e-06,
+      "loss": 0.7728,
+      "step": 347
+    },
+    {
+      "epoch": 0.4175164967006599,
+      "grad_norm": 0.8245470625798709,
+      "learning_rate": 9.954171066338346e-06,
+      "loss": 0.7399,
+      "step": 348
+    },
+    {
+      "epoch": 0.41871625674865026,
+      "grad_norm": 0.7171137486202372,
+      "learning_rate": 9.95322275282963e-06,
+      "loss": 0.7375,
+      "step": 349
+    },
+    {
+      "epoch": 0.41991601679664065,
+      "grad_norm": 0.8337532347907729,
+      "learning_rate": 9.952264774170368e-06,
+      "loss": 0.8555,
+      "step": 350
+    },
+    {
+      "epoch": 0.4211157768446311,
+      "grad_norm": 0.8802297139352729,
+      "learning_rate": 9.95129713222985e-06,
+      "loss": 0.7439,
+      "step": 351
+    },
+    {
+      "epoch": 0.42231553689262147,
+      "grad_norm": 0.871172679451122,
+      "learning_rate": 9.95031982889622e-06,
+      "loss": 0.7549,
+      "step": 352
+    },
+    {
+      "epoch": 0.42351529694061185,
+      "grad_norm": 0.8808416904757895,
+      "learning_rate": 9.949332866076475e-06,
+      "loss": 0.7092,
+      "step": 353
+    },
+    {
+      "epoch": 0.4247150569886023,
+      "grad_norm": 0.8948904416493283,
+      "learning_rate": 9.948336245696461e-06,
+      "loss": 0.7929,
+      "step": 354
+    },
+    {
+      "epoch": 0.4259148170365927,
+      "grad_norm": 0.7020740256232711,
+      "learning_rate": 9.94732996970087e-06,
+      "loss": 0.7058,
+      "step": 355
+    },
+    {
+      "epoch": 0.42711457708458306,
+      "grad_norm": 0.7923237942501681,
+      "learning_rate": 9.946314040053234e-06,
+      "loss": 0.7346,
+      "step": 356
+    },
+    {
+      "epoch": 0.4283143371325735,
+      "grad_norm": 0.7968620685418168,
+      "learning_rate": 9.945288458735919e-06,
+      "loss": 0.781,
+      "step": 357
+    },
+    {
+      "epoch": 0.4295140971805639,
+      "grad_norm": 0.8543323867178862,
+      "learning_rate": 9.94425322775013e-06,
+      "loss": 0.816,
+      "step": 358
+    },
+    {
+      "epoch": 0.43071385722855426,
+      "grad_norm": 0.9090120863126556,
+      "learning_rate": 9.943208349115893e-06,
+      "loss": 0.7682,
+      "step": 359
+    },
+    {
+      "epoch": 0.4319136172765447,
+      "grad_norm": 0.816617430755013,
+      "learning_rate": 9.94215382487207e-06,
+      "loss": 0.7139,
+      "step": 360
+    },
+    {
+      "epoch": 0.4331133773245351,
+      "grad_norm": 0.7208369485157738,
+      "learning_rate": 9.941089657076334e-06,
+      "loss": 0.7975,
+      "step": 361
+    },
+    {
+      "epoch": 0.43431313737252547,
+      "grad_norm": 0.7129151616725876,
+      "learning_rate": 9.940015847805183e-06,
+      "loss": 0.761,
+      "step": 362
+    },
+    {
+      "epoch": 0.4355128974205159,
+      "grad_norm": 0.8814622211192246,
+      "learning_rate": 9.938932399153924e-06,
+      "loss": 0.8132,
+      "step": 363
+    },
+    {
+      "epoch": 0.4367126574685063,
+      "grad_norm": 0.7244971715773292,
+      "learning_rate": 9.937839313236675e-06,
+      "loss": 0.7698,
+      "step": 364
+    },
+    {
+      "epoch": 0.4379124175164967,
+      "grad_norm": 0.8575978639150249,
+      "learning_rate": 9.936736592186357e-06,
+      "loss": 0.8339,
+      "step": 365
+    },
+    {
+      "epoch": 0.4391121775644871,
+      "grad_norm": 0.8461733576826898,
+      "learning_rate": 9.935624238154696e-06,
+      "loss": 0.7724,
+      "step": 366
+    },
+    {
+      "epoch": 0.4403119376124775,
+      "grad_norm": 0.7924350476401562,
+      "learning_rate": 9.934502253312211e-06,
+      "loss": 0.6884,
+      "step": 367
+    },
+    {
+      "epoch": 0.4415116976604679,
+      "grad_norm": 0.7096586005883623,
+      "learning_rate": 9.93337063984821e-06,
+      "loss": 0.7898,
+      "step": 368
+    },
+    {
+      "epoch": 0.4427114577084583,
+      "grad_norm": 0.8066448342385316,
+      "learning_rate": 9.932229399970801e-06,
+      "loss": 0.7898,
+      "step": 369
+    },
+    {
+      "epoch": 0.4439112177564487,
+      "grad_norm": 0.8256485114482758,
+      "learning_rate": 9.931078535906863e-06,
+      "loss": 0.839,
+      "step": 370
+    },
+    {
+      "epoch": 0.4451109778044391,
+      "grad_norm": 0.7650729169319722,
+      "learning_rate": 9.929918049902062e-06,
+      "loss": 0.7912,
+      "step": 371
+    },
+    {
+      "epoch": 0.44631073785242953,
+      "grad_norm": 0.7731359524210224,
+      "learning_rate": 9.92874794422084e-06,
+      "loss": 0.7476,
+      "step": 372
+    },
+    {
+      "epoch": 0.4475104979004199,
+      "grad_norm": 0.8188756125557376,
+      "learning_rate": 9.927568221146401e-06,
+      "loss": 0.8029,
+      "step": 373
+    },
+    {
+      "epoch": 0.4487102579484103,
+      "grad_norm": 0.8780583103974543,
+      "learning_rate": 9.926378882980728e-06,
+      "loss": 0.8268,
+      "step": 374
+    },
+    {
+      "epoch": 0.44991001799640074,
+      "grad_norm": 0.8976930887917707,
+      "learning_rate": 9.925179932044552e-06,
+      "loss": 0.7914,
+      "step": 375
+    },
+    {
+      "epoch": 0.4511097780443911,
+      "grad_norm": 0.790418734596251,
+      "learning_rate": 9.923971370677375e-06,
+      "loss": 0.7754,
+      "step": 376
+    },
+    {
+      "epoch": 0.4523095380923815,
+      "grad_norm": 0.7540539012125508,
+      "learning_rate": 9.922753201237441e-06,
+      "loss": 0.8271,
+      "step": 377
+    },
+    {
+      "epoch": 0.45350929814037194,
+      "grad_norm": 0.8114545834788608,
+      "learning_rate": 9.921525426101745e-06,
+      "loss": 0.7598,
+      "step": 378
+    },
+    {
+      "epoch": 0.4547090581883623,
+      "grad_norm": 0.8148141210694403,
+      "learning_rate": 9.920288047666031e-06,
+      "loss": 0.7773,
+      "step": 379
+    },
+    {
+      "epoch": 0.4559088182363527,
+      "grad_norm": 0.7760783698206488,
+      "learning_rate": 9.919041068344774e-06,
+      "loss": 0.7527,
+      "step": 380
+    },
+    {
+      "epoch": 0.45710857828434315,
+      "grad_norm": 0.7331163486277703,
+      "learning_rate": 9.917784490571188e-06,
+      "loss": 0.7976,
+      "step": 381
+    },
+    {
+      "epoch": 0.45830833833233353,
+      "grad_norm": 0.8178911597875124,
+      "learning_rate": 9.916518316797212e-06,
+      "loss": 0.7989,
+      "step": 382
+    },
+    {
+      "epoch": 0.4595080983803239,
+      "grad_norm": 0.7708390841568354,
+      "learning_rate": 9.915242549493513e-06,
+      "loss": 0.7658,
+      "step": 383
+    },
+    {
+      "epoch": 0.46070785842831435,
+      "grad_norm": 0.7346178145033802,
+      "learning_rate": 9.91395719114948e-06,
+      "loss": 0.7896,
+      "step": 384
+    },
+    {
+      "epoch": 0.46190761847630474,
+      "grad_norm": 0.7831346158780439,
+      "learning_rate": 9.912662244273213e-06,
+      "loss": 0.7912,
+      "step": 385
+    },
+    {
+      "epoch": 0.4631073785242951,
+      "grad_norm": 0.7685210278278765,
+      "learning_rate": 9.911357711391519e-06,
+      "loss": 0.7363,
+      "step": 386
+    },
+    {
+      "epoch": 0.46430713857228556,
+      "grad_norm": 0.8417914054386219,
+      "learning_rate": 9.910043595049918e-06,
+      "loss": 0.8178,
+      "step": 387
+    },
+    {
+      "epoch": 0.46550689862027594,
+      "grad_norm": 0.7672805808964465,
+      "learning_rate": 9.908719897812622e-06,
+      "loss": 0.7157,
+      "step": 388
+    },
+    {
+      "epoch": 0.4667066586682663,
+      "grad_norm": 0.8649515383741985,
+      "learning_rate": 9.907386622262547e-06,
+      "loss": 0.8493,
+      "step": 389
+    },
+    {
+      "epoch": 0.46790641871625677,
+      "grad_norm": 0.7738870524107353,
+      "learning_rate": 9.906043771001289e-06,
+      "loss": 0.8194,
+      "step": 390
+    },
+    {
+      "epoch": 0.46910617876424715,
+      "grad_norm": 0.8084944877365888,
+      "learning_rate": 9.904691346649137e-06,
+      "loss": 0.7827,
+      "step": 391
+    },
+    {
+      "epoch": 0.47030593881223753,
+      "grad_norm": 0.8645182910322615,
+      "learning_rate": 9.903329351845055e-06,
+      "loss": 0.7888,
+      "step": 392
+    },
+    {
+      "epoch": 0.471505698860228,
+      "grad_norm": 0.7753940783542463,
+      "learning_rate": 9.901957789246683e-06,
+      "loss": 0.708,
+      "step": 393
+    },
+    {
+      "epoch": 0.47270545890821836,
+      "grad_norm": 0.7535155544641968,
+      "learning_rate": 9.900576661530334e-06,
+      "loss": 0.758,
+      "step": 394
+    },
+    {
+      "epoch": 0.47390521895620874,
+      "grad_norm": 0.7990248815331882,
+      "learning_rate": 9.89918597139098e-06,
+      "loss": 0.7712,
+      "step": 395
+    },
+    {
+      "epoch": 0.4751049790041992,
+      "grad_norm": 0.7872305119483683,
+      "learning_rate": 9.89778572154225e-06,
+      "loss": 0.8311,
+      "step": 396
+    },
+    {
+      "epoch": 0.47630473905218956,
+      "grad_norm": 0.790424891523676,
+      "learning_rate": 9.89637591471644e-06,
+      "loss": 0.7826,
+      "step": 397
+    },
+    {
+      "epoch": 0.47750449910017995,
+      "grad_norm": 0.9628608337412052,
+      "learning_rate": 9.894956553664479e-06,
+      "loss": 0.8336,
+      "step": 398
+    },
+    {
+      "epoch": 0.4787042591481704,
+      "grad_norm": 0.9252288172565076,
+      "learning_rate": 9.893527641155945e-06,
+      "loss": 0.7676,
+      "step": 399
+    },
+    {
+      "epoch": 0.47990401919616077,
+      "grad_norm": 0.8662616102690699,
+      "learning_rate": 9.892089179979056e-06,
+      "loss": 0.7821,
+      "step": 400
+    },
+    {
+      "epoch": 0.48110377924415115,
+      "grad_norm": 0.7515686991319386,
+      "learning_rate": 9.89064117294066e-06,
+      "loss": 0.7426,
+      "step": 401
+    },
+    {
+      "epoch": 0.4823035392921416,
+      "grad_norm": 1.0924459166312195,
+      "learning_rate": 9.889183622866232e-06,
+      "loss": 0.8409,
+      "step": 402
+    },
+    {
+      "epoch": 0.483503299340132,
+      "grad_norm": 0.9035652732859879,
+      "learning_rate": 9.887716532599868e-06,
+      "loss": 0.7565,
+      "step": 403
+    },
+    {
+      "epoch": 0.48470305938812236,
+      "grad_norm": 0.8786197862572435,
+      "learning_rate": 9.886239905004277e-06,
+      "loss": 0.832,
+      "step": 404
+    },
+    {
+      "epoch": 0.4859028194361128,
+      "grad_norm": 0.7910128169701829,
+      "learning_rate": 9.884753742960784e-06,
+      "loss": 0.7388,
+      "step": 405
+    },
+    {
+      "epoch": 0.4871025794841032,
+      "grad_norm": 0.9082566293088858,
+      "learning_rate": 9.883258049369313e-06,
+      "loss": 0.7749,
+      "step": 406
+    },
+    {
+      "epoch": 0.48830233953209357,
+      "grad_norm": 5.393295746353944,
+      "learning_rate": 9.881752827148391e-06,
+      "loss": 0.6713,
+      "step": 407
+    },
+    {
+      "epoch": 0.489502099580084,
+      "grad_norm": 0.8808053301797711,
+      "learning_rate": 9.880238079235134e-06,
+      "loss": 0.7838,
+      "step": 408
+    },
+    {
+      "epoch": 0.4907018596280744,
+      "grad_norm": 0.8549263656066898,
+      "learning_rate": 9.878713808585247e-06,
+      "loss": 0.7639,
+      "step": 409
+    },
+    {
+      "epoch": 0.49190161967606477,
+      "grad_norm": 0.7105156902384112,
+      "learning_rate": 9.877180018173019e-06,
+      "loss": 0.7337,
+      "step": 410
+    },
+    {
+      "epoch": 0.4931013797240552,
+      "grad_norm": 0.7739159427592175,
+      "learning_rate": 9.87563671099131e-06,
+      "loss": 0.7576,
+      "step": 411
+    },
+    {
+      "epoch": 0.4943011397720456,
+      "grad_norm": 0.8040600363561748,
+      "learning_rate": 9.874083890051553e-06,
+      "loss": 0.7507,
+      "step": 412
+    },
+    {
+      "epoch": 0.495500899820036,
+      "grad_norm": 0.9575415088088016,
+      "learning_rate": 9.872521558383748e-06,
+      "loss": 0.758,
+      "step": 413
+    },
+    {
+      "epoch": 0.4967006598680264,
+      "grad_norm": 0.8925178432779028,
+      "learning_rate": 9.870949719036446e-06,
+      "loss": 0.7158,
+      "step": 414
+    },
+    {
+      "epoch": 0.4979004199160168,
+      "grad_norm": 0.8368122703058055,
+      "learning_rate": 9.869368375076755e-06,
+      "loss": 0.7411,
+      "step": 415
+    },
+    {
+      "epoch": 0.4991001799640072,
+      "grad_norm": 0.8523083439993232,
+      "learning_rate": 9.86777752959033e-06,
+      "loss": 0.75,
+      "step": 416
+    },
+    {
+      "epoch": 0.5002999400119976,
+      "grad_norm": 0.7765675505221259,
+      "learning_rate": 9.866177185681361e-06,
+      "loss": 0.7819,
+      "step": 417
+    },
+    {
+      "epoch": 0.501499700059988,
+      "grad_norm": 0.8129680747900232,
+      "learning_rate": 9.864567346472578e-06,
+      "loss": 0.7734,
+      "step": 418
+    },
+    {
+      "epoch": 0.5026994601079784,
+      "grad_norm": 0.7888227168815115,
+      "learning_rate": 9.862948015105233e-06,
+      "loss": 0.7784,
+      "step": 419
+    },
+    {
+      "epoch": 0.5038992201559688,
+      "grad_norm": 0.7485734740834548,
+      "learning_rate": 9.861319194739109e-06,
+      "loss": 0.7798,
+      "step": 420
+    },
+    {
+      "epoch": 0.5050989802039592,
+      "grad_norm": 0.783877019342748,
+      "learning_rate": 9.859680888552497e-06,
+      "loss": 0.7235,
+      "step": 421
+    },
+    {
+      "epoch": 0.5062987402519497,
+      "grad_norm": 0.933538619958105,
+      "learning_rate": 9.858033099742197e-06,
+      "loss": 0.7915,
+      "step": 422
+    },
+    {
+      "epoch": 0.50749850029994,
+      "grad_norm": 0.6991884169094089,
+      "learning_rate": 9.856375831523517e-06,
+      "loss": 0.7194,
+      "step": 423
+    },
+    {
+      "epoch": 0.5086982603479304,
+      "grad_norm": 0.8599435760285431,
+      "learning_rate": 9.854709087130261e-06,
+      "loss": 0.7195,
+      "step": 424
+    },
+    {
+      "epoch": 0.5098980203959208,
+      "grad_norm": 0.7761136844143638,
+      "learning_rate": 9.853032869814721e-06,
+      "loss": 0.7657,
+      "step": 425
+    },
+    {
+      "epoch": 0.5110977804439112,
+      "grad_norm": 1.193889743974957,
+      "learning_rate": 9.851347182847678e-06,
+      "loss": 0.8358,
+      "step": 426
+    },
+    {
+      "epoch": 0.5122975404919016,
+      "grad_norm": 0.7678561049828674,
+      "learning_rate": 9.849652029518384e-06,
+      "loss": 0.7357,
+      "step": 427
+    },
+    {
+      "epoch": 0.5134973005398921,
+      "grad_norm": 0.7865687358146413,
+      "learning_rate": 9.847947413134568e-06,
+      "loss": 0.7414,
+      "step": 428
+    },
+    {
+      "epoch": 0.5146970605878824,
+      "grad_norm": 0.7346520114156417,
+      "learning_rate": 9.846233337022427e-06,
+      "loss": 0.7832,
+      "step": 429
+    },
+    {
+      "epoch": 0.5158968206358728,
+      "grad_norm": 0.9350176986428819,
+      "learning_rate": 9.844509804526606e-06,
+      "loss": 0.7505,
+      "step": 430
+    },
+    {
+      "epoch": 0.5170965806838632,
+      "grad_norm": 0.8296219956459046,
+      "learning_rate": 9.842776819010213e-06,
+      "loss": 0.7832,
+      "step": 431
+    },
+    {
+      "epoch": 0.5182963407318536,
+      "grad_norm": 0.8476492086914537,
+      "learning_rate": 9.841034383854795e-06,
+      "loss": 0.7001,
+      "step": 432
+    },
+    {
+      "epoch": 0.519496100779844,
+      "grad_norm": 0.7240816239335365,
+      "learning_rate": 9.83928250246034e-06,
+      "loss": 0.6933,
+      "step": 433
+    },
+    {
+      "epoch": 0.5206958608278345,
+      "grad_norm": 0.7405341057000968,
+      "learning_rate": 9.837521178245273e-06,
+      "loss": 0.7561,
+      "step": 434
+    },
+    {
+      "epoch": 0.5218956208758249,
+      "grad_norm": 0.7692071981948461,
+      "learning_rate": 9.835750414646433e-06,
+      "loss": 0.693,
+      "step": 435
+    },
+    {
+      "epoch": 0.5230953809238152,
+      "grad_norm": 0.8287045028686297,
+      "learning_rate": 9.833970215119088e-06,
+      "loss": 0.7489,
+      "step": 436
+    },
+    {
+      "epoch": 0.5242951409718056,
+      "grad_norm": 0.795961787816489,
+      "learning_rate": 9.832180583136916e-06,
+      "loss": 0.7446,
+      "step": 437
+    },
+    {
+      "epoch": 0.525494901019796,
+      "grad_norm": 0.7581333038407717,
+      "learning_rate": 9.830381522191997e-06,
+      "loss": 0.7231,
+      "step": 438
+    },
+    {
+      "epoch": 0.5266946610677864,
+      "grad_norm": 0.9303998090134032,
+      "learning_rate": 9.828573035794815e-06,
+      "loss": 0.8329,
+      "step": 439
+    },
+    {
+      "epoch": 0.5278944211157769,
+      "grad_norm": 0.6566928859534891,
+      "learning_rate": 9.826755127474241e-06,
+      "loss": 0.6941,
+      "step": 440
+    },
+    {
+      "epoch": 0.5290941811637673,
+      "grad_norm": 0.7135091281559202,
+      "learning_rate": 9.824927800777535e-06,
+      "loss": 0.7822,
+      "step": 441
+    },
+    {
+      "epoch": 0.5302939412117577,
+      "grad_norm": 0.7534927462043957,
+      "learning_rate": 9.823091059270329e-06,
+      "loss": 0.8006,
+      "step": 442
+    },
+    {
+      "epoch": 0.531493701259748,
+      "grad_norm": 0.7406618710423156,
+      "learning_rate": 9.821244906536633e-06,
+      "loss": 0.7802,
+      "step": 443
+    },
+    {
+      "epoch": 0.5326934613077384,
+      "grad_norm": 0.7491038976546027,
+      "learning_rate": 9.819389346178814e-06,
+      "loss": 0.7735,
+      "step": 444
+    },
+    {
+      "epoch": 0.5338932213557288,
+      "grad_norm": 0.7822185677172192,
+      "learning_rate": 9.817524381817603e-06,
+      "loss": 0.7542,
+      "step": 445
+    },
+    {
+      "epoch": 0.5350929814037193,
+      "grad_norm": 0.7128977961072269,
+      "learning_rate": 9.815650017092078e-06,
+      "loss": 0.7778,
+      "step": 446
+    },
+    {
+      "epoch": 0.5362927414517097,
+      "grad_norm": 0.6904057914443502,
+      "learning_rate": 9.813766255659654e-06,
+      "loss": 0.734,
+      "step": 447
+    },
+    {
+      "epoch": 0.5374925014997001,
+      "grad_norm": 0.8427999544381056,
+      "learning_rate": 9.811873101196093e-06,
+      "loss": 0.7819,
+      "step": 448
+    },
+    {
+      "epoch": 0.5386922615476905,
+      "grad_norm": 0.7866934841281968,
+      "learning_rate": 9.809970557395476e-06,
+      "loss": 0.7153,
+      "step": 449
+    },
+    {
+      "epoch": 0.5398920215956808,
+      "grad_norm": 0.8032501027155347,
+      "learning_rate": 9.80805862797021e-06,
+      "loss": 0.7641,
+      "step": 450
+    },
+    {
+      "epoch": 0.5410917816436712,
+      "grad_norm": 0.7686919182685255,
+      "learning_rate": 9.806137316651013e-06,
+      "loss": 0.7456,
+      "step": 451
+    },
+    {
+      "epoch": 0.5422915416916617,
+      "grad_norm": 0.8177582066471052,
+      "learning_rate": 9.804206627186911e-06,
+      "loss": 0.7931,
+      "step": 452
+    },
+    {
+      "epoch": 0.5434913017396521,
+      "grad_norm": 0.9006022094358479,
+      "learning_rate": 9.802266563345235e-06,
+      "loss": 0.8071,
+      "step": 453
+    },
+    {
+      "epoch": 0.5446910617876425,
+      "grad_norm": 0.7518170142436402,
+      "learning_rate": 9.8003171289116e-06,
+      "loss": 0.8075,
+      "step": 454
+    },
+    {
+      "epoch": 0.5458908218356329,
+      "grad_norm": 0.7514481594832335,
+      "learning_rate": 9.798358327689906e-06,
+      "loss": 0.6912,
+      "step": 455
+    },
+    {
+      "epoch": 0.5470905818836233,
+      "grad_norm": 0.7915234537215006,
+      "learning_rate": 9.796390163502334e-06,
+      "loss": 0.7001,
+      "step": 456
+    },
+    {
+      "epoch": 0.5482903419316136,
+      "grad_norm": 0.8439870437962147,
+      "learning_rate": 9.794412640189336e-06,
+      "loss": 0.7564,
+      "step": 457
+    },
+    {
+      "epoch": 0.5494901019796041,
+      "grad_norm": 0.7780134611809413,
+      "learning_rate": 9.792425761609623e-06,
+      "loss": 0.7974,
+      "step": 458
+    },
+    {
+      "epoch": 0.5506898620275945,
+      "grad_norm": 0.9166536013833003,
+      "learning_rate": 9.790429531640163e-06,
+      "loss": 0.8225,
+      "step": 459
+    },
+    {
+      "epoch": 0.5518896220755849,
+      "grad_norm": 0.9079256575884883,
+      "learning_rate": 9.788423954176166e-06,
+      "loss": 0.7794,
+      "step": 460
+    },
+    {
+      "epoch": 0.5530893821235753,
+      "grad_norm": 0.7992584892279518,
+      "learning_rate": 9.786409033131092e-06,
+      "loss": 0.7594,
+      "step": 461
+    },
+    {
+      "epoch": 0.5542891421715657,
+      "grad_norm": 0.7316252955597614,
+      "learning_rate": 9.784384772436624e-06,
+      "loss": 0.7028,
+      "step": 462
+    },
+    {
+      "epoch": 0.555488902219556,
+      "grad_norm": 0.8776441197086907,
+      "learning_rate": 9.782351176042668e-06,
+      "loss": 0.7024,
+      "step": 463
+    },
+    {
+      "epoch": 0.5566886622675465,
+      "grad_norm": 0.8485818018512938,
+      "learning_rate": 9.780308247917355e-06,
+      "loss": 0.816,
+      "step": 464
+    },
+    {
+      "epoch": 0.5578884223155369,
+      "grad_norm": 0.7292056958121589,
+      "learning_rate": 9.778255992047024e-06,
+      "loss": 0.7556,
+      "step": 465
+    },
+    {
+      "epoch": 0.5590881823635273,
+      "grad_norm": 0.886905755147264,
+      "learning_rate": 9.776194412436202e-06,
+      "loss": 0.7597,
+      "step": 466
+    },
+    {
+      "epoch": 0.5602879424115177,
+      "grad_norm": 0.8700121060467031,
+      "learning_rate": 9.774123513107628e-06,
+      "loss": 0.7258,
+      "step": 467
+    },
+    {
+      "epoch": 0.5614877024595081,
+      "grad_norm": 0.8348518898244337,
+      "learning_rate": 9.77204329810221e-06,
+      "loss": 0.6959,
+      "step": 468
+    },
+    {
+      "epoch": 0.5626874625074985,
+      "grad_norm": 0.8063492786549548,
+      "learning_rate": 9.769953771479048e-06,
+      "loss": 0.7216,
+      "step": 469
+    },
+    {
+      "epoch": 0.563887222555489,
+      "grad_norm": 15.790661371937125,
+      "learning_rate": 9.767854937315398e-06,
+      "loss": 0.9547,
+      "step": 470
+    },
+    {
+      "epoch": 0.5650869826034793,
+      "grad_norm": 0.7455923646791719,
+      "learning_rate": 9.765746799706688e-06,
+      "loss": 0.7034,
+      "step": 471
+    },
+    {
+      "epoch": 0.5662867426514697,
+      "grad_norm": 0.7161568027914286,
+      "learning_rate": 9.763629362766495e-06,
+      "loss": 0.7697,
+      "step": 472
+    },
+    {
+      "epoch": 0.5674865026994601,
+      "grad_norm": 0.7960011292767376,
+      "learning_rate": 9.761502630626544e-06,
+      "loss": 0.74,
+      "step": 473
+    },
+    {
+      "epoch": 0.5686862627474505,
+      "grad_norm": 0.8386461066974339,
+      "learning_rate": 9.759366607436694e-06,
+      "loss": 0.7591,
+      "step": 474
+    },
+    {
+      "epoch": 0.5698860227954409,
+      "grad_norm": 0.8219184010947729,
+      "learning_rate": 9.757221297364936e-06,
+      "loss": 0.7985,
+      "step": 475
+    },
+    {
+      "epoch": 0.5710857828434314,
+      "grad_norm": 0.8526023917338333,
+      "learning_rate": 9.75506670459738e-06,
+      "loss": 0.7363,
+      "step": 476
+    },
+    {
+      "epoch": 0.5722855428914218,
+      "grad_norm": 0.7158601779007123,
+      "learning_rate": 9.752902833338256e-06,
+      "loss": 0.784,
+      "step": 477
+    },
+    {
+      "epoch": 0.5734853029394121,
+      "grad_norm": 0.9009368283263316,
+      "learning_rate": 9.75072968780989e-06,
+      "loss": 0.76,
+      "step": 478
+    },
+    {
+      "epoch": 0.5746850629874025,
+      "grad_norm": 0.844246397660529,
+      "learning_rate": 9.74854727225271e-06,
+      "loss": 0.7456,
+      "step": 479
+    },
+    {
+      "epoch": 0.5758848230353929,
+      "grad_norm": 0.9418669055461818,
+      "learning_rate": 9.746355590925231e-06,
+      "loss": 0.7488,
+      "step": 480
+    },
+    {
+      "epoch": 0.5770845830833833,
+      "grad_norm": 0.8489283035012338,
+      "learning_rate": 9.744154648104049e-06,
+      "loss": 0.7222,
+      "step": 481
+    },
+    {
+      "epoch": 0.5782843431313738,
+      "grad_norm": 0.7383667346126159,
+      "learning_rate": 9.74194444808383e-06,
+      "loss": 0.7387,
+      "step": 482
+    },
+    {
+      "epoch": 0.5794841031793642,
+      "grad_norm": 0.838663810803923,
+      "learning_rate": 9.739724995177309e-06,
+      "loss": 0.7478,
+      "step": 483
+    },
+    {
+      "epoch": 0.5806838632273545,
+      "grad_norm": 0.7732760292403525,
+      "learning_rate": 9.737496293715268e-06,
+      "loss": 0.7577,
+      "step": 484
+    },
+    {
+      "epoch": 0.5818836232753449,
+      "grad_norm": 0.8358539665976709,
+      "learning_rate": 9.735258348046538e-06,
+      "loss": 0.7199,
+      "step": 485
+    },
+    {
+      "epoch": 0.5830833833233353,
+      "grad_norm": 0.931301011982876,
+      "learning_rate": 9.73301116253799e-06,
+      "loss": 0.751,
+      "step": 486
+    },
+    {
+      "epoch": 0.5842831433713257,
+      "grad_norm": 0.8287516764022196,
+      "learning_rate": 9.730754741574529e-06,
+      "loss": 0.7583,
+      "step": 487
+    },
+    {
+      "epoch": 0.5854829034193162,
+      "grad_norm": 0.824571405077036,
+      "learning_rate": 9.72848908955907e-06,
+      "loss": 0.7018,
+      "step": 488
+    },
+    {
+      "epoch": 0.5866826634673066,
+      "grad_norm": 0.8092772472268522,
+      "learning_rate": 9.726214210912549e-06,
+      "loss": 0.7899,
+      "step": 489
+    },
+    {
+      "epoch": 0.587882423515297,
+      "grad_norm": 0.7579661335080632,
+      "learning_rate": 9.723930110073902e-06,
+      "loss": 0.6777,
+      "step": 490
+    },
+    {
+      "epoch": 0.5890821835632873,
+      "grad_norm": 0.9801294633681042,
+      "learning_rate": 9.721636791500065e-06,
+      "loss": 0.8294,
+      "step": 491
+    },
+    {
+      "epoch": 0.5902819436112777,
+      "grad_norm": 0.8218005486888227,
+      "learning_rate": 9.719334259665951e-06,
+      "loss": 0.7115,
+      "step": 492
+    },
+    {
+      "epoch": 0.5914817036592681,
+      "grad_norm": 0.8044698534033967,
+      "learning_rate": 9.71702251906446e-06,
+      "loss": 0.7619,
+      "step": 493
+    },
+    {
+      "epoch": 0.5926814637072585,
+      "grad_norm": 0.7541226470722168,
+      "learning_rate": 9.714701574206457e-06,
+      "loss": 0.7595,
+      "step": 494
+    },
+    {
+      "epoch": 0.593881223755249,
+      "grad_norm": 0.8013622981520949,
+      "learning_rate": 9.712371429620767e-06,
+      "loss": 0.7405,
+      "step": 495
+    },
+    {
+      "epoch": 0.5950809838032394,
+      "grad_norm": 0.6836569320702858,
+      "learning_rate": 9.710032089854166e-06,
+      "loss": 0.6942,
+      "step": 496
+    },
+    {
+      "epoch": 0.5962807438512298,
+      "grad_norm": 0.8065057972585213,
+      "learning_rate": 9.707683559471374e-06,
+      "loss": 0.8385,
+      "step": 497
+    },
+    {
+      "epoch": 0.5974805038992201,
+      "grad_norm": 0.7580471644835146,
+      "learning_rate": 9.705325843055045e-06,
+      "loss": 0.7433,
+      "step": 498
+    },
+    {
+      "epoch": 0.5986802639472105,
+      "grad_norm": 0.8606247011714901,
+      "learning_rate": 9.702958945205753e-06,
+      "loss": 0.7035,
+      "step": 499
+    },
+    {
+      "epoch": 0.5998800239952009,
+      "grad_norm": 1.5950934849613543,
+      "learning_rate": 9.700582870541995e-06,
+      "loss": 0.7909,
+      "step": 500
+    },
+    {
+      "epoch": 0.6010797840431914,
+      "grad_norm": 0.7731565625014056,
+      "learning_rate": 9.698197623700168e-06,
+      "loss": 0.7965,
+      "step": 501
+    },
+    {
+      "epoch": 0.6022795440911818,
+      "grad_norm": 0.8349098467362701,
+      "learning_rate": 9.69580320933457e-06,
+      "loss": 0.7261,
+      "step": 502
+    },
+    {
+      "epoch": 0.6034793041391722,
+      "grad_norm": 0.754057594438619,
+      "learning_rate": 9.693399632117385e-06,
+      "loss": 0.6971,
+      "step": 503
+    },
+    {
+      "epoch": 0.6046790641871626,
+      "grad_norm": 0.8023211344060523,
+      "learning_rate": 9.69098689673868e-06,
+      "loss": 0.7167,
+      "step": 504
+    },
+    {
+      "epoch": 0.6058788242351529,
+      "grad_norm": 1.0058172149504045,
+      "learning_rate": 9.688565007906387e-06,
+      "loss": 0.7819,
+      "step": 505
+    },
+    {
+      "epoch": 0.6070785842831433,
+      "grad_norm": 0.7242215853525499,
+      "learning_rate": 9.686133970346305e-06,
+      "loss": 0.6986,
+      "step": 506
+    },
+    {
+      "epoch": 0.6082783443311338,
+      "grad_norm": 0.7901538256806555,
+      "learning_rate": 9.68369378880208e-06,
+      "loss": 0.7646,
+      "step": 507
+    },
+    {
+      "epoch": 0.6094781043791242,
+      "grad_norm": 0.7503104666431076,
+      "learning_rate": 9.6812444680352e-06,
+      "loss": 0.7561,
+      "step": 508
+    },
+    {
+      "epoch": 0.6106778644271146,
+      "grad_norm": 0.7788288768425731,
+      "learning_rate": 9.678786012824995e-06,
+      "loss": 0.7616,
+      "step": 509
+    },
+    {
+      "epoch": 0.611877624475105,
+      "grad_norm": 0.7603265669161717,
+      "learning_rate": 9.676318427968608e-06,
+      "loss": 0.7311,
+      "step": 510
+    },
+    {
+      "epoch": 0.6130773845230953,
+      "grad_norm": 0.7078386385583856,
+      "learning_rate": 9.673841718281e-06,
+      "loss": 0.6879,
+      "step": 511
+    },
+    {
+      "epoch": 0.6142771445710857,
+      "grad_norm": 0.8055381057080542,
+      "learning_rate": 9.67135588859494e-06,
+      "loss": 0.7147,
+      "step": 512
+    },
+    {
+      "epoch": 0.6154769046190762,
+      "grad_norm": 0.7631878298047712,
+      "learning_rate": 9.668860943760992e-06,
+      "loss": 0.6773,
+      "step": 513
+    },
+    {
+      "epoch": 0.6166766646670666,
+      "grad_norm": 0.6465663248068814,
+      "learning_rate": 9.6663568886475e-06,
+      "loss": 0.7829,
+      "step": 514
+    },
+    {
+      "epoch": 0.617876424715057,
+      "grad_norm": 0.8571734522586008,
+      "learning_rate": 9.663843728140597e-06,
+      "loss": 0.7633,
+      "step": 515
+    },
+    {
+      "epoch": 0.6190761847630474,
+      "grad_norm": 0.8017497846966288,
+      "learning_rate": 9.661321467144172e-06,
+      "loss": 0.7087,
+      "step": 516
+    },
+    {
+      "epoch": 0.6202759448110378,
+      "grad_norm": 0.7759613883579511,
+      "learning_rate": 9.658790110579876e-06,
+      "loss": 0.7867,
+      "step": 517
+    },
+    {
+      "epoch": 0.6214757048590281,
+      "grad_norm": 0.8554914259201436,
+      "learning_rate": 9.656249663387107e-06,
+      "loss": 0.6902,
+      "step": 518
+    },
+    {
+      "epoch": 0.6226754649070186,
+      "grad_norm": 0.8565972170846713,
+      "learning_rate": 9.653700130523004e-06,
+      "loss": 0.816,
+      "step": 519
+    },
+    {
+      "epoch": 0.623875224955009,
+      "grad_norm": 0.8071076828122373,
+      "learning_rate": 9.651141516962431e-06,
+      "loss": 0.7227,
+      "step": 520
+    },
+    {
+      "epoch": 0.6250749850029994,
+      "grad_norm": 0.7660995482695625,
+      "learning_rate": 9.648573827697974e-06,
+      "loss": 0.7861,
+      "step": 521
+    },
+    {
+      "epoch": 0.6262747450509898,
+      "grad_norm": 1.0283097998930535,
+      "learning_rate": 9.645997067739929e-06,
+      "loss": 0.7303,
+      "step": 522
+    },
+    {
+      "epoch": 0.6274745050989802,
+      "grad_norm": 0.7745683261007134,
+      "learning_rate": 9.643411242116286e-06,
+      "loss": 0.6725,
+      "step": 523
+    },
+    {
+      "epoch": 0.6286742651469706,
+      "grad_norm": 0.8609117214762857,
+      "learning_rate": 9.64081635587273e-06,
+      "loss": 0.796,
+      "step": 524
+    },
+    {
+      "epoch": 0.629874025194961,
+      "grad_norm": 0.7864348864977745,
+      "learning_rate": 9.638212414072625e-06,
+      "loss": 0.7589,
+      "step": 525
+    },
+    {
+      "epoch": 0.6310737852429514,
+      "grad_norm": 0.7066437056556076,
+      "learning_rate": 9.635599421797004e-06,
+      "loss": 0.7414,
+      "step": 526
+    },
+    {
+      "epoch": 0.6322735452909418,
+      "grad_norm": 0.7634711558250465,
+      "learning_rate": 9.63297738414456e-06,
+      "loss": 0.7523,
+      "step": 527
+    },
+    {
+      "epoch": 0.6334733053389322,
+      "grad_norm": 0.7460018288720783,
+      "learning_rate": 9.630346306231637e-06,
+      "loss": 0.6864,
+      "step": 528
+    },
+    {
+      "epoch": 0.6346730653869226,
+      "grad_norm": 0.7895072856945539,
+      "learning_rate": 9.627706193192219e-06,
+      "loss": 0.7236,
+      "step": 529
+    },
+    {
+      "epoch": 0.635872825434913,
+      "grad_norm": 0.8164390104749236,
+      "learning_rate": 9.625057050177918e-06,
+      "loss": 0.6658,
+      "step": 530
+    },
+    {
+      "epoch": 0.6370725854829035,
+      "grad_norm": 0.65954773123744,
+      "learning_rate": 9.622398882357968e-06,
+      "loss": 0.7595,
+      "step": 531
+    },
+    {
+      "epoch": 0.6382723455308938,
+      "grad_norm": 0.9010898382276785,
+      "learning_rate": 9.619731694919212e-06,
+      "loss": 0.7702,
+      "step": 532
+    },
+    {
+      "epoch": 0.6394721055788842,
+      "grad_norm": 0.8386415071967853,
+      "learning_rate": 9.617055493066097e-06,
+      "loss": 0.6642,
+      "step": 533
+    },
+    {
+      "epoch": 0.6406718656268746,
+      "grad_norm": 0.8006666484897155,
+      "learning_rate": 9.614370282020652e-06,
+      "loss": 0.7895,
+      "step": 534
+    },
+    {
+      "epoch": 0.641871625674865,
+      "grad_norm": 0.6746633069935659,
+      "learning_rate": 9.611676067022492e-06,
+      "loss": 0.7577,
+      "step": 535
+    },
+    {
+      "epoch": 0.6430713857228554,
+      "grad_norm": 0.8148039263840821,
+      "learning_rate": 9.6089728533288e-06,
+      "loss": 0.7268,
+      "step": 536
+    },
+    {
+      "epoch": 0.6442711457708459,
+      "grad_norm": 0.8016142578877647,
+      "learning_rate": 9.606260646214314e-06,
+      "loss": 0.6941,
+      "step": 537
+    },
+    {
+      "epoch": 0.6454709058188363,
+      "grad_norm": 0.718756601136374,
+      "learning_rate": 9.603539450971326e-06,
+      "loss": 0.7473,
+      "step": 538
+    },
+    {
+      "epoch": 0.6466706658668266,
+      "grad_norm": 0.6757552684273367,
+      "learning_rate": 9.600809272909663e-06,
+      "loss": 0.7058,
+      "step": 539
+    },
+    {
+      "epoch": 0.647870425914817,
+      "grad_norm": 0.8704025257703171,
+      "learning_rate": 9.598070117356684e-06,
+      "loss": 0.8438,
+      "step": 540
+    },
+    {
+      "epoch": 0.6490701859628074,
+      "grad_norm": 0.667088769868094,
+      "learning_rate": 9.59532198965726e-06,
+      "loss": 0.7659,
+      "step": 541
+    },
+    {
+      "epoch": 0.6502699460107978,
+      "grad_norm": 0.9260561307588099,
+      "learning_rate": 9.59256489517377e-06,
+      "loss": 0.8057,
+      "step": 542
+    },
+    {
+      "epoch": 0.6514697060587883,
+      "grad_norm": 0.8775666263474623,
+      "learning_rate": 9.589798839286098e-06,
+      "loss": 0.7711,
+      "step": 543
+    },
+    {
+      "epoch": 0.6526694661067787,
+      "grad_norm": 0.6877550211780249,
+      "learning_rate": 9.587023827391603e-06,
+      "loss": 0.7835,
+      "step": 544
+    },
+    {
+      "epoch": 0.6538692261547691,
+      "grad_norm": 0.7349863305923766,
+      "learning_rate": 9.584239864905128e-06,
+      "loss": 0.7296,
+      "step": 545
+    },
+    {
+      "epoch": 0.6550689862027594,
+      "grad_norm": 0.8218995233024665,
+      "learning_rate": 9.581446957258975e-06,
+      "loss": 0.7829,
+      "step": 546
+    },
+    {
+      "epoch": 0.6562687462507498,
+      "grad_norm": 0.8134993669043972,
+      "learning_rate": 9.578645109902904e-06,
+      "loss": 0.765,
+      "step": 547
+    },
+    {
+      "epoch": 0.6574685062987402,
+      "grad_norm": 0.8557360032972663,
+      "learning_rate": 9.575834328304121e-06,
+      "loss": 0.7334,
+      "step": 548
+    },
+    {
+      "epoch": 0.6586682663467307,
+      "grad_norm": 0.7914397185522426,
+      "learning_rate": 9.57301461794726e-06,
+      "loss": 0.7127,
+      "step": 549
+    },
+    {
+      "epoch": 0.6598680263947211,
+      "grad_norm": 0.7734053557921267,
+      "learning_rate": 9.570185984334383e-06,
+      "loss": 0.7695,
+      "step": 550
+    },
+    {
+      "epoch": 0.6610677864427115,
+      "grad_norm": 0.880648269242868,
+      "learning_rate": 9.567348432984957e-06,
+      "loss": 0.7167,
+      "step": 551
+    },
+    {
+      "epoch": 0.6622675464907019,
+      "grad_norm": 0.7940200917011428,
+      "learning_rate": 9.56450196943586e-06,
+      "loss": 0.7364,
+      "step": 552
+    },
+    {
+      "epoch": 0.6634673065386922,
+      "grad_norm": 0.6781100978857986,
+      "learning_rate": 9.561646599241349e-06,
+      "loss": 0.6847,
+      "step": 553
+    },
+    {
+      "epoch": 0.6646670665866826,
+      "grad_norm": 0.7121065187027151,
+      "learning_rate": 9.558782327973068e-06,
+      "loss": 0.759,
+      "step": 554
+    },
+    {
+      "epoch": 0.6658668266346731,
+      "grad_norm": 0.9493717892282374,
+      "learning_rate": 9.555909161220025e-06,
+      "loss": 0.8074,
+      "step": 555
+    },
+    {
+      "epoch": 0.6670665866826635,
+      "grad_norm": 0.8350439792211257,
+      "learning_rate": 9.553027104588593e-06,
+      "loss": 0.7336,
+      "step": 556
+    },
+    {
+      "epoch": 0.6682663467306539,
+      "grad_norm": 0.7868865491202452,
+      "learning_rate": 9.550136163702482e-06,
+      "loss": 0.7655,
+      "step": 557
+    },
+    {
+      "epoch": 0.6694661067786443,
+      "grad_norm": 0.7655933329400806,
+      "learning_rate": 9.547236344202744e-06,
+      "loss": 0.7437,
+      "step": 558
+    },
+    {
+      "epoch": 0.6706658668266346,
+      "grad_norm": 0.9690309207482182,
+      "learning_rate": 9.544327651747753e-06,
+      "loss": 0.8109,
+      "step": 559
+    },
+    {
+      "epoch": 0.671865626874625,
+      "grad_norm": 0.7032506626870428,
+      "learning_rate": 9.541410092013201e-06,
+      "loss": 0.6855,
+      "step": 560
+    },
+    {
+      "epoch": 0.6730653869226155,
+      "grad_norm": 0.7619673285229566,
+      "learning_rate": 9.538483670692076e-06,
+      "loss": 0.6919,
+      "step": 561
+    },
+    {
+      "epoch": 0.6742651469706059,
+      "grad_norm": 0.7359622527888945,
+      "learning_rate": 9.535548393494661e-06,
+      "loss": 0.7804,
+      "step": 562
+    },
+    {
+      "epoch": 0.6754649070185963,
+      "grad_norm": 0.7216172279516673,
+      "learning_rate": 9.532604266148521e-06,
+      "loss": 0.7119,
+      "step": 563
+    },
+    {
+      "epoch": 0.6766646670665867,
+      "grad_norm": 0.7947602490045758,
+      "learning_rate": 9.529651294398485e-06,
+      "loss": 0.7116,
+      "step": 564
+    },
+    {
+      "epoch": 0.6778644271145771,
+      "grad_norm": 0.7938540244579113,
+      "learning_rate": 9.526689484006649e-06,
+      "loss": 0.7139,
+      "step": 565
+    },
+    {
+      "epoch": 0.6790641871625674,
+      "grad_norm": 0.7557012771757716,
+      "learning_rate": 9.523718840752343e-06,
+      "loss": 0.7737,
+      "step": 566
+    },
+    {
+      "epoch": 0.6802639472105579,
+      "grad_norm": 0.7130200527476641,
+      "learning_rate": 9.520739370432143e-06,
+      "loss": 0.7284,
+      "step": 567
+    },
+    {
+      "epoch": 0.6814637072585483,
+      "grad_norm": 0.8339374298828425,
+      "learning_rate": 9.517751078859848e-06,
+      "loss": 0.759,
+      "step": 568
+    },
+    {
+      "epoch": 0.6826634673065387,
+      "grad_norm": 0.7492909405177227,
+      "learning_rate": 9.514753971866463e-06,
+      "loss": 0.7311,
+      "step": 569
+    },
+    {
+      "epoch": 0.6838632273545291,
+      "grad_norm": 0.9066211146549824,
+      "learning_rate": 9.511748055300201e-06,
+      "loss": 0.8017,
+      "step": 570
+    },
+    {
+      "epoch": 0.6850629874025195,
+      "grad_norm": 0.720161620025462,
+      "learning_rate": 9.508733335026462e-06,
+      "loss": 0.773,
+      "step": 571
+    },
+    {
+      "epoch": 0.6862627474505099,
+      "grad_norm": 0.6967812860809943,
+      "learning_rate": 9.505709816927823e-06,
+      "loss": 0.7461,
+      "step": 572
+    },
+    {
+      "epoch": 0.6874625074985004,
+      "grad_norm": 0.7860373824966781,
+      "learning_rate": 9.502677506904035e-06,
+      "loss": 0.7483,
+      "step": 573
+    },
+    {
+      "epoch": 0.6886622675464907,
+      "grad_norm": 0.7275359182122862,
+      "learning_rate": 9.499636410871994e-06,
+      "loss": 0.8021,
+      "step": 574
+    },
+    {
+      "epoch": 0.6898620275944811,
+      "grad_norm": 0.7520613473149373,
+      "learning_rate": 9.496586534765752e-06,
+      "loss": 0.7328,
+      "step": 575
+    },
+    {
+      "epoch": 0.6910617876424715,
+      "grad_norm": 0.7301675898688976,
+      "learning_rate": 9.493527884536487e-06,
+      "loss": 0.6881,
+      "step": 576
+    },
+    {
+      "epoch": 0.6922615476904619,
+      "grad_norm": 0.7678401412943116,
+      "learning_rate": 9.490460466152491e-06,
+      "loss": 0.7324,
+      "step": 577
+    },
+    {
+      "epoch": 0.6934613077384523,
+      "grad_norm": 0.8612062026391113,
+      "learning_rate": 9.487384285599179e-06,
+      "loss": 0.768,
+      "step": 578
+    },
+    {
+      "epoch": 0.6946610677864428,
+      "grad_norm": 0.8147151721250593,
+      "learning_rate": 9.484299348879055e-06,
+      "loss": 0.6829,
+      "step": 579
+    },
+    {
+      "epoch": 0.6958608278344331,
+      "grad_norm": 0.8738236748718973,
+      "learning_rate": 9.48120566201171e-06,
+      "loss": 0.82,
+      "step": 580
+    },
+    {
+      "epoch": 0.6970605878824235,
+      "grad_norm": 0.8986436192452325,
+      "learning_rate": 9.478103231033808e-06,
+      "loss": 0.6793,
+      "step": 581
+    },
+    {
+      "epoch": 0.6982603479304139,
+      "grad_norm": 0.8904464145598673,
+      "learning_rate": 9.474992061999081e-06,
+      "loss": 0.7572,
+      "step": 582
+    },
+    {
+      "epoch": 0.6994601079784043,
+      "grad_norm": 0.6574883638307059,
+      "learning_rate": 9.471872160978303e-06,
+      "loss": 0.6443,
+      "step": 583
+    },
+    {
+      "epoch": 0.7006598680263947,
+      "grad_norm": 0.8985559380717492,
+      "learning_rate": 9.468743534059295e-06,
+      "loss": 0.7783,
+      "step": 584
+    },
+    {
+      "epoch": 0.7018596280743852,
+      "grad_norm": 0.7431036219557079,
+      "learning_rate": 9.465606187346897e-06,
+      "loss": 0.6254,
+      "step": 585
+    },
+    {
+      "epoch": 0.7030593881223756,
+      "grad_norm": 0.7728865866981772,
+      "learning_rate": 9.46246012696297e-06,
+      "loss": 0.7561,
+      "step": 586
+    },
+    {
+      "epoch": 0.704259148170366,
+      "grad_norm": 0.7501647430361906,
+      "learning_rate": 9.459305359046372e-06,
+      "loss": 0.6625,
+      "step": 587
+    },
+    {
+      "epoch": 0.7054589082183563,
+      "grad_norm": 0.7594681233852942,
+      "learning_rate": 9.456141889752959e-06,
+      "loss": 0.7741,
+      "step": 588
+    },
+    {
+      "epoch": 0.7066586682663467,
+      "grad_norm": 0.8174670353526049,
+      "learning_rate": 9.452969725255558e-06,
+      "loss": 0.7341,
+      "step": 589
+    },
+    {
+      "epoch": 0.7078584283143371,
+      "grad_norm": 0.7618253465793706,
+      "learning_rate": 9.449788871743972e-06,
+      "loss": 0.7444,
+      "step": 590
+    },
+    {
+      "epoch": 0.7090581883623275,
+      "grad_norm": 0.7474309464649311,
+      "learning_rate": 9.446599335424948e-06,
+      "loss": 0.7273,
+      "step": 591
+    },
+    {
+      "epoch": 0.710257948410318,
+      "grad_norm": 0.7187702974171543,
+      "learning_rate": 9.443401122522185e-06,
+      "loss": 0.7459,
+      "step": 592
+    },
+    {
+      "epoch": 0.7114577084583084,
+      "grad_norm": 0.7185805793407908,
+      "learning_rate": 9.440194239276308e-06,
+      "loss": 0.6859,
+      "step": 593
+    },
+    {
+      "epoch": 0.7126574685062987,
+      "grad_norm": 0.8506171645399014,
+      "learning_rate": 9.43697869194486e-06,
+      "loss": 0.6996,
+      "step": 594
+    },
+    {
+      "epoch": 0.7138572285542891,
+      "grad_norm": 0.9093408934214373,
+      "learning_rate": 9.433754486802291e-06,
+      "loss": 0.7084,
+      "step": 595
+    },
+    {
+      "epoch": 0.7150569886022795,
+      "grad_norm": 0.7899145578993906,
+      "learning_rate": 9.430521630139946e-06,
+      "loss": 0.6313,
+      "step": 596
+    },
+    {
+      "epoch": 0.7162567486502699,
+      "grad_norm": 0.871030688739406,
+      "learning_rate": 9.427280128266049e-06,
+      "loss": 0.7541,
+      "step": 597
+    },
+    {
+      "epoch": 0.7174565086982604,
+      "grad_norm": 0.7650457947902857,
+      "learning_rate": 9.424029987505699e-06,
+      "loss": 0.6741,
+      "step": 598
+    },
+    {
+      "epoch": 0.7186562687462508,
+      "grad_norm": 0.7472119241841967,
+      "learning_rate": 9.420771214200843e-06,
+      "loss": 0.7183,
+      "step": 599
+    },
+    {
+      "epoch": 0.7198560287942412,
+      "grad_norm": 0.8238613774473978,
+      "learning_rate": 9.417503814710277e-06,
+      "loss": 0.7379,
+      "step": 600
+    },
+    {
+      "epoch": 0.7210557888422315,
+      "grad_norm": 0.6805508800539326,
+      "learning_rate": 9.414227795409634e-06,
+      "loss": 0.6928,
+      "step": 601
+    },
+    {
+      "epoch": 0.7222555488902219,
+      "grad_norm": 0.8624131473450718,
+      "learning_rate": 9.410943162691359e-06,
+      "loss": 0.7423,
+      "step": 602
+    },
+    {
+      "epoch": 0.7234553089382123,
+      "grad_norm": 0.747745459350849,
+      "learning_rate": 9.407649922964708e-06,
+      "loss": 0.6918,
+      "step": 603
+    },
+    {
+      "epoch": 0.7246550689862028,
+      "grad_norm": 0.7044757854005275,
+      "learning_rate": 9.404348082655732e-06,
+      "loss": 0.7604,
+      "step": 604
+    },
+    {
+      "epoch": 0.7258548290341932,
+      "grad_norm": 0.7140315477847774,
+      "learning_rate": 9.40103764820726e-06,
+      "loss": 0.7346,
+      "step": 605
+    },
+    {
+      "epoch": 0.7270545890821836,
+      "grad_norm": 0.7921784179018161,
+      "learning_rate": 9.3977186260789e-06,
+      "loss": 0.6944,
+      "step": 606
+    },
+    {
+      "epoch": 0.728254349130174,
+      "grad_norm": 0.6872232338475704,
+      "learning_rate": 9.394391022747005e-06,
+      "loss": 0.7004,
+      "step": 607
+    },
+    {
+      "epoch": 0.7294541091781643,
+      "grad_norm": 0.8873585465398308,
+      "learning_rate": 9.39105484470468e-06,
+      "loss": 0.7299,
+      "step": 608
+    },
+    {
+      "epoch": 0.7306538692261547,
+      "grad_norm": 0.7736862318160932,
+      "learning_rate": 9.387710098461765e-06,
+      "loss": 0.7765,
+      "step": 609
+    },
+    {
+      "epoch": 0.7318536292741452,
+      "grad_norm": 0.6927992969964133,
+      "learning_rate": 9.384356790544807e-06,
+      "loss": 0.7472,
+      "step": 610
+    },
+    {
+      "epoch": 0.7330533893221356,
+      "grad_norm": 0.8082225325172666,
+      "learning_rate": 9.380994927497068e-06,
+      "loss": 0.7758,
+      "step": 611
+    },
+    {
+      "epoch": 0.734253149370126,
+      "grad_norm": 0.8724820015048785,
+      "learning_rate": 9.377624515878504e-06,
+      "loss": 0.7804,
+      "step": 612
+    },
+    {
+      "epoch": 0.7354529094181164,
+      "grad_norm": 0.82067275868822,
+      "learning_rate": 9.37424556226575e-06,
+      "loss": 0.7215,
+      "step": 613
+    },
+    {
+      "epoch": 0.7366526694661067,
+      "grad_norm": 0.7802391975377598,
+      "learning_rate": 9.370858073252104e-06,
+      "loss": 0.7595,
+      "step": 614
+    },
+    {
+      "epoch": 0.7378524295140971,
+      "grad_norm": 0.695511573939589,
+      "learning_rate": 9.367462055447528e-06,
+      "loss": 0.7495,
+      "step": 615
+    },
+    {
+      "epoch": 0.7390521895620876,
+      "grad_norm": 0.7210971676399367,
+      "learning_rate": 9.36405751547862e-06,
+      "loss": 0.7504,
+      "step": 616
+    },
+    {
+      "epoch": 0.740251949610078,
+      "grad_norm": 0.7127870994797852,
+      "learning_rate": 9.360644459988606e-06,
+      "loss": 0.6989,
+      "step": 617
+    },
+    {
+      "epoch": 0.7414517096580684,
+      "grad_norm": 0.8252434951513874,
+      "learning_rate": 9.357222895637337e-06,
+      "loss": 0.7435,
+      "step": 618
+    },
+    {
+      "epoch": 0.7426514697060588,
+      "grad_norm": 0.6777237241417108,
+      "learning_rate": 9.353792829101255e-06,
+      "loss": 0.7181,
+      "step": 619
+    },
+    {
+      "epoch": 0.7438512297540492,
+      "grad_norm": 0.7246319587284746,
+      "learning_rate": 9.3503542670734e-06,
+      "loss": 0.7542,
+      "step": 620
+    },
+    {
+      "epoch": 0.7450509898020395,
+      "grad_norm": 0.7233632732400332,
+      "learning_rate": 9.34690721626339e-06,
+      "loss": 0.6969,
+      "step": 621
+    },
+    {
+      "epoch": 0.74625074985003,
+      "grad_norm": 0.6981036869918213,
+      "learning_rate": 9.343451683397402e-06,
+      "loss": 0.6936,
+      "step": 622
+    },
+    {
+      "epoch": 0.7474505098980204,
+      "grad_norm": 0.6881147839478524,
+      "learning_rate": 9.339987675218165e-06,
+      "loss": 0.7398,
+      "step": 623
+    },
+    {
+      "epoch": 0.7486502699460108,
+      "grad_norm": 0.8089969231321731,
+      "learning_rate": 9.33651519848495e-06,
+      "loss": 0.7676,
+      "step": 624
+    },
+    {
+      "epoch": 0.7498500299940012,
+      "grad_norm": 0.7172080435840271,
+      "learning_rate": 9.333034259973551e-06,
+      "loss": 0.7167,
+      "step": 625
+    },
+    {
+      "epoch": 0.7510497900419916,
+      "grad_norm": 0.7256057710440875,
+      "learning_rate": 9.329544866476265e-06,
+      "loss": 0.6851,
+      "step": 626
+    },
+    {
+      "epoch": 0.752249550089982,
+      "grad_norm": 0.8188587519438086,
+      "learning_rate": 9.326047024801903e-06,
+      "loss": 0.7775,
+      "step": 627
+    },
+    {
+      "epoch": 0.7534493101379725,
+      "grad_norm": 0.762491836568235,
+      "learning_rate": 9.322540741775745e-06,
+      "loss": 0.7412,
+      "step": 628
+    },
+    {
+      "epoch": 0.7546490701859628,
+      "grad_norm": 0.6861742416181088,
+      "learning_rate": 9.319026024239552e-06,
+      "loss": 0.7172,
+      "step": 629
+    },
+    {
+      "epoch": 0.7558488302339532,
+      "grad_norm": 0.7425371085335192,
+      "learning_rate": 9.315502879051541e-06,
+      "loss": 0.744,
+      "step": 630
+    },
+    {
+      "epoch": 0.7570485902819436,
+      "grad_norm": 0.860120319975177,
+      "learning_rate": 9.311971313086371e-06,
+      "loss": 0.6923,
+      "step": 631
+    },
+    {
+      "epoch": 0.758248350329934,
+      "grad_norm": 0.7346323032138261,
+      "learning_rate": 9.30843133323514e-06,
+      "loss": 0.7782,
+      "step": 632
+    },
+    {
+      "epoch": 0.7594481103779244,
+      "grad_norm": 0.7544285377877584,
+      "learning_rate": 9.304882946405352e-06,
+      "loss": 0.6255,
+      "step": 633
+    },
+    {
+      "epoch": 0.7606478704259149,
+      "grad_norm": 0.736655907630535,
+      "learning_rate": 9.301326159520925e-06,
+      "loss": 0.6643,
+      "step": 634
+    },
+    {
+      "epoch": 0.7618476304739052,
+      "grad_norm": 0.7427962509814233,
+      "learning_rate": 9.297760979522166e-06,
+      "loss": 0.791,
+      "step": 635
+    },
+    {
+      "epoch": 0.7630473905218956,
+      "grad_norm": 0.748355158635932,
+      "learning_rate": 9.294187413365755e-06,
+      "loss": 0.8099,
+      "step": 636
+    },
+    {
+      "epoch": 0.764247150569886,
+      "grad_norm": 0.6991205935285665,
+      "learning_rate": 9.290605468024742e-06,
+      "loss": 0.7769,
+      "step": 637
+    },
+    {
+      "epoch": 0.7654469106178764,
+      "grad_norm": 0.7202702597169961,
+      "learning_rate": 9.287015150488522e-06,
+      "loss": 0.707,
+      "step": 638
+    },
+    {
+      "epoch": 0.7666466706658668,
+      "grad_norm": 0.8358327107940074,
+      "learning_rate": 9.28341646776283e-06,
+      "loss": 0.7207,
+      "step": 639
+    },
+    {
+      "epoch": 0.7678464307138573,
+      "grad_norm": 0.7159463048010999,
+      "learning_rate": 9.279809426869723e-06,
+      "loss": 0.6979,
+      "step": 640
+    },
+    {
+      "epoch": 0.7690461907618477,
+      "grad_norm": 0.861983749589215,
+      "learning_rate": 9.276194034847565e-06,
+      "loss": 0.8149,
+      "step": 641
+    },
+    {
+      "epoch": 0.770245950809838,
+      "grad_norm": 0.7787392983292577,
+      "learning_rate": 9.272570298751019e-06,
+      "loss": 0.6745,
+      "step": 642
+    },
+    {
+      "epoch": 0.7714457108578284,
+      "grad_norm": 0.7104910752402336,
+      "learning_rate": 9.268938225651027e-06,
+      "loss": 0.6715,
+      "step": 643
+    },
+    {
+      "epoch": 0.7726454709058188,
+      "grad_norm": 0.8049250722530529,
+      "learning_rate": 9.265297822634798e-06,
+      "loss": 0.7453,
+      "step": 644
+    },
+    {
+      "epoch": 0.7738452309538092,
+      "grad_norm": 0.6857302660826016,
+      "learning_rate": 9.261649096805798e-06,
+      "loss": 0.7502,
+      "step": 645
+    },
+    {
+      "epoch": 0.7750449910017997,
+      "grad_norm": 0.6553885497808106,
+      "learning_rate": 9.257992055283735e-06,
+      "loss": 0.7036,
+      "step": 646
+    },
+    {
+      "epoch": 0.7762447510497901,
+      "grad_norm": 0.7577789705457922,
+      "learning_rate": 9.254326705204534e-06,
+      "loss": 0.8082,
+      "step": 647
+    },
+    {
+      "epoch": 0.7774445110977805,
+      "grad_norm": 0.7691264141555266,
+      "learning_rate": 9.250653053720345e-06,
+      "loss": 0.8038,
+      "step": 648
+    },
+    {
+      "epoch": 0.7786442711457708,
+      "grad_norm": 0.7499407632048067,
+      "learning_rate": 9.246971107999505e-06,
+      "loss": 0.8091,
+      "step": 649
+    },
+    {
+      "epoch": 0.7798440311937612,
+      "grad_norm": 0.7775805611228338,
+      "learning_rate": 9.243280875226544e-06,
+      "loss": 0.8139,
+      "step": 650
+    },
+    {
+      "epoch": 0.7810437912417516,
+      "grad_norm": 0.6563095758055505,
+      "learning_rate": 9.239582362602155e-06,
+      "loss": 0.6856,
+      "step": 651
+    },
+    {
+      "epoch": 0.7822435512897421,
+      "grad_norm": 0.9321681932072071,
+      "learning_rate": 9.235875577343195e-06,
+      "loss": 0.7461,
+      "step": 652
+    },
+    {
+      "epoch": 0.7834433113377325,
+      "grad_norm": 0.870840392576577,
+      "learning_rate": 9.232160526682659e-06,
+      "loss": 0.7332,
+      "step": 653
+    },
+    {
+      "epoch": 0.7846430713857229,
+      "grad_norm": 0.8479916292898527,
+      "learning_rate": 9.228437217869668e-06,
+      "loss": 0.7548,
+      "step": 654
+    },
+    {
+      "epoch": 0.7858428314337133,
+      "grad_norm": 0.7577898264784557,
+      "learning_rate": 9.22470565816946e-06,
+      "loss": 0.7112,
+      "step": 655
+    },
+    {
+      "epoch": 0.7870425914817036,
+      "grad_norm": 0.7300874136086064,
+      "learning_rate": 9.220965854863375e-06,
+      "loss": 0.7645,
+      "step": 656
+    },
+    {
+      "epoch": 0.788242351529694,
+      "grad_norm": 0.8105013430146889,
+      "learning_rate": 9.217217815248834e-06,
+      "loss": 0.7181,
+      "step": 657
+    },
+    {
+      "epoch": 0.7894421115776845,
+      "grad_norm": 0.6988905401190743,
+      "learning_rate": 9.213461546639334e-06,
+      "loss": 0.8008,
+      "step": 658
+    },
+    {
+      "epoch": 0.7906418716256749,
+      "grad_norm": 0.9367110211078484,
+      "learning_rate": 9.209697056364421e-06,
+      "loss": 0.6649,
+      "step": 659
+    },
+    {
+      "epoch": 0.7918416316736653,
+      "grad_norm": 0.8415790618546976,
+      "learning_rate": 9.205924351769695e-06,
+      "loss": 0.6454,
+      "step": 660
+    },
+    {
+      "epoch": 0.7930413917216557,
+      "grad_norm": 0.779708965565287,
+      "learning_rate": 9.202143440216776e-06,
+      "loss": 0.7348,
+      "step": 661
+    },
+    {
+      "epoch": 0.794241151769646,
+      "grad_norm": 0.7584769779508698,
+      "learning_rate": 9.198354329083304e-06,
+      "loss": 0.7771,
+      "step": 662
+    },
+    {
+      "epoch": 0.7954409118176364,
+      "grad_norm": 0.752690256345137,
+      "learning_rate": 9.194557025762911e-06,
+      "loss": 0.6826,
+      "step": 663
+    },
+    {
+      "epoch": 0.7966406718656269,
+      "grad_norm": 0.6987843714144564,
+      "learning_rate": 9.190751537665222e-06,
+      "loss": 0.7299,
+      "step": 664
+    },
+    {
+      "epoch": 0.7978404319136173,
+      "grad_norm": 0.7684852943780471,
+      "learning_rate": 9.186937872215831e-06,
+      "loss": 0.6788,
+      "step": 665
+    },
+    {
+      "epoch": 0.7990401919616077,
+      "grad_norm": 0.7291940251139142,
+      "learning_rate": 9.183116036856283e-06,
+      "loss": 0.7404,
+      "step": 666
+    },
+    {
+      "epoch": 0.8002399520095981,
+      "grad_norm": 0.7059763113053509,
+      "learning_rate": 9.179286039044072e-06,
+      "loss": 0.7853,
+      "step": 667
+    },
+    {
+      "epoch": 0.8014397120575885,
+      "grad_norm": 0.8840293556447637,
+      "learning_rate": 9.175447886252618e-06,
+      "loss": 0.7489,
+      "step": 668
+    },
+    {
+      "epoch": 0.8026394721055788,
+      "grad_norm": 0.8385407380699331,
+      "learning_rate": 9.171601585971248e-06,
+      "loss": 0.75,
+      "step": 669
+    },
+    {
+      "epoch": 0.8038392321535693,
+      "grad_norm": 0.7991173151829092,
+      "learning_rate": 9.167747145705195e-06,
+      "loss": 0.6904,
+      "step": 670
+    },
+    {
+      "epoch": 0.8050389922015597,
+      "grad_norm": 0.6799107722840952,
+      "learning_rate": 9.163884572975566e-06,
+      "loss": 0.7309,
+      "step": 671
+    },
+    {
+      "epoch": 0.8062387522495501,
+      "grad_norm": 0.8564949850489171,
+      "learning_rate": 9.160013875319348e-06,
+      "loss": 0.6968,
+      "step": 672
+    },
+    {
+      "epoch": 0.8074385122975405,
+      "grad_norm": 0.797552540085641,
+      "learning_rate": 9.156135060289374e-06,
+      "loss": 0.6622,
+      "step": 673
+    },
+    {
+      "epoch": 0.8086382723455309,
+      "grad_norm": 0.7142660355705535,
+      "learning_rate": 9.152248135454315e-06,
+      "loss": 0.7295,
+      "step": 674
+    },
+    {
+      "epoch": 0.8098380323935213,
+      "grad_norm": 0.7496127091364704,
+      "learning_rate": 9.148353108398677e-06,
+      "loss": 0.7389,
+      "step": 675
+    },
+    {
+      "epoch": 0.8110377924415118,
+      "grad_norm": 0.8373650209653336,
+      "learning_rate": 9.144449986722764e-06,
+      "loss": 0.6162,
+      "step": 676
+    },
+    {
+      "epoch": 0.8122375524895021,
+      "grad_norm": 0.7508306658345354,
+      "learning_rate": 9.140538778042682e-06,
+      "loss": 0.814,
+      "step": 677
+    },
+    {
+      "epoch": 0.8134373125374925,
+      "grad_norm": 0.7764431964930828,
+      "learning_rate": 9.136619489990313e-06,
+      "loss": 0.6989,
+      "step": 678
+    },
+    {
+      "epoch": 0.8146370725854829,
+      "grad_norm": 0.7645974259736653,
+      "learning_rate": 9.132692130213307e-06,
+      "loss": 0.7694,
+      "step": 679
+    },
+    {
+      "epoch": 0.8158368326334733,
+      "grad_norm": 0.7556080783634802,
+      "learning_rate": 9.128756706375065e-06,
+      "loss": 0.7806,
+      "step": 680
+    },
+    {
+      "epoch": 0.8170365926814637,
+      "grad_norm": 0.7970652457497799,
+      "learning_rate": 9.124813226154719e-06,
+      "loss": 0.748,
+      "step": 681
+    },
+    {
+      "epoch": 0.8182363527294542,
+      "grad_norm": 1.0254876320720594,
+      "learning_rate": 9.120861697247125e-06,
+      "loss": 0.7371,
+      "step": 682
+    },
+    {
+      "epoch": 0.8194361127774445,
+      "grad_norm": 0.7158919072080201,
+      "learning_rate": 9.116902127362841e-06,
+      "loss": 0.7136,
+      "step": 683
+    },
+    {
+      "epoch": 0.8206358728254349,
+      "grad_norm": 0.8533405560260696,
+      "learning_rate": 9.11293452422812e-06,
+      "loss": 0.7545,
+      "step": 684
+    },
+    {
+      "epoch": 0.8218356328734253,
+      "grad_norm": 0.8259141057451911,
+      "learning_rate": 9.108958895584888e-06,
+      "loss": 0.7221,
+      "step": 685
+    },
+    {
+      "epoch": 0.8230353929214157,
+      "grad_norm": 0.8168844953059569,
+      "learning_rate": 9.104975249190728e-06,
+      "loss": 0.7437,
+      "step": 686
+    },
+    {
+      "epoch": 0.8242351529694061,
+      "grad_norm": 0.7572408428890092,
+      "learning_rate": 9.100983592818873e-06,
+      "loss": 0.7147,
+      "step": 687
+    },
+    {
+      "epoch": 0.8254349130173965,
+      "grad_norm": 0.7417806524544673,
+      "learning_rate": 9.09698393425818e-06,
+      "loss": 0.7825,
+      "step": 688
+    },
+    {
+      "epoch": 0.826634673065387,
+      "grad_norm": 0.9109315486736959,
+      "learning_rate": 9.092976281313127e-06,
+      "loss": 0.6836,
+      "step": 689
+    },
+    {
+      "epoch": 0.8278344331133773,
+      "grad_norm": 0.7700193136501021,
+      "learning_rate": 9.088960641803786e-06,
+      "loss": 0.7101,
+      "step": 690
+    },
+    {
+      "epoch": 0.8290341931613677,
+      "grad_norm": 0.7063368254506233,
+      "learning_rate": 9.084937023565817e-06,
+      "loss": 0.7607,
+      "step": 691
+    },
+    {
+      "epoch": 0.8302339532093581,
+      "grad_norm": 0.7335228407782676,
+      "learning_rate": 9.080905434450446e-06,
+      "loss": 0.793,
+      "step": 692
+    },
+    {
+      "epoch": 0.8314337132573485,
+      "grad_norm": 0.757472933993561,
+      "learning_rate": 9.076865882324453e-06,
+      "loss": 0.7252,
+      "step": 693
+    },
+    {
+      "epoch": 0.8326334733053389,
+      "grad_norm": 0.7775955900228193,
+      "learning_rate": 9.072818375070156e-06,
+      "loss": 0.7111,
+      "step": 694
+    },
+    {
+      "epoch": 0.8338332333533294,
+      "grad_norm": 0.7222525868463516,
+      "learning_rate": 9.068762920585399e-06,
+      "loss": 0.6625,
+      "step": 695
+    },
+    {
+      "epoch": 0.8350329934013198,
+      "grad_norm": 0.7291752779428845,
+      "learning_rate": 9.064699526783527e-06,
+      "loss": 0.7153,
+      "step": 696
+    },
+    {
+      "epoch": 0.8362327534493101,
+      "grad_norm": 0.7526637151804322,
+      "learning_rate": 9.060628201593383e-06,
+      "loss": 0.7601,
+      "step": 697
+    },
+    {
+      "epoch": 0.8374325134973005,
+      "grad_norm": 0.8495898517083398,
+      "learning_rate": 9.056548952959283e-06,
+      "loss": 0.663,
+      "step": 698
+    },
+    {
+      "epoch": 0.8386322735452909,
+      "grad_norm": 0.7919350740237497,
+      "learning_rate": 9.052461788841005e-06,
+      "loss": 0.7653,
+      "step": 699
+    },
+    {
+      "epoch": 0.8398320335932813,
+      "grad_norm": 0.6139199206157331,
+      "learning_rate": 9.048366717213772e-06,
+      "loss": 0.6735,
+      "step": 700
+    },
+    {
+      "epoch": 0.8410317936412718,
+      "grad_norm": 0.6954314325669156,
+      "learning_rate": 9.044263746068236e-06,
+      "loss": 0.7085,
+      "step": 701
+    },
+    {
+      "epoch": 0.8422315536892622,
+      "grad_norm": 0.7703628098346507,
+      "learning_rate": 9.040152883410465e-06,
+      "loss": 0.7534,
+      "step": 702
+    },
+    {
+      "epoch": 0.8434313137372526,
+      "grad_norm": 0.7962857304460215,
+      "learning_rate": 9.036034137261924e-06,
+      "loss": 0.6811,
+      "step": 703
+    },
+    {
+      "epoch": 0.8446310737852429,
+      "grad_norm": 0.6355076254938227,
+      "learning_rate": 9.031907515659464e-06,
+      "loss": 0.7239,
+      "step": 704
+    },
+    {
+      "epoch": 0.8458308338332333,
+      "grad_norm": 0.6583975844454037,
+      "learning_rate": 9.027773026655298e-06,
+      "loss": 0.716,
+      "step": 705
+    },
+    {
+      "epoch": 0.8470305938812237,
+      "grad_norm": 0.7138960727557581,
+      "learning_rate": 9.023630678316994e-06,
+      "loss": 0.7629,
+      "step": 706
+    },
+    {
+      "epoch": 0.8482303539292142,
+      "grad_norm": 0.759517500163122,
+      "learning_rate": 9.019480478727458e-06,
+      "loss": 0.712,
+      "step": 707
+    },
+    {
+      "epoch": 0.8494301139772046,
+      "grad_norm": 0.7617781662555174,
+      "learning_rate": 9.015322435984909e-06,
+      "loss": 0.6963,
+      "step": 708
+    },
+    {
+      "epoch": 0.850629874025195,
+      "grad_norm": 0.8017913620380468,
+      "learning_rate": 9.011156558202877e-06,
+      "loss": 0.75,
+      "step": 709
+    },
+    {
+      "epoch": 0.8518296340731853,
+      "grad_norm": 0.6428736050904517,
+      "learning_rate": 9.006982853510177e-06,
+      "loss": 0.7252,
+      "step": 710
+    },
+    {
+      "epoch": 0.8530293941211757,
+      "grad_norm": 0.6897099281544877,
+      "learning_rate": 9.0028013300509e-06,
+      "loss": 0.7084,
+      "step": 711
+    },
+    {
+      "epoch": 0.8542291541691661,
+      "grad_norm": 0.8692697405988467,
+      "learning_rate": 8.998611995984388e-06,
+      "loss": 0.7003,
+      "step": 712
+    },
+    {
+      "epoch": 0.8554289142171566,
+      "grad_norm": 0.6913961519935883,
+      "learning_rate": 8.994414859485228e-06,
+      "loss": 0.7201,
+      "step": 713
+    },
+    {
+      "epoch": 0.856628674265147,
+      "grad_norm": 0.8601053884938681,
+      "learning_rate": 8.99020992874323e-06,
+      "loss": 0.7144,
+      "step": 714
+    },
+    {
+      "epoch": 0.8578284343131374,
+      "grad_norm": 0.6869846626792715,
+      "learning_rate": 8.985997211963414e-06,
+      "loss": 0.6746,
+      "step": 715
+    },
+    {
+      "epoch": 0.8590281943611278,
+      "grad_norm": 0.8118651300115735,
+      "learning_rate": 8.981776717365991e-06,
+      "loss": 0.6991,
+      "step": 716
+    },
+    {
+      "epoch": 0.8602279544091181,
+      "grad_norm": 0.8515786561839657,
+      "learning_rate": 8.977548453186354e-06,
+      "loss": 0.6962,
+      "step": 717
+    },
+    {
+      "epoch": 0.8614277144571085,
+      "grad_norm": 0.6595328779321165,
+      "learning_rate": 8.973312427675047e-06,
+      "loss": 0.7054,
+      "step": 718
+    },
+    {
+      "epoch": 0.862627474505099,
+      "grad_norm": 0.7348146209601154,
+      "learning_rate": 8.969068649097766e-06,
+      "loss": 0.7633,
+      "step": 719
+    },
+    {
+      "epoch": 0.8638272345530894,
+      "grad_norm": 0.8617055777943434,
+      "learning_rate": 8.964817125735337e-06,
+      "loss": 0.776,
+      "step": 720
+    },
+    {
+      "epoch": 0.8650269946010798,
+      "grad_norm": 0.8590136330003421,
+      "learning_rate": 8.96055786588369e-06,
+      "loss": 0.7528,
+      "step": 721
+    },
+    {
+      "epoch": 0.8662267546490702,
+      "grad_norm": 0.7281656999282983,
+      "learning_rate": 8.956290877853857e-06,
+      "loss": 0.657,
+      "step": 722
+    },
+    {
+      "epoch": 0.8674265146970606,
+      "grad_norm": 0.7673981705874487,
+      "learning_rate": 8.95201616997195e-06,
+      "loss": 0.7062,
+      "step": 723
+    },
+    {
+      "epoch": 0.8686262747450509,
+      "grad_norm": 0.8162483335235116,
+      "learning_rate": 8.947733750579147e-06,
+      "loss": 0.7224,
+      "step": 724
+    },
+    {
+      "epoch": 0.8698260347930414,
+      "grad_norm": 0.7829593234370885,
+      "learning_rate": 8.943443628031663e-06,
+      "loss": 0.7262,
+      "step": 725
+    },
+    {
+      "epoch": 0.8710257948410318,
+      "grad_norm": 0.8091531239120493,
+      "learning_rate": 8.939145810700755e-06,
+      "loss": 0.6739,
+      "step": 726
+    },
+    {
+      "epoch": 0.8722255548890222,
+      "grad_norm": 0.7558985566945485,
+      "learning_rate": 8.93484030697269e-06,
+      "loss": 0.7302,
+      "step": 727
+    },
+    {
+      "epoch": 0.8734253149370126,
+      "grad_norm": 0.836085347122929,
+      "learning_rate": 8.930527125248733e-06,
+      "loss": 0.7361,
+      "step": 728
+    },
+    {
+      "epoch": 0.874625074985003,
+      "grad_norm": 0.7320634563463685,
+      "learning_rate": 8.926206273945133e-06,
+      "loss": 0.7481,
+      "step": 729
+    },
+    {
+      "epoch": 0.8758248350329934,
+      "grad_norm": 0.7362413271525446,
+      "learning_rate": 8.921877761493103e-06,
+      "loss": 0.6921,
+      "step": 730
+    },
+    {
+      "epoch": 0.8770245950809838,
+      "grad_norm": 0.7337908243444188,
+      "learning_rate": 8.917541596338806e-06,
+      "loss": 0.7598,
+      "step": 731
+    },
+    {
+      "epoch": 0.8782243551289742,
+      "grad_norm": 0.7532399005677989,
+      "learning_rate": 8.913197786943335e-06,
+      "loss": 0.671,
+      "step": 732
+    },
+    {
+      "epoch": 0.8794241151769646,
+      "grad_norm": 0.720119023479152,
+      "learning_rate": 8.908846341782705e-06,
+      "loss": 0.6855,
+      "step": 733
+    },
+    {
+      "epoch": 0.880623875224955,
+      "grad_norm": 0.8340213195042716,
+      "learning_rate": 8.904487269347824e-06,
+      "loss": 0.8062,
+      "step": 734
+    },
+    {
+      "epoch": 0.8818236352729454,
+      "grad_norm": 0.8537544717020235,
+      "learning_rate": 8.900120578144487e-06,
+      "loss": 0.7572,
+      "step": 735
+    },
+    {
+      "epoch": 0.8830233953209358,
+      "grad_norm": 0.7711195264883798,
+      "learning_rate": 8.895746276693354e-06,
+      "loss": 0.6446,
+      "step": 736
+    },
+    {
+      "epoch": 0.8842231553689263,
+      "grad_norm": 0.6887520424872645,
+      "learning_rate": 8.891364373529933e-06,
+      "loss": 0.7289,
+      "step": 737
+    },
+    {
+      "epoch": 0.8854229154169166,
+      "grad_norm": 0.8719145209704985,
+      "learning_rate": 8.886974877204571e-06,
+      "loss": 0.7299,
+      "step": 738
+    },
+    {
+      "epoch": 0.886622675464907,
+      "grad_norm": 0.6511567150951192,
+      "learning_rate": 8.882577796282423e-06,
+      "loss": 0.775,
+      "step": 739
+    },
+    {
+      "epoch": 0.8878224355128974,
+      "grad_norm": 0.6579623722688875,
+      "learning_rate": 8.87817313934345e-06,
+      "loss": 0.6634,
+      "step": 740
+    },
+    {
+      "epoch": 0.8890221955608878,
+      "grad_norm": 0.7008815480410427,
+      "learning_rate": 8.873760914982398e-06,
+      "loss": 0.6636,
+      "step": 741
+    },
+    {
+      "epoch": 0.8902219556088782,
+      "grad_norm": 1.4279455662876566,
+      "learning_rate": 8.86934113180877e-06,
+      "loss": 0.7955,
+      "step": 742
+    },
+    {
+      "epoch": 0.8914217156568687,
+      "grad_norm": 0.7713950135088316,
+      "learning_rate": 8.864913798446825e-06,
+      "loss": 0.7636,
+      "step": 743
+    },
+    {
+      "epoch": 0.8926214757048591,
+      "grad_norm": 0.8260504764129432,
+      "learning_rate": 8.860478923535556e-06,
+      "loss": 0.6612,
+      "step": 744
+    },
+    {
+      "epoch": 0.8938212357528494,
+      "grad_norm": 0.7201716538315065,
+      "learning_rate": 8.856036515728666e-06,
+      "loss": 0.6913,
+      "step": 745
+    },
+    {
+      "epoch": 0.8950209958008398,
+      "grad_norm": 0.7214346438900352,
+      "learning_rate": 8.85158658369456e-06,
+      "loss": 0.768,
+      "step": 746
+    },
+    {
+      "epoch": 0.8962207558488302,
+      "grad_norm": 0.7259238301199039,
+      "learning_rate": 8.847129136116326e-06,
+      "loss": 0.6861,
+      "step": 747
+    },
+    {
+      "epoch": 0.8974205158968206,
+      "grad_norm": 1.5664870972162603,
+      "learning_rate": 8.842664181691716e-06,
+      "loss": 0.7551,
+      "step": 748
+    },
+    {
+      "epoch": 0.8986202759448111,
+      "grad_norm": 0.7374710992889468,
+      "learning_rate": 8.83819172913313e-06,
+      "loss": 0.6725,
+      "step": 749
+    },
+    {
+      "epoch": 0.8998200359928015,
+      "grad_norm": 0.6867731164372635,
+      "learning_rate": 8.833711787167596e-06,
+      "loss": 0.7708,
+      "step": 750
+    },
+    {
+      "epoch": 0.9010197960407919,
+      "grad_norm": 0.7433539004061774,
+      "learning_rate": 8.829224364536762e-06,
+      "loss": 0.66,
+      "step": 751
+    },
+    {
+      "epoch": 0.9022195560887822,
+      "grad_norm": 0.6976954630092328,
+      "learning_rate": 8.82472946999687e-06,
+      "loss": 0.7192,
+      "step": 752
+    },
+    {
+      "epoch": 0.9034193161367726,
+      "grad_norm": 0.7243742893502679,
+      "learning_rate": 8.820227112318737e-06,
+      "loss": 0.6881,
+      "step": 753
+    },
+    {
+      "epoch": 0.904619076184763,
+      "grad_norm": 0.6601153532474624,
+      "learning_rate": 8.815717300287751e-06,
+      "loss": 0.681,
+      "step": 754
+    },
+    {
+      "epoch": 0.9058188362327535,
+      "grad_norm": 0.7442440370029324,
+      "learning_rate": 8.81120004270384e-06,
+      "loss": 0.7264,
+      "step": 755
+    },
+    {
+      "epoch": 0.9070185962807439,
+      "grad_norm": 0.7407327729335063,
+      "learning_rate": 8.806675348381465e-06,
+      "loss": 0.6697,
+      "step": 756
+    },
+    {
+      "epoch": 0.9082183563287343,
+      "grad_norm": 0.6749557172369749,
+      "learning_rate": 8.80214322614959e-06,
+      "loss": 0.7387,
+      "step": 757
+    },
+    {
+      "epoch": 0.9094181163767247,
+      "grad_norm": 0.7888777311895881,
+      "learning_rate": 8.797603684851685e-06,
+      "loss": 0.7901,
+      "step": 758
+    },
+    {
+      "epoch": 0.910617876424715,
+      "grad_norm": 0.8381921280616869,
+      "learning_rate": 8.793056733345685e-06,
+      "loss": 0.7785,
+      "step": 759
+    },
+    {
+      "epoch": 0.9118176364727054,
+      "grad_norm": 0.9634129064332837,
+      "learning_rate": 8.788502380503991e-06,
+      "loss": 0.6772,
+      "step": 760
+    },
+    {
+      "epoch": 0.9130173965206959,
+      "grad_norm": 0.6957936617015104,
+      "learning_rate": 8.783940635213443e-06,
+      "loss": 0.7519,
+      "step": 761
+    },
+    {
+      "epoch": 0.9142171565686863,
+      "grad_norm": 0.8261484130782292,
+      "learning_rate": 8.779371506375312e-06,
+      "loss": 0.7726,
+      "step": 762
+    },
+    {
+      "epoch": 0.9154169166166767,
+      "grad_norm": 0.8443144653539115,
+      "learning_rate": 8.774795002905266e-06,
+      "loss": 0.7667,
+      "step": 763
+    },
+    {
+      "epoch": 0.9166166766646671,
+      "grad_norm": 0.6997642773177474,
+      "learning_rate": 8.770211133733373e-06,
+      "loss": 0.6865,
+      "step": 764
+    },
+    {
+      "epoch": 0.9178164367126574,
+      "grad_norm": 0.9107408755390968,
+      "learning_rate": 8.765619907804068e-06,
+      "loss": 0.7522,
+      "step": 765
+    },
+    {
+      "epoch": 0.9190161967606478,
+      "grad_norm": 0.7972813404134526,
+      "learning_rate": 8.76102133407614e-06,
+      "loss": 0.7496,
+      "step": 766
+    },
+    {
+      "epoch": 0.9202159568086383,
+      "grad_norm": 0.8575564826675797,
+      "learning_rate": 8.756415421522723e-06,
+      "loss": 0.7115,
+      "step": 767
+    },
+    {
+      "epoch": 0.9214157168566287,
+      "grad_norm": 0.784169782895324,
+      "learning_rate": 8.751802179131261e-06,
+      "loss": 0.7411,
+      "step": 768
+    },
+    {
+      "epoch": 0.9226154769046191,
+      "grad_norm": 0.8257712462719169,
+      "learning_rate": 8.74718161590351e-06,
+      "loss": 0.722,
+      "step": 769
+    },
+    {
+      "epoch": 0.9238152369526095,
+      "grad_norm": 1.1175941018251747,
+      "learning_rate": 8.742553740855507e-06,
+      "loss": 0.8218,
+      "step": 770
+    },
+    {
+      "epoch": 0.9250149970005999,
+      "grad_norm": 0.9189804422676542,
+      "learning_rate": 8.737918563017553e-06,
+      "loss": 0.6823,
+      "step": 771
+    },
+    {
+      "epoch": 0.9262147570485902,
+      "grad_norm": 0.7037201795139563,
+      "learning_rate": 8.733276091434204e-06,
+      "loss": 0.7116,
+      "step": 772
+    },
+    {
+      "epoch": 0.9274145170965807,
+      "grad_norm": 0.7875872081694147,
+      "learning_rate": 8.728626335164247e-06,
+      "loss": 0.7094,
+      "step": 773
+    },
+    {
+      "epoch": 0.9286142771445711,
+      "grad_norm": 0.6815464687911802,
+      "learning_rate": 8.723969303280682e-06,
+      "loss": 0.7418,
+      "step": 774
+    },
+    {
+      "epoch": 0.9298140371925615,
+      "grad_norm": 0.7265610401869881,
+      "learning_rate": 8.719305004870705e-06,
+      "loss": 0.7806,
+      "step": 775
+    },
+    {
+      "epoch": 0.9310137972405519,
+      "grad_norm": 0.717266987553133,
+      "learning_rate": 8.714633449035699e-06,
+      "loss": 0.7155,
+      "step": 776
+    },
+    {
+      "epoch": 0.9322135572885423,
+      "grad_norm": 0.7642623821787721,
+      "learning_rate": 8.709954644891195e-06,
+      "loss": 0.7438,
+      "step": 777
+    },
+    {
+      "epoch": 0.9334133173365327,
+      "grad_norm": 0.8504999037092563,
+      "learning_rate": 8.705268601566877e-06,
+      "loss": 0.7389,
+      "step": 778
+    },
+    {
+      "epoch": 0.9346130773845231,
+      "grad_norm": 0.8259898115474855,
+      "learning_rate": 8.700575328206554e-06,
+      "loss": 0.7779,
+      "step": 779
+    },
+    {
+      "epoch": 0.9358128374325135,
+      "grad_norm": 0.6695114476759497,
+      "learning_rate": 8.695874833968136e-06,
+      "loss": 0.7475,
+      "step": 780
+    },
+    {
+      "epoch": 0.9370125974805039,
+      "grad_norm": 0.7666424504704847,
+      "learning_rate": 8.691167128023637e-06,
+      "loss": 0.6744,
+      "step": 781
+    },
+    {
+      "epoch": 0.9382123575284943,
+      "grad_norm": 0.6682439375073063,
+      "learning_rate": 8.686452219559125e-06,
+      "loss": 0.651,
+      "step": 782
+    },
+    {
+      "epoch": 0.9394121175764847,
+      "grad_norm": 0.751684373428545,
+      "learning_rate": 8.681730117774736e-06,
+      "loss": 0.7075,
+      "step": 783
+    },
+    {
+      "epoch": 0.9406118776244751,
+      "grad_norm": 0.7024570630082877,
+      "learning_rate": 8.677000831884639e-06,
+      "loss": 0.6648,
+      "step": 784
+    },
+    {
+      "epoch": 0.9418116376724655,
+      "grad_norm": 0.7583152269181472,
+      "learning_rate": 8.672264371117016e-06,
+      "loss": 0.751,
+      "step": 785
+    },
+    {
+      "epoch": 0.943011397720456,
+      "grad_norm": 0.7316407315135922,
+      "learning_rate": 8.667520744714055e-06,
+      "loss": 0.7648,
+      "step": 786
+    },
+    {
+      "epoch": 0.9442111577684463,
+      "grad_norm": 0.8269882990695296,
+      "learning_rate": 8.662769961931926e-06,
+      "loss": 0.6869,
+      "step": 787
+    },
+    {
+      "epoch": 0.9454109178164367,
+      "grad_norm": 0.9002478494140085,
+      "learning_rate": 8.658012032040758e-06,
+      "loss": 0.7247,
+      "step": 788
+    },
+    {
+      "epoch": 0.9466106778644271,
+      "grad_norm": 0.7721188467093039,
+      "learning_rate": 8.653246964324631e-06,
+      "loss": 0.6905,
+      "step": 789
+    },
+    {
+      "epoch": 0.9478104379124175,
+      "grad_norm": 0.7001233970231492,
+      "learning_rate": 8.648474768081553e-06,
+      "loss": 0.7291,
+      "step": 790
+    },
+    {
+      "epoch": 0.9490101979604079,
+      "grad_norm": 0.6839392449808562,
+      "learning_rate": 8.643695452623437e-06,
+      "loss": 0.6951,
+      "step": 791
+    },
+    {
+      "epoch": 0.9502099580083984,
+      "grad_norm": 0.7310041234694719,
+      "learning_rate": 8.638909027276093e-06,
+      "loss": 0.6974,
+      "step": 792
+    },
+    {
+      "epoch": 0.9514097180563887,
+      "grad_norm": 0.7067703460134739,
+      "learning_rate": 8.634115501379202e-06,
+      "loss": 0.6696,
+      "step": 793
+    },
+    {
+      "epoch": 0.9526094781043791,
+      "grad_norm": 0.7156477215190924,
+      "learning_rate": 8.629314884286299e-06,
+      "loss": 0.7316,
+      "step": 794
+    },
+    {
+      "epoch": 0.9538092381523695,
+      "grad_norm": 0.743906326807422,
+      "learning_rate": 8.62450718536476e-06,
+      "loss": 0.7036,
+      "step": 795
+    },
+    {
+      "epoch": 0.9550089982003599,
+      "grad_norm": 0.8704938925080471,
+      "learning_rate": 8.619692413995775e-06,
+      "loss": 0.6215,
+      "step": 796
+    },
+    {
+      "epoch": 0.9562087582483503,
+      "grad_norm": 0.7599568654719119,
+      "learning_rate": 8.614870579574338e-06,
+      "loss": 0.6672,
+      "step": 797
+    },
+    {
+      "epoch": 0.9574085182963408,
+      "grad_norm": 0.8316313659598846,
+      "learning_rate": 8.61004169150922e-06,
+      "loss": 0.754,
+      "step": 798
+    },
+    {
+      "epoch": 0.9586082783443312,
+      "grad_norm": 0.7315128118767747,
+      "learning_rate": 8.605205759222963e-06,
+      "loss": 0.7195,
+      "step": 799
+    },
+    {
+      "epoch": 0.9598080383923215,
+      "grad_norm": 0.7171949806066087,
+      "learning_rate": 8.600362792151849e-06,
+      "loss": 0.6753,
+      "step": 800
+    },
+    {
+      "epoch": 0.9610077984403119,
+      "grad_norm": 0.7630107755408194,
+      "learning_rate": 8.595512799745887e-06,
+      "loss": 0.7561,
+      "step": 801
+    },
+    {
+      "epoch": 0.9622075584883023,
+      "grad_norm": 0.6960112659463641,
+      "learning_rate": 8.590655791468798e-06,
+      "loss": 0.7178,
+      "step": 802
+    },
+    {
+      "epoch": 0.9634073185362927,
+      "grad_norm": 0.7785762158178057,
+      "learning_rate": 8.585791776797989e-06,
+      "loss": 0.775,
+      "step": 803
+    },
+    {
+      "epoch": 0.9646070785842832,
+      "grad_norm": 0.7625428223240929,
+      "learning_rate": 8.580920765224541e-06,
+      "loss": 0.6772,
+      "step": 804
+    },
+    {
+      "epoch": 0.9658068386322736,
+      "grad_norm": 0.6918983117535726,
+      "learning_rate": 8.576042766253186e-06,
+      "loss": 0.7432,
+      "step": 805
+    },
+    {
+      "epoch": 0.967006598680264,
+      "grad_norm": 0.6725832160053664,
+      "learning_rate": 8.571157789402292e-06,
+      "loss": 0.7227,
+      "step": 806
+    },
+    {
+      "epoch": 0.9682063587282543,
+      "grad_norm": 0.8274038768697642,
+      "learning_rate": 8.566265844203843e-06,
+      "loss": 0.6806,
+      "step": 807
+    },
+    {
+      "epoch": 0.9694061187762447,
+      "grad_norm": 0.8338463853279005,
+      "learning_rate": 8.56136694020342e-06,
+      "loss": 0.7171,
+      "step": 808
+    },
+    {
+      "epoch": 0.9706058788242351,
+      "grad_norm": 0.7590280630412745,
+      "learning_rate": 8.556461086960179e-06,
+      "loss": 0.6099,
+      "step": 809
+    },
+    {
+      "epoch": 0.9718056388722256,
+      "grad_norm": 0.6679038582059358,
+      "learning_rate": 8.551548294046843e-06,
+      "loss": 0.6973,
+      "step": 810
+    },
+    {
+      "epoch": 0.973005398920216,
+      "grad_norm": 0.909079297394524,
+      "learning_rate": 8.546628571049671e-06,
+      "loss": 0.7315,
+      "step": 811
+    },
+    {
+      "epoch": 0.9742051589682064,
+      "grad_norm": 0.9645613467378517,
+      "learning_rate": 8.541701927568444e-06,
+      "loss": 0.7391,
+      "step": 812
+    },
+    {
+      "epoch": 0.9754049190161967,
+      "grad_norm": 0.7697405806994291,
+      "learning_rate": 8.536768373216454e-06,
+      "loss": 0.6878,
+      "step": 813
+    },
+    {
+      "epoch": 0.9766046790641871,
+      "grad_norm": 0.7701896004261258,
+      "learning_rate": 8.531827917620466e-06,
+      "loss": 0.6798,
+      "step": 814
+    },
+    {
+      "epoch": 0.9778044391121775,
+      "grad_norm": 0.7096781450038352,
+      "learning_rate": 8.526880570420721e-06,
+      "loss": 0.7076,
+      "step": 815
+    },
+    {
+      "epoch": 0.979004199160168,
+      "grad_norm": 0.6409247185038676,
+      "learning_rate": 8.521926341270908e-06,
+      "loss": 0.7069,
+      "step": 816
+    },
+    {
+      "epoch": 0.9802039592081584,
+      "grad_norm": 0.706560411878609,
+      "learning_rate": 8.516965239838137e-06,
+      "loss": 0.7162,
+      "step": 817
+    },
+    {
+      "epoch": 0.9814037192561488,
+      "grad_norm": 0.9947129450737373,
+      "learning_rate": 8.511997275802934e-06,
+      "loss": 0.7489,
+      "step": 818
+    },
+    {
+      "epoch": 0.9826034793041392,
+      "grad_norm": 0.8735057070309135,
+      "learning_rate": 8.507022458859215e-06,
+      "loss": 0.7346,
+      "step": 819
+    },
+    {
+      "epoch": 0.9838032393521295,
+      "grad_norm": 0.6508191410059432,
+      "learning_rate": 8.502040798714263e-06,
+      "loss": 0.6725,
+      "step": 820
+    },
+    {
+      "epoch": 0.9850029994001199,
+      "grad_norm": 0.7443473259703014,
+      "learning_rate": 8.497052305088722e-06,
+      "loss": 0.6889,
+      "step": 821
+    },
+    {
+      "epoch": 0.9862027594481104,
+      "grad_norm": 1.2455924883181906,
+      "learning_rate": 8.492056987716566e-06,
+      "loss": 0.6847,
+      "step": 822
+    },
+    {
+      "epoch": 0.9874025194961008,
+      "grad_norm": 0.7944880426727919,
+      "learning_rate": 8.487054856345081e-06,
+      "loss": 0.749,
+      "step": 823
+    },
+    {
+      "epoch": 0.9886022795440912,
+      "grad_norm": 0.7262988013969764,
+      "learning_rate": 8.482045920734855e-06,
+      "loss": 0.6876,
+      "step": 824
+    },
+    {
+      "epoch": 0.9898020395920816,
+      "grad_norm": 0.7812167032222277,
+      "learning_rate": 8.47703019065975e-06,
+      "loss": 0.6905,
+      "step": 825
+    },
+    {
+      "epoch": 0.991001799640072,
+      "grad_norm": 0.8887583534089065,
+      "learning_rate": 8.472007675906883e-06,
+      "loss": 0.701,
+      "step": 826
+    },
+    {
+      "epoch": 0.9922015596880623,
+      "grad_norm": 0.8180330273352874,
+      "learning_rate": 8.466978386276618e-06,
+      "loss": 0.6619,
+      "step": 827
+    },
+    {
+      "epoch": 0.9934013197360528,
+      "grad_norm": 0.7550253827882137,
+      "learning_rate": 8.46194233158253e-06,
+      "loss": 0.6879,
+      "step": 828
+    },
+    {
+      "epoch": 0.9946010797840432,
+      "grad_norm": 0.8094156521246557,
+      "learning_rate": 8.456899521651401e-06,
+      "loss": 0.7122,
+      "step": 829
+    },
+    {
+      "epoch": 0.9958008398320336,
+      "grad_norm": 0.7764238822386331,
+      "learning_rate": 8.451849966323189e-06,
+      "loss": 0.7269,
+      "step": 830
+    },
+    {
+      "epoch": 0.997000599880024,
+      "grad_norm": 0.6943839690282205,
+      "learning_rate": 8.446793675451017e-06,
+      "loss": 0.6991,
+      "step": 831
+    },
+    {
+      "epoch": 0.9982003599280144,
+      "grad_norm": 0.7890307349482584,
+      "learning_rate": 8.441730658901154e-06,
+      "loss": 0.722,
+      "step": 832
+    },
+    {
+      "epoch": 0.9994001199760048,
+      "grad_norm": 0.7177227139393145,
+      "learning_rate": 8.436660926552987e-06,
+      "loss": 0.6527,
+      "step": 833
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7177227139393145,
+      "learning_rate": 8.431584488299009e-06,
+      "loss": 0.6803,
+      "step": 834
+    },
+    {
+      "epoch": 1.0011997600479905,
+      "grad_norm": 1.2017300303291012,
+      "learning_rate": 8.426501354044802e-06,
+      "loss": 0.7494,
+      "step": 835
+    },
+    {
+      "epoch": 1.0023995200959808,
+      "grad_norm": 1.0195681094503213,
+      "learning_rate": 8.42141153370901e-06,
+      "loss": 0.6218,
+      "step": 836
+    },
+    {
+      "epoch": 1.0035992801439713,
+      "grad_norm": 1.2864757513523697,
+      "learning_rate": 8.416315037223322e-06,
+      "loss": 0.7156,
+      "step": 837
+    },
+    {
+      "epoch": 1.0047990401919615,
+      "grad_norm": 0.7178806399171327,
+      "learning_rate": 8.41121187453246e-06,
+      "loss": 0.6571,
+      "step": 838
+    },
+    {
+      "epoch": 1.005998800239952,
+      "grad_norm": 0.8446232852801754,
+      "learning_rate": 8.406102055594148e-06,
+      "loss": 0.6132,
+      "step": 839
+    },
+    {
+      "epoch": 1.0071985602879423,
+      "grad_norm": 0.6838103195827099,
+      "learning_rate": 8.400985590379102e-06,
+      "loss": 0.5841,
+      "step": 840
+    },
+    {
+      "epoch": 1.0083983203359328,
+      "grad_norm": 0.6677261351678965,
+      "learning_rate": 8.395862488871003e-06,
+      "loss": 0.6029,
+      "step": 841
+    },
+    {
+      "epoch": 1.0095980803839233,
+      "grad_norm": 0.812967264899385,
+      "learning_rate": 8.390732761066484e-06,
+      "loss": 0.6174,
+      "step": 842
+    },
+    {
+      "epoch": 1.0107978404319136,
+      "grad_norm": 0.8837166380270197,
+      "learning_rate": 8.38559641697511e-06,
+      "loss": 0.6194,
+      "step": 843
+    },
+    {
+      "epoch": 1.011997600479904,
+      "grad_norm": 0.5840251307941942,
+      "learning_rate": 8.38045346661935e-06,
+      "loss": 0.6831,
+      "step": 844
+    },
+    {
+      "epoch": 1.0131973605278943,
+      "grad_norm": 0.8540882477821887,
+      "learning_rate": 8.375303920034567e-06,
+      "loss": 0.6423,
+      "step": 845
+    },
+    {
+      "epoch": 1.0143971205758848,
+      "grad_norm": 0.8173963023070573,
+      "learning_rate": 8.370147787269001e-06,
+      "loss": 0.6051,
+      "step": 846
+    },
+    {
+      "epoch": 1.0155968806238753,
+      "grad_norm": 0.7778251780116223,
+      "learning_rate": 8.364985078383729e-06,
+      "loss": 0.5686,
+      "step": 847
+    },
+    {
+      "epoch": 1.0167966406718656,
+      "grad_norm": 0.7777169141031703,
+      "learning_rate": 8.359815803452677e-06,
+      "loss": 0.6723,
+      "step": 848
+    },
+    {
+      "epoch": 1.017996400719856,
+      "grad_norm": 0.7375802737519649,
+      "learning_rate": 8.35463997256257e-06,
+      "loss": 0.6736,
+      "step": 849
+    },
+    {
+      "epoch": 1.0191961607678464,
+      "grad_norm": 0.5812046566626188,
+      "learning_rate": 8.349457595812934e-06,
+      "loss": 0.5867,
+      "step": 850
+    },
+    {
+      "epoch": 1.0203959208158369,
+      "grad_norm": 0.7775044961284155,
+      "learning_rate": 8.34426868331606e-06,
+      "loss": 0.6513,
+      "step": 851
+    },
+    {
+      "epoch": 1.0215956808638271,
+      "grad_norm": 0.6633412981042944,
+      "learning_rate": 8.339073245197001e-06,
+      "loss": 0.6703,
+      "step": 852
+    },
+    {
+      "epoch": 1.0227954409118176,
+      "grad_norm": 0.8696629740675844,
+      "learning_rate": 8.333871291593534e-06,
+      "loss": 0.6518,
+      "step": 853
+    },
+    {
+      "epoch": 1.0239952009598081,
+      "grad_norm": 0.6240149906719957,
+      "learning_rate": 8.328662832656159e-06,
+      "loss": 0.5998,
+      "step": 854
+    },
+    {
+      "epoch": 1.0251949610077984,
+      "grad_norm": 0.8402359587949699,
+      "learning_rate": 8.323447878548063e-06,
+      "loss": 0.6399,
+      "step": 855
+    },
+    {
+      "epoch": 1.0263947210557889,
+      "grad_norm": 0.9284063049240554,
+      "learning_rate": 8.318226439445107e-06,
+      "loss": 0.5422,
+      "step": 856
+    },
+    {
+      "epoch": 1.0275944811037792,
+      "grad_norm": 0.6819000269369228,
+      "learning_rate": 8.312998525535813e-06,
+      "loss": 0.5781,
+      "step": 857
+    },
+    {
+      "epoch": 1.0287942411517697,
+      "grad_norm": 0.6889190567412302,
+      "learning_rate": 8.307764147021328e-06,
+      "loss": 0.6427,
+      "step": 858
+    },
+    {
+      "epoch": 1.0299940011997601,
+      "grad_norm": 0.8990639088926144,
+      "learning_rate": 8.302523314115421e-06,
+      "loss": 0.5896,
+      "step": 859
+    },
+    {
+      "epoch": 1.0311937612477504,
+      "grad_norm": 0.6367860307845866,
+      "learning_rate": 8.297276037044451e-06,
+      "loss": 0.6469,
+      "step": 860
+    },
+    {
+      "epoch": 1.032393521295741,
+      "grad_norm": 0.628397539390034,
+      "learning_rate": 8.29202232604735e-06,
+      "loss": 0.6532,
+      "step": 861
+    },
+    {
+      "epoch": 1.0335932813437312,
+      "grad_norm": 0.7168283069895726,
+      "learning_rate": 8.28676219137561e-06,
+      "loss": 0.6416,
+      "step": 862
+    },
+    {
+      "epoch": 1.0347930413917217,
+      "grad_norm": 0.7469325210173965,
+      "learning_rate": 8.281495643293254e-06,
+      "loss": 0.6107,
+      "step": 863
+    },
+    {
+      "epoch": 1.035992801439712,
+      "grad_norm": 0.7806686875957286,
+      "learning_rate": 8.276222692076817e-06,
+      "loss": 0.642,
+      "step": 864
+    },
+    {
+      "epoch": 1.0371925614877024,
+      "grad_norm": 0.7758783574735141,
+      "learning_rate": 8.270943348015333e-06,
+      "loss": 0.6457,
+      "step": 865
+    },
+    {
+      "epoch": 1.038392321535693,
+      "grad_norm": 0.6626976713714866,
+      "learning_rate": 8.265657621410308e-06,
+      "loss": 0.6046,
+      "step": 866
+    },
+    {
+      "epoch": 1.0395920815836832,
+      "grad_norm": 0.7180729110513119,
+      "learning_rate": 8.2603655225757e-06,
+      "loss": 0.6085,
+      "step": 867
+    },
+    {
+      "epoch": 1.0407918416316737,
+      "grad_norm": 0.9268142513774826,
+      "learning_rate": 8.255067061837908e-06,
+      "loss": 0.5917,
+      "step": 868
+    },
+    {
+      "epoch": 1.041991601679664,
+      "grad_norm": 0.7650068515227395,
+      "learning_rate": 8.249762249535738e-06,
+      "loss": 0.6483,
+      "step": 869
+    },
+    {
+      "epoch": 1.0431913617276545,
+      "grad_norm": 0.6173847612200957,
+      "learning_rate": 8.244451096020392e-06,
+      "loss": 0.6258,
+      "step": 870
+    },
+    {
+      "epoch": 1.044391121775645,
+      "grad_norm": 0.5401877202785061,
+      "learning_rate": 8.239133611655445e-06,
+      "loss": 0.6582,
+      "step": 871
+    },
+    {
+      "epoch": 1.0455908818236352,
+      "grad_norm": 0.8339913450126345,
+      "learning_rate": 8.233809806816826e-06,
+      "loss": 0.702,
+      "step": 872
+    },
+    {
+      "epoch": 1.0467906418716257,
+      "grad_norm": 0.7421783902223927,
+      "learning_rate": 8.228479691892798e-06,
+      "loss": 0.6043,
+      "step": 873
+    },
+    {
+      "epoch": 1.047990401919616,
+      "grad_norm": 0.6485869690651025,
+      "learning_rate": 8.223143277283937e-06,
+      "loss": 0.6802,
+      "step": 874
+    },
+    {
+      "epoch": 1.0491901619676065,
+      "grad_norm": 0.8376775424330809,
+      "learning_rate": 8.217800573403105e-06,
+      "loss": 0.5938,
+      "step": 875
+    },
+    {
+      "epoch": 1.0503899220155968,
+      "grad_norm": 0.6241377838561342,
+      "learning_rate": 8.212451590675445e-06,
+      "loss": 0.6256,
+      "step": 876
+    },
+    {
+      "epoch": 1.0515896820635873,
+      "grad_norm": 0.7333742917325202,
+      "learning_rate": 8.20709633953835e-06,
+      "loss": 0.6947,
+      "step": 877
+    },
+    {
+      "epoch": 1.0527894421115778,
+      "grad_norm": 0.7659926461397386,
+      "learning_rate": 8.20173483044144e-06,
+      "loss": 0.5769,
+      "step": 878
+    },
+    {
+      "epoch": 1.053989202159568,
+      "grad_norm": 0.5499942656083657,
+      "learning_rate": 8.196367073846548e-06,
+      "loss": 0.7144,
+      "step": 879
+    },
+    {
+      "epoch": 1.0551889622075585,
+      "grad_norm": 0.8596591269856485,
+      "learning_rate": 8.190993080227699e-06,
+      "loss": 0.6613,
+      "step": 880
+    },
+    {
+      "epoch": 1.0563887222555488,
+      "grad_norm": 0.799981814356054,
+      "learning_rate": 8.185612860071088e-06,
+      "loss": 0.5757,
+      "step": 881
+    },
+    {
+      "epoch": 1.0575884823035393,
+      "grad_norm": 0.5822804320615319,
+      "learning_rate": 8.18022642387506e-06,
+      "loss": 0.6399,
+      "step": 882
+    },
+    {
+      "epoch": 1.0587882423515298,
+      "grad_norm": 0.7513579482146487,
+      "learning_rate": 8.174833782150087e-06,
+      "loss": 0.6252,
+      "step": 883
+    },
+    {
+      "epoch": 1.05998800239952,
+      "grad_norm": 0.747723338953572,
+      "learning_rate": 8.169434945418753e-06,
+      "loss": 0.6896,
+      "step": 884
+    },
+    {
+      "epoch": 1.0611877624475106,
+      "grad_norm": 0.9348483655270661,
+      "learning_rate": 8.164029924215726e-06,
+      "loss": 0.7288,
+      "step": 885
+    },
+    {
+      "epoch": 1.0623875224955008,
+      "grad_norm": 0.8017838449602085,
+      "learning_rate": 8.158618729087746e-06,
+      "loss": 0.6362,
+      "step": 886
+    },
+    {
+      "epoch": 1.0635872825434913,
+      "grad_norm": 0.6526350086776308,
+      "learning_rate": 8.1532013705936e-06,
+      "loss": 0.6259,
+      "step": 887
+    },
+    {
+      "epoch": 1.0647870425914816,
+      "grad_norm": 0.7686722113617305,
+      "learning_rate": 8.147777859304095e-06,
+      "loss": 0.6296,
+      "step": 888
+    },
+    {
+      "epoch": 1.065986802639472,
+      "grad_norm": 0.7302411966675888,
+      "learning_rate": 8.142348205802053e-06,
+      "loss": 0.6991,
+      "step": 889
+    },
+    {
+      "epoch": 1.0671865626874626,
+      "grad_norm": 0.5601035737922527,
+      "learning_rate": 8.136912420682275e-06,
+      "loss": 0.6498,
+      "step": 890
+    },
+    {
+      "epoch": 1.0683863227354529,
+      "grad_norm": 0.805045152553524,
+      "learning_rate": 8.13147051455153e-06,
+      "loss": 0.7064,
+      "step": 891
+    },
+    {
+      "epoch": 1.0695860827834434,
+      "grad_norm": 0.5928026959639249,
+      "learning_rate": 8.126022498028527e-06,
+      "loss": 0.6043,
+      "step": 892
+    },
+    {
+      "epoch": 1.0707858428314336,
+      "grad_norm": 0.7909305590908955,
+      "learning_rate": 8.120568381743901e-06,
+      "loss": 0.6051,
+      "step": 893
+    },
+    {
+      "epoch": 1.0719856028794241,
+      "grad_norm": 0.8191799284024129,
+      "learning_rate": 8.11510817634019e-06,
+      "loss": 0.6285,
+      "step": 894
+    },
+    {
+      "epoch": 1.0731853629274146,
+      "grad_norm": 0.6985090097745722,
+      "learning_rate": 8.10964189247181e-06,
+      "loss": 0.5839,
+      "step": 895
+    },
+    {
+      "epoch": 1.074385122975405,
+      "grad_norm": 0.7633055880376761,
+      "learning_rate": 8.10416954080504e-06,
+      "loss": 0.6305,
+      "step": 896
+    },
+    {
+      "epoch": 1.0755848830233954,
+      "grad_norm": 0.7557111172959706,
+      "learning_rate": 8.098691132018004e-06,
+      "loss": 0.5791,
+      "step": 897
+    },
+    {
+      "epoch": 1.0767846430713857,
+      "grad_norm": 0.5709527709412924,
+      "learning_rate": 8.093206676800635e-06,
+      "loss": 0.6336,
+      "step": 898
+    },
+    {
+      "epoch": 1.0779844031193762,
+      "grad_norm": 0.6652917879121284,
+      "learning_rate": 8.087716185854673e-06,
+      "loss": 0.6635,
+      "step": 899
+    },
+    {
+      "epoch": 1.0791841631673664,
+      "grad_norm": 0.7647659312405806,
+      "learning_rate": 8.08221966989363e-06,
+      "loss": 0.6738,
+      "step": 900
+    },
+    {
+      "epoch": 1.080383923215357,
+      "grad_norm": 0.790523330347467,
+      "learning_rate": 8.076717139642775e-06,
+      "loss": 0.5743,
+      "step": 901
+    },
+    {
+      "epoch": 1.0815836832633474,
+      "grad_norm": 0.7415457762653487,
+      "learning_rate": 8.071208605839118e-06,
+      "loss": 0.6239,
+      "step": 902
+    },
+    {
+      "epoch": 1.0827834433113377,
+      "grad_norm": 0.6364375401106082,
+      "learning_rate": 8.065694079231379e-06,
+      "loss": 0.5959,
+      "step": 903
+    },
+    {
+      "epoch": 1.0839832033593282,
+      "grad_norm": 0.7322831814610964,
+      "learning_rate": 8.06017357057997e-06,
+      "loss": 0.7109,
+      "step": 904
+    },
+    {
+      "epoch": 1.0851829634073185,
+      "grad_norm": 0.7326664889329088,
+      "learning_rate": 8.05464709065698e-06,
+      "loss": 0.5572,
+      "step": 905
+    },
+    {
+      "epoch": 1.086382723455309,
+      "grad_norm": 0.8097688842435833,
+      "learning_rate": 8.049114650246149e-06,
+      "loss": 0.5968,
+      "step": 906
+    },
+    {
+      "epoch": 1.0875824835032994,
+      "grad_norm": 0.8082297761876786,
+      "learning_rate": 8.043576260142844e-06,
+      "loss": 0.5504,
+      "step": 907
+    },
+    {
+      "epoch": 1.0887822435512897,
+      "grad_norm": 0.7700931670572386,
+      "learning_rate": 8.038031931154044e-06,
+      "loss": 0.6451,
+      "step": 908
+    },
+    {
+      "epoch": 1.0899820035992802,
+      "grad_norm": 0.6369168250832012,
+      "learning_rate": 8.03248167409832e-06,
+      "loss": 0.6843,
+      "step": 909
+    },
+    {
+      "epoch": 1.0911817636472705,
+      "grad_norm": 0.7380681634425306,
+      "learning_rate": 8.026925499805802e-06,
+      "loss": 0.6699,
+      "step": 910
+    },
+    {
+      "epoch": 1.092381523695261,
+      "grad_norm": 0.6481000672770302,
+      "learning_rate": 8.021363419118173e-06,
+      "loss": 0.6482,
+      "step": 911
+    },
+    {
+      "epoch": 1.0935812837432513,
+      "grad_norm": 0.6254888477119547,
+      "learning_rate": 8.01579544288864e-06,
+      "loss": 0.6728,
+      "step": 912
+    },
+    {
+      "epoch": 1.0947810437912417,
+      "grad_norm": 0.8215650406361228,
+      "learning_rate": 8.010221581981912e-06,
+      "loss": 0.6579,
+      "step": 913
+    },
+    {
+      "epoch": 1.0959808038392322,
+      "grad_norm": 0.7334573953644762,
+      "learning_rate": 8.004641847274182e-06,
+      "loss": 0.6592,
+      "step": 914
+    },
+    {
+      "epoch": 1.0971805638872225,
+      "grad_norm": 0.64502117270513,
+      "learning_rate": 7.999056249653106e-06,
+      "loss": 0.67,
+      "step": 915
+    },
+    {
+      "epoch": 1.098380323935213,
+      "grad_norm": 0.7423289771175321,
+      "learning_rate": 7.993464800017774e-06,
+      "loss": 0.6587,
+      "step": 916
+    },
+    {
+      "epoch": 1.0995800839832033,
+      "grad_norm": 0.6328514541805109,
+      "learning_rate": 7.9878675092787e-06,
+      "loss": 0.6558,
+      "step": 917
+    },
+    {
+      "epoch": 1.1007798440311938,
+      "grad_norm": 0.6315258751159936,
+      "learning_rate": 7.9822643883578e-06,
+      "loss": 0.6311,
+      "step": 918
+    },
+    {
+      "epoch": 1.101979604079184,
+      "grad_norm": 0.7830994090026224,
+      "learning_rate": 7.976655448188356e-06,
+      "loss": 0.7285,
+      "step": 919
+    },
+    {
+      "epoch": 1.1031793641271745,
+      "grad_norm": 0.656409629518676,
+      "learning_rate": 7.971040699715008e-06,
+      "loss": 0.5734,
+      "step": 920
+    },
+    {
+      "epoch": 1.104379124175165,
+      "grad_norm": 0.7995809349422637,
+      "learning_rate": 7.965420153893741e-06,
+      "loss": 0.6572,
+      "step": 921
+    },
+    {
+      "epoch": 1.1055788842231553,
+      "grad_norm": 0.6045764821458152,
+      "learning_rate": 7.959793821691838e-06,
+      "loss": 0.6165,
+      "step": 922
+    },
+    {
+      "epoch": 1.1067786442711458,
+      "grad_norm": 0.6254737760421022,
+      "learning_rate": 7.954161714087876e-06,
+      "loss": 0.6607,
+      "step": 923
+    },
+    {
+      "epoch": 1.107978404319136,
+      "grad_norm": 0.8767118032126044,
+      "learning_rate": 7.948523842071707e-06,
+      "loss": 0.6039,
+      "step": 924
+    },
+    {
+      "epoch": 1.1091781643671266,
+      "grad_norm": 0.6164747668994374,
+      "learning_rate": 7.942880216644426e-06,
+      "loss": 0.6112,
+      "step": 925
+    },
+    {
+      "epoch": 1.110377924415117,
+      "grad_norm": 0.7393019211191909,
+      "learning_rate": 7.937230848818355e-06,
+      "loss": 0.5887,
+      "step": 926
+    },
+    {
+      "epoch": 1.1115776844631073,
+      "grad_norm": 0.7090151969865134,
+      "learning_rate": 7.931575749617027e-06,
+      "loss": 0.6331,
+      "step": 927
+    },
+    {
+      "epoch": 1.1127774445110978,
+      "grad_norm": 0.8196729176899866,
+      "learning_rate": 7.925914930075147e-06,
+      "loss": 0.645,
+      "step": 928
+    },
+    {
+      "epoch": 1.113977204559088,
+      "grad_norm": 0.5956719830391075,
+      "learning_rate": 7.920248401238593e-06,
+      "loss": 0.5781,
+      "step": 929
+    },
+    {
+      "epoch": 1.1151769646070786,
+      "grad_norm": 0.6762218315155455,
+      "learning_rate": 7.914576174164378e-06,
+      "loss": 0.6568,
+      "step": 930
+    },
+    {
+      "epoch": 1.116376724655069,
+      "grad_norm": 0.636496405234365,
+      "learning_rate": 7.908898259920636e-06,
+      "loss": 0.7224,
+      "step": 931
+    },
+    {
+      "epoch": 1.1175764847030594,
+      "grad_norm": 0.865234836801879,
+      "learning_rate": 7.903214669586596e-06,
+      "loss": 0.6938,
+      "step": 932
+    },
+    {
+      "epoch": 1.1187762447510499,
+      "grad_norm": 0.8873977059438501,
+      "learning_rate": 7.897525414252564e-06,
+      "loss": 0.6277,
+      "step": 933
+    },
+    {
+      "epoch": 1.1199760047990401,
+      "grad_norm": 0.7977055991778375,
+      "learning_rate": 7.891830505019905e-06,
+      "loss": 0.6089,
+      "step": 934
+    },
+    {
+      "epoch": 1.1211757648470306,
+      "grad_norm": 0.7985938779341601,
+      "learning_rate": 7.886129953001002e-06,
+      "loss": 0.6699,
+      "step": 935
+    },
+    {
+      "epoch": 1.122375524895021,
+      "grad_norm": 0.7529036412800081,
+      "learning_rate": 7.880423769319265e-06,
+      "loss": 0.6115,
+      "step": 936
+    },
+    {
+      "epoch": 1.1235752849430114,
+      "grad_norm": 0.7728641750560808,
+      "learning_rate": 7.874711965109085e-06,
+      "loss": 0.5787,
+      "step": 937
+    },
+    {
+      "epoch": 1.124775044991002,
+      "grad_norm": 0.7603933905232403,
+      "learning_rate": 7.86899455151582e-06,
+      "loss": 0.6112,
+      "step": 938
+    },
+    {
+      "epoch": 1.1259748050389922,
+      "grad_norm": 0.8582898047056604,
+      "learning_rate": 7.863271539695777e-06,
+      "loss": 0.6396,
+      "step": 939
+    },
+    {
+      "epoch": 1.1271745650869827,
+      "grad_norm": 0.5942730046106529,
+      "learning_rate": 7.857542940816183e-06,
+      "loss": 0.5919,
+      "step": 940
+    },
+    {
+      "epoch": 1.128374325134973,
+      "grad_norm": 0.6646323437342261,
+      "learning_rate": 7.85180876605517e-06,
+      "loss": 0.576,
+      "step": 941
+    },
+    {
+      "epoch": 1.1295740851829634,
+      "grad_norm": 0.7939071629516148,
+      "learning_rate": 7.846069026601745e-06,
+      "loss": 0.6983,
+      "step": 942
+    },
+    {
+      "epoch": 1.1307738452309537,
+      "grad_norm": 0.7676942847848861,
+      "learning_rate": 7.84032373365578e-06,
+      "loss": 0.6937,
+      "step": 943
+    },
+    {
+      "epoch": 1.1319736052789442,
+      "grad_norm": 0.6185861404100774,
+      "learning_rate": 7.83457289842798e-06,
+      "loss": 0.6797,
+      "step": 944
+    },
+    {
+      "epoch": 1.1331733653269347,
+      "grad_norm": 0.8840067772768776,
+      "learning_rate": 7.828816532139866e-06,
+      "loss": 0.6588,
+      "step": 945
+    },
+    {
+      "epoch": 1.134373125374925,
+      "grad_norm": 0.720664166016145,
+      "learning_rate": 7.823054646023749e-06,
+      "loss": 0.6417,
+      "step": 946
+    },
+    {
+      "epoch": 1.1355728854229155,
+      "grad_norm": 0.655102377172491,
+      "learning_rate": 7.817287251322712e-06,
+      "loss": 0.6567,
+      "step": 947
+    },
+    {
+      "epoch": 1.1367726454709057,
+      "grad_norm": 0.7295884468134075,
+      "learning_rate": 7.811514359290591e-06,
+      "loss": 0.6704,
+      "step": 948
+    },
+    {
+      "epoch": 1.1379724055188962,
+      "grad_norm": 0.6055585752339337,
+      "learning_rate": 7.805735981191939e-06,
+      "loss": 0.6365,
+      "step": 949
+    },
+    {
+      "epoch": 1.1391721655668867,
+      "grad_norm": 0.9045297618528259,
+      "learning_rate": 7.799952128302027e-06,
+      "loss": 0.6148,
+      "step": 950
+    },
+    {
+      "epoch": 1.140371925614877,
+      "grad_norm": 0.7573009176463668,
+      "learning_rate": 7.794162811906798e-06,
+      "loss": 0.6422,
+      "step": 951
+    },
+    {
+      "epoch": 1.1415716856628675,
+      "grad_norm": 0.5189738295277829,
+      "learning_rate": 7.78836804330286e-06,
+      "loss": 0.6972,
+      "step": 952
+    },
+    {
+      "epoch": 1.1427714457108578,
+      "grad_norm": 0.761896631397838,
+      "learning_rate": 7.782567833797458e-06,
+      "loss": 0.6136,
+      "step": 953
+    },
+    {
+      "epoch": 1.1439712057588483,
+      "grad_norm": 1.1206412330960442,
+      "learning_rate": 7.776762194708459e-06,
+      "loss": 0.6781,
+      "step": 954
+    },
+    {
+      "epoch": 1.1451709658068387,
+      "grad_norm": 0.582336141346009,
+      "learning_rate": 7.770951137364317e-06,
+      "loss": 0.6204,
+      "step": 955
+    },
+    {
+      "epoch": 1.146370725854829,
+      "grad_norm": 0.711849105566516,
+      "learning_rate": 7.765134673104065e-06,
+      "loss": 0.7008,
+      "step": 956
+    },
+    {
+      "epoch": 1.1475704859028195,
+      "grad_norm": 0.7048388709901285,
+      "learning_rate": 7.759312813277285e-06,
+      "loss": 0.6661,
+      "step": 957
+    },
+    {
+      "epoch": 1.1487702459508098,
+      "grad_norm": 0.7260048434291833,
+      "learning_rate": 7.753485569244083e-06,
+      "loss": 0.6246,
+      "step": 958
+    },
+    {
+      "epoch": 1.1499700059988003,
+      "grad_norm": 0.6785224096803231,
+      "learning_rate": 7.747652952375078e-06,
+      "loss": 0.613,
+      "step": 959
+    },
+    {
+      "epoch": 1.1511697660467906,
+      "grad_norm": 0.7560505404058467,
+      "learning_rate": 7.741814974051366e-06,
+      "loss": 0.5887,
+      "step": 960
+    },
+    {
+      "epoch": 1.152369526094781,
+      "grad_norm": 0.7965630156892449,
+      "learning_rate": 7.73597164566451e-06,
+      "loss": 0.6341,
+      "step": 961
+    },
+    {
+      "epoch": 1.1535692861427713,
+      "grad_norm": 0.6923330802478767,
+      "learning_rate": 7.730122978616511e-06,
+      "loss": 0.6093,
+      "step": 962
+    },
+    {
+      "epoch": 1.1547690461907618,
+      "grad_norm": 0.6879908653795777,
+      "learning_rate": 7.724268984319785e-06,
+      "loss": 0.6887,
+      "step": 963
+    },
+    {
+      "epoch": 1.1559688062387523,
+      "grad_norm": 0.6108013120938313,
+      "learning_rate": 7.718409674197147e-06,
+      "loss": 0.6261,
+      "step": 964
+    },
+    {
+      "epoch": 1.1571685662867426,
+      "grad_norm": 0.7592662810855833,
+      "learning_rate": 7.71254505968178e-06,
+      "loss": 0.5789,
+      "step": 965
+    },
+    {
+      "epoch": 1.158368326334733,
+      "grad_norm": 0.803730538996608,
+      "learning_rate": 7.70667515221722e-06,
+      "loss": 0.6657,
+      "step": 966
+    },
+    {
+      "epoch": 1.1595680863827234,
+      "grad_norm": 0.9674342724300058,
+      "learning_rate": 7.700799963257335e-06,
+      "loss": 0.6145,
+      "step": 967
+    },
+    {
+      "epoch": 1.1607678464307138,
+      "grad_norm": 0.5854325160848438,
+      "learning_rate": 7.69491950426629e-06,
+      "loss": 0.6678,
+      "step": 968
+    },
+    {
+      "epoch": 1.1619676064787043,
+      "grad_norm": 0.6718578895332913,
+      "learning_rate": 7.689033786718539e-06,
+      "loss": 0.6125,
+      "step": 969
+    },
+    {
+      "epoch": 1.1631673665266946,
+      "grad_norm": 0.7289857412083824,
+      "learning_rate": 7.683142822098795e-06,
+      "loss": 0.6835,
+      "step": 970
+    },
+    {
+      "epoch": 1.164367126574685,
+      "grad_norm": 0.5641877014830333,
+      "learning_rate": 7.677246621902014e-06,
+      "loss": 0.6334,
+      "step": 971
+    },
+    {
+      "epoch": 1.1655668866226754,
+      "grad_norm": 0.8539371224490931,
+      "learning_rate": 7.671345197633361e-06,
+      "loss": 0.6436,
+      "step": 972
+    },
+    {
+      "epoch": 1.1667666466706659,
+      "grad_norm": 0.8045170582715613,
+      "learning_rate": 7.6654385608082e-06,
+      "loss": 0.6559,
+      "step": 973
+    },
+    {
+      "epoch": 1.1679664067186564,
+      "grad_norm": 0.565700728135143,
+      "learning_rate": 7.659526722952066e-06,
+      "loss": 0.5513,
+      "step": 974
+    },
+    {
+      "epoch": 1.1691661667666466,
+      "grad_norm": 0.8743238461492279,
+      "learning_rate": 7.653609695600636e-06,
+      "loss": 0.6148,
+      "step": 975
+    },
+    {
+      "epoch": 1.1703659268146371,
+      "grad_norm": 0.6122517836173725,
+      "learning_rate": 7.647687490299723e-06,
+      "loss": 0.7412,
+      "step": 976
+    },
+    {
+      "epoch": 1.1715656868626274,
+      "grad_norm": 0.7449414224622495,
+      "learning_rate": 7.641760118605237e-06,
+      "loss": 0.6942,
+      "step": 977
+    },
+    {
+      "epoch": 1.172765446910618,
+      "grad_norm": 0.877399381789804,
+      "learning_rate": 7.635827592083168e-06,
+      "loss": 0.6523,
+      "step": 978
+    },
+    {
+      "epoch": 1.1739652069586084,
+      "grad_norm": 0.6643133900992996,
+      "learning_rate": 7.629889922309576e-06,
+      "loss": 0.5856,
+      "step": 979
+    },
+    {
+      "epoch": 1.1751649670065987,
+      "grad_norm": 0.728087789468,
+      "learning_rate": 7.6239471208705404e-06,
+      "loss": 0.6325,
+      "step": 980
+    },
+    {
+      "epoch": 1.1763647270545892,
+      "grad_norm": 1.2168375793316153,
+      "learning_rate": 7.617999199362166e-06,
+      "loss": 0.6837,
+      "step": 981
+    },
+    {
+      "epoch": 1.1775644871025794,
+      "grad_norm": 0.5792305435721297,
+      "learning_rate": 7.612046169390543e-06,
+      "loss": 0.6752,
+      "step": 982
+    },
+    {
+      "epoch": 1.17876424715057,
+      "grad_norm": 0.758139899265491,
+      "learning_rate": 7.606088042571731e-06,
+      "loss": 0.5868,
+      "step": 983
+    },
+    {
+      "epoch": 1.1799640071985602,
+      "grad_norm": 0.7826574118053469,
+      "learning_rate": 7.600124830531737e-06,
+      "loss": 0.5982,
+      "step": 984
+    },
+    {
+      "epoch": 1.1811637672465507,
+      "grad_norm": 0.6000356797051724,
+      "learning_rate": 7.594156544906484e-06,
+      "loss": 0.6364,
+      "step": 985
+    },
+    {
+      "epoch": 1.182363527294541,
+      "grad_norm": 0.9687253001599708,
+      "learning_rate": 7.588183197341805e-06,
+      "loss": 0.6904,
+      "step": 986
+    },
+    {
+      "epoch": 1.1835632873425315,
+      "grad_norm": 0.6188352397748479,
+      "learning_rate": 7.582204799493402e-06,
+      "loss": 0.7003,
+      "step": 987
+    },
+    {
+      "epoch": 1.184763047390522,
+      "grad_norm": 0.7202260204090514,
+      "learning_rate": 7.576221363026834e-06,
+      "loss": 0.684,
+      "step": 988
+    },
+    {
+      "epoch": 1.1859628074385122,
+      "grad_norm": 0.7586597687175531,
+      "learning_rate": 7.570232899617497e-06,
+      "loss": 0.6204,
+      "step": 989
+    },
+    {
+      "epoch": 1.1871625674865027,
+      "grad_norm": 0.6651953021064884,
+      "learning_rate": 7.564239420950587e-06,
+      "loss": 0.718,
+      "step": 990
+    },
+    {
+      "epoch": 1.188362327534493,
+      "grad_norm": 0.8721655660530172,
+      "learning_rate": 7.558240938721091e-06,
+      "loss": 0.6177,
+      "step": 991
+    },
+    {
+      "epoch": 1.1895620875824835,
+      "grad_norm": 0.8435311574009753,
+      "learning_rate": 7.55223746463376e-06,
+      "loss": 0.6586,
+      "step": 992
+    },
+    {
+      "epoch": 1.190761847630474,
+      "grad_norm": 0.622077937555054,
+      "learning_rate": 7.546229010403086e-06,
+      "loss": 0.6699,
+      "step": 993
+    },
+    {
+      "epoch": 1.1919616076784643,
+      "grad_norm": 0.5482286323460653,
+      "learning_rate": 7.540215587753274e-06,
+      "loss": 0.5602,
+      "step": 994
+    },
+    {
+      "epoch": 1.1931613677264548,
+      "grad_norm": 0.7358949985009436,
+      "learning_rate": 7.534197208418228e-06,
+      "loss": 0.6869,
+      "step": 995
+    },
+    {
+      "epoch": 1.194361127774445,
+      "grad_norm": 0.6697867463290572,
+      "learning_rate": 7.5281738841415256e-06,
+      "loss": 0.6525,
+      "step": 996
+    },
+    {
+      "epoch": 1.1955608878224355,
+      "grad_norm": 0.8182406510517298,
+      "learning_rate": 7.522145626676386e-06,
+      "loss": 0.6842,
+      "step": 997
+    },
+    {
+      "epoch": 1.196760647870426,
+      "grad_norm": 0.5186713299722027,
+      "learning_rate": 7.5161124477856636e-06,
+      "loss": 0.613,
+      "step": 998
+    },
+    {
+      "epoch": 1.1979604079184163,
+      "grad_norm": 0.8244793743836701,
+      "learning_rate": 7.510074359241808e-06,
+      "loss": 0.6283,
+      "step": 999
+    },
+    {
+      "epoch": 1.1991601679664068,
+      "grad_norm": 0.6792712021878325,
+      "learning_rate": 7.5040313728268544e-06,
+      "loss": 0.6142,
+      "step": 1000
+    },
+    {
+      "epoch": 1.200359928014397,
+      "grad_norm": 0.6717251990019677,
+      "learning_rate": 7.497983500332392e-06,
+      "loss": 0.6295,
+      "step": 1001
+    },
+    {
+      "epoch": 1.2015596880623876,
+      "grad_norm": 0.5937700301061093,
+      "learning_rate": 7.491930753559547e-06,
+      "loss": 0.5831,
+      "step": 1002
+    },
+    {
+      "epoch": 1.202759448110378,
+      "grad_norm": 0.743402154852896,
+      "learning_rate": 7.4858731443189526e-06,
+      "loss": 0.6018,
+      "step": 1003
+    },
+    {
+      "epoch": 1.2039592081583683,
+      "grad_norm": 0.6412286431039477,
+      "learning_rate": 7.4798106844307325e-06,
+      "loss": 0.7176,
+      "step": 1004
+    },
+    {
+      "epoch": 1.2051589682063588,
+      "grad_norm": 0.712722659936385,
+      "learning_rate": 7.473743385724478e-06,
+      "loss": 0.6298,
+      "step": 1005
+    },
+    {
+      "epoch": 1.206358728254349,
+      "grad_norm": 0.6689458591312148,
+      "learning_rate": 7.467671260039218e-06,
+      "loss": 0.6576,
+      "step": 1006
+    },
+    {
+      "epoch": 1.2075584883023396,
+      "grad_norm": 0.7929731468061676,
+      "learning_rate": 7.4615943192234e-06,
+      "loss": 0.6488,
+      "step": 1007
+    },
+    {
+      "epoch": 1.2087582483503299,
+      "grad_norm": 0.8070861175008949,
+      "learning_rate": 7.4555125751348735e-06,
+      "loss": 0.6181,
+      "step": 1008
+    },
+    {
+      "epoch": 1.2099580083983204,
+      "grad_norm": 0.6500828766089807,
+      "learning_rate": 7.449426039640853e-06,
+      "loss": 0.6571,
+      "step": 1009
+    },
+    {
+      "epoch": 1.2111577684463106,
+      "grad_norm": 0.6679890203007678,
+      "learning_rate": 7.443334724617904e-06,
+      "loss": 0.6086,
+      "step": 1010
+    },
+    {
+      "epoch": 1.2123575284943011,
+      "grad_norm": 0.5709805644821978,
+      "learning_rate": 7.437238641951922e-06,
+      "loss": 0.6289,
+      "step": 1011
+    },
+    {
+      "epoch": 1.2135572885422916,
+      "grad_norm": 0.6959911040226187,
+      "learning_rate": 7.431137803538105e-06,
+      "loss": 0.7094,
+      "step": 1012
+    },
+    {
+      "epoch": 1.2147570485902819,
+      "grad_norm": 0.6389305357842437,
+      "learning_rate": 7.425032221280925e-06,
+      "loss": 0.6112,
+      "step": 1013
+    },
+    {
+      "epoch": 1.2159568086382724,
+      "grad_norm": 0.6980973604370527,
+      "learning_rate": 7.418921907094117e-06,
+      "loss": 0.6821,
+      "step": 1014
+    },
+    {
+      "epoch": 1.2171565686862627,
+      "grad_norm": 0.8056526394131966,
+      "learning_rate": 7.412806872900648e-06,
+      "loss": 0.5926,
+      "step": 1015
+    },
+    {
+      "epoch": 1.2183563287342531,
+      "grad_norm": 0.5509146634760903,
+      "learning_rate": 7.406687130632692e-06,
+      "loss": 0.6358,
+      "step": 1016
+    },
+    {
+      "epoch": 1.2195560887822436,
+      "grad_norm": 0.6980703149971923,
+      "learning_rate": 7.400562692231611e-06,
+      "loss": 0.6635,
+      "step": 1017
+    },
+    {
+      "epoch": 1.220755848830234,
+      "grad_norm": 0.6236557163845787,
+      "learning_rate": 7.394433569647935e-06,
+      "loss": 0.669,
+      "step": 1018
+    },
+    {
+      "epoch": 1.2219556088782244,
+      "grad_norm": 0.6846420411643493,
+      "learning_rate": 7.3882997748413285e-06,
+      "loss": 0.6305,
+      "step": 1019
+    },
+    {
+      "epoch": 1.2231553689262147,
+      "grad_norm": 0.7357860502422784,
+      "learning_rate": 7.382161319780574e-06,
+      "loss": 0.719,
+      "step": 1020
+    },
+    {
+      "epoch": 1.2243551289742052,
+      "grad_norm": 0.6768304157113985,
+      "learning_rate": 7.37601821644355e-06,
+      "loss": 0.6303,
+      "step": 1021
+    },
+    {
+      "epoch": 1.2255548890221957,
+      "grad_norm": 0.7669717553744075,
+      "learning_rate": 7.3698704768172026e-06,
+      "loss": 0.7411,
+      "step": 1022
+    },
+    {
+      "epoch": 1.226754649070186,
+      "grad_norm": 0.7701810105912146,
+      "learning_rate": 7.363718112897525e-06,
+      "loss": 0.6904,
+      "step": 1023
+    },
+    {
+      "epoch": 1.2279544091181764,
+      "grad_norm": 0.6981186558383136,
+      "learning_rate": 7.357561136689536e-06,
+      "loss": 0.6152,
+      "step": 1024
+    },
+    {
+      "epoch": 1.2291541691661667,
+      "grad_norm": 0.6666100717443457,
+      "learning_rate": 7.351399560207253e-06,
+      "loss": 0.6399,
+      "step": 1025
+    },
+    {
+      "epoch": 1.2303539292141572,
+      "grad_norm": 0.6869300699310232,
+      "learning_rate": 7.345233395473664e-06,
+      "loss": 0.7237,
+      "step": 1026
+    },
+    {
+      "epoch": 1.2315536892621475,
+      "grad_norm": 0.6965197593345103,
+      "learning_rate": 7.339062654520724e-06,
+      "loss": 0.6339,
+      "step": 1027
+    },
+    {
+      "epoch": 1.232753449310138,
+      "grad_norm": 0.6292232680942303,
+      "learning_rate": 7.332887349389301e-06,
+      "loss": 0.6196,
+      "step": 1028
+    },
+    {
+      "epoch": 1.2339532093581285,
+      "grad_norm": 0.7687665423496974,
+      "learning_rate": 7.3267074921291795e-06,
+      "loss": 0.6551,
+      "step": 1029
+    },
+    {
+      "epoch": 1.2351529694061187,
+      "grad_norm": 0.6249840034898358,
+      "learning_rate": 7.320523094799025e-06,
+      "loss": 0.6302,
+      "step": 1030
+    },
+    {
+      "epoch": 1.2363527294541092,
+      "grad_norm": 0.6468018329037355,
+      "learning_rate": 7.3143341694663604e-06,
+      "loss": 0.6683,
+      "step": 1031
+    },
+    {
+      "epoch": 1.2375524895020995,
+      "grad_norm": 0.7867545650766382,
+      "learning_rate": 7.308140728207544e-06,
+      "loss": 0.6594,
+      "step": 1032
+    },
+    {
+      "epoch": 1.23875224955009,
+      "grad_norm": 0.7323728579960705,
+      "learning_rate": 7.301942783107747e-06,
+      "loss": 0.6151,
+      "step": 1033
+    },
+    {
+      "epoch": 1.2399520095980803,
+      "grad_norm": 0.7095345504633144,
+      "learning_rate": 7.2957403462609276e-06,
+      "loss": 0.6424,
+      "step": 1034
+    },
+    {
+      "epoch": 1.2411517696460708,
+      "grad_norm": 0.7230751286642569,
+      "learning_rate": 7.289533429769812e-06,
+      "loss": 0.6643,
+      "step": 1035
+    },
+    {
+      "epoch": 1.2423515296940613,
+      "grad_norm": 0.7438701965456714,
+      "learning_rate": 7.2833220457458596e-06,
+      "loss": 0.639,
+      "step": 1036
+    },
+    {
+      "epoch": 1.2435512897420515,
+      "grad_norm": 0.6573622376094815,
+      "learning_rate": 7.277106206309258e-06,
+      "loss": 0.6497,
+      "step": 1037
+    },
+    {
+      "epoch": 1.244751049790042,
+      "grad_norm": 0.7426817035433573,
+      "learning_rate": 7.270885923588879e-06,
+      "loss": 0.6933,
+      "step": 1038
+    },
+    {
+      "epoch": 1.2459508098380323,
+      "grad_norm": 0.6219922286802084,
+      "learning_rate": 7.26466120972227e-06,
+      "loss": 0.6609,
+      "step": 1039
+    },
+    {
+      "epoch": 1.2471505698860228,
+      "grad_norm": 0.7422857469589963,
+      "learning_rate": 7.258432076855625e-06,
+      "loss": 0.6354,
+      "step": 1040
+    },
+    {
+      "epoch": 1.2483503299340133,
+      "grad_norm": 0.8573633191476585,
+      "learning_rate": 7.2521985371437575e-06,
+      "loss": 0.6304,
+      "step": 1041
+    },
+    {
+      "epoch": 1.2495500899820036,
+      "grad_norm": 0.8093256278783655,
+      "learning_rate": 7.245960602750081e-06,
+      "loss": 0.6303,
+      "step": 1042
+    },
+    {
+      "epoch": 1.250749850029994,
+      "grad_norm": 0.581070725446635,
+      "learning_rate": 7.239718285846586e-06,
+      "loss": 0.581,
+      "step": 1043
+    },
+    {
+      "epoch": 1.2519496100779843,
+      "grad_norm": 0.6549868295461385,
+      "learning_rate": 7.233471598613815e-06,
+      "loss": 0.6951,
+      "step": 1044
+    },
+    {
+      "epoch": 1.2531493701259748,
+      "grad_norm": 0.9138529056050184,
+      "learning_rate": 7.227220553240835e-06,
+      "loss": 0.6326,
+      "step": 1045
+    },
+    {
+      "epoch": 1.2543491301739653,
+      "grad_norm": 0.704514262769512,
+      "learning_rate": 7.220965161925215e-06,
+      "loss": 0.5955,
+      "step": 1046
+    },
+    {
+      "epoch": 1.2555488902219556,
+      "grad_norm": 0.813534183420755,
+      "learning_rate": 7.214705436873016e-06,
+      "loss": 0.6434,
+      "step": 1047
+    },
+    {
+      "epoch": 1.256748650269946,
+      "grad_norm": 0.7262085566877929,
+      "learning_rate": 7.208441390298742e-06,
+      "loss": 0.6835,
+      "step": 1048
+    },
+    {
+      "epoch": 1.2579484103179364,
+      "grad_norm": 0.550271216349488,
+      "learning_rate": 7.202173034425334e-06,
+      "loss": 0.6457,
+      "step": 1049
+    },
+    {
+      "epoch": 1.2591481703659269,
+      "grad_norm": 0.7468724934410274,
+      "learning_rate": 7.195900381484144e-06,
+      "loss": 0.6574,
+      "step": 1050
+    },
+    {
+      "epoch": 1.2603479304139174,
+      "grad_norm": 0.572209941881098,
+      "learning_rate": 7.189623443714908e-06,
+      "loss": 0.6297,
+      "step": 1051
+    },
+    {
+      "epoch": 1.2615476904619076,
+      "grad_norm": 0.8468906643732181,
+      "learning_rate": 7.183342233365721e-06,
+      "loss": 0.6229,
+      "step": 1052
+    },
+    {
+      "epoch": 1.262747450509898,
+      "grad_norm": 0.6943785807969055,
+      "learning_rate": 7.177056762693018e-06,
+      "loss": 0.6624,
+      "step": 1053
+    },
+    {
+      "epoch": 1.2639472105578884,
+      "grad_norm": 0.6925625991995398,
+      "learning_rate": 7.170767043961542e-06,
+      "loss": 0.6311,
+      "step": 1054
+    },
+    {
+      "epoch": 1.2651469706058789,
+      "grad_norm": 0.6762621801494739,
+      "learning_rate": 7.1644730894443305e-06,
+      "loss": 0.5871,
+      "step": 1055
+    },
+    {
+      "epoch": 1.2663467306538692,
+      "grad_norm": 0.6459618641769452,
+      "learning_rate": 7.158174911422685e-06,
+      "loss": 0.6532,
+      "step": 1056
+    },
+    {
+      "epoch": 1.2675464907018597,
+      "grad_norm": 0.8083476414846833,
+      "learning_rate": 7.151872522186147e-06,
+      "loss": 0.594,
+      "step": 1057
+    },
+    {
+      "epoch": 1.26874625074985,
+      "grad_norm": 0.6951131175229895,
+      "learning_rate": 7.145565934032472e-06,
+      "loss": 0.624,
+      "step": 1058
+    },
+    {
+      "epoch": 1.2699460107978404,
+      "grad_norm": 0.5042926125416562,
+      "learning_rate": 7.139255159267619e-06,
+      "loss": 0.6358,
+      "step": 1059
+    },
+    {
+      "epoch": 1.271145770845831,
+      "grad_norm": 0.9478685294428895,
+      "learning_rate": 7.132940210205705e-06,
+      "loss": 0.5738,
+      "step": 1060
+    },
+    {
+      "epoch": 1.2723455308938212,
+      "grad_norm": 0.694451255964263,
+      "learning_rate": 7.126621099168998e-06,
+      "loss": 0.6865,
+      "step": 1061
+    },
+    {
+      "epoch": 1.2735452909418117,
+      "grad_norm": 0.6123286837725577,
+      "learning_rate": 7.120297838487887e-06,
+      "loss": 0.622,
+      "step": 1062
+    },
+    {
+      "epoch": 1.274745050989802,
+      "grad_norm": 0.8242058769471933,
+      "learning_rate": 7.113970440500858e-06,
+      "loss": 0.6465,
+      "step": 1063
+    },
+    {
+      "epoch": 1.2759448110377924,
+      "grad_norm": 0.6682649397103331,
+      "learning_rate": 7.107638917554468e-06,
+      "loss": 0.6296,
+      "step": 1064
+    },
+    {
+      "epoch": 1.277144571085783,
+      "grad_norm": 0.5673261780539186,
+      "learning_rate": 7.101303282003324e-06,
+      "loss": 0.585,
+      "step": 1065
+    },
+    {
+      "epoch": 1.2783443311337732,
+      "grad_norm": 0.7034650786971746,
+      "learning_rate": 7.094963546210059e-06,
+      "loss": 0.6322,
+      "step": 1066
+    },
+    {
+      "epoch": 1.2795440911817637,
+      "grad_norm": 0.6856469553412491,
+      "learning_rate": 7.088619722545307e-06,
+      "loss": 0.6143,
+      "step": 1067
+    },
+    {
+      "epoch": 1.280743851229754,
+      "grad_norm": 0.6117789618536958,
+      "learning_rate": 7.082271823387675e-06,
+      "loss": 0.6641,
+      "step": 1068
+    },
+    {
+      "epoch": 1.2819436112777445,
+      "grad_norm": 0.7097706181718433,
+      "learning_rate": 7.07591986112373e-06,
+      "loss": 0.7064,
+      "step": 1069
+    },
+    {
+      "epoch": 1.283143371325735,
+      "grad_norm": 0.6841325747730488,
+      "learning_rate": 7.0695638481479565e-06,
+      "loss": 0.6184,
+      "step": 1070
+    },
+    {
+      "epoch": 1.2843431313737252,
+      "grad_norm": 0.8318445369442227,
+      "learning_rate": 7.063203796862752e-06,
+      "loss": 0.6399,
+      "step": 1071
+    },
+    {
+      "epoch": 1.2855428914217157,
+      "grad_norm": 0.4578231661291545,
+      "learning_rate": 7.056839719678391e-06,
+      "loss": 0.6473,
+      "step": 1072
+    },
+    {
+      "epoch": 1.286742651469706,
+      "grad_norm": 0.8006195088069488,
+      "learning_rate": 7.050471629013003e-06,
+      "loss": 0.6145,
+      "step": 1073
+    },
+    {
+      "epoch": 1.2879424115176965,
+      "grad_norm": 0.6727364270890672,
+      "learning_rate": 7.044099537292548e-06,
+      "loss": 0.6281,
+      "step": 1074
+    },
+    {
+      "epoch": 1.289142171565687,
+      "grad_norm": 1.0538331322539352,
+      "learning_rate": 7.037723456950797e-06,
+      "loss": 0.6266,
+      "step": 1075
+    },
+    {
+      "epoch": 1.2903419316136773,
+      "grad_norm": 0.626784875264221,
+      "learning_rate": 7.0313434004293015e-06,
+      "loss": 0.6819,
+      "step": 1076
+    },
+    {
+      "epoch": 1.2915416916616675,
+      "grad_norm": 0.5464614778448484,
+      "learning_rate": 7.02495938017737e-06,
+      "loss": 0.586,
+      "step": 1077
+    },
+    {
+      "epoch": 1.292741451709658,
+      "grad_norm": 0.6690811552578267,
+      "learning_rate": 7.018571408652045e-06,
+      "loss": 0.6915,
+      "step": 1078
+    },
+    {
+      "epoch": 1.2939412117576485,
+      "grad_norm": 0.730862264750046,
+      "learning_rate": 7.012179498318088e-06,
+      "loss": 0.6646,
+      "step": 1079
+    },
+    {
+      "epoch": 1.2951409718056388,
+      "grad_norm": 0.5329908348747406,
+      "learning_rate": 7.0057836616479334e-06,
+      "loss": 0.7027,
+      "step": 1080
+    },
+    {
+      "epoch": 1.2963407318536293,
+      "grad_norm": 0.8233431844664252,
+      "learning_rate": 6.9993839111216865e-06,
+      "loss": 0.6272,
+      "step": 1081
+    },
+    {
+      "epoch": 1.2975404919016196,
+      "grad_norm": 0.6808125676138477,
+      "learning_rate": 6.992980259227085e-06,
+      "loss": 0.6851,
+      "step": 1082
+    },
+    {
+      "epoch": 1.29874025194961,
+      "grad_norm": 0.9643095276335473,
+      "learning_rate": 6.986572718459479e-06,
+      "loss": 0.6963,
+      "step": 1083
+    },
+    {
+      "epoch": 1.2999400119976006,
+      "grad_norm": 0.44399583003612864,
+      "learning_rate": 6.98016130132181e-06,
+      "loss": 0.5905,
+      "step": 1084
+    },
+    {
+      "epoch": 1.3011397720455908,
+      "grad_norm": 0.7485852209698809,
+      "learning_rate": 6.973746020324581e-06,
+      "loss": 0.6267,
+      "step": 1085
+    },
+    {
+      "epoch": 1.3023395320935813,
+      "grad_norm": 0.6343967189087504,
+      "learning_rate": 6.967326887985837e-06,
+      "loss": 0.5925,
+      "step": 1086
+    },
+    {
+      "epoch": 1.3035392921415716,
+      "grad_norm": 0.8508973665476871,
+      "learning_rate": 6.960903916831132e-06,
+      "loss": 0.6478,
+      "step": 1087
+    },
+    {
+      "epoch": 1.304739052189562,
+      "grad_norm": 0.49693666140528264,
+      "learning_rate": 6.954477119393521e-06,
+      "loss": 0.6109,
+      "step": 1088
+    },
+    {
+      "epoch": 1.3059388122375526,
+      "grad_norm": 0.768961587133166,
+      "learning_rate": 6.948046508213515e-06,
+      "loss": 0.6555,
+      "step": 1089
+    },
+    {
+      "epoch": 1.3071385722855429,
+      "grad_norm": 0.7314059891873381,
+      "learning_rate": 6.94161209583907e-06,
+      "loss": 0.6874,
+      "step": 1090
+    },
+    {
+      "epoch": 1.3083383323335334,
+      "grad_norm": 0.7742341576454875,
+      "learning_rate": 6.9351738948255645e-06,
+      "loss": 0.6901,
+      "step": 1091
+    },
+    {
+      "epoch": 1.3095380923815236,
+      "grad_norm": 0.5975730969202686,
+      "learning_rate": 6.9287319177357615e-06,
+      "loss": 0.7018,
+      "step": 1092
+    },
+    {
+      "epoch": 1.3107378524295141,
+      "grad_norm": 0.764953793274789,
+      "learning_rate": 6.922286177139797e-06,
+      "loss": 0.5935,
+      "step": 1093
+    },
+    {
+      "epoch": 1.3119376124775046,
+      "grad_norm": 0.6493804506074498,
+      "learning_rate": 6.915836685615149e-06,
+      "loss": 0.6654,
+      "step": 1094
+    },
+    {
+      "epoch": 1.313137372525495,
+      "grad_norm": 0.7936377495763912,
+      "learning_rate": 6.9093834557466164e-06,
+      "loss": 0.6357,
+      "step": 1095
+    },
+    {
+      "epoch": 1.3143371325734852,
+      "grad_norm": 0.5796146895088004,
+      "learning_rate": 6.902926500126292e-06,
+      "loss": 0.6089,
+      "step": 1096
+    },
+    {
+      "epoch": 1.3155368926214757,
+      "grad_norm": 0.746242101409277,
+      "learning_rate": 6.896465831353535e-06,
+      "loss": 0.7047,
+      "step": 1097
+    },
+    {
+      "epoch": 1.3167366526694662,
+      "grad_norm": 0.5367920357525211,
+      "learning_rate": 6.890001462034957e-06,
+      "loss": 0.6199,
+      "step": 1098
+    },
+    {
+      "epoch": 1.3179364127174564,
+      "grad_norm": 0.7319144508274227,
+      "learning_rate": 6.883533404784384e-06,
+      "loss": 0.5796,
+      "step": 1099
+    },
+    {
+      "epoch": 1.319136172765447,
+      "grad_norm": 0.8063686483869226,
+      "learning_rate": 6.877061672222842e-06,
+      "loss": 0.6073,
+      "step": 1100
+    },
+    {
+      "epoch": 1.3203359328134372,
+      "grad_norm": 0.7251354738596527,
+      "learning_rate": 6.870586276978526e-06,
+      "loss": 0.6583,
+      "step": 1101
+    },
+    {
+      "epoch": 1.3215356928614277,
+      "grad_norm": 0.6011859592461312,
+      "learning_rate": 6.864107231686781e-06,
+      "loss": 0.6082,
+      "step": 1102
+    },
+    {
+      "epoch": 1.3227354529094182,
+      "grad_norm": 0.6832987527658642,
+      "learning_rate": 6.857624548990071e-06,
+      "loss": 0.6734,
+      "step": 1103
+    },
+    {
+      "epoch": 1.3239352129574085,
+      "grad_norm": 0.8621672410348915,
+      "learning_rate": 6.851138241537961e-06,
+      "loss": 0.6446,
+      "step": 1104
+    },
+    {
+      "epoch": 1.325134973005399,
+      "grad_norm": 0.6048231145547539,
+      "learning_rate": 6.8446483219870855e-06,
+      "loss": 0.6603,
+      "step": 1105
+    },
+    {
+      "epoch": 1.3263347330533892,
+      "grad_norm": 0.8589489889623938,
+      "learning_rate": 6.838154803001131e-06,
+      "loss": 0.7011,
+      "step": 1106
+    },
+    {
+      "epoch": 1.3275344931013797,
+      "grad_norm": 0.6661232123237989,
+      "learning_rate": 6.831657697250801e-06,
+      "loss": 0.637,
+      "step": 1107
+    },
+    {
+      "epoch": 1.3287342531493702,
+      "grad_norm": 0.6147162030404598,
+      "learning_rate": 6.825157017413808e-06,
+      "loss": 0.5849,
+      "step": 1108
+    },
+    {
+      "epoch": 1.3299340131973605,
+      "grad_norm": 0.7662173418303564,
+      "learning_rate": 6.818652776174828e-06,
+      "loss": 0.7015,
+      "step": 1109
+    },
+    {
+      "epoch": 1.331133773245351,
+      "grad_norm": 0.45937740526986315,
+      "learning_rate": 6.812144986225493e-06,
+      "loss": 0.6484,
+      "step": 1110
+    },
+    {
+      "epoch": 1.3323335332933413,
+      "grad_norm": 0.5298575907336551,
+      "learning_rate": 6.805633660264358e-06,
+      "loss": 0.6436,
+      "step": 1111
+    },
+    {
+      "epoch": 1.3335332933413317,
+      "grad_norm": 0.7406696355689192,
+      "learning_rate": 6.7991188109968766e-06,
+      "loss": 0.671,
+      "step": 1112
+    },
+    {
+      "epoch": 1.3347330533893222,
+      "grad_norm": 0.6034998910180078,
+      "learning_rate": 6.792600451135378e-06,
+      "loss": 0.5948,
+      "step": 1113
+    },
+    {
+      "epoch": 1.3359328134373125,
+      "grad_norm": 0.6245548231844243,
+      "learning_rate": 6.786078593399041e-06,
+      "loss": 0.6184,
+      "step": 1114
+    },
+    {
+      "epoch": 1.337132573485303,
+      "grad_norm": 0.7830690088927433,
+      "learning_rate": 6.779553250513875e-06,
+      "loss": 0.6353,
+      "step": 1115
+    },
+    {
+      "epoch": 1.3383323335332933,
+      "grad_norm": 0.5056698958171423,
+      "learning_rate": 6.773024435212678e-06,
+      "loss": 0.6272,
+      "step": 1116
+    },
+    {
+      "epoch": 1.3395320935812838,
+      "grad_norm": 0.8131497915203822,
+      "learning_rate": 6.766492160235037e-06,
+      "loss": 0.5875,
+      "step": 1117
+    },
+    {
+      "epoch": 1.3407318536292743,
+      "grad_norm": 0.7013985136781519,
+      "learning_rate": 6.759956438327283e-06,
+      "loss": 0.6156,
+      "step": 1118
+    },
+    {
+      "epoch": 1.3419316136772645,
+      "grad_norm": 0.5893473658594538,
+      "learning_rate": 6.75341728224247e-06,
+      "loss": 0.6636,
+      "step": 1119
+    },
+    {
+      "epoch": 1.3431313737252548,
+      "grad_norm": 0.7317031927045132,
+      "learning_rate": 6.746874704740363e-06,
+      "loss": 0.6148,
+      "step": 1120
+    },
+    {
+      "epoch": 1.3443311337732453,
+      "grad_norm": 0.8112527565928205,
+      "learning_rate": 6.74032871858739e-06,
+      "loss": 0.644,
+      "step": 1121
+    },
+    {
+      "epoch": 1.3455308938212358,
+      "grad_norm": 0.6641460501561001,
+      "learning_rate": 6.733779336556643e-06,
+      "loss": 0.5878,
+      "step": 1122
+    },
+    {
+      "epoch": 1.346730653869226,
+      "grad_norm": 0.7381532106113696,
+      "learning_rate": 6.727226571427832e-06,
+      "loss": 0.6282,
+      "step": 1123
+    },
+    {
+      "epoch": 1.3479304139172166,
+      "grad_norm": 0.6770212013493809,
+      "learning_rate": 6.7206704359872706e-06,
+      "loss": 0.6388,
+      "step": 1124
+    },
+    {
+      "epoch": 1.3491301739652068,
+      "grad_norm": 0.5750273845268145,
+      "learning_rate": 6.714110943027853e-06,
+      "loss": 0.556,
+      "step": 1125
+    },
+    {
+      "epoch": 1.3503299340131973,
+      "grad_norm": 0.9108358628622016,
+      "learning_rate": 6.707548105349016e-06,
+      "loss": 0.6073,
+      "step": 1126
+    },
+    {
+      "epoch": 1.3515296940611878,
+      "grad_norm": 0.7322466065246941,
+      "learning_rate": 6.700981935756732e-06,
+      "loss": 0.5416,
+      "step": 1127
+    },
+    {
+      "epoch": 1.352729454109178,
+      "grad_norm": 0.6079719258021722,
+      "learning_rate": 6.694412447063467e-06,
+      "loss": 0.6685,
+      "step": 1128
+    },
+    {
+      "epoch": 1.3539292141571686,
+      "grad_norm": 0.8370293574026989,
+      "learning_rate": 6.68783965208817e-06,
+      "loss": 0.6738,
+      "step": 1129
+    },
+    {
+      "epoch": 1.3551289742051589,
+      "grad_norm": 0.565962142703862,
+      "learning_rate": 6.68126356365624e-06,
+      "loss": 0.58,
+      "step": 1130
+    },
+    {
+      "epoch": 1.3563287342531494,
+      "grad_norm": 0.6388448863130164,
+      "learning_rate": 6.6746841945995e-06,
+      "loss": 0.6354,
+      "step": 1131
+    },
+    {
+      "epoch": 1.3575284943011399,
+      "grad_norm": 0.6653541085891204,
+      "learning_rate": 6.668101557756176e-06,
+      "loss": 0.6083,
+      "step": 1132
+    },
+    {
+      "epoch": 1.3587282543491301,
+      "grad_norm": 0.7599696151343085,
+      "learning_rate": 6.661515665970869e-06,
+      "loss": 0.6793,
+      "step": 1133
+    },
+    {
+      "epoch": 1.3599280143971206,
+      "grad_norm": 0.6741621620992139,
+      "learning_rate": 6.6549265320945345e-06,
+      "loss": 0.5406,
+      "step": 1134
+    },
+    {
+      "epoch": 1.361127774445111,
+      "grad_norm": 0.6832030336773401,
+      "learning_rate": 6.648334168984452e-06,
+      "loss": 0.6509,
+      "step": 1135
+    },
+    {
+      "epoch": 1.3623275344931014,
+      "grad_norm": 0.7042153457501232,
+      "learning_rate": 6.641738589504202e-06,
+      "loss": 0.6252,
+      "step": 1136
+    },
+    {
+      "epoch": 1.363527294541092,
+      "grad_norm": 0.7129225998390416,
+      "learning_rate": 6.635139806523642e-06,
+      "loss": 0.6705,
+      "step": 1137
+    },
+    {
+      "epoch": 1.3647270545890822,
+      "grad_norm": 0.6633480074567827,
+      "learning_rate": 6.6285378329188784e-06,
+      "loss": 0.6667,
+      "step": 1138
+    },
+    {
+      "epoch": 1.3659268146370727,
+      "grad_norm": 0.7298944730666421,
+      "learning_rate": 6.621932681572245e-06,
+      "loss": 0.6684,
+      "step": 1139
+    },
+    {
+      "epoch": 1.367126574685063,
+      "grad_norm": 0.7535335352629341,
+      "learning_rate": 6.615324365372281e-06,
+      "loss": 0.6122,
+      "step": 1140
+    },
+    {
+      "epoch": 1.3683263347330534,
+      "grad_norm": 0.6883059661899033,
+      "learning_rate": 6.608712897213691e-06,
+      "loss": 0.6674,
+      "step": 1141
+    },
+    {
+      "epoch": 1.369526094781044,
+      "grad_norm": 0.7904357771826949,
+      "learning_rate": 6.602098289997336e-06,
+      "loss": 0.6171,
+      "step": 1142
+    },
+    {
+      "epoch": 1.3707258548290342,
+      "grad_norm": 0.7843313126783715,
+      "learning_rate": 6.5954805566302035e-06,
+      "loss": 0.6219,
+      "step": 1143
+    },
+    {
+      "epoch": 1.3719256148770245,
+      "grad_norm": 0.8312569687830124,
+      "learning_rate": 6.588859710025378e-06,
+      "loss": 0.61,
+      "step": 1144
+    },
+    {
+      "epoch": 1.373125374925015,
+      "grad_norm": 0.5669030307035022,
+      "learning_rate": 6.5822357631020205e-06,
+      "loss": 0.6347,
+      "step": 1145
+    },
+    {
+      "epoch": 1.3743251349730055,
+      "grad_norm": 0.6927951715863442,
+      "learning_rate": 6.575608728785341e-06,
+      "loss": 0.6055,
+      "step": 1146
+    },
+    {
+      "epoch": 1.3755248950209957,
+      "grad_norm": 0.7869796862407736,
+      "learning_rate": 6.568978620006574e-06,
+      "loss": 0.68,
+      "step": 1147
+    },
+    {
+      "epoch": 1.3767246550689862,
+      "grad_norm": 0.6329606513199126,
+      "learning_rate": 6.562345449702952e-06,
+      "loss": 0.6729,
+      "step": 1148
+    },
+    {
+      "epoch": 1.3779244151169765,
+      "grad_norm": 0.7470692094435768,
+      "learning_rate": 6.555709230817683e-06,
+      "loss": 0.6922,
+      "step": 1149
+    },
+    {
+      "epoch": 1.379124175164967,
+      "grad_norm": 0.6397194190507423,
+      "learning_rate": 6.5490699762999226e-06,
+      "loss": 0.6609,
+      "step": 1150
+    },
+    {
+      "epoch": 1.3803239352129575,
+      "grad_norm": 0.7495978599469907,
+      "learning_rate": 6.54242769910475e-06,
+      "loss": 0.6081,
+      "step": 1151
+    },
+    {
+      "epoch": 1.3815236952609478,
+      "grad_norm": 0.7735457645872734,
+      "learning_rate": 6.535782412193146e-06,
+      "loss": 0.6591,
+      "step": 1152
+    },
+    {
+      "epoch": 1.3827234553089383,
+      "grad_norm": 0.6756589768594755,
+      "learning_rate": 6.529134128531956e-06,
+      "loss": 0.6439,
+      "step": 1153
+    },
+    {
+      "epoch": 1.3839232153569285,
+      "grad_norm": 0.7556235859296331,
+      "learning_rate": 6.522482861093882e-06,
+      "loss": 0.6511,
+      "step": 1154
+    },
+    {
+      "epoch": 1.385122975404919,
+      "grad_norm": 0.8105361367882571,
+      "learning_rate": 6.515828622857443e-06,
+      "loss": 0.5748,
+      "step": 1155
+    },
+    {
+      "epoch": 1.3863227354529095,
+      "grad_norm": 0.5900221272166272,
+      "learning_rate": 6.509171426806954e-06,
+      "loss": 0.6394,
+      "step": 1156
+    },
+    {
+      "epoch": 1.3875224955008998,
+      "grad_norm": 0.49277049926647637,
+      "learning_rate": 6.502511285932507e-06,
+      "loss": 0.6577,
+      "step": 1157
+    },
+    {
+      "epoch": 1.3887222555488903,
+      "grad_norm": 0.865199565640867,
+      "learning_rate": 6.495848213229933e-06,
+      "loss": 0.6189,
+      "step": 1158
+    },
+    {
+      "epoch": 1.3899220155968806,
+      "grad_norm": 0.6212103321219965,
+      "learning_rate": 6.4891822217007904e-06,
+      "loss": 0.6561,
+      "step": 1159
+    },
+    {
+      "epoch": 1.391121775644871,
+      "grad_norm": 0.650567408341297,
+      "learning_rate": 6.482513324352328e-06,
+      "loss": 0.6505,
+      "step": 1160
+    },
+    {
+      "epoch": 1.3923215356928615,
+      "grad_norm": 0.8464647346033654,
+      "learning_rate": 6.4758415341974705e-06,
+      "loss": 0.6155,
+      "step": 1161
+    },
+    {
+      "epoch": 1.3935212957408518,
+      "grad_norm": 0.6799298795578801,
+      "learning_rate": 6.469166864254779e-06,
+      "loss": 0.6417,
+      "step": 1162
+    },
+    {
+      "epoch": 1.3947210557888423,
+      "grad_norm": 0.5391827736330186,
+      "learning_rate": 6.462489327548442e-06,
+      "loss": 0.6379,
+      "step": 1163
+    },
+    {
+      "epoch": 1.3959208158368326,
+      "grad_norm": 0.7018397816358515,
+      "learning_rate": 6.455808937108237e-06,
+      "loss": 0.7058,
+      "step": 1164
+    },
+    {
+      "epoch": 1.397120575884823,
+      "grad_norm": 0.840076044364674,
+      "learning_rate": 6.4491257059695116e-06,
+      "loss": 0.667,
+      "step": 1165
+    },
+    {
+      "epoch": 1.3983203359328136,
+      "grad_norm": 0.6515030138359833,
+      "learning_rate": 6.442439647173155e-06,
+      "loss": 0.5647,
+      "step": 1166
+    },
+    {
+      "epoch": 1.3995200959808038,
+      "grad_norm": 0.5717150990683747,
+      "learning_rate": 6.435750773765579e-06,
+      "loss": 0.6355,
+      "step": 1167
+    },
+    {
+      "epoch": 1.4007198560287941,
+      "grad_norm": 0.7895379356993755,
+      "learning_rate": 6.429059098798679e-06,
+      "loss": 0.6481,
+      "step": 1168
+    },
+    {
+      "epoch": 1.4019196160767846,
+      "grad_norm": 0.698749126512253,
+      "learning_rate": 6.4223646353298264e-06,
+      "loss": 0.6304,
+      "step": 1169
+    },
+    {
+      "epoch": 1.403119376124775,
+      "grad_norm": 0.6577509725302786,
+      "learning_rate": 6.415667396421825e-06,
+      "loss": 0.6315,
+      "step": 1170
+    },
+    {
+      "epoch": 1.4043191361727654,
+      "grad_norm": 0.6922907089938489,
+      "learning_rate": 6.4089673951429e-06,
+      "loss": 0.6636,
+      "step": 1171
+    },
+    {
+      "epoch": 1.4055188962207559,
+      "grad_norm": 0.617477414124189,
+      "learning_rate": 6.402264644566665e-06,
+      "loss": 0.6193,
+      "step": 1172
+    },
+    {
+      "epoch": 1.4067186562687461,
+      "grad_norm": 0.6576230106720494,
+      "learning_rate": 6.395559157772098e-06,
+      "loss": 0.6295,
+      "step": 1173
+    },
+    {
+      "epoch": 1.4079184163167366,
+      "grad_norm": 0.6715729203526053,
+      "learning_rate": 6.388850947843517e-06,
+      "loss": 0.6243,
+      "step": 1174
+    },
+    {
+      "epoch": 1.4091181763647271,
+      "grad_norm": 0.7679879928940876,
+      "learning_rate": 6.382140027870553e-06,
+      "loss": 0.6648,
+      "step": 1175
+    },
+    {
+      "epoch": 1.4103179364127174,
+      "grad_norm": 0.8037472260249561,
+      "learning_rate": 6.375426410948128e-06,
+      "loss": 0.6517,
+      "step": 1176
+    },
+    {
+      "epoch": 1.411517696460708,
+      "grad_norm": 0.6349627119181712,
+      "learning_rate": 6.3687101101764185e-06,
+      "loss": 0.676,
+      "step": 1177
+    },
+    {
+      "epoch": 1.4127174565086982,
+      "grad_norm": 0.6801899051106781,
+      "learning_rate": 6.361991138660846e-06,
+      "loss": 0.6266,
+      "step": 1178
+    },
+    {
+      "epoch": 1.4139172165566887,
+      "grad_norm": 0.5350755164375657,
+      "learning_rate": 6.355269509512042e-06,
+      "loss": 0.6375,
+      "step": 1179
+    },
+    {
+      "epoch": 1.4151169766046792,
+      "grad_norm": 0.7239692027078118,
+      "learning_rate": 6.3485452358458185e-06,
+      "loss": 0.6139,
+      "step": 1180
+    },
+    {
+      "epoch": 1.4163167366526694,
+      "grad_norm": 0.695469325517881,
+      "learning_rate": 6.341818330783156e-06,
+      "loss": 0.625,
+      "step": 1181
+    },
+    {
+      "epoch": 1.41751649670066,
+      "grad_norm": 0.4780893249383196,
+      "learning_rate": 6.335088807450163e-06,
+      "loss": 0.7342,
+      "step": 1182
+    },
+    {
+      "epoch": 1.4187162567486502,
+      "grad_norm": 0.927274137868549,
+      "learning_rate": 6.328356678978058e-06,
+      "loss": 0.5985,
+      "step": 1183
+    },
+    {
+      "epoch": 1.4199160167966407,
+      "grad_norm": 0.72395308753094,
+      "learning_rate": 6.321621958503146e-06,
+      "loss": 0.6434,
+      "step": 1184
+    },
+    {
+      "epoch": 1.4211157768446312,
+      "grad_norm": 0.8789638735851689,
+      "learning_rate": 6.3148846591667865e-06,
+      "loss": 0.6734,
+      "step": 1185
+    },
+    {
+      "epoch": 1.4223155368926215,
+      "grad_norm": 0.5581994386487651,
+      "learning_rate": 6.308144794115374e-06,
+      "loss": 0.6156,
+      "step": 1186
+    },
+    {
+      "epoch": 1.4235152969406117,
+      "grad_norm": 0.6272793758583736,
+      "learning_rate": 6.301402376500306e-06,
+      "loss": 0.6515,
+      "step": 1187
+    },
+    {
+      "epoch": 1.4247150569886022,
+      "grad_norm": 0.6447701993231002,
+      "learning_rate": 6.2946574194779645e-06,
+      "loss": 0.6583,
+      "step": 1188
+    },
+    {
+      "epoch": 1.4259148170365927,
+      "grad_norm": 0.6209843693848559,
+      "learning_rate": 6.287909936209682e-06,
+      "loss": 0.6319,
+      "step": 1189
+    },
+    {
+      "epoch": 1.427114577084583,
+      "grad_norm": 0.6912573791199007,
+      "learning_rate": 6.281159939861725e-06,
+      "loss": 0.643,
+      "step": 1190
+    },
+    {
+      "epoch": 1.4283143371325735,
+      "grad_norm": 0.6415838122000773,
+      "learning_rate": 6.274407443605264e-06,
+      "loss": 0.6287,
+      "step": 1191
+    },
+    {
+      "epoch": 1.4295140971805638,
+      "grad_norm": 0.7097184462078711,
+      "learning_rate": 6.26765246061634e-06,
+      "loss": 0.6561,
+      "step": 1192
+    },
+    {
+      "epoch": 1.4307138572285543,
+      "grad_norm": 0.7213328291617336,
+      "learning_rate": 6.2608950040758575e-06,
+      "loss": 0.6163,
+      "step": 1193
+    },
+    {
+      "epoch": 1.4319136172765448,
+      "grad_norm": 0.6342851785100996,
+      "learning_rate": 6.254135087169537e-06,
+      "loss": 0.6688,
+      "step": 1194
+    },
+    {
+      "epoch": 1.433113377324535,
+      "grad_norm": 0.6609542753691291,
+      "learning_rate": 6.2473727230879076e-06,
+      "loss": 0.6745,
+      "step": 1195
+    },
+    {
+      "epoch": 1.4343131373725255,
+      "grad_norm": 0.7165029004297052,
+      "learning_rate": 6.240607925026271e-06,
+      "loss": 0.6552,
+      "step": 1196
+    },
+    {
+      "epoch": 1.4355128974205158,
+      "grad_norm": 0.65710539720105,
+      "learning_rate": 6.2338407061846765e-06,
+      "loss": 0.5844,
+      "step": 1197
+    },
+    {
+      "epoch": 1.4367126574685063,
+      "grad_norm": 0.6067918304499893,
+      "learning_rate": 6.227071079767899e-06,
+      "loss": 0.6143,
+      "step": 1198
+    },
+    {
+      "epoch": 1.4379124175164968,
+      "grad_norm": 0.6505049285655436,
+      "learning_rate": 6.2202990589854095e-06,
+      "loss": 0.5458,
+      "step": 1199
+    },
+    {
+      "epoch": 1.439112177564487,
+      "grad_norm": 0.7325345130366029,
+      "learning_rate": 6.213524657051354e-06,
+      "loss": 0.6436,
+      "step": 1200
+    },
+    {
+      "epoch": 1.4403119376124776,
+      "grad_norm": 0.589110413780781,
+      "learning_rate": 6.2067478871845236e-06,
+      "loss": 0.6423,
+      "step": 1201
+    },
+    {
+      "epoch": 1.4415116976604678,
+      "grad_norm": 0.7616775529555003,
+      "learning_rate": 6.199968762608326e-06,
+      "loss": 0.6762,
+      "step": 1202
+    },
+    {
+      "epoch": 1.4427114577084583,
+      "grad_norm": 0.5459706578178842,
+      "learning_rate": 6.193187296550771e-06,
+      "loss": 0.6944,
+      "step": 1203
+    },
+    {
+      "epoch": 1.4439112177564488,
+      "grad_norm": 0.622979747002905,
+      "learning_rate": 6.1864035022444325e-06,
+      "loss": 0.6007,
+      "step": 1204
+    },
+    {
+      "epoch": 1.445110977804439,
+      "grad_norm": 0.7495542665123462,
+      "learning_rate": 6.179617392926427e-06,
+      "loss": 0.6425,
+      "step": 1205
+    },
+    {
+      "epoch": 1.4463107378524296,
+      "grad_norm": 0.8002049020027295,
+      "learning_rate": 6.17282898183839e-06,
+      "loss": 0.5747,
+      "step": 1206
+    },
+    {
+      "epoch": 1.4475104979004199,
+      "grad_norm": 0.49958690427656804,
+      "learning_rate": 6.166038282226449e-06,
+      "loss": 0.6046,
+      "step": 1207
+    },
+    {
+      "epoch": 1.4487102579484104,
+      "grad_norm": 0.7221357839193259,
+      "learning_rate": 6.159245307341195e-06,
+      "loss": 0.6211,
+      "step": 1208
+    },
+    {
+      "epoch": 1.4499100179964008,
+      "grad_norm": 0.8801610973062568,
+      "learning_rate": 6.152450070437659e-06,
+      "loss": 0.6477,
+      "step": 1209
+    },
+    {
+      "epoch": 1.4511097780443911,
+      "grad_norm": 0.6659853219207728,
+      "learning_rate": 6.14565258477529e-06,
+      "loss": 0.6441,
+      "step": 1210
+    },
+    {
+      "epoch": 1.4523095380923814,
+      "grad_norm": 0.6760897828611526,
+      "learning_rate": 6.138852863617918e-06,
+      "loss": 0.569,
+      "step": 1211
+    },
+    {
+      "epoch": 1.4535092981403719,
+      "grad_norm": 0.6098179229000052,
+      "learning_rate": 6.132050920233739e-06,
+      "loss": 0.6635,
+      "step": 1212
+    },
+    {
+      "epoch": 1.4547090581883624,
+      "grad_norm": 0.813858537026656,
+      "learning_rate": 6.125246767895287e-06,
+      "loss": 0.6488,
+      "step": 1213
+    },
+    {
+      "epoch": 1.4559088182363527,
+      "grad_norm": 0.5711502698398734,
+      "learning_rate": 6.118440419879402e-06,
+      "loss": 0.6072,
+      "step": 1214
+    },
+    {
+      "epoch": 1.4571085782843431,
+      "grad_norm": 0.6667021617430856,
+      "learning_rate": 6.111631889467213e-06,
+      "loss": 0.6966,
+      "step": 1215
+    },
+    {
+      "epoch": 1.4583083383323334,
+      "grad_norm": 0.7557909481255707,
+      "learning_rate": 6.104821189944102e-06,
+      "loss": 0.67,
+      "step": 1216
+    },
+    {
+      "epoch": 1.459508098380324,
+      "grad_norm": 0.5640947675429846,
+      "learning_rate": 6.098008334599689e-06,
+      "loss": 0.6869,
+      "step": 1217
+    },
+    {
+      "epoch": 1.4607078584283144,
+      "grad_norm": 0.7139040670249647,
+      "learning_rate": 6.0911933367277984e-06,
+      "loss": 0.6422,
+      "step": 1218
+    },
+    {
+      "epoch": 1.4619076184763047,
+      "grad_norm": 0.6342341991664133,
+      "learning_rate": 6.084376209626433e-06,
+      "loss": 0.667,
+      "step": 1219
+    },
+    {
+      "epoch": 1.4631073785242952,
+      "grad_norm": 0.7585758162626316,
+      "learning_rate": 6.077556966597753e-06,
+      "loss": 0.6154,
+      "step": 1220
+    },
+    {
+      "epoch": 1.4643071385722854,
+      "grad_norm": 0.5965885737089633,
+      "learning_rate": 6.070735620948049e-06,
+      "loss": 0.6181,
+      "step": 1221
+    },
+    {
+      "epoch": 1.465506898620276,
+      "grad_norm": 0.7167205284378334,
+      "learning_rate": 6.063912185987708e-06,
+      "loss": 0.7338,
+      "step": 1222
+    },
+    {
+      "epoch": 1.4667066586682664,
+      "grad_norm": 0.6249955365345098,
+      "learning_rate": 6.057086675031202e-06,
+      "loss": 0.6121,
+      "step": 1223
+    },
+    {
+      "epoch": 1.4679064187162567,
+      "grad_norm": 0.7612387736488212,
+      "learning_rate": 6.050259101397048e-06,
+      "loss": 0.6384,
+      "step": 1224
+    },
+    {
+      "epoch": 1.4691061787642472,
+      "grad_norm": 0.6565025743667317,
+      "learning_rate": 6.0434294784077905e-06,
+      "loss": 0.6335,
+      "step": 1225
+    },
+    {
+      "epoch": 1.4703059388122375,
+      "grad_norm": 0.5840172467296441,
+      "learning_rate": 6.036597819389972e-06,
+      "loss": 0.5546,
+      "step": 1226
+    },
+    {
+      "epoch": 1.471505698860228,
+      "grad_norm": 0.8134562582624975,
+      "learning_rate": 6.02976413767411e-06,
+      "loss": 0.6825,
+      "step": 1227
+    },
+    {
+      "epoch": 1.4727054589082185,
+      "grad_norm": 0.7013684567318601,
+      "learning_rate": 6.022928446594662e-06,
+      "loss": 0.6696,
+      "step": 1228
+    },
+    {
+      "epoch": 1.4739052189562087,
+      "grad_norm": 0.476068932069486,
+      "learning_rate": 6.016090759490014e-06,
+      "loss": 0.6119,
+      "step": 1229
+    },
+    {
+      "epoch": 1.4751049790041992,
+      "grad_norm": 0.7243788321830064,
+      "learning_rate": 6.009251089702447e-06,
+      "loss": 0.6987,
+      "step": 1230
+    },
+    {
+      "epoch": 1.4763047390521895,
+      "grad_norm": 0.7913274804741819,
+      "learning_rate": 6.002409450578104e-06,
+      "loss": 0.6777,
+      "step": 1231
+    },
+    {
+      "epoch": 1.47750449910018,
+      "grad_norm": 0.6735369529497145,
+      "learning_rate": 5.995565855466974e-06,
+      "loss": 0.6325,
+      "step": 1232
+    },
+    {
+      "epoch": 1.4787042591481705,
+      "grad_norm": 0.5554440981407642,
+      "learning_rate": 5.988720317722866e-06,
+      "loss": 0.6515,
+      "step": 1233
+    },
+    {
+      "epoch": 1.4799040191961608,
+      "grad_norm": 0.6834658022330459,
+      "learning_rate": 5.981872850703376e-06,
+      "loss": 0.5841,
+      "step": 1234
+    },
+    {
+      "epoch": 1.481103779244151,
+      "grad_norm": 0.6933068366611141,
+      "learning_rate": 5.975023467769865e-06,
+      "loss": 0.6415,
+      "step": 1235
+    },
+    {
+      "epoch": 1.4823035392921415,
+      "grad_norm": 0.7325088249095414,
+      "learning_rate": 5.968172182287433e-06,
+      "loss": 0.5846,
+      "step": 1236
+    },
+    {
+      "epoch": 1.483503299340132,
+      "grad_norm": 0.7036937352178297,
+      "learning_rate": 5.9613190076248935e-06,
+      "loss": 0.6728,
+      "step": 1237
+    },
+    {
+      "epoch": 1.4847030593881223,
+      "grad_norm": 0.6323787187969813,
+      "learning_rate": 5.954463957154743e-06,
+      "loss": 0.5662,
+      "step": 1238
+    },
+    {
+      "epoch": 1.4859028194361128,
+      "grad_norm": 0.737462376786475,
+      "learning_rate": 5.947607044253142e-06,
+      "loss": 0.6503,
+      "step": 1239
+    },
+    {
+      "epoch": 1.487102579484103,
+      "grad_norm": 0.6486858074761898,
+      "learning_rate": 5.940748282299885e-06,
+      "loss": 0.5894,
+      "step": 1240
+    },
+    {
+      "epoch": 1.4883023395320936,
+      "grad_norm": 0.6100466907434766,
+      "learning_rate": 5.933887684678369e-06,
+      "loss": 0.5828,
+      "step": 1241
+    },
+    {
+      "epoch": 1.489502099580084,
+      "grad_norm": 0.8409578952042881,
+      "learning_rate": 5.927025264775581e-06,
+      "loss": 0.589,
+      "step": 1242
+    },
+    {
+      "epoch": 1.4907018596280743,
+      "grad_norm": 0.582264752223249,
+      "learning_rate": 5.920161035982058e-06,
+      "loss": 0.6669,
+      "step": 1243
+    },
+    {
+      "epoch": 1.4919016196760648,
+      "grad_norm": 0.6079716398734951,
+      "learning_rate": 5.913295011691869e-06,
+      "loss": 0.6101,
+      "step": 1244
+    },
+    {
+      "epoch": 1.493101379724055,
+      "grad_norm": 0.7467107653736639,
+      "learning_rate": 5.906427205302584e-06,
+      "loss": 0.6203,
+      "step": 1245
+    },
+    {
+      "epoch": 1.4943011397720456,
+      "grad_norm": 0.623686442929491,
+      "learning_rate": 5.899557630215256e-06,
+      "loss": 0.624,
+      "step": 1246
+    },
+    {
+      "epoch": 1.495500899820036,
+      "grad_norm": 0.6732483651744011,
+      "learning_rate": 5.892686299834382e-06,
+      "loss": 0.6519,
+      "step": 1247
+    },
+    {
+      "epoch": 1.4967006598680264,
+      "grad_norm": 0.8058906996878299,
+      "learning_rate": 5.88581322756789e-06,
+      "loss": 0.6071,
+      "step": 1248
+    },
+    {
+      "epoch": 1.4979004199160169,
+      "grad_norm": 0.5320089904676821,
+      "learning_rate": 5.878938426827106e-06,
+      "loss": 0.648,
+      "step": 1249
+    },
+    {
+      "epoch": 1.4991001799640071,
+      "grad_norm": 0.7532237642342396,
+      "learning_rate": 5.872061911026723e-06,
+      "loss": 0.5871,
+      "step": 1250
+    },
+    {
+      "epoch": 1.5002999400119976,
+      "grad_norm": 0.6431714703702812,
+      "learning_rate": 5.865183693584786e-06,
+      "loss": 0.6295,
+      "step": 1251
+    },
+    {
+      "epoch": 1.5014997000599881,
+      "grad_norm": 0.6235215837163856,
+      "learning_rate": 5.858303787922663e-06,
+      "loss": 0.7042,
+      "step": 1252
+    },
+    {
+      "epoch": 1.5026994601079784,
+      "grad_norm": 0.7383379478172799,
+      "learning_rate": 5.851422207465009e-06,
+      "loss": 0.6177,
+      "step": 1253
+    },
+    {
+      "epoch": 1.5038992201559687,
+      "grad_norm": 0.6367437664682997,
+      "learning_rate": 5.8445389656397525e-06,
+      "loss": 0.7393,
+      "step": 1254
+    },
+    {
+      "epoch": 1.5050989802039592,
+      "grad_norm": 0.6840499847058636,
+      "learning_rate": 5.83765407587806e-06,
+      "loss": 0.6664,
+      "step": 1255
+    },
+    {
+      "epoch": 1.5062987402519497,
+      "grad_norm": 0.7082182417706232,
+      "learning_rate": 5.830767551614316e-06,
+      "loss": 0.6624,
+      "step": 1256
+    },
+    {
+      "epoch": 1.5074985002999401,
+      "grad_norm": 0.7448387921989272,
+      "learning_rate": 5.823879406286095e-06,
+      "loss": 0.6458,
+      "step": 1257
+    },
+    {
+      "epoch": 1.5086982603479304,
+      "grad_norm": 0.6314122065500002,
+      "learning_rate": 5.816989653334131e-06,
+      "loss": 0.6356,
+      "step": 1258
+    },
+    {
+      "epoch": 1.5098980203959207,
+      "grad_norm": 0.7077408943810496,
+      "learning_rate": 5.810098306202301e-06,
+      "loss": 0.6376,
+      "step": 1259
+    },
+    {
+      "epoch": 1.5110977804439112,
+      "grad_norm": 0.5255442417829472,
+      "learning_rate": 5.803205378337586e-06,
+      "loss": 0.6054,
+      "step": 1260
+    },
+    {
+      "epoch": 1.5122975404919017,
+      "grad_norm": 0.7423020316453411,
+      "learning_rate": 5.796310883190055e-06,
+      "loss": 0.5346,
+      "step": 1261
+    },
+    {
+      "epoch": 1.5134973005398922,
+      "grad_norm": 0.72373854141275,
+      "learning_rate": 5.789414834212837e-06,
+      "loss": 0.6513,
+      "step": 1262
+    },
+    {
+      "epoch": 1.5146970605878824,
+      "grad_norm": 0.6268195313504005,
+      "learning_rate": 5.782517244862087e-06,
+      "loss": 0.6355,
+      "step": 1263
+    },
+    {
+      "epoch": 1.5158968206358727,
+      "grad_norm": 0.7470521652894679,
+      "learning_rate": 5.775618128596972e-06,
+      "loss": 0.6039,
+      "step": 1264
+    },
+    {
+      "epoch": 1.5170965806838632,
+      "grad_norm": 0.581894195138229,
+      "learning_rate": 5.768717498879635e-06,
+      "loss": 0.6927,
+      "step": 1265
+    },
+    {
+      "epoch": 1.5182963407318537,
+      "grad_norm": 0.641456779328077,
+      "learning_rate": 5.761815369175175e-06,
+      "loss": 0.6586,
+      "step": 1266
+    },
+    {
+      "epoch": 1.519496100779844,
+      "grad_norm": 0.5650571897087318,
+      "learning_rate": 5.754911752951612e-06,
+      "loss": 0.5913,
+      "step": 1267
+    },
+    {
+      "epoch": 1.5206958608278345,
+      "grad_norm": 0.7021394162807891,
+      "learning_rate": 5.7480066636798735e-06,
+      "loss": 0.6405,
+      "step": 1268
+    },
+    {
+      "epoch": 1.5218956208758248,
+      "grad_norm": 0.7911515180750498,
+      "learning_rate": 5.74110011483376e-06,
+      "loss": 0.5796,
+      "step": 1269
+    },
+    {
+      "epoch": 1.5230953809238152,
+      "grad_norm": 0.7166122110519555,
+      "learning_rate": 5.734192119889913e-06,
+      "loss": 0.6079,
+      "step": 1270
+    },
+    {
+      "epoch": 1.5242951409718057,
+      "grad_norm": 0.5612041476172737,
+      "learning_rate": 5.7272826923278065e-06,
+      "loss": 0.6609,
+      "step": 1271
+    },
+    {
+      "epoch": 1.525494901019796,
+      "grad_norm": 0.6213354163393838,
+      "learning_rate": 5.720371845629703e-06,
+      "loss": 0.6125,
+      "step": 1272
+    },
+    {
+      "epoch": 1.5266946610677863,
+      "grad_norm": 0.6474864927657513,
+      "learning_rate": 5.713459593280635e-06,
+      "loss": 0.6444,
+      "step": 1273
+    },
+    {
+      "epoch": 1.5278944211157768,
+      "grad_norm": 0.5908040200841372,
+      "learning_rate": 5.706545948768378e-06,
+      "loss": 0.5962,
+      "step": 1274
+    },
+    {
+      "epoch": 1.5290941811637673,
+      "grad_norm": 0.6227188076704627,
+      "learning_rate": 5.699630925583426e-06,
+      "loss": 0.6565,
+      "step": 1275
+    },
+    {
+      "epoch": 1.5302939412117578,
+      "grad_norm": 0.7599470245412667,
+      "learning_rate": 5.692714537218963e-06,
+      "loss": 0.6916,
+      "step": 1276
+    },
+    {
+      "epoch": 1.531493701259748,
+      "grad_norm": 0.7502609001812292,
+      "learning_rate": 5.685796797170831e-06,
+      "loss": 0.6004,
+      "step": 1277
+    },
+    {
+      "epoch": 1.5326934613077383,
+      "grad_norm": 0.5411325249265353,
+      "learning_rate": 5.67887771893752e-06,
+      "loss": 0.6619,
+      "step": 1278
+    },
+    {
+      "epoch": 1.5338932213557288,
+      "grad_norm": 0.7761314801897223,
+      "learning_rate": 5.671957316020123e-06,
+      "loss": 0.6192,
+      "step": 1279
+    },
+    {
+      "epoch": 1.5350929814037193,
+      "grad_norm": 0.6862836201660087,
+      "learning_rate": 5.665035601922317e-06,
+      "loss": 0.5928,
+      "step": 1280
+    },
+    {
+      "epoch": 1.5362927414517098,
+      "grad_norm": 0.7866975019317517,
+      "learning_rate": 5.658112590150348e-06,
+      "loss": 0.6476,
+      "step": 1281
+    },
+    {
+      "epoch": 1.5374925014997,
+      "grad_norm": 0.48330894009913766,
+      "learning_rate": 5.651188294212979e-06,
+      "loss": 0.6323,
+      "step": 1282
+    },
+    {
+      "epoch": 1.5386922615476903,
+      "grad_norm": 0.8543107937867963,
+      "learning_rate": 5.644262727621491e-06,
+      "loss": 0.6185,
+      "step": 1283
+    },
+    {
+      "epoch": 1.5398920215956808,
+      "grad_norm": 0.7477891341738985,
+      "learning_rate": 5.637335903889639e-06,
+      "loss": 0.6001,
+      "step": 1284
+    },
+    {
+      "epoch": 1.5410917816436713,
+      "grad_norm": 0.525523421014255,
+      "learning_rate": 5.630407836533632e-06,
+      "loss": 0.6311,
+      "step": 1285
+    },
+    {
+      "epoch": 1.5422915416916618,
+      "grad_norm": 1.0471427028757878,
+      "learning_rate": 5.623478539072106e-06,
+      "loss": 0.5697,
+      "step": 1286
+    },
+    {
+      "epoch": 1.543491301739652,
+      "grad_norm": 0.76087236543128,
+      "learning_rate": 5.616548025026096e-06,
+      "loss": 0.6693,
+      "step": 1287
+    },
+    {
+      "epoch": 1.5446910617876424,
+      "grad_norm": 0.6268895731868466,
+      "learning_rate": 5.609616307919015e-06,
+      "loss": 0.6079,
+      "step": 1288
+    },
+    {
+      "epoch": 1.5458908218356329,
+      "grad_norm": 0.6984978410322382,
+      "learning_rate": 5.6026834012766155e-06,
+      "loss": 0.6549,
+      "step": 1289
+    },
+    {
+      "epoch": 1.5470905818836234,
+      "grad_norm": 0.5098259790983851,
+      "learning_rate": 5.59574931862698e-06,
+      "loss": 0.6849,
+      "step": 1290
+    },
+    {
+      "epoch": 1.5482903419316136,
+      "grad_norm": 0.74458734342187,
+      "learning_rate": 5.588814073500481e-06,
+      "loss": 0.6239,
+      "step": 1291
+    },
+    {
+      "epoch": 1.5494901019796041,
+      "grad_norm": 0.5624204820252244,
+      "learning_rate": 5.581877679429758e-06,
+      "loss": 0.6346,
+      "step": 1292
+    },
+    {
+      "epoch": 1.5506898620275944,
+      "grad_norm": 0.8345891432596939,
+      "learning_rate": 5.574940149949697e-06,
+      "loss": 0.6148,
+      "step": 1293
+    },
+    {
+      "epoch": 1.551889622075585,
+      "grad_norm": 0.7357972047845589,
+      "learning_rate": 5.568001498597395e-06,
+      "loss": 0.5902,
+      "step": 1294
+    },
+    {
+      "epoch": 1.5530893821235754,
+      "grad_norm": 0.6544965040507694,
+      "learning_rate": 5.561061738912143e-06,
+      "loss": 0.6137,
+      "step": 1295
+    },
+    {
+      "epoch": 1.5542891421715657,
+      "grad_norm": 0.542534811709464,
+      "learning_rate": 5.5541208844353875e-06,
+      "loss": 0.6494,
+      "step": 1296
+    },
+    {
+      "epoch": 1.555488902219556,
+      "grad_norm": 0.7380424075936614,
+      "learning_rate": 5.547178948710719e-06,
+      "loss": 0.6584,
+      "step": 1297
+    },
+    {
+      "epoch": 1.5566886622675464,
+      "grad_norm": 0.6423461323690198,
+      "learning_rate": 5.540235945283835e-06,
+      "loss": 0.6675,
+      "step": 1298
+    },
+    {
+      "epoch": 1.557888422315537,
+      "grad_norm": 0.7083491579052724,
+      "learning_rate": 5.533291887702512e-06,
+      "loss": 0.606,
+      "step": 1299
+    },
+    {
+      "epoch": 1.5590881823635274,
+      "grad_norm": 0.6406839795704589,
+      "learning_rate": 5.526346789516591e-06,
+      "loss": 0.601,
+      "step": 1300
+    },
+    {
+      "epoch": 1.5602879424115177,
+      "grad_norm": 0.6174645592082069,
+      "learning_rate": 5.519400664277936e-06,
+      "loss": 0.7061,
+      "step": 1301
+    },
+    {
+      "epoch": 1.561487702459508,
+      "grad_norm": 0.704019949782006,
+      "learning_rate": 5.512453525540421e-06,
+      "loss": 0.6713,
+      "step": 1302
+    },
+    {
+      "epoch": 1.5626874625074985,
+      "grad_norm": 0.6448062770664398,
+      "learning_rate": 5.505505386859898e-06,
+      "loss": 0.6642,
+      "step": 1303
+    },
+    {
+      "epoch": 1.563887222555489,
+      "grad_norm": 0.7065362159390188,
+      "learning_rate": 5.498556261794161e-06,
+      "loss": 0.6278,
+      "step": 1304
+    },
+    {
+      "epoch": 1.5650869826034794,
+      "grad_norm": 0.7260098687088066,
+      "learning_rate": 5.491606163902942e-06,
+      "loss": 0.5843,
+      "step": 1305
+    },
+    {
+      "epoch": 1.5662867426514697,
+      "grad_norm": 0.6334867367558583,
+      "learning_rate": 5.4846551067478595e-06,
+      "loss": 0.6244,
+      "step": 1306
+    },
+    {
+      "epoch": 1.56748650269946,
+      "grad_norm": 0.6941300206724687,
+      "learning_rate": 5.477703103892412e-06,
+      "loss": 0.6038,
+      "step": 1307
+    },
+    {
+      "epoch": 1.5686862627474505,
+      "grad_norm": 0.7633621862009159,
+      "learning_rate": 5.47075016890194e-06,
+      "loss": 0.5862,
+      "step": 1308
+    },
+    {
+      "epoch": 1.569886022795441,
+      "grad_norm": 0.9369076335853657,
+      "learning_rate": 5.4637963153436e-06,
+      "loss": 0.6297,
+      "step": 1309
+    },
+    {
+      "epoch": 1.5710857828434315,
+      "grad_norm": 0.5555133609295394,
+      "learning_rate": 5.4568415567863485e-06,
+      "loss": 0.6275,
+      "step": 1310
+    },
+    {
+      "epoch": 1.5722855428914218,
+      "grad_norm": 0.7770497074003324,
+      "learning_rate": 5.449885906800901e-06,
+      "loss": 0.6219,
+      "step": 1311
+    },
+    {
+      "epoch": 1.573485302939412,
+      "grad_norm": 0.6168582319141467,
+      "learning_rate": 5.442929378959713e-06,
+      "loss": 0.6439,
+      "step": 1312
+    },
+    {
+      "epoch": 1.5746850629874025,
+      "grad_norm": 0.6524715422067409,
+      "learning_rate": 5.435971986836958e-06,
+      "loss": 0.6573,
+      "step": 1313
+    },
+    {
+      "epoch": 1.575884823035393,
+      "grad_norm": 0.7842695350118289,
+      "learning_rate": 5.429013744008491e-06,
+      "loss": 0.6485,
+      "step": 1314
+    },
+    {
+      "epoch": 1.5770845830833833,
+      "grad_norm": 0.5362782200254684,
+      "learning_rate": 5.42205466405183e-06,
+      "loss": 0.6324,
+      "step": 1315
+    },
+    {
+      "epoch": 1.5782843431313738,
+      "grad_norm": 0.7618827490812607,
+      "learning_rate": 5.4150947605461225e-06,
+      "loss": 0.6343,
+      "step": 1316
+    },
+    {
+      "epoch": 1.579484103179364,
+      "grad_norm": 0.5253757301879358,
+      "learning_rate": 5.4081340470721286e-06,
+      "loss": 0.6574,
+      "step": 1317
+    },
+    {
+      "epoch": 1.5806838632273545,
+      "grad_norm": 0.5647718553683049,
+      "learning_rate": 5.401172537212183e-06,
+      "loss": 0.6078,
+      "step": 1318
+    },
+    {
+      "epoch": 1.581883623275345,
+      "grad_norm": 0.7617002418019216,
+      "learning_rate": 5.3942102445501795e-06,
+      "loss": 0.6249,
+      "step": 1319
+    },
+    {
+      "epoch": 1.5830833833233353,
+      "grad_norm": 0.7260704174604635,
+      "learning_rate": 5.387247182671539e-06,
+      "loss": 0.6495,
+      "step": 1320
+    },
+    {
+      "epoch": 1.5842831433713256,
+      "grad_norm": 0.7150652300708271,
+      "learning_rate": 5.380283365163175e-06,
+      "loss": 0.6716,
+      "step": 1321
+    },
+    {
+      "epoch": 1.585482903419316,
+      "grad_norm": 0.6810037087806773,
+      "learning_rate": 5.373318805613489e-06,
+      "loss": 0.6083,
+      "step": 1322
+    },
+    {
+      "epoch": 1.5866826634673066,
+      "grad_norm": 0.7478140636419214,
+      "learning_rate": 5.366353517612319e-06,
+      "loss": 0.5957,
+      "step": 1323
+    },
+    {
+      "epoch": 1.587882423515297,
+      "grad_norm": 0.7161538437371096,
+      "learning_rate": 5.35938751475093e-06,
+      "loss": 0.6425,
+      "step": 1324
+    },
+    {
+      "epoch": 1.5890821835632873,
+      "grad_norm": 0.6063804079438616,
+      "learning_rate": 5.352420810621981e-06,
+      "loss": 0.6298,
+      "step": 1325
+    },
+    {
+      "epoch": 1.5902819436112776,
+      "grad_norm": 0.5841230926203578,
+      "learning_rate": 5.345453418819499e-06,
+      "loss": 0.6152,
+      "step": 1326
+    },
+    {
+      "epoch": 1.591481703659268,
+      "grad_norm": 0.6787445465536185,
+      "learning_rate": 5.338485352938854e-06,
+      "loss": 0.5985,
+      "step": 1327
+    },
+    {
+      "epoch": 1.5926814637072586,
+      "grad_norm": 0.6924927583320081,
+      "learning_rate": 5.3315166265767274e-06,
+      "loss": 0.6117,
+      "step": 1328
+    },
+    {
+      "epoch": 1.593881223755249,
+      "grad_norm": 0.7352710603579419,
+      "learning_rate": 5.324547253331094e-06,
+      "loss": 0.5997,
+      "step": 1329
+    },
+    {
+      "epoch": 1.5950809838032394,
+      "grad_norm": 0.6977482174008609,
+      "learning_rate": 5.31757724680119e-06,
+      "loss": 0.6009,
+      "step": 1330
+    },
+    {
+      "epoch": 1.5962807438512296,
+      "grad_norm": 0.6382832332344649,
+      "learning_rate": 5.310606620587484e-06,
+      "loss": 0.5573,
+      "step": 1331
+    },
+    {
+      "epoch": 1.5974805038992201,
+      "grad_norm": 0.6971060593932868,
+      "learning_rate": 5.303635388291659e-06,
+      "loss": 0.5733,
+      "step": 1332
+    },
+    {
+      "epoch": 1.5986802639472106,
+      "grad_norm": 0.6316557545076512,
+      "learning_rate": 5.296663563516576e-06,
+      "loss": 0.6096,
+      "step": 1333
+    },
+    {
+      "epoch": 1.599880023995201,
+      "grad_norm": 0.7608501254340002,
+      "learning_rate": 5.289691159866254e-06,
+      "loss": 0.6572,
+      "step": 1334
+    },
+    {
+      "epoch": 1.6010797840431914,
+      "grad_norm": 0.8037027202981949,
+      "learning_rate": 5.282718190945842e-06,
+      "loss": 0.7289,
+      "step": 1335
+    },
+    {
+      "epoch": 1.6022795440911817,
+      "grad_norm": 0.7421398229841776,
+      "learning_rate": 5.275744670361591e-06,
+      "loss": 0.6547,
+      "step": 1336
+    },
+    {
+      "epoch": 1.6034793041391722,
+      "grad_norm": 0.6625938252607363,
+      "learning_rate": 5.268770611720831e-06,
+      "loss": 0.6152,
+      "step": 1337
+    },
+    {
+      "epoch": 1.6046790641871627,
+      "grad_norm": 0.8199782776015508,
+      "learning_rate": 5.261796028631935e-06,
+      "loss": 0.6317,
+      "step": 1338
+    },
+    {
+      "epoch": 1.605878824235153,
+      "grad_norm": 0.49806684477338437,
+      "learning_rate": 5.254820934704309e-06,
+      "loss": 0.6433,
+      "step": 1339
+    },
+    {
+      "epoch": 1.6070785842831432,
+      "grad_norm": 0.7916961270191544,
+      "learning_rate": 5.24784534354835e-06,
+      "loss": 0.6764,
+      "step": 1340
+    },
+    {
+      "epoch": 1.6082783443311337,
+      "grad_norm": 0.6868484859480074,
+      "learning_rate": 5.240869268775422e-06,
+      "loss": 0.6777,
+      "step": 1341
+    },
+    {
+      "epoch": 1.6094781043791242,
+      "grad_norm": 0.7051861937478411,
+      "learning_rate": 5.2338927239978446e-06,
+      "loss": 0.6278,
+      "step": 1342
+    },
+    {
+      "epoch": 1.6106778644271147,
+      "grad_norm": 0.6532318157109314,
+      "learning_rate": 5.22691572282884e-06,
+      "loss": 0.5676,
+      "step": 1343
+    },
+    {
+      "epoch": 1.611877624475105,
+      "grad_norm": 0.9654876160289116,
+      "learning_rate": 5.21993827888253e-06,
+      "loss": 0.6753,
+      "step": 1344
+    },
+    {
+      "epoch": 1.6130773845230952,
+      "grad_norm": 0.5877057698919808,
+      "learning_rate": 5.2129604057739e-06,
+      "loss": 0.5589,
+      "step": 1345
+    },
+    {
+      "epoch": 1.6142771445710857,
+      "grad_norm": 0.6121179015957432,
+      "learning_rate": 5.205982117118768e-06,
+      "loss": 0.6621,
+      "step": 1346
+    },
+    {
+      "epoch": 1.6154769046190762,
+      "grad_norm": 0.7429376189213517,
+      "learning_rate": 5.199003426533767e-06,
+      "loss": 0.6336,
+      "step": 1347
+    },
+    {
+      "epoch": 1.6166766646670667,
+      "grad_norm": 0.6341338353766705,
+      "learning_rate": 5.192024347636313e-06,
+      "loss": 0.6051,
+      "step": 1348
+    },
+    {
+      "epoch": 1.617876424715057,
+      "grad_norm": 0.7824479742838273,
+      "learning_rate": 5.185044894044582e-06,
+      "loss": 0.6389,
+      "step": 1349
+    },
+    {
+      "epoch": 1.6190761847630473,
+      "grad_norm": 0.8642004701087276,
+      "learning_rate": 5.178065079377473e-06,
+      "loss": 0.5368,
+      "step": 1350
+    },
+    {
+      "epoch": 1.6202759448110378,
+      "grad_norm": 0.6679893599433782,
+      "learning_rate": 5.171084917254602e-06,
+      "loss": 0.6081,
+      "step": 1351
+    },
+    {
+      "epoch": 1.6214757048590283,
+      "grad_norm": 0.7236938887490497,
+      "learning_rate": 5.164104421296254e-06,
+      "loss": 0.5892,
+      "step": 1352
+    },
+    {
+      "epoch": 1.6226754649070187,
+      "grad_norm": 0.6099431977403261,
+      "learning_rate": 5.157123605123366e-06,
+      "loss": 0.5639,
+      "step": 1353
+    },
+    {
+      "epoch": 1.623875224955009,
+      "grad_norm": 0.7748564110848425,
+      "learning_rate": 5.150142482357505e-06,
+      "loss": 0.5525,
+      "step": 1354
+    },
+    {
+      "epoch": 1.6250749850029993,
+      "grad_norm": 0.6297212838369077,
+      "learning_rate": 5.14316106662083e-06,
+      "loss": 0.6638,
+      "step": 1355
+    },
+    {
+      "epoch": 1.6262747450509898,
+      "grad_norm": 0.9192127022014425,
+      "learning_rate": 5.136179371536076e-06,
+      "loss": 0.6518,
+      "step": 1356
+    },
+    {
+      "epoch": 1.6274745050989803,
+      "grad_norm": 0.5657648988393104,
+      "learning_rate": 5.129197410726522e-06,
+      "loss": 0.6914,
+      "step": 1357
+    },
+    {
+      "epoch": 1.6286742651469706,
+      "grad_norm": 0.779290014963476,
+      "learning_rate": 5.122215197815965e-06,
+      "loss": 0.6128,
+      "step": 1358
+    },
+    {
+      "epoch": 1.629874025194961,
+      "grad_norm": 0.743576865686689,
+      "learning_rate": 5.115232746428694e-06,
+      "loss": 0.6733,
+      "step": 1359
+    },
+    {
+      "epoch": 1.6310737852429513,
+      "grad_norm": 0.7455700611458127,
+      "learning_rate": 5.108250070189462e-06,
+      "loss": 0.6036,
+      "step": 1360
+    },
+    {
+      "epoch": 1.6322735452909418,
+      "grad_norm": 0.7410112167182892,
+      "learning_rate": 5.1012671827234664e-06,
+      "loss": 0.6891,
+      "step": 1361
+    },
+    {
+      "epoch": 1.6334733053389323,
+      "grad_norm": 0.7360720344296737,
+      "learning_rate": 5.094284097656308e-06,
+      "loss": 0.6906,
+      "step": 1362
+    },
+    {
+      "epoch": 1.6346730653869226,
+      "grad_norm": 0.8699066333508229,
+      "learning_rate": 5.087300828613978e-06,
+      "loss": 0.64,
+      "step": 1363
+    },
+    {
+      "epoch": 1.6358728254349129,
+      "grad_norm": 0.5635231414162302,
+      "learning_rate": 5.08031738922283e-06,
+      "loss": 0.6577,
+      "step": 1364
+    },
+    {
+      "epoch": 1.6370725854829034,
+      "grad_norm": 0.5908428663621056,
+      "learning_rate": 5.073333793109545e-06,
+      "loss": 0.678,
+      "step": 1365
+    },
+    {
+      "epoch": 1.6382723455308938,
+      "grad_norm": 0.8734239583999815,
+      "learning_rate": 5.066350053901109e-06,
+      "loss": 0.5877,
+      "step": 1366
+    },
+    {
+      "epoch": 1.6394721055788843,
+      "grad_norm": 0.7409036971656349,
+      "learning_rate": 5.059366185224791e-06,
+      "loss": 0.6378,
+      "step": 1367
+    },
+    {
+      "epoch": 1.6406718656268746,
+      "grad_norm": 0.6545105385418288,
+      "learning_rate": 5.052382200708111e-06,
+      "loss": 0.6184,
+      "step": 1368
+    },
+    {
+      "epoch": 1.6418716256748649,
+      "grad_norm": 0.697345909377518,
+      "learning_rate": 5.045398113978816e-06,
+      "loss": 0.6545,
+      "step": 1369
+    },
+    {
+      "epoch": 1.6430713857228554,
+      "grad_norm": 0.5220491623520667,
+      "learning_rate": 5.038413938664849e-06,
+      "loss": 0.5996,
+      "step": 1370
+    },
+    {
+      "epoch": 1.6442711457708459,
+      "grad_norm": 0.5669323994294941,
+      "learning_rate": 5.03142968839433e-06,
+      "loss": 0.6126,
+      "step": 1371
+    },
+    {
+      "epoch": 1.6454709058188364,
+      "grad_norm": 0.6387534834755425,
+      "learning_rate": 5.024445376795524e-06,
+      "loss": 0.5562,
+      "step": 1372
+    },
+    {
+      "epoch": 1.6466706658668266,
+      "grad_norm": 0.7008635230548818,
+      "learning_rate": 5.017461017496814e-06,
+      "loss": 0.6342,
+      "step": 1373
+    },
+    {
+      "epoch": 1.647870425914817,
+      "grad_norm": 0.6515918866550156,
+      "learning_rate": 5.010476624126678e-06,
+      "loss": 0.6251,
+      "step": 1374
+    },
+    {
+      "epoch": 1.6490701859628074,
+      "grad_norm": 0.5081521953065723,
+      "learning_rate": 5.00349221031366e-06,
+      "loss": 0.663,
+      "step": 1375
+    },
+    {
+      "epoch": 1.650269946010798,
+      "grad_norm": 0.7432546157072372,
+      "learning_rate": 4.996507789686342e-06,
+      "loss": 0.642,
+      "step": 1376
+    },
+    {
+      "epoch": 1.6514697060587884,
+      "grad_norm": 0.552207404022524,
+      "learning_rate": 4.989523375873325e-06,
+      "loss": 0.6261,
+      "step": 1377
+    },
+    {
+      "epoch": 1.6526694661067787,
+      "grad_norm": 0.6493854028503266,
+      "learning_rate": 4.982538982503188e-06,
+      "loss": 0.6748,
+      "step": 1378
+    },
+    {
+      "epoch": 1.653869226154769,
+      "grad_norm": 0.7591599367795824,
+      "learning_rate": 4.975554623204478e-06,
+      "loss": 0.7019,
+      "step": 1379
+    },
+    {
+      "epoch": 1.6550689862027594,
+      "grad_norm": 0.5672636560414229,
+      "learning_rate": 4.9685703116056715e-06,
+      "loss": 0.6207,
+      "step": 1380
+    },
+    {
+      "epoch": 1.65626874625075,
+      "grad_norm": 0.6634334406291118,
+      "learning_rate": 4.961586061335153e-06,
+      "loss": 0.6331,
+      "step": 1381
+    },
+    {
+      "epoch": 1.6574685062987402,
+      "grad_norm": 0.5677574979637521,
+      "learning_rate": 4.9546018860211845e-06,
+      "loss": 0.6381,
+      "step": 1382
+    },
+    {
+      "epoch": 1.6586682663467307,
+      "grad_norm": 0.5986843581839568,
+      "learning_rate": 4.94761779929189e-06,
+      "loss": 0.6833,
+      "step": 1383
+    },
+    {
+      "epoch": 1.659868026394721,
+      "grad_norm": 0.6519456536042523,
+      "learning_rate": 4.9406338147752094e-06,
+      "loss": 0.5493,
+      "step": 1384
+    },
+    {
+      "epoch": 1.6610677864427115,
+      "grad_norm": 0.6893166733936632,
+      "learning_rate": 4.933649946098892e-06,
+      "loss": 0.6549,
+      "step": 1385
+    },
+    {
+      "epoch": 1.662267546490702,
+      "grad_norm": 0.7457906133470886,
+      "learning_rate": 4.926666206890457e-06,
+      "loss": 0.6356,
+      "step": 1386
+    },
+    {
+      "epoch": 1.6634673065386922,
+      "grad_norm": 0.5601863451041117,
+      "learning_rate": 4.91968261077717e-06,
+      "loss": 0.6127,
+      "step": 1387
+    },
+    {
+      "epoch": 1.6646670665866825,
+      "grad_norm": 0.6284409463777942,
+      "learning_rate": 4.912699171386023e-06,
+      "loss": 0.6172,
+      "step": 1388
+    },
+    {
+      "epoch": 1.665866826634673,
+      "grad_norm": 0.5949620807781346,
+      "learning_rate": 4.905715902343695e-06,
+      "loss": 0.6274,
+      "step": 1389
+    },
+    {
+      "epoch": 1.6670665866826635,
+      "grad_norm": 0.8324715418723123,
+      "learning_rate": 4.898732817276535e-06,
+      "loss": 0.6103,
+      "step": 1390
+    },
+    {
+      "epoch": 1.668266346730654,
+      "grad_norm": 0.6575519307953094,
+      "learning_rate": 4.891749929810538e-06,
+      "loss": 0.6174,
+      "step": 1391
+    },
+    {
+      "epoch": 1.6694661067786443,
+      "grad_norm": 0.7097139058095568,
+      "learning_rate": 4.884767253571308e-06,
+      "loss": 0.6218,
+      "step": 1392
+    },
+    {
+      "epoch": 1.6706658668266345,
+      "grad_norm": 0.8040567508035438,
+      "learning_rate": 4.877784802184037e-06,
+      "loss": 0.6645,
+      "step": 1393
+    },
+    {
+      "epoch": 1.671865626874625,
+      "grad_norm": 0.7822401523231658,
+      "learning_rate": 4.870802589273478e-06,
+      "loss": 0.5774,
+      "step": 1394
+    },
+    {
+      "epoch": 1.6730653869226155,
+      "grad_norm": 0.616670401217597,
+      "learning_rate": 4.863820628463925e-06,
+      "loss": 0.6616,
+      "step": 1395
+    },
+    {
+      "epoch": 1.674265146970606,
+      "grad_norm": 0.5030978714474055,
+      "learning_rate": 4.856838933379171e-06,
+      "loss": 0.6501,
+      "step": 1396
+    },
+    {
+      "epoch": 1.6754649070185963,
+      "grad_norm": 0.6174359209713869,
+      "learning_rate": 4.8498575176424955e-06,
+      "loss": 0.5841,
+      "step": 1397
+    },
+    {
+      "epoch": 1.6766646670665866,
+      "grad_norm": 0.7981596232184236,
+      "learning_rate": 4.842876394876636e-06,
+      "loss": 0.5693,
+      "step": 1398
+    },
+    {
+      "epoch": 1.677864427114577,
+      "grad_norm": 0.5773863788433636,
+      "learning_rate": 4.8358955787037476e-06,
+      "loss": 0.5577,
+      "step": 1399
+    },
+    {
+      "epoch": 1.6790641871625676,
+      "grad_norm": 0.8147173648555551,
+      "learning_rate": 4.828915082745398e-06,
+      "loss": 0.6019,
+      "step": 1400
+    },
+    {
+      "epoch": 1.680263947210558,
+      "grad_norm": 0.6998856064970432,
+      "learning_rate": 4.821934920622528e-06,
+      "loss": 0.6435,
+      "step": 1401
+    },
+    {
+      "epoch": 1.6814637072585483,
+      "grad_norm": 0.6301096966575788,
+      "learning_rate": 4.81495510595542e-06,
+      "loss": 0.5991,
+      "step": 1402
+    },
+    {
+      "epoch": 1.6826634673065386,
+      "grad_norm": 0.5376164948997293,
+      "learning_rate": 4.807975652363687e-06,
+      "loss": 0.6178,
+      "step": 1403
+    },
+    {
+      "epoch": 1.683863227354529,
+      "grad_norm": 0.6960603742114515,
+      "learning_rate": 4.8009965734662345e-06,
+      "loss": 0.6492,
+      "step": 1404
+    },
+    {
+      "epoch": 1.6850629874025196,
+      "grad_norm": 0.8949959217017778,
+      "learning_rate": 4.794017882881233e-06,
+      "loss": 0.6691,
+      "step": 1405
+    },
+    {
+      "epoch": 1.6862627474505099,
+      "grad_norm": 0.6958345832251298,
+      "learning_rate": 4.7870395942261005e-06,
+      "loss": 0.6615,
+      "step": 1406
+    },
+    {
+      "epoch": 1.6874625074985004,
+      "grad_norm": 0.586209277636213,
+      "learning_rate": 4.7800617211174705e-06,
+      "loss": 0.6434,
+      "step": 1407
+    },
+    {
+      "epoch": 1.6886622675464906,
+      "grad_norm": 0.73110751774531,
+      "learning_rate": 4.773084277171161e-06,
+      "loss": 0.6084,
+      "step": 1408
+    },
+    {
+      "epoch": 1.6898620275944811,
+      "grad_norm": 0.7239656781429334,
+      "learning_rate": 4.766107276002158e-06,
+      "loss": 0.5835,
+      "step": 1409
+    },
+    {
+      "epoch": 1.6910617876424716,
+      "grad_norm": 0.6031929392948729,
+      "learning_rate": 4.759130731224579e-06,
+      "loss": 0.6424,
+      "step": 1410
+    },
+    {
+      "epoch": 1.6922615476904619,
+      "grad_norm": 0.8232354870495764,
+      "learning_rate": 4.752154656451652e-06,
+      "loss": 0.5513,
+      "step": 1411
+    },
+    {
+      "epoch": 1.6934613077384522,
+      "grad_norm": 0.7640962339418518,
+      "learning_rate": 4.745179065295692e-06,
+      "loss": 0.671,
+      "step": 1412
+    },
+    {
+      "epoch": 1.6946610677864427,
+      "grad_norm": 0.7514613522929842,
+      "learning_rate": 4.738203971368066e-06,
+      "loss": 0.6191,
+      "step": 1413
+    },
+    {
+      "epoch": 1.6958608278344331,
+      "grad_norm": 0.5324176220078973,
+      "learning_rate": 4.73122938827917e-06,
+      "loss": 0.6231,
+      "step": 1414
+    },
+    {
+      "epoch": 1.6970605878824236,
+      "grad_norm": 0.7887179892123806,
+      "learning_rate": 4.724255329638411e-06,
+      "loss": 0.617,
+      "step": 1415
+    },
+    {
+      "epoch": 1.698260347930414,
+      "grad_norm": 0.5051402718166775,
+      "learning_rate": 4.717281809054159e-06,
+      "loss": 0.5679,
+      "step": 1416
+    },
+    {
+      "epoch": 1.6994601079784042,
+      "grad_norm": 0.7305596217490118,
+      "learning_rate": 4.710308840133747e-06,
+      "loss": 0.6135,
+      "step": 1417
+    },
+    {
+      "epoch": 1.7006598680263947,
+      "grad_norm": 0.718026410793312,
+      "learning_rate": 4.703336436483426e-06,
+      "loss": 0.6765,
+      "step": 1418
+    },
+    {
+      "epoch": 1.7018596280743852,
+      "grad_norm": 0.752441248195329,
+      "learning_rate": 4.696364611708342e-06,
+      "loss": 0.6097,
+      "step": 1419
+    },
+    {
+      "epoch": 1.7030593881223757,
+      "grad_norm": 0.6721419890454678,
+      "learning_rate": 4.689393379412517e-06,
+      "loss": 0.6639,
+      "step": 1420
+    },
+    {
+      "epoch": 1.704259148170366,
+      "grad_norm": 0.7818993504772956,
+      "learning_rate": 4.682422753198812e-06,
+      "loss": 0.6074,
+      "step": 1421
+    },
+    {
+      "epoch": 1.7054589082183562,
+      "grad_norm": 0.6065727544273684,
+      "learning_rate": 4.6754527466689086e-06,
+      "loss": 0.6049,
+      "step": 1422
+    },
+    {
+      "epoch": 1.7066586682663467,
+      "grad_norm": 0.6877103580556603,
+      "learning_rate": 4.668483373423273e-06,
+      "loss": 0.6457,
+      "step": 1423
+    },
+    {
+      "epoch": 1.7078584283143372,
+      "grad_norm": 0.6648773775650588,
+      "learning_rate": 4.661514647061149e-06,
+      "loss": 0.5718,
+      "step": 1424
+    },
+    {
+      "epoch": 1.7090581883623275,
+      "grad_norm": 0.7531066166259784,
+      "learning_rate": 4.654546581180502e-06,
+      "loss": 0.6491,
+      "step": 1425
+    },
+    {
+      "epoch": 1.710257948410318,
+      "grad_norm": 0.7177465989254774,
+      "learning_rate": 4.647579189378019e-06,
+      "loss": 0.6496,
+      "step": 1426
+    },
+    {
+      "epoch": 1.7114577084583082,
+      "grad_norm": 0.6506887097762082,
+      "learning_rate": 4.640612485249072e-06,
+      "loss": 0.6098,
+      "step": 1427
+    },
+    {
+      "epoch": 1.7126574685062987,
+      "grad_norm": 0.6342494598879174,
+      "learning_rate": 4.6336464823876826e-06,
+      "loss": 0.585,
+      "step": 1428
+    },
+    {
+      "epoch": 1.7138572285542892,
+      "grad_norm": 0.7498553556828876,
+      "learning_rate": 4.6266811943865124e-06,
+      "loss": 0.6213,
+      "step": 1429
+    },
+    {
+      "epoch": 1.7150569886022795,
+      "grad_norm": 0.5775583134180416,
+      "learning_rate": 4.619716634836826e-06,
+      "loss": 0.6555,
+      "step": 1430
+    },
+    {
+      "epoch": 1.7162567486502698,
+      "grad_norm": 0.6937580595546063,
+      "learning_rate": 4.612752817328463e-06,
+      "loss": 0.6028,
+      "step": 1431
+    },
+    {
+      "epoch": 1.7174565086982603,
+      "grad_norm": 0.7321126076487201,
+      "learning_rate": 4.60578975544982e-06,
+      "loss": 0.6546,
+      "step": 1432
+    },
+    {
+      "epoch": 1.7186562687462508,
+      "grad_norm": 0.8260819299546153,
+      "learning_rate": 4.5988274627878175e-06,
+      "loss": 0.684,
+      "step": 1433
+    },
+    {
+      "epoch": 1.7198560287942413,
+      "grad_norm": 0.6295125038484417,
+      "learning_rate": 4.591865952927873e-06,
+      "loss": 0.6312,
+      "step": 1434
+    },
+    {
+      "epoch": 1.7210557888422315,
+      "grad_norm": 0.7764441386243321,
+      "learning_rate": 4.5849052394538775e-06,
+      "loss": 0.6707,
+      "step": 1435
+    },
+    {
+      "epoch": 1.7222555488902218,
+      "grad_norm": 0.5909491911998985,
+      "learning_rate": 4.577945335948172e-06,
+      "loss": 0.5888,
+      "step": 1436
+    },
+    {
+      "epoch": 1.7234553089382123,
+      "grad_norm": 0.6214967272966967,
+      "learning_rate": 4.570986255991509e-06,
+      "loss": 0.686,
+      "step": 1437
+    },
+    {
+      "epoch": 1.7246550689862028,
+      "grad_norm": 0.7019193004149545,
+      "learning_rate": 4.564028013163043e-06,
+      "loss": 0.6136,
+      "step": 1438
+    },
+    {
+      "epoch": 1.7258548290341933,
+      "grad_norm": 0.6201886115016094,
+      "learning_rate": 4.5570706210402874e-06,
+      "loss": 0.5633,
+      "step": 1439
+    },
+    {
+      "epoch": 1.7270545890821836,
+      "grad_norm": 0.8722538070364205,
+      "learning_rate": 4.5501140931991e-06,
+      "loss": 0.5883,
+      "step": 1440
+    },
+    {
+      "epoch": 1.7282543491301738,
+      "grad_norm": 0.6686952418096178,
+      "learning_rate": 4.543158443213653e-06,
+      "loss": 0.6282,
+      "step": 1441
+    },
+    {
+      "epoch": 1.7294541091781643,
+      "grad_norm": 0.7937448111380455,
+      "learning_rate": 4.5362036846564015e-06,
+      "loss": 0.5691,
+      "step": 1442
+    },
+    {
+      "epoch": 1.7306538692261548,
+      "grad_norm": 0.6396496808668831,
+      "learning_rate": 4.529249831098062e-06,
+      "loss": 0.6303,
+      "step": 1443
+    },
+    {
+      "epoch": 1.7318536292741453,
+      "grad_norm": 0.6642865948086335,
+      "learning_rate": 4.52229689610759e-06,
+      "loss": 0.6523,
+      "step": 1444
+    },
+    {
+      "epoch": 1.7330533893221356,
+      "grad_norm": 0.7621455130419058,
+      "learning_rate": 4.515344893252142e-06,
+      "loss": 0.6516,
+      "step": 1445
+    },
+    {
+      "epoch": 1.7342531493701259,
+      "grad_norm": 0.5789503449826376,
+      "learning_rate": 4.508393836097061e-06,
+      "loss": 0.6032,
+      "step": 1446
+    },
+    {
+      "epoch": 1.7354529094181164,
+      "grad_norm": 0.6277333953284421,
+      "learning_rate": 4.501443738205841e-06,
+      "loss": 0.6478,
+      "step": 1447
+    },
+    {
+      "epoch": 1.7366526694661069,
+      "grad_norm": 0.6816407236510691,
+      "learning_rate": 4.494494613140105e-06,
+      "loss": 0.6952,
+      "step": 1448
+    },
+    {
+      "epoch": 1.7378524295140971,
+      "grad_norm": 0.5982908395568053,
+      "learning_rate": 4.487546474459579e-06,
+      "loss": 0.5816,
+      "step": 1449
+    },
+    {
+      "epoch": 1.7390521895620876,
+      "grad_norm": 0.7660687758767524,
+      "learning_rate": 4.480599335722066e-06,
+      "loss": 0.619,
+      "step": 1450
+    },
+    {
+      "epoch": 1.740251949610078,
+      "grad_norm": 0.7223849213293781,
+      "learning_rate": 4.4736532104834105e-06,
+      "loss": 0.6366,
+      "step": 1451
+    },
+    {
+      "epoch": 1.7414517096580684,
+      "grad_norm": 0.5146072094360288,
+      "learning_rate": 4.466708112297489e-06,
+      "loss": 0.6983,
+      "step": 1452
+    },
+    {
+      "epoch": 1.7426514697060589,
+      "grad_norm": 0.7024831535983507,
+      "learning_rate": 4.459764054716167e-06,
+      "loss": 0.6105,
+      "step": 1453
+    },
+    {
+      "epoch": 1.7438512297540492,
+      "grad_norm": 0.4880549556650625,
+      "learning_rate": 4.4528210512892814e-06,
+      "loss": 0.5857,
+      "step": 1454
+    },
+    {
+      "epoch": 1.7450509898020394,
+      "grad_norm": 0.8018785675884215,
+      "learning_rate": 4.445879115564612e-06,
+      "loss": 0.6069,
+      "step": 1455
+    },
+    {
+      "epoch": 1.74625074985003,
+      "grad_norm": 0.571914332481684,
+      "learning_rate": 4.438938261087859e-06,
+      "loss": 0.6243,
+      "step": 1456
+    },
+    {
+      "epoch": 1.7474505098980204,
+      "grad_norm": 0.5807256525801378,
+      "learning_rate": 4.431998501402605e-06,
+      "loss": 0.6511,
+      "step": 1457
+    },
+    {
+      "epoch": 1.748650269946011,
+      "grad_norm": 0.7345672940171202,
+      "learning_rate": 4.4250598500503045e-06,
+      "loss": 0.555,
+      "step": 1458
+    },
+    {
+      "epoch": 1.7498500299940012,
+      "grad_norm": 0.7940960330578659,
+      "learning_rate": 4.4181223205702435e-06,
+      "loss": 0.61,
+      "step": 1459
+    },
+    {
+      "epoch": 1.7510497900419915,
+      "grad_norm": 0.7993824179585033,
+      "learning_rate": 4.41118592649952e-06,
+      "loss": 0.6065,
+      "step": 1460
+    },
+    {
+      "epoch": 1.752249550089982,
+      "grad_norm": 0.6358705050170755,
+      "learning_rate": 4.404250681373021e-06,
+      "loss": 0.6099,
+      "step": 1461
+    },
+    {
+      "epoch": 1.7534493101379725,
+      "grad_norm": 0.6965743907502828,
+      "learning_rate": 4.397316598723385e-06,
+      "loss": 0.6742,
+      "step": 1462
+    },
+    {
+      "epoch": 1.754649070185963,
+      "grad_norm": 0.6328725130943839,
+      "learning_rate": 4.390383692080986e-06,
+      "loss": 0.5806,
+      "step": 1463
+    },
+    {
+      "epoch": 1.7558488302339532,
+      "grad_norm": 0.6311566335345002,
+      "learning_rate": 4.383451974973903e-06,
+      "loss": 0.6053,
+      "step": 1464
+    },
+    {
+      "epoch": 1.7570485902819435,
+      "grad_norm": 0.7052395609633253,
+      "learning_rate": 4.376521460927896e-06,
+      "loss": 0.6321,
+      "step": 1465
+    },
+    {
+      "epoch": 1.758248350329934,
+      "grad_norm": 0.7535406881872222,
+      "learning_rate": 4.36959216346637e-06,
+      "loss": 0.5487,
+      "step": 1466
+    },
+    {
+      "epoch": 1.7594481103779245,
+      "grad_norm": 0.6338229748455944,
+      "learning_rate": 4.362664096110361e-06,
+      "loss": 0.7053,
+      "step": 1467
+    },
+    {
+      "epoch": 1.760647870425915,
+      "grad_norm": 0.6309233539496764,
+      "learning_rate": 4.35573727237851e-06,
+      "loss": 0.6439,
+      "step": 1468
+    },
+    {
+      "epoch": 1.7618476304739052,
+      "grad_norm": 0.6150574415134877,
+      "learning_rate": 4.3488117057870225e-06,
+      "loss": 0.6238,
+      "step": 1469
+    },
+    {
+      "epoch": 1.7630473905218955,
+      "grad_norm": 0.5685790958617888,
+      "learning_rate": 4.341887409849656e-06,
+      "loss": 0.6298,
+      "step": 1470
+    },
+    {
+      "epoch": 1.764247150569886,
+      "grad_norm": 0.6752528106945532,
+      "learning_rate": 4.334964398077684e-06,
+      "loss": 0.5655,
+      "step": 1471
+    },
+    {
+      "epoch": 1.7654469106178765,
+      "grad_norm": 0.7649358108814712,
+      "learning_rate": 4.328042683979878e-06,
+      "loss": 0.6546,
+      "step": 1472
+    },
+    {
+      "epoch": 1.7666466706658668,
+      "grad_norm": 0.7905896064810997,
+      "learning_rate": 4.321122281062481e-06,
+      "loss": 0.6105,
+      "step": 1473
+    },
+    {
+      "epoch": 1.7678464307138573,
+      "grad_norm": 0.5660683589294571,
+      "learning_rate": 4.3142032028291695e-06,
+      "loss": 0.5924,
+      "step": 1474
+    },
+    {
+      "epoch": 1.7690461907618475,
+      "grad_norm": 0.7846283436256514,
+      "learning_rate": 4.307285462781038e-06,
+      "loss": 0.6607,
+      "step": 1475
+    },
+    {
+      "epoch": 1.770245950809838,
+      "grad_norm": 0.6580703214396916,
+      "learning_rate": 4.3003690744165765e-06,
+      "loss": 0.578,
+      "step": 1476
+    },
+    {
+      "epoch": 1.7714457108578285,
+      "grad_norm": 0.5237471095482176,
+      "learning_rate": 4.293454051231624e-06,
+      "loss": 0.5871,
+      "step": 1477
+    },
+    {
+      "epoch": 1.7726454709058188,
+      "grad_norm": 0.6327731192095628,
+      "learning_rate": 4.286540406719367e-06,
+      "loss": 0.6445,
+      "step": 1478
+    },
+    {
+      "epoch": 1.773845230953809,
+      "grad_norm": 0.8070222294599864,
+      "learning_rate": 4.2796281543703e-06,
+      "loss": 0.63,
+      "step": 1479
+    },
+    {
+      "epoch": 1.7750449910017996,
+      "grad_norm": 0.6062725964397185,
+      "learning_rate": 4.272717307672194e-06,
+      "loss": 0.6437,
+      "step": 1480
+    },
+    {
+      "epoch": 1.77624475104979,
+      "grad_norm": 0.6722768114469376,
+      "learning_rate": 4.265807880110087e-06,
+      "loss": 0.5731,
+      "step": 1481
+    },
+    {
+      "epoch": 1.7774445110977806,
+      "grad_norm": 0.5678106199265854,
+      "learning_rate": 4.258899885166242e-06,
+      "loss": 0.6429,
+      "step": 1482
+    },
+    {
+      "epoch": 1.7786442711457708,
+      "grad_norm": 0.5571672309900889,
+      "learning_rate": 4.251993336320127e-06,
+      "loss": 0.6058,
+      "step": 1483
+    },
+    {
+      "epoch": 1.7798440311937611,
+      "grad_norm": 0.7036183857353746,
+      "learning_rate": 4.245088247048388e-06,
+      "loss": 0.6477,
+      "step": 1484
+    },
+    {
+      "epoch": 1.7810437912417516,
+      "grad_norm": 0.7123575403550069,
+      "learning_rate": 4.238184630824828e-06,
+      "loss": 0.6218,
+      "step": 1485
+    },
+    {
+      "epoch": 1.782243551289742,
+      "grad_norm": 0.7751307578636013,
+      "learning_rate": 4.231282501120366e-06,
+      "loss": 0.6096,
+      "step": 1486
+    },
+    {
+      "epoch": 1.7834433113377326,
+      "grad_norm": 0.5388334838746758,
+      "learning_rate": 4.224381871403028e-06,
+      "loss": 0.608,
+      "step": 1487
+    },
+    {
+      "epoch": 1.7846430713857229,
+      "grad_norm": 0.7485815957038062,
+      "learning_rate": 4.217482755137916e-06,
+      "loss": 0.5487,
+      "step": 1488
+    },
+    {
+      "epoch": 1.7858428314337131,
+      "grad_norm": 0.6686635243767864,
+      "learning_rate": 4.210585165787166e-06,
+      "loss": 0.5603,
+      "step": 1489
+    },
+    {
+      "epoch": 1.7870425914817036,
+      "grad_norm": 0.7980471135553686,
+      "learning_rate": 4.203689116809946e-06,
+      "loss": 0.5965,
+      "step": 1490
+    },
+    {
+      "epoch": 1.7882423515296941,
+      "grad_norm": 0.7711362904906689,
+      "learning_rate": 4.196794621662417e-06,
+      "loss": 0.6021,
+      "step": 1491
+    },
+    {
+      "epoch": 1.7894421115776846,
+      "grad_norm": 0.6478401612668694,
+      "learning_rate": 4.1899016937977e-06,
+      "loss": 0.5521,
+      "step": 1492
+    },
+    {
+      "epoch": 1.790641871625675,
+      "grad_norm": 0.7301532095236675,
+      "learning_rate": 4.183010346665869e-06,
+      "loss": 0.5973,
+      "step": 1493
+    },
+    {
+      "epoch": 1.7918416316736652,
+      "grad_norm": 0.7504012219009591,
+      "learning_rate": 4.1761205937139075e-06,
+      "loss": 0.5623,
+      "step": 1494
+    },
+    {
+      "epoch": 1.7930413917216557,
+      "grad_norm": 0.8575530566266752,
+      "learning_rate": 4.169232448385685e-06,
+      "loss": 0.6651,
+      "step": 1495
+    },
+    {
+      "epoch": 1.7942411517696462,
+      "grad_norm": 0.5991870908982022,
+      "learning_rate": 4.162345924121941e-06,
+      "loss": 0.5979,
+      "step": 1496
+    },
+    {
+      "epoch": 1.7954409118176364,
+      "grad_norm": 0.727844764668071,
+      "learning_rate": 4.15546103436025e-06,
+      "loss": 0.6081,
+      "step": 1497
+    },
+    {
+      "epoch": 1.796640671865627,
+      "grad_norm": 0.7160612221748286,
+      "learning_rate": 4.148577792534992e-06,
+      "loss": 0.6317,
+      "step": 1498
+    },
+    {
+      "epoch": 1.7978404319136172,
+      "grad_norm": 0.7569049873211127,
+      "learning_rate": 4.14169621207734e-06,
+      "loss": 0.5891,
+      "step": 1499
+    },
+    {
+      "epoch": 1.7990401919616077,
+      "grad_norm": 0.626522678372816,
+      "learning_rate": 4.134816306415216e-06,
+      "loss": 0.6095,
+      "step": 1500
+    },
+    {
+      "epoch": 1.8002399520095982,
+      "grad_norm": 0.6719967537153403,
+      "learning_rate": 4.127938088973279e-06,
+      "loss": 0.6506,
+      "step": 1501
+    },
+    {
+      "epoch": 1.8014397120575885,
+      "grad_norm": 0.5512021941132703,
+      "learning_rate": 4.121061573172898e-06,
+      "loss": 0.6205,
+      "step": 1502
+    },
+    {
+      "epoch": 1.8026394721055787,
+      "grad_norm": 0.7297583764268756,
+      "learning_rate": 4.1141867724321115e-06,
+      "loss": 0.6474,
+      "step": 1503
+    },
+    {
+      "epoch": 1.8038392321535692,
+      "grad_norm": 0.45406727311456535,
+      "learning_rate": 4.1073137001656184e-06,
+      "loss": 0.665,
+      "step": 1504
+    },
+    {
+      "epoch": 1.8050389922015597,
+      "grad_norm": 0.770605560110556,
+      "learning_rate": 4.100442369784746e-06,
+      "loss": 0.6455,
+      "step": 1505
+    },
+    {
+      "epoch": 1.8062387522495502,
+      "grad_norm": 0.6900802474911933,
+      "learning_rate": 4.0935727946974176e-06,
+      "loss": 0.6682,
+      "step": 1506
+    },
+    {
+      "epoch": 1.8074385122975405,
+      "grad_norm": 0.5556687244755016,
+      "learning_rate": 4.086704988308134e-06,
+      "loss": 0.6034,
+      "step": 1507
+    },
+    {
+      "epoch": 1.8086382723455308,
+      "grad_norm": 0.6873341697662002,
+      "learning_rate": 4.079838964017945e-06,
+      "loss": 0.5547,
+      "step": 1508
+    },
+    {
+      "epoch": 1.8098380323935213,
+      "grad_norm": 0.6874227965615172,
+      "learning_rate": 4.072974735224421e-06,
+      "loss": 0.6071,
+      "step": 1509
+    },
+    {
+      "epoch": 1.8110377924415118,
+      "grad_norm": 0.6122799863116688,
+      "learning_rate": 4.066112315321632e-06,
+      "loss": 0.6301,
+      "step": 1510
+    },
+    {
+      "epoch": 1.8122375524895022,
+      "grad_norm": 0.6156415651630566,
+      "learning_rate": 4.059251717700118e-06,
+      "loss": 0.6434,
+      "step": 1511
+    },
+    {
+      "epoch": 1.8134373125374925,
+      "grad_norm": 0.7039193843462168,
+      "learning_rate": 4.05239295574686e-06,
+      "loss": 0.5995,
+      "step": 1512
+    },
+    {
+      "epoch": 1.8146370725854828,
+      "grad_norm": 0.7226905662344233,
+      "learning_rate": 4.045536042845257e-06,
+      "loss": 0.6785,
+      "step": 1513
+    },
+    {
+      "epoch": 1.8158368326334733,
+      "grad_norm": 0.7656410951997662,
+      "learning_rate": 4.038680992375108e-06,
+      "loss": 0.5813,
+      "step": 1514
+    },
+    {
+      "epoch": 1.8170365926814638,
+      "grad_norm": 0.6446736657739984,
+      "learning_rate": 4.031827817712568e-06,
+      "loss": 0.5885,
+      "step": 1515
+    },
+    {
+      "epoch": 1.8182363527294543,
+      "grad_norm": 0.6008413183656197,
+      "learning_rate": 4.024976532230135e-06,
+      "loss": 0.626,
+      "step": 1516
+    },
+    {
+      "epoch": 1.8194361127774445,
+      "grad_norm": 0.6617307462453432,
+      "learning_rate": 4.018127149296626e-06,
+      "loss": 0.6027,
+      "step": 1517
+    },
+    {
+      "epoch": 1.8206358728254348,
+      "grad_norm": 0.6356581631265987,
+      "learning_rate": 4.011279682277135e-06,
+      "loss": 0.5932,
+      "step": 1518
+    },
+    {
+      "epoch": 1.8218356328734253,
+      "grad_norm": 0.6686838163259288,
+      "learning_rate": 4.004434144533027e-06,
+      "loss": 0.6644,
+      "step": 1519
+    },
+    {
+      "epoch": 1.8230353929214158,
+      "grad_norm": 0.7226742772232984,
+      "learning_rate": 3.9975905494218995e-06,
+      "loss": 0.6478,
+      "step": 1520
+    },
+    {
+      "epoch": 1.824235152969406,
+      "grad_norm": 0.6254638417669428,
+      "learning_rate": 3.9907489102975546e-06,
+      "loss": 0.5728,
+      "step": 1521
+    },
+    {
+      "epoch": 1.8254349130173964,
+      "grad_norm": 0.7162896495952091,
+      "learning_rate": 3.9839092405099856e-06,
+      "loss": 0.6315,
+      "step": 1522
+    },
+    {
+      "epoch": 1.8266346730653868,
+      "grad_norm": 0.6635180291450986,
+      "learning_rate": 3.97707155340534e-06,
+      "loss": 0.5864,
+      "step": 1523
+    },
+    {
+      "epoch": 1.8278344331133773,
+      "grad_norm": 0.5896170786586302,
+      "learning_rate": 3.970235862325892e-06,
+      "loss": 0.6109,
+      "step": 1524
+    },
+    {
+      "epoch": 1.8290341931613678,
+      "grad_norm": 0.6741198016269524,
+      "learning_rate": 3.963402180610028e-06,
+      "loss": 0.6371,
+      "step": 1525
+    },
+    {
+      "epoch": 1.830233953209358,
+      "grad_norm": 0.6608823149096729,
+      "learning_rate": 3.95657052159221e-06,
+      "loss": 0.6146,
+      "step": 1526
+    },
+    {
+      "epoch": 1.8314337132573484,
+      "grad_norm": 0.6141701612238413,
+      "learning_rate": 3.949740898602953e-06,
+      "loss": 0.6445,
+      "step": 1527
+    },
+    {
+      "epoch": 1.8326334733053389,
+      "grad_norm": 0.5381899289584886,
+      "learning_rate": 3.942913324968799e-06,
+      "loss": 0.6448,
+      "step": 1528
+    },
+    {
+      "epoch": 1.8338332333533294,
+      "grad_norm": 0.6289407950918525,
+      "learning_rate": 3.936087814012293e-06,
+      "loss": 0.5984,
+      "step": 1529
+    },
+    {
+      "epoch": 1.8350329934013199,
+      "grad_norm": 0.7430299369735149,
+      "learning_rate": 3.929264379051954e-06,
+      "loss": 0.615,
+      "step": 1530
+    },
+    {
+      "epoch": 1.8362327534493101,
+      "grad_norm": 0.6034032294138236,
+      "learning_rate": 3.922443033402249e-06,
+      "loss": 0.6413,
+      "step": 1531
+    },
+    {
+      "epoch": 1.8374325134973004,
+      "grad_norm": 0.5886494650231635,
+      "learning_rate": 3.91562379037357e-06,
+      "loss": 0.5852,
+      "step": 1532
+    },
+    {
+      "epoch": 1.838632273545291,
+      "grad_norm": 0.5068433432677715,
+      "learning_rate": 3.908806663272202e-06,
+      "loss": 0.6635,
+      "step": 1533
+    },
+    {
+      "epoch": 1.8398320335932814,
+      "grad_norm": 0.6607006051476574,
+      "learning_rate": 3.901991665400312e-06,
+      "loss": 0.6946,
+      "step": 1534
+    },
+    {
+      "epoch": 1.841031793641272,
+      "grad_norm": 0.7126897580438574,
+      "learning_rate": 3.895178810055899e-06,
+      "loss": 0.6821,
+      "step": 1535
+    },
+    {
+      "epoch": 1.8422315536892622,
+      "grad_norm": 0.6167775006271451,
+      "learning_rate": 3.888368110532788e-06,
+      "loss": 0.6649,
+      "step": 1536
+    },
+    {
+      "epoch": 1.8434313137372524,
+      "grad_norm": 0.7069325448529817,
+      "learning_rate": 3.8815595801206e-06,
+      "loss": 0.7081,
+      "step": 1537
+    },
+    {
+      "epoch": 1.844631073785243,
+      "grad_norm": 0.5739397130561915,
+      "learning_rate": 3.874753232104714e-06,
+      "loss": 0.5863,
+      "step": 1538
+    },
+    {
+      "epoch": 1.8458308338332334,
+      "grad_norm": 0.6899181546698574,
+      "learning_rate": 3.867949079766262e-06,
+      "loss": 0.5816,
+      "step": 1539
+    },
+    {
+      "epoch": 1.8470305938812237,
+      "grad_norm": 0.5441258043275793,
+      "learning_rate": 3.861147136382085e-06,
+      "loss": 0.5964,
+      "step": 1540
+    },
+    {
+      "epoch": 1.8482303539292142,
+      "grad_norm": 0.8088124006931219,
+      "learning_rate": 3.854347415224712e-06,
+      "loss": 0.6622,
+      "step": 1541
+    },
+    {
+      "epoch": 1.8494301139772045,
+      "grad_norm": 0.7350261458150595,
+      "learning_rate": 3.847549929562341e-06,
+      "loss": 0.614,
+      "step": 1542
+    },
+    {
+      "epoch": 1.850629874025195,
+      "grad_norm": 0.5648819785524932,
+      "learning_rate": 3.840754692658807e-06,
+      "loss": 0.5854,
+      "step": 1543
+    },
+    {
+      "epoch": 1.8518296340731855,
+      "grad_norm": 0.5970755135089987,
+      "learning_rate": 3.833961717773554e-06,
+      "loss": 0.6509,
+      "step": 1544
+    },
+    {
+      "epoch": 1.8530293941211757,
+      "grad_norm": 0.7247750021092256,
+      "learning_rate": 3.8271710181616106e-06,
+      "loss": 0.5358,
+      "step": 1545
+    },
+    {
+      "epoch": 1.854229154169166,
+      "grad_norm": 0.7132939979536987,
+      "learning_rate": 3.820382607073575e-06,
+      "loss": 0.5935,
+      "step": 1546
+    },
+    {
+      "epoch": 1.8554289142171565,
+      "grad_norm": 0.6616048635749737,
+      "learning_rate": 3.8135964977555696e-06,
+      "loss": 0.6382,
+      "step": 1547
+    },
+    {
+      "epoch": 1.856628674265147,
+      "grad_norm": 0.5171970812831564,
+      "learning_rate": 3.8068127034492287e-06,
+      "loss": 0.6577,
+      "step": 1548
+    },
+    {
+      "epoch": 1.8578284343131375,
+      "grad_norm": 0.7182240258341588,
+      "learning_rate": 3.800031237391676e-06,
+      "loss": 0.628,
+      "step": 1549
+    },
+    {
+      "epoch": 1.8590281943611278,
+      "grad_norm": 0.5717078220686386,
+      "learning_rate": 3.7932521128154785e-06,
+      "loss": 0.6608,
+      "step": 1550
+    },
+    {
+      "epoch": 1.860227954409118,
+      "grad_norm": 0.6749749924395927,
+      "learning_rate": 3.7864753429486475e-06,
+      "loss": 0.5582,
+      "step": 1551
+    },
+    {
+      "epoch": 1.8614277144571085,
+      "grad_norm": 0.6451774623688347,
+      "learning_rate": 3.779700941014593e-06,
+      "loss": 0.6287,
+      "step": 1552
+    },
+    {
+      "epoch": 1.862627474505099,
+      "grad_norm": 0.6596254606412011,
+      "learning_rate": 3.772928920232103e-06,
+      "loss": 0.6363,
+      "step": 1553
+    },
+    {
+      "epoch": 1.8638272345530895,
+      "grad_norm": 0.7688241705078183,
+      "learning_rate": 3.7661592938153248e-06,
+      "loss": 0.6424,
+      "step": 1554
+    },
+    {
+      "epoch": 1.8650269946010798,
+      "grad_norm": 0.639799196330518,
+      "learning_rate": 3.7593920749737307e-06,
+      "loss": 0.6,
+      "step": 1555
+    },
+    {
+      "epoch": 1.86622675464907,
+      "grad_norm": 0.6901288327510051,
+      "learning_rate": 3.7526272769120933e-06,
+      "loss": 0.6375,
+      "step": 1556
+    },
+    {
+      "epoch": 1.8674265146970606,
+      "grad_norm": 0.6477346369269456,
+      "learning_rate": 3.745864912830463e-06,
+      "loss": 0.6143,
+      "step": 1557
+    },
+    {
+      "epoch": 1.868626274745051,
+      "grad_norm": 0.7668025609143018,
+      "learning_rate": 3.739104995924144e-06,
+      "loss": 0.6667,
+      "step": 1558
+    },
+    {
+      "epoch": 1.8698260347930415,
+      "grad_norm": 0.5984068174104991,
+      "learning_rate": 3.7323475393836606e-06,
+      "loss": 0.6416,
+      "step": 1559
+    },
+    {
+      "epoch": 1.8710257948410318,
+      "grad_norm": 0.7428673455418627,
+      "learning_rate": 3.725592556394737e-06,
+      "loss": 0.615,
+      "step": 1560
+    },
+    {
+      "epoch": 1.872225554889022,
+      "grad_norm": 0.6109891944291239,
+      "learning_rate": 3.7188400601382768e-06,
+      "loss": 0.6209,
+      "step": 1561
+    },
+    {
+      "epoch": 1.8734253149370126,
+      "grad_norm": 0.6302325104487722,
+      "learning_rate": 3.712090063790319e-06,
+      "loss": 0.5695,
+      "step": 1562
+    },
+    {
+      "epoch": 1.874625074985003,
+      "grad_norm": 0.6689292962337307,
+      "learning_rate": 3.7053425805220385e-06,
+      "loss": 0.6115,
+      "step": 1563
+    },
+    {
+      "epoch": 1.8758248350329934,
+      "grad_norm": 0.7362877890092404,
+      "learning_rate": 3.6985976234996957e-06,
+      "loss": 0.5754,
+      "step": 1564
+    },
+    {
+      "epoch": 1.8770245950809838,
+      "grad_norm": 0.5993796529437634,
+      "learning_rate": 3.691855205884627e-06,
+      "loss": 0.6283,
+      "step": 1565
+    },
+    {
+      "epoch": 1.8782243551289741,
+      "grad_norm": 0.6836726848169472,
+      "learning_rate": 3.685115340833215e-06,
+      "loss": 0.622,
+      "step": 1566
+    },
+    {
+      "epoch": 1.8794241151769646,
+      "grad_norm": 0.679556193178331,
+      "learning_rate": 3.678378041496856e-06,
+      "loss": 0.6357,
+      "step": 1567
+    },
+    {
+      "epoch": 1.880623875224955,
+      "grad_norm": 0.6254658107874946,
+      "learning_rate": 3.6716433210219437e-06,
+      "loss": 0.618,
+      "step": 1568
+    },
+    {
+      "epoch": 1.8818236352729454,
+      "grad_norm": 0.7162368032094661,
+      "learning_rate": 3.66491119254984e-06,
+      "loss": 0.6344,
+      "step": 1569
+    },
+    {
+      "epoch": 1.8830233953209357,
+      "grad_norm": 0.5923415229221276,
+      "learning_rate": 3.6581816692168454e-06,
+      "loss": 0.6638,
+      "step": 1570
+    },
+    {
+      "epoch": 1.8842231553689262,
+      "grad_norm": 0.5789337203250047,
+      "learning_rate": 3.651454764154182e-06,
+      "loss": 0.6832,
+      "step": 1571
+    },
+    {
+      "epoch": 1.8854229154169166,
+      "grad_norm": 0.6858589129555117,
+      "learning_rate": 3.644730490487961e-06,
+      "loss": 0.6212,
+      "step": 1572
+    },
+    {
+      "epoch": 1.8866226754649071,
+      "grad_norm": 0.7585622324613859,
+      "learning_rate": 3.638008861339156e-06,
+      "loss": 0.599,
+      "step": 1573
+    },
+    {
+      "epoch": 1.8878224355128974,
+      "grad_norm": 0.6791926906418987,
+      "learning_rate": 3.6312898898235828e-06,
+      "loss": 0.6323,
+      "step": 1574
+    },
+    {
+      "epoch": 1.8890221955608877,
+      "grad_norm": 0.6789403422597844,
+      "learning_rate": 3.6245735890518746e-06,
+      "loss": 0.647,
+      "step": 1575
+    },
+    {
+      "epoch": 1.8902219556088782,
+      "grad_norm": 0.6475201273403531,
+      "learning_rate": 3.6178599721294476e-06,
+      "loss": 0.5911,
+      "step": 1576
+    },
+    {
+      "epoch": 1.8914217156568687,
+      "grad_norm": 0.9632926156368794,
+      "learning_rate": 3.611149052156483e-06,
+      "loss": 0.6901,
+      "step": 1577
+    },
+    {
+      "epoch": 1.8926214757048592,
+      "grad_norm": 0.9992430340877584,
+      "learning_rate": 3.604440842227904e-06,
+      "loss": 0.6734,
+      "step": 1578
+    },
+    {
+      "epoch": 1.8938212357528494,
+      "grad_norm": 0.5662345641902004,
+      "learning_rate": 3.5977353554333373e-06,
+      "loss": 0.6472,
+      "step": 1579
+    },
+    {
+      "epoch": 1.8950209958008397,
+      "grad_norm": 0.6833948607274497,
+      "learning_rate": 3.591032604857102e-06,
+      "loss": 0.618,
+      "step": 1580
+    },
+    {
+      "epoch": 1.8962207558488302,
+      "grad_norm": 0.5504537043144545,
+      "learning_rate": 3.584332603578178e-06,
+      "loss": 0.6248,
+      "step": 1581
+    },
+    {
+      "epoch": 1.8974205158968207,
+      "grad_norm": 0.5973518111867986,
+      "learning_rate": 3.5776353646701757e-06,
+      "loss": 0.6231,
+      "step": 1582
+    },
+    {
+      "epoch": 1.8986202759448112,
+      "grad_norm": 0.6890278607358444,
+      "learning_rate": 3.5709409012013207e-06,
+      "loss": 0.6166,
+      "step": 1583
+    },
+    {
+      "epoch": 1.8998200359928015,
+      "grad_norm": 0.6893800529744167,
+      "learning_rate": 3.564249226234423e-06,
+      "loss": 0.577,
+      "step": 1584
+    },
+    {
+      "epoch": 1.9010197960407917,
+      "grad_norm": 0.6950537843364714,
+      "learning_rate": 3.5575603528268454e-06,
+      "loss": 0.654,
+      "step": 1585
+    },
+    {
+      "epoch": 1.9022195560887822,
+      "grad_norm": 0.6343195501551226,
+      "learning_rate": 3.5508742940304893e-06,
+      "loss": 0.6121,
+      "step": 1586
+    },
+    {
+      "epoch": 1.9034193161367727,
+      "grad_norm": 0.6572797354811389,
+      "learning_rate": 3.5441910628917646e-06,
+      "loss": 0.6396,
+      "step": 1587
+    },
+    {
+      "epoch": 1.904619076184763,
+      "grad_norm": 0.7168995016643164,
+      "learning_rate": 3.5375106724515594e-06,
+      "loss": 0.6449,
+      "step": 1588
+    },
+    {
+      "epoch": 1.9058188362327535,
+      "grad_norm": 0.6522301267570335,
+      "learning_rate": 3.5308331357452206e-06,
+      "loss": 0.6397,
+      "step": 1589
+    },
+    {
+      "epoch": 1.9070185962807438,
+      "grad_norm": 0.5185299455083228,
+      "learning_rate": 3.524158465802531e-06,
+      "loss": 0.5938,
+      "step": 1590
+    },
+    {
+      "epoch": 1.9082183563287343,
+      "grad_norm": 0.6507834551656557,
+      "learning_rate": 3.517486675647672e-06,
+      "loss": 0.6077,
+      "step": 1591
+    },
+    {
+      "epoch": 1.9094181163767248,
+      "grad_norm": 0.5007079150029254,
+      "learning_rate": 3.5108177782992112e-06,
+      "loss": 0.5797,
+      "step": 1592
+    },
+    {
+      "epoch": 1.910617876424715,
+      "grad_norm": 0.6518483853622233,
+      "learning_rate": 3.5041517867700686e-06,
+      "loss": 0.6257,
+      "step": 1593
+    },
+    {
+      "epoch": 1.9118176364727053,
+      "grad_norm": 0.6171506924153792,
+      "learning_rate": 3.497488714067494e-06,
+      "loss": 0.6237,
+      "step": 1594
+    },
+    {
+      "epoch": 1.9130173965206958,
+      "grad_norm": 0.5877533330592362,
+      "learning_rate": 3.4908285731930467e-06,
+      "loss": 0.6141,
+      "step": 1595
+    },
+    {
+      "epoch": 1.9142171565686863,
+      "grad_norm": 0.6652036567138812,
+      "learning_rate": 3.484171377142559e-06,
+      "loss": 0.6327,
+      "step": 1596
+    },
+    {
+      "epoch": 1.9154169166166768,
+      "grad_norm": 0.526253481313353,
+      "learning_rate": 3.4775171389061193e-06,
+      "loss": 0.6375,
+      "step": 1597
+    },
+    {
+      "epoch": 1.916616676664667,
+      "grad_norm": 0.7534939416117871,
+      "learning_rate": 3.470865871468046e-06,
+      "loss": 0.6776,
+      "step": 1598
+    },
+    {
+      "epoch": 1.9178164367126573,
+      "grad_norm": 0.6263147900496997,
+      "learning_rate": 3.4642175878068562e-06,
+      "loss": 0.6927,
+      "step": 1599
+    },
+    {
+      "epoch": 1.9190161967606478,
+      "grad_norm": 0.7617577055615982,
+      "learning_rate": 3.4575723008952503e-06,
+      "loss": 0.6378,
+      "step": 1600
+    },
+    {
+      "epoch": 1.9202159568086383,
+      "grad_norm": 0.6064889940890927,
+      "learning_rate": 3.45093002370008e-06,
+      "loss": 0.6825,
+      "step": 1601
+    },
+    {
+      "epoch": 1.9214157168566288,
+      "grad_norm": 0.8528410463243201,
+      "learning_rate": 3.4442907691823193e-06,
+      "loss": 0.6226,
+      "step": 1602
+    },
+    {
+      "epoch": 1.922615476904619,
+      "grad_norm": 0.5302074932770431,
+      "learning_rate": 3.437654550297049e-06,
+      "loss": 0.6454,
+      "step": 1603
+    },
+    {
+      "epoch": 1.9238152369526094,
+      "grad_norm": 0.5529349035112056,
+      "learning_rate": 3.431021379993428e-06,
+      "loss": 0.6355,
+      "step": 1604
+    },
+    {
+      "epoch": 1.9250149970005999,
+      "grad_norm": 0.7120435058914695,
+      "learning_rate": 3.4243912712146606e-06,
+      "loss": 0.6604,
+      "step": 1605
+    },
+    {
+      "epoch": 1.9262147570485904,
+      "grad_norm": 0.7601134981983861,
+      "learning_rate": 3.417764236897979e-06,
+      "loss": 0.6125,
+      "step": 1606
+    },
+    {
+      "epoch": 1.9274145170965808,
+      "grad_norm": 0.7066131442900028,
+      "learning_rate": 3.4111402899746226e-06,
+      "loss": 0.6166,
+      "step": 1607
+    },
+    {
+      "epoch": 1.9286142771445711,
+      "grad_norm": 0.8372429490650506,
+      "learning_rate": 3.404519443369798e-06,
+      "loss": 0.611,
+      "step": 1608
+    },
+    {
+      "epoch": 1.9298140371925614,
+      "grad_norm": 0.5658543889572308,
+      "learning_rate": 3.3979017100026645e-06,
+      "loss": 0.6069,
+      "step": 1609
+    },
+    {
+      "epoch": 1.9310137972405519,
+      "grad_norm": 0.6589818674295754,
+      "learning_rate": 3.391287102786312e-06,
+      "loss": 0.5221,
+      "step": 1610
+    },
+    {
+      "epoch": 1.9322135572885424,
+      "grad_norm": 0.6756834191950231,
+      "learning_rate": 3.384675634627721e-06,
+      "loss": 0.6089,
+      "step": 1611
+    },
+    {
+      "epoch": 1.9334133173365327,
+      "grad_norm": 0.63275711529377,
+      "learning_rate": 3.3780673184277547e-06,
+      "loss": 0.6663,
+      "step": 1612
+    },
+    {
+      "epoch": 1.9346130773845231,
+      "grad_norm": 0.8138389230087799,
+      "learning_rate": 3.3714621670811236e-06,
+      "loss": 0.651,
+      "step": 1613
+    },
+    {
+      "epoch": 1.9358128374325134,
+      "grad_norm": 0.6481188306140845,
+      "learning_rate": 3.364860193476359e-06,
+      "loss": 0.6578,
+      "step": 1614
+    },
+    {
+      "epoch": 1.937012597480504,
+      "grad_norm": 0.5926771545560223,
+      "learning_rate": 3.358261410495799e-06,
+      "loss": 0.5994,
+      "step": 1615
+    },
+    {
+      "epoch": 1.9382123575284944,
+      "grad_norm": 0.5468743795966091,
+      "learning_rate": 3.3516658310155493e-06,
+      "loss": 0.6623,
+      "step": 1616
+    },
+    {
+      "epoch": 1.9394121175764847,
+      "grad_norm": 0.8235416182606894,
+      "learning_rate": 3.3450734679054663e-06,
+      "loss": 0.5928,
+      "step": 1617
+    },
+    {
+      "epoch": 1.940611877624475,
+      "grad_norm": 0.7211676880287002,
+      "learning_rate": 3.3384843340291308e-06,
+      "loss": 0.6084,
+      "step": 1618
+    },
+    {
+      "epoch": 1.9418116376724655,
+      "grad_norm": 0.6389688609459044,
+      "learning_rate": 3.331898442243826e-06,
+      "loss": 0.5666,
+      "step": 1619
+    },
+    {
+      "epoch": 1.943011397720456,
+      "grad_norm": 0.659744054555252,
+      "learning_rate": 3.3253158054005015e-06,
+      "loss": 0.6762,
+      "step": 1620
+    },
+    {
+      "epoch": 1.9442111577684464,
+      "grad_norm": 0.7401291382849171,
+      "learning_rate": 3.3187364363437604e-06,
+      "loss": 0.599,
+      "step": 1621
+    },
+    {
+      "epoch": 1.9454109178164367,
+      "grad_norm": 0.5857916376908441,
+      "learning_rate": 3.3121603479118315e-06,
+      "loss": 0.625,
+      "step": 1622
+    },
+    {
+      "epoch": 1.946610677864427,
+      "grad_norm": 0.6566329837613745,
+      "learning_rate": 3.3055875529365346e-06,
+      "loss": 0.613,
+      "step": 1623
+    },
+    {
+      "epoch": 1.9478104379124175,
+      "grad_norm": 0.4807287603523042,
+      "learning_rate": 3.2990180642432712e-06,
+      "loss": 0.6251,
+      "step": 1624
+    },
+    {
+      "epoch": 1.949010197960408,
+      "grad_norm": 0.6483447443970377,
+      "learning_rate": 3.292451894650986e-06,
+      "loss": 0.6355,
+      "step": 1625
+    },
+    {
+      "epoch": 1.9502099580083985,
+      "grad_norm": 0.6429062695240559,
+      "learning_rate": 3.285889056972148e-06,
+      "loss": 0.6118,
+      "step": 1626
+    },
+    {
+      "epoch": 1.9514097180563887,
+      "grad_norm": 0.7961405043126901,
+      "learning_rate": 3.279329564012731e-06,
+      "loss": 0.5941,
+      "step": 1627
+    },
+    {
+      "epoch": 1.952609478104379,
+      "grad_norm": 0.6899134331766983,
+      "learning_rate": 3.272773428572169e-06,
+      "loss": 0.6666,
+      "step": 1628
+    },
+    {
+      "epoch": 1.9538092381523695,
+      "grad_norm": 0.5947428608906596,
+      "learning_rate": 3.266220663443358e-06,
+      "loss": 0.6597,
+      "step": 1629
+    },
+    {
+      "epoch": 1.95500899820036,
+      "grad_norm": 0.8022747878699585,
+      "learning_rate": 3.259671281412612e-06,
+      "loss": 0.6407,
+      "step": 1630
+    },
+    {
+      "epoch": 1.9562087582483503,
+      "grad_norm": 0.6435528226556113,
+      "learning_rate": 3.253125295259639e-06,
+      "loss": 0.6103,
+      "step": 1631
+    },
+    {
+      "epoch": 1.9574085182963408,
+      "grad_norm": 0.6160566369973391,
+      "learning_rate": 3.2465827177575302e-06,
+      "loss": 0.7168,
+      "step": 1632
+    },
+    {
+      "epoch": 1.958608278344331,
+      "grad_norm": 0.6646541607513289,
+      "learning_rate": 3.2400435616727198e-06,
+      "loss": 0.6397,
+      "step": 1633
+    },
+    {
+      "epoch": 1.9598080383923215,
+      "grad_norm": 0.6301402280538672,
+      "learning_rate": 3.2335078397649646e-06,
+      "loss": 0.5901,
+      "step": 1634
+    },
+    {
+      "epoch": 1.961007798440312,
+      "grad_norm": 0.7735228178846464,
+      "learning_rate": 3.226975564787322e-06,
+      "loss": 0.6473,
+      "step": 1635
+    },
+    {
+      "epoch": 1.9622075584883023,
+      "grad_norm": 0.5614535765622817,
+      "learning_rate": 3.220446749486128e-06,
+      "loss": 0.6503,
+      "step": 1636
+    },
+    {
+      "epoch": 1.9634073185362926,
+      "grad_norm": 0.5106206875601105,
+      "learning_rate": 3.2139214066009595e-06,
+      "loss": 0.7028,
+      "step": 1637
+    },
+    {
+      "epoch": 1.964607078584283,
+      "grad_norm": 0.7156699211997962,
+      "learning_rate": 3.2073995488646225e-06,
+      "loss": 0.6614,
+      "step": 1638
+    },
+    {
+      "epoch": 1.9658068386322736,
+      "grad_norm": 0.642822707358646,
+      "learning_rate": 3.200881189003127e-06,
+      "loss": 0.6372,
+      "step": 1639
+    },
+    {
+      "epoch": 1.967006598680264,
+      "grad_norm": 0.655597372390054,
+      "learning_rate": 3.1943663397356437e-06,
+      "loss": 0.6272,
+      "step": 1640
+    },
+    {
+      "epoch": 1.9682063587282543,
+      "grad_norm": 0.7004155570951615,
+      "learning_rate": 3.187855013774508e-06,
+      "loss": 0.625,
+      "step": 1641
+    },
+    {
+      "epoch": 1.9694061187762446,
+      "grad_norm": 0.5392430860345746,
+      "learning_rate": 3.1813472238251742e-06,
+      "loss": 0.6249,
+      "step": 1642
+    },
+    {
+      "epoch": 1.970605878824235,
+      "grad_norm": 0.69825795555644,
+      "learning_rate": 3.1748429825861937e-06,
+      "loss": 0.659,
+      "step": 1643
+    },
+    {
+      "epoch": 1.9718056388722256,
+      "grad_norm": 0.5836952294843116,
+      "learning_rate": 3.1683423027491987e-06,
+      "loss": 0.6314,
+      "step": 1644
+    },
+    {
+      "epoch": 1.973005398920216,
+      "grad_norm": 0.7106771954157693,
+      "learning_rate": 3.1618451969988716e-06,
+      "loss": 0.6216,
+      "step": 1645
+    },
+    {
+      "epoch": 1.9742051589682064,
+      "grad_norm": 0.6079799375663077,
+      "learning_rate": 3.1553516780129157e-06,
+      "loss": 0.6255,
+      "step": 1646
+    },
+    {
+      "epoch": 1.9754049190161966,
+      "grad_norm": 0.6533033933369583,
+      "learning_rate": 3.1488617584620396e-06,
+      "loss": 0.6016,
+      "step": 1647
+    },
+    {
+      "epoch": 1.9766046790641871,
+      "grad_norm": 0.647873061009936,
+      "learning_rate": 3.1423754510099304e-06,
+      "loss": 0.6823,
+      "step": 1648
+    },
+    {
+      "epoch": 1.9778044391121776,
+      "grad_norm": 0.6100569670661736,
+      "learning_rate": 3.13589276831322e-06,
+      "loss": 0.6436,
+      "step": 1649
+    },
+    {
+      "epoch": 1.9790041991601681,
+      "grad_norm": 0.7053313473498782,
+      "learning_rate": 3.129413723021474e-06,
+      "loss": 0.6495,
+      "step": 1650
+    },
+    {
+      "epoch": 1.9802039592081584,
+      "grad_norm": 0.6057256781443613,
+      "learning_rate": 3.1229383277771595e-06,
+      "loss": 0.6333,
+      "step": 1651
+    },
+    {
+      "epoch": 1.9814037192561487,
+      "grad_norm": 0.6372818155791644,
+      "learning_rate": 3.116466595215617e-06,
+      "loss": 0.6111,
+      "step": 1652
+    },
+    {
+      "epoch": 1.9826034793041392,
+      "grad_norm": 0.7174431458091668,
+      "learning_rate": 3.1099985379650454e-06,
+      "loss": 0.6921,
+      "step": 1653
+    },
+    {
+      "epoch": 1.9838032393521297,
+      "grad_norm": 0.7915955539043774,
+      "learning_rate": 3.1035341686464664e-06,
+      "loss": 0.6267,
+      "step": 1654
+    },
+    {
+      "epoch": 1.98500299940012,
+      "grad_norm": 0.6050217020619578,
+      "learning_rate": 3.0970734998737095e-06,
+      "loss": 0.6578,
+      "step": 1655
+    },
+    {
+      "epoch": 1.9862027594481104,
+      "grad_norm": 0.7268494061378847,
+      "learning_rate": 3.090616544253385e-06,
+      "loss": 0.5408,
+      "step": 1656
+    },
+    {
+      "epoch": 1.9874025194961007,
+      "grad_norm": 0.6127371326001422,
+      "learning_rate": 3.084163314384852e-06,
+      "loss": 0.6581,
+      "step": 1657
+    },
+    {
+      "epoch": 1.9886022795440912,
+      "grad_norm": 0.6323484798333604,
+      "learning_rate": 3.0777138228602045e-06,
+      "loss": 0.6357,
+      "step": 1658
+    },
+    {
+      "epoch": 1.9898020395920817,
+      "grad_norm": 0.7300351753850884,
+      "learning_rate": 3.071268082264241e-06,
+      "loss": 0.6547,
+      "step": 1659
+    },
+    {
+      "epoch": 1.991001799640072,
+      "grad_norm": 0.41278449269264283,
+      "learning_rate": 3.0648261051744367e-06,
+      "loss": 0.6348,
+      "step": 1660
+    },
+    {
+      "epoch": 1.9922015596880622,
+      "grad_norm": 0.6580197911643658,
+      "learning_rate": 3.05838790416093e-06,
+      "loss": 0.6357,
+      "step": 1661
+    },
+    {
+      "epoch": 1.9934013197360527,
+      "grad_norm": 0.813218363737704,
+      "learning_rate": 3.051953491786487e-06,
+      "loss": 0.5861,
+      "step": 1662
+    },
+    {
+      "epoch": 1.9946010797840432,
+      "grad_norm": 0.5836776878674278,
+      "learning_rate": 3.04552288060648e-06,
+      "loss": 0.6315,
+      "step": 1663
+    },
+    {
+      "epoch": 1.9958008398320337,
+      "grad_norm": 0.9820496254307773,
+      "learning_rate": 3.0390960831688676e-06,
+      "loss": 0.6234,
+      "step": 1664
+    },
+    {
+      "epoch": 1.997000599880024,
+      "grad_norm": 0.6584441414557831,
+      "learning_rate": 3.0326731120141663e-06,
+      "loss": 0.5642,
+      "step": 1665
+    },
+    {
+      "epoch": 1.9982003599280143,
+      "grad_norm": 0.4613016565652139,
+      "learning_rate": 3.026253979675421e-06,
+      "loss": 0.6433,
+      "step": 1666
+    },
+    {
+      "epoch": 1.9994001199760048,
+      "grad_norm": 0.5993227231828745,
+      "learning_rate": 3.019838698678191e-06,
+      "loss": 0.6251,
+      "step": 1667
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.1286287068986351,
+      "learning_rate": 3.013427281540523e-06,
+      "loss": 0.6533,
+      "step": 1668
+    },
+    {
+      "epoch": 2.0011997600479905,
+      "grad_norm": 0.6553687343168106,
+      "learning_rate": 3.0070197407729174e-06,
+      "loss": 0.546,
+      "step": 1669
+    },
+    {
+      "epoch": 2.002399520095981,
+      "grad_norm": 0.7232483187926956,
+      "learning_rate": 3.000616088878315e-06,
+      "loss": 0.5026,
+      "step": 1670
+    },
+    {
+      "epoch": 2.003599280143971,
+      "grad_norm": 0.635507837834464,
+      "learning_rate": 2.9942163383520682e-06,
+      "loss": 0.5566,
+      "step": 1671
+    },
+    {
+      "epoch": 2.0047990401919615,
+      "grad_norm": 0.586070052604526,
+      "learning_rate": 2.9878205016819133e-06,
+      "loss": 0.5692,
+      "step": 1672
+    },
+    {
+      "epoch": 2.005998800239952,
+      "grad_norm": 0.5744491184365953,
+      "learning_rate": 2.981428591347955e-06,
+      "loss": 0.6042,
+      "step": 1673
+    },
+    {
+      "epoch": 2.0071985602879425,
+      "grad_norm": 0.5533414219666576,
+      "learning_rate": 2.9750406198226333e-06,
+      "loss": 0.5313,
+      "step": 1674
+    },
+    {
+      "epoch": 2.008398320335933,
+      "grad_norm": 0.527974323446745,
+      "learning_rate": 2.9686565995707006e-06,
+      "loss": 0.5716,
+      "step": 1675
+    },
+    {
+      "epoch": 2.009598080383923,
+      "grad_norm": 0.6581060126537326,
+      "learning_rate": 2.962276543049204e-06,
+      "loss": 0.5912,
+      "step": 1676
+    },
+    {
+      "epoch": 2.0107978404319136,
+      "grad_norm": 0.6284377755426946,
+      "learning_rate": 2.9559004627074535e-06,
+      "loss": 0.6311,
+      "step": 1677
+    },
+    {
+      "epoch": 2.011997600479904,
+      "grad_norm": 0.6136973438415546,
+      "learning_rate": 2.9495283709869993e-06,
+      "loss": 0.563,
+      "step": 1678
+    },
+    {
+      "epoch": 2.0131973605278946,
+      "grad_norm": 0.7255219368682616,
+      "learning_rate": 2.94316028032161e-06,
+      "loss": 0.5456,
+      "step": 1679
+    },
+    {
+      "epoch": 2.0143971205758846,
+      "grad_norm": 0.6531829585887703,
+      "learning_rate": 2.936796203137249e-06,
+      "loss": 0.607,
+      "step": 1680
+    },
+    {
+      "epoch": 2.015596880623875,
+      "grad_norm": 0.5763977931898144,
+      "learning_rate": 2.9304361518520447e-06,
+      "loss": 0.5683,
+      "step": 1681
+    },
+    {
+      "epoch": 2.0167966406718656,
+      "grad_norm": 0.5910528949546794,
+      "learning_rate": 2.9240801388762707e-06,
+      "loss": 0.5357,
+      "step": 1682
+    },
+    {
+      "epoch": 2.017996400719856,
+      "grad_norm": 0.5898805007247918,
+      "learning_rate": 2.9177281766123254e-06,
+      "loss": 0.4886,
+      "step": 1683
+    },
+    {
+      "epoch": 2.0191961607678466,
+      "grad_norm": 0.5960173782729209,
+      "learning_rate": 2.911380277454695e-06,
+      "loss": 0.5925,
+      "step": 1684
+    },
+    {
+      "epoch": 2.0203959208158366,
+      "grad_norm": 0.6167640235101897,
+      "learning_rate": 2.905036453789942e-06,
+      "loss": 0.5941,
+      "step": 1685
+    },
+    {
+      "epoch": 2.021595680863827,
+      "grad_norm": 0.6138548143862488,
+      "learning_rate": 2.898696717996678e-06,
+      "loss": 0.5909,
+      "step": 1686
+    },
+    {
+      "epoch": 2.0227954409118176,
+      "grad_norm": 0.5839889378518093,
+      "learning_rate": 2.892361082445534e-06,
+      "loss": 0.516,
+      "step": 1687
+    },
+    {
+      "epoch": 2.023995200959808,
+      "grad_norm": 0.7247860919816828,
+      "learning_rate": 2.886029559499144e-06,
+      "loss": 0.5563,
+      "step": 1688
+    },
+    {
+      "epoch": 2.0251949610077986,
+      "grad_norm": 0.7298128666554536,
+      "learning_rate": 2.8797021615121134e-06,
+      "loss": 0.5459,
+      "step": 1689
+    },
+    {
+      "epoch": 2.0263947210557887,
+      "grad_norm": 0.7316860323201141,
+      "learning_rate": 2.8733789008310033e-06,
+      "loss": 0.5754,
+      "step": 1690
+    },
+    {
+      "epoch": 2.027594481103779,
+      "grad_norm": 0.6870139028142205,
+      "learning_rate": 2.8670597897942964e-06,
+      "loss": 0.5671,
+      "step": 1691
+    },
+    {
+      "epoch": 2.0287942411517697,
+      "grad_norm": 0.5965131558498986,
+      "learning_rate": 2.8607448407323824e-06,
+      "loss": 0.5275,
+      "step": 1692
+    },
+    {
+      "epoch": 2.02999400119976,
+      "grad_norm": 0.6436418134198707,
+      "learning_rate": 2.854434065967528e-06,
+      "loss": 0.5991,
+      "step": 1693
+    },
+    {
+      "epoch": 2.0311937612477506,
+      "grad_norm": 0.7620520831649701,
+      "learning_rate": 2.8481274778138567e-06,
+      "loss": 0.5569,
+      "step": 1694
+    },
+    {
+      "epoch": 2.0323935212957407,
+      "grad_norm": 0.620553773544572,
+      "learning_rate": 2.8418250885773156e-06,
+      "loss": 0.6529,
+      "step": 1695
+    },
+    {
+      "epoch": 2.033593281343731,
+      "grad_norm": 0.5900016668327399,
+      "learning_rate": 2.8355269105556695e-06,
+      "loss": 0.5883,
+      "step": 1696
+    },
+    {
+      "epoch": 2.0347930413917217,
+      "grad_norm": 0.5751741742801739,
+      "learning_rate": 2.82923295603846e-06,
+      "loss": 0.54,
+      "step": 1697
+    },
+    {
+      "epoch": 2.035992801439712,
+      "grad_norm": 0.7322012852297586,
+      "learning_rate": 2.822943237306983e-06,
+      "loss": 0.5295,
+      "step": 1698
+    },
+    {
+      "epoch": 2.0371925614877027,
+      "grad_norm": 0.6465495010276673,
+      "learning_rate": 2.8166577666342786e-06,
+      "loss": 0.5136,
+      "step": 1699
+    },
+    {
+      "epoch": 2.0383923215356927,
+      "grad_norm": 0.5951509571107764,
+      "learning_rate": 2.8103765562850942e-06,
+      "loss": 0.5782,
+      "step": 1700
+    },
+    {
+      "epoch": 2.039592081583683,
+      "grad_norm": 0.7817384358122911,
+      "learning_rate": 2.804099618515858e-06,
+      "loss": 0.5821,
+      "step": 1701
+    },
+    {
+      "epoch": 2.0407918416316737,
+      "grad_norm": 0.6031361130877024,
+      "learning_rate": 2.797826965574667e-06,
+      "loss": 0.5593,
+      "step": 1702
+    },
+    {
+      "epoch": 2.041991601679664,
+      "grad_norm": 0.5957045599311164,
+      "learning_rate": 2.7915586097012613e-06,
+      "loss": 0.5458,
+      "step": 1703
+    },
+    {
+      "epoch": 2.0431913617276543,
+      "grad_norm": 0.6360695872826579,
+      "learning_rate": 2.785294563126987e-06,
+      "loss": 0.5656,
+      "step": 1704
+    },
+    {
+      "epoch": 2.0443911217756447,
+      "grad_norm": 0.5655789747161657,
+      "learning_rate": 2.7790348380747832e-06,
+      "loss": 0.5162,
+      "step": 1705
+    },
+    {
+      "epoch": 2.0455908818236352,
+      "grad_norm": 0.6266463387114029,
+      "learning_rate": 2.7727794467591682e-06,
+      "loss": 0.5836,
+      "step": 1706
+    },
+    {
+      "epoch": 2.0467906418716257,
+      "grad_norm": 0.6397476354687799,
+      "learning_rate": 2.766528401386187e-06,
+      "loss": 0.5469,
+      "step": 1707
+    },
+    {
+      "epoch": 2.0479904019196162,
+      "grad_norm": 0.6424712078878715,
+      "learning_rate": 2.760281714153414e-06,
+      "loss": 0.6244,
+      "step": 1708
+    },
+    {
+      "epoch": 2.0491901619676063,
+      "grad_norm": 0.6454459575829394,
+      "learning_rate": 2.7540393972499203e-06,
+      "loss": 0.5179,
+      "step": 1709
+    },
+    {
+      "epoch": 2.0503899220155968,
+      "grad_norm": 0.6701297156785533,
+      "learning_rate": 2.747801462856244e-06,
+      "loss": 0.5488,
+      "step": 1710
+    },
+    {
+      "epoch": 2.0515896820635873,
+      "grad_norm": 0.6389965775957791,
+      "learning_rate": 2.741567923144376e-06,
+      "loss": 0.5138,
+      "step": 1711
+    },
+    {
+      "epoch": 2.0527894421115778,
+      "grad_norm": 0.63361030302146,
+      "learning_rate": 2.7353387902777304e-06,
+      "loss": 0.5385,
+      "step": 1712
+    },
+    {
+      "epoch": 2.0539892021595683,
+      "grad_norm": 0.6214392044652081,
+      "learning_rate": 2.7291140764111223e-06,
+      "loss": 0.6765,
+      "step": 1713
+    },
+    {
+      "epoch": 2.0551889622075583,
+      "grad_norm": 0.6284153913170616,
+      "learning_rate": 2.722893793690744e-06,
+      "loss": 0.6256,
+      "step": 1714
+    },
+    {
+      "epoch": 2.056388722255549,
+      "grad_norm": 0.6045362882434572,
+      "learning_rate": 2.716677954254141e-06,
+      "loss": 0.5945,
+      "step": 1715
+    },
+    {
+      "epoch": 2.0575884823035393,
+      "grad_norm": 0.7832123459310987,
+      "learning_rate": 2.7104665702301895e-06,
+      "loss": 0.5517,
+      "step": 1716
+    },
+    {
+      "epoch": 2.05878824235153,
+      "grad_norm": 0.599016337999235,
+      "learning_rate": 2.704259653739074e-06,
+      "loss": 0.6268,
+      "step": 1717
+    },
+    {
+      "epoch": 2.0599880023995203,
+      "grad_norm": 0.594891933671642,
+      "learning_rate": 2.6980572168922555e-06,
+      "loss": 0.5233,
+      "step": 1718
+    },
+    {
+      "epoch": 2.0611877624475103,
+      "grad_norm": 0.6773121292151333,
+      "learning_rate": 2.6918592717924563e-06,
+      "loss": 0.5777,
+      "step": 1719
+    },
+    {
+      "epoch": 2.062387522495501,
+      "grad_norm": 0.7031827325029097,
+      "learning_rate": 2.685665830533642e-06,
+      "loss": 0.5999,
+      "step": 1720
+    },
+    {
+      "epoch": 2.0635872825434913,
+      "grad_norm": 0.6371739899003336,
+      "learning_rate": 2.679476905200977e-06,
+      "loss": 0.5485,
+      "step": 1721
+    },
+    {
+      "epoch": 2.064787042591482,
+      "grad_norm": 0.6566325322148608,
+      "learning_rate": 2.6732925078708205e-06,
+      "loss": 0.5708,
+      "step": 1722
+    },
+    {
+      "epoch": 2.0659868026394723,
+      "grad_norm": 0.5948043795465717,
+      "learning_rate": 2.667112650610702e-06,
+      "loss": 0.6146,
+      "step": 1723
+    },
+    {
+      "epoch": 2.0671865626874624,
+      "grad_norm": 0.5999379851132288,
+      "learning_rate": 2.6609373454792797e-06,
+      "loss": 0.588,
+      "step": 1724
+    },
+    {
+      "epoch": 2.068386322735453,
+      "grad_norm": 0.6423600146443728,
+      "learning_rate": 2.6547666045263365e-06,
+      "loss": 0.5946,
+      "step": 1725
+    },
+    {
+      "epoch": 2.0695860827834434,
+      "grad_norm": 0.5936111859457325,
+      "learning_rate": 2.64860043979275e-06,
+      "loss": 0.5742,
+      "step": 1726
+    },
+    {
+      "epoch": 2.070785842831434,
+      "grad_norm": 0.6003496127801714,
+      "learning_rate": 2.6424388633104646e-06,
+      "loss": 0.6131,
+      "step": 1727
+    },
+    {
+      "epoch": 2.071985602879424,
+      "grad_norm": 0.5663334196056212,
+      "learning_rate": 2.6362818871024754e-06,
+      "loss": 0.6251,
+      "step": 1728
+    },
+    {
+      "epoch": 2.0731853629274144,
+      "grad_norm": 0.5997400925623815,
+      "learning_rate": 2.6301295231827983e-06,
+      "loss": 0.5446,
+      "step": 1729
+    },
+    {
+      "epoch": 2.074385122975405,
+      "grad_norm": 0.6463961773967529,
+      "learning_rate": 2.6239817835564514e-06,
+      "loss": 0.5936,
+      "step": 1730
+    },
+    {
+      "epoch": 2.0755848830233954,
+      "grad_norm": 0.6755373596792922,
+      "learning_rate": 2.6178386802194267e-06,
+      "loss": 0.5836,
+      "step": 1731
+    },
+    {
+      "epoch": 2.076784643071386,
+      "grad_norm": 0.6141145857607571,
+      "learning_rate": 2.611700225158673e-06,
+      "loss": 0.5602,
+      "step": 1732
+    },
+    {
+      "epoch": 2.077984403119376,
+      "grad_norm": 0.6293863917640159,
+      "learning_rate": 2.6055664303520655e-06,
+      "loss": 0.586,
+      "step": 1733
+    },
+    {
+      "epoch": 2.0791841631673664,
+      "grad_norm": 0.6042059842491126,
+      "learning_rate": 2.599437307768389e-06,
+      "loss": 0.5728,
+      "step": 1734
+    },
+    {
+      "epoch": 2.080383923215357,
+      "grad_norm": 0.6425208877547268,
+      "learning_rate": 2.5933128693673093e-06,
+      "loss": 0.5447,
+      "step": 1735
+    },
+    {
+      "epoch": 2.0815836832633474,
+      "grad_norm": 0.6252508061699878,
+      "learning_rate": 2.5871931270993533e-06,
+      "loss": 0.5833,
+      "step": 1736
+    },
+    {
+      "epoch": 2.082783443311338,
+      "grad_norm": 0.5753613975248165,
+      "learning_rate": 2.581078092905883e-06,
+      "loss": 0.5572,
+      "step": 1737
+    },
+    {
+      "epoch": 2.083983203359328,
+      "grad_norm": 0.6190082489851423,
+      "learning_rate": 2.5749677787190773e-06,
+      "loss": 0.5895,
+      "step": 1738
+    },
+    {
+      "epoch": 2.0851829634073185,
+      "grad_norm": 0.6483439527059782,
+      "learning_rate": 2.568862196461896e-06,
+      "loss": 0.5809,
+      "step": 1739
+    },
+    {
+      "epoch": 2.086382723455309,
+      "grad_norm": 0.6049753604492797,
+      "learning_rate": 2.5627613580480777e-06,
+      "loss": 0.5745,
+      "step": 1740
+    },
+    {
+      "epoch": 2.0875824835032994,
+      "grad_norm": 0.6019511442996845,
+      "learning_rate": 2.5566652753820986e-06,
+      "loss": 0.5828,
+      "step": 1741
+    },
+    {
+      "epoch": 2.08878224355129,
+      "grad_norm": 0.5637693698154113,
+      "learning_rate": 2.5505739603591485e-06,
+      "loss": 0.5205,
+      "step": 1742
+    },
+    {
+      "epoch": 2.08998200359928,
+      "grad_norm": 0.5664060694906904,
+      "learning_rate": 2.544487424865127e-06,
+      "loss": 0.5925,
+      "step": 1743
+    },
+    {
+      "epoch": 2.0911817636472705,
+      "grad_norm": 0.7064759661824541,
+      "learning_rate": 2.5384056807766013e-06,
+      "loss": 0.5034,
+      "step": 1744
+    },
+    {
+      "epoch": 2.092381523695261,
+      "grad_norm": 0.6981691256033558,
+      "learning_rate": 2.5323287399607844e-06,
+      "loss": 0.6892,
+      "step": 1745
+    },
+    {
+      "epoch": 2.0935812837432515,
+      "grad_norm": 0.7045947459947126,
+      "learning_rate": 2.526256614275524e-06,
+      "loss": 0.6233,
+      "step": 1746
+    },
+    {
+      "epoch": 2.094781043791242,
+      "grad_norm": 0.648809316236553,
+      "learning_rate": 2.5201893155692687e-06,
+      "loss": 0.6659,
+      "step": 1747
+    },
+    {
+      "epoch": 2.095980803839232,
+      "grad_norm": 0.662631188788291,
+      "learning_rate": 2.5141268556810495e-06,
+      "loss": 0.5075,
+      "step": 1748
+    },
+    {
+      "epoch": 2.0971805638872225,
+      "grad_norm": 0.6193505883910716,
+      "learning_rate": 2.5080692464404557e-06,
+      "loss": 0.5616,
+      "step": 1749
+    },
+    {
+      "epoch": 2.098380323935213,
+      "grad_norm": 0.7122822339221859,
+      "learning_rate": 2.5020164996676087e-06,
+      "loss": 0.5347,
+      "step": 1750
+    },
+    {
+      "epoch": 2.0995800839832035,
+      "grad_norm": 0.5625146971918583,
+      "learning_rate": 2.4959686271731472e-06,
+      "loss": 0.5681,
+      "step": 1751
+    },
+    {
+      "epoch": 2.1007798440311936,
+      "grad_norm": 0.6549716986169428,
+      "learning_rate": 2.489925640758193e-06,
+      "loss": 0.5175,
+      "step": 1752
+    },
+    {
+      "epoch": 2.101979604079184,
+      "grad_norm": 0.577542516177496,
+      "learning_rate": 2.483887552214338e-06,
+      "loss": 0.6132,
+      "step": 1753
+    },
+    {
+      "epoch": 2.1031793641271745,
+      "grad_norm": 0.6344390216058583,
+      "learning_rate": 2.4778543733236144e-06,
+      "loss": 0.5836,
+      "step": 1754
+    },
+    {
+      "epoch": 2.104379124175165,
+      "grad_norm": 0.6597128044852273,
+      "learning_rate": 2.471826115858478e-06,
+      "loss": 0.5651,
+      "step": 1755
+    },
+    {
+      "epoch": 2.1055788842231555,
+      "grad_norm": 0.6004318635958817,
+      "learning_rate": 2.4658027915817717e-06,
+      "loss": 0.562,
+      "step": 1756
+    },
+    {
+      "epoch": 2.1067786442711456,
+      "grad_norm": 0.6826868530541879,
+      "learning_rate": 2.4597844122467267e-06,
+      "loss": 0.5516,
+      "step": 1757
+    },
+    {
+      "epoch": 2.107978404319136,
+      "grad_norm": 0.6731413540794284,
+      "learning_rate": 2.453770989596917e-06,
+      "loss": 0.5182,
+      "step": 1758
+    },
+    {
+      "epoch": 2.1091781643671266,
+      "grad_norm": 0.6275693290935841,
+      "learning_rate": 2.44776253536624e-06,
+      "loss": 0.5585,
+      "step": 1759
+    },
+    {
+      "epoch": 2.110377924415117,
+      "grad_norm": 0.5874746966605109,
+      "learning_rate": 2.4417590612789094e-06,
+      "loss": 0.5465,
+      "step": 1760
+    },
+    {
+      "epoch": 2.1115776844631076,
+      "grad_norm": 0.5974898845267843,
+      "learning_rate": 2.4357605790494158e-06,
+      "loss": 0.6207,
+      "step": 1761
+    },
+    {
+      "epoch": 2.1127774445110976,
+      "grad_norm": 0.6061825734549284,
+      "learning_rate": 2.4297671003825058e-06,
+      "loss": 0.4986,
+      "step": 1762
+    },
+    {
+      "epoch": 2.113977204559088,
+      "grad_norm": 0.6439695766776677,
+      "learning_rate": 2.4237786369731655e-06,
+      "loss": 0.5817,
+      "step": 1763
+    },
+    {
+      "epoch": 2.1151769646070786,
+      "grad_norm": 0.6447339343313097,
+      "learning_rate": 2.4177952005066006e-06,
+      "loss": 0.5764,
+      "step": 1764
+    },
+    {
+      "epoch": 2.116376724655069,
+      "grad_norm": 0.6139730437589196,
+      "learning_rate": 2.4118168026581976e-06,
+      "loss": 0.5558,
+      "step": 1765
+    },
+    {
+      "epoch": 2.1175764847030596,
+      "grad_norm": 0.5680150493815513,
+      "learning_rate": 2.4058434550935163e-06,
+      "loss": 0.6039,
+      "step": 1766
+    },
+    {
+      "epoch": 2.1187762447510496,
+      "grad_norm": 0.6345243664025577,
+      "learning_rate": 2.3998751694682666e-06,
+      "loss": 0.5321,
+      "step": 1767
+    },
+    {
+      "epoch": 2.11997600479904,
+      "grad_norm": 0.764950759999074,
+      "learning_rate": 2.3939119574282706e-06,
+      "loss": 0.5613,
+      "step": 1768
+    },
+    {
+      "epoch": 2.1211757648470306,
+      "grad_norm": 0.6816883523166991,
+      "learning_rate": 2.3879538306094587e-06,
+      "loss": 0.5428,
+      "step": 1769
+    },
+    {
+      "epoch": 2.122375524895021,
+      "grad_norm": 0.6669704645020077,
+      "learning_rate": 2.3820008006378355e-06,
+      "loss": 0.6003,
+      "step": 1770
+    },
+    {
+      "epoch": 2.123575284943011,
+      "grad_norm": 0.6234045791660564,
+      "learning_rate": 2.37605287912946e-06,
+      "loss": 0.56,
+      "step": 1771
+    },
+    {
+      "epoch": 2.1247750449910017,
+      "grad_norm": 0.5950638939401581,
+      "learning_rate": 2.370110077690425e-06,
+      "loss": 0.5683,
+      "step": 1772
+    },
+    {
+      "epoch": 2.125974805038992,
+      "grad_norm": 0.6512885551567529,
+      "learning_rate": 2.3641724079168303e-06,
+      "loss": 0.5468,
+      "step": 1773
+    },
+    {
+      "epoch": 2.1271745650869827,
+      "grad_norm": 0.5747798498609408,
+      "learning_rate": 2.3582398813947642e-06,
+      "loss": 0.5684,
+      "step": 1774
+    },
+    {
+      "epoch": 2.128374325134973,
+      "grad_norm": 0.6611106061274544,
+      "learning_rate": 2.352312509700278e-06,
+      "loss": 0.6094,
+      "step": 1775
+    },
+    {
+      "epoch": 2.129574085182963,
+      "grad_norm": 0.6917106334768783,
+      "learning_rate": 2.3463903043993643e-06,
+      "loss": 0.5543,
+      "step": 1776
+    },
+    {
+      "epoch": 2.1307738452309537,
+      "grad_norm": 0.6511261294062615,
+      "learning_rate": 2.3404732770479355e-06,
+      "loss": 0.5143,
+      "step": 1777
+    },
+    {
+      "epoch": 2.131973605278944,
+      "grad_norm": 0.5835121301193615,
+      "learning_rate": 2.3345614391918013e-06,
+      "loss": 0.5327,
+      "step": 1778
+    },
+    {
+      "epoch": 2.1331733653269347,
+      "grad_norm": 0.5995561327066398,
+      "learning_rate": 2.328654802366641e-06,
+      "loss": 0.549,
+      "step": 1779
+    },
+    {
+      "epoch": 2.134373125374925,
+      "grad_norm": 0.6270294499218602,
+      "learning_rate": 2.3227533780979866e-06,
+      "loss": 0.5662,
+      "step": 1780
+    },
+    {
+      "epoch": 2.1355728854229152,
+      "grad_norm": 0.6066283235428948,
+      "learning_rate": 2.3168571779012073e-06,
+      "loss": 0.5755,
+      "step": 1781
+    },
+    {
+      "epoch": 2.1367726454709057,
+      "grad_norm": 0.6018231343746554,
+      "learning_rate": 2.310966213281465e-06,
+      "loss": 0.5028,
+      "step": 1782
+    },
+    {
+      "epoch": 2.1379724055188962,
+      "grad_norm": 0.6771968625455023,
+      "learning_rate": 2.305080495733712e-06,
+      "loss": 0.6507,
+      "step": 1783
+    },
+    {
+      "epoch": 2.1391721655668867,
+      "grad_norm": 0.5914151795227278,
+      "learning_rate": 2.299200036742668e-06,
+      "loss": 0.5812,
+      "step": 1784
+    },
+    {
+      "epoch": 2.140371925614877,
+      "grad_norm": 0.6638265537528982,
+      "learning_rate": 2.2933248477827814e-06,
+      "loss": 0.5296,
+      "step": 1785
+    },
+    {
+      "epoch": 2.1415716856628673,
+      "grad_norm": 0.6743505439103027,
+      "learning_rate": 2.2874549403182224e-06,
+      "loss": 0.6193,
+      "step": 1786
+    },
+    {
+      "epoch": 2.1427714457108578,
+      "grad_norm": 0.6093655720042745,
+      "learning_rate": 2.2815903258028554e-06,
+      "loss": 0.5518,
+      "step": 1787
+    },
+    {
+      "epoch": 2.1439712057588483,
+      "grad_norm": 0.6607671944638438,
+      "learning_rate": 2.2757310156802158e-06,
+      "loss": 0.5959,
+      "step": 1788
+    },
+    {
+      "epoch": 2.1451709658068387,
+      "grad_norm": 0.6864152138905839,
+      "learning_rate": 2.2698770213834904e-06,
+      "loss": 0.4978,
+      "step": 1789
+    },
+    {
+      "epoch": 2.1463707258548292,
+      "grad_norm": 0.6756950013938815,
+      "learning_rate": 2.2640283543354907e-06,
+      "loss": 0.5446,
+      "step": 1790
+    },
+    {
+      "epoch": 2.1475704859028193,
+      "grad_norm": 0.6961267024749692,
+      "learning_rate": 2.258185025948635e-06,
+      "loss": 0.4971,
+      "step": 1791
+    },
+    {
+      "epoch": 2.14877024595081,
+      "grad_norm": 0.5991236172755423,
+      "learning_rate": 2.2523470476249232e-06,
+      "loss": 0.5322,
+      "step": 1792
+    },
+    {
+      "epoch": 2.1499700059988003,
+      "grad_norm": 0.7345552094596932,
+      "learning_rate": 2.2465144307559172e-06,
+      "loss": 0.5695,
+      "step": 1793
+    },
+    {
+      "epoch": 2.1511697660467908,
+      "grad_norm": 0.6706955037760028,
+      "learning_rate": 2.2406871867227163e-06,
+      "loss": 0.5616,
+      "step": 1794
+    },
+    {
+      "epoch": 2.1523695260947813,
+      "grad_norm": 0.5887206055969618,
+      "learning_rate": 2.2348653268959353e-06,
+      "loss": 0.511,
+      "step": 1795
+    },
+    {
+      "epoch": 2.1535692861427713,
+      "grad_norm": 0.7224310469698312,
+      "learning_rate": 2.2290488626356836e-06,
+      "loss": 0.5691,
+      "step": 1796
+    },
+    {
+      "epoch": 2.154769046190762,
+      "grad_norm": 0.7170794492801699,
+      "learning_rate": 2.2232378052915428e-06,
+      "loss": 0.5402,
+      "step": 1797
+    },
+    {
+      "epoch": 2.1559688062387523,
+      "grad_norm": 0.6126640435973594,
+      "learning_rate": 2.2174321662025427e-06,
+      "loss": 0.6472,
+      "step": 1798
+    },
+    {
+      "epoch": 2.157168566286743,
+      "grad_norm": 0.5886685204838491,
+      "learning_rate": 2.2116319566971433e-06,
+      "loss": 0.5853,
+      "step": 1799
+    },
+    {
+      "epoch": 2.158368326334733,
+      "grad_norm": 0.7093491324672906,
+      "learning_rate": 2.2058371880932028e-06,
+      "loss": 0.5987,
+      "step": 1800
+    },
+    {
+      "epoch": 2.1595680863827234,
+      "grad_norm": 0.5759850720608728,
+      "learning_rate": 2.2000478716979727e-06,
+      "loss": 0.608,
+      "step": 1801
+    },
+    {
+      "epoch": 2.160767846430714,
+      "grad_norm": 0.6527246415338012,
+      "learning_rate": 2.194264018808061e-06,
+      "loss": 0.6342,
+      "step": 1802
+    },
+    {
+      "epoch": 2.1619676064787043,
+      "grad_norm": 0.9156170459502739,
+      "learning_rate": 2.188485640709411e-06,
+      "loss": 0.5886,
+      "step": 1803
+    },
+    {
+      "epoch": 2.163167366526695,
+      "grad_norm": 0.5950925395063658,
+      "learning_rate": 2.182712748677287e-06,
+      "loss": 0.537,
+      "step": 1804
+    },
+    {
+      "epoch": 2.164367126574685,
+      "grad_norm": 0.6694812889645247,
+      "learning_rate": 2.176945353976253e-06,
+      "loss": 0.5663,
+      "step": 1805
+    },
+    {
+      "epoch": 2.1655668866226754,
+      "grad_norm": 0.5956168088449492,
+      "learning_rate": 2.171183467860136e-06,
+      "loss": 0.569,
+      "step": 1806
+    },
+    {
+      "epoch": 2.166766646670666,
+      "grad_norm": 0.6008079381234008,
+      "learning_rate": 2.1654271015720213e-06,
+      "loss": 0.5288,
+      "step": 1807
+    },
+    {
+      "epoch": 2.1679664067186564,
+      "grad_norm": 0.702051221306401,
+      "learning_rate": 2.159676266344222e-06,
+      "loss": 0.5867,
+      "step": 1808
+    },
+    {
+      "epoch": 2.169166166766647,
+      "grad_norm": 0.6182868356474166,
+      "learning_rate": 2.1539309733982574e-06,
+      "loss": 0.5342,
+      "step": 1809
+    },
+    {
+      "epoch": 2.170365926814637,
+      "grad_norm": 0.5258453454474016,
+      "learning_rate": 2.148191233944833e-06,
+      "loss": 0.5732,
+      "step": 1810
+    },
+    {
+      "epoch": 2.1715656868626274,
+      "grad_norm": 0.6285807905161801,
+      "learning_rate": 2.1424570591838184e-06,
+      "loss": 0.5281,
+      "step": 1811
+    },
+    {
+      "epoch": 2.172765446910618,
+      "grad_norm": 0.7210098999907698,
+      "learning_rate": 2.136728460304224e-06,
+      "loss": 0.5957,
+      "step": 1812
+    },
+    {
+      "epoch": 2.1739652069586084,
+      "grad_norm": 0.6136648287223856,
+      "learning_rate": 2.1310054484841804e-06,
+      "loss": 0.5206,
+      "step": 1813
+    },
+    {
+      "epoch": 2.175164967006599,
+      "grad_norm": 0.5721710134462407,
+      "learning_rate": 2.1252880348909154e-06,
+      "loss": 0.5455,
+      "step": 1814
+    },
+    {
+      "epoch": 2.176364727054589,
+      "grad_norm": 0.5829787754472344,
+      "learning_rate": 2.1195762306807354e-06,
+      "loss": 0.5988,
+      "step": 1815
+    },
+    {
+      "epoch": 2.1775644871025794,
+      "grad_norm": 0.6073107932285269,
+      "learning_rate": 2.113870046999e-06,
+      "loss": 0.5854,
+      "step": 1816
+    },
+    {
+      "epoch": 2.17876424715057,
+      "grad_norm": 0.6634043222962149,
+      "learning_rate": 2.1081694949800972e-06,
+      "loss": 0.5948,
+      "step": 1817
+    },
+    {
+      "epoch": 2.1799640071985604,
+      "grad_norm": 0.5952194802441421,
+      "learning_rate": 2.102474585747435e-06,
+      "loss": 0.566,
+      "step": 1818
+    },
+    {
+      "epoch": 2.1811637672465505,
+      "grad_norm": 0.6401399556311126,
+      "learning_rate": 2.096785330413406e-06,
+      "loss": 0.6252,
+      "step": 1819
+    },
+    {
+      "epoch": 2.182363527294541,
+      "grad_norm": 0.5965514929018709,
+      "learning_rate": 2.091101740079364e-06,
+      "loss": 0.5862,
+      "step": 1820
+    },
+    {
+      "epoch": 2.1835632873425315,
+      "grad_norm": 0.6449129682843583,
+      "learning_rate": 2.085423825835622e-06,
+      "loss": 0.5684,
+      "step": 1821
+    },
+    {
+      "epoch": 2.184763047390522,
+      "grad_norm": 0.6076189884120574,
+      "learning_rate": 2.0797515987614084e-06,
+      "loss": 0.5475,
+      "step": 1822
+    },
+    {
+      "epoch": 2.1859628074385125,
+      "grad_norm": 0.6850039662885326,
+      "learning_rate": 2.074085069924855e-06,
+      "loss": 0.5261,
+      "step": 1823
+    },
+    {
+      "epoch": 2.1871625674865025,
+      "grad_norm": 0.5817182709374599,
+      "learning_rate": 2.068424250382974e-06,
+      "loss": 0.5711,
+      "step": 1824
+    },
+    {
+      "epoch": 2.188362327534493,
+      "grad_norm": 0.7508580988668893,
+      "learning_rate": 2.0627691511816453e-06,
+      "loss": 0.5366,
+      "step": 1825
+    },
+    {
+      "epoch": 2.1895620875824835,
+      "grad_norm": 0.6464934239515318,
+      "learning_rate": 2.0571197833555756e-06,
+      "loss": 0.5638,
+      "step": 1826
+    },
+    {
+      "epoch": 2.190761847630474,
+      "grad_norm": 0.6322795581774937,
+      "learning_rate": 2.0514761579282926e-06,
+      "loss": 0.5796,
+      "step": 1827
+    },
+    {
+      "epoch": 2.1919616076784645,
+      "grad_norm": 0.6151031350395569,
+      "learning_rate": 2.045838285912125e-06,
+      "loss": 0.5823,
+      "step": 1828
+    },
+    {
+      "epoch": 2.1931613677264545,
+      "grad_norm": 0.9645888747249992,
+      "learning_rate": 2.040206178308164e-06,
+      "loss": 0.4862,
+      "step": 1829
+    },
+    {
+      "epoch": 2.194361127774445,
+      "grad_norm": 0.6181538881404581,
+      "learning_rate": 2.0345798461062594e-06,
+      "loss": 0.5455,
+      "step": 1830
+    },
+    {
+      "epoch": 2.1955608878224355,
+      "grad_norm": 0.7168161618227379,
+      "learning_rate": 2.028959300284991e-06,
+      "loss": 0.4991,
+      "step": 1831
+    },
+    {
+      "epoch": 2.196760647870426,
+      "grad_norm": 0.5593225461837725,
+      "learning_rate": 2.023344551811647e-06,
+      "loss": 0.5718,
+      "step": 1832
+    },
+    {
+      "epoch": 2.1979604079184165,
+      "grad_norm": 0.6297978896378902,
+      "learning_rate": 2.0177356116422025e-06,
+      "loss": 0.586,
+      "step": 1833
+    },
+    {
+      "epoch": 2.1991601679664066,
+      "grad_norm": 0.6019293943984955,
+      "learning_rate": 2.0121324907213003e-06,
+      "loss": 0.5756,
+      "step": 1834
+    },
+    {
+      "epoch": 2.200359928014397,
+      "grad_norm": 0.5867637519763697,
+      "learning_rate": 2.006535199982228e-06,
+      "loss": 0.5619,
+      "step": 1835
+    },
+    {
+      "epoch": 2.2015596880623876,
+      "grad_norm": 0.5787485857506453,
+      "learning_rate": 2.000943750346896e-06,
+      "loss": 0.6323,
+      "step": 1836
+    },
+    {
+      "epoch": 2.202759448110378,
+      "grad_norm": 0.6453884378409138,
+      "learning_rate": 1.995358152725818e-06,
+      "loss": 0.5689,
+      "step": 1837
+    },
+    {
+      "epoch": 2.203959208158368,
+      "grad_norm": 0.7534093765701781,
+      "learning_rate": 1.989778418018088e-06,
+      "loss": 0.6421,
+      "step": 1838
+    },
+    {
+      "epoch": 2.2051589682063586,
+      "grad_norm": 0.6859433373701794,
+      "learning_rate": 1.9842045571113623e-06,
+      "loss": 0.5641,
+      "step": 1839
+    },
+    {
+      "epoch": 2.206358728254349,
+      "grad_norm": 0.6513989642933808,
+      "learning_rate": 1.978636580881829e-06,
+      "loss": 0.5017,
+      "step": 1840
+    },
+    {
+      "epoch": 2.2075584883023396,
+      "grad_norm": 0.5656305153730823,
+      "learning_rate": 1.973074500194199e-06,
+      "loss": 0.5369,
+      "step": 1841
+    },
+    {
+      "epoch": 2.20875824835033,
+      "grad_norm": 0.5825701804911828,
+      "learning_rate": 1.967518325901683e-06,
+      "loss": 0.6088,
+      "step": 1842
+    },
+    {
+      "epoch": 2.20995800839832,
+      "grad_norm": 0.6557230732903017,
+      "learning_rate": 1.9619680688459574e-06,
+      "loss": 0.6454,
+      "step": 1843
+    },
+    {
+      "epoch": 2.2111577684463106,
+      "grad_norm": 0.6135776891881046,
+      "learning_rate": 1.9564237398571565e-06,
+      "loss": 0.5716,
+      "step": 1844
+    },
+    {
+      "epoch": 2.212357528494301,
+      "grad_norm": 0.5859319693232549,
+      "learning_rate": 1.9508853497538533e-06,
+      "loss": 0.6105,
+      "step": 1845
+    },
+    {
+      "epoch": 2.2135572885422916,
+      "grad_norm": 0.5963528780810635,
+      "learning_rate": 1.945352909343021e-06,
+      "loss": 0.5885,
+      "step": 1846
+    },
+    {
+      "epoch": 2.214757048590282,
+      "grad_norm": 0.6196620299527998,
+      "learning_rate": 1.9398264294200315e-06,
+      "loss": 0.5963,
+      "step": 1847
+    },
+    {
+      "epoch": 2.215956808638272,
+      "grad_norm": 0.690298706312048,
+      "learning_rate": 1.9343059207686233e-06,
+      "loss": 0.5237,
+      "step": 1848
+    },
+    {
+      "epoch": 2.2171565686862627,
+      "grad_norm": 0.6464422212185456,
+      "learning_rate": 1.9287913941608833e-06,
+      "loss": 0.5286,
+      "step": 1849
+    },
+    {
+      "epoch": 2.218356328734253,
+      "grad_norm": 0.6701580975562744,
+      "learning_rate": 1.9232828603572255e-06,
+      "loss": 0.4676,
+      "step": 1850
+    },
+    {
+      "epoch": 2.2195560887822436,
+      "grad_norm": 0.7331563604028161,
+      "learning_rate": 1.9177803301063726e-06,
+      "loss": 0.6865,
+      "step": 1851
+    },
+    {
+      "epoch": 2.220755848830234,
+      "grad_norm": 0.6404123124096934,
+      "learning_rate": 1.9122838141453287e-06,
+      "loss": 0.5835,
+      "step": 1852
+    },
+    {
+      "epoch": 2.221955608878224,
+      "grad_norm": 0.6725869376483263,
+      "learning_rate": 1.9067933231993656e-06,
+      "loss": 0.5538,
+      "step": 1853
+    },
+    {
+      "epoch": 2.2231553689262147,
+      "grad_norm": 0.618777309548578,
+      "learning_rate": 1.9013088679819968e-06,
+      "loss": 0.6463,
+      "step": 1854
+    },
+    {
+      "epoch": 2.224355128974205,
+      "grad_norm": 0.601708285350291,
+      "learning_rate": 1.8958304591949594e-06,
+      "loss": 0.5473,
+      "step": 1855
+    },
+    {
+      "epoch": 2.2255548890221957,
+      "grad_norm": 0.6762916720677586,
+      "learning_rate": 1.8903581075281912e-06,
+      "loss": 0.5765,
+      "step": 1856
+    },
+    {
+      "epoch": 2.226754649070186,
+      "grad_norm": 0.5819918345353873,
+      "learning_rate": 1.8848918236598112e-06,
+      "loss": 0.5443,
+      "step": 1857
+    },
+    {
+      "epoch": 2.227954409118176,
+      "grad_norm": 0.5970533348240304,
+      "learning_rate": 1.8794316182560994e-06,
+      "loss": 0.507,
+      "step": 1858
+    },
+    {
+      "epoch": 2.2291541691661667,
+      "grad_norm": 0.592778211817268,
+      "learning_rate": 1.8739775019714735e-06,
+      "loss": 0.5656,
+      "step": 1859
+    },
+    {
+      "epoch": 2.230353929214157,
+      "grad_norm": 0.6466163136909957,
+      "learning_rate": 1.868529485448472e-06,
+      "loss": 0.5749,
+      "step": 1860
+    },
+    {
+      "epoch": 2.2315536892621477,
+      "grad_norm": 0.6461789354205344,
+      "learning_rate": 1.8630875793177245e-06,
+      "loss": 0.5744,
+      "step": 1861
+    },
+    {
+      "epoch": 2.232753449310138,
+      "grad_norm": 0.5990001671730018,
+      "learning_rate": 1.8576517941979466e-06,
+      "loss": 0.5665,
+      "step": 1862
+    },
+    {
+      "epoch": 2.2339532093581282,
+      "grad_norm": 0.6786058775623247,
+      "learning_rate": 1.8522221406959063e-06,
+      "loss": 0.5481,
+      "step": 1863
+    },
+    {
+      "epoch": 2.2351529694061187,
+      "grad_norm": 0.5890200409884031,
+      "learning_rate": 1.8467986294064029e-06,
+      "loss": 0.5426,
+      "step": 1864
+    },
+    {
+      "epoch": 2.2363527294541092,
+      "grad_norm": 0.566197078035473,
+      "learning_rate": 1.8413812709122535e-06,
+      "loss": 0.5251,
+      "step": 1865
+    },
+    {
+      "epoch": 2.2375524895020997,
+      "grad_norm": 0.5416952672510823,
+      "learning_rate": 1.8359700757842757e-06,
+      "loss": 0.5652,
+      "step": 1866
+    },
+    {
+      "epoch": 2.23875224955009,
+      "grad_norm": 0.6852543834616891,
+      "learning_rate": 1.8305650545812499e-06,
+      "loss": 0.5909,
+      "step": 1867
+    },
+    {
+      "epoch": 2.2399520095980803,
+      "grad_norm": 0.811544879193826,
+      "learning_rate": 1.8251662178499136e-06,
+      "loss": 0.5932,
+      "step": 1868
+    },
+    {
+      "epoch": 2.2411517696460708,
+      "grad_norm": 0.5762869874446908,
+      "learning_rate": 1.8197735761249424e-06,
+      "loss": 0.6263,
+      "step": 1869
+    },
+    {
+      "epoch": 2.2423515296940613,
+      "grad_norm": 0.5488879273291238,
+      "learning_rate": 1.8143871399289137e-06,
+      "loss": 0.5617,
+      "step": 1870
+    },
+    {
+      "epoch": 2.2435512897420518,
+      "grad_norm": 0.5629041301403368,
+      "learning_rate": 1.8090069197723032e-06,
+      "loss": 0.5402,
+      "step": 1871
+    },
+    {
+      "epoch": 2.244751049790042,
+      "grad_norm": 0.613623805105581,
+      "learning_rate": 1.8036329261534546e-06,
+      "loss": 0.5841,
+      "step": 1872
+    },
+    {
+      "epoch": 2.2459508098380323,
+      "grad_norm": 0.530537229404665,
+      "learning_rate": 1.7982651695585623e-06,
+      "loss": 0.51,
+      "step": 1873
+    },
+    {
+      "epoch": 2.247150569886023,
+      "grad_norm": 0.573926051224772,
+      "learning_rate": 1.7929036604616518e-06,
+      "loss": 0.5295,
+      "step": 1874
+    },
+    {
+      "epoch": 2.2483503299340133,
+      "grad_norm": 0.6723771525346085,
+      "learning_rate": 1.7875484093245555e-06,
+      "loss": 0.5417,
+      "step": 1875
+    },
+    {
+      "epoch": 2.249550089982004,
+      "grad_norm": 0.6500871435131167,
+      "learning_rate": 1.7821994265968962e-06,
+      "loss": 0.6398,
+      "step": 1876
+    },
+    {
+      "epoch": 2.250749850029994,
+      "grad_norm": 0.6648278777836503,
+      "learning_rate": 1.776856722716067e-06,
+      "loss": 0.5293,
+      "step": 1877
+    },
+    {
+      "epoch": 2.2519496100779843,
+      "grad_norm": 0.5759898973309917,
+      "learning_rate": 1.7715203081072025e-06,
+      "loss": 0.6463,
+      "step": 1878
+    },
+    {
+      "epoch": 2.253149370125975,
+      "grad_norm": 0.6066032014974669,
+      "learning_rate": 1.766190193183175e-06,
+      "loss": 0.514,
+      "step": 1879
+    },
+    {
+      "epoch": 2.2543491301739653,
+      "grad_norm": 0.6287796350012775,
+      "learning_rate": 1.7608663883445582e-06,
+      "loss": 0.5563,
+      "step": 1880
+    },
+    {
+      "epoch": 2.255548890221956,
+      "grad_norm": 0.6187048867104146,
+      "learning_rate": 1.7555489039796093e-06,
+      "loss": 0.5151,
+      "step": 1881
+    },
+    {
+      "epoch": 2.256748650269946,
+      "grad_norm": 0.5403178157587972,
+      "learning_rate": 1.7502377504642631e-06,
+      "loss": 0.5299,
+      "step": 1882
+    },
+    {
+      "epoch": 2.2579484103179364,
+      "grad_norm": 0.6300808545625,
+      "learning_rate": 1.7449329381620938e-06,
+      "loss": 0.595,
+      "step": 1883
+    },
+    {
+      "epoch": 2.259148170365927,
+      "grad_norm": 0.5671233702525605,
+      "learning_rate": 1.7396344774243013e-06,
+      "loss": 0.5507,
+      "step": 1884
+    },
+    {
+      "epoch": 2.2603479304139174,
+      "grad_norm": 0.5235190922216324,
+      "learning_rate": 1.7343423785896934e-06,
+      "loss": 0.5609,
+      "step": 1885
+    },
+    {
+      "epoch": 2.2615476904619074,
+      "grad_norm": 0.7559982150412795,
+      "learning_rate": 1.7290566519846697e-06,
+      "loss": 0.5258,
+      "step": 1886
+    },
+    {
+      "epoch": 2.262747450509898,
+      "grad_norm": 0.6092653856814914,
+      "learning_rate": 1.723777307923185e-06,
+      "loss": 0.5163,
+      "step": 1887
+    },
+    {
+      "epoch": 2.2639472105578884,
+      "grad_norm": 0.5677836138018956,
+      "learning_rate": 1.7185043567067466e-06,
+      "loss": 0.5546,
+      "step": 1888
+    },
+    {
+      "epoch": 2.265146970605879,
+      "grad_norm": 0.6187893833043113,
+      "learning_rate": 1.7132378086243907e-06,
+      "loss": 0.5551,
+      "step": 1889
+    },
+    {
+      "epoch": 2.2663467306538694,
+      "grad_norm": 0.708337066742395,
+      "learning_rate": 1.7079776739526505e-06,
+      "loss": 0.5433,
+      "step": 1890
+    },
+    {
+      "epoch": 2.26754649070186,
+      "grad_norm": 0.563463133541644,
+      "learning_rate": 1.7027239629555503e-06,
+      "loss": 0.567,
+      "step": 1891
+    },
+    {
+      "epoch": 2.26874625074985,
+      "grad_norm": 0.6579698507083996,
+      "learning_rate": 1.6974766858845792e-06,
+      "loss": 0.5555,
+      "step": 1892
+    },
+    {
+      "epoch": 2.2699460107978404,
+      "grad_norm": 0.6877758286059601,
+      "learning_rate": 1.692235852978672e-06,
+      "loss": 0.5324,
+      "step": 1893
+    },
+    {
+      "epoch": 2.271145770845831,
+      "grad_norm": 0.6399348315226369,
+      "learning_rate": 1.6870014744641882e-06,
+      "loss": 0.5188,
+      "step": 1894
+    },
+    {
+      "epoch": 2.2723455308938214,
+      "grad_norm": 0.5486291401853911,
+      "learning_rate": 1.6817735605548936e-06,
+      "loss": 0.6143,
+      "step": 1895
+    },
+    {
+      "epoch": 2.2735452909418115,
+      "grad_norm": 0.5805810155866001,
+      "learning_rate": 1.6765521214519393e-06,
+      "loss": 0.5345,
+      "step": 1896
+    },
+    {
+      "epoch": 2.274745050989802,
+      "grad_norm": 0.7942331671780003,
+      "learning_rate": 1.6713371673438427e-06,
+      "loss": 0.6989,
+      "step": 1897
+    },
+    {
+      "epoch": 2.2759448110377924,
+      "grad_norm": 0.6246634234338206,
+      "learning_rate": 1.666128708406467e-06,
+      "loss": 0.5657,
+      "step": 1898
+    },
+    {
+      "epoch": 2.277144571085783,
+      "grad_norm": 0.6477679105114238,
+      "learning_rate": 1.6609267548030011e-06,
+      "loss": 0.5589,
+      "step": 1899
+    },
+    {
+      "epoch": 2.2783443311337734,
+      "grad_norm": 1.5909813424098906,
+      "learning_rate": 1.6557313166839429e-06,
+      "loss": 0.6089,
+      "step": 1900
+    },
+    {
+      "epoch": 2.2795440911817635,
+      "grad_norm": 0.5726043180383195,
+      "learning_rate": 1.6505424041870693e-06,
+      "loss": 0.6009,
+      "step": 1901
+    },
+    {
+      "epoch": 2.280743851229754,
+      "grad_norm": 0.6290546283029272,
+      "learning_rate": 1.64536002743743e-06,
+      "loss": 0.5536,
+      "step": 1902
+    },
+    {
+      "epoch": 2.2819436112777445,
+      "grad_norm": 0.6065446291792649,
+      "learning_rate": 1.6401841965473248e-06,
+      "loss": 0.5703,
+      "step": 1903
+    },
+    {
+      "epoch": 2.283143371325735,
+      "grad_norm": 0.6387978612018363,
+      "learning_rate": 1.635014921616272e-06,
+      "loss": 0.5529,
+      "step": 1904
+    },
+    {
+      "epoch": 2.284343131373725,
+      "grad_norm": 0.5525726365824832,
+      "learning_rate": 1.6298522127310002e-06,
+      "loss": 0.596,
+      "step": 1905
+    },
+    {
+      "epoch": 2.2855428914217155,
+      "grad_norm": 0.6718164458112008,
+      "learning_rate": 1.6246960799654337e-06,
+      "loss": 0.5944,
+      "step": 1906
+    },
+    {
+      "epoch": 2.286742651469706,
+      "grad_norm": 0.6021082270319865,
+      "learning_rate": 1.619546533380652e-06,
+      "loss": 0.5542,
+      "step": 1907
+    },
+    {
+      "epoch": 2.2879424115176965,
+      "grad_norm": 0.5195251990068274,
+      "learning_rate": 1.614403583024892e-06,
+      "loss": 0.6108,
+      "step": 1908
+    },
+    {
+      "epoch": 2.289142171565687,
+      "grad_norm": 0.6037277381699504,
+      "learning_rate": 1.6092672389335169e-06,
+      "loss": 0.5581,
+      "step": 1909
+    },
+    {
+      "epoch": 2.2903419316136775,
+      "grad_norm": 0.6132642852995489,
+      "learning_rate": 1.6041375111289987e-06,
+      "loss": 0.6014,
+      "step": 1910
+    },
+    {
+      "epoch": 2.2915416916616675,
+      "grad_norm": 0.5869390782704332,
+      "learning_rate": 1.5990144096208998e-06,
+      "loss": 0.6195,
+      "step": 1911
+    },
+    {
+      "epoch": 2.292741451709658,
+      "grad_norm": 0.7480947459686417,
+      "learning_rate": 1.5938979444058527e-06,
+      "loss": 0.4999,
+      "step": 1912
+    },
+    {
+      "epoch": 2.2939412117576485,
+      "grad_norm": 0.6503912322641925,
+      "learning_rate": 1.5887881254675414e-06,
+      "loss": 0.5348,
+      "step": 1913
+    },
+    {
+      "epoch": 2.295140971805639,
+      "grad_norm": 0.5927596306455847,
+      "learning_rate": 1.5836849627766787e-06,
+      "loss": 0.6018,
+      "step": 1914
+    },
+    {
+      "epoch": 2.296340731853629,
+      "grad_norm": 0.5863027916401812,
+      "learning_rate": 1.5785884662909917e-06,
+      "loss": 0.5754,
+      "step": 1915
+    },
+    {
+      "epoch": 2.2975404919016196,
+      "grad_norm": 0.5287286367507306,
+      "learning_rate": 1.5734986459551994e-06,
+      "loss": 0.5533,
+      "step": 1916
+    },
+    {
+      "epoch": 2.29874025194961,
+      "grad_norm": 0.683394217140752,
+      "learning_rate": 1.568415511700992e-06,
+      "loss": 0.623,
+      "step": 1917
+    },
+    {
+      "epoch": 2.2999400119976006,
+      "grad_norm": 0.5975790778744007,
+      "learning_rate": 1.563339073447016e-06,
+      "loss": 0.5726,
+      "step": 1918
+    },
+    {
+      "epoch": 2.301139772045591,
+      "grad_norm": 0.6729930876497099,
+      "learning_rate": 1.5582693410988469e-06,
+      "loss": 0.5701,
+      "step": 1919
+    },
+    {
+      "epoch": 2.302339532093581,
+      "grad_norm": 0.6390721570846494,
+      "learning_rate": 1.553206324548983e-06,
+      "loss": 0.5314,
+      "step": 1920
+    },
+    {
+      "epoch": 2.3035392921415716,
+      "grad_norm": 0.6375196247502823,
+      "learning_rate": 1.548150033676814e-06,
+      "loss": 0.5239,
+      "step": 1921
+    },
+    {
+      "epoch": 2.304739052189562,
+      "grad_norm": 0.633062354492987,
+      "learning_rate": 1.5431004783486004e-06,
+      "loss": 0.5109,
+      "step": 1922
+    },
+    {
+      "epoch": 2.3059388122375526,
+      "grad_norm": 0.6168277970639483,
+      "learning_rate": 1.5380576684174698e-06,
+      "loss": 0.5909,
+      "step": 1923
+    },
+    {
+      "epoch": 2.3071385722855426,
+      "grad_norm": 0.6394106835366958,
+      "learning_rate": 1.5330216137233838e-06,
+      "loss": 0.562,
+      "step": 1924
+    },
+    {
+      "epoch": 2.308338332333533,
+      "grad_norm": 0.5875706845731299,
+      "learning_rate": 1.5279923240931176e-06,
+      "loss": 0.6031,
+      "step": 1925
+    },
+    {
+      "epoch": 2.3095380923815236,
+      "grad_norm": 0.5722294455291087,
+      "learning_rate": 1.5229698093402505e-06,
+      "loss": 0.5993,
+      "step": 1926
+    },
+    {
+      "epoch": 2.310737852429514,
+      "grad_norm": 0.5564432556484449,
+      "learning_rate": 1.5179540792651464e-06,
+      "loss": 0.5627,
+      "step": 1927
+    },
+    {
+      "epoch": 2.3119376124775046,
+      "grad_norm": 0.641828291089028,
+      "learning_rate": 1.5129451436549203e-06,
+      "loss": 0.5383,
+      "step": 1928
+    },
+    {
+      "epoch": 2.313137372525495,
+      "grad_norm": 0.5726544285669013,
+      "learning_rate": 1.5079430122834343e-06,
+      "loss": 0.5737,
+      "step": 1929
+    },
+    {
+      "epoch": 2.314337132573485,
+      "grad_norm": 0.8717905620475225,
+      "learning_rate": 1.5029476949112787e-06,
+      "loss": 0.6155,
+      "step": 1930
+    },
+    {
+      "epoch": 2.3155368926214757,
+      "grad_norm": 0.6063397584759121,
+      "learning_rate": 1.4979592012857375e-06,
+      "loss": 0.5923,
+      "step": 1931
+    },
+    {
+      "epoch": 2.316736652669466,
+      "grad_norm": 0.6089932443469553,
+      "learning_rate": 1.4929775411407871e-06,
+      "loss": 0.6011,
+      "step": 1932
+    },
+    {
+      "epoch": 2.3179364127174567,
+      "grad_norm": 0.6062511474466317,
+      "learning_rate": 1.4880027241970668e-06,
+      "loss": 0.5431,
+      "step": 1933
+    },
+    {
+      "epoch": 2.3191361727654467,
+      "grad_norm": 0.59304838560386,
+      "learning_rate": 1.4830347601618639e-06,
+      "loss": 0.5945,
+      "step": 1934
+    },
+    {
+      "epoch": 2.320335932813437,
+      "grad_norm": 0.643578110697501,
+      "learning_rate": 1.4780736587290927e-06,
+      "loss": 0.5187,
+      "step": 1935
+    },
+    {
+      "epoch": 2.3215356928614277,
+      "grad_norm": 0.5817383170621565,
+      "learning_rate": 1.4731194295792788e-06,
+      "loss": 0.5428,
+      "step": 1936
+    },
+    {
+      "epoch": 2.322735452909418,
+      "grad_norm": 0.6605528620423613,
+      "learning_rate": 1.4681720823795353e-06,
+      "loss": 0.5942,
+      "step": 1937
+    },
+    {
+      "epoch": 2.3239352129574087,
+      "grad_norm": 0.5914015474201617,
+      "learning_rate": 1.46323162678355e-06,
+      "loss": 0.5933,
+      "step": 1938
+    },
+    {
+      "epoch": 2.3251349730053987,
+      "grad_norm": 0.542788224900796,
+      "learning_rate": 1.458298072431556e-06,
+      "loss": 0.5954,
+      "step": 1939
+    },
+    {
+      "epoch": 2.3263347330533892,
+      "grad_norm": 0.569600511641598,
+      "learning_rate": 1.4533714289503297e-06,
+      "loss": 0.5392,
+      "step": 1940
+    },
+    {
+      "epoch": 2.3275344931013797,
+      "grad_norm": 0.592861908888043,
+      "learning_rate": 1.4484517059531588e-06,
+      "loss": 0.5719,
+      "step": 1941
+    },
+    {
+      "epoch": 2.32873425314937,
+      "grad_norm": 0.6265611773472458,
+      "learning_rate": 1.4435389130398208e-06,
+      "loss": 0.58,
+      "step": 1942
+    },
+    {
+      "epoch": 2.3299340131973607,
+      "grad_norm": 0.6304403704739078,
+      "learning_rate": 1.438633059796581e-06,
+      "loss": 0.5154,
+      "step": 1943
+    },
+    {
+      "epoch": 2.3311337732453508,
+      "grad_norm": 0.5543815663073327,
+      "learning_rate": 1.4337341557961587e-06,
+      "loss": 0.5859,
+      "step": 1944
+    },
+    {
+      "epoch": 2.3323335332933413,
+      "grad_norm": 0.605843532077587,
+      "learning_rate": 1.4288422105977096e-06,
+      "loss": 0.5499,
+      "step": 1945
+    },
+    {
+      "epoch": 2.3335332933413317,
+      "grad_norm": 0.6619172150696679,
+      "learning_rate": 1.4239572337468144e-06,
+      "loss": 0.528,
+      "step": 1946
+    },
+    {
+      "epoch": 2.3347330533893222,
+      "grad_norm": 0.6475994414381125,
+      "learning_rate": 1.4190792347754612e-06,
+      "loss": 0.4947,
+      "step": 1947
+    },
+    {
+      "epoch": 2.3359328134373127,
+      "grad_norm": 0.6041381388503334,
+      "learning_rate": 1.4142082232020127e-06,
+      "loss": 0.5507,
+      "step": 1948
+    },
+    {
+      "epoch": 2.337132573485303,
+      "grad_norm": 0.5982926224915779,
+      "learning_rate": 1.4093442085312036e-06,
+      "loss": 0.5908,
+      "step": 1949
+    },
+    {
+      "epoch": 2.3383323335332933,
+      "grad_norm": 0.6189361684892732,
+      "learning_rate": 1.4044872002541137e-06,
+      "loss": 0.5565,
+      "step": 1950
+    },
+    {
+      "epoch": 2.339532093581284,
+      "grad_norm": 0.6693290788425857,
+      "learning_rate": 1.3996372078481524e-06,
+      "loss": 0.5737,
+      "step": 1951
+    },
+    {
+      "epoch": 2.3407318536292743,
+      "grad_norm": 0.6402845104708695,
+      "learning_rate": 1.3947942407770382e-06,
+      "loss": 0.5414,
+      "step": 1952
+    },
+    {
+      "epoch": 2.3419316136772643,
+      "grad_norm": 0.616135532565906,
+      "learning_rate": 1.3899583084907808e-06,
+      "loss": 0.5517,
+      "step": 1953
+    },
+    {
+      "epoch": 2.343131373725255,
+      "grad_norm": 0.6461494511051894,
+      "learning_rate": 1.3851294204256638e-06,
+      "loss": 0.5728,
+      "step": 1954
+    },
+    {
+      "epoch": 2.3443311337732453,
+      "grad_norm": 0.6861833846520821,
+      "learning_rate": 1.3803075860042258e-06,
+      "loss": 0.5806,
+      "step": 1955
+    },
+    {
+      "epoch": 2.345530893821236,
+      "grad_norm": 0.6014821091393561,
+      "learning_rate": 1.3754928146352408e-06,
+      "loss": 0.6158,
+      "step": 1956
+    },
+    {
+      "epoch": 2.3467306538692263,
+      "grad_norm": 0.5475636439705546,
+      "learning_rate": 1.3706851157137008e-06,
+      "loss": 0.5431,
+      "step": 1957
+    },
+    {
+      "epoch": 2.347930413917217,
+      "grad_norm": 0.635894078553141,
+      "learning_rate": 1.3658844986207987e-06,
+      "loss": 0.5246,
+      "step": 1958
+    },
+    {
+      "epoch": 2.349130173965207,
+      "grad_norm": 0.6110431358884112,
+      "learning_rate": 1.3610909727239074e-06,
+      "loss": 0.6095,
+      "step": 1959
+    },
+    {
+      "epoch": 2.3503299340131973,
+      "grad_norm": 0.6877673486069413,
+      "learning_rate": 1.3563045473765634e-06,
+      "loss": 0.5491,
+      "step": 1960
+    },
+    {
+      "epoch": 2.351529694061188,
+      "grad_norm": 0.5904116741620707,
+      "learning_rate": 1.3515252319184497e-06,
+      "loss": 0.5307,
+      "step": 1961
+    },
+    {
+      "epoch": 2.3527294541091783,
+      "grad_norm": 0.5757789707012982,
+      "learning_rate": 1.3467530356753705e-06,
+      "loss": 0.545,
+      "step": 1962
+    },
+    {
+      "epoch": 2.3539292141571684,
+      "grad_norm": 0.6074692348087055,
+      "learning_rate": 1.3419879679592423e-06,
+      "loss": 0.5854,
+      "step": 1963
+    },
+    {
+      "epoch": 2.355128974205159,
+      "grad_norm": 0.6340604837018596,
+      "learning_rate": 1.3372300380680765e-06,
+      "loss": 0.5274,
+      "step": 1964
+    },
+    {
+      "epoch": 2.3563287342531494,
+      "grad_norm": 0.5595653101922524,
+      "learning_rate": 1.3324792552859456e-06,
+      "loss": 0.5705,
+      "step": 1965
+    },
+    {
+      "epoch": 2.35752849430114,
+      "grad_norm": 0.6038322319383336,
+      "learning_rate": 1.327735628882984e-06,
+      "loss": 0.5967,
+      "step": 1966
+    },
+    {
+      "epoch": 2.3587282543491304,
+      "grad_norm": 0.6053575074959013,
+      "learning_rate": 1.3229991681153632e-06,
+      "loss": 0.5921,
+      "step": 1967
+    },
+    {
+      "epoch": 2.3599280143971204,
+      "grad_norm": 0.7111278599605177,
+      "learning_rate": 1.3182698822252648e-06,
+      "loss": 0.531,
+      "step": 1968
+    },
+    {
+      "epoch": 2.361127774445111,
+      "grad_norm": 0.6305780252744821,
+      "learning_rate": 1.3135477804408763e-06,
+      "loss": 0.5621,
+      "step": 1969
+    },
+    {
+      "epoch": 2.3623275344931014,
+      "grad_norm": 0.6419371901578995,
+      "learning_rate": 1.308832871976366e-06,
+      "loss": 0.5162,
+      "step": 1970
+    },
+    {
+      "epoch": 2.363527294541092,
+      "grad_norm": 0.5597564694748165,
+      "learning_rate": 1.3041251660318643e-06,
+      "loss": 0.5992,
+      "step": 1971
+    },
+    {
+      "epoch": 2.364727054589082,
+      "grad_norm": 0.6223790891939087,
+      "learning_rate": 1.2994246717934488e-06,
+      "loss": 0.5711,
+      "step": 1972
+    },
+    {
+      "epoch": 2.3659268146370724,
+      "grad_norm": 0.7933541406230165,
+      "learning_rate": 1.2947313984331244e-06,
+      "loss": 0.5178,
+      "step": 1973
+    },
+    {
+      "epoch": 2.367126574685063,
+      "grad_norm": 0.5654858787055241,
+      "learning_rate": 1.2900453551088071e-06,
+      "loss": 0.5364,
+      "step": 1974
+    },
+    {
+      "epoch": 2.3683263347330534,
+      "grad_norm": 0.6515659326283025,
+      "learning_rate": 1.2853665509643032e-06,
+      "loss": 0.5745,
+      "step": 1975
+    },
+    {
+      "epoch": 2.369526094781044,
+      "grad_norm": 0.6874196766144196,
+      "learning_rate": 1.2806949951292945e-06,
+      "loss": 0.563,
+      "step": 1976
+    },
+    {
+      "epoch": 2.3707258548290344,
+      "grad_norm": 0.6517047305231014,
+      "learning_rate": 1.2760306967193191e-06,
+      "loss": 0.5392,
+      "step": 1977
+    },
+    {
+      "epoch": 2.3719256148770245,
+      "grad_norm": 0.7324117552434645,
+      "learning_rate": 1.2713736648357545e-06,
+      "loss": 0.5036,
+      "step": 1978
+    },
+    {
+      "epoch": 2.373125374925015,
+      "grad_norm": 0.6192932403368865,
+      "learning_rate": 1.2667239085657983e-06,
+      "loss": 0.5777,
+      "step": 1979
+    },
+    {
+      "epoch": 2.3743251349730055,
+      "grad_norm": 0.5380945260489934,
+      "learning_rate": 1.262081436982448e-06,
+      "loss": 0.5881,
+      "step": 1980
+    },
+    {
+      "epoch": 2.375524895020996,
+      "grad_norm": 0.7003327766357957,
+      "learning_rate": 1.257446259144494e-06,
+      "loss": 0.519,
+      "step": 1981
+    },
+    {
+      "epoch": 2.376724655068986,
+      "grad_norm": 0.6167640023652006,
+      "learning_rate": 1.2528183840964913e-06,
+      "loss": 0.5325,
+      "step": 1982
+    },
+    {
+      "epoch": 2.3779244151169765,
+      "grad_norm": 0.5850934143132089,
+      "learning_rate": 1.248197820868739e-06,
+      "loss": 0.5664,
+      "step": 1983
+    },
+    {
+      "epoch": 2.379124175164967,
+      "grad_norm": 0.6297352724034515,
+      "learning_rate": 1.2435845784772783e-06,
+      "loss": 0.5641,
+      "step": 1984
+    },
+    {
+      "epoch": 2.3803239352129575,
+      "grad_norm": 0.5442818523295648,
+      "learning_rate": 1.238978665923861e-06,
+      "loss": 0.5983,
+      "step": 1985
+    },
+    {
+      "epoch": 2.381523695260948,
+      "grad_norm": 0.7001314080471295,
+      "learning_rate": 1.2343800921959348e-06,
+      "loss": 0.6551,
+      "step": 1986
+    },
+    {
+      "epoch": 2.382723455308938,
+      "grad_norm": 0.5938647577049746,
+      "learning_rate": 1.2297888662666275e-06,
+      "loss": 0.5488,
+      "step": 1987
+    },
+    {
+      "epoch": 2.3839232153569285,
+      "grad_norm": 0.6845150089494614,
+      "learning_rate": 1.2252049970947355e-06,
+      "loss": 0.4832,
+      "step": 1988
+    },
+    {
+      "epoch": 2.385122975404919,
+      "grad_norm": 0.5537726896596256,
+      "learning_rate": 1.2206284936246903e-06,
+      "loss": 0.621,
+      "step": 1989
+    },
+    {
+      "epoch": 2.3863227354529095,
+      "grad_norm": 0.5956832825946745,
+      "learning_rate": 1.216059364786556e-06,
+      "loss": 0.6012,
+      "step": 1990
+    },
+    {
+      "epoch": 2.3875224955009,
+      "grad_norm": 0.5883351697255591,
+      "learning_rate": 1.211497619496011e-06,
+      "loss": 0.5384,
+      "step": 1991
+    },
+    {
+      "epoch": 2.38872225554889,
+      "grad_norm": 0.609285464500927,
+      "learning_rate": 1.2069432666543167e-06,
+      "loss": 0.5799,
+      "step": 1992
+    },
+    {
+      "epoch": 2.3899220155968806,
+      "grad_norm": 0.6326702607786753,
+      "learning_rate": 1.2023963151483165e-06,
+      "loss": 0.5632,
+      "step": 1993
+    },
+    {
+      "epoch": 2.391121775644871,
+      "grad_norm": 0.6649244538257834,
+      "learning_rate": 1.1978567738504099e-06,
+      "loss": 0.5383,
+      "step": 1994
+    },
+    {
+      "epoch": 2.3923215356928615,
+      "grad_norm": 0.5446010496554285,
+      "learning_rate": 1.1933246516185372e-06,
+      "loss": 0.553,
+      "step": 1995
+    },
+    {
+      "epoch": 2.393521295740852,
+      "grad_norm": 0.6395049348350018,
+      "learning_rate": 1.1887999572961605e-06,
+      "loss": 0.5323,
+      "step": 1996
+    },
+    {
+      "epoch": 2.394721055788842,
+      "grad_norm": 0.5624977453415554,
+      "learning_rate": 1.1842826997122504e-06,
+      "loss": 0.5583,
+      "step": 1997
+    },
+    {
+      "epoch": 2.3959208158368326,
+      "grad_norm": 0.6597021331336324,
+      "learning_rate": 1.1797728876812643e-06,
+      "loss": 0.5321,
+      "step": 1998
+    },
+    {
+      "epoch": 2.397120575884823,
+      "grad_norm": 0.6019638446046344,
+      "learning_rate": 1.1752705300031336e-06,
+      "loss": 0.5535,
+      "step": 1999
+    },
+    {
+      "epoch": 2.3983203359328136,
+      "grad_norm": 0.6228151763589116,
+      "learning_rate": 1.1707756354632382e-06,
+      "loss": 0.5591,
+      "step": 2000
+    },
+    {
+      "epoch": 2.3995200959808036,
+      "grad_norm": 0.6312014950596989,
+      "learning_rate": 1.1662882128324044e-06,
+      "loss": 0.5698,
+      "step": 2001
+    },
+    {
+      "epoch": 2.400719856028794,
+      "grad_norm": 0.7290597527034558,
+      "learning_rate": 1.1618082708668725e-06,
+      "loss": 0.5857,
+      "step": 2002
+    },
+    {
+      "epoch": 2.4019196160767846,
+      "grad_norm": 0.5542132677584134,
+      "learning_rate": 1.1573358183082861e-06,
+      "loss": 0.5852,
+      "step": 2003
+    },
+    {
+      "epoch": 2.403119376124775,
+      "grad_norm": 0.5953507728815691,
+      "learning_rate": 1.1528708638836744e-06,
+      "loss": 0.5464,
+      "step": 2004
+    },
+    {
+      "epoch": 2.4043191361727656,
+      "grad_norm": 0.5587521569960043,
+      "learning_rate": 1.1484134163054423e-06,
+      "loss": 0.5384,
+      "step": 2005
+    },
+    {
+      "epoch": 2.405518896220756,
+      "grad_norm": 0.5002152020323153,
+      "learning_rate": 1.1439634842713371e-06,
+      "loss": 0.6208,
+      "step": 2006
+    },
+    {
+      "epoch": 2.406718656268746,
+      "grad_norm": 0.5942264281550809,
+      "learning_rate": 1.1395210764644454e-06,
+      "loss": 0.6291,
+      "step": 2007
+    },
+    {
+      "epoch": 2.4079184163167366,
+      "grad_norm": 0.6363610661455074,
+      "learning_rate": 1.1350862015531766e-06,
+      "loss": 0.615,
+      "step": 2008
+    },
+    {
+      "epoch": 2.409118176364727,
+      "grad_norm": 0.6811779985331536,
+      "learning_rate": 1.1306588681912318e-06,
+      "loss": 0.5607,
+      "step": 2009
+    },
+    {
+      "epoch": 2.4103179364127176,
+      "grad_norm": 0.6209782474828403,
+      "learning_rate": 1.1262390850176042e-06,
+      "loss": 0.541,
+      "step": 2010
+    },
+    {
+      "epoch": 2.4115176964607077,
+      "grad_norm": 0.5541181532088056,
+      "learning_rate": 1.1218268606565497e-06,
+      "loss": 0.5851,
+      "step": 2011
+    },
+    {
+      "epoch": 2.412717456508698,
+      "grad_norm": 0.650537648990647,
+      "learning_rate": 1.117422203717578e-06,
+      "loss": 0.5038,
+      "step": 2012
+    },
+    {
+      "epoch": 2.4139172165566887,
+      "grad_norm": 0.670922801022453,
+      "learning_rate": 1.1130251227954313e-06,
+      "loss": 0.5934,
+      "step": 2013
+    },
+    {
+      "epoch": 2.415116976604679,
+      "grad_norm": 0.6601396976807905,
+      "learning_rate": 1.1086356264700683e-06,
+      "loss": 0.5824,
+      "step": 2014
+    },
+    {
+      "epoch": 2.4163167366526697,
+      "grad_norm": 0.5633180549832768,
+      "learning_rate": 1.104253723306648e-06,
+      "loss": 0.5778,
+      "step": 2015
+    },
+    {
+      "epoch": 2.4175164967006597,
+      "grad_norm": 0.5997911293984551,
+      "learning_rate": 1.0998794218555141e-06,
+      "loss": 0.5644,
+      "step": 2016
+    },
+    {
+      "epoch": 2.41871625674865,
+      "grad_norm": 0.5152887028635624,
+      "learning_rate": 1.0955127306521768e-06,
+      "loss": 0.5438,
+      "step": 2017
+    },
+    {
+      "epoch": 2.4199160167966407,
+      "grad_norm": 0.6229782943785084,
+      "learning_rate": 1.091153658217296e-06,
+      "loss": 0.5625,
+      "step": 2018
+    },
+    {
+      "epoch": 2.421115776844631,
+      "grad_norm": 0.5777102576408649,
+      "learning_rate": 1.0868022130566652e-06,
+      "loss": 0.6141,
+      "step": 2019
+    },
+    {
+      "epoch": 2.4223155368926212,
+      "grad_norm": 0.519362477250475,
+      "learning_rate": 1.0824584036611952e-06,
+      "loss": 0.5948,
+      "step": 2020
+    },
+    {
+      "epoch": 2.4235152969406117,
+      "grad_norm": 0.6323279030159387,
+      "learning_rate": 1.0781222385068974e-06,
+      "loss": 0.5099,
+      "step": 2021
+    },
+    {
+      "epoch": 2.4247150569886022,
+      "grad_norm": 0.6096341213942442,
+      "learning_rate": 1.0737937260548675e-06,
+      "loss": 0.5425,
+      "step": 2022
+    },
+    {
+      "epoch": 2.4259148170365927,
+      "grad_norm": 0.5847459771908484,
+      "learning_rate": 1.0694728747512688e-06,
+      "loss": 0.5576,
+      "step": 2023
+    },
+    {
+      "epoch": 2.4271145770845832,
+      "grad_norm": 0.6310530150131601,
+      "learning_rate": 1.0651596930273107e-06,
+      "loss": 0.5516,
+      "step": 2024
+    },
+    {
+      "epoch": 2.4283143371325737,
+      "grad_norm": 0.6063344779401464,
+      "learning_rate": 1.0608541892992464e-06,
+      "loss": 0.563,
+      "step": 2025
+    },
+    {
+      "epoch": 2.4295140971805638,
+      "grad_norm": 0.6688338028777848,
+      "learning_rate": 1.0565563719683387e-06,
+      "loss": 0.5099,
+      "step": 2026
+    },
+    {
+      "epoch": 2.4307138572285543,
+      "grad_norm": 0.828679681934732,
+      "learning_rate": 1.0522662494208535e-06,
+      "loss": 0.5441,
+      "step": 2027
+    },
+    {
+      "epoch": 2.4319136172765448,
+      "grad_norm": 0.5121616235981272,
+      "learning_rate": 1.0479838300280499e-06,
+      "loss": 0.6223,
+      "step": 2028
+    },
+    {
+      "epoch": 2.4331133773245353,
+      "grad_norm": 0.6018457847836182,
+      "learning_rate": 1.0437091221461438e-06,
+      "loss": 0.5202,
+      "step": 2029
+    },
+    {
+      "epoch": 2.4343131373725253,
+      "grad_norm": 0.5854513098809797,
+      "learning_rate": 1.0394421341163123e-06,
+      "loss": 0.5088,
+      "step": 2030
+    },
+    {
+      "epoch": 2.435512897420516,
+      "grad_norm": 0.6316958864507666,
+      "learning_rate": 1.0351828742646658e-06,
+      "loss": 0.5061,
+      "step": 2031
+    },
+    {
+      "epoch": 2.4367126574685063,
+      "grad_norm": 0.6489235005842126,
+      "learning_rate": 1.030931350902235e-06,
+      "loss": 0.6063,
+      "step": 2032
+    },
+    {
+      "epoch": 2.437912417516497,
+      "grad_norm": 0.5824017683498028,
+      "learning_rate": 1.0266875723249547e-06,
+      "loss": 0.5601,
+      "step": 2033
+    },
+    {
+      "epoch": 2.4391121775644873,
+      "grad_norm": 0.6200243422536237,
+      "learning_rate": 1.0224515468136481e-06,
+      "loss": 0.5506,
+      "step": 2034
+    },
+    {
+      "epoch": 2.4403119376124773,
+      "grad_norm": 0.5720863280705479,
+      "learning_rate": 1.0182232826340095e-06,
+      "loss": 0.5558,
+      "step": 2035
+    },
+    {
+      "epoch": 2.441511697660468,
+      "grad_norm": 0.6205568182114685,
+      "learning_rate": 1.014002788036587e-06,
+      "loss": 0.5641,
+      "step": 2036
+    },
+    {
+      "epoch": 2.4427114577084583,
+      "grad_norm": 0.6206921709526924,
+      "learning_rate": 1.0097900712567716e-06,
+      "loss": 0.4522,
+      "step": 2037
+    },
+    {
+      "epoch": 2.443911217756449,
+      "grad_norm": 0.5979265289150865,
+      "learning_rate": 1.005585140514773e-06,
+      "loss": 0.5393,
+      "step": 2038
+    },
+    {
+      "epoch": 2.445110977804439,
+      "grad_norm": 0.6106838156167552,
+      "learning_rate": 1.001388004015613e-06,
+      "loss": 0.5683,
+      "step": 2039
+    },
+    {
+      "epoch": 2.4463107378524294,
+      "grad_norm": 0.608285288407351,
+      "learning_rate": 9.97198669949102e-07,
+      "loss": 0.5913,
+      "step": 2040
+    },
+    {
+      "epoch": 2.44751049790042,
+      "grad_norm": 0.5438934210875334,
+      "learning_rate": 9.930171464898224e-07,
+      "loss": 0.6163,
+      "step": 2041
+    },
+    {
+      "epoch": 2.4487102579484104,
+      "grad_norm": 0.5543471267140387,
+      "learning_rate": 9.888434417971231e-07,
+      "loss": 0.5392,
+      "step": 2042
+    },
+    {
+      "epoch": 2.449910017996401,
+      "grad_norm": 0.6119791506740563,
+      "learning_rate": 9.846775640150929e-07,
+      "loss": 0.5483,
+      "step": 2043
+    },
+    {
+      "epoch": 2.4511097780443913,
+      "grad_norm": 0.626406196205996,
+      "learning_rate": 9.805195212725432e-07,
+      "loss": 0.5641,
+      "step": 2044
+    },
+    {
+      "epoch": 2.4523095380923814,
+      "grad_norm": 0.5706517719297208,
+      "learning_rate": 9.763693216830055e-07,
+      "loss": 0.5396,
+      "step": 2045
+    },
+    {
+      "epoch": 2.453509298140372,
+      "grad_norm": 0.6232083000658682,
+      "learning_rate": 9.722269733447038e-07,
+      "loss": 0.5458,
+      "step": 2046
+    },
+    {
+      "epoch": 2.4547090581883624,
+      "grad_norm": 0.6360627530329451,
+      "learning_rate": 9.680924843405377e-07,
+      "loss": 0.5637,
+      "step": 2047
+    },
+    {
+      "epoch": 2.455908818236353,
+      "grad_norm": 0.5964220775139456,
+      "learning_rate": 9.639658627380755e-07,
+      "loss": 0.5884,
+      "step": 2048
+    },
+    {
+      "epoch": 2.457108578284343,
+      "grad_norm": 0.639571658876441,
+      "learning_rate": 9.598471165895361e-07,
+      "loss": 0.5329,
+      "step": 2049
+    },
+    {
+      "epoch": 2.4583083383323334,
+      "grad_norm": 0.585403973447277,
+      "learning_rate": 9.557362539317656e-07,
+      "loss": 0.5455,
+      "step": 2050
+    },
+    {
+      "epoch": 2.459508098380324,
+      "grad_norm": 0.5243149220065292,
+      "learning_rate": 9.516332827862284e-07,
+      "loss": 0.5215,
+      "step": 2051
+    },
+    {
+      "epoch": 2.4607078584283144,
+      "grad_norm": 0.6437331191238298,
+      "learning_rate": 9.475382111589965e-07,
+      "loss": 0.5345,
+      "step": 2052
+    },
+    {
+      "epoch": 2.461907618476305,
+      "grad_norm": 0.6279857956904481,
+      "learning_rate": 9.434510470407182e-07,
+      "loss": 0.6214,
+      "step": 2053
+    },
+    {
+      "epoch": 2.463107378524295,
+      "grad_norm": 0.5517051975103123,
+      "learning_rate": 9.393717984066181e-07,
+      "loss": 0.5662,
+      "step": 2054
+    },
+    {
+      "epoch": 2.4643071385722854,
+      "grad_norm": 0.5512005435549526,
+      "learning_rate": 9.353004732164739e-07,
+      "loss": 0.5303,
+      "step": 2055
+    },
+    {
+      "epoch": 2.465506898620276,
+      "grad_norm": 0.72031566792074,
+      "learning_rate": 9.312370794146025e-07,
+      "loss": 0.4986,
+      "step": 2056
+    },
+    {
+      "epoch": 2.4667066586682664,
+      "grad_norm": 0.6545253273196747,
+      "learning_rate": 9.271816249298449e-07,
+      "loss": 0.5527,
+      "step": 2057
+    },
+    {
+      "epoch": 2.467906418716257,
+      "grad_norm": 0.5422272008238539,
+      "learning_rate": 9.231341176755487e-07,
+      "loss": 0.5918,
+      "step": 2058
+    },
+    {
+      "epoch": 2.469106178764247,
+      "grad_norm": 0.5958161709563157,
+      "learning_rate": 9.190945655495553e-07,
+      "loss": 0.5588,
+      "step": 2059
+    },
+    {
+      "epoch": 2.4703059388122375,
+      "grad_norm": 0.7328783616345088,
+      "learning_rate": 9.15062976434185e-07,
+      "loss": 0.5479,
+      "step": 2060
+    },
+    {
+      "epoch": 2.471505698860228,
+      "grad_norm": 0.6278689600428524,
+      "learning_rate": 9.110393581962146e-07,
+      "loss": 0.5794,
+      "step": 2061
+    },
+    {
+      "epoch": 2.4727054589082185,
+      "grad_norm": 0.6492780524869001,
+      "learning_rate": 9.070237186868741e-07,
+      "loss": 0.5853,
+      "step": 2062
+    },
+    {
+      "epoch": 2.473905218956209,
+      "grad_norm": 0.5522041253309613,
+      "learning_rate": 9.030160657418218e-07,
+      "loss": 0.6022,
+      "step": 2063
+    },
+    {
+      "epoch": 2.475104979004199,
+      "grad_norm": 0.6684264219716441,
+      "learning_rate": 8.990164071811297e-07,
+      "loss": 0.5468,
+      "step": 2064
+    },
+    {
+      "epoch": 2.4763047390521895,
+      "grad_norm": 0.6135364422730409,
+      "learning_rate": 8.950247508092724e-07,
+      "loss": 0.5553,
+      "step": 2065
+    },
+    {
+      "epoch": 2.47750449910018,
+      "grad_norm": 0.5842931943708874,
+      "learning_rate": 8.910411044151135e-07,
+      "loss": 0.6056,
+      "step": 2066
+    },
+    {
+      "epoch": 2.4787042591481705,
+      "grad_norm": 0.6777460725695204,
+      "learning_rate": 8.870654757718805e-07,
+      "loss": 0.5567,
+      "step": 2067
+    },
+    {
+      "epoch": 2.4799040191961605,
+      "grad_norm": 0.6673476797934147,
+      "learning_rate": 8.830978726371586e-07,
+      "loss": 0.5377,
+      "step": 2068
+    },
+    {
+      "epoch": 2.481103779244151,
+      "grad_norm": 0.6182567506606317,
+      "learning_rate": 8.791383027528772e-07,
+      "loss": 0.5056,
+      "step": 2069
+    },
+    {
+      "epoch": 2.4823035392921415,
+      "grad_norm": 0.5761850561096625,
+      "learning_rate": 8.75186773845283e-07,
+      "loss": 0.5573,
+      "step": 2070
+    },
+    {
+      "epoch": 2.483503299340132,
+      "grad_norm": 0.606360236325203,
+      "learning_rate": 8.712432936249365e-07,
+      "loss": 0.6207,
+      "step": 2071
+    },
+    {
+      "epoch": 2.4847030593881225,
+      "grad_norm": 0.603595466271828,
+      "learning_rate": 8.673078697866933e-07,
+      "loss": 0.6215,
+      "step": 2072
+    },
+    {
+      "epoch": 2.485902819436113,
+      "grad_norm": 0.6749154407409601,
+      "learning_rate": 8.633805100096881e-07,
+      "loss": 0.5485,
+      "step": 2073
+    },
+    {
+      "epoch": 2.487102579484103,
+      "grad_norm": 0.7045002622056066,
+      "learning_rate": 8.594612219573195e-07,
+      "loss": 0.5678,
+      "step": 2074
+    },
+    {
+      "epoch": 2.4883023395320936,
+      "grad_norm": 0.5950945639339188,
+      "learning_rate": 8.555500132772371e-07,
+      "loss": 0.5867,
+      "step": 2075
+    },
+    {
+      "epoch": 2.489502099580084,
+      "grad_norm": 0.5732665057845471,
+      "learning_rate": 8.516468916013243e-07,
+      "loss": 0.5325,
+      "step": 2076
+    },
+    {
+      "epoch": 2.4907018596280746,
+      "grad_norm": 0.6730573537624812,
+      "learning_rate": 8.47751864545685e-07,
+      "loss": 0.5712,
+      "step": 2077
+    },
+    {
+      "epoch": 2.4919016196760646,
+      "grad_norm": 0.6103788351739281,
+      "learning_rate": 8.438649397106285e-07,
+      "loss": 0.5506,
+      "step": 2078
+    },
+    {
+      "epoch": 2.493101379724055,
+      "grad_norm": 0.6331513008649531,
+      "learning_rate": 8.399861246806534e-07,
+      "loss": 0.5603,
+      "step": 2079
+    },
+    {
+      "epoch": 2.4943011397720456,
+      "grad_norm": 0.6559565087341892,
+      "learning_rate": 8.361154270244342e-07,
+      "loss": 0.4952,
+      "step": 2080
+    },
+    {
+      "epoch": 2.495500899820036,
+      "grad_norm": 0.6259786029877843,
+      "learning_rate": 8.322528542948067e-07,
+      "loss": 0.5476,
+      "step": 2081
+    },
+    {
+      "epoch": 2.4967006598680266,
+      "grad_norm": 0.6397549274955587,
+      "learning_rate": 8.283984140287521e-07,
+      "loss": 0.5335,
+      "step": 2082
+    },
+    {
+      "epoch": 2.4979004199160166,
+      "grad_norm": 0.7385966603303945,
+      "learning_rate": 8.245521137473822e-07,
+      "loss": 0.647,
+      "step": 2083
+    },
+    {
+      "epoch": 2.499100179964007,
+      "grad_norm": 0.6137631792373935,
+      "learning_rate": 8.207139609559284e-07,
+      "loss": 0.5315,
+      "step": 2084
+    },
+    {
+      "epoch": 2.5002999400119976,
+      "grad_norm": 0.6350875453667628,
+      "learning_rate": 8.16883963143717e-07,
+      "loss": 0.5511,
+      "step": 2085
+    },
+    {
+      "epoch": 2.501499700059988,
+      "grad_norm": 0.6032552621237215,
+      "learning_rate": 8.130621277841716e-07,
+      "loss": 0.5984,
+      "step": 2086
+    },
+    {
+      "epoch": 2.502699460107978,
+      "grad_norm": 0.5810034597194302,
+      "learning_rate": 8.09248462334779e-07,
+      "loss": 0.5514,
+      "step": 2087
+    },
+    {
+      "epoch": 2.5038992201559687,
+      "grad_norm": 0.564087058663493,
+      "learning_rate": 8.054429742370906e-07,
+      "loss": 0.5841,
+      "step": 2088
+    },
+    {
+      "epoch": 2.505098980203959,
+      "grad_norm": 0.62931338569948,
+      "learning_rate": 8.016456709166986e-07,
+      "loss": 0.5231,
+      "step": 2089
+    },
+    {
+      "epoch": 2.5062987402519497,
+      "grad_norm": 0.5686764121520594,
+      "learning_rate": 7.97856559783225e-07,
+      "loss": 0.5367,
+      "step": 2090
+    },
+    {
+      "epoch": 2.50749850029994,
+      "grad_norm": 0.6061303901603898,
+      "learning_rate": 7.940756482303064e-07,
+      "loss": 0.5871,
+      "step": 2091
+    },
+    {
+      "epoch": 2.5086982603479306,
+      "grad_norm": 0.637376733778738,
+      "learning_rate": 7.903029436355802e-07,
+      "loss": 0.6376,
+      "step": 2092
+    },
+    {
+      "epoch": 2.5098980203959207,
+      "grad_norm": 0.6070290680506467,
+      "learning_rate": 7.86538453360669e-07,
+      "loss": 0.5408,
+      "step": 2093
+    },
+    {
+      "epoch": 2.511097780443911,
+      "grad_norm": 0.6605654652869697,
+      "learning_rate": 7.827821847511668e-07,
+      "loss": 0.6269,
+      "step": 2094
+    },
+    {
+      "epoch": 2.5122975404919017,
+      "grad_norm": 0.6225011194684036,
+      "learning_rate": 7.790341451366263e-07,
+      "loss": 0.5601,
+      "step": 2095
+    },
+    {
+      "epoch": 2.513497300539892,
+      "grad_norm": 0.5458392403706084,
+      "learning_rate": 7.752943418305408e-07,
+      "loss": 0.5221,
+      "step": 2096
+    },
+    {
+      "epoch": 2.5146970605878822,
+      "grad_norm": 0.6360773622632925,
+      "learning_rate": 7.71562782130334e-07,
+      "loss": 0.6115,
+      "step": 2097
+    },
+    {
+      "epoch": 2.5158968206358727,
+      "grad_norm": 0.5782806887863192,
+      "learning_rate": 7.678394733173427e-07,
+      "loss": 0.5208,
+      "step": 2098
+    },
+    {
+      "epoch": 2.517096580683863,
+      "grad_norm": 0.5044811505319163,
+      "learning_rate": 7.641244226568056e-07,
+      "loss": 0.6212,
+      "step": 2099
+    },
+    {
+      "epoch": 2.5182963407318537,
+      "grad_norm": 0.6499090066208069,
+      "learning_rate": 7.60417637397845e-07,
+      "loss": 0.6085,
+      "step": 2100
+    },
+    {
+      "epoch": 2.519496100779844,
+      "grad_norm": 0.546825659889867,
+      "learning_rate": 7.567191247734589e-07,
+      "loss": 0.5505,
+      "step": 2101
+    },
+    {
+      "epoch": 2.5206958608278347,
+      "grad_norm": 0.5661118081548666,
+      "learning_rate": 7.530288920004957e-07,
+      "loss": 0.5503,
+      "step": 2102
+    },
+    {
+      "epoch": 2.5218956208758248,
+      "grad_norm": 0.5758245642104611,
+      "learning_rate": 7.493469462796565e-07,
+      "loss": 0.5188,
+      "step": 2103
+    },
+    {
+      "epoch": 2.5230953809238152,
+      "grad_norm": 0.65196002526569,
+      "learning_rate": 7.45673294795467e-07,
+      "loss": 0.6145,
+      "step": 2104
+    },
+    {
+      "epoch": 2.5242951409718057,
+      "grad_norm": 0.6387451554071346,
+      "learning_rate": 7.420079447162665e-07,
+      "loss": 0.5421,
+      "step": 2105
+    },
+    {
+      "epoch": 2.525494901019796,
+      "grad_norm": 0.6660052812587164,
+      "learning_rate": 7.383509031942016e-07,
+      "loss": 0.5441,
+      "step": 2106
+    },
+    {
+      "epoch": 2.5266946610677863,
+      "grad_norm": 0.6643620782254298,
+      "learning_rate": 7.347021773652035e-07,
+      "loss": 0.5121,
+      "step": 2107
+    },
+    {
+      "epoch": 2.527894421115777,
+      "grad_norm": 0.5942439165853299,
+      "learning_rate": 7.31061774348975e-07,
+      "loss": 0.5815,
+      "step": 2108
+    },
+    {
+      "epoch": 2.5290941811637673,
+      "grad_norm": 0.6514630475774669,
+      "learning_rate": 7.274297012489812e-07,
+      "loss": 0.5866,
+      "step": 2109
+    },
+    {
+      "epoch": 2.5302939412117578,
+      "grad_norm": 0.6417301950385145,
+      "learning_rate": 7.238059651524354e-07,
+      "loss": 0.5927,
+      "step": 2110
+    },
+    {
+      "epoch": 2.5314937012597483,
+      "grad_norm": 0.6574104218840121,
+      "learning_rate": 7.201905731302777e-07,
+      "loss": 0.6499,
+      "step": 2111
+    },
+    {
+      "epoch": 2.5326934613077383,
+      "grad_norm": 0.6806120471137596,
+      "learning_rate": 7.165835322371689e-07,
+      "loss": 0.5817,
+      "step": 2112
+    },
+    {
+      "epoch": 2.533893221355729,
+      "grad_norm": 0.6370789687688718,
+      "learning_rate": 7.129848495114783e-07,
+      "loss": 0.5126,
+      "step": 2113
+    },
+    {
+      "epoch": 2.5350929814037193,
+      "grad_norm": 0.6451567782463036,
+      "learning_rate": 7.093945319752588e-07,
+      "loss": 0.5185,
+      "step": 2114
+    },
+    {
+      "epoch": 2.53629274145171,
+      "grad_norm": 0.6322934547855595,
+      "learning_rate": 7.058125866342453e-07,
+      "loss": 0.6352,
+      "step": 2115
+    },
+    {
+      "epoch": 2.5374925014997,
+      "grad_norm": 0.5974029199120204,
+      "learning_rate": 7.022390204778351e-07,
+      "loss": 0.5667,
+      "step": 2116
+    },
+    {
+      "epoch": 2.5386922615476903,
+      "grad_norm": 0.8295297775898799,
+      "learning_rate": 6.986738404790755e-07,
+      "loss": 0.5237,
+      "step": 2117
+    },
+    {
+      "epoch": 2.539892021595681,
+      "grad_norm": 0.5713910785693439,
+      "learning_rate": 6.9511705359465e-07,
+      "loss": 0.6225,
+      "step": 2118
+    },
+    {
+      "epoch": 2.5410917816436713,
+      "grad_norm": 0.566923531902045,
+      "learning_rate": 6.915686667648619e-07,
+      "loss": 0.5867,
+      "step": 2119
+    },
+    {
+      "epoch": 2.542291541691662,
+      "grad_norm": 0.6492417746075824,
+      "learning_rate": 6.880286869136288e-07,
+      "loss": 0.5182,
+      "step": 2120
+    },
+    {
+      "epoch": 2.5434913017396523,
+      "grad_norm": 0.7956927113017503,
+      "learning_rate": 6.844971209484613e-07,
+      "loss": 0.5674,
+      "step": 2121
+    },
+    {
+      "epoch": 2.5446910617876424,
+      "grad_norm": 0.5641376870362562,
+      "learning_rate": 6.809739757604494e-07,
+      "loss": 0.5253,
+      "step": 2122
+    },
+    {
+      "epoch": 2.545890821835633,
+      "grad_norm": 0.5912812937401165,
+      "learning_rate": 6.774592582242567e-07,
+      "loss": 0.5383,
+      "step": 2123
+    },
+    {
+      "epoch": 2.5470905818836234,
+      "grad_norm": 0.5580922133198004,
+      "learning_rate": 6.739529751981e-07,
+      "loss": 0.5861,
+      "step": 2124
+    },
+    {
+      "epoch": 2.5482903419316134,
+      "grad_norm": 0.6298162645996003,
+      "learning_rate": 6.704551335237358e-07,
+      "loss": 0.5748,
+      "step": 2125
+    },
+    {
+      "epoch": 2.549490101979604,
+      "grad_norm": 0.5868891574360139,
+      "learning_rate": 6.669657400264512e-07,
+      "loss": 0.574,
+      "step": 2126
+    },
+    {
+      "epoch": 2.5506898620275944,
+      "grad_norm": 0.5976636824154793,
+      "learning_rate": 6.634848015150508e-07,
+      "loss": 0.545,
+      "step": 2127
+    },
+    {
+      "epoch": 2.551889622075585,
+      "grad_norm": 0.6020475860573156,
+      "learning_rate": 6.600123247818357e-07,
+      "loss": 0.5106,
+      "step": 2128
+    },
+    {
+      "epoch": 2.5530893821235754,
+      "grad_norm": 0.5902722858962282,
+      "learning_rate": 6.565483166025987e-07,
+      "loss": 0.5626,
+      "step": 2129
+    },
+    {
+      "epoch": 2.554289142171566,
+      "grad_norm": 0.5587928729210574,
+      "learning_rate": 6.530927837366113e-07,
+      "loss": 0.5722,
+      "step": 2130
+    },
+    {
+      "epoch": 2.555488902219556,
+      "grad_norm": 0.632512991280109,
+      "learning_rate": 6.496457329266003e-07,
+      "loss": 0.5234,
+      "step": 2131
+    },
+    {
+      "epoch": 2.5566886622675464,
+      "grad_norm": 0.6377743211690943,
+      "learning_rate": 6.462071708987467e-07,
+      "loss": 0.5435,
+      "step": 2132
+    },
+    {
+      "epoch": 2.557888422315537,
+      "grad_norm": 0.5842896842093335,
+      "learning_rate": 6.427771043626646e-07,
+      "loss": 0.5487,
+      "step": 2133
+    },
+    {
+      "epoch": 2.5590881823635274,
+      "grad_norm": 0.5957269315639774,
+      "learning_rate": 6.393555400113938e-07,
+      "loss": 0.5712,
+      "step": 2134
+    },
+    {
+      "epoch": 2.5602879424115175,
+      "grad_norm": 0.607664147978379,
+      "learning_rate": 6.359424845213819e-07,
+      "loss": 0.51,
+      "step": 2135
+    },
+    {
+      "epoch": 2.561487702459508,
+      "grad_norm": 0.6481620130889848,
+      "learning_rate": 6.325379445524732e-07,
+      "loss": 0.5201,
+      "step": 2136
+    },
+    {
+      "epoch": 2.5626874625074985,
+      "grad_norm": 0.5655635655291085,
+      "learning_rate": 6.291419267478965e-07,
+      "loss": 0.5335,
+      "step": 2137
+    },
+    {
+      "epoch": 2.563887222555489,
+      "grad_norm": 0.777957284332089,
+      "learning_rate": 6.257544377342523e-07,
+      "loss": 0.5303,
+      "step": 2138
+    },
+    {
+      "epoch": 2.5650869826034794,
+      "grad_norm": 0.62107869014827,
+      "learning_rate": 6.223754841214969e-07,
+      "loss": 0.5255,
+      "step": 2139
+    },
+    {
+      "epoch": 2.56628674265147,
+      "grad_norm": 0.6111004858053942,
+      "learning_rate": 6.190050725029328e-07,
+      "loss": 0.5165,
+      "step": 2140
+    },
+    {
+      "epoch": 2.56748650269946,
+      "grad_norm": 0.6142744373421746,
+      "learning_rate": 6.156432094551951e-07,
+      "loss": 0.593,
+      "step": 2141
+    },
+    {
+      "epoch": 2.5686862627474505,
+      "grad_norm": 0.6028771942191737,
+      "learning_rate": 6.122899015382372e-07,
+      "loss": 0.5702,
+      "step": 2142
+    },
+    {
+      "epoch": 2.569886022795441,
+      "grad_norm": 0.6055817514723558,
+      "learning_rate": 6.089451552953191e-07,
+      "loss": 0.5333,
+      "step": 2143
+    },
+    {
+      "epoch": 2.5710857828434315,
+      "grad_norm": 0.6190509492084894,
+      "learning_rate": 6.056089772529955e-07,
+      "loss": 0.5611,
+      "step": 2144
+    },
+    {
+      "epoch": 2.5722855428914215,
+      "grad_norm": 0.5863993150494622,
+      "learning_rate": 6.022813739211025e-07,
+      "loss": 0.5748,
+      "step": 2145
+    },
+    {
+      "epoch": 2.573485302939412,
+      "grad_norm": 0.6463470257981921,
+      "learning_rate": 5.989623517927395e-07,
+      "loss": 0.5035,
+      "step": 2146
+    },
+    {
+      "epoch": 2.5746850629874025,
+      "grad_norm": 0.5974658144499296,
+      "learning_rate": 5.956519173442704e-07,
+      "loss": 0.5744,
+      "step": 2147
+    },
+    {
+      "epoch": 2.575884823035393,
+      "grad_norm": 0.6047039800838729,
+      "learning_rate": 5.923500770352935e-07,
+      "loss": 0.5541,
+      "step": 2148
+    },
+    {
+      "epoch": 2.5770845830833835,
+      "grad_norm": 0.6412243636579218,
+      "learning_rate": 5.890568373086425e-07,
+      "loss": 0.5502,
+      "step": 2149
+    },
+    {
+      "epoch": 2.578284343131374,
+      "grad_norm": 0.6420558939025194,
+      "learning_rate": 5.857722045903674e-07,
+      "loss": 0.6135,
+      "step": 2150
+    },
+    {
+      "epoch": 2.579484103179364,
+      "grad_norm": 0.6692959800290453,
+      "learning_rate": 5.824961852897231e-07,
+      "loss": 0.5778,
+      "step": 2151
+    },
+    {
+      "epoch": 2.5806838632273545,
+      "grad_norm": 0.5791690455093836,
+      "learning_rate": 5.792287857991591e-07,
+      "loss": 0.6185,
+      "step": 2152
+    },
+    {
+      "epoch": 2.581883623275345,
+      "grad_norm": 0.5876943201180187,
+      "learning_rate": 5.759700124943029e-07,
+      "loss": 0.5751,
+      "step": 2153
+    },
+    {
+      "epoch": 2.583083383323335,
+      "grad_norm": 0.6354825100788556,
+      "learning_rate": 5.727198717339511e-07,
+      "loss": 0.5939,
+      "step": 2154
+    },
+    {
+      "epoch": 2.5842831433713256,
+      "grad_norm": 0.5648854282439252,
+      "learning_rate": 5.694783698600548e-07,
+      "loss": 0.5738,
+      "step": 2155
+    },
+    {
+      "epoch": 2.585482903419316,
+      "grad_norm": 0.5688007543004594,
+      "learning_rate": 5.662455131977102e-07,
+      "loss": 0.5758,
+      "step": 2156
+    },
+    {
+      "epoch": 2.5866826634673066,
+      "grad_norm": 0.5622931780417261,
+      "learning_rate": 5.63021308055141e-07,
+      "loss": 0.5811,
+      "step": 2157
+    },
+    {
+      "epoch": 2.587882423515297,
+      "grad_norm": 0.5936911421046386,
+      "learning_rate": 5.598057607236928e-07,
+      "loss": 0.5717,
+      "step": 2158
+    },
+    {
+      "epoch": 2.5890821835632876,
+      "grad_norm": 0.634008275839511,
+      "learning_rate": 5.565988774778153e-07,
+      "loss": 0.5215,
+      "step": 2159
+    },
+    {
+      "epoch": 2.5902819436112776,
+      "grad_norm": 0.6018185932926242,
+      "learning_rate": 5.534006645750523e-07,
+      "loss": 0.5197,
+      "step": 2160
+    },
+    {
+      "epoch": 2.591481703659268,
+      "grad_norm": 0.7005719436138388,
+      "learning_rate": 5.502111282560291e-07,
+      "loss": 0.5887,
+      "step": 2161
+    },
+    {
+      "epoch": 2.5926814637072586,
+      "grad_norm": 0.7301197459611951,
+      "learning_rate": 5.470302747444428e-07,
+      "loss": 0.6252,
+      "step": 2162
+    },
+    {
+      "epoch": 2.593881223755249,
+      "grad_norm": 0.6942438861370627,
+      "learning_rate": 5.438581102470419e-07,
+      "loss": 0.5182,
+      "step": 2163
+    },
+    {
+      "epoch": 2.595080983803239,
+      "grad_norm": 0.6917749064390649,
+      "learning_rate": 5.406946409536284e-07,
+      "loss": 0.6221,
+      "step": 2164
+    },
+    {
+      "epoch": 2.5962807438512296,
+      "grad_norm": 0.6215078882375504,
+      "learning_rate": 5.375398730370323e-07,
+      "loss": 0.5242,
+      "step": 2165
+    },
+    {
+      "epoch": 2.59748050389922,
+      "grad_norm": 0.6185351010861437,
+      "learning_rate": 5.343938126531034e-07,
+      "loss": 0.5276,
+      "step": 2166
+    },
+    {
+      "epoch": 2.5986802639472106,
+      "grad_norm": 0.6043673893011502,
+      "learning_rate": 5.312564659407054e-07,
+      "loss": 0.6079,
+      "step": 2167
+    },
+    {
+      "epoch": 2.599880023995201,
+      "grad_norm": 0.5734256754088662,
+      "learning_rate": 5.281278390216976e-07,
+      "loss": 0.5979,
+      "step": 2168
+    },
+    {
+      "epoch": 2.6010797840431916,
+      "grad_norm": 0.5800707816384436,
+      "learning_rate": 5.250079380009205e-07,
+      "loss": 0.568,
+      "step": 2169
+    },
+    {
+      "epoch": 2.6022795440911817,
+      "grad_norm": 0.5313291939600484,
+      "learning_rate": 5.218967689661914e-07,
+      "loss": 0.5976,
+      "step": 2170
+    },
+    {
+      "epoch": 2.603479304139172,
+      "grad_norm": 0.5736923527583822,
+      "learning_rate": 5.18794337988292e-07,
+      "loss": 0.5387,
+      "step": 2171
+    },
+    {
+      "epoch": 2.6046790641871627,
+      "grad_norm": 0.6999475757676541,
+      "learning_rate": 5.157006511209467e-07,
+      "loss": 0.4878,
+      "step": 2172
+    },
+    {
+      "epoch": 2.6058788242351527,
+      "grad_norm": 0.628131409078769,
+      "learning_rate": 5.126157144008209e-07,
+      "loss": 0.4561,
+      "step": 2173
+    },
+    {
+      "epoch": 2.607078584283143,
+      "grad_norm": 0.6090807924772929,
+      "learning_rate": 5.095395338475095e-07,
+      "loss": 0.547,
+      "step": 2174
+    },
+    {
+      "epoch": 2.6082783443311337,
+      "grad_norm": 0.6568840037108555,
+      "learning_rate": 5.064721154635155e-07,
+      "loss": 0.5179,
+      "step": 2175
+    },
+    {
+      "epoch": 2.609478104379124,
+      "grad_norm": 0.6077345432598468,
+      "learning_rate": 5.034134652342477e-07,
+      "loss": 0.5896,
+      "step": 2176
+    },
+    {
+      "epoch": 2.6106778644271147,
+      "grad_norm": 0.5683475107240232,
+      "learning_rate": 5.003635891280051e-07,
+      "loss": 0.577,
+      "step": 2177
+    },
+    {
+      "epoch": 2.611877624475105,
+      "grad_norm": 0.5940425103459605,
+      "learning_rate": 4.973224930959669e-07,
+      "loss": 0.5775,
+      "step": 2178
+    },
+    {
+      "epoch": 2.6130773845230952,
+      "grad_norm": 0.6807554600883522,
+      "learning_rate": 4.942901830721786e-07,
+      "loss": 0.5676,
+      "step": 2179
+    },
+    {
+      "epoch": 2.6142771445710857,
+      "grad_norm": 0.6501107217177522,
+      "learning_rate": 4.912666649735403e-07,
+      "loss": 0.5832,
+      "step": 2180
+    },
+    {
+      "epoch": 2.6154769046190762,
+      "grad_norm": 0.6529578897175722,
+      "learning_rate": 4.882519446998007e-07,
+      "loss": 0.5626,
+      "step": 2181
+    },
+    {
+      "epoch": 2.6166766646670667,
+      "grad_norm": 0.5592171901908932,
+      "learning_rate": 4.852460281335386e-07,
+      "loss": 0.5317,
+      "step": 2182
+    },
+    {
+      "epoch": 2.6178764247150568,
+      "grad_norm": 0.5899871633157431,
+      "learning_rate": 4.822489211401526e-07,
+      "loss": 0.5589,
+      "step": 2183
+    },
+    {
+      "epoch": 2.6190761847630473,
+      "grad_norm": 0.6203897260792866,
+      "learning_rate": 4.792606295678564e-07,
+      "loss": 0.5781,
+      "step": 2184
+    },
+    {
+      "epoch": 2.6202759448110378,
+      "grad_norm": 0.5954957072253819,
+      "learning_rate": 4.762811592476585e-07,
+      "loss": 0.5792,
+      "step": 2185
+    },
+    {
+      "epoch": 2.6214757048590283,
+      "grad_norm": 0.6842825105432413,
+      "learning_rate": 4.7331051599335377e-07,
+      "loss": 0.5621,
+      "step": 2186
+    },
+    {
+      "epoch": 2.6226754649070187,
+      "grad_norm": 0.5835272067634589,
+      "learning_rate": 4.7034870560151435e-07,
+      "loss": 0.607,
+      "step": 2187
+    },
+    {
+      "epoch": 2.6238752249550092,
+      "grad_norm": 0.6096228214681819,
+      "learning_rate": 4.673957338514812e-07,
+      "loss": 0.5736,
+      "step": 2188
+    },
+    {
+      "epoch": 2.6250749850029993,
+      "grad_norm": 0.566074085668047,
+      "learning_rate": 4.644516065053406e-07,
+      "loss": 0.591,
+      "step": 2189
+    },
+    {
+      "epoch": 2.62627474505099,
+      "grad_norm": 0.6278033549800017,
+      "learning_rate": 4.61516329307925e-07,
+      "loss": 0.6223,
+      "step": 2190
+    },
+    {
+      "epoch": 2.6274745050989803,
+      "grad_norm": 0.666495192770457,
+      "learning_rate": 4.585899079868006e-07,
+      "loss": 0.5843,
+      "step": 2191
+    },
+    {
+      "epoch": 2.6286742651469703,
+      "grad_norm": 0.6002593159128293,
+      "learning_rate": 4.5567234825224736e-07,
+      "loss": 0.5601,
+      "step": 2192
+    },
+    {
+      "epoch": 2.629874025194961,
+      "grad_norm": 0.6059102396663301,
+      "learning_rate": 4.52763655797257e-07,
+      "loss": 0.5704,
+      "step": 2193
+    },
+    {
+      "epoch": 2.6310737852429513,
+      "grad_norm": 0.5962329324448424,
+      "learning_rate": 4.498638362975194e-07,
+      "loss": 0.6028,
+      "step": 2194
+    },
+    {
+      "epoch": 2.632273545290942,
+      "grad_norm": 0.5800985716686567,
+      "learning_rate": 4.469728954114083e-07,
+      "loss": 0.573,
+      "step": 2195
+    },
+    {
+      "epoch": 2.6334733053389323,
+      "grad_norm": 0.550969607398562,
+      "learning_rate": 4.440908387799753e-07,
+      "loss": 0.5627,
+      "step": 2196
+    },
+    {
+      "epoch": 2.634673065386923,
+      "grad_norm": 0.5858661147935802,
+      "learning_rate": 4.41217672026934e-07,
+      "loss": 0.5717,
+      "step": 2197
+    },
+    {
+      "epoch": 2.635872825434913,
+      "grad_norm": 0.565563272849404,
+      "learning_rate": 4.3835340075865263e-07,
+      "loss": 0.567,
+      "step": 2198
+    },
+    {
+      "epoch": 2.6370725854829034,
+      "grad_norm": 0.7154679915472364,
+      "learning_rate": 4.354980305641415e-07,
+      "loss": 0.4806,
+      "step": 2199
+    },
+    {
+      "epoch": 2.638272345530894,
+      "grad_norm": 0.6277957077530519,
+      "learning_rate": 4.3265156701504274e-07,
+      "loss": 0.5637,
+      "step": 2200
+    },
+    {
+      "epoch": 2.6394721055788843,
+      "grad_norm": 0.5817270389510264,
+      "learning_rate": 4.298140156656178e-07,
+      "loss": 0.5409,
+      "step": 2201
+    },
+    {
+      "epoch": 2.6406718656268744,
+      "grad_norm": 0.5737173928522017,
+      "learning_rate": 4.269853820527403e-07,
+      "loss": 0.5886,
+      "step": 2202
+    },
+    {
+      "epoch": 2.641871625674865,
+      "grad_norm": 0.6314989133285192,
+      "learning_rate": 4.2416567169588086e-07,
+      "loss": 0.5606,
+      "step": 2203
+    },
+    {
+      "epoch": 2.6430713857228554,
+      "grad_norm": 0.6795554928383936,
+      "learning_rate": 4.213548900970965e-07,
+      "loss": 0.4729,
+      "step": 2204
+    },
+    {
+      "epoch": 2.644271145770846,
+      "grad_norm": 0.6145336139358205,
+      "learning_rate": 4.185530427410267e-07,
+      "loss": 0.6001,
+      "step": 2205
+    },
+    {
+      "epoch": 2.6454709058188364,
+      "grad_norm": 0.5157672326801428,
+      "learning_rate": 4.157601350948748e-07,
+      "loss": 0.5769,
+      "step": 2206
+    },
+    {
+      "epoch": 2.646670665866827,
+      "grad_norm": 0.5875060652510343,
+      "learning_rate": 4.1297617260839753e-07,
+      "loss": 0.5676,
+      "step": 2207
+    },
+    {
+      "epoch": 2.647870425914817,
+      "grad_norm": 0.7477248765689066,
+      "learning_rate": 4.102011607139039e-07,
+      "loss": 0.5785,
+      "step": 2208
+    },
+    {
+      "epoch": 2.6490701859628074,
+      "grad_norm": 0.6293565517750546,
+      "learning_rate": 4.074351048262304e-07,
+      "loss": 0.5925,
+      "step": 2209
+    },
+    {
+      "epoch": 2.650269946010798,
+      "grad_norm": 0.5569274950930724,
+      "learning_rate": 4.046780103427422e-07,
+      "loss": 0.6122,
+      "step": 2210
+    },
+    {
+      "epoch": 2.6514697060587884,
+      "grad_norm": 0.6226911550571225,
+      "learning_rate": 4.0192988264331765e-07,
+      "loss": 0.5728,
+      "step": 2211
+    },
+    {
+      "epoch": 2.6526694661067785,
+      "grad_norm": 0.7467205426182734,
+      "learning_rate": 3.9919072709033724e-07,
+      "loss": 0.5054,
+      "step": 2212
+    },
+    {
+      "epoch": 2.653869226154769,
+      "grad_norm": 0.5628270181417315,
+      "learning_rate": 3.9646054902867495e-07,
+      "loss": 0.5441,
+      "step": 2213
+    },
+    {
+      "epoch": 2.6550689862027594,
+      "grad_norm": 0.691376073275601,
+      "learning_rate": 3.937393537856871e-07,
+      "loss": 0.5514,
+      "step": 2214
+    },
+    {
+      "epoch": 2.65626874625075,
+      "grad_norm": 0.6045653367982666,
+      "learning_rate": 3.9102714667120177e-07,
+      "loss": 0.6203,
+      "step": 2215
+    },
+    {
+      "epoch": 2.6574685062987404,
+      "grad_norm": 0.6399541518504948,
+      "learning_rate": 3.8832393297750905e-07,
+      "loss": 0.533,
+      "step": 2216
+    },
+    {
+      "epoch": 2.658668266346731,
+      "grad_norm": 0.5815400951783926,
+      "learning_rate": 3.8562971797934947e-07,
+      "loss": 0.5676,
+      "step": 2217
+    },
+    {
+      "epoch": 2.659868026394721,
+      "grad_norm": 0.6824231088462228,
+      "learning_rate": 3.8294450693390496e-07,
+      "loss": 0.5527,
+      "step": 2218
+    },
+    {
+      "epoch": 2.6610677864427115,
+      "grad_norm": 0.630113655846623,
+      "learning_rate": 3.8026830508078873e-07,
+      "loss": 0.5274,
+      "step": 2219
+    },
+    {
+      "epoch": 2.662267546490702,
+      "grad_norm": 0.5548091593081005,
+      "learning_rate": 3.776011176420341e-07,
+      "loss": 0.5691,
+      "step": 2220
+    },
+    {
+      "epoch": 2.663467306538692,
+      "grad_norm": 0.7411474172071046,
+      "learning_rate": 3.749429498220841e-07,
+      "loss": 0.5111,
+      "step": 2221
+    },
+    {
+      "epoch": 2.6646670665866825,
+      "grad_norm": 0.5731552377828735,
+      "learning_rate": 3.722938068077825e-07,
+      "loss": 0.6043,
+      "step": 2222
+    },
+    {
+      "epoch": 2.665866826634673,
+      "grad_norm": 0.5699287440850611,
+      "learning_rate": 3.6965369376836427e-07,
+      "loss": 0.5435,
+      "step": 2223
+    },
+    {
+      "epoch": 2.6670665866826635,
+      "grad_norm": 0.6513608062882078,
+      "learning_rate": 3.670226158554396e-07,
+      "loss": 0.499,
+      "step": 2224
+    },
+    {
+      "epoch": 2.668266346730654,
+      "grad_norm": 0.575845986646372,
+      "learning_rate": 3.6440057820299577e-07,
+      "loss": 0.5676,
+      "step": 2225
+    },
+    {
+      "epoch": 2.6694661067786445,
+      "grad_norm": 0.5656523199158348,
+      "learning_rate": 3.617875859273756e-07,
+      "loss": 0.5739,
+      "step": 2226
+    },
+    {
+      "epoch": 2.6706658668266345,
+      "grad_norm": 0.5928884294996334,
+      "learning_rate": 3.5918364412727004e-07,
+      "loss": 0.5755,
+      "step": 2227
+    },
+    {
+      "epoch": 2.671865626874625,
+      "grad_norm": 0.6500031397126891,
+      "learning_rate": 3.565887578837152e-07,
+      "loss": 0.5345,
+      "step": 2228
+    },
+    {
+      "epoch": 2.6730653869226155,
+      "grad_norm": 0.6618145889563011,
+      "learning_rate": 3.5400293226007296e-07,
+      "loss": 0.473,
+      "step": 2229
+    },
+    {
+      "epoch": 2.674265146970606,
+      "grad_norm": 0.5874887604233205,
+      "learning_rate": 3.514261723020268e-07,
+      "loss": 0.5536,
+      "step": 2230
+    },
+    {
+      "epoch": 2.675464907018596,
+      "grad_norm": 0.7232082935107603,
+      "learning_rate": 3.4885848303756963e-07,
+      "loss": 0.4964,
+      "step": 2231
+    },
+    {
+      "epoch": 2.6766646670665866,
+      "grad_norm": 0.6363229555016645,
+      "learning_rate": 3.462998694769976e-07,
+      "loss": 0.5929,
+      "step": 2232
+    },
+    {
+      "epoch": 2.677864427114577,
+      "grad_norm": 0.5800160738933541,
+      "learning_rate": 3.4375033661289413e-07,
+      "loss": 0.5662,
+      "step": 2233
+    },
+    {
+      "epoch": 2.6790641871625676,
+      "grad_norm": 0.8804672806059494,
+      "learning_rate": 3.412098894201255e-07,
+      "loss": 0.6181,
+      "step": 2234
+    },
+    {
+      "epoch": 2.680263947210558,
+      "grad_norm": 0.524971145657382,
+      "learning_rate": 3.3867853285582894e-07,
+      "loss": 0.5407,
+      "step": 2235
+    },
+    {
+      "epoch": 2.6814637072585485,
+      "grad_norm": 0.6100957559110444,
+      "learning_rate": 3.361562718594036e-07,
+      "loss": 0.504,
+      "step": 2236
+    },
+    {
+      "epoch": 2.6826634673065386,
+      "grad_norm": 0.5354284577217275,
+      "learning_rate": 3.336431113524996e-07,
+      "loss": 0.556,
+      "step": 2237
+    },
+    {
+      "epoch": 2.683863227354529,
+      "grad_norm": 0.5678265948723596,
+      "learning_rate": 3.311390562390099e-07,
+      "loss": 0.6007,
+      "step": 2238
+    },
+    {
+      "epoch": 2.6850629874025196,
+      "grad_norm": 0.685672769832599,
+      "learning_rate": 3.2864411140506094e-07,
+      "loss": 0.6002,
+      "step": 2239
+    },
+    {
+      "epoch": 2.6862627474505096,
+      "grad_norm": 0.6067651086831962,
+      "learning_rate": 3.2615828171900234e-07,
+      "loss": 0.5837,
+      "step": 2240
+    },
+    {
+      "epoch": 2.6874625074985,
+      "grad_norm": 0.5540343155655241,
+      "learning_rate": 3.2368157203139404e-07,
+      "loss": 0.6169,
+      "step": 2241
+    },
+    {
+      "epoch": 2.6886622675464906,
+      "grad_norm": 0.5243216125695884,
+      "learning_rate": 3.212139871750064e-07,
+      "loss": 0.5947,
+      "step": 2242
+    },
+    {
+      "epoch": 2.689862027594481,
+      "grad_norm": 0.5166214552324191,
+      "learning_rate": 3.1875553196480056e-07,
+      "loss": 0.5611,
+      "step": 2243
+    },
+    {
+      "epoch": 2.6910617876424716,
+      "grad_norm": 0.5380333721802236,
+      "learning_rate": 3.1630621119792203e-07,
+      "loss": 0.4938,
+      "step": 2244
+    },
+    {
+      "epoch": 2.692261547690462,
+      "grad_norm": 0.5420827111585474,
+      "learning_rate": 3.138660296536966e-07,
+      "loss": 0.5731,
+      "step": 2245
+    },
+    {
+      "epoch": 2.693461307738452,
+      "grad_norm": 0.6244285648915006,
+      "learning_rate": 3.114349920936144e-07,
+      "loss": 0.5156,
+      "step": 2246
+    },
+    {
+      "epoch": 2.6946610677864427,
+      "grad_norm": 0.6573495030556705,
+      "learning_rate": 3.090131032613225e-07,
+      "loss": 0.547,
+      "step": 2247
+    },
+    {
+      "epoch": 2.695860827834433,
+      "grad_norm": 0.5442357185305635,
+      "learning_rate": 3.066003678826152e-07,
+      "loss": 0.5845,
+      "step": 2248
+    },
+    {
+      "epoch": 2.6970605878824236,
+      "grad_norm": 0.544062706584872,
+      "learning_rate": 3.041967906654314e-07,
+      "loss": 0.5996,
+      "step": 2249
+    },
+    {
+      "epoch": 2.6982603479304137,
+      "grad_norm": 0.9578707468901981,
+      "learning_rate": 3.018023762998329e-07,
+      "loss": 0.5733,
+      "step": 2250
+    },
+    {
+      "epoch": 2.699460107978404,
+      "grad_norm": 0.5953350417695669,
+      "learning_rate": 2.994171294580045e-07,
+      "loss": 0.5834,
+      "step": 2251
+    },
+    {
+      "epoch": 2.7006598680263947,
+      "grad_norm": 0.5622708003061522,
+      "learning_rate": 2.9704105479424705e-07,
+      "loss": 0.5551,
+      "step": 2252
+    },
+    {
+      "epoch": 2.701859628074385,
+      "grad_norm": 0.5305691541375912,
+      "learning_rate": 2.946741569449563e-07,
+      "loss": 0.5684,
+      "step": 2253
+    },
+    {
+      "epoch": 2.7030593881223757,
+      "grad_norm": 0.5249312174708018,
+      "learning_rate": 2.923164405286266e-07,
+      "loss": 0.5662,
+      "step": 2254
+    },
+    {
+      "epoch": 2.704259148170366,
+      "grad_norm": 0.6204603022129775,
+      "learning_rate": 2.8996791014583523e-07,
+      "loss": 0.613,
+      "step": 2255
+    },
+    {
+      "epoch": 2.705458908218356,
+      "grad_norm": 0.6978090453715121,
+      "learning_rate": 2.8762857037923475e-07,
+      "loss": 0.5123,
+      "step": 2256
+    },
+    {
+      "epoch": 2.7066586682663467,
+      "grad_norm": 0.5800125254374642,
+      "learning_rate": 2.852984257935448e-07,
+      "loss": 0.5672,
+      "step": 2257
+    },
+    {
+      "epoch": 2.707858428314337,
+      "grad_norm": 1.2967018180608374,
+      "learning_rate": 2.8297748093554125e-07,
+      "loss": 0.6101,
+      "step": 2258
+    },
+    {
+      "epoch": 2.7090581883623273,
+      "grad_norm": 0.5642756332587243,
+      "learning_rate": 2.8066574033405014e-07,
+      "loss": 0.6105,
+      "step": 2259
+    },
+    {
+      "epoch": 2.7102579484103178,
+      "grad_norm": 0.7239876886430663,
+      "learning_rate": 2.783632084999366e-07,
+      "loss": 0.5856,
+      "step": 2260
+    },
+    {
+      "epoch": 2.7114577084583082,
+      "grad_norm": 0.5860625340774893,
+      "learning_rate": 2.7606988992609773e-07,
+      "loss": 0.6328,
+      "step": 2261
+    },
+    {
+      "epoch": 2.7126574685062987,
+      "grad_norm": 0.5655581547072107,
+      "learning_rate": 2.737857890874518e-07,
+      "loss": 0.5996,
+      "step": 2262
+    },
+    {
+      "epoch": 2.7138572285542892,
+      "grad_norm": 16.76224385771302,
+      "learning_rate": 2.7151091044093083e-07,
+      "loss": 0.9452,
+      "step": 2263
+    },
+    {
+      "epoch": 2.7150569886022797,
+      "grad_norm": 0.6382218253421764,
+      "learning_rate": 2.6924525842547286e-07,
+      "loss": 0.5567,
+      "step": 2264
+    },
+    {
+      "epoch": 2.71625674865027,
+      "grad_norm": 0.6517148416150133,
+      "learning_rate": 2.669888374620094e-07,
+      "loss": 0.5537,
+      "step": 2265
+    },
+    {
+      "epoch": 2.7174565086982603,
+      "grad_norm": 0.5644663327138808,
+      "learning_rate": 2.6474165195346346e-07,
+      "loss": 0.5069,
+      "step": 2266
+    },
+    {
+      "epoch": 2.7186562687462508,
+      "grad_norm": 0.6619100893862593,
+      "learning_rate": 2.6250370628473465e-07,
+      "loss": 0.5838,
+      "step": 2267
+    },
+    {
+      "epoch": 2.7198560287942413,
+      "grad_norm": 0.6201250184997715,
+      "learning_rate": 2.6027500482269153e-07,
+      "loss": 0.5597,
+      "step": 2268
+    },
+    {
+      "epoch": 2.7210557888422313,
+      "grad_norm": 0.6363157832150368,
+      "learning_rate": 2.580555519161698e-07,
+      "loss": 0.5632,
+      "step": 2269
+    },
+    {
+      "epoch": 2.722255548890222,
+      "grad_norm": 0.6064969590377077,
+      "learning_rate": 2.5584535189595193e-07,
+      "loss": 0.5688,
+      "step": 2270
+    },
+    {
+      "epoch": 2.7234553089382123,
+      "grad_norm": 0.591823448277897,
+      "learning_rate": 2.536444090747703e-07,
+      "loss": 0.5613,
+      "step": 2271
+    },
+    {
+      "epoch": 2.724655068986203,
+      "grad_norm": 0.6261058741329462,
+      "learning_rate": 2.514527277472917e-07,
+      "loss": 0.5932,
+      "step": 2272
+    },
+    {
+      "epoch": 2.7258548290341933,
+      "grad_norm": 0.5497707289720722,
+      "learning_rate": 2.4927031219011133e-07,
+      "loss": 0.5965,
+      "step": 2273
+    },
+    {
+      "epoch": 2.727054589082184,
+      "grad_norm": 0.6196870230550074,
+      "learning_rate": 2.470971666617455e-07,
+      "loss": 0.5696,
+      "step": 2274
+    },
+    {
+      "epoch": 2.728254349130174,
+      "grad_norm": 0.6006894934472979,
+      "learning_rate": 2.449332954026201e-07,
+      "loss": 0.5852,
+      "step": 2275
+    },
+    {
+      "epoch": 2.7294541091781643,
+      "grad_norm": 0.5761744307987052,
+      "learning_rate": 2.427787026350659e-07,
+      "loss": 0.5648,
+      "step": 2276
+    },
+    {
+      "epoch": 2.730653869226155,
+      "grad_norm": 0.6287395982883524,
+      "learning_rate": 2.4063339256330695e-07,
+      "loss": 0.5975,
+      "step": 2277
+    },
+    {
+      "epoch": 2.7318536292741453,
+      "grad_norm": 0.6167967352976029,
+      "learning_rate": 2.3849736937345657e-07,
+      "loss": 0.5998,
+      "step": 2278
+    },
+    {
+      "epoch": 2.7330533893221354,
+      "grad_norm": 0.5976809733384075,
+      "learning_rate": 2.363706372335045e-07,
+      "loss": 0.6361,
+      "step": 2279
+    },
+    {
+      "epoch": 2.734253149370126,
+      "grad_norm": 0.5309139536680249,
+      "learning_rate": 2.3425320029331177e-07,
+      "loss": 0.6062,
+      "step": 2280
+    },
+    {
+      "epoch": 2.7354529094181164,
+      "grad_norm": 0.5643678683098926,
+      "learning_rate": 2.3214506268460224e-07,
+      "loss": 0.5474,
+      "step": 2281
+    },
+    {
+      "epoch": 2.736652669466107,
+      "grad_norm": 0.6645838525987101,
+      "learning_rate": 2.3004622852095372e-07,
+      "loss": 0.4881,
+      "step": 2282
+    },
+    {
+      "epoch": 2.7378524295140974,
+      "grad_norm": 0.5958003697026015,
+      "learning_rate": 2.279567018977902e-07,
+      "loss": 0.5989,
+      "step": 2283
+    },
+    {
+      "epoch": 2.739052189562088,
+      "grad_norm": 0.6578014953406578,
+      "learning_rate": 2.2587648689237461e-07,
+      "loss": 0.6215,
+      "step": 2284
+    },
+    {
+      "epoch": 2.740251949610078,
+      "grad_norm": 0.6076524618518147,
+      "learning_rate": 2.2380558756379788e-07,
+      "loss": 0.5891,
+      "step": 2285
+    },
+    {
+      "epoch": 2.7414517096580684,
+      "grad_norm": 0.6184516057759724,
+      "learning_rate": 2.2174400795297814e-07,
+      "loss": 0.5446,
+      "step": 2286
+    },
+    {
+      "epoch": 2.742651469706059,
+      "grad_norm": 0.5844299039413995,
+      "learning_rate": 2.196917520826447e-07,
+      "loss": 0.5247,
+      "step": 2287
+    },
+    {
+      "epoch": 2.743851229754049,
+      "grad_norm": 0.5706212187640874,
+      "learning_rate": 2.1764882395733266e-07,
+      "loss": 0.5421,
+      "step": 2288
+    },
+    {
+      "epoch": 2.7450509898020394,
+      "grad_norm": 0.63442375775085,
+      "learning_rate": 2.1561522756337773e-07,
+      "loss": 0.6292,
+      "step": 2289
+    },
+    {
+      "epoch": 2.74625074985003,
+      "grad_norm": 0.6434067374513008,
+      "learning_rate": 2.1359096686890845e-07,
+      "loss": 0.5952,
+      "step": 2290
+    },
+    {
+      "epoch": 2.7474505098980204,
+      "grad_norm": 0.6033121054333428,
+      "learning_rate": 2.1157604582383308e-07,
+      "loss": 0.5912,
+      "step": 2291
+    },
+    {
+      "epoch": 2.748650269946011,
+      "grad_norm": 0.6444524526744027,
+      "learning_rate": 2.0957046835983764e-07,
+      "loss": 0.5709,
+      "step": 2292
+    },
+    {
+      "epoch": 2.7498500299940014,
+      "grad_norm": 0.6249267289836973,
+      "learning_rate": 2.0757423839037728e-07,
+      "loss": 0.5662,
+      "step": 2293
+    },
+    {
+      "epoch": 2.7510497900419915,
+      "grad_norm": 0.6596582768889766,
+      "learning_rate": 2.0558735981066445e-07,
+      "loss": 0.566,
+      "step": 2294
+    },
+    {
+      "epoch": 2.752249550089982,
+      "grad_norm": 0.6574313280601249,
+      "learning_rate": 2.0360983649766685e-07,
+      "loss": 0.6004,
+      "step": 2295
+    },
+    {
+      "epoch": 2.7534493101379725,
+      "grad_norm": 0.6199832270141975,
+      "learning_rate": 2.0164167231009612e-07,
+      "loss": 0.5257,
+      "step": 2296
+    },
+    {
+      "epoch": 2.754649070185963,
+      "grad_norm": 0.6445693406967722,
+      "learning_rate": 1.9968287108840246e-07,
+      "loss": 0.5262,
+      "step": 2297
+    },
+    {
+      "epoch": 2.755848830233953,
+      "grad_norm": 0.6079282676234945,
+      "learning_rate": 1.9773343665476563e-07,
+      "loss": 0.545,
+      "step": 2298
+    },
+    {
+      "epoch": 2.7570485902819435,
+      "grad_norm": 0.6526380557697038,
+      "learning_rate": 1.957933728130884e-07,
+      "loss": 0.5556,
+      "step": 2299
+    },
+    {
+      "epoch": 2.758248350329934,
+      "grad_norm": 0.6277477478771757,
+      "learning_rate": 1.9386268334898817e-07,
+      "loss": 0.5109,
+      "step": 2300
+    },
+    {
+      "epoch": 2.7594481103779245,
+      "grad_norm": 0.9598483982453556,
+      "learning_rate": 1.9194137202979246e-07,
+      "loss": 0.5581,
+      "step": 2301
+    },
+    {
+      "epoch": 2.760647870425915,
+      "grad_norm": 0.5113688585036608,
+      "learning_rate": 1.9002944260452517e-07,
+      "loss": 0.5743,
+      "step": 2302
+    },
+    {
+      "epoch": 2.7618476304739055,
+      "grad_norm": 0.529723461042736,
+      "learning_rate": 1.8812689880390812e-07,
+      "loss": 0.6466,
+      "step": 2303
+    },
+    {
+      "epoch": 2.7630473905218955,
+      "grad_norm": 0.5898958642033482,
+      "learning_rate": 1.862337443403467e-07,
+      "loss": 0.5929,
+      "step": 2304
+    },
+    {
+      "epoch": 2.764247150569886,
+      "grad_norm": 0.6898063775649348,
+      "learning_rate": 1.8434998290792373e-07,
+      "loss": 0.5399,
+      "step": 2305
+    },
+    {
+      "epoch": 2.7654469106178765,
+      "grad_norm": 0.5779799023026893,
+      "learning_rate": 1.8247561818239723e-07,
+      "loss": 0.5641,
+      "step": 2306
+    },
+    {
+      "epoch": 2.7666466706658666,
+      "grad_norm": 0.6202471124125889,
+      "learning_rate": 1.806106538211866e-07,
+      "loss": 0.5772,
+      "step": 2307
+    },
+    {
+      "epoch": 2.767846430713857,
+      "grad_norm": 0.5851019466251818,
+      "learning_rate": 1.7875509346336917e-07,
+      "loss": 0.5765,
+      "step": 2308
+    },
+    {
+      "epoch": 2.7690461907618475,
+      "grad_norm": 0.6024107261524307,
+      "learning_rate": 1.7690894072967146e-07,
+      "loss": 0.581,
+      "step": 2309
+    },
+    {
+      "epoch": 2.770245950809838,
+      "grad_norm": 0.5510832385681983,
+      "learning_rate": 1.7507219922246688e-07,
+      "loss": 0.5625,
+      "step": 2310
+    },
+    {
+      "epoch": 2.7714457108578285,
+      "grad_norm": 0.589613404873875,
+      "learning_rate": 1.7324487252575907e-07,
+      "loss": 0.5761,
+      "step": 2311
+    },
+    {
+      "epoch": 2.772645470905819,
+      "grad_norm": 0.5682011883468383,
+      "learning_rate": 1.7142696420518523e-07,
+      "loss": 0.5825,
+      "step": 2312
+    },
+    {
+      "epoch": 2.773845230953809,
+      "grad_norm": 0.5897501239285461,
+      "learning_rate": 1.6961847780800346e-07,
+      "loss": 0.607,
+      "step": 2313
+    },
+    {
+      "epoch": 2.7750449910017996,
+      "grad_norm": 0.4809755268612014,
+      "learning_rate": 1.678194168630859e-07,
+      "loss": 0.5588,
+      "step": 2314
+    },
+    {
+      "epoch": 2.77624475104979,
+      "grad_norm": 0.6118928390334075,
+      "learning_rate": 1.6602978488091336e-07,
+      "loss": 0.6085,
+      "step": 2315
+    },
+    {
+      "epoch": 2.7774445110977806,
+      "grad_norm": 0.6209341488451864,
+      "learning_rate": 1.642495853535686e-07,
+      "loss": 0.5666,
+      "step": 2316
+    },
+    {
+      "epoch": 2.7786442711457706,
+      "grad_norm": 0.586087352341755,
+      "learning_rate": 1.6247882175472906e-07,
+      "loss": 0.5868,
+      "step": 2317
+    },
+    {
+      "epoch": 2.779844031193761,
+      "grad_norm": 0.5936490028024038,
+      "learning_rate": 1.6071749753965914e-07,
+      "loss": 0.5987,
+      "step": 2318
+    },
+    {
+      "epoch": 2.7810437912417516,
+      "grad_norm": 0.5932244561766362,
+      "learning_rate": 1.5896561614520578e-07,
+      "loss": 0.5096,
+      "step": 2319
+    },
+    {
+      "epoch": 2.782243551289742,
+      "grad_norm": 0.6959133173132499,
+      "learning_rate": 1.5722318098978783e-07,
+      "loss": 0.5321,
+      "step": 2320
+    },
+    {
+      "epoch": 2.7834433113377326,
+      "grad_norm": 0.5360416449049131,
+      "learning_rate": 1.5549019547339506e-07,
+      "loss": 0.5818,
+      "step": 2321
+    },
+    {
+      "epoch": 2.784643071385723,
+      "grad_norm": 0.5863954015074843,
+      "learning_rate": 1.537666629775747e-07,
+      "loss": 0.5521,
+      "step": 2322
+    },
+    {
+      "epoch": 2.785842831433713,
+      "grad_norm": 0.4947675242004059,
+      "learning_rate": 1.5205258686543212e-07,
+      "loss": 0.5907,
+      "step": 2323
+    },
+    {
+      "epoch": 2.7870425914817036,
+      "grad_norm": 0.5860510775749931,
+      "learning_rate": 1.5034797048161742e-07,
+      "loss": 0.5598,
+      "step": 2324
+    },
+    {
+      "epoch": 2.788242351529694,
+      "grad_norm": 0.6242552752538505,
+      "learning_rate": 1.4865281715232437e-07,
+      "loss": 0.6003,
+      "step": 2325
+    },
+    {
+      "epoch": 2.7894421115776846,
+      "grad_norm": 0.5525242477859322,
+      "learning_rate": 1.4696713018527932e-07,
+      "loss": 0.5539,
+      "step": 2326
+    },
+    {
+      "epoch": 2.7906418716256747,
+      "grad_norm": 0.5416778071383951,
+      "learning_rate": 1.4529091286973994e-07,
+      "loss": 0.5364,
+      "step": 2327
+    },
+    {
+      "epoch": 2.791841631673665,
+      "grad_norm": 0.5563327851110774,
+      "learning_rate": 1.4362416847648387e-07,
+      "loss": 0.6012,
+      "step": 2328
+    },
+    {
+      "epoch": 2.7930413917216557,
+      "grad_norm": 0.6265224455532714,
+      "learning_rate": 1.419669002578039e-07,
+      "loss": 0.5578,
+      "step": 2329
+    },
+    {
+      "epoch": 2.794241151769646,
+      "grad_norm": 0.6064562054059573,
+      "learning_rate": 1.4031911144750443e-07,
+      "loss": 0.6092,
+      "step": 2330
+    },
+    {
+      "epoch": 2.7954409118176367,
+      "grad_norm": 0.6221941054992836,
+      "learning_rate": 1.386808052608918e-07,
+      "loss": 0.5907,
+      "step": 2331
+    },
+    {
+      "epoch": 2.796640671865627,
+      "grad_norm": 0.6617135526417072,
+      "learning_rate": 1.3705198489476712e-07,
+      "loss": 0.6039,
+      "step": 2332
+    },
+    {
+      "epoch": 2.797840431913617,
+      "grad_norm": 0.5885914853696609,
+      "learning_rate": 1.3543265352742418e-07,
+      "loss": 0.5131,
+      "step": 2333
+    },
+    {
+      "epoch": 2.7990401919616077,
+      "grad_norm": 0.5369215495584861,
+      "learning_rate": 1.338228143186404e-07,
+      "loss": 0.5668,
+      "step": 2334
+    },
+    {
+      "epoch": 2.800239952009598,
+      "grad_norm": 0.7015199653828252,
+      "learning_rate": 1.3222247040967195e-07,
+      "loss": 0.539,
+      "step": 2335
+    },
+    {
+      "epoch": 2.8014397120575882,
+      "grad_norm": 0.6373537615543214,
+      "learning_rate": 1.3063162492324533e-07,
+      "loss": 0.6098,
+      "step": 2336
+    },
+    {
+      "epoch": 2.8026394721055787,
+      "grad_norm": 0.6182980392879887,
+      "learning_rate": 1.2905028096355465e-07,
+      "loss": 0.5617,
+      "step": 2337
+    },
+    {
+      "epoch": 2.8038392321535692,
+      "grad_norm": 0.656616322139744,
+      "learning_rate": 1.2747844161625277e-07,
+      "loss": 0.5108,
+      "step": 2338
+    },
+    {
+      "epoch": 2.8050389922015597,
+      "grad_norm": 0.6934710935426451,
+      "learning_rate": 1.2591610994844682e-07,
+      "loss": 0.6423,
+      "step": 2339
+    },
+    {
+      "epoch": 2.80623875224955,
+      "grad_norm": 0.6103801170457688,
+      "learning_rate": 1.2436328900869098e-07,
+      "loss": 0.5763,
+      "step": 2340
+    },
+    {
+      "epoch": 2.8074385122975407,
+      "grad_norm": 0.5358157555678392,
+      "learning_rate": 1.228199818269826e-07,
+      "loss": 0.6331,
+      "step": 2341
+    },
+    {
+      "epoch": 2.8086382723455308,
+      "grad_norm": 0.5499433489191992,
+      "learning_rate": 1.2128619141475394e-07,
+      "loss": 0.5712,
+      "step": 2342
+    },
+    {
+      "epoch": 2.8098380323935213,
+      "grad_norm": 0.7654236897587715,
+      "learning_rate": 1.1976192076486703e-07,
+      "loss": 0.5075,
+      "step": 2343
+    },
+    {
+      "epoch": 2.8110377924415118,
+      "grad_norm": 0.5805466981639074,
+      "learning_rate": 1.1824717285160992e-07,
+      "loss": 0.5489,
+      "step": 2344
+    },
+    {
+      "epoch": 2.8122375524895022,
+      "grad_norm": 0.5859214634697024,
+      "learning_rate": 1.1674195063068771e-07,
+      "loss": 0.5515,
+      "step": 2345
+    },
+    {
+      "epoch": 2.8134373125374923,
+      "grad_norm": 0.615826387549419,
+      "learning_rate": 1.1524625703921654e-07,
+      "loss": 0.5766,
+      "step": 2346
+    },
+    {
+      "epoch": 2.814637072585483,
+      "grad_norm": 0.6412381130303643,
+      "learning_rate": 1.137600949957235e-07,
+      "loss": 0.6163,
+      "step": 2347
+    },
+    {
+      "epoch": 2.8158368326334733,
+      "grad_norm": 0.5747136469427545,
+      "learning_rate": 1.1228346740013385e-07,
+      "loss": 0.6252,
+      "step": 2348
+    },
+    {
+      "epoch": 2.817036592681464,
+      "grad_norm": 0.6039502257496154,
+      "learning_rate": 1.1081637713376892e-07,
+      "loss": 0.5536,
+      "step": 2349
+    },
+    {
+      "epoch": 2.8182363527294543,
+      "grad_norm": 0.5639887024242807,
+      "learning_rate": 1.0935882705933987e-07,
+      "loss": 0.5199,
+      "step": 2350
+    },
+    {
+      "epoch": 2.8194361127774448,
+      "grad_norm": 0.6164265957182835,
+      "learning_rate": 1.0791082002094444e-07,
+      "loss": 0.5222,
+      "step": 2351
+    },
+    {
+      "epoch": 2.820635872825435,
+      "grad_norm": 0.5936091059699564,
+      "learning_rate": 1.0647235884405638e-07,
+      "loss": 0.6108,
+      "step": 2352
+    },
+    {
+      "epoch": 2.8218356328734253,
+      "grad_norm": 0.5932654147724347,
+      "learning_rate": 1.0504344633552265e-07,
+      "loss": 0.5337,
+      "step": 2353
+    },
+    {
+      "epoch": 2.823035392921416,
+      "grad_norm": 0.5184278300606726,
+      "learning_rate": 1.0362408528356127e-07,
+      "loss": 0.5649,
+      "step": 2354
+    },
+    {
+      "epoch": 2.824235152969406,
+      "grad_norm": 0.5589941396527119,
+      "learning_rate": 1.0221427845774956e-07,
+      "loss": 0.5803,
+      "step": 2355
+    },
+    {
+      "epoch": 2.8254349130173964,
+      "grad_norm": 0.671790775421853,
+      "learning_rate": 1.0081402860902257e-07,
+      "loss": 0.5729,
+      "step": 2356
+    },
+    {
+      "epoch": 2.826634673065387,
+      "grad_norm": 0.5891561264062914,
+      "learning_rate": 9.942333846966745e-08,
+      "loss": 0.6147,
+      "step": 2357
+    },
+    {
+      "epoch": 2.8278344331133773,
+      "grad_norm": 0.654399315559274,
+      "learning_rate": 9.804221075331688e-08,
+      "loss": 0.5565,
+      "step": 2358
+    },
+    {
+      "epoch": 2.829034193161368,
+      "grad_norm": 0.5758731263307098,
+      "learning_rate": 9.66706481549462e-08,
+      "loss": 0.5397,
+      "step": 2359
+    },
+    {
+      "epoch": 2.8302339532093583,
+      "grad_norm": 0.7214093965701747,
+      "learning_rate": 9.530865335086403e-08,
+      "loss": 0.5648,
+      "step": 2360
+    },
+    {
+      "epoch": 2.8314337132573484,
+      "grad_norm": 0.6203261667989486,
+      "learning_rate": 9.395622899871171e-08,
+      "loss": 0.5816,
+      "step": 2361
+    },
+    {
+      "epoch": 2.832633473305339,
+      "grad_norm": 0.5976424430809402,
+      "learning_rate": 9.261337773745437e-08,
+      "loss": 0.5973,
+      "step": 2362
+    },
+    {
+      "epoch": 2.8338332333533294,
+      "grad_norm": 0.5973645134564878,
+      "learning_rate": 9.128010218737771e-08,
+      "loss": 0.5933,
+      "step": 2363
+    },
+    {
+      "epoch": 2.83503299340132,
+      "grad_norm": 0.6093648571259452,
+      "learning_rate": 8.99564049500834e-08,
+      "loss": 0.6046,
+      "step": 2364
+    },
+    {
+      "epoch": 2.83623275344931,
+      "grad_norm": 0.6355614257808114,
+      "learning_rate": 8.8642288608482e-08,
+      "loss": 0.5631,
+      "step": 2365
+    },
+    {
+      "epoch": 2.8374325134973004,
+      "grad_norm": 0.6078813029467055,
+      "learning_rate": 8.733775572678849e-08,
+      "loss": 0.5822,
+      "step": 2366
+    },
+    {
+      "epoch": 2.838632273545291,
+      "grad_norm": 0.6311747353130119,
+      "learning_rate": 8.604280885051996e-08,
+      "loss": 0.5456,
+      "step": 2367
+    },
+    {
+      "epoch": 2.8398320335932814,
+      "grad_norm": 0.6648083150307312,
+      "learning_rate": 8.475745050648687e-08,
+      "loss": 0.5614,
+      "step": 2368
+    },
+    {
+      "epoch": 2.841031793641272,
+      "grad_norm": 0.5762667824699731,
+      "learning_rate": 8.34816832027896e-08,
+      "loss": 0.6001,
+      "step": 2369
+    },
+    {
+      "epoch": 2.8422315536892624,
+      "grad_norm": 0.5498158183758537,
+      "learning_rate": 8.221550942881406e-08,
+      "loss": 0.5935,
+      "step": 2370
+    },
+    {
+      "epoch": 2.8434313137372524,
+      "grad_norm": 0.6219391351167715,
+      "learning_rate": 8.095893165522728e-08,
+      "loss": 0.5799,
+      "step": 2371
+    },
+    {
+      "epoch": 2.844631073785243,
+      "grad_norm": 0.6413960554893885,
+      "learning_rate": 7.971195233397011e-08,
+      "loss": 0.5447,
+      "step": 2372
+    },
+    {
+      "epoch": 2.8458308338332334,
+      "grad_norm": 0.5827133421232401,
+      "learning_rate": 7.847457389825507e-08,
+      "loss": 0.5537,
+      "step": 2373
+    },
+    {
+      "epoch": 2.8470305938812235,
+      "grad_norm": 0.6326857277662814,
+      "learning_rate": 7.724679876256024e-08,
+      "loss": 0.5823,
+      "step": 2374
+    },
+    {
+      "epoch": 2.848230353929214,
+      "grad_norm": 0.6354885512998139,
+      "learning_rate": 7.602862932262589e-08,
+      "loss": 0.5863,
+      "step": 2375
+    },
+    {
+      "epoch": 2.8494301139772045,
+      "grad_norm": 0.6748462246435095,
+      "learning_rate": 7.482006795544783e-08,
+      "loss": 0.5455,
+      "step": 2376
+    },
+    {
+      "epoch": 2.850629874025195,
+      "grad_norm": 0.5673661368960419,
+      "learning_rate": 7.362111701927354e-08,
+      "loss": 0.5251,
+      "step": 2377
+    },
+    {
+      "epoch": 2.8518296340731855,
+      "grad_norm": 0.5556661562224987,
+      "learning_rate": 7.243177885359831e-08,
+      "loss": 0.5417,
+      "step": 2378
+    },
+    {
+      "epoch": 2.853029394121176,
+      "grad_norm": 0.6581598350545725,
+      "learning_rate": 7.125205577916072e-08,
+      "loss": 0.5949,
+      "step": 2379
+    },
+    {
+      "epoch": 2.854229154169166,
+      "grad_norm": 0.5916519804123358,
+      "learning_rate": 7.008195009793717e-08,
+      "loss": 0.6273,
+      "step": 2380
+    },
+    {
+      "epoch": 2.8554289142171565,
+      "grad_norm": 0.6805425376195103,
+      "learning_rate": 6.892146409313738e-08,
+      "loss": 0.5406,
+      "step": 2381
+    },
+    {
+      "epoch": 2.856628674265147,
+      "grad_norm": 0.5976912878047412,
+      "learning_rate": 6.77706000292e-08,
+      "loss": 0.5711,
+      "step": 2382
+    },
+    {
+      "epoch": 2.8578284343131375,
+      "grad_norm": 0.5729419043066288,
+      "learning_rate": 6.662936015178978e-08,
+      "loss": 0.6223,
+      "step": 2383
+    },
+    {
+      "epoch": 2.8590281943611275,
+      "grad_norm": 0.6842825240551157,
+      "learning_rate": 6.549774668779097e-08,
+      "loss": 0.5314,
+      "step": 2384
+    },
+    {
+      "epoch": 2.860227954409118,
+      "grad_norm": 0.5627449880625564,
+      "learning_rate": 6.437576184530447e-08,
+      "loss": 0.5852,
+      "step": 2385
+    },
+    {
+      "epoch": 2.8614277144571085,
+      "grad_norm": 0.6166357064036266,
+      "learning_rate": 6.326340781364238e-08,
+      "loss": 0.5233,
+      "step": 2386
+    },
+    {
+      "epoch": 2.862627474505099,
+      "grad_norm": 0.6309257685754714,
+      "learning_rate": 6.21606867633251e-08,
+      "loss": 0.5371,
+      "step": 2387
+    },
+    {
+      "epoch": 2.8638272345530895,
+      "grad_norm": 0.5931088942775689,
+      "learning_rate": 6.106760084607589e-08,
+      "loss": 0.5298,
+      "step": 2388
+    },
+    {
+      "epoch": 2.86502699460108,
+      "grad_norm": 0.6740502744496198,
+      "learning_rate": 5.998415219481746e-08,
+      "loss": 0.5738,
+      "step": 2389
+    },
+    {
+      "epoch": 2.86622675464907,
+      "grad_norm": 0.5451730412062963,
+      "learning_rate": 5.8910342923666486e-08,
+      "loss": 0.5101,
+      "step": 2390
+    },
+    {
+      "epoch": 2.8674265146970606,
+      "grad_norm": 0.5889645164578342,
+      "learning_rate": 5.784617512793134e-08,
+      "loss": 0.5556,
+      "step": 2391
+    },
+    {
+      "epoch": 2.868626274745051,
+      "grad_norm": 0.5443396605516243,
+      "learning_rate": 5.679165088410765e-08,
+      "loss": 0.6002,
+      "step": 2392
+    },
+    {
+      "epoch": 2.8698260347930415,
+      "grad_norm": 0.5962558841554856,
+      "learning_rate": 5.5746772249872797e-08,
+      "loss": 0.5479,
+      "step": 2393
+    },
+    {
+      "epoch": 2.8710257948410316,
+      "grad_norm": 0.6561710368807768,
+      "learning_rate": 5.4711541264082516e-08,
+      "loss": 0.563,
+      "step": 2394
+    },
+    {
+      "epoch": 2.872225554889022,
+      "grad_norm": 0.5435010579982906,
+      "learning_rate": 5.368595994676762e-08,
+      "loss": 0.5816,
+      "step": 2395
+    },
+    {
+      "epoch": 2.8734253149370126,
+      "grad_norm": 0.5893099934206002,
+      "learning_rate": 5.267003029913065e-08,
+      "loss": 0.5841,
+      "step": 2396
+    },
+    {
+      "epoch": 2.874625074985003,
+      "grad_norm": 0.5737803336900801,
+      "learning_rate": 5.166375430353976e-08,
+      "loss": 0.5323,
+      "step": 2397
+    },
+    {
+      "epoch": 2.8758248350329936,
+      "grad_norm": 0.6553772837801549,
+      "learning_rate": 5.066713392352651e-08,
+      "loss": 0.4994,
+      "step": 2398
+    },
+    {
+      "epoch": 2.877024595080984,
+      "grad_norm": 0.6562504677178241,
+      "learning_rate": 4.9680171103781425e-08,
+      "loss": 0.6203,
+      "step": 2399
+    },
+    {
+      "epoch": 2.878224355128974,
+      "grad_norm": 0.5349033934562512,
+      "learning_rate": 4.8702867770150655e-08,
+      "loss": 0.5742,
+      "step": 2400
+    },
+    {
+      "epoch": 2.8794241151769646,
+      "grad_norm": 0.5570332821601537,
+      "learning_rate": 4.7735225829632084e-08,
+      "loss": 0.5724,
+      "step": 2401
+    },
+    {
+      "epoch": 2.880623875224955,
+      "grad_norm": 0.6557660398768234,
+      "learning_rate": 4.677724717037035e-08,
+      "loss": 0.6064,
+      "step": 2402
+    },
+    {
+      "epoch": 2.881823635272945,
+      "grad_norm": 0.5688158694700273,
+      "learning_rate": 4.582893366165631e-08,
+      "loss": 0.5199,
+      "step": 2403
+    },
+    {
+      "epoch": 2.8830233953209357,
+      "grad_norm": 0.7807570238399966,
+      "learning_rate": 4.489028715391808e-08,
+      "loss": 0.5327,
+      "step": 2404
+    },
+    {
+      "epoch": 2.884223155368926,
+      "grad_norm": 0.5401166517202306,
+      "learning_rate": 4.396130947872501e-08,
+      "loss": 0.5663,
+      "step": 2405
+    },
+    {
+      "epoch": 2.8854229154169166,
+      "grad_norm": 0.6043981734915697,
+      "learning_rate": 4.304200244877654e-08,
+      "loss": 0.5759,
+      "step": 2406
+    },
+    {
+      "epoch": 2.886622675464907,
+      "grad_norm": 0.593386744178214,
+      "learning_rate": 4.213236785790275e-08,
+      "loss": 0.5814,
+      "step": 2407
+    },
+    {
+      "epoch": 2.8878224355128976,
+      "grad_norm": 0.5952394648525249,
+      "learning_rate": 4.12324074810605e-08,
+      "loss": 0.6064,
+      "step": 2408
+    },
+    {
+      "epoch": 2.8890221955608877,
+      "grad_norm": 0.5647464562397199,
+      "learning_rate": 4.03421230743295e-08,
+      "loss": 0.5369,
+      "step": 2409
+    },
+    {
+      "epoch": 2.890221955608878,
+      "grad_norm": 0.5457714401248307,
+      "learning_rate": 3.946151637490792e-08,
+      "loss": 0.5919,
+      "step": 2410
+    },
+    {
+      "epoch": 2.8914217156568687,
+      "grad_norm": 0.6228834155462756,
+      "learning_rate": 3.859058910111069e-08,
+      "loss": 0.6165,
+      "step": 2411
+    },
+    {
+      "epoch": 2.892621475704859,
+      "grad_norm": 0.6845194732842805,
+      "learning_rate": 3.772934295236619e-08,
+      "loss": 0.5278,
+      "step": 2412
+    },
+    {
+      "epoch": 2.893821235752849,
+      "grad_norm": 0.7300397421212467,
+      "learning_rate": 3.6877779609210686e-08,
+      "loss": 0.5539,
+      "step": 2413
+    },
+    {
+      "epoch": 2.8950209958008397,
+      "grad_norm": 0.615417448310366,
+      "learning_rate": 3.6035900733286667e-08,
+      "loss": 0.5605,
+      "step": 2414
+    },
+    {
+      "epoch": 2.89622075584883,
+      "grad_norm": 0.5390319632170048,
+      "learning_rate": 3.5203707967341184e-08,
+      "loss": 0.6184,
+      "step": 2415
+    },
+    {
+      "epoch": 2.8974205158968207,
+      "grad_norm": 0.6135132318318581,
+      "learning_rate": 3.438120293521918e-08,
+      "loss": 0.4843,
+      "step": 2416
+    },
+    {
+      "epoch": 2.898620275944811,
+      "grad_norm": 0.6615575451967282,
+      "learning_rate": 3.35683872418624e-08,
+      "loss": 0.6066,
+      "step": 2417
+    },
+    {
+      "epoch": 2.8998200359928017,
+      "grad_norm": 0.7081757860107213,
+      "learning_rate": 3.276526247330602e-08,
+      "loss": 0.6354,
+      "step": 2418
+    },
+    {
+      "epoch": 2.9010197960407917,
+      "grad_norm": 0.6064346017295151,
+      "learning_rate": 3.197183019667594e-08,
+      "loss": 0.5965,
+      "step": 2419
+    },
+    {
+      "epoch": 2.9022195560887822,
+      "grad_norm": 0.6021962892961614,
+      "learning_rate": 3.118809196018424e-08,
+      "loss": 0.5247,
+      "step": 2420
+    },
+    {
+      "epoch": 2.9034193161367727,
+      "grad_norm": 0.5059717113947267,
+      "learning_rate": 3.0414049293128744e-08,
+      "loss": 0.6033,
+      "step": 2421
+    },
+    {
+      "epoch": 2.904619076184763,
+      "grad_norm": 0.5794323760670155,
+      "learning_rate": 2.964970370588738e-08,
+      "loss": 0.4934,
+      "step": 2422
+    },
+    {
+      "epoch": 2.9058188362327533,
+      "grad_norm": 0.6680157195088374,
+      "learning_rate": 2.889505668991599e-08,
+      "loss": 0.5562,
+      "step": 2423
+    },
+    {
+      "epoch": 2.9070185962807438,
+      "grad_norm": 0.669071371015542,
+      "learning_rate": 2.8150109717745566e-08,
+      "loss": 0.538,
+      "step": 2424
+    },
+    {
+      "epoch": 2.9082183563287343,
+      "grad_norm": 0.6467881845154935,
+      "learning_rate": 2.741486424298112e-08,
+      "loss": 0.5068,
+      "step": 2425
+    },
+    {
+      "epoch": 2.9094181163767248,
+      "grad_norm": 0.5615104696063138,
+      "learning_rate": 2.6689321700296144e-08,
+      "loss": 0.6049,
+      "step": 2426
+    },
+    {
+      "epoch": 2.9106178764247153,
+      "grad_norm": 0.5337904628569098,
+      "learning_rate": 2.597348350543094e-08,
+      "loss": 0.5577,
+      "step": 2427
+    },
+    {
+      "epoch": 2.9118176364727053,
+      "grad_norm": 0.6039259193524817,
+      "learning_rate": 2.5267351055190404e-08,
+      "loss": 0.6093,
+      "step": 2428
+    },
+    {
+      "epoch": 2.913017396520696,
+      "grad_norm": 0.5885852772977639,
+      "learning_rate": 2.4570925727439576e-08,
+      "loss": 0.5727,
+      "step": 2429
+    },
+    {
+      "epoch": 2.9142171565686863,
+      "grad_norm": 0.5597545658591652,
+      "learning_rate": 2.3884208881103654e-08,
+      "loss": 0.5568,
+      "step": 2430
+    },
+    {
+      "epoch": 2.915416916616677,
+      "grad_norm": 0.7074707681715321,
+      "learning_rate": 2.3207201856162432e-08,
+      "loss": 0.5879,
+      "step": 2431
+    },
+    {
+      "epoch": 2.916616676664667,
+      "grad_norm": 0.5535623547416776,
+      "learning_rate": 2.2539905973650855e-08,
+      "loss": 0.6107,
+      "step": 2432
+    },
+    {
+      "epoch": 2.9178164367126573,
+      "grad_norm": 0.6002257711241739,
+      "learning_rate": 2.1882322535652923e-08,
+      "loss": 0.5723,
+      "step": 2433
+    },
+    {
+      "epoch": 2.919016196760648,
+      "grad_norm": 0.7927136704041635,
+      "learning_rate": 2.1234452825300565e-08,
+      "loss": 0.5036,
+      "step": 2434
+    },
+    {
+      "epoch": 2.9202159568086383,
+      "grad_norm": 0.6268379548387949,
+      "learning_rate": 2.0596298106774214e-08,
+      "loss": 0.5084,
+      "step": 2435
+    },
+    {
+      "epoch": 2.921415716856629,
+      "grad_norm": 0.5571291168516115,
+      "learning_rate": 1.9967859625293905e-08,
+      "loss": 0.5239,
+      "step": 2436
+    },
+    {
+      "epoch": 2.9226154769046193,
+      "grad_norm": 0.5562198335939383,
+      "learning_rate": 1.934913860712373e-08,
+      "loss": 0.5042,
+      "step": 2437
+    },
+    {
+      "epoch": 2.9238152369526094,
+      "grad_norm": 0.6745555259181267,
+      "learning_rate": 1.8740136259564612e-08,
+      "loss": 0.573,
+      "step": 2438
+    },
+    {
+      "epoch": 2.9250149970006,
+      "grad_norm": 0.5578468264382872,
+      "learning_rate": 1.8140853770953758e-08,
+      "loss": 0.5526,
+      "step": 2439
+    },
+    {
+      "epoch": 2.9262147570485904,
+      "grad_norm": 0.48928727530060995,
+      "learning_rate": 1.7551292310661883e-08,
+      "loss": 0.5701,
+      "step": 2440
+    },
+    {
+      "epoch": 2.927414517096581,
+      "grad_norm": 0.6335310178542912,
+      "learning_rate": 1.697145302909209e-08,
+      "loss": 0.5229,
+      "step": 2441
+    },
+    {
+      "epoch": 2.928614277144571,
+      "grad_norm": 0.6410518832303463,
+      "learning_rate": 1.6401337057675437e-08,
+      "loss": 0.6148,
+      "step": 2442
+    },
+    {
+      "epoch": 2.9298140371925614,
+      "grad_norm": 0.6393388135655955,
+      "learning_rate": 1.584094550887205e-08,
+      "loss": 0.5227,
+      "step": 2443
+    },
+    {
+      "epoch": 2.931013797240552,
+      "grad_norm": 0.7323633673448885,
+      "learning_rate": 1.5290279476165014e-08,
+      "loss": 0.5459,
+      "step": 2444
+    },
+    {
+      "epoch": 2.9322135572885424,
+      "grad_norm": 0.6045893347103232,
+      "learning_rate": 1.4749340034060367e-08,
+      "loss": 0.5908,
+      "step": 2445
+    },
+    {
+      "epoch": 2.933413317336533,
+      "grad_norm": 0.6857979509852705,
+      "learning_rate": 1.4218128238086548e-08,
+      "loss": 0.5103,
+      "step": 2446
+    },
+    {
+      "epoch": 2.9346130773845234,
+      "grad_norm": 0.6239867493238006,
+      "learning_rate": 1.36966451247883e-08,
+      "loss": 0.5371,
+      "step": 2447
+    },
+    {
+      "epoch": 2.9358128374325134,
+      "grad_norm": 0.5863751459074298,
+      "learning_rate": 1.3184891711727766e-08,
+      "loss": 0.5873,
+      "step": 2448
+    },
+    {
+      "epoch": 2.937012597480504,
+      "grad_norm": 0.6548800154129585,
+      "learning_rate": 1.2682868997482279e-08,
+      "loss": 0.5607,
+      "step": 2449
+    },
+    {
+      "epoch": 2.9382123575284944,
+      "grad_norm": 0.6205175088081284,
+      "learning_rate": 1.2190577961641581e-08,
+      "loss": 0.546,
+      "step": 2450
+    },
+    {
+      "epoch": 2.9394121175764845,
+      "grad_norm": 0.5730112925542632,
+      "learning_rate": 1.1708019564805607e-08,
+      "loss": 0.4884,
+      "step": 2451
+    },
+    {
+      "epoch": 2.940611877624475,
+      "grad_norm": 0.5576631496485588,
+      "learning_rate": 1.1235194748583367e-08,
+      "loss": 0.6219,
+      "step": 2452
+    },
+    {
+      "epoch": 2.9418116376724655,
+      "grad_norm": 0.6177446793222537,
+      "learning_rate": 1.0772104435591291e-08,
+      "loss": 0.5779,
+      "step": 2453
+    },
+    {
+      "epoch": 2.943011397720456,
+      "grad_norm": 0.5488751270134903,
+      "learning_rate": 1.0318749529450445e-08,
+      "loss": 0.5111,
+      "step": 2454
+    },
+    {
+      "epoch": 2.9442111577684464,
+      "grad_norm": 0.5853988884195377,
+      "learning_rate": 9.875130914785424e-09,
+      "loss": 0.4794,
+      "step": 2455
+    },
+    {
+      "epoch": 2.945410917816437,
+      "grad_norm": 0.5746464436940514,
+      "learning_rate": 9.44124945722269e-09,
+      "loss": 0.5213,
+      "step": 2456
+    },
+    {
+      "epoch": 2.946610677864427,
+      "grad_norm": 0.5958493842054852,
+      "learning_rate": 9.017106003389453e-09,
+      "loss": 0.5468,
+      "step": 2457
+    },
+    {
+      "epoch": 2.9478104379124175,
+      "grad_norm": 0.5536530116313324,
+      "learning_rate": 8.602701380909795e-09,
+      "loss": 0.6127,
+      "step": 2458
+    },
+    {
+      "epoch": 2.949010197960408,
+      "grad_norm": 0.6009468479285284,
+      "learning_rate": 8.198036398405217e-09,
+      "loss": 0.5661,
+      "step": 2459
+    },
+    {
+      "epoch": 2.9502099580083985,
+      "grad_norm": 0.5506053937698066,
+      "learning_rate": 7.803111845493539e-09,
+      "loss": 0.5562,
+      "step": 2460
+    },
+    {
+      "epoch": 2.9514097180563885,
+      "grad_norm": 0.6695927647551572,
+      "learning_rate": 7.417928492784443e-09,
+      "loss": 0.5259,
+      "step": 2461
+    },
+    {
+      "epoch": 2.952609478104379,
+      "grad_norm": 0.6277211369298169,
+      "learning_rate": 7.042487091880046e-09,
+      "loss": 0.5805,
+      "step": 2462
+    },
+    {
+      "epoch": 2.9538092381523695,
+      "grad_norm": 0.6151065472376642,
+      "learning_rate": 6.676788375374887e-09,
+      "loss": 0.5123,
+      "step": 2463
+    },
+    {
+      "epoch": 2.95500899820036,
+      "grad_norm": 0.6197422431335191,
+      "learning_rate": 6.320833056850939e-09,
+      "loss": 0.5593,
+      "step": 2464
+    },
+    {
+      "epoch": 2.9562087582483505,
+      "grad_norm": 0.5881896045872286,
+      "learning_rate": 5.9746218308781576e-09,
+      "loss": 0.5372,
+      "step": 2465
+    },
+    {
+      "epoch": 2.957408518296341,
+      "grad_norm": 0.5461184226742699,
+      "learning_rate": 5.638155373013931e-09,
+      "loss": 0.6064,
+      "step": 2466
+    },
+    {
+      "epoch": 2.958608278344331,
+      "grad_norm": 0.5740321047604405,
+      "learning_rate": 5.311434339799748e-09,
+      "loss": 0.5821,
+      "step": 2467
+    },
+    {
+      "epoch": 2.9598080383923215,
+      "grad_norm": 0.5209288885463851,
+      "learning_rate": 4.994459368762306e-09,
+      "loss": 0.5358,
+      "step": 2468
+    },
+    {
+      "epoch": 2.961007798440312,
+      "grad_norm": 0.6130283562317994,
+      "learning_rate": 4.6872310784096266e-09,
+      "loss": 0.582,
+      "step": 2469
+    },
+    {
+      "epoch": 2.962207558488302,
+      "grad_norm": 0.6166228177199247,
+      "learning_rate": 4.389750068231613e-09,
+      "loss": 0.5364,
+      "step": 2470
+    },
+    {
+      "epoch": 2.9634073185362926,
+      "grad_norm": 0.6014700381489093,
+      "learning_rate": 4.102016918698937e-09,
+      "loss": 0.5706,
+      "step": 2471
+    },
+    {
+      "epoch": 2.964607078584283,
+      "grad_norm": 0.5523814773020451,
+      "learning_rate": 3.824032191261373e-09,
+      "loss": 0.5369,
+      "step": 2472
+    },
+    {
+      "epoch": 2.9658068386322736,
+      "grad_norm": 0.5692347250551049,
+      "learning_rate": 3.5557964283455814e-09,
+      "loss": 0.5781,
+      "step": 2473
+    },
+    {
+      "epoch": 2.967006598680264,
+      "grad_norm": 0.6636113231878251,
+      "learning_rate": 3.2973101533567698e-09,
+      "loss": 0.5406,
+      "step": 2474
+    },
+    {
+      "epoch": 2.9682063587282546,
+      "grad_norm": 0.5418052771926479,
+      "learning_rate": 3.048573870674809e-09,
+      "loss": 0.5626,
+      "step": 2475
+    },
+    {
+      "epoch": 2.9694061187762446,
+      "grad_norm": 0.552204788057103,
+      "learning_rate": 2.809588065655899e-09,
+      "loss": 0.5345,
+      "step": 2476
+    },
+    {
+      "epoch": 2.970605878824235,
+      "grad_norm": 0.601451785587039,
+      "learning_rate": 2.580353204628683e-09,
+      "loss": 0.5226,
+      "step": 2477
+    },
+    {
+      "epoch": 2.9718056388722256,
+      "grad_norm": 0.6129126790989069,
+      "learning_rate": 2.360869734896465e-09,
+      "loss": 0.5999,
+      "step": 2478
+    },
+    {
+      "epoch": 2.973005398920216,
+      "grad_norm": 0.560396843514585,
+      "learning_rate": 2.1511380847333287e-09,
+      "loss": 0.5495,
+      "step": 2479
+    },
+    {
+      "epoch": 2.974205158968206,
+      "grad_norm": 0.6612438320348933,
+      "learning_rate": 1.9511586633863545e-09,
+      "loss": 0.5255,
+      "step": 2480
+    },
+    {
+      "epoch": 2.9754049190161966,
+      "grad_norm": 0.566045451367403,
+      "learning_rate": 1.7609318610722902e-09,
+      "loss": 0.5438,
+      "step": 2481
+    },
+    {
+      "epoch": 2.976604679064187,
+      "grad_norm": 0.5837424311079994,
+      "learning_rate": 1.580458048976996e-09,
+      "loss": 0.5038,
+      "step": 2482
+    },
+    {
+      "epoch": 2.9778044391121776,
+      "grad_norm": 0.5534621281042921,
+      "learning_rate": 1.409737579257664e-09,
+      "loss": 0.6024,
+      "step": 2483
+    },
+    {
+      "epoch": 2.979004199160168,
+      "grad_norm": 0.6026104152195847,
+      "learning_rate": 1.2487707850378227e-09,
+      "loss": 0.6059,
+      "step": 2484
+    },
+    {
+      "epoch": 2.9802039592081586,
+      "grad_norm": 0.6385190115377162,
+      "learning_rate": 1.0975579804101132e-09,
+      "loss": 0.5992,
+      "step": 2485
+    },
+    {
+      "epoch": 2.9814037192561487,
+      "grad_norm": 0.649119435899077,
+      "learning_rate": 9.560994604329576e-10,
+      "loss": 0.5002,
+      "step": 2486
+    },
+    {
+      "epoch": 2.982603479304139,
+      "grad_norm": 0.5676956087255925,
+      "learning_rate": 8.243955011333349e-10,
+      "loss": 0.5228,
+      "step": 2487
+    },
+    {
+      "epoch": 2.9838032393521297,
+      "grad_norm": 0.6041311035308607,
+      "learning_rate": 7.024463595028952e-10,
+      "loss": 0.5831,
+      "step": 2488
+    },
+    {
+      "epoch": 2.9850029994001197,
+      "grad_norm": 0.6071102452722618,
+      "learning_rate": 5.90252273498515e-10,
+      "loss": 0.5929,
+      "step": 2489
+    },
+    {
+      "epoch": 2.98620275944811,
+      "grad_norm": 0.5810758766628411,
+      "learning_rate": 4.878134620439623e-10,
+      "loss": 0.5855,
+      "step": 2490
+    },
+    {
+      "epoch": 2.9874025194961007,
+      "grad_norm": 0.5902052789115674,
+      "learning_rate": 3.951301250260109e-10,
+      "loss": 0.5904,
+      "step": 2491
+    },
+    {
+      "epoch": 2.988602279544091,
+      "grad_norm": 0.5486510541340834,
+      "learning_rate": 3.122024432961057e-10,
+      "loss": 0.6075,
+      "step": 2492
+    },
+    {
+      "epoch": 2.9898020395920817,
+      "grad_norm": 0.6679678462887061,
+      "learning_rate": 2.390305786698077e-10,
+      "loss": 0.5877,
+      "step": 2493
+    },
+    {
+      "epoch": 2.991001799640072,
+      "grad_norm": 0.6475178912393099,
+      "learning_rate": 1.7561467392679387e-10,
+      "loss": 0.5602,
+      "step": 2494
+    },
+    {
+      "epoch": 2.9922015596880622,
+      "grad_norm": 0.6573905852671234,
+      "learning_rate": 1.2195485280974694e-10,
+      "loss": 0.5506,
+      "step": 2495
+    },
+    {
+      "epoch": 2.9934013197360527,
+      "grad_norm": 0.6839301032257503,
+      "learning_rate": 7.805122002324527e-11,
+      "loss": 0.5639,
+      "step": 2496
+    },
+    {
+      "epoch": 2.994601079784043,
+      "grad_norm": 0.6982652385995234,
+      "learning_rate": 4.3903861236538336e-11,
+      "loss": 0.6393,
+      "step": 2497
+    },
+    {
+      "epoch": 2.9958008398320337,
+      "grad_norm": 0.6221337547855504,
+      "learning_rate": 1.951284308077117e-11,
+      "loss": 0.5426,
+      "step": 2498
+    },
+    {
+      "epoch": 2.9970005998800238,
+      "grad_norm": 0.5416322473354866,
+      "learning_rate": 4.878213150094624e-12,
+      "loss": 0.6198,
+      "step": 2499
+    },
+    {
+      "epoch": 2.9970005998800238,
+      "step": 2499,
+      "total_flos": 461172498612224.0,
+      "train_loss": 0.7641865797355777,
+      "train_runtime": 21559.3093,
+      "train_samples_per_second": 1.392,
+      "train_steps_per_second": 0.116
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2499,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 461172498612224.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}