diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,13407 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.998676016280096,
+  "eval_steps": 500,
+  "global_step": 1911,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0015691658902564605,
+      "grad_norm": 11.272488257964362,
+      "learning_rate": 3.4482758620689656e-07,
+      "loss": 1.4352,
+      "step": 1
+    },
+    {
+      "epoch": 0.003138331780512921,
+      "grad_norm": 11.0931606459403,
+      "learning_rate": 6.896551724137931e-07,
+      "loss": 1.4128,
+      "step": 2
+    },
+    {
+      "epoch": 0.004707497670769382,
+      "grad_norm": 10.643291417717766,
+      "learning_rate": 1.0344827586206898e-06,
+      "loss": 1.3831,
+      "step": 3
+    },
+    {
+      "epoch": 0.006276663561025842,
+      "grad_norm": 11.830761571467178,
+      "learning_rate": 1.3793103448275862e-06,
+      "loss": 1.403,
+      "step": 4
+    },
+    {
+      "epoch": 0.007845829451282304,
+      "grad_norm": 10.521045392166496,
+      "learning_rate": 1.724137931034483e-06,
+      "loss": 1.4114,
+      "step": 5
+    },
+    {
+      "epoch": 0.009414995341538763,
+      "grad_norm": 9.521913823560627,
+      "learning_rate": 2.0689655172413796e-06,
+      "loss": 1.3692,
+      "step": 6
+    },
+    {
+      "epoch": 0.010984161231795225,
+      "grad_norm": 5.232682953056989,
+      "learning_rate": 2.4137931034482762e-06,
+      "loss": 1.3208,
+      "step": 7
+    },
+    {
+      "epoch": 0.012553327122051684,
+      "grad_norm": 4.476804897678295,
+      "learning_rate": 2.7586206896551725e-06,
+      "loss": 1.2706,
+      "step": 8
+    },
+    {
+      "epoch": 0.014122493012308146,
+      "grad_norm": 2.7967062768481865,
+      "learning_rate": 3.103448275862069e-06,
+      "loss": 1.1767,
+      "step": 9
+    },
+    {
+      "epoch": 0.015691658902564607,
+      "grad_norm": 2.5441620398429463,
+      "learning_rate": 3.448275862068966e-06,
+      "loss": 1.1364,
+      "step": 10
+    },
+    {
+      "epoch": 0.017260824792821067,
+      "grad_norm": 2.500120686959644,
+      "learning_rate": 3.793103448275862e-06,
+      "loss": 1.1842,
+      "step": 11
+    },
+    {
+      "epoch": 0.018829990683077526,
+      "grad_norm": 2.740073197861995,
+      "learning_rate": 4.137931034482759e-06,
+      "loss": 1.1003,
+      "step": 12
+    },
+    {
+      "epoch": 0.020399156573333986,
+      "grad_norm": 2.3918313029109917,
+      "learning_rate": 4.482758620689656e-06,
+      "loss": 1.1322,
+      "step": 13
+    },
+    {
+      "epoch": 0.02196832246359045,
+      "grad_norm": 2.2662733556116312,
+      "learning_rate": 4.8275862068965525e-06,
+      "loss": 1.1162,
+      "step": 14
+    },
+    {
+      "epoch": 0.02353748835384691,
+      "grad_norm": 2.1915447862782527,
+      "learning_rate": 5.172413793103449e-06,
+      "loss": 1.0935,
+      "step": 15
+    },
+    {
+      "epoch": 0.02510665424410337,
+      "grad_norm": 2.0421259400073,
+      "learning_rate": 5.517241379310345e-06,
+      "loss": 1.0534,
+      "step": 16
+    },
+    {
+      "epoch": 0.026675820134359828,
+      "grad_norm": 2.4390986023165158,
+      "learning_rate": 5.862068965517242e-06,
+      "loss": 1.0959,
+      "step": 17
+    },
+    {
+      "epoch": 0.02824498602461629,
+      "grad_norm": 1.8985984381891956,
+      "learning_rate": 6.206896551724138e-06,
+      "loss": 1.056,
+      "step": 18
+    },
+    {
+      "epoch": 0.02981415191487275,
+      "grad_norm": 1.6963799376959103,
+      "learning_rate": 6.551724137931035e-06,
+      "loss": 1.0459,
+      "step": 19
+    },
+    {
+      "epoch": 0.031383317805129214,
+      "grad_norm": 1.6281550972419943,
+      "learning_rate": 6.896551724137932e-06,
+      "loss": 1.042,
+      "step": 20
+    },
+    {
+      "epoch": 0.03295248369538567,
+      "grad_norm": 1.4261844326769773,
+      "learning_rate": 7.241379310344828e-06,
+      "loss": 1.0209,
+      "step": 21
+    },
+    {
+      "epoch": 0.034521649585642133,
+      "grad_norm": 1.3836614317055176,
+      "learning_rate": 7.586206896551724e-06,
+      "loss": 1.0288,
+      "step": 22
+    },
+    {
+      "epoch": 0.03609081547589859,
+      "grad_norm": 1.3366047431421009,
+      "learning_rate": 7.93103448275862e-06,
+      "loss": 1.0478,
+      "step": 23
+    },
+    {
+      "epoch": 0.03765998136615505,
+      "grad_norm": 1.3544976854281805,
+      "learning_rate": 8.275862068965518e-06,
+      "loss": 1.0472,
+      "step": 24
+    },
+    {
+      "epoch": 0.039229147256411516,
+      "grad_norm": 1.37085590741488,
+      "learning_rate": 8.620689655172414e-06,
+      "loss": 1.0243,
+      "step": 25
+    },
+    {
+      "epoch": 0.04079831314666797,
+      "grad_norm": 1.3399659661475207,
+      "learning_rate": 8.965517241379312e-06,
+      "loss": 1.0067,
+      "step": 26
+    },
+    {
+      "epoch": 0.042367479036924435,
+      "grad_norm": 1.1338660079511698,
+      "learning_rate": 9.310344827586207e-06,
+      "loss": 0.9967,
+      "step": 27
+    },
+    {
+      "epoch": 0.0439366449271809,
+      "grad_norm": 1.1473655664327627,
+      "learning_rate": 9.655172413793105e-06,
+      "loss": 0.989,
+      "step": 28
+    },
+    {
+      "epoch": 0.045505810817437355,
+      "grad_norm": 1.1519251849162189,
+      "learning_rate": 1e-05,
+      "loss": 0.9928,
+      "step": 29
+    },
+    {
+      "epoch": 0.04707497670769382,
+      "grad_norm": 1.1092645739751517,
+      "learning_rate": 1.0344827586206898e-05,
+      "loss": 0.9679,
+      "step": 30
+    },
+    {
+      "epoch": 0.048644142597950274,
+      "grad_norm": 1.0638236897062967,
+      "learning_rate": 1.0689655172413792e-05,
+      "loss": 0.9589,
+      "step": 31
+    },
+    {
+      "epoch": 0.05021330848820674,
+      "grad_norm": 0.9312615409095046,
+      "learning_rate": 1.103448275862069e-05,
+      "loss": 0.9473,
+      "step": 32
+    },
+    {
+      "epoch": 0.0517824743784632,
+      "grad_norm": 0.9660355103833085,
+      "learning_rate": 1.1379310344827587e-05,
+      "loss": 0.9828,
+      "step": 33
+    },
+    {
+      "epoch": 0.053351640268719656,
+      "grad_norm": 1.0466827883694991,
+      "learning_rate": 1.1724137931034483e-05,
+      "loss": 0.9772,
+      "step": 34
+    },
+    {
+      "epoch": 0.05492080615897612,
+      "grad_norm": 0.8931657272159612,
+      "learning_rate": 1.206896551724138e-05,
+      "loss": 0.9646,
+      "step": 35
+    },
+    {
+      "epoch": 0.05648997204923258,
+      "grad_norm": 0.8019826823207468,
+      "learning_rate": 1.2413793103448277e-05,
+      "loss": 0.9663,
+      "step": 36
+    },
+    {
+      "epoch": 0.05805913793948904,
+      "grad_norm": 0.8276045830440536,
+      "learning_rate": 1.2758620689655174e-05,
+      "loss": 0.9662,
+      "step": 37
+    },
+    {
+      "epoch": 0.0596283038297455,
+      "grad_norm": 0.923971173195591,
+      "learning_rate": 1.310344827586207e-05,
+      "loss": 0.9707,
+      "step": 38
+    },
+    {
+      "epoch": 0.06119746972000196,
+      "grad_norm": 0.8625258254200348,
+      "learning_rate": 1.3448275862068967e-05,
+      "loss": 0.9785,
+      "step": 39
+    },
+    {
+      "epoch": 0.06276663561025843,
+      "grad_norm": 0.7743708162100692,
+      "learning_rate": 1.3793103448275863e-05,
+      "loss": 0.9679,
+      "step": 40
+    },
+    {
+      "epoch": 0.06433580150051488,
+      "grad_norm": 0.8627254879884673,
+      "learning_rate": 1.4137931034482759e-05,
+      "loss": 0.9393,
+      "step": 41
+    },
+    {
+      "epoch": 0.06590496739077134,
+      "grad_norm": 0.9276663241128001,
+      "learning_rate": 1.4482758620689657e-05,
+      "loss": 0.965,
+      "step": 42
+    },
+    {
+      "epoch": 0.0674741332810278,
+      "grad_norm": 0.8040970997079206,
+      "learning_rate": 1.4827586206896554e-05,
+      "loss": 0.9449,
+      "step": 43
+    },
+    {
+      "epoch": 0.06904329917128427,
+      "grad_norm": 0.7654771283178015,
+      "learning_rate": 1.5172413793103448e-05,
+      "loss": 0.9292,
+      "step": 44
+    },
+    {
+      "epoch": 0.07061246506154073,
+      "grad_norm": 0.8331404099529315,
+      "learning_rate": 1.5517241379310346e-05,
+      "loss": 0.9264,
+      "step": 45
+    },
+    {
+      "epoch": 0.07218163095179718,
+      "grad_norm": 0.8582155957880822,
+      "learning_rate": 1.586206896551724e-05,
+      "loss": 0.96,
+      "step": 46
+    },
+    {
+      "epoch": 0.07375079684205364,
+      "grad_norm": 0.984701936656816,
+      "learning_rate": 1.6206896551724137e-05,
+      "loss": 0.9365,
+      "step": 47
+    },
+    {
+      "epoch": 0.0753199627323101,
+      "grad_norm": 0.8543319754765061,
+      "learning_rate": 1.6551724137931037e-05,
+      "loss": 0.9387,
+      "step": 48
+    },
+    {
+      "epoch": 0.07688912862256657,
+      "grad_norm": 0.7354318449970685,
+      "learning_rate": 1.6896551724137932e-05,
+      "loss": 0.9425,
+      "step": 49
+    },
+    {
+      "epoch": 0.07845829451282303,
+      "grad_norm": 0.8213529344491065,
+      "learning_rate": 1.7241379310344828e-05,
+      "loss": 0.9348,
+      "step": 50
+    },
+    {
+      "epoch": 0.08002746040307948,
+      "grad_norm": 1.1067533114051213,
+      "learning_rate": 1.7586206896551724e-05,
+      "loss": 0.9473,
+      "step": 51
+    },
+    {
+      "epoch": 0.08159662629333594,
+      "grad_norm": 0.8400328281550804,
+      "learning_rate": 1.7931034482758623e-05,
+      "loss": 0.9301,
+      "step": 52
+    },
+    {
+      "epoch": 0.08316579218359241,
+      "grad_norm": 0.8370778003878797,
+      "learning_rate": 1.827586206896552e-05,
+      "loss": 0.9317,
+      "step": 53
+    },
+    {
+      "epoch": 0.08473495807384887,
+      "grad_norm": 0.8908306556589923,
+      "learning_rate": 1.8620689655172415e-05,
+      "loss": 0.9133,
+      "step": 54
+    },
+    {
+      "epoch": 0.08630412396410533,
+      "grad_norm": 0.83407744299919,
+      "learning_rate": 1.896551724137931e-05,
+      "loss": 0.9302,
+      "step": 55
+    },
+    {
+      "epoch": 0.0878732898543618,
+      "grad_norm": 0.7635292113005164,
+      "learning_rate": 1.931034482758621e-05,
+      "loss": 0.9181,
+      "step": 56
+    },
+    {
+      "epoch": 0.08944245574461825,
+      "grad_norm": 0.7652021267207634,
+      "learning_rate": 1.9655172413793106e-05,
+      "loss": 0.8823,
+      "step": 57
+    },
+    {
+      "epoch": 0.09101162163487471,
+      "grad_norm": 0.8904272775089193,
+      "learning_rate": 2e-05,
+      "loss": 0.9123,
+      "step": 58
+    },
+    {
+      "epoch": 0.09258078752513117,
+      "grad_norm": 0.8381840657381086,
+      "learning_rate": 1.9999985627947104e-05,
+      "loss": 0.9399,
+      "step": 59
+    },
+    {
+      "epoch": 0.09414995341538764,
+      "grad_norm": 0.7381922221059936,
+      "learning_rate": 1.9999942511829714e-05,
+      "loss": 0.8994,
+      "step": 60
+    },
+    {
+      "epoch": 0.0957191193056441,
+      "grad_norm": 0.7611813511178191,
+      "learning_rate": 1.999987065177177e-05,
+      "loss": 0.9294,
+      "step": 61
+    },
+    {
+      "epoch": 0.09728828519590055,
+      "grad_norm": 0.7566970543114174,
+      "learning_rate": 1.9999770047979828e-05,
+      "loss": 0.9168,
+      "step": 62
+    },
+    {
+      "epoch": 0.09885745108615701,
+      "grad_norm": 0.7454082428140971,
+      "learning_rate": 1.9999640700743063e-05,
+      "loss": 0.9244,
+      "step": 63
+    },
+    {
+      "epoch": 0.10042661697641347,
+      "grad_norm": 0.723587008022164,
+      "learning_rate": 1.9999482610433275e-05,
+      "loss": 0.9097,
+      "step": 64
+    },
+    {
+      "epoch": 0.10199578286666994,
+      "grad_norm": 0.7751102363742651,
+      "learning_rate": 1.999929577750487e-05,
+      "loss": 0.9254,
+      "step": 65
+    },
+    {
+      "epoch": 0.1035649487569264,
+      "grad_norm": 0.7564755143947245,
+      "learning_rate": 1.99990802024949e-05,
+      "loss": 0.9375,
+      "step": 66
+    },
+    {
+      "epoch": 0.10513411464718286,
+      "grad_norm": 0.8377969755997268,
+      "learning_rate": 1.9998835886022997e-05,
+      "loss": 0.9067,
+      "step": 67
+    },
+    {
+      "epoch": 0.10670328053743931,
+      "grad_norm": 0.8026025545637548,
+      "learning_rate": 1.999856282879144e-05,
+      "loss": 0.8952,
+      "step": 68
+    },
+    {
+      "epoch": 0.10827244642769578,
+      "grad_norm": 0.7015852417445009,
+      "learning_rate": 1.9998261031585104e-05,
+      "loss": 0.9073,
+      "step": 69
+    },
+    {
+      "epoch": 0.10984161231795224,
+      "grad_norm": 0.8132394285529749,
+      "learning_rate": 1.999793049527148e-05,
+      "loss": 0.9144,
+      "step": 70
+    },
+    {
+      "epoch": 0.1114107782082087,
+      "grad_norm": 0.7960059801435135,
+      "learning_rate": 1.999757122080066e-05,
+      "loss": 0.8968,
+      "step": 71
+    },
+    {
+      "epoch": 0.11297994409846517,
+      "grad_norm": 0.783264987001189,
+      "learning_rate": 1.999718320920535e-05,
+      "loss": 0.8967,
+      "step": 72
+    },
+    {
+      "epoch": 0.11454910998872161,
+      "grad_norm": 0.6914436170942413,
+      "learning_rate": 1.9996766461600855e-05,
+      "loss": 0.9005,
+      "step": 73
+    },
+    {
+      "epoch": 0.11611827587897808,
+      "grad_norm": 0.7636496635969408,
+      "learning_rate": 1.9996320979185074e-05,
+      "loss": 0.9016,
+      "step": 74
+    },
+    {
+      "epoch": 0.11768744176923454,
+      "grad_norm": 0.7664674319956424,
+      "learning_rate": 1.9995846763238514e-05,
+      "loss": 0.9106,
+      "step": 75
+    },
+    {
+      "epoch": 0.119256607659491,
+      "grad_norm": 0.7641192104240221,
+      "learning_rate": 1.999534381512426e-05,
+      "loss": 0.9179,
+      "step": 76
+    },
+    {
+      "epoch": 0.12082577354974747,
+      "grad_norm": 0.6854691940498465,
+      "learning_rate": 1.9994812136287994e-05,
+      "loss": 0.8938,
+      "step": 77
+    },
+    {
+      "epoch": 0.12239493944000392,
+      "grad_norm": 0.8147579468081143,
+      "learning_rate": 1.9994251728257978e-05,
+      "loss": 0.906,
+      "step": 78
+    },
+    {
+      "epoch": 0.12396410533026038,
+      "grad_norm": 0.7927705378197664,
+      "learning_rate": 1.9993662592645057e-05,
+      "loss": 0.899,
+      "step": 79
+    },
+    {
+      "epoch": 0.12553327122051686,
+      "grad_norm": 0.6887191201827737,
+      "learning_rate": 1.999304473114265e-05,
+      "loss": 0.8593,
+      "step": 80
+    },
+    {
+      "epoch": 0.1271024371107733,
+      "grad_norm": 0.7761908068261424,
+      "learning_rate": 1.9992398145526738e-05,
+      "loss": 0.9124,
+      "step": 81
+    },
+    {
+      "epoch": 0.12867160300102976,
+      "grad_norm": 0.770274666552246,
+      "learning_rate": 1.999172283765588e-05,
+      "loss": 0.9079,
+      "step": 82
+    },
+    {
+      "epoch": 0.13024076889128622,
+      "grad_norm": 0.7241292852716699,
+      "learning_rate": 1.999101880947118e-05,
+      "loss": 0.9007,
+      "step": 83
+    },
+    {
+      "epoch": 0.13180993478154268,
+      "grad_norm": 0.7622797843473729,
+      "learning_rate": 1.9990286062996316e-05,
+      "loss": 0.877,
+      "step": 84
+    },
+    {
+      "epoch": 0.13337910067179914,
+      "grad_norm": 0.7786328812817573,
+      "learning_rate": 1.9989524600337493e-05,
+      "loss": 0.8843,
+      "step": 85
+    },
+    {
+      "epoch": 0.1349482665620556,
+      "grad_norm": 0.7229907044002635,
+      "learning_rate": 1.998873442368347e-05,
+      "loss": 0.8947,
+      "step": 86
+    },
+    {
+      "epoch": 0.13651743245231207,
+      "grad_norm": 0.6751340757105616,
+      "learning_rate": 1.998791553530554e-05,
+      "loss": 0.8847,
+      "step": 87
+    },
+    {
+      "epoch": 0.13808659834256853,
+      "grad_norm": 0.7662400723909407,
+      "learning_rate": 1.9987067937557527e-05,
+      "loss": 0.9022,
+      "step": 88
+    },
+    {
+      "epoch": 0.139655764232825,
+      "grad_norm": 0.7538616345769723,
+      "learning_rate": 1.9986191632875768e-05,
+      "loss": 0.8758,
+      "step": 89
+    },
+    {
+      "epoch": 0.14122493012308146,
+      "grad_norm": 0.6826897699341007,
+      "learning_rate": 1.9985286623779126e-05,
+      "loss": 0.89,
+      "step": 90
+    },
+    {
+      "epoch": 0.14279409601333792,
+      "grad_norm": 0.7905043333256622,
+      "learning_rate": 1.9984352912868973e-05,
+      "loss": 0.8574,
+      "step": 91
+    },
+    {
+      "epoch": 0.14436326190359436,
+      "grad_norm": 0.7367870689453241,
+      "learning_rate": 1.9983390502829168e-05,
+      "loss": 0.8601,
+      "step": 92
+    },
+    {
+      "epoch": 0.14593242779385082,
+      "grad_norm": 0.7362573090751993,
+      "learning_rate": 1.998239939642608e-05,
+      "loss": 0.8926,
+      "step": 93
+    },
+    {
+      "epoch": 0.14750159368410728,
+      "grad_norm": 0.7226330158275116,
+      "learning_rate": 1.998137959650855e-05,
+      "loss": 0.9153,
+      "step": 94
+    },
+    {
+      "epoch": 0.14907075957436375,
+      "grad_norm": 0.730058433948825,
+      "learning_rate": 1.9980331106007912e-05,
+      "loss": 0.8993,
+      "step": 95
+    },
+    {
+      "epoch": 0.1506399254646202,
+      "grad_norm": 0.753974038787166,
+      "learning_rate": 1.9979253927937948e-05,
+      "loss": 0.9024,
+      "step": 96
+    },
+    {
+      "epoch": 0.15220909135487667,
+      "grad_norm": 0.7227900960487232,
+      "learning_rate": 1.9978148065394917e-05,
+      "loss": 0.8931,
+      "step": 97
+    },
+    {
+      "epoch": 0.15377825724513314,
+      "grad_norm": 0.7433440664504725,
+      "learning_rate": 1.9977013521557516e-05,
+      "loss": 0.8645,
+      "step": 98
+    },
+    {
+      "epoch": 0.1553474231353896,
+      "grad_norm": 0.7077760276480354,
+      "learning_rate": 1.9975850299686894e-05,
+      "loss": 0.8559,
+      "step": 99
+    },
+    {
+      "epoch": 0.15691658902564606,
+      "grad_norm": 0.7576132903661229,
+      "learning_rate": 1.9974658403126624e-05,
+      "loss": 0.8759,
+      "step": 100
+    },
+    {
+      "epoch": 0.15848575491590253,
+      "grad_norm": 0.7240351543413873,
+      "learning_rate": 1.997343783530271e-05,
+      "loss": 0.8694,
+      "step": 101
+    },
+    {
+      "epoch": 0.16005492080615896,
+      "grad_norm": 0.7254639461474256,
+      "learning_rate": 1.9972188599723564e-05,
+      "loss": 0.8989,
+      "step": 102
+    },
+    {
+      "epoch": 0.16162408669641543,
+      "grad_norm": 0.7851603911250958,
+      "learning_rate": 1.997091069998e-05,
+      "loss": 0.8763,
+      "step": 103
+    },
+    {
+      "epoch": 0.1631932525866719,
+      "grad_norm": 0.7705234916818976,
+      "learning_rate": 1.9969604139745227e-05,
+      "loss": 0.8642,
+      "step": 104
+    },
+    {
+      "epoch": 0.16476241847692835,
+      "grad_norm": 0.7699032365333096,
+      "learning_rate": 1.9968268922774842e-05,
+      "loss": 0.8931,
+      "step": 105
+    },
+    {
+      "epoch": 0.16633158436718481,
+      "grad_norm": 0.8631933068348875,
+      "learning_rate": 1.99669050529068e-05,
+      "loss": 0.8982,
+      "step": 106
+    },
+    {
+      "epoch": 0.16790075025744128,
+      "grad_norm": 0.8233206296932418,
+      "learning_rate": 1.9965512534061426e-05,
+      "loss": 0.8745,
+      "step": 107
+    },
+    {
+      "epoch": 0.16946991614769774,
+      "grad_norm": 0.8348117365026448,
+      "learning_rate": 1.9964091370241384e-05,
+      "loss": 0.8942,
+      "step": 108
+    },
+    {
+      "epoch": 0.1710390820379542,
+      "grad_norm": 0.7198099494003858,
+      "learning_rate": 1.9962641565531694e-05,
+      "loss": 0.8899,
+      "step": 109
+    },
+    {
+      "epoch": 0.17260824792821067,
+      "grad_norm": 0.8670933610588272,
+      "learning_rate": 1.996116312409968e-05,
+      "loss": 0.8812,
+      "step": 110
+    },
+    {
+      "epoch": 0.17417741381846713,
+      "grad_norm": 0.8491765986431412,
+      "learning_rate": 1.9959656050194994e-05,
+      "loss": 0.8715,
+      "step": 111
+    },
+    {
+      "epoch": 0.1757465797087236,
+      "grad_norm": 0.7046196782092339,
+      "learning_rate": 1.9958120348149585e-05,
+      "loss": 0.8721,
+      "step": 112
+    },
+    {
+      "epoch": 0.17731574559898003,
+      "grad_norm": 0.7369560395354052,
+      "learning_rate": 1.9956556022377693e-05,
+      "loss": 0.872,
+      "step": 113
+    },
+    {
+      "epoch": 0.1788849114892365,
+      "grad_norm": 0.7802191792216598,
+      "learning_rate": 1.995496307737583e-05,
+      "loss": 0.8833,
+      "step": 114
+    },
+    {
+      "epoch": 0.18045407737949296,
+      "grad_norm": 0.816528253592501,
+      "learning_rate": 1.9953341517722773e-05,
+      "loss": 0.8688,
+      "step": 115
+    },
+    {
+      "epoch": 0.18202324326974942,
+      "grad_norm": 0.7298517686996712,
+      "learning_rate": 1.9951691348079553e-05,
+      "loss": 0.8691,
+      "step": 116
+    },
+    {
+      "epoch": 0.18359240916000588,
+      "grad_norm": 0.7604405768637559,
+      "learning_rate": 1.995001257318943e-05,
+      "loss": 0.8786,
+      "step": 117
+    },
+    {
+      "epoch": 0.18516157505026234,
+      "grad_norm": 0.7836159442393097,
+      "learning_rate": 1.99483051978779e-05,
+      "loss": 0.8649,
+      "step": 118
+    },
+    {
+      "epoch": 0.1867307409405188,
+      "grad_norm": 0.7299028152447427,
+      "learning_rate": 1.994656922705266e-05,
+      "loss": 0.856,
+      "step": 119
+    },
+    {
+      "epoch": 0.18829990683077527,
+      "grad_norm": 0.7548713186339182,
+      "learning_rate": 1.9944804665703593e-05,
+      "loss": 0.8823,
+      "step": 120
+    },
+    {
+      "epoch": 0.18986907272103173,
+      "grad_norm": 0.7573653085180222,
+      "learning_rate": 1.994301151890278e-05,
+      "loss": 0.8675,
+      "step": 121
+    },
+    {
+      "epoch": 0.1914382386112882,
+      "grad_norm": 0.8885442741020362,
+      "learning_rate": 1.9941189791804466e-05,
+      "loss": 0.8886,
+      "step": 122
+    },
+    {
+      "epoch": 0.19300740450154466,
+      "grad_norm": 0.7381405086867769,
+      "learning_rate": 1.9939339489645034e-05,
+      "loss": 0.8717,
+      "step": 123
+    },
+    {
+      "epoch": 0.1945765703918011,
+      "grad_norm": 0.7463307222711407,
+      "learning_rate": 1.9937460617743013e-05,
+      "loss": 0.8744,
+      "step": 124
+    },
+    {
+      "epoch": 0.19614573628205756,
+      "grad_norm": 0.8654609652602285,
+      "learning_rate": 1.993555318149906e-05,
+      "loss": 0.889,
+      "step": 125
+    },
+    {
+      "epoch": 0.19771490217231402,
+      "grad_norm": 0.7849912689324775,
+      "learning_rate": 1.9933617186395917e-05,
+      "loss": 0.8762,
+      "step": 126
+    },
+    {
+      "epoch": 0.19928406806257049,
+      "grad_norm": 0.8002241791369005,
+      "learning_rate": 1.9931652637998443e-05,
+      "loss": 0.8751,
+      "step": 127
+    },
+    {
+      "epoch": 0.20085323395282695,
+      "grad_norm": 0.7068996354932181,
+      "learning_rate": 1.9929659541953548e-05,
+      "loss": 0.8712,
+      "step": 128
+    },
+    {
+      "epoch": 0.2024223998430834,
+      "grad_norm": 0.7889080948860275,
+      "learning_rate": 1.992763790399021e-05,
+      "loss": 0.859,
+      "step": 129
+    },
+    {
+      "epoch": 0.20399156573333987,
+      "grad_norm": 0.7822335872655365,
+      "learning_rate": 1.9925587729919447e-05,
+      "loss": 0.8811,
+      "step": 130
+    },
+    {
+      "epoch": 0.20556073162359634,
+      "grad_norm": 0.8059774406273587,
+      "learning_rate": 1.99235090256343e-05,
+      "loss": 0.8858,
+      "step": 131
+    },
+    {
+      "epoch": 0.2071298975138528,
+      "grad_norm": 0.7540895868089319,
+      "learning_rate": 1.992140179710982e-05,
+      "loss": 0.8875,
+      "step": 132
+    },
+    {
+      "epoch": 0.20869906340410926,
+      "grad_norm": 0.7846220864667804,
+      "learning_rate": 1.991926605040305e-05,
+      "loss": 0.8458,
+      "step": 133
+    },
+    {
+      "epoch": 0.21026822929436573,
+      "grad_norm": 0.7047087658235964,
+      "learning_rate": 1.9917101791652997e-05,
+      "loss": 0.8738,
+      "step": 134
+    },
+    {
+      "epoch": 0.21183739518462216,
+      "grad_norm": 0.7196516438286433,
+      "learning_rate": 1.991490902708063e-05,
+      "loss": 0.8692,
+      "step": 135
+    },
+    {
+      "epoch": 0.21340656107487863,
+      "grad_norm": 0.7883176612843322,
+      "learning_rate": 1.9912687762988855e-05,
+      "loss": 0.879,
+      "step": 136
+    },
+    {
+      "epoch": 0.2149757269651351,
+      "grad_norm": 0.6751868528232794,
+      "learning_rate": 1.9910438005762503e-05,
+      "loss": 0.8713,
+      "step": 137
+    },
+    {
+      "epoch": 0.21654489285539155,
+      "grad_norm": 0.7266397934419208,
+      "learning_rate": 1.9908159761868292e-05,
+      "loss": 0.8769,
+      "step": 138
+    },
+    {
+      "epoch": 0.21811405874564802,
+      "grad_norm": 0.7471264193008941,
+      "learning_rate": 1.9905853037854834e-05,
+      "loss": 0.8802,
+      "step": 139
+    },
+    {
+      "epoch": 0.21968322463590448,
+      "grad_norm": 0.715724210142443,
+      "learning_rate": 1.9903517840352605e-05,
+      "loss": 0.8587,
+      "step": 140
+    },
+    {
+      "epoch": 0.22125239052616094,
+      "grad_norm": 0.7561328229150864,
+      "learning_rate": 1.9901154176073914e-05,
+      "loss": 0.8683,
+      "step": 141
+    },
+    {
+      "epoch": 0.2228215564164174,
+      "grad_norm": 0.686487994915129,
+      "learning_rate": 1.9898762051812906e-05,
+      "loss": 0.8487,
+      "step": 142
+    },
+    {
+      "epoch": 0.22439072230667387,
+      "grad_norm": 0.8089172767827801,
+      "learning_rate": 1.9896341474445526e-05,
+      "loss": 0.8892,
+      "step": 143
+    },
+    {
+      "epoch": 0.22595988819693033,
+      "grad_norm": 0.8740625586916276,
+      "learning_rate": 1.989389245092951e-05,
+      "loss": 0.8639,
+      "step": 144
+    },
+    {
+      "epoch": 0.22752905408718677,
+      "grad_norm": 0.7390304199448842,
+      "learning_rate": 1.9891414988304357e-05,
+      "loss": 0.8464,
+      "step": 145
+    },
+    {
+      "epoch": 0.22909821997744323,
+      "grad_norm": 0.8288692386757024,
+      "learning_rate": 1.988890909369131e-05,
+      "loss": 0.8557,
+      "step": 146
+    },
+    {
+      "epoch": 0.2306673858676997,
+      "grad_norm": 0.8406054645692689,
+      "learning_rate": 1.988637477429334e-05,
+      "loss": 0.867,
+      "step": 147
+    },
+    {
+      "epoch": 0.23223655175795616,
+      "grad_norm": 0.686442460999749,
+      "learning_rate": 1.9883812037395126e-05,
+      "loss": 0.8862,
+      "step": 148
+    },
+    {
+      "epoch": 0.23380571764821262,
+      "grad_norm": 0.8295574504292643,
+      "learning_rate": 1.988122089036302e-05,
+      "loss": 0.8442,
+      "step": 149
+    },
+    {
+      "epoch": 0.23537488353846908,
+      "grad_norm": 0.6853562810053988,
+      "learning_rate": 1.987860134064504e-05,
+      "loss": 0.8517,
+      "step": 150
+    },
+    {
+      "epoch": 0.23694404942872554,
+      "grad_norm": 0.7775758820403763,
+      "learning_rate": 1.987595339577085e-05,
+      "loss": 0.8663,
+      "step": 151
+    },
+    {
+      "epoch": 0.238513215318982,
+      "grad_norm": 0.7067263950544675,
+      "learning_rate": 1.9873277063351737e-05,
+      "loss": 0.8619,
+      "step": 152
+    },
+    {
+      "epoch": 0.24008238120923847,
+      "grad_norm": 0.8089442938328503,
+      "learning_rate": 1.9870572351080573e-05,
+      "loss": 0.8715,
+      "step": 153
+    },
+    {
+      "epoch": 0.24165154709949493,
+      "grad_norm": 0.8755005657177319,
+      "learning_rate": 1.9867839266731813e-05,
+      "loss": 0.8681,
+      "step": 154
+    },
+    {
+      "epoch": 0.2432207129897514,
+      "grad_norm": 0.7162575447869155,
+      "learning_rate": 1.9865077818161464e-05,
+      "loss": 0.8515,
+      "step": 155
+    },
+    {
+      "epoch": 0.24478987888000783,
+      "grad_norm": 0.7787457082502274,
+      "learning_rate": 1.9862288013307062e-05,
+      "loss": 0.8871,
+      "step": 156
+    },
+    {
+      "epoch": 0.2463590447702643,
+      "grad_norm": 0.8095480592704093,
+      "learning_rate": 1.9859469860187654e-05,
+      "loss": 0.8641,
+      "step": 157
+    },
+    {
+      "epoch": 0.24792821066052076,
+      "grad_norm": 0.8070091072897313,
+      "learning_rate": 1.9856623366903765e-05,
+      "loss": 0.8728,
+      "step": 158
+    },
+    {
+      "epoch": 0.24949737655077722,
+      "grad_norm": 0.8021783356149979,
+      "learning_rate": 1.9853748541637387e-05,
+      "loss": 0.8826,
+      "step": 159
+    },
+    {
+      "epoch": 0.2510665424410337,
+      "grad_norm": 0.886216903978757,
+      "learning_rate": 1.985084539265195e-05,
+      "loss": 0.8672,
+      "step": 160
+    },
+    {
+      "epoch": 0.2526357083312901,
+      "grad_norm": 0.7893526987955237,
+      "learning_rate": 1.9847913928292297e-05,
+      "loss": 0.8408,
+      "step": 161
+    },
+    {
+      "epoch": 0.2542048742215466,
+      "grad_norm": 0.7433259267981747,
+      "learning_rate": 1.9844954156984656e-05,
+      "loss": 0.8467,
+      "step": 162
+    },
+    {
+      "epoch": 0.25577404011180305,
+      "grad_norm": 0.7949982916591,
+      "learning_rate": 1.9841966087236626e-05,
+      "loss": 0.8534,
+      "step": 163
+    },
+    {
+      "epoch": 0.2573432060020595,
+      "grad_norm": 0.75524371392311,
+      "learning_rate": 1.983894972763715e-05,
+      "loss": 0.8985,
+      "step": 164
+    },
+    {
+      "epoch": 0.258912371892316,
+      "grad_norm": 0.7557283683559115,
+      "learning_rate": 1.983590508685648e-05,
+      "loss": 0.863,
+      "step": 165
+    },
+    {
+      "epoch": 0.26048153778257244,
+      "grad_norm": 0.7663216799290998,
+      "learning_rate": 1.9832832173646164e-05,
+      "loss": 0.8538,
+      "step": 166
+    },
+    {
+      "epoch": 0.2620507036728289,
+      "grad_norm": 0.7646094386491168,
+      "learning_rate": 1.982973099683902e-05,
+      "loss": 0.8451,
+      "step": 167
+    },
+    {
+      "epoch": 0.26361986956308536,
+      "grad_norm": 0.7370525737792721,
+      "learning_rate": 1.98266015653491e-05,
+      "loss": 0.8488,
+      "step": 168
+    },
+    {
+      "epoch": 0.2651890354533418,
+      "grad_norm": 0.7807270711651909,
+      "learning_rate": 1.9823443888171676e-05,
+      "loss": 0.8916,
+      "step": 169
+    },
+    {
+      "epoch": 0.2667582013435983,
+      "grad_norm": 0.7585536651749992,
+      "learning_rate": 1.9820257974383205e-05,
+      "loss": 0.8711,
+      "step": 170
+    },
+    {
+      "epoch": 0.26832736723385475,
+      "grad_norm": 0.7660249493583986,
+      "learning_rate": 1.9817043833141317e-05,
+      "loss": 0.8759,
+      "step": 171
+    },
+    {
+      "epoch": 0.2698965331241112,
+      "grad_norm": 0.9119784301784903,
+      "learning_rate": 1.981380147368477e-05,
+      "loss": 0.8904,
+      "step": 172
+    },
+    {
+      "epoch": 0.2714656990143677,
+      "grad_norm": 0.6806506174100767,
+      "learning_rate": 1.9810530905333437e-05,
+      "loss": 0.8657,
+      "step": 173
+    },
+    {
+      "epoch": 0.27303486490462414,
+      "grad_norm": 0.7868453586139633,
+      "learning_rate": 1.980723213748828e-05,
+      "loss": 0.8823,
+      "step": 174
+    },
+    {
+      "epoch": 0.2746040307948806,
+      "grad_norm": 0.8644204420772899,
+      "learning_rate": 1.9803905179631303e-05,
+      "loss": 0.8546,
+      "step": 175
+    },
+    {
+      "epoch": 0.27617319668513707,
+      "grad_norm": 0.7160963839532088,
+      "learning_rate": 1.9800550041325553e-05,
+      "loss": 0.8538,
+      "step": 176
+    },
+    {
+      "epoch": 0.27774236257539353,
+      "grad_norm": 0.7636955168568258,
+      "learning_rate": 1.9797166732215078e-05,
+      "loss": 0.8687,
+      "step": 177
+    },
+    {
+      "epoch": 0.27931152846565,
+      "grad_norm": 0.81733852619721,
+      "learning_rate": 1.979375526202489e-05,
+      "loss": 0.8541,
+      "step": 178
+    },
+    {
+      "epoch": 0.28088069435590646,
+      "grad_norm": 0.8057135487952796,
+      "learning_rate": 1.9790315640560962e-05,
+      "loss": 0.8346,
+      "step": 179
+    },
+    {
+      "epoch": 0.2824498602461629,
+      "grad_norm": 0.7793925721093384,
+      "learning_rate": 1.9786847877710177e-05,
+      "loss": 0.8558,
+      "step": 180
+    },
+    {
+      "epoch": 0.2840190261364194,
+      "grad_norm": 0.7542240585851123,
+      "learning_rate": 1.9783351983440306e-05,
+      "loss": 0.8648,
+      "step": 181
+    },
+    {
+      "epoch": 0.28558819202667585,
+      "grad_norm": 0.7354379173828268,
+      "learning_rate": 1.9779827967799985e-05,
+      "loss": 0.8538,
+      "step": 182
+    },
+    {
+      "epoch": 0.28715735791693225,
+      "grad_norm": 0.7651881586610363,
+      "learning_rate": 1.9776275840918685e-05,
+      "loss": 0.8342,
+      "step": 183
+    },
+    {
+      "epoch": 0.2887265238071887,
+      "grad_norm": 0.9027663878612077,
+      "learning_rate": 1.9772695613006676e-05,
+      "loss": 0.861,
+      "step": 184
+    },
+    {
+      "epoch": 0.2902956896974452,
+      "grad_norm": 0.773452599555622,
+      "learning_rate": 1.9769087294355006e-05,
+      "loss": 0.8578,
+      "step": 185
+    },
+    {
+      "epoch": 0.29186485558770164,
+      "grad_norm": 0.7674065376911422,
+      "learning_rate": 1.9765450895335454e-05,
+      "loss": 0.8403,
+      "step": 186
+    },
+    {
+      "epoch": 0.2934340214779581,
+      "grad_norm": 0.8319274210080099,
+      "learning_rate": 1.9761786426400532e-05,
+      "loss": 0.8558,
+      "step": 187
+    },
+    {
+      "epoch": 0.29500318736821457,
+      "grad_norm": 0.748304024100806,
+      "learning_rate": 1.975809389808343e-05,
+      "loss": 0.8521,
+      "step": 188
+    },
+    {
+      "epoch": 0.29657235325847103,
+      "grad_norm": 0.7667066312347395,
+      "learning_rate": 1.9754373320997987e-05,
+      "loss": 0.8744,
+      "step": 189
+    },
+    {
+      "epoch": 0.2981415191487275,
+      "grad_norm": 0.7863025391114771,
+      "learning_rate": 1.9750624705838666e-05,
+      "loss": 0.8575,
+      "step": 190
+    },
+    {
+      "epoch": 0.29971068503898396,
+      "grad_norm": 0.7808819255998675,
+      "learning_rate": 1.974684806338053e-05,
+      "loss": 0.8417,
+      "step": 191
+    },
+    {
+      "epoch": 0.3012798509292404,
+      "grad_norm": 0.677894377823736,
+      "learning_rate": 1.9743043404479203e-05,
+      "loss": 0.8646,
+      "step": 192
+    },
+    {
+      "epoch": 0.3028490168194969,
+      "grad_norm": 0.7694146043646329,
+      "learning_rate": 1.9739210740070833e-05,
+      "loss": 0.8662,
+      "step": 193
+    },
+    {
+      "epoch": 0.30441818270975335,
+      "grad_norm": 0.9255283581381419,
+      "learning_rate": 1.973535008117207e-05,
+      "loss": 0.843,
+      "step": 194
+    },
+    {
+      "epoch": 0.3059873486000098,
+      "grad_norm": 0.6817104055346374,
+      "learning_rate": 1.9731461438880034e-05,
+      "loss": 0.83,
+      "step": 195
+    },
+    {
+      "epoch": 0.3075565144902663,
+      "grad_norm": 0.8302031790838414,
+      "learning_rate": 1.972754482437228e-05,
+      "loss": 0.8469,
+      "step": 196
+    },
+    {
+      "epoch": 0.30912568038052274,
+      "grad_norm": 0.8489783993839521,
+      "learning_rate": 1.9723600248906766e-05,
+      "loss": 0.8591,
+      "step": 197
+    },
+    {
+      "epoch": 0.3106948462707792,
+      "grad_norm": 0.813303706179681,
+      "learning_rate": 1.9719627723821818e-05,
+      "loss": 0.8701,
+      "step": 198
+    },
+    {
+      "epoch": 0.31226401216103566,
+      "grad_norm": 0.8224094205940473,
+      "learning_rate": 1.971562726053611e-05,
+      "loss": 0.8514,
+      "step": 199
+    },
+    {
+      "epoch": 0.3138331780512921,
+      "grad_norm": 0.768413372749551,
+      "learning_rate": 1.971159887054862e-05,
+      "loss": 0.8391,
+      "step": 200
+    },
+    {
+      "epoch": 0.3154023439415486,
+      "grad_norm": 0.9123999473550879,
+      "learning_rate": 1.970754256543858e-05,
+      "loss": 0.8549,
+      "step": 201
+    },
+    {
+      "epoch": 0.31697150983180505,
+      "grad_norm": 0.7812643761892657,
+      "learning_rate": 1.9703458356865484e-05,
+      "loss": 0.8523,
+      "step": 202
+    },
+    {
+      "epoch": 0.3185406757220615,
+      "grad_norm": 0.7040191756905371,
+      "learning_rate": 1.9699346256569033e-05,
+      "loss": 0.8478,
+      "step": 203
+    },
+    {
+      "epoch": 0.3201098416123179,
+      "grad_norm": 0.7115085906671391,
+      "learning_rate": 1.9695206276369077e-05,
+      "loss": 0.8522,
+      "step": 204
+    },
+    {
+      "epoch": 0.3216790075025744,
+      "grad_norm": 0.8202197104252852,
+      "learning_rate": 1.969103842816563e-05,
+      "loss": 0.8461,
+      "step": 205
+    },
+    {
+      "epoch": 0.32324817339283085,
+      "grad_norm": 0.7631221688538243,
+      "learning_rate": 1.968684272393879e-05,
+      "loss": 0.8708,
+      "step": 206
+    },
+    {
+      "epoch": 0.3248173392830873,
+      "grad_norm": 0.7388642425826576,
+      "learning_rate": 1.968261917574874e-05,
+      "loss": 0.8636,
+      "step": 207
+    },
+    {
+      "epoch": 0.3263865051733438,
+      "grad_norm": 0.7633365283209427,
+      "learning_rate": 1.967836779573569e-05,
+      "loss": 0.8417,
+      "step": 208
+    },
+    {
+      "epoch": 0.32795567106360024,
+      "grad_norm": 0.7832600718655486,
+      "learning_rate": 1.9674088596119853e-05,
+      "loss": 0.8552,
+      "step": 209
+    },
+    {
+      "epoch": 0.3295248369538567,
+      "grad_norm": 0.8340710850147682,
+      "learning_rate": 1.96697815892014e-05,
+      "loss": 0.8319,
+      "step": 210
+    },
+    {
+      "epoch": 0.33109400284411317,
+      "grad_norm": 0.7636784436230482,
+      "learning_rate": 1.9665446787360444e-05,
+      "loss": 0.8688,
+      "step": 211
+    },
+    {
+      "epoch": 0.33266316873436963,
+      "grad_norm": 0.842898500420848,
+      "learning_rate": 1.9661084203056983e-05,
+      "loss": 0.8472,
+      "step": 212
+    },
+    {
+      "epoch": 0.3342323346246261,
+      "grad_norm": 0.7981515604884727,
+      "learning_rate": 1.965669384883087e-05,
+      "loss": 0.8474,
+      "step": 213
+    },
+    {
+      "epoch": 0.33580150051488256,
+      "grad_norm": 0.7248564359387966,
+      "learning_rate": 1.9652275737301797e-05,
+      "loss": 0.8688,
+      "step": 214
+    },
+    {
+      "epoch": 0.337370666405139,
+      "grad_norm": 0.9367555688235145,
+      "learning_rate": 1.9647829881169223e-05,
+      "loss": 0.8414,
+      "step": 215
+    },
+    {
+      "epoch": 0.3389398322953955,
+      "grad_norm": 0.7512923513745018,
+      "learning_rate": 1.9643356293212363e-05,
+      "loss": 0.8381,
+      "step": 216
+    },
+    {
+      "epoch": 0.34050899818565195,
+      "grad_norm": 0.8292116707814458,
+      "learning_rate": 1.963885498629015e-05,
+      "loss": 0.8672,
+      "step": 217
+    },
+    {
+      "epoch": 0.3420781640759084,
+      "grad_norm": 0.7571748129831849,
+      "learning_rate": 1.9634325973341183e-05,
+      "loss": 0.8687,
+      "step": 218
+    },
+    {
+      "epoch": 0.34364732996616487,
+      "grad_norm": 0.8033335331307723,
+      "learning_rate": 1.962976926738371e-05,
+      "loss": 0.8503,
+      "step": 219
+    },
+    {
+      "epoch": 0.34521649585642133,
+      "grad_norm": 0.7702958013328455,
+      "learning_rate": 1.962518488151557e-05,
+      "loss": 0.8572,
+      "step": 220
+    },
+    {
+      "epoch": 0.3467856617466778,
+      "grad_norm": 0.7773309211600538,
+      "learning_rate": 1.9620572828914177e-05,
+      "loss": 0.8564,
+      "step": 221
+    },
+    {
+      "epoch": 0.34835482763693426,
+      "grad_norm": 0.7172600760847799,
+      "learning_rate": 1.9615933122836454e-05,
+      "loss": 0.8574,
+      "step": 222
+    },
+    {
+      "epoch": 0.3499239935271907,
+      "grad_norm": 0.8070121659356566,
+      "learning_rate": 1.961126577661883e-05,
+      "loss": 0.8812,
+      "step": 223
+    },
+    {
+      "epoch": 0.3514931594174472,
+      "grad_norm": 0.7659770573898073,
+      "learning_rate": 1.960657080367717e-05,
+      "loss": 0.8211,
+      "step": 224
+    },
+    {
+      "epoch": 0.35306232530770365,
+      "grad_norm": 0.7005613157252402,
+      "learning_rate": 1.9601848217506755e-05,
+      "loss": 0.8629,
+      "step": 225
+    },
+    {
+      "epoch": 0.35463149119796006,
+      "grad_norm": 0.7772168734601813,
+      "learning_rate": 1.9597098031682238e-05,
+      "loss": 0.847,
+      "step": 226
+    },
+    {
+      "epoch": 0.3562006570882165,
+      "grad_norm": 0.7242043799321884,
+      "learning_rate": 1.9592320259857603e-05,
+      "loss": 0.8416,
+      "step": 227
+    },
+    {
+      "epoch": 0.357769822978473,
+      "grad_norm": 0.7007408152068441,
+      "learning_rate": 1.9587514915766124e-05,
+      "loss": 0.8443,
+      "step": 228
+    },
+    {
+      "epoch": 0.35933898886872945,
+      "grad_norm": 0.7071228511997497,
+      "learning_rate": 1.9582682013220337e-05,
+      "loss": 0.8788,
+      "step": 229
+    },
+    {
+      "epoch": 0.3609081547589859,
+      "grad_norm": 0.7185437861339777,
+      "learning_rate": 1.957782156611199e-05,
+      "loss": 0.8523,
+      "step": 230
+    },
+    {
+      "epoch": 0.3624773206492424,
+      "grad_norm": 0.7472842949489382,
+      "learning_rate": 1.9572933588411997e-05,
+      "loss": 0.848,
+      "step": 231
+    },
+    {
+      "epoch": 0.36404648653949884,
+      "grad_norm": 0.775608056134747,
+      "learning_rate": 1.9568018094170415e-05,
+      "loss": 0.8396,
+      "step": 232
+    },
+    {
+      "epoch": 0.3656156524297553,
+      "grad_norm": 0.7521110831245076,
+      "learning_rate": 1.9563075097516397e-05,
+      "loss": 0.8264,
+      "step": 233
+    },
+    {
+      "epoch": 0.36718481832001176,
+      "grad_norm": 0.7901869260162981,
+      "learning_rate": 1.9558104612658137e-05,
+      "loss": 0.871,
+      "step": 234
+    },
+    {
+      "epoch": 0.3687539842102682,
+      "grad_norm": 0.8020075798033522,
+      "learning_rate": 1.955310665388286e-05,
+      "loss": 0.8172,
+      "step": 235
+    },
+    {
+      "epoch": 0.3703231501005247,
+      "grad_norm": 0.7891001984416371,
+      "learning_rate": 1.954808123555674e-05,
+      "loss": 0.8221,
+      "step": 236
+    },
+    {
+      "epoch": 0.37189231599078115,
+      "grad_norm": 0.7864844060886398,
+      "learning_rate": 1.9543028372124903e-05,
+      "loss": 0.8435,
+      "step": 237
+    },
+    {
+      "epoch": 0.3734614818810376,
+      "grad_norm": 0.8322594692248915,
+      "learning_rate": 1.9537948078111348e-05,
+      "loss": 0.8591,
+      "step": 238
+    },
+    {
+      "epoch": 0.3750306477712941,
+      "grad_norm": 0.7861781276737043,
+      "learning_rate": 1.953284036811893e-05,
+      "loss": 0.8424,
+      "step": 239
+    },
+    {
+      "epoch": 0.37659981366155054,
+      "grad_norm": 0.7587824109421507,
+      "learning_rate": 1.9527705256829293e-05,
+      "loss": 0.8455,
+      "step": 240
+    },
+    {
+      "epoch": 0.378168979551807,
+      "grad_norm": 0.8054301495528072,
+      "learning_rate": 1.9522542759002866e-05,
+      "loss": 0.837,
+      "step": 241
+    },
+    {
+      "epoch": 0.37973814544206347,
+      "grad_norm": 0.7435044879925587,
+      "learning_rate": 1.9517352889478787e-05,
+      "loss": 0.845,
+      "step": 242
+    },
+    {
+      "epoch": 0.38130731133231993,
+      "grad_norm": 0.7020482850404819,
+      "learning_rate": 1.951213566317487e-05,
+      "loss": 0.8352,
+      "step": 243
+    },
+    {
+      "epoch": 0.3828764772225764,
+      "grad_norm": 0.81693994407979,
+      "learning_rate": 1.9506891095087563e-05,
+      "loss": 0.8346,
+      "step": 244
+    },
+    {
+      "epoch": 0.38444564311283286,
+      "grad_norm": 0.7303764658558978,
+      "learning_rate": 1.950161920029191e-05,
+      "loss": 0.8578,
+      "step": 245
+    },
+    {
+      "epoch": 0.3860148090030893,
+      "grad_norm": 0.7351241094383767,
+      "learning_rate": 1.9496319993941498e-05,
+      "loss": 0.849,
+      "step": 246
+    },
+    {
+      "epoch": 0.38758397489334573,
+      "grad_norm": 0.686380690779999,
+      "learning_rate": 1.949099349126843e-05,
+      "loss": 0.867,
+      "step": 247
+    },
+    {
+      "epoch": 0.3891531407836022,
+      "grad_norm": 0.7471915013075717,
+      "learning_rate": 1.9485639707583255e-05,
+      "loss": 0.8319,
+      "step": 248
+    },
+    {
+      "epoch": 0.39072230667385865,
+      "grad_norm": 0.6882834803145461,
+      "learning_rate": 1.9480258658274947e-05,
+      "loss": 0.8277,
+      "step": 249
+    },
+    {
+      "epoch": 0.3922914725641151,
+      "grad_norm": 0.7468679664828046,
+      "learning_rate": 1.947485035881085e-05,
+      "loss": 0.8456,
+      "step": 250
+    },
+    {
+      "epoch": 0.3938606384543716,
+      "grad_norm": 0.6936300092711225,
+      "learning_rate": 1.946941482473664e-05,
+      "loss": 0.8439,
+      "step": 251
+    },
+    {
+      "epoch": 0.39542980434462804,
+      "grad_norm": 0.7452972132793442,
+      "learning_rate": 1.9463952071676268e-05,
+      "loss": 0.8529,
+      "step": 252
+    },
+    {
+      "epoch": 0.3969989702348845,
+      "grad_norm": 0.7041663843901055,
+      "learning_rate": 1.9458462115331938e-05,
+      "loss": 0.8463,
+      "step": 253
+    },
+    {
+      "epoch": 0.39856813612514097,
+      "grad_norm": 0.6960468670599904,
+      "learning_rate": 1.945294497148403e-05,
+      "loss": 0.8655,
+      "step": 254
+    },
+    {
+      "epoch": 0.40013730201539743,
+      "grad_norm": 0.7269589347441181,
+      "learning_rate": 1.9447400655991086e-05,
+      "loss": 0.8402,
+      "step": 255
+    },
+    {
+      "epoch": 0.4017064679056539,
+      "grad_norm": 0.6911247809680555,
+      "learning_rate": 1.9441829184789745e-05,
+      "loss": 0.8378,
+      "step": 256
+    },
+    {
+      "epoch": 0.40327563379591036,
+      "grad_norm": 0.7300455734199899,
+      "learning_rate": 1.94362305738947e-05,
+      "loss": 0.8283,
+      "step": 257
+    },
+    {
+      "epoch": 0.4048447996861668,
+      "grad_norm": 0.6804187941038841,
+      "learning_rate": 1.9430604839398657e-05,
+      "loss": 0.8502,
+      "step": 258
+    },
+    {
+      "epoch": 0.4064139655764233,
+      "grad_norm": 0.6745629125019782,
+      "learning_rate": 1.942495199747229e-05,
+      "loss": 0.8309,
+      "step": 259
+    },
+    {
+      "epoch": 0.40798313146667975,
+      "grad_norm": 0.666356591402976,
+      "learning_rate": 1.9419272064364186e-05,
+      "loss": 0.8482,
+      "step": 260
+    },
+    {
+      "epoch": 0.4095522973569362,
+      "grad_norm": 0.7201969961891354,
+      "learning_rate": 1.9413565056400804e-05,
+      "loss": 0.8516,
+      "step": 261
+    },
+    {
+      "epoch": 0.4111214632471927,
+      "grad_norm": 0.7064391607371447,
+      "learning_rate": 1.940783098998643e-05,
+      "loss": 0.858,
+      "step": 262
+    },
+    {
+      "epoch": 0.41269062913744914,
+      "grad_norm": 0.6844556992776223,
+      "learning_rate": 1.9402069881603123e-05,
+      "loss": 0.8503,
+      "step": 263
+    },
+    {
+      "epoch": 0.4142597950277056,
+      "grad_norm": 0.686600947246657,
+      "learning_rate": 1.939628174781068e-05,
+      "loss": 0.8441,
+      "step": 264
+    },
+    {
+      "epoch": 0.41582896091796206,
+      "grad_norm": 0.6792094536874498,
+      "learning_rate": 1.939046660524656e-05,
+      "loss": 0.8286,
+      "step": 265
+    },
+    {
+      "epoch": 0.41739812680821853,
+      "grad_norm": 0.734167478877529,
+      "learning_rate": 1.9384624470625884e-05,
+      "loss": 0.8434,
+      "step": 266
+    },
+    {
+      "epoch": 0.418967292698475,
+      "grad_norm": 0.654205695446543,
+      "learning_rate": 1.9378755360741342e-05,
+      "loss": 0.8177,
+      "step": 267
+    },
+    {
+      "epoch": 0.42053645858873145,
+      "grad_norm": 0.7225090616344176,
+      "learning_rate": 1.9372859292463163e-05,
+      "loss": 0.8503,
+      "step": 268
+    },
+    {
+      "epoch": 0.42210562447898786,
+      "grad_norm": 0.67515867367939,
+      "learning_rate": 1.9366936282739067e-05,
+      "loss": 0.8393,
+      "step": 269
+    },
+    {
+      "epoch": 0.4236747903692443,
+      "grad_norm": 0.7261627103914131,
+      "learning_rate": 1.9360986348594226e-05,
+      "loss": 0.8639,
+      "step": 270
+    },
+    {
+      "epoch": 0.4252439562595008,
+      "grad_norm": 0.695718762198511,
+      "learning_rate": 1.935500950713118e-05,
+      "loss": 0.8422,
+      "step": 271
+    },
+    {
+      "epoch": 0.42681312214975725,
+      "grad_norm": 0.7195710303151387,
+      "learning_rate": 1.934900577552983e-05,
+      "loss": 0.8397,
+      "step": 272
+    },
+    {
+      "epoch": 0.4283822880400137,
+      "grad_norm": 0.7035679541108324,
+      "learning_rate": 1.934297517104737e-05,
+      "loss": 0.8082,
+      "step": 273
+    },
+    {
+      "epoch": 0.4299514539302702,
+      "grad_norm": 0.7268112865575431,
+      "learning_rate": 1.9336917711018227e-05,
+      "loss": 0.8351,
+      "step": 274
+    },
+    {
+      "epoch": 0.43152061982052664,
+      "grad_norm": 0.6561930178599715,
+      "learning_rate": 1.9330833412854035e-05,
+      "loss": 0.8317,
+      "step": 275
+    },
+    {
+      "epoch": 0.4330897857107831,
+      "grad_norm": 0.845233409356619,
+      "learning_rate": 1.932472229404356e-05,
+      "loss": 0.8459,
+      "step": 276
+    },
+    {
+      "epoch": 0.43465895160103957,
+      "grad_norm": 0.7468477924420288,
+      "learning_rate": 1.9318584372152668e-05,
+      "loss": 0.845,
+      "step": 277
+    },
+    {
+      "epoch": 0.43622811749129603,
+      "grad_norm": 0.6898115886093594,
+      "learning_rate": 1.9312419664824268e-05,
+      "loss": 0.8457,
+      "step": 278
+    },
+    {
+      "epoch": 0.4377972833815525,
+      "grad_norm": 0.74066723201382,
+      "learning_rate": 1.9306228189778255e-05,
+      "loss": 0.8355,
+      "step": 279
+    },
+    {
+      "epoch": 0.43936644927180896,
+      "grad_norm": 0.769753190022015,
+      "learning_rate": 1.9300009964811475e-05,
+      "loss": 0.8635,
+      "step": 280
+    },
+    {
+      "epoch": 0.4409356151620654,
+      "grad_norm": 0.7530379937412007,
+      "learning_rate": 1.9293765007797662e-05,
+      "loss": 0.8336,
+      "step": 281
+    },
+    {
+      "epoch": 0.4425047810523219,
+      "grad_norm": 0.7286877598832755,
+      "learning_rate": 1.928749333668738e-05,
+      "loss": 0.8343,
+      "step": 282
+    },
+    {
+      "epoch": 0.44407394694257835,
+      "grad_norm": 0.7288783062280276,
+      "learning_rate": 1.928119496950799e-05,
+      "loss": 0.8067,
+      "step": 283
+    },
+    {
+      "epoch": 0.4456431128328348,
+      "grad_norm": 0.802597143867939,
+      "learning_rate": 1.9274869924363585e-05,
+      "loss": 0.8544,
+      "step": 284
+    },
+    {
+      "epoch": 0.44721227872309127,
+      "grad_norm": 0.7162852675919648,
+      "learning_rate": 1.9268518219434943e-05,
+      "loss": 0.8453,
+      "step": 285
+    },
+    {
+      "epoch": 0.44878144461334774,
+      "grad_norm": 0.867720949977952,
+      "learning_rate": 1.9262139872979474e-05,
+      "loss": 0.8433,
+      "step": 286
+    },
+    {
+      "epoch": 0.4503506105036042,
+      "grad_norm": 0.797679948164506,
+      "learning_rate": 1.9255734903331157e-05,
+      "loss": 0.8304,
+      "step": 287
+    },
+    {
+      "epoch": 0.45191977639386066,
+      "grad_norm": 0.713966312590489,
+      "learning_rate": 1.9249303328900505e-05,
+      "loss": 0.8335,
+      "step": 288
+    },
+    {
+      "epoch": 0.4534889422841171,
+      "grad_norm": 0.8107308721937179,
+      "learning_rate": 1.9242845168174514e-05,
+      "loss": 0.8505,
+      "step": 289
+    },
+    {
+      "epoch": 0.45505810817437353,
+      "grad_norm": 0.7241037125351435,
+      "learning_rate": 1.923636043971658e-05,
+      "loss": 0.843,
+      "step": 290
+    },
+    {
+      "epoch": 0.45662727406463,
+      "grad_norm": 0.7808231867201284,
+      "learning_rate": 1.9229849162166478e-05,
+      "loss": 0.8504,
+      "step": 291
+    },
+    {
+      "epoch": 0.45819643995488646,
+      "grad_norm": 0.7327029734065018,
+      "learning_rate": 1.922331135424029e-05,
+      "loss": 0.8519,
+      "step": 292
+    },
+    {
+      "epoch": 0.4597656058451429,
+      "grad_norm": 0.8722600730108714,
+      "learning_rate": 1.921674703473037e-05,
+      "loss": 0.8213,
+      "step": 293
+    },
+    {
+      "epoch": 0.4613347717353994,
+      "grad_norm": 0.830604449039856,
+      "learning_rate": 1.921015622250526e-05,
+      "loss": 0.8139,
+      "step": 294
+    },
+    {
+      "epoch": 0.46290393762565585,
+      "grad_norm": 0.7186326357087109,
+      "learning_rate": 1.920353893650966e-05,
+      "loss": 0.833,
+      "step": 295
+    },
+    {
+      "epoch": 0.4644731035159123,
+      "grad_norm": 0.8148732054144233,
+      "learning_rate": 1.9196895195764363e-05,
+      "loss": 0.8263,
+      "step": 296
+    },
+    {
+      "epoch": 0.4660422694061688,
+      "grad_norm": 0.7113937997307276,
+      "learning_rate": 1.919022501936622e-05,
+      "loss": 0.8353,
+      "step": 297
+    },
+    {
+      "epoch": 0.46761143529642524,
+      "grad_norm": 0.6805837993776361,
+      "learning_rate": 1.918352842648804e-05,
+      "loss": 0.8519,
+      "step": 298
+    },
+    {
+      "epoch": 0.4691806011866817,
+      "grad_norm": 0.7755578389831994,
+      "learning_rate": 1.9176805436378597e-05,
+      "loss": 0.8234,
+      "step": 299
+    },
+    {
+      "epoch": 0.47074976707693816,
+      "grad_norm": 0.7456227489277198,
+      "learning_rate": 1.9170056068362516e-05,
+      "loss": 0.8304,
+      "step": 300
+    },
+    {
+      "epoch": 0.4723189329671946,
+      "grad_norm": 0.6758692862571203,
+      "learning_rate": 1.9163280341840253e-05,
+      "loss": 0.8354,
+      "step": 301
+    },
+    {
+      "epoch": 0.4738880988574511,
+      "grad_norm": 0.7940370146646629,
+      "learning_rate": 1.9156478276288027e-05,
+      "loss": 0.8422,
+      "step": 302
+    },
+    {
+      "epoch": 0.47545726474770755,
+      "grad_norm": 0.6963204258292274,
+      "learning_rate": 1.9149649891257772e-05,
+      "loss": 0.8248,
+      "step": 303
+    },
+    {
+      "epoch": 0.477026430637964,
+      "grad_norm": 0.7186445641137688,
+      "learning_rate": 1.9142795206377066e-05,
+      "loss": 0.8343,
+      "step": 304
+    },
+    {
+      "epoch": 0.4785955965282205,
+      "grad_norm": 0.8126238965138781,
+      "learning_rate": 1.9135914241349088e-05,
+      "loss": 0.8435,
+      "step": 305
+    },
+    {
+      "epoch": 0.48016476241847694,
+      "grad_norm": 0.6629313852535808,
+      "learning_rate": 1.9129007015952556e-05,
+      "loss": 0.8392,
+      "step": 306
+    },
+    {
+      "epoch": 0.4817339283087334,
+      "grad_norm": 0.784882838437951,
+      "learning_rate": 1.9122073550041675e-05,
+      "loss": 0.833,
+      "step": 307
+    },
+    {
+      "epoch": 0.48330309419898987,
+      "grad_norm": 0.7791688559441003,
+      "learning_rate": 1.911511386354607e-05,
+      "loss": 0.8295,
+      "step": 308
+    },
+    {
+      "epoch": 0.48487226008924633,
+      "grad_norm": 0.7227436466913426,
+      "learning_rate": 1.9108127976470737e-05,
+      "loss": 0.8342,
+      "step": 309
+    },
+    {
+      "epoch": 0.4864414259795028,
+      "grad_norm": 0.8750790818261531,
+      "learning_rate": 1.9101115908895985e-05,
+      "loss": 0.8268,
+      "step": 310
+    },
+    {
+      "epoch": 0.48801059186975926,
+      "grad_norm": 0.7111097650814941,
+      "learning_rate": 1.9094077680977377e-05,
+      "loss": 0.8704,
+      "step": 311
+    },
+    {
+      "epoch": 0.48957975776001567,
+      "grad_norm": 0.8435025682165305,
+      "learning_rate": 1.9087013312945668e-05,
+      "loss": 0.8143,
+      "step": 312
+    },
+    {
+      "epoch": 0.49114892365027213,
+      "grad_norm": 0.8688751647388246,
+      "learning_rate": 1.907992282510675e-05,
+      "loss": 0.8566,
+      "step": 313
+    },
+    {
+      "epoch": 0.4927180895405286,
+      "grad_norm": 0.801427654894205,
+      "learning_rate": 1.90728062378416e-05,
+      "loss": 0.8354,
+      "step": 314
+    },
+    {
+      "epoch": 0.49428725543078506,
+      "grad_norm": 0.8195367234878244,
+      "learning_rate": 1.906566357160621e-05,
+      "loss": 0.8159,
+      "step": 315
+    },
+    {
+      "epoch": 0.4958564213210415,
+      "grad_norm": 0.883355320234134,
+      "learning_rate": 1.9058494846931538e-05,
+      "loss": 0.8253,
+      "step": 316
+    },
+    {
+      "epoch": 0.497425587211298,
+      "grad_norm": 0.8406215564534029,
+      "learning_rate": 1.9051300084423433e-05,
+      "loss": 0.8071,
+      "step": 317
+    },
+    {
+      "epoch": 0.49899475310155444,
+      "grad_norm": 0.8578724375502428,
+      "learning_rate": 1.904407930476261e-05,
+      "loss": 0.8505,
+      "step": 318
+    },
+    {
+      "epoch": 0.500563918991811,
+      "grad_norm": 0.9793981210398741,
+      "learning_rate": 1.9036832528704547e-05,
+      "loss": 0.8615,
+      "step": 319
+    },
+    {
+      "epoch": 0.5021330848820674,
+      "grad_norm": 0.7779210504068751,
+      "learning_rate": 1.902955977707945e-05,
+      "loss": 0.8281,
+      "step": 320
+    },
+    {
+      "epoch": 0.5037022507723239,
+      "grad_norm": 0.7872790047793906,
+      "learning_rate": 1.9022261070792197e-05,
+      "loss": 0.8754,
+      "step": 321
+    },
+    {
+      "epoch": 0.5052714166625802,
+      "grad_norm": 0.8536393043955611,
+      "learning_rate": 1.901493643082227e-05,
+      "loss": 0.8055,
+      "step": 322
+    },
+    {
+      "epoch": 0.5068405825528367,
+      "grad_norm": 0.8295119413410734,
+      "learning_rate": 1.900758587822369e-05,
+      "loss": 0.8266,
+      "step": 323
+    },
+    {
+      "epoch": 0.5084097484430932,
+      "grad_norm": 0.7784805521191259,
+      "learning_rate": 1.900020943412496e-05,
+      "loss": 0.846,
+      "step": 324
+    },
+    {
+      "epoch": 0.5099789143333496,
+      "grad_norm": 0.8722231513721744,
+      "learning_rate": 1.8992807119729013e-05,
+      "loss": 0.8034,
+      "step": 325
+    },
+    {
+      "epoch": 0.5115480802236061,
+      "grad_norm": 0.7190110457603671,
+      "learning_rate": 1.8985378956313133e-05,
+      "loss": 0.8111,
+      "step": 326
+    },
+    {
+      "epoch": 0.5131172461138626,
+      "grad_norm": 0.8173731238597682,
+      "learning_rate": 1.8977924965228923e-05,
+      "loss": 0.8142,
+      "step": 327
+    },
+    {
+      "epoch": 0.514686412004119,
+      "grad_norm": 0.8855477796622119,
+      "learning_rate": 1.89704451679022e-05,
+      "loss": 0.8201,
+      "step": 328
+    },
+    {
+      "epoch": 0.5162555778943755,
+      "grad_norm": 0.698335340514232,
+      "learning_rate": 1.8962939585832985e-05,
+      "loss": 0.8295,
+      "step": 329
+    },
+    {
+      "epoch": 0.517824743784632,
+      "grad_norm": 0.753903170150091,
+      "learning_rate": 1.8955408240595396e-05,
+      "loss": 0.8338,
+      "step": 330
+    },
+    {
+      "epoch": 0.5193939096748884,
+      "grad_norm": 0.868205977372359,
+      "learning_rate": 1.894785115383761e-05,
+      "loss": 0.8303,
+      "step": 331
+    },
+    {
+      "epoch": 0.5209630755651449,
+      "grad_norm": 0.7253986099083615,
+      "learning_rate": 1.8940268347281802e-05,
+      "loss": 0.8258,
+      "step": 332
+    },
+    {
+      "epoch": 0.5225322414554013,
+      "grad_norm": 0.6871576278756758,
+      "learning_rate": 1.8932659842724067e-05,
+      "loss": 0.8416,
+      "step": 333
+    },
+    {
+      "epoch": 0.5241014073456578,
+      "grad_norm": 0.8083398963531386,
+      "learning_rate": 1.8925025662034376e-05,
+      "loss": 0.8347,
+      "step": 334
+    },
+    {
+      "epoch": 0.5256705732359143,
+      "grad_norm": 0.7667380824139266,
+      "learning_rate": 1.8917365827156493e-05,
+      "loss": 0.8313,
+      "step": 335
+    },
+    {
+      "epoch": 0.5272397391261707,
+      "grad_norm": 0.7210432342675546,
+      "learning_rate": 1.890968036010793e-05,
+      "loss": 0.8399,
+      "step": 336
+    },
+    {
+      "epoch": 0.5288089050164272,
+      "grad_norm": 0.7462258544818293,
+      "learning_rate": 1.8901969282979882e-05,
+      "loss": 0.8226,
+      "step": 337
+    },
+    {
+      "epoch": 0.5303780709066837,
+      "grad_norm": 0.7527879259370438,
+      "learning_rate": 1.889423261793714e-05,
+      "loss": 0.8243,
+      "step": 338
+    },
+    {
+      "epoch": 0.5319472367969401,
+      "grad_norm": 0.7166200704322698,
+      "learning_rate": 1.888647038721806e-05,
+      "loss": 0.8257,
+      "step": 339
+    },
+    {
+      "epoch": 0.5335164026871966,
+      "grad_norm": 0.8800857116890911,
+      "learning_rate": 1.8878682613134486e-05,
+      "loss": 0.8309,
+      "step": 340
+    },
+    {
+      "epoch": 0.535085568577453,
+      "grad_norm": 0.7056883535212769,
+      "learning_rate": 1.8870869318071667e-05,
+      "loss": 0.8403,
+      "step": 341
+    },
+    {
+      "epoch": 0.5366547344677095,
+      "grad_norm": 0.724266957402652,
+      "learning_rate": 1.886303052448823e-05,
+      "loss": 0.8411,
+      "step": 342
+    },
+    {
+      "epoch": 0.538223900357966,
+      "grad_norm": 0.7478114130480105,
+      "learning_rate": 1.8855166254916082e-05,
+      "loss": 0.8532,
+      "step": 343
+    },
+    {
+      "epoch": 0.5397930662482224,
+      "grad_norm": 0.7236268168310761,
+      "learning_rate": 1.8847276531960363e-05,
+      "loss": 0.8373,
+      "step": 344
+    },
+    {
+      "epoch": 0.5413622321384789,
+      "grad_norm": 0.7708356058648402,
+      "learning_rate": 1.8839361378299376e-05,
+      "loss": 0.8433,
+      "step": 345
+    },
+    {
+      "epoch": 0.5429313980287354,
+      "grad_norm": 0.7122535332724641,
+      "learning_rate": 1.8831420816684528e-05,
+      "loss": 0.8494,
+      "step": 346
+    },
+    {
+      "epoch": 0.5445005639189918,
+      "grad_norm": 0.6818118735087734,
+      "learning_rate": 1.8823454869940243e-05,
+      "loss": 0.8096,
+      "step": 347
+    },
+    {
+      "epoch": 0.5460697298092483,
+      "grad_norm": 0.6892639418959253,
+      "learning_rate": 1.8815463560963932e-05,
+      "loss": 0.8313,
+      "step": 348
+    },
+    {
+      "epoch": 0.5476388956995047,
+      "grad_norm": 0.6231698332420209,
+      "learning_rate": 1.8807446912725893e-05,
+      "loss": 0.816,
+      "step": 349
+    },
+    {
+      "epoch": 0.5492080615897612,
+      "grad_norm": 0.69177416881062,
+      "learning_rate": 1.8799404948269267e-05,
+      "loss": 0.8386,
+      "step": 350
+    },
+    {
+      "epoch": 0.5507772274800177,
+      "grad_norm": 0.676144378396341,
+      "learning_rate": 1.879133769070996e-05,
+      "loss": 0.8294,
+      "step": 351
+    },
+    {
+      "epoch": 0.5523463933702741,
+      "grad_norm": 0.6822340175503191,
+      "learning_rate": 1.878324516323658e-05,
+      "loss": 0.809,
+      "step": 352
+    },
+    {
+      "epoch": 0.5539155592605306,
+      "grad_norm": 0.7360218095662615,
+      "learning_rate": 1.877512738911038e-05,
+      "loss": 0.8096,
+      "step": 353
+    },
+    {
+      "epoch": 0.5554847251507871,
+      "grad_norm": 0.6271317377013338,
+      "learning_rate": 1.8766984391665173e-05,
+      "loss": 0.8235,
+      "step": 354
+    },
+    {
+      "epoch": 0.5570538910410435,
+      "grad_norm": 0.6640421778175191,
+      "learning_rate": 1.8758816194307272e-05,
+      "loss": 0.829,
+      "step": 355
+    },
+    {
+      "epoch": 0.5586230569313,
+      "grad_norm": 0.7029292433271469,
+      "learning_rate": 1.875062282051544e-05,
+      "loss": 0.844,
+      "step": 356
+    },
+    {
+      "epoch": 0.5601922228215565,
+      "grad_norm": 0.6373488534328882,
+      "learning_rate": 1.8742404293840784e-05,
+      "loss": 0.8339,
+      "step": 357
+    },
+    {
+      "epoch": 0.5617613887118129,
+      "grad_norm": 0.7044723707421129,
+      "learning_rate": 1.8734160637906738e-05,
+      "loss": 0.8331,
+      "step": 358
+    },
+    {
+      "epoch": 0.5633305546020694,
+      "grad_norm": 0.7584335727897751,
+      "learning_rate": 1.8725891876408946e-05,
+      "loss": 0.8213,
+      "step": 359
+    },
+    {
+      "epoch": 0.5648997204923258,
+      "grad_norm": 0.6599185658756294,
+      "learning_rate": 1.8717598033115224e-05,
+      "loss": 0.8569,
+      "step": 360
+    },
+    {
+      "epoch": 0.5664688863825823,
+      "grad_norm": 0.704365458509732,
+      "learning_rate": 1.8709279131865486e-05,
+      "loss": 0.8265,
+      "step": 361
+    },
+    {
+      "epoch": 0.5680380522728388,
+      "grad_norm": 0.7824297627538636,
+      "learning_rate": 1.8700935196571666e-05,
+      "loss": 0.8294,
+      "step": 362
+    },
+    {
+      "epoch": 0.5696072181630952,
+      "grad_norm": 0.6910388517838051,
+      "learning_rate": 1.8692566251217665e-05,
+      "loss": 0.8159,
+      "step": 363
+    },
+    {
+      "epoch": 0.5711763840533517,
+      "grad_norm": 0.9141759575207105,
+      "learning_rate": 1.8684172319859258e-05,
+      "loss": 0.8543,
+      "step": 364
+    },
+    {
+      "epoch": 0.572745549943608,
+      "grad_norm": 0.6923514907759056,
+      "learning_rate": 1.867575342662406e-05,
+      "loss": 0.8308,
+      "step": 365
+    },
+    {
+      "epoch": 0.5743147158338645,
+      "grad_norm": 0.7871755486736863,
+      "learning_rate": 1.8667309595711426e-05,
+      "loss": 0.8195,
+      "step": 366
+    },
+    {
+      "epoch": 0.575883881724121,
+      "grad_norm": 0.736697098608543,
+      "learning_rate": 1.8658840851392386e-05,
+      "loss": 0.8137,
+      "step": 367
+    },
+    {
+      "epoch": 0.5774530476143774,
+      "grad_norm": 0.7015610378221094,
+      "learning_rate": 1.8650347218009596e-05,
+      "loss": 0.8524,
+      "step": 368
+    },
+    {
+      "epoch": 0.5790222135046339,
+      "grad_norm": 0.7357617134103119,
+      "learning_rate": 1.8641828719977244e-05,
+      "loss": 0.8345,
+      "step": 369
+    },
+    {
+      "epoch": 0.5805913793948904,
+      "grad_norm": 0.7109966959279804,
+      "learning_rate": 1.8633285381780987e-05,
+      "loss": 0.8353,
+      "step": 370
+    },
+    {
+      "epoch": 0.5821605452851468,
+      "grad_norm": 0.7024674889311903,
+      "learning_rate": 1.862471722797789e-05,
+      "loss": 0.8432,
+      "step": 371
+    },
+    {
+      "epoch": 0.5837297111754033,
+      "grad_norm": 0.6999505622425886,
+      "learning_rate": 1.861612428319634e-05,
+      "loss": 0.8411,
+      "step": 372
+    },
+    {
+      "epoch": 0.5852988770656598,
+      "grad_norm": 0.6982150916537049,
+      "learning_rate": 1.8607506572135994e-05,
+      "loss": 0.7959,
+      "step": 373
+    },
+    {
+      "epoch": 0.5868680429559162,
+      "grad_norm": 0.7020596340950395,
+      "learning_rate": 1.8598864119567693e-05,
+      "loss": 0.8017,
+      "step": 374
+    },
+    {
+      "epoch": 0.5884372088461727,
+      "grad_norm": 0.6686275845549737,
+      "learning_rate": 1.859019695033339e-05,
+      "loss": 0.802,
+      "step": 375
+    },
+    {
+      "epoch": 0.5900063747364291,
+      "grad_norm": 0.6645612734866118,
+      "learning_rate": 1.858150508934609e-05,
+      "loss": 0.8262,
+      "step": 376
+    },
+    {
+      "epoch": 0.5915755406266856,
+      "grad_norm": 0.6923508947440026,
+      "learning_rate": 1.8572788561589766e-05,
+      "loss": 0.8193,
+      "step": 377
+    },
+    {
+      "epoch": 0.5931447065169421,
+      "grad_norm": 0.7561539114886064,
+      "learning_rate": 1.8564047392119306e-05,
+      "loss": 0.8336,
+      "step": 378
+    },
+    {
+      "epoch": 0.5947138724071985,
+      "grad_norm": 0.6750352883555016,
+      "learning_rate": 1.855528160606041e-05,
+      "loss": 0.8167,
+      "step": 379
+    },
+    {
+      "epoch": 0.596283038297455,
+      "grad_norm": 0.7493837600956766,
+      "learning_rate": 1.8546491228609552e-05,
+      "loss": 0.8188,
+      "step": 380
+    },
+    {
+      "epoch": 0.5978522041877115,
+      "grad_norm": 0.731594857129808,
+      "learning_rate": 1.8537676285033886e-05,
+      "loss": 0.8324,
+      "step": 381
+    },
+    {
+      "epoch": 0.5994213700779679,
+      "grad_norm": 0.7001893897171816,
+      "learning_rate": 1.852883680067118e-05,
+      "loss": 0.8178,
+      "step": 382
+    },
+    {
+      "epoch": 0.6009905359682244,
+      "grad_norm": 0.6852142446532932,
+      "learning_rate": 1.851997280092974e-05,
+      "loss": 0.8102,
+      "step": 383
+    },
+    {
+      "epoch": 0.6025597018584808,
+      "grad_norm": 0.6983434794976544,
+      "learning_rate": 1.8511084311288338e-05,
+      "loss": 0.8332,
+      "step": 384
+    },
+    {
+      "epoch": 0.6041288677487373,
+      "grad_norm": 0.6933691219197556,
+      "learning_rate": 1.8502171357296144e-05,
+      "loss": 0.8146,
+      "step": 385
+    },
+    {
+      "epoch": 0.6056980336389938,
+      "grad_norm": 0.700602487961442,
+      "learning_rate": 1.8493233964572648e-05,
+      "loss": 0.8135,
+      "step": 386
+    },
+    {
+      "epoch": 0.6072671995292502,
+      "grad_norm": 0.6744943633457979,
+      "learning_rate": 1.8484272158807587e-05,
+      "loss": 0.8376,
+      "step": 387
+    },
+    {
+      "epoch": 0.6088363654195067,
+      "grad_norm": 0.7865201124937491,
+      "learning_rate": 1.847528596576087e-05,
+      "loss": 0.8117,
+      "step": 388
+    },
+    {
+      "epoch": 0.6104055313097632,
+      "grad_norm": 0.6433086814077168,
+      "learning_rate": 1.8466275411262504e-05,
+      "loss": 0.8155,
+      "step": 389
+    },
+    {
+      "epoch": 0.6119746972000196,
+      "grad_norm": 0.7076424626661005,
+      "learning_rate": 1.8457240521212525e-05,
+      "loss": 0.8166,
+      "step": 390
+    },
+    {
+      "epoch": 0.6135438630902761,
+      "grad_norm": 0.8078938878284891,
+      "learning_rate": 1.8448181321580916e-05,
+      "loss": 0.8459,
+      "step": 391
+    },
+    {
+      "epoch": 0.6151130289805326,
+      "grad_norm": 0.7461616818713168,
+      "learning_rate": 1.8439097838407532e-05,
+      "loss": 0.8049,
+      "step": 392
+    },
+    {
+      "epoch": 0.616682194870789,
+      "grad_norm": 0.7152268552139599,
+      "learning_rate": 1.8429990097802037e-05,
+      "loss": 0.808,
+      "step": 393
+    },
+    {
+      "epoch": 0.6182513607610455,
+      "grad_norm": 0.6846357455674837,
+      "learning_rate": 1.8420858125943818e-05,
+      "loss": 0.8045,
+      "step": 394
+    },
+    {
+      "epoch": 0.6198205266513019,
+      "grad_norm": 0.6983269194639065,
+      "learning_rate": 1.8411701949081906e-05,
+      "loss": 0.8105,
+      "step": 395
+    },
+    {
+      "epoch": 0.6213896925415584,
+      "grad_norm": 0.7605464379002871,
+      "learning_rate": 1.840252159353492e-05,
+      "loss": 0.797,
+      "step": 396
+    },
+    {
+      "epoch": 0.6229588584318149,
+      "grad_norm": 0.6983797607098158,
+      "learning_rate": 1.8393317085690964e-05,
+      "loss": 0.8569,
+      "step": 397
+    },
+    {
+      "epoch": 0.6245280243220713,
+      "grad_norm": 0.6748691689781784,
+      "learning_rate": 1.838408845200758e-05,
+      "loss": 0.8208,
+      "step": 398
+    },
+    {
+      "epoch": 0.6260971902123278,
+      "grad_norm": 0.8098834293213556,
+      "learning_rate": 1.837483571901164e-05,
+      "loss": 0.8147,
+      "step": 399
+    },
+    {
+      "epoch": 0.6276663561025843,
+      "grad_norm": 0.7181360293484808,
+      "learning_rate": 1.8365558913299307e-05,
+      "loss": 0.8299,
+      "step": 400
+    },
+    {
+      "epoch": 0.6292355219928407,
+      "grad_norm": 0.6969158627505524,
+      "learning_rate": 1.835625806153593e-05,
+      "loss": 0.8234,
+      "step": 401
+    },
+    {
+      "epoch": 0.6308046878830972,
+      "grad_norm": 0.754159219293913,
+      "learning_rate": 1.834693319045597e-05,
+      "loss": 0.8247,
+      "step": 402
+    },
+    {
+      "epoch": 0.6323738537733536,
+      "grad_norm": 0.7319986358100911,
+      "learning_rate": 1.8337584326862935e-05,
+      "loss": 0.8152,
+      "step": 403
+    },
+    {
+      "epoch": 0.6339430196636101,
+      "grad_norm": 0.7452151204389407,
+      "learning_rate": 1.8328211497629304e-05,
+      "loss": 0.8148,
+      "step": 404
+    },
+    {
+      "epoch": 0.6355121855538666,
+      "grad_norm": 0.6651465044606695,
+      "learning_rate": 1.8318814729696427e-05,
+      "loss": 0.824,
+      "step": 405
+    },
+    {
+      "epoch": 0.637081351444123,
+      "grad_norm": 0.6770145443430586,
+      "learning_rate": 1.8309394050074482e-05,
+      "loss": 0.8237,
+      "step": 406
+    },
+    {
+      "epoch": 0.6386505173343795,
+      "grad_norm": 0.6727769957144305,
+      "learning_rate": 1.8299949485842366e-05,
+      "loss": 0.8377,
+      "step": 407
+    },
+    {
+      "epoch": 0.6402196832246358,
+      "grad_norm": 0.8010705157322117,
+      "learning_rate": 1.8290481064147637e-05,
+      "loss": 0.8135,
+      "step": 408
+    },
+    {
+      "epoch": 0.6417888491148923,
+      "grad_norm": 0.6733077616751881,
+      "learning_rate": 1.828098881220642e-05,
+      "loss": 0.8432,
+      "step": 409
+    },
+    {
+      "epoch": 0.6433580150051488,
+      "grad_norm": 0.7089761572504684,
+      "learning_rate": 1.8271472757303352e-05,
+      "loss": 0.8379,
+      "step": 410
+    },
+    {
+      "epoch": 0.6449271808954052,
+      "grad_norm": 0.7752473611159495,
+      "learning_rate": 1.8261932926791477e-05,
+      "loss": 0.8398,
+      "step": 411
+    },
+    {
+      "epoch": 0.6464963467856617,
+      "grad_norm": 0.734051159422781,
+      "learning_rate": 1.8252369348092184e-05,
+      "loss": 0.8234,
+      "step": 412
+    },
+    {
+      "epoch": 0.6480655126759182,
+      "grad_norm": 0.6468926275147576,
+      "learning_rate": 1.824278204869513e-05,
+      "loss": 0.8038,
+      "step": 413
+    },
+    {
+      "epoch": 0.6496346785661746,
+      "grad_norm": 0.6950725330894157,
+      "learning_rate": 1.823317105615815e-05,
+      "loss": 0.8208,
+      "step": 414
+    },
+    {
+      "epoch": 0.6512038444564311,
+      "grad_norm": 0.6753962572420696,
+      "learning_rate": 1.8223536398107177e-05,
+      "loss": 0.8391,
+      "step": 415
+    },
+    {
+      "epoch": 0.6527730103466876,
+      "grad_norm": 0.7042901367806785,
+      "learning_rate": 1.8213878102236177e-05,
+      "loss": 0.8146,
+      "step": 416
+    },
+    {
+      "epoch": 0.654342176236944,
+      "grad_norm": 0.7500913478458661,
+      "learning_rate": 1.8204196196307058e-05,
+      "loss": 0.8314,
+      "step": 417
+    },
+    {
+      "epoch": 0.6559113421272005,
+      "grad_norm": 0.6957807209230921,
+      "learning_rate": 1.819449070814959e-05,
+      "loss": 0.8001,
+      "step": 418
+    },
+    {
+      "epoch": 0.6574805080174569,
+      "grad_norm": 0.6457468023907118,
+      "learning_rate": 1.8184761665661335e-05,
+      "loss": 0.7992,
+      "step": 419
+    },
+    {
+      "epoch": 0.6590496739077134,
+      "grad_norm": 0.6955793263668436,
+      "learning_rate": 1.8175009096807557e-05,
+      "loss": 0.8134,
+      "step": 420
+    },
+    {
+      "epoch": 0.6606188397979699,
+      "grad_norm": 0.7545609633979165,
+      "learning_rate": 1.8165233029621136e-05,
+      "loss": 0.8164,
+      "step": 421
+    },
+    {
+      "epoch": 0.6621880056882263,
+      "grad_norm": 0.6554743383140347,
+      "learning_rate": 1.8155433492202507e-05,
+      "loss": 0.8259,
+      "step": 422
+    },
+    {
+      "epoch": 0.6637571715784828,
+      "grad_norm": 0.7139814219154882,
+      "learning_rate": 1.8145610512719565e-05,
+      "loss": 0.8333,
+      "step": 423
+    },
+    {
+      "epoch": 0.6653263374687393,
+      "grad_norm": 0.6950404797930613,
+      "learning_rate": 1.8135764119407585e-05,
+      "loss": 0.8127,
+      "step": 424
+    },
+    {
+      "epoch": 0.6668955033589957,
+      "grad_norm": 0.641920314446041,
+      "learning_rate": 1.8125894340569145e-05,
+      "loss": 0.8045,
+      "step": 425
+    },
+    {
+      "epoch": 0.6684646692492522,
+      "grad_norm": 0.7501186303259437,
+      "learning_rate": 1.8116001204574042e-05,
+      "loss": 0.8251,
+      "step": 426
+    },
+    {
+      "epoch": 0.6700338351395086,
+      "grad_norm": 0.7091587706100728,
+      "learning_rate": 1.8106084739859206e-05,
+      "loss": 0.8217,
+      "step": 427
+    },
+    {
+      "epoch": 0.6716030010297651,
+      "grad_norm": 0.6999075333919983,
+      "learning_rate": 1.8096144974928635e-05,
+      "loss": 0.8167,
+      "step": 428
+    },
+    {
+      "epoch": 0.6731721669200216,
+      "grad_norm": 0.8425405072714384,
+      "learning_rate": 1.808618193835329e-05,
+      "loss": 0.8065,
+      "step": 429
+    },
+    {
+      "epoch": 0.674741332810278,
+      "grad_norm": 0.7656723967224829,
+      "learning_rate": 1.8076195658771033e-05,
+      "loss": 0.8271,
+      "step": 430
+    },
+    {
+      "epoch": 0.6763104987005345,
+      "grad_norm": 0.7407720143672063,
+      "learning_rate": 1.8066186164886525e-05,
+      "loss": 0.8063,
+      "step": 431
+    },
+    {
+      "epoch": 0.677879664590791,
+      "grad_norm": 0.8941476261659167,
+      "learning_rate": 1.8056153485471167e-05,
+      "loss": 0.8047,
+      "step": 432
+    },
+    {
+      "epoch": 0.6794488304810474,
+      "grad_norm": 0.6360637102703494,
+      "learning_rate": 1.8046097649362994e-05,
+      "loss": 0.8094,
+      "step": 433
+    },
+    {
+      "epoch": 0.6810179963713039,
+      "grad_norm": 0.7238862166213049,
+      "learning_rate": 1.8036018685466614e-05,
+      "loss": 0.8101,
+      "step": 434
+    },
+    {
+      "epoch": 0.6825871622615604,
+      "grad_norm": 0.7574138784308379,
+      "learning_rate": 1.80259166227531e-05,
+      "loss": 0.8079,
+      "step": 435
+    },
+    {
+      "epoch": 0.6841563281518168,
+      "grad_norm": 0.7023579698448427,
+      "learning_rate": 1.8015791490259932e-05,
+      "loss": 0.8145,
+      "step": 436
+    },
+    {
+      "epoch": 0.6857254940420733,
+      "grad_norm": 0.8225272967989602,
+      "learning_rate": 1.80056433170909e-05,
+      "loss": 0.82,
+      "step": 437
+    },
+    {
+      "epoch": 0.6872946599323297,
+      "grad_norm": 0.64159432837912,
+      "learning_rate": 1.799547213241602e-05,
+      "loss": 0.8296,
+      "step": 438
+    },
+    {
+      "epoch": 0.6888638258225862,
+      "grad_norm": 0.8057821733572857,
+      "learning_rate": 1.798527796547145e-05,
+      "loss": 0.8003,
+      "step": 439
+    },
+    {
+      "epoch": 0.6904329917128427,
+      "grad_norm": 0.716957403699838,
+      "learning_rate": 1.7975060845559412e-05,
+      "loss": 0.7932,
+      "step": 440
+    },
+    {
+      "epoch": 0.6920021576030991,
+      "grad_norm": 0.6642423022463745,
+      "learning_rate": 1.79648208020481e-05,
+      "loss": 0.8097,
+      "step": 441
+    },
+    {
+      "epoch": 0.6935713234933556,
+      "grad_norm": 0.7709982710378553,
+      "learning_rate": 1.7954557864371615e-05,
+      "loss": 0.8034,
+      "step": 442
+    },
+    {
+      "epoch": 0.6951404893836121,
+      "grad_norm": 0.6386502331455692,
+      "learning_rate": 1.794427206202984e-05,
+      "loss": 0.7855,
+      "step": 443
+    },
+    {
+      "epoch": 0.6967096552738685,
+      "grad_norm": 0.6497383877425416,
+      "learning_rate": 1.79339634245884e-05,
+      "loss": 0.8157,
+      "step": 444
+    },
+    {
+      "epoch": 0.698278821164125,
+      "grad_norm": 0.7699030331105429,
+      "learning_rate": 1.7923631981678552e-05,
+      "loss": 0.8272,
+      "step": 445
+    },
+    {
+      "epoch": 0.6998479870543814,
+      "grad_norm": 0.6970165818064721,
+      "learning_rate": 1.7913277762997105e-05,
+      "loss": 0.8118,
+      "step": 446
+    },
+    {
+      "epoch": 0.7014171529446379,
+      "grad_norm": 0.7099113528250288,
+      "learning_rate": 1.7902900798306334e-05,
+      "loss": 0.8209,
+      "step": 447
+    },
+    {
+      "epoch": 0.7029863188348944,
+      "grad_norm": 0.6720400867479696,
+      "learning_rate": 1.7892501117433896e-05,
+      "loss": 0.7994,
+      "step": 448
+    },
+    {
+      "epoch": 0.7045554847251508,
+      "grad_norm": 0.7888958270253592,
+      "learning_rate": 1.788207875027274e-05,
+      "loss": 0.8367,
+      "step": 449
+    },
+    {
+      "epoch": 0.7061246506154073,
+      "grad_norm": 0.7290525822038607,
+      "learning_rate": 1.7871633726781042e-05,
+      "loss": 0.813,
+      "step": 450
+    },
+    {
+      "epoch": 0.7076938165056637,
+      "grad_norm": 0.6698872656305315,
+      "learning_rate": 1.786116607698207e-05,
+      "loss": 0.799,
+      "step": 451
+    },
+    {
+      "epoch": 0.7092629823959201,
+      "grad_norm": 0.6393617197451871,
+      "learning_rate": 1.785067583096416e-05,
+      "loss": 0.8196,
+      "step": 452
+    },
+    {
+      "epoch": 0.7108321482861766,
+      "grad_norm": 0.7159540730903501,
+      "learning_rate": 1.784016301888058e-05,
+      "loss": 0.7878,
+      "step": 453
+    },
+    {
+      "epoch": 0.712401314176433,
+      "grad_norm": 0.7368282115740014,
+      "learning_rate": 1.7829627670949468e-05,
+      "loss": 0.8086,
+      "step": 454
+    },
+    {
+      "epoch": 0.7139704800666895,
+      "grad_norm": 0.6476538158940627,
+      "learning_rate": 1.7819069817453746e-05,
+      "loss": 0.7988,
+      "step": 455
+    },
+    {
+      "epoch": 0.715539645956946,
+      "grad_norm": 0.784552921400997,
+      "learning_rate": 1.7808489488741013e-05,
+      "loss": 0.8233,
+      "step": 456
+    },
+    {
+      "epoch": 0.7171088118472024,
+      "grad_norm": 0.7178294530644177,
+      "learning_rate": 1.779788671522348e-05,
+      "loss": 0.8073,
+      "step": 457
+    },
+    {
+      "epoch": 0.7186779777374589,
+      "grad_norm": 0.6730171014633602,
+      "learning_rate": 1.778726152737787e-05,
+      "loss": 0.8221,
+      "step": 458
+    },
+    {
+      "epoch": 0.7202471436277154,
+      "grad_norm": 0.8214261337376169,
+      "learning_rate": 1.7776613955745337e-05,
+      "loss": 0.8297,
+      "step": 459
+    },
+    {
+      "epoch": 0.7218163095179718,
+      "grad_norm": 0.6893321165565585,
+      "learning_rate": 1.776594403093138e-05,
+      "loss": 0.8321,
+      "step": 460
+    },
+    {
+      "epoch": 0.7233854754082283,
+      "grad_norm": 0.8516099663226969,
+      "learning_rate": 1.7755251783605733e-05,
+      "loss": 0.7984,
+      "step": 461
+    },
+    {
+      "epoch": 0.7249546412984847,
+      "grad_norm": 0.7839754746050254,
+      "learning_rate": 1.774453724450231e-05,
+      "loss": 0.8425,
+      "step": 462
+    },
+    {
+      "epoch": 0.7265238071887412,
+      "grad_norm": 0.6537865642848764,
+      "learning_rate": 1.7733800444419088e-05,
+      "loss": 0.8158,
+      "step": 463
+    },
+    {
+      "epoch": 0.7280929730789977,
+      "grad_norm": 0.7533825963172626,
+      "learning_rate": 1.7723041414218053e-05,
+      "loss": 0.8107,
+      "step": 464
+    },
+    {
+      "epoch": 0.7296621389692541,
+      "grad_norm": 0.7109718264788607,
+      "learning_rate": 1.7712260184825063e-05,
+      "loss": 0.8251,
+      "step": 465
+    },
+    {
+      "epoch": 0.7312313048595106,
+      "grad_norm": 0.6721597266694148,
+      "learning_rate": 1.7701456787229805e-05,
+      "loss": 0.8117,
+      "step": 466
+    },
+    {
+      "epoch": 0.7328004707497671,
+      "grad_norm": 0.6881719830558359,
+      "learning_rate": 1.7690631252485676e-05,
+      "loss": 0.7967,
+      "step": 467
+    },
+    {
+      "epoch": 0.7343696366400235,
+      "grad_norm": 0.6445607899227929,
+      "learning_rate": 1.767978361170971e-05,
+      "loss": 0.8097,
+      "step": 468
+    },
+    {
+      "epoch": 0.73593880253028,
+      "grad_norm": 0.7504824832613186,
+      "learning_rate": 1.766891389608248e-05,
+      "loss": 0.835,
+      "step": 469
+    },
+    {
+      "epoch": 0.7375079684205365,
+      "grad_norm": 0.7866067337434499,
+      "learning_rate": 1.7658022136848008e-05,
+      "loss": 0.8014,
+      "step": 470
+    },
+    {
+      "epoch": 0.7390771343107929,
+      "grad_norm": 0.6303009056377867,
+      "learning_rate": 1.7647108365313683e-05,
+      "loss": 0.8233,
+      "step": 471
+    },
+    {
+      "epoch": 0.7406463002010494,
+      "grad_norm": 0.7147872336913142,
+      "learning_rate": 1.763617261285017e-05,
+      "loss": 0.8154,
+      "step": 472
+    },
+    {
+      "epoch": 0.7422154660913058,
+      "grad_norm": 0.7829534690519674,
+      "learning_rate": 1.7625214910891307e-05,
+      "loss": 0.8299,
+      "step": 473
+    },
+    {
+      "epoch": 0.7437846319815623,
+      "grad_norm": 0.675903984438671,
+      "learning_rate": 1.761423529093403e-05,
+      "loss": 0.8245,
+      "step": 474
+    },
+    {
+      "epoch": 0.7453537978718188,
+      "grad_norm": 0.7220838854114594,
+      "learning_rate": 1.7603233784538275e-05,
+      "loss": 0.8274,
+      "step": 475
+    },
+    {
+      "epoch": 0.7469229637620752,
+      "grad_norm": 0.7917437857037457,
+      "learning_rate": 1.759221042332689e-05,
+      "loss": 0.8475,
+      "step": 476
+    },
+    {
+      "epoch": 0.7484921296523317,
+      "grad_norm": 0.7149355792999409,
+      "learning_rate": 1.7581165238985533e-05,
+      "loss": 0.809,
+      "step": 477
+    },
+    {
+      "epoch": 0.7500612955425882,
+      "grad_norm": 0.6910650547613271,
+      "learning_rate": 1.7570098263262608e-05,
+      "loss": 0.8132,
+      "step": 478
+    },
+    {
+      "epoch": 0.7516304614328446,
+      "grad_norm": 0.7650525336029764,
+      "learning_rate": 1.7559009527969145e-05,
+      "loss": 0.8107,
+      "step": 479
+    },
+    {
+      "epoch": 0.7531996273231011,
+      "grad_norm": 0.7678212906690784,
+      "learning_rate": 1.754789906497872e-05,
+      "loss": 0.8109,
+      "step": 480
+    },
+    {
+      "epoch": 0.7547687932133575,
+      "grad_norm": 0.6383707499049117,
+      "learning_rate": 1.7536766906227365e-05,
+      "loss": 0.8063,
+      "step": 481
+    },
+    {
+      "epoch": 0.756337959103614,
+      "grad_norm": 0.8522670527346118,
+      "learning_rate": 1.752561308371348e-05,
+      "loss": 0.8157,
+      "step": 482
+    },
+    {
+      "epoch": 0.7579071249938705,
+      "grad_norm": 0.7930403589499919,
+      "learning_rate": 1.751443762949772e-05,
+      "loss": 0.8029,
+      "step": 483
+    },
+    {
+      "epoch": 0.7594762908841269,
+      "grad_norm": 0.6288113764581796,
+      "learning_rate": 1.750324057570294e-05,
+      "loss": 0.8252,
+      "step": 484
+    },
+    {
+      "epoch": 0.7610454567743834,
+      "grad_norm": 0.7031530416277852,
+      "learning_rate": 1.7492021954514068e-05,
+      "loss": 0.8077,
+      "step": 485
+    },
+    {
+      "epoch": 0.7626146226646399,
+      "grad_norm": 0.7708218065173724,
+      "learning_rate": 1.748078179817802e-05,
+      "loss": 0.8023,
+      "step": 486
+    },
+    {
+      "epoch": 0.7641837885548963,
+      "grad_norm": 0.6799372475491817,
+      "learning_rate": 1.7469520139003627e-05,
+      "loss": 0.8203,
+      "step": 487
+    },
+    {
+      "epoch": 0.7657529544451528,
+      "grad_norm": 0.6992511409953582,
+      "learning_rate": 1.745823700936152e-05,
+      "loss": 0.831,
+      "step": 488
+    },
+    {
+      "epoch": 0.7673221203354093,
+      "grad_norm": 0.6781285850790751,
+      "learning_rate": 1.7446932441684044e-05,
+      "loss": 0.8379,
+      "step": 489
+    },
+    {
+      "epoch": 0.7688912862256657,
+      "grad_norm": 0.7003178265211839,
+      "learning_rate": 1.743560646846517e-05,
+      "loss": 0.8156,
+      "step": 490
+    },
+    {
+      "epoch": 0.7704604521159222,
+      "grad_norm": 0.6281699122180051,
+      "learning_rate": 1.7424259122260396e-05,
+      "loss": 0.8118,
+      "step": 491
+    },
+    {
+      "epoch": 0.7720296180061786,
+      "grad_norm": 0.6555713044357072,
+      "learning_rate": 1.741289043568665e-05,
+      "loss": 0.8137,
+      "step": 492
+    },
+    {
+      "epoch": 0.7735987838964351,
+      "grad_norm": 0.7316387142927496,
+      "learning_rate": 1.740150044142221e-05,
+      "loss": 0.8007,
+      "step": 493
+    },
+    {
+      "epoch": 0.7751679497866915,
+      "grad_norm": 0.722609826172713,
+      "learning_rate": 1.7390089172206594e-05,
+      "loss": 0.8345,
+      "step": 494
+    },
+    {
+      "epoch": 0.7767371156769479,
+      "grad_norm": 0.6531002990849316,
+      "learning_rate": 1.7378656660840475e-05,
+      "loss": 0.8249,
+      "step": 495
+    },
+    {
+      "epoch": 0.7783062815672044,
+      "grad_norm": 0.6741354485887595,
+      "learning_rate": 1.7367202940185583e-05,
+      "loss": 0.8341,
+      "step": 496
+    },
+    {
+      "epoch": 0.7798754474574608,
+      "grad_norm": 0.6709988543956592,
+      "learning_rate": 1.7355728043164613e-05,
+      "loss": 0.7946,
+      "step": 497
+    },
+    {
+      "epoch": 0.7814446133477173,
+      "grad_norm": 0.6595885302529915,
+      "learning_rate": 1.7344232002761138e-05,
+      "loss": 0.8125,
+      "step": 498
+    },
+    {
+      "epoch": 0.7830137792379738,
+      "grad_norm": 0.710979420911621,
+      "learning_rate": 1.7332714852019487e-05,
+      "loss": 0.8054,
+      "step": 499
+    },
+    {
+      "epoch": 0.7845829451282302,
+      "grad_norm": 0.6533184911971489,
+      "learning_rate": 1.732117662404469e-05,
+      "loss": 0.8434,
+      "step": 500
+    },
+    {
+      "epoch": 0.7861521110184867,
+      "grad_norm": 0.6659869923020332,
+      "learning_rate": 1.7309617352002343e-05,
+      "loss": 0.7999,
+      "step": 501
+    },
+    {
+      "epoch": 0.7877212769087432,
+      "grad_norm": 0.6568986059667483,
+      "learning_rate": 1.7298037069118546e-05,
+      "loss": 0.8009,
+      "step": 502
+    },
+    {
+      "epoch": 0.7892904427989996,
+      "grad_norm": 0.6354201540803367,
+      "learning_rate": 1.7286435808679787e-05,
+      "loss": 0.8188,
+      "step": 503
+    },
+    {
+      "epoch": 0.7908596086892561,
+      "grad_norm": 0.6672100749606731,
+      "learning_rate": 1.7274813604032847e-05,
+      "loss": 0.7892,
+      "step": 504
+    },
+    {
+      "epoch": 0.7924287745795126,
+      "grad_norm": 0.6371444526615044,
+      "learning_rate": 1.7263170488584717e-05,
+      "loss": 0.8131,
+      "step": 505
+    },
+    {
+      "epoch": 0.793997940469769,
+      "grad_norm": 0.7456005982963471,
+      "learning_rate": 1.725150649580249e-05,
+      "loss": 0.8383,
+      "step": 506
+    },
+    {
+      "epoch": 0.7955671063600255,
+      "grad_norm": 0.6511784387683395,
+      "learning_rate": 1.7239821659213272e-05,
+      "loss": 0.8088,
+      "step": 507
+    },
+    {
+      "epoch": 0.7971362722502819,
+      "grad_norm": 0.6793395853342363,
+      "learning_rate": 1.7228116012404085e-05,
+      "loss": 0.8165,
+      "step": 508
+    },
+    {
+      "epoch": 0.7987054381405384,
+      "grad_norm": 0.6691807054058334,
+      "learning_rate": 1.721638958902175e-05,
+      "loss": 0.8315,
+      "step": 509
+    },
+    {
+      "epoch": 0.8002746040307949,
+      "grad_norm": 0.6504268301547628,
+      "learning_rate": 1.7204642422772836e-05,
+      "loss": 0.8096,
+      "step": 510
+    },
+    {
+      "epoch": 0.8018437699210513,
+      "grad_norm": 0.6534928735312999,
+      "learning_rate": 1.7192874547423514e-05,
+      "loss": 0.84,
+      "step": 511
+    },
+    {
+      "epoch": 0.8034129358113078,
+      "grad_norm": 0.6838591730669165,
+      "learning_rate": 1.71810859967995e-05,
+      "loss": 0.8014,
+      "step": 512
+    },
+    {
+      "epoch": 0.8049821017015643,
+      "grad_norm": 0.6165450255519258,
+      "learning_rate": 1.716927680478591e-05,
+      "loss": 0.7928,
+      "step": 513
+    },
+    {
+      "epoch": 0.8065512675918207,
+      "grad_norm": 0.6587483637904128,
+      "learning_rate": 1.715744700532723e-05,
+      "loss": 0.8181,
+      "step": 514
+    },
+    {
+      "epoch": 0.8081204334820772,
+      "grad_norm": 0.6524271238302766,
+      "learning_rate": 1.714559663242715e-05,
+      "loss": 0.8101,
+      "step": 515
+    },
+    {
+      "epoch": 0.8096895993723336,
+      "grad_norm": 0.6403769753872768,
+      "learning_rate": 1.7133725720148513e-05,
+      "loss": 0.7953,
+      "step": 516
+    },
+    {
+      "epoch": 0.8112587652625901,
+      "grad_norm": 0.6357272635711984,
+      "learning_rate": 1.712183430261319e-05,
+      "loss": 0.8136,
+      "step": 517
+    },
+    {
+      "epoch": 0.8128279311528466,
+      "grad_norm": 0.6058535344892332,
+      "learning_rate": 1.7109922414001997e-05,
+      "loss": 0.8116,
+      "step": 518
+    },
+    {
+      "epoch": 0.814397097043103,
+      "grad_norm": 0.7117480045479407,
+      "learning_rate": 1.7097990088554595e-05,
+      "loss": 0.7945,
+      "step": 519
+    },
+    {
+      "epoch": 0.8159662629333595,
+      "grad_norm": 0.640144578639951,
+      "learning_rate": 1.7086037360569386e-05,
+      "loss": 0.818,
+      "step": 520
+    },
+    {
+      "epoch": 0.817535428823616,
+      "grad_norm": 0.6492922119177317,
+      "learning_rate": 1.707406426440342e-05,
+      "loss": 0.8093,
+      "step": 521
+    },
+    {
+      "epoch": 0.8191045947138724,
+      "grad_norm": 0.7197118156272728,
+      "learning_rate": 1.7062070834472287e-05,
+      "loss": 0.808,
+      "step": 522
+    },
+    {
+      "epoch": 0.8206737606041289,
+      "grad_norm": 0.7818181859268165,
+      "learning_rate": 1.7050057105250034e-05,
+      "loss": 0.8161,
+      "step": 523
+    },
+    {
+      "epoch": 0.8222429264943854,
+      "grad_norm": 0.6941809311994893,
+      "learning_rate": 1.7038023111269045e-05,
+      "loss": 0.8247,
+      "step": 524
+    },
+    {
+      "epoch": 0.8238120923846418,
+      "grad_norm": 0.7386763962568045,
+      "learning_rate": 1.7025968887119962e-05,
+      "loss": 0.8178,
+      "step": 525
+    },
+    {
+      "epoch": 0.8253812582748983,
+      "grad_norm": 0.7077831748661425,
+      "learning_rate": 1.701389446745158e-05,
+      "loss": 0.8161,
+      "step": 526
+    },
+    {
+      "epoch": 0.8269504241651547,
+      "grad_norm": 0.6526885632115078,
+      "learning_rate": 1.7001799886970735e-05,
+      "loss": 0.799,
+      "step": 527
+    },
+    {
+      "epoch": 0.8285195900554112,
+      "grad_norm": 0.7639975857248498,
+      "learning_rate": 1.6989685180442212e-05,
+      "loss": 0.781,
+      "step": 528
+    },
+    {
+      "epoch": 0.8300887559456677,
+      "grad_norm": 0.6684469040627832,
+      "learning_rate": 1.697755038268866e-05,
+      "loss": 0.7856,
+      "step": 529
+    },
+    {
+      "epoch": 0.8316579218359241,
+      "grad_norm": 0.7058117508872238,
+      "learning_rate": 1.696539552859046e-05,
+      "loss": 0.8014,
+      "step": 530
+    },
+    {
+      "epoch": 0.8332270877261806,
+      "grad_norm": 0.6836163548650119,
+      "learning_rate": 1.6953220653085662e-05,
+      "loss": 0.7861,
+      "step": 531
+    },
+    {
+      "epoch": 0.8347962536164371,
+      "grad_norm": 0.6527971166004355,
+      "learning_rate": 1.6941025791169855e-05,
+      "loss": 0.7992,
+      "step": 532
+    },
+    {
+      "epoch": 0.8363654195066935,
+      "grad_norm": 0.6978339614853983,
+      "learning_rate": 1.692881097789608e-05,
+      "loss": 0.794,
+      "step": 533
+    },
+    {
+      "epoch": 0.83793458539695,
+      "grad_norm": 0.6821421763392262,
+      "learning_rate": 1.691657624837472e-05,
+      "loss": 0.8234,
+      "step": 534
+    },
+    {
+      "epoch": 0.8395037512872064,
+      "grad_norm": 0.6675300069385321,
+      "learning_rate": 1.6904321637773414e-05,
+      "loss": 0.8032,
+      "step": 535
+    },
+    {
+      "epoch": 0.8410729171774629,
+      "grad_norm": 0.7346622800030119,
+      "learning_rate": 1.6892047181316952e-05,
+      "loss": 0.7911,
+      "step": 536
+    },
+    {
+      "epoch": 0.8426420830677193,
+      "grad_norm": 0.739051301794649,
+      "learning_rate": 1.687975291428715e-05,
+      "loss": 0.7891,
+      "step": 537
+    },
+    {
+      "epoch": 0.8442112489579757,
+      "grad_norm": 0.6373011156184432,
+      "learning_rate": 1.6867438872022782e-05,
+      "loss": 0.8069,
+      "step": 538
+    },
+    {
+      "epoch": 0.8457804148482322,
+      "grad_norm": 0.6709585938330882,
+      "learning_rate": 1.6855105089919467e-05,
+      "loss": 0.8332,
+      "step": 539
+    },
+    {
+      "epoch": 0.8473495807384887,
+      "grad_norm": 0.677293117839924,
+      "learning_rate": 1.684275160342955e-05,
+      "loss": 0.7958,
+      "step": 540
+    },
+    {
+      "epoch": 0.8489187466287451,
+      "grad_norm": 0.6761789735191491,
+      "learning_rate": 1.683037844806203e-05,
+      "loss": 0.8076,
+      "step": 541
+    },
+    {
+      "epoch": 0.8504879125190016,
+      "grad_norm": 0.6408496083925778,
+      "learning_rate": 1.6817985659382437e-05,
+      "loss": 0.8117,
+      "step": 542
+    },
+    {
+      "epoch": 0.852057078409258,
+      "grad_norm": 0.6430311453864617,
+      "learning_rate": 1.680557327301273e-05,
+      "loss": 0.8161,
+      "step": 543
+    },
+    {
+      "epoch": 0.8536262442995145,
+      "grad_norm": 0.7397332654739386,
+      "learning_rate": 1.67931413246312e-05,
+      "loss": 0.8106,
+      "step": 544
+    },
+    {
+      "epoch": 0.855195410189771,
+      "grad_norm": 0.7299719224754243,
+      "learning_rate": 1.678068984997238e-05,
+      "loss": 0.8189,
+      "step": 545
+    },
+    {
+      "epoch": 0.8567645760800274,
+      "grad_norm": 0.6491507227287643,
+      "learning_rate": 1.6768218884826916e-05,
+      "loss": 0.8176,
+      "step": 546
+    },
+    {
+      "epoch": 0.8583337419702839,
+      "grad_norm": 0.7289415570036244,
+      "learning_rate": 1.675572846504148e-05,
+      "loss": 0.8036,
+      "step": 547
+    },
+    {
+      "epoch": 0.8599029078605404,
+      "grad_norm": 0.7030586652496121,
+      "learning_rate": 1.674321862651867e-05,
+      "loss": 0.783,
+      "step": 548
+    },
+    {
+      "epoch": 0.8614720737507968,
+      "grad_norm": 0.6497840412025223,
+      "learning_rate": 1.6730689405216895e-05,
+      "loss": 0.8181,
+      "step": 549
+    },
+    {
+      "epoch": 0.8630412396410533,
+      "grad_norm": 0.6676400549113347,
+      "learning_rate": 1.6718140837150286e-05,
+      "loss": 0.8343,
+      "step": 550
+    },
+    {
+      "epoch": 0.8646104055313097,
+      "grad_norm": 0.7593329134013538,
+      "learning_rate": 1.6705572958388576e-05,
+      "loss": 0.8094,
+      "step": 551
+    },
+    {
+      "epoch": 0.8661795714215662,
+      "grad_norm": 0.8003932084095836,
+      "learning_rate": 1.669298580505701e-05,
+      "loss": 0.7976,
+      "step": 552
+    },
+    {
+      "epoch": 0.8677487373118227,
+      "grad_norm": 0.6574327020780073,
+      "learning_rate": 1.668037941333623e-05,
+      "loss": 0.7995,
+      "step": 553
+    },
+    {
+      "epoch": 0.8693179032020791,
+      "grad_norm": 0.7513090844076649,
+      "learning_rate": 1.6667753819462196e-05,
+      "loss": 0.8091,
+      "step": 554
+    },
+    {
+      "epoch": 0.8708870690923356,
+      "grad_norm": 0.7905519835629078,
+      "learning_rate": 1.665510905972603e-05,
+      "loss": 0.7994,
+      "step": 555
+    },
+    {
+      "epoch": 0.8724562349825921,
+      "grad_norm": 0.6830400188674131,
+      "learning_rate": 1.664244517047398e-05,
+      "loss": 0.8211,
+      "step": 556
+    },
+    {
+      "epoch": 0.8740254008728485,
+      "grad_norm": 0.6631120909160505,
+      "learning_rate": 1.6629762188107248e-05,
+      "loss": 0.7918,
+      "step": 557
+    },
+    {
+      "epoch": 0.875594566763105,
+      "grad_norm": 0.7167243386372935,
+      "learning_rate": 1.6617060149081944e-05,
+      "loss": 0.8367,
+      "step": 558
+    },
+    {
+      "epoch": 0.8771637326533614,
+      "grad_norm": 0.696860779606996,
+      "learning_rate": 1.6604339089908935e-05,
+      "loss": 0.8103,
+      "step": 559
+    },
+    {
+      "epoch": 0.8787328985436179,
+      "grad_norm": 0.6807416652503484,
+      "learning_rate": 1.6591599047153778e-05,
+      "loss": 0.7905,
+      "step": 560
+    },
+    {
+      "epoch": 0.8803020644338744,
+      "grad_norm": 0.6744850894294322,
+      "learning_rate": 1.657884005743658e-05,
+      "loss": 0.7875,
+      "step": 561
+    },
+    {
+      "epoch": 0.8818712303241308,
+      "grad_norm": 0.6892485774873763,
+      "learning_rate": 1.6566062157431915e-05,
+      "loss": 0.8014,
+      "step": 562
+    },
+    {
+      "epoch": 0.8834403962143873,
+      "grad_norm": 0.6506627437686835,
+      "learning_rate": 1.655326538386872e-05,
+      "loss": 0.8068,
+      "step": 563
+    },
+    {
+      "epoch": 0.8850095621046438,
+      "grad_norm": 0.872940706391606,
+      "learning_rate": 1.6540449773530168e-05,
+      "loss": 0.8227,
+      "step": 564
+    },
+    {
+      "epoch": 0.8865787279949002,
+      "grad_norm": 0.7041284488554426,
+      "learning_rate": 1.652761536325359e-05,
+      "loss": 0.7926,
+      "step": 565
+    },
+    {
+      "epoch": 0.8881478938851567,
+      "grad_norm": 0.7439493389344818,
+      "learning_rate": 1.6514762189930353e-05,
+      "loss": 0.8106,
+      "step": 566
+    },
+    {
+      "epoch": 0.8897170597754132,
+      "grad_norm": 0.6983851282263962,
+      "learning_rate": 1.650189029050575e-05,
+      "loss": 0.7708,
+      "step": 567
+    },
+    {
+      "epoch": 0.8912862256656696,
+      "grad_norm": 0.6917605366023396,
+      "learning_rate": 1.6488999701978905e-05,
+      "loss": 0.7659,
+      "step": 568
+    },
+    {
+      "epoch": 0.8928553915559261,
+      "grad_norm": 0.6377545035956164,
+      "learning_rate": 1.6476090461402664e-05,
+      "loss": 0.7944,
+      "step": 569
+    },
+    {
+      "epoch": 0.8944245574461825,
+      "grad_norm": 0.7392678410611455,
+      "learning_rate": 1.6463162605883485e-05,
+      "loss": 0.794,
+      "step": 570
+    },
+    {
+      "epoch": 0.895993723336439,
+      "grad_norm": 0.705170379300143,
+      "learning_rate": 1.645021617258133e-05,
+      "loss": 0.7925,
+      "step": 571
+    },
+    {
+      "epoch": 0.8975628892266955,
+      "grad_norm": 0.6128144490754132,
+      "learning_rate": 1.6437251198709567e-05,
+      "loss": 0.8006,
+      "step": 572
+    },
+    {
+      "epoch": 0.8991320551169519,
+      "grad_norm": 0.7117533384983966,
+      "learning_rate": 1.6424267721534856e-05,
+      "loss": 0.8134,
+      "step": 573
+    },
+    {
+      "epoch": 0.9007012210072084,
+      "grad_norm": 0.8039404801729146,
+      "learning_rate": 1.641126577837703e-05,
+      "loss": 0.7725,
+      "step": 574
+    },
+    {
+      "epoch": 0.9022703868974649,
+      "grad_norm": 0.659058227738277,
+      "learning_rate": 1.6398245406609027e-05,
+      "loss": 0.7845,
+      "step": 575
+    },
+    {
+      "epoch": 0.9038395527877213,
+      "grad_norm": 0.6551723308376716,
+      "learning_rate": 1.638520664365673e-05,
+      "loss": 0.8051,
+      "step": 576
+    },
+    {
+      "epoch": 0.9054087186779778,
+      "grad_norm": 0.7339846146712853,
+      "learning_rate": 1.6372149526998905e-05,
+      "loss": 0.7969,
+      "step": 577
+    },
+    {
+      "epoch": 0.9069778845682342,
+      "grad_norm": 0.6972162560558418,
+      "learning_rate": 1.635907409416706e-05,
+      "loss": 0.7813,
+      "step": 578
+    },
+    {
+      "epoch": 0.9085470504584907,
+      "grad_norm": 0.6665523295553141,
+      "learning_rate": 1.634598038274536e-05,
+      "loss": 0.8136,
+      "step": 579
+    },
+    {
+      "epoch": 0.9101162163487471,
+      "grad_norm": 0.7291147795349994,
+      "learning_rate": 1.6332868430370508e-05,
+      "loss": 0.8109,
+      "step": 580
+    },
+    {
+      "epoch": 0.9116853822390035,
+      "grad_norm": 0.8210953729995312,
+      "learning_rate": 1.631973827473164e-05,
+      "loss": 0.7945,
+      "step": 581
+    },
+    {
+      "epoch": 0.91325454812926,
+      "grad_norm": 0.637417375253505,
+      "learning_rate": 1.6306589953570207e-05,
+      "loss": 0.7873,
+      "step": 582
+    },
+    {
+      "epoch": 0.9148237140195165,
+      "grad_norm": 0.6572399735162374,
+      "learning_rate": 1.6293423504679898e-05,
+      "loss": 0.8011,
+      "step": 583
+    },
+    {
+      "epoch": 0.9163928799097729,
+      "grad_norm": 0.8253311695850781,
+      "learning_rate": 1.6280238965906474e-05,
+      "loss": 0.8112,
+      "step": 584
+    },
+    {
+      "epoch": 0.9179620458000294,
+      "grad_norm": 0.6770273612358756,
+      "learning_rate": 1.6267036375147728e-05,
+      "loss": 0.8016,
+      "step": 585
+    },
+    {
+      "epoch": 0.9195312116902858,
+      "grad_norm": 0.7928913502491136,
+      "learning_rate": 1.6253815770353316e-05,
+      "loss": 0.8001,
+      "step": 586
+    },
+    {
+      "epoch": 0.9211003775805423,
+      "grad_norm": 0.7135756659960236,
+      "learning_rate": 1.6240577189524688e-05,
+      "loss": 0.8056,
+      "step": 587
+    },
+    {
+      "epoch": 0.9226695434707988,
+      "grad_norm": 0.7348035046236058,
+      "learning_rate": 1.6227320670714966e-05,
+      "loss": 0.7864,
+      "step": 588
+    },
+    {
+      "epoch": 0.9242387093610552,
+      "grad_norm": 0.7994395457219496,
+      "learning_rate": 1.6214046252028823e-05,
+      "loss": 0.8077,
+      "step": 589
+    },
+    {
+      "epoch": 0.9258078752513117,
+      "grad_norm": 0.7586472260929925,
+      "learning_rate": 1.6200753971622387e-05,
+      "loss": 0.8155,
+      "step": 590
+    },
+    {
+      "epoch": 0.9273770411415682,
+      "grad_norm": 0.6969293077110756,
+      "learning_rate": 1.618744386770313e-05,
+      "loss": 0.7827,
+      "step": 591
+    },
+    {
+      "epoch": 0.9289462070318246,
+      "grad_norm": 0.7310085843341342,
+      "learning_rate": 1.617411597852976e-05,
+      "loss": 0.7897,
+      "step": 592
+    },
+    {
+      "epoch": 0.9305153729220811,
+      "grad_norm": 0.8143514280731344,
+      "learning_rate": 1.6160770342412097e-05,
+      "loss": 0.7906,
+      "step": 593
+    },
+    {
+      "epoch": 0.9320845388123375,
+      "grad_norm": 0.7111896304465191,
+      "learning_rate": 1.614740699771098e-05,
+      "loss": 0.8025,
+      "step": 594
+    },
+    {
+      "epoch": 0.933653704702594,
+      "grad_norm": 0.7037008978412734,
+      "learning_rate": 1.6134025982838153e-05,
+      "loss": 0.8099,
+      "step": 595
+    },
+    {
+      "epoch": 0.9352228705928505,
+      "grad_norm": 0.6685839806742361,
+      "learning_rate": 1.6120627336256142e-05,
+      "loss": 0.7937,
+      "step": 596
+    },
+    {
+      "epoch": 0.9367920364831069,
+      "grad_norm": 0.6827532038787048,
+      "learning_rate": 1.610721109647816e-05,
+      "loss": 0.8105,
+      "step": 597
+    },
+    {
+      "epoch": 0.9383612023733634,
+      "grad_norm": 0.7455307978046383,
+      "learning_rate": 1.609377730206799e-05,
+      "loss": 0.8081,
+      "step": 598
+    },
+    {
+      "epoch": 0.9399303682636199,
+      "grad_norm": 0.6851142967985461,
+      "learning_rate": 1.6080325991639866e-05,
+      "loss": 0.8064,
+      "step": 599
+    },
+    {
+      "epoch": 0.9414995341538763,
+      "grad_norm": 0.8055141185224926,
+      "learning_rate": 1.6066857203858387e-05,
+      "loss": 0.8292,
+      "step": 600
+    },
+    {
+      "epoch": 0.9430687000441328,
+      "grad_norm": 0.6564242605325623,
+      "learning_rate": 1.605337097743837e-05,
+      "loss": 0.8237,
+      "step": 601
+    },
+    {
+      "epoch": 0.9446378659343893,
+      "grad_norm": 0.7054701247398754,
+      "learning_rate": 1.6039867351144778e-05,
+      "loss": 0.7874,
+      "step": 602
+    },
+    {
+      "epoch": 0.9462070318246457,
+      "grad_norm": 0.7358471397571945,
+      "learning_rate": 1.6026346363792565e-05,
+      "loss": 0.7977,
+      "step": 603
+    },
+    {
+      "epoch": 0.9477761977149022,
+      "grad_norm": 0.6529066045354731,
+      "learning_rate": 1.6012808054246608e-05,
+      "loss": 0.8131,
+      "step": 604
+    },
+    {
+      "epoch": 0.9493453636051586,
+      "grad_norm": 0.6962657017653682,
+      "learning_rate": 1.5999252461421564e-05,
+      "loss": 0.7983,
+      "step": 605
+    },
+    {
+      "epoch": 0.9509145294954151,
+      "grad_norm": 0.6459694871618518,
+      "learning_rate": 1.5985679624281772e-05,
+      "loss": 0.7979,
+      "step": 606
+    },
+    {
+      "epoch": 0.9524836953856716,
+      "grad_norm": 0.6978067782862442,
+      "learning_rate": 1.597208958184114e-05,
+      "loss": 0.7997,
+      "step": 607
+    },
+    {
+      "epoch": 0.954052861275928,
+      "grad_norm": 0.6068382673996924,
+      "learning_rate": 1.595848237316303e-05,
+      "loss": 0.8017,
+      "step": 608
+    },
+    {
+      "epoch": 0.9556220271661845,
+      "grad_norm": 0.6789099200038117,
+      "learning_rate": 1.5944858037360145e-05,
+      "loss": 0.8076,
+      "step": 609
+    },
+    {
+      "epoch": 0.957191193056441,
+      "grad_norm": 0.6596958956112973,
+      "learning_rate": 1.5931216613594426e-05,
+      "loss": 0.7958,
+      "step": 610
+    },
+    {
+      "epoch": 0.9587603589466974,
+      "grad_norm": 0.6296782433758215,
+      "learning_rate": 1.5917558141076915e-05,
+      "loss": 0.7926,
+      "step": 611
+    },
+    {
+      "epoch": 0.9603295248369539,
+      "grad_norm": 0.6805938856593269,
+      "learning_rate": 1.590388265906768e-05,
+      "loss": 0.8197,
+      "step": 612
+    },
+    {
+      "epoch": 0.9618986907272103,
+      "grad_norm": 0.6410835407854243,
+      "learning_rate": 1.5890190206875663e-05,
+      "loss": 0.784,
+      "step": 613
+    },
+    {
+      "epoch": 0.9634678566174668,
+      "grad_norm": 0.6277299047590931,
+      "learning_rate": 1.58764808238586e-05,
+      "loss": 0.7902,
+      "step": 614
+    },
+    {
+      "epoch": 0.9650370225077233,
+      "grad_norm": 0.6772710416091088,
+      "learning_rate": 1.5862754549422884e-05,
+      "loss": 0.7988,
+      "step": 615
+    },
+    {
+      "epoch": 0.9666061883979797,
+      "grad_norm": 0.6181929496022875,
+      "learning_rate": 1.5849011423023463e-05,
+      "loss": 0.7938,
+      "step": 616
+    },
+    {
+      "epoch": 0.9681753542882362,
+      "grad_norm": 0.6247195756779031,
+      "learning_rate": 1.5835251484163728e-05,
+      "loss": 0.8203,
+      "step": 617
+    },
+    {
+      "epoch": 0.9697445201784927,
+      "grad_norm": 0.6604799598320987,
+      "learning_rate": 1.5821474772395385e-05,
+      "loss": 0.8037,
+      "step": 618
+    },
+    {
+      "epoch": 0.9713136860687491,
+      "grad_norm": 0.6873045775586228,
+      "learning_rate": 1.5807681327318372e-05,
+      "loss": 0.7987,
+      "step": 619
+    },
+    {
+      "epoch": 0.9728828519590056,
+      "grad_norm": 0.6508798780264626,
+      "learning_rate": 1.5793871188580704e-05,
+      "loss": 0.8004,
+      "step": 620
+    },
+    {
+      "epoch": 0.974452017849262,
+      "grad_norm": 0.7422396353508237,
+      "learning_rate": 1.578004439587839e-05,
+      "loss": 0.8181,
+      "step": 621
+    },
+    {
+      "epoch": 0.9760211837395185,
+      "grad_norm": 0.7019113344967481,
+      "learning_rate": 1.5766200988955315e-05,
+      "loss": 0.7923,
+      "step": 622
+    },
+    {
+      "epoch": 0.9775903496297749,
+      "grad_norm": 0.708015835537674,
+      "learning_rate": 1.575234100760311e-05,
+      "loss": 0.8026,
+      "step": 623
+    },
+    {
+      "epoch": 0.9791595155200313,
+      "grad_norm": 0.6597727898719461,
+      "learning_rate": 1.5738464491661054e-05,
+      "loss": 0.7918,
+      "step": 624
+    },
+    {
+      "epoch": 0.9807286814102878,
+      "grad_norm": 0.7287073025117725,
+      "learning_rate": 1.5724571481015953e-05,
+      "loss": 0.7694,
+      "step": 625
+    },
+    {
+      "epoch": 0.9822978473005443,
+      "grad_norm": 0.6803212996836236,
+      "learning_rate": 1.5710662015602016e-05,
+      "loss": 0.8232,
+      "step": 626
+    },
+    {
+      "epoch": 0.9838670131908007,
+      "grad_norm": 0.6634984218475023,
+      "learning_rate": 1.5696736135400765e-05,
+      "loss": 0.7971,
+      "step": 627
+    },
+    {
+      "epoch": 0.9854361790810572,
+      "grad_norm": 0.69890043399333,
+      "learning_rate": 1.5682793880440895e-05,
+      "loss": 0.8126,
+      "step": 628
+    },
+    {
+      "epoch": 0.9870053449713136,
+      "grad_norm": 0.651169439597212,
+      "learning_rate": 1.566883529079817e-05,
+      "loss": 0.7793,
+      "step": 629
+    },
+    {
+      "epoch": 0.9885745108615701,
+      "grad_norm": 0.760055267233693,
+      "learning_rate": 1.5654860406595312e-05,
+      "loss": 0.8125,
+      "step": 630
+    },
+    {
+      "epoch": 0.9901436767518266,
+      "grad_norm": 0.7551011983188946,
+      "learning_rate": 1.5640869268001875e-05,
+      "loss": 0.8128,
+      "step": 631
+    },
+    {
+      "epoch": 0.991712842642083,
+      "grad_norm": 0.6831265112531708,
+      "learning_rate": 1.562686191523413e-05,
+      "loss": 0.7988,
+      "step": 632
+    },
+    {
+      "epoch": 0.9932820085323395,
+      "grad_norm": 0.6219245966811059,
+      "learning_rate": 1.5612838388554966e-05,
+      "loss": 0.798,
+      "step": 633
+    },
+    {
+      "epoch": 0.994851174422596,
+      "grad_norm": 0.6964076748363703,
+      "learning_rate": 1.5598798728273753e-05,
+      "loss": 0.7923,
+      "step": 634
+    },
+    {
+      "epoch": 0.9964203403128524,
+      "grad_norm": 0.770883927331421,
+      "learning_rate": 1.5584742974746242e-05,
+      "loss": 0.815,
+      "step": 635
+    },
+    {
+      "epoch": 0.9979895062031089,
+      "grad_norm": 0.6772179151384508,
+      "learning_rate": 1.557067116837444e-05,
+      "loss": 0.8173,
+      "step": 636
+    },
+    {
+      "epoch": 0.9995586720933654,
+      "grad_norm": 0.668345093008785,
+      "learning_rate": 1.5556583349606492e-05,
+      "loss": 0.7858,
+      "step": 637
+    },
+    {
+      "epoch": 1.001127837983622,
+      "grad_norm": 0.7803388353156092,
+      "learning_rate": 1.5542479558936572e-05,
+      "loss": 0.7356,
+      "step": 638
+    },
+    {
+      "epoch": 1.0026970038738783,
+      "grad_norm": 0.7541533192368013,
+      "learning_rate": 1.5528359836904774e-05,
+      "loss": 0.7531,
+      "step": 639
+    },
+    {
+      "epoch": 1.0042661697641349,
+      "grad_norm": 0.6935113060988236,
+      "learning_rate": 1.551422422409697e-05,
+      "loss": 0.7336,
+      "step": 640
+    },
+    {
+      "epoch": 1.0058353356543912,
+      "grad_norm": 0.7637416776658377,
+      "learning_rate": 1.5500072761144712e-05,
+      "loss": 0.7449,
+      "step": 641
+    },
+    {
+      "epoch": 1.0074045015446478,
+      "grad_norm": 0.748625067791586,
+      "learning_rate": 1.5485905488725126e-05,
+      "loss": 0.7301,
+      "step": 642
+    },
+    {
+      "epoch": 1.0089736674349041,
+      "grad_norm": 0.7528740869999863,
+      "learning_rate": 1.5471722447560757e-05,
+      "loss": 0.75,
+      "step": 643
+    },
+    {
+      "epoch": 1.0105428333251605,
+      "grad_norm": 0.6412858721686633,
+      "learning_rate": 1.5457523678419498e-05,
+      "loss": 0.7241,
+      "step": 644
+    },
+    {
+      "epoch": 1.012111999215417,
+      "grad_norm": 0.826256044071756,
+      "learning_rate": 1.5443309222114434e-05,
+      "loss": 0.756,
+      "step": 645
+    },
+    {
+      "epoch": 1.0136811651056734,
+      "grad_norm": 0.7145799807315482,
+      "learning_rate": 1.5429079119503752e-05,
+      "loss": 0.7259,
+      "step": 646
+    },
+    {
+      "epoch": 1.01525033099593,
+      "grad_norm": 0.6475158763865285,
+      "learning_rate": 1.5414833411490612e-05,
+      "loss": 0.7477,
+      "step": 647
+    },
+    {
+      "epoch": 1.0168194968861863,
+      "grad_norm": 0.7281077002979637,
+      "learning_rate": 1.5400572139023026e-05,
+      "loss": 0.7218,
+      "step": 648
+    },
+    {
+      "epoch": 1.018388662776443,
+      "grad_norm": 0.745008142531903,
+      "learning_rate": 1.5386295343093743e-05,
+      "loss": 0.7278,
+      "step": 649
+    },
+    {
+      "epoch": 1.0199578286666993,
+      "grad_norm": 0.596703405304876,
+      "learning_rate": 1.537200306474014e-05,
+      "loss": 0.7168,
+      "step": 650
+    },
+    {
+      "epoch": 1.0215269945569558,
+      "grad_norm": 0.66818392318934,
+      "learning_rate": 1.5357695345044096e-05,
+      "loss": 0.7111,
+      "step": 651
+    },
+    {
+      "epoch": 1.0230961604472122,
+      "grad_norm": 0.6890967493522583,
+      "learning_rate": 1.5343372225131863e-05,
+      "loss": 0.7301,
+      "step": 652
+    },
+    {
+      "epoch": 1.0246653263374688,
+      "grad_norm": 0.7105057085919756,
+      "learning_rate": 1.5329033746173975e-05,
+      "loss": 0.7028,
+      "step": 653
+    },
+    {
+      "epoch": 1.0262344922277251,
+      "grad_norm": 0.6490853329771772,
+      "learning_rate": 1.5314679949385105e-05,
+      "loss": 0.7267,
+      "step": 654
+    },
+    {
+      "epoch": 1.0278036581179817,
+      "grad_norm": 0.7135115488094335,
+      "learning_rate": 1.530031087602396e-05,
+      "loss": 0.7226,
+      "step": 655
+    },
+    {
+      "epoch": 1.029372824008238,
+      "grad_norm": 0.6752809293469212,
+      "learning_rate": 1.5285926567393158e-05,
+      "loss": 0.7197,
+      "step": 656
+    },
+    {
+      "epoch": 1.0309419898984946,
+      "grad_norm": 0.6330814107063447,
+      "learning_rate": 1.5271527064839107e-05,
+      "loss": 0.7489,
+      "step": 657
+    },
+    {
+      "epoch": 1.032511155788751,
+      "grad_norm": 0.6878492533579073,
+      "learning_rate": 1.5257112409751886e-05,
+      "loss": 0.7377,
+      "step": 658
+    },
+    {
+      "epoch": 1.0340803216790075,
+      "grad_norm": 0.6674789984104226,
+      "learning_rate": 1.5242682643565135e-05,
+      "loss": 0.7208,
+      "step": 659
+    },
+    {
+      "epoch": 1.035649487569264,
+      "grad_norm": 0.657117739008243,
+      "learning_rate": 1.5228237807755925e-05,
+      "loss": 0.7303,
+      "step": 660
+    },
+    {
+      "epoch": 1.0372186534595205,
+      "grad_norm": 0.6767574155048245,
+      "learning_rate": 1.5213777943844648e-05,
+      "loss": 0.7407,
+      "step": 661
+    },
+    {
+      "epoch": 1.0387878193497768,
+      "grad_norm": 0.6912916163166987,
+      "learning_rate": 1.5199303093394883e-05,
+      "loss": 0.727,
+      "step": 662
+    },
+    {
+      "epoch": 1.0403569852400334,
+      "grad_norm": 0.6699302597943366,
+      "learning_rate": 1.5184813298013304e-05,
+      "loss": 0.7347,
+      "step": 663
+    },
+    {
+      "epoch": 1.0419261511302897,
+      "grad_norm": 0.6620724775997375,
+      "learning_rate": 1.5170308599349524e-05,
+      "loss": 0.7454,
+      "step": 664
+    },
+    {
+      "epoch": 1.0434953170205463,
+      "grad_norm": 0.6846086093807906,
+      "learning_rate": 1.5155789039096005e-05,
+      "loss": 0.7305,
+      "step": 665
+    },
+    {
+      "epoch": 1.0450644829108027,
+      "grad_norm": 0.6974910569549667,
+      "learning_rate": 1.5141254658987922e-05,
+      "loss": 0.7182,
+      "step": 666
+    },
+    {
+      "epoch": 1.0466336488010592,
+      "grad_norm": 0.6761012200383468,
+      "learning_rate": 1.5126705500803055e-05,
+      "loss": 0.7431,
+      "step": 667
+    },
+    {
+      "epoch": 1.0482028146913156,
+      "grad_norm": 0.7116609578466914,
+      "learning_rate": 1.5112141606361657e-05,
+      "loss": 0.7052,
+      "step": 668
+    },
+    {
+      "epoch": 1.0497719805815722,
+      "grad_norm": 0.6486505551529763,
+      "learning_rate": 1.5097563017526336e-05,
+      "loss": 0.7362,
+      "step": 669
+    },
+    {
+      "epoch": 1.0513411464718285,
+      "grad_norm": 0.7280159021660275,
+      "learning_rate": 1.5082969776201948e-05,
+      "loss": 0.7384,
+      "step": 670
+    },
+    {
+      "epoch": 1.052910312362085,
+      "grad_norm": 0.5974562161059254,
+      "learning_rate": 1.5068361924335457e-05,
+      "loss": 0.7164,
+      "step": 671
+    },
+    {
+      "epoch": 1.0544794782523415,
+      "grad_norm": 0.654157948660559,
+      "learning_rate": 1.505373950391583e-05,
+      "loss": 0.7062,
+      "step": 672
+    },
+    {
+      "epoch": 1.056048644142598,
+      "grad_norm": 0.7288734094958,
+      "learning_rate": 1.5039102556973899e-05,
+      "loss": 0.7237,
+      "step": 673
+    },
+    {
+      "epoch": 1.0576178100328544,
+      "grad_norm": 0.7272863503548945,
+      "learning_rate": 1.5024451125582271e-05,
+      "loss": 0.7189,
+      "step": 674
+    },
+    {
+      "epoch": 1.059186975923111,
+      "grad_norm": 0.6597557887721957,
+      "learning_rate": 1.5009785251855164e-05,
+      "loss": 0.7231,
+      "step": 675
+    },
+    {
+      "epoch": 1.0607561418133673,
+      "grad_norm": 0.6888057732787886,
+      "learning_rate": 1.4995104977948331e-05,
+      "loss": 0.7448,
+      "step": 676
+    },
+    {
+      "epoch": 1.0623253077036239,
+      "grad_norm": 0.7517206654601427,
+      "learning_rate": 1.49804103460589e-05,
+      "loss": 0.7238,
+      "step": 677
+    },
+    {
+      "epoch": 1.0638944735938802,
+      "grad_norm": 0.7536335190220884,
+      "learning_rate": 1.4965701398425275e-05,
+      "loss": 0.7255,
+      "step": 678
+    },
+    {
+      "epoch": 1.0654636394841368,
+      "grad_norm": 0.7202992167808299,
+      "learning_rate": 1.4950978177327016e-05,
+      "loss": 0.722,
+      "step": 679
+    },
+    {
+      "epoch": 1.0670328053743932,
+      "grad_norm": 0.723904254278952,
+      "learning_rate": 1.4936240725084701e-05,
+      "loss": 0.7354,
+      "step": 680
+    },
+    {
+      "epoch": 1.0686019712646497,
+      "grad_norm": 0.7439928284873452,
+      "learning_rate": 1.4921489084059823e-05,
+      "loss": 0.7129,
+      "step": 681
+    },
+    {
+      "epoch": 1.070171137154906,
+      "grad_norm": 0.6888363367701373,
+      "learning_rate": 1.4906723296654652e-05,
+      "loss": 0.7384,
+      "step": 682
+    },
+    {
+      "epoch": 1.0717403030451627,
+      "grad_norm": 0.6967227398393554,
+      "learning_rate": 1.4891943405312126e-05,
+      "loss": 0.7353,
+      "step": 683
+    },
+    {
+      "epoch": 1.073309468935419,
+      "grad_norm": 0.794666467584343,
+      "learning_rate": 1.4877149452515718e-05,
+      "loss": 0.7132,
+      "step": 684
+    },
+    {
+      "epoch": 1.0748786348256756,
+      "grad_norm": 0.6602649754123938,
+      "learning_rate": 1.4862341480789322e-05,
+      "loss": 0.7146,
+      "step": 685
+    },
+    {
+      "epoch": 1.076447800715932,
+      "grad_norm": 0.8894682823538541,
+      "learning_rate": 1.484751953269713e-05,
+      "loss": 0.7375,
+      "step": 686
+    },
+    {
+      "epoch": 1.0780169666061883,
+      "grad_norm": 0.7585247138160331,
+      "learning_rate": 1.483268365084351e-05,
+      "loss": 0.7157,
+      "step": 687
+    },
+    {
+      "epoch": 1.0795861324964449,
+      "grad_norm": 0.7117654864119813,
+      "learning_rate": 1.481783387787287e-05,
+      "loss": 0.7052,
+      "step": 688
+    },
+    {
+      "epoch": 1.0811552983867014,
+      "grad_norm": 0.7344324037870501,
+      "learning_rate": 1.4802970256469564e-05,
+      "loss": 0.7241,
+      "step": 689
+    },
+    {
+      "epoch": 1.0827244642769578,
+      "grad_norm": 0.8246376521345272,
+      "learning_rate": 1.4788092829357736e-05,
+      "loss": 0.7255,
+      "step": 690
+    },
+    {
+      "epoch": 1.0842936301672141,
+      "grad_norm": 0.7457152420569799,
+      "learning_rate": 1.4773201639301223e-05,
+      "loss": 0.7239,
+      "step": 691
+    },
+    {
+      "epoch": 1.0858627960574707,
+      "grad_norm": 0.6845725156795525,
+      "learning_rate": 1.4758296729103413e-05,
+      "loss": 0.7202,
+      "step": 692
+    },
+    {
+      "epoch": 1.087431961947727,
+      "grad_norm": 0.6635657637485262,
+      "learning_rate": 1.4743378141607146e-05,
+      "loss": 0.7221,
+      "step": 693
+    },
+    {
+      "epoch": 1.0890011278379836,
+      "grad_norm": 0.7144072518590094,
+      "learning_rate": 1.4728445919694563e-05,
+      "loss": 0.7106,
+      "step": 694
+    },
+    {
+      "epoch": 1.09057029372824,
+      "grad_norm": 0.6914091269695335,
+      "learning_rate": 1.4713500106287e-05,
+      "loss": 0.7142,
+      "step": 695
+    },
+    {
+      "epoch": 1.0921394596184966,
+      "grad_norm": 0.6442449050355384,
+      "learning_rate": 1.4698540744344864e-05,
+      "loss": 0.7291,
+      "step": 696
+    },
+    {
+      "epoch": 1.093708625508753,
+      "grad_norm": 0.7010092912473106,
+      "learning_rate": 1.4683567876867503e-05,
+      "loss": 0.7278,
+      "step": 697
+    },
+    {
+      "epoch": 1.0952777913990095,
+      "grad_norm": 0.6958420274973538,
+      "learning_rate": 1.4668581546893086e-05,
+      "loss": 0.7157,
+      "step": 698
+    },
+    {
+      "epoch": 1.0968469572892658,
+      "grad_norm": 0.7332870445914398,
+      "learning_rate": 1.4653581797498477e-05,
+      "loss": 0.7189,
+      "step": 699
+    },
+    {
+      "epoch": 1.0984161231795224,
+      "grad_norm": 0.6665929969073299,
+      "learning_rate": 1.4638568671799113e-05,
+      "loss": 0.7281,
+      "step": 700
+    },
+    {
+      "epoch": 1.0999852890697788,
+      "grad_norm": 0.6833765726881289,
+      "learning_rate": 1.4623542212948887e-05,
+      "loss": 0.7269,
+      "step": 701
+    },
+    {
+      "epoch": 1.1015544549600353,
+      "grad_norm": 0.7928659779727998,
+      "learning_rate": 1.4608502464140002e-05,
+      "loss": 0.7094,
+      "step": 702
+    },
+    {
+      "epoch": 1.1031236208502917,
+      "grad_norm": 0.6567329845650816,
+      "learning_rate": 1.4593449468602878e-05,
+      "loss": 0.7345,
+      "step": 703
+    },
+    {
+      "epoch": 1.1046927867405483,
+      "grad_norm": 0.6578233587658882,
+      "learning_rate": 1.4578383269606004e-05,
+      "loss": 0.7194,
+      "step": 704
+    },
+    {
+      "epoch": 1.1062619526308046,
+      "grad_norm": 0.678695684331327,
+      "learning_rate": 1.4563303910455821e-05,
+      "loss": 0.7255,
+      "step": 705
+    },
+    {
+      "epoch": 1.1078311185210612,
+      "grad_norm": 0.6779685173206671,
+      "learning_rate": 1.4548211434496598e-05,
+      "loss": 0.722,
+      "step": 706
+    },
+    {
+      "epoch": 1.1094002844113175,
+      "grad_norm": 0.6314744333887462,
+      "learning_rate": 1.4533105885110307e-05,
+      "loss": 0.7216,
+      "step": 707
+    },
+    {
+      "epoch": 1.1109694503015741,
+      "grad_norm": 0.6502234664163404,
+      "learning_rate": 1.4517987305716502e-05,
+      "loss": 0.7073,
+      "step": 708
+    },
+    {
+      "epoch": 1.1125386161918305,
+      "grad_norm": 0.7175775943504926,
+      "learning_rate": 1.4502855739772184e-05,
+      "loss": 0.7144,
+      "step": 709
+    },
+    {
+      "epoch": 1.114107782082087,
+      "grad_norm": 0.5783185520260196,
+      "learning_rate": 1.4487711230771686e-05,
+      "loss": 0.7234,
+      "step": 710
+    },
+    {
+      "epoch": 1.1156769479723434,
+      "grad_norm": 0.6231893108916796,
+      "learning_rate": 1.4472553822246548e-05,
+      "loss": 0.7153,
+      "step": 711
+    },
+    {
+      "epoch": 1.1172461138626,
+      "grad_norm": 0.632956161157449,
+      "learning_rate": 1.4457383557765385e-05,
+      "loss": 0.7358,
+      "step": 712
+    },
+    {
+      "epoch": 1.1188152797528563,
+      "grad_norm": 0.6553940813975944,
+      "learning_rate": 1.4442200480933761e-05,
+      "loss": 0.7216,
+      "step": 713
+    },
+    {
+      "epoch": 1.120384445643113,
+      "grad_norm": 0.6897533234088822,
+      "learning_rate": 1.442700463539408e-05,
+      "loss": 0.7311,
+      "step": 714
+    },
+    {
+      "epoch": 1.1219536115333693,
+      "grad_norm": 0.652847984685953,
+      "learning_rate": 1.4411796064825436e-05,
+      "loss": 0.7292,
+      "step": 715
+    },
+    {
+      "epoch": 1.1235227774236258,
+      "grad_norm": 0.6700420140424015,
+      "learning_rate": 1.439657481294351e-05,
+      "loss": 0.7303,
+      "step": 716
+    },
+    {
+      "epoch": 1.1250919433138822,
+      "grad_norm": 0.6808545629120182,
+      "learning_rate": 1.4381340923500422e-05,
+      "loss": 0.7236,
+      "step": 717
+    },
+    {
+      "epoch": 1.1266611092041388,
+      "grad_norm": 0.6536634238814496,
+      "learning_rate": 1.4366094440284632e-05,
+      "loss": 0.717,
+      "step": 718
+    },
+    {
+      "epoch": 1.128230275094395,
+      "grad_norm": 0.6088038345099579,
+      "learning_rate": 1.435083540712079e-05,
+      "loss": 0.7274,
+      "step": 719
+    },
+    {
+      "epoch": 1.1297994409846517,
+      "grad_norm": 0.6884030984569609,
+      "learning_rate": 1.4335563867869618e-05,
+      "loss": 0.7274,
+      "step": 720
+    },
+    {
+      "epoch": 1.131368606874908,
+      "grad_norm": 0.6862114045899801,
+      "learning_rate": 1.4320279866427798e-05,
+      "loss": 0.7402,
+      "step": 721
+    },
+    {
+      "epoch": 1.1329377727651646,
+      "grad_norm": 0.663924831018658,
+      "learning_rate": 1.430498344672782e-05,
+      "loss": 0.7245,
+      "step": 722
+    },
+    {
+      "epoch": 1.134506938655421,
+      "grad_norm": 0.7009174871348716,
+      "learning_rate": 1.4289674652737877e-05,
+      "loss": 0.7195,
+      "step": 723
+    },
+    {
+      "epoch": 1.1360761045456775,
+      "grad_norm": 0.6401414921019333,
+      "learning_rate": 1.4274353528461728e-05,
+      "loss": 0.7251,
+      "step": 724
+    },
+    {
+      "epoch": 1.1376452704359339,
+      "grad_norm": 0.6623338594853233,
+      "learning_rate": 1.4259020117938574e-05,
+      "loss": 0.7143,
+      "step": 725
+    },
+    {
+      "epoch": 1.1392144363261902,
+      "grad_norm": 0.7038453216768193,
+      "learning_rate": 1.4243674465242934e-05,
+      "loss": 0.7413,
+      "step": 726
+    },
+    {
+      "epoch": 1.1407836022164468,
+      "grad_norm": 0.6001545347663756,
+      "learning_rate": 1.4228316614484511e-05,
+      "loss": 0.7197,
+      "step": 727
+    },
+    {
+      "epoch": 1.1423527681067034,
+      "grad_norm": 0.6238452842054542,
+      "learning_rate": 1.4212946609808079e-05,
+      "loss": 0.7233,
+      "step": 728
+    },
+    {
+      "epoch": 1.1439219339969597,
+      "grad_norm": 0.6693838531750909,
+      "learning_rate": 1.4197564495393334e-05,
+      "loss": 0.7203,
+      "step": 729
+    },
+    {
+      "epoch": 1.145491099887216,
+      "grad_norm": 0.6342589745019122,
+      "learning_rate": 1.4182170315454798e-05,
+      "loss": 0.7326,
+      "step": 730
+    },
+    {
+      "epoch": 1.1470602657774727,
+      "grad_norm": 0.812157778897247,
+      "learning_rate": 1.4166764114241657e-05,
+      "loss": 0.7144,
+      "step": 731
+    },
+    {
+      "epoch": 1.1486294316677292,
+      "grad_norm": 0.6446321313076545,
+      "learning_rate": 1.4151345936037664e-05,
+      "loss": 0.7132,
+      "step": 732
+    },
+    {
+      "epoch": 1.1501985975579856,
+      "grad_norm": 0.6147445437009036,
+      "learning_rate": 1.413591582516099e-05,
+      "loss": 0.7384,
+      "step": 733
+    },
+    {
+      "epoch": 1.151767763448242,
+      "grad_norm": 0.6342181435018887,
+      "learning_rate": 1.4120473825964112e-05,
+      "loss": 0.737,
+      "step": 734
+    },
+    {
+      "epoch": 1.1533369293384985,
+      "grad_norm": 0.6587877437587489,
+      "learning_rate": 1.4105019982833674e-05,
+      "loss": 0.7151,
+      "step": 735
+    },
+    {
+      "epoch": 1.1549060952287549,
+      "grad_norm": 0.6537503190247131,
+      "learning_rate": 1.4089554340190365e-05,
+      "loss": 0.725,
+      "step": 736
+    },
+    {
+      "epoch": 1.1564752611190114,
+      "grad_norm": 0.6142318456564848,
+      "learning_rate": 1.4074076942488794e-05,
+      "loss": 0.7119,
+      "step": 737
+    },
+    {
+      "epoch": 1.1580444270092678,
+      "grad_norm": 0.6378727308612256,
+      "learning_rate": 1.4058587834217356e-05,
+      "loss": 0.7139,
+      "step": 738
+    },
+    {
+      "epoch": 1.1596135928995244,
+      "grad_norm": 0.6808954236194737,
+      "learning_rate": 1.404308705989811e-05,
+      "loss": 0.7317,
+      "step": 739
+    },
+    {
+      "epoch": 1.1611827587897807,
+      "grad_norm": 0.6202247666014872,
+      "learning_rate": 1.4027574664086641e-05,
+      "loss": 0.7253,
+      "step": 740
+    },
+    {
+      "epoch": 1.1627519246800373,
+      "grad_norm": 0.6120356659088082,
+      "learning_rate": 1.4012050691371948e-05,
+      "loss": 0.7209,
+      "step": 741
+    },
+    {
+      "epoch": 1.1643210905702936,
+      "grad_norm": 0.699382990304684,
+      "learning_rate": 1.3996515186376298e-05,
+      "loss": 0.7004,
+      "step": 742
+    },
+    {
+      "epoch": 1.1658902564605502,
+      "grad_norm": 0.6534250771619439,
+      "learning_rate": 1.3980968193755115e-05,
+      "loss": 0.723,
+      "step": 743
+    },
+    {
+      "epoch": 1.1674594223508066,
+      "grad_norm": 0.6360644274914368,
+      "learning_rate": 1.3965409758196838e-05,
+      "loss": 0.7123,
+      "step": 744
+    },
+    {
+      "epoch": 1.1690285882410631,
+      "grad_norm": 0.7111862414760431,
+      "learning_rate": 1.3949839924422798e-05,
+      "loss": 0.7189,
+      "step": 745
+    },
+    {
+      "epoch": 1.1705977541313195,
+      "grad_norm": 0.8081990423548935,
+      "learning_rate": 1.3934258737187087e-05,
+      "loss": 0.7223,
+      "step": 746
+    },
+    {
+      "epoch": 1.172166920021576,
+      "grad_norm": 0.6891489914761177,
+      "learning_rate": 1.3918666241276442e-05,
+      "loss": 0.7118,
+      "step": 747
+    },
+    {
+      "epoch": 1.1737360859118324,
+      "grad_norm": 0.7017431769982081,
+      "learning_rate": 1.390306248151009e-05,
+      "loss": 0.7293,
+      "step": 748
+    },
+    {
+      "epoch": 1.175305251802089,
+      "grad_norm": 0.8561422174851111,
+      "learning_rate": 1.3887447502739649e-05,
+      "loss": 0.733,
+      "step": 749
+    },
+    {
+      "epoch": 1.1768744176923454,
+      "grad_norm": 0.7801804711838448,
+      "learning_rate": 1.3871821349848977e-05,
+      "loss": 0.7166,
+      "step": 750
+    },
+    {
+      "epoch": 1.178443583582602,
+      "grad_norm": 0.6844665558684861,
+      "learning_rate": 1.3856184067754053e-05,
+      "loss": 0.7311,
+      "step": 751
+    },
+    {
+      "epoch": 1.1800127494728583,
+      "grad_norm": 0.6838754450457964,
+      "learning_rate": 1.3840535701402847e-05,
+      "loss": 0.725,
+      "step": 752
+    },
+    {
+      "epoch": 1.1815819153631149,
+      "grad_norm": 0.8355096684243345,
+      "learning_rate": 1.3824876295775187e-05,
+      "loss": 0.7054,
+      "step": 753
+    },
+    {
+      "epoch": 1.1831510812533712,
+      "grad_norm": 0.6994375840647716,
+      "learning_rate": 1.3809205895882635e-05,
+      "loss": 0.7439,
+      "step": 754
+    },
+    {
+      "epoch": 1.1847202471436278,
+      "grad_norm": 0.625280968043297,
+      "learning_rate": 1.3793524546768358e-05,
+      "loss": 0.7363,
+      "step": 755
+    },
+    {
+      "epoch": 1.1862894130338841,
+      "grad_norm": 0.7952663185665346,
+      "learning_rate": 1.3777832293506985e-05,
+      "loss": 0.7299,
+      "step": 756
+    },
+    {
+      "epoch": 1.1878585789241407,
+      "grad_norm": 0.7136619606631258,
+      "learning_rate": 1.3762129181204502e-05,
+      "loss": 0.7178,
+      "step": 757
+    },
+    {
+      "epoch": 1.189427744814397,
+      "grad_norm": 0.7151037409561691,
+      "learning_rate": 1.3746415254998095e-05,
+      "loss": 0.7234,
+      "step": 758
+    },
+    {
+      "epoch": 1.1909969107046536,
+      "grad_norm": 0.7534371098019876,
+      "learning_rate": 1.3730690560056043e-05,
+      "loss": 0.7358,
+      "step": 759
+    },
+    {
+      "epoch": 1.19256607659491,
+      "grad_norm": 0.6561493218889439,
+      "learning_rate": 1.3714955141577573e-05,
+      "loss": 0.7012,
+      "step": 760
+    },
+    {
+      "epoch": 1.1941352424851666,
+      "grad_norm": 0.6467586929020817,
+      "learning_rate": 1.3699209044792743e-05,
+      "loss": 0.7178,
+      "step": 761
+    },
+    {
+      "epoch": 1.195704408375423,
+      "grad_norm": 0.7512650743410423,
+      "learning_rate": 1.3683452314962295e-05,
+      "loss": 0.7178,
+      "step": 762
+    },
+    {
+      "epoch": 1.1972735742656795,
+      "grad_norm": 0.6525744254059402,
+      "learning_rate": 1.3667684997377542e-05,
+      "loss": 0.6999,
+      "step": 763
+    },
+    {
+      "epoch": 1.1988427401559358,
+      "grad_norm": 0.6412166724199153,
+      "learning_rate": 1.365190713736023e-05,
+      "loss": 0.7277,
+      "step": 764
+    },
+    {
+      "epoch": 1.2004119060461924,
+      "grad_norm": 0.6524839068324342,
+      "learning_rate": 1.3636118780262405e-05,
+      "loss": 0.7145,
+      "step": 765
+    },
+    {
+      "epoch": 1.2019810719364488,
+      "grad_norm": 0.6941780975807582,
+      "learning_rate": 1.3620319971466288e-05,
+      "loss": 0.6978,
+      "step": 766
+    },
+    {
+      "epoch": 1.2035502378267053,
+      "grad_norm": 0.6528914685779267,
+      "learning_rate": 1.3604510756384141e-05,
+      "loss": 0.726,
+      "step": 767
+    },
+    {
+      "epoch": 1.2051194037169617,
+      "grad_norm": 0.6728320802721812,
+      "learning_rate": 1.3588691180458143e-05,
+      "loss": 0.7346,
+      "step": 768
+    },
+    {
+      "epoch": 1.206688569607218,
+      "grad_norm": 0.6796777970406698,
+      "learning_rate": 1.3572861289160245e-05,
+      "loss": 0.7067,
+      "step": 769
+    },
+    {
+      "epoch": 1.2082577354974746,
+      "grad_norm": 0.633779910604212,
+      "learning_rate": 1.3557021127992055e-05,
+      "loss": 0.7008,
+      "step": 770
+    },
+    {
+      "epoch": 1.2098269013877312,
+      "grad_norm": 0.6428649575602099,
+      "learning_rate": 1.3541170742484705e-05,
+      "loss": 0.7212,
+      "step": 771
+    },
+    {
+      "epoch": 1.2113960672779875,
+      "grad_norm": 0.6547919283959451,
+      "learning_rate": 1.3525310178198707e-05,
+      "loss": 0.7426,
+      "step": 772
+    },
+    {
+      "epoch": 1.212965233168244,
+      "grad_norm": 0.6758542651299493,
+      "learning_rate": 1.3509439480723835e-05,
+      "loss": 0.7277,
+      "step": 773
+    },
+    {
+      "epoch": 1.2145343990585005,
+      "grad_norm": 0.6632989746423631,
+      "learning_rate": 1.3493558695678992e-05,
+      "loss": 0.7351,
+      "step": 774
+    },
+    {
+      "epoch": 1.216103564948757,
+      "grad_norm": 0.6737948427199923,
+      "learning_rate": 1.347766786871207e-05,
+      "loss": 0.7132,
+      "step": 775
+    },
+    {
+      "epoch": 1.2176727308390134,
+      "grad_norm": 0.613664926888893,
+      "learning_rate": 1.3461767045499834e-05,
+      "loss": 0.7085,
+      "step": 776
+    },
+    {
+      "epoch": 1.2192418967292697,
+      "grad_norm": 0.7189106691619672,
+      "learning_rate": 1.3445856271747777e-05,
+      "loss": 0.7202,
+      "step": 777
+    },
+    {
+      "epoch": 1.2208110626195263,
+      "grad_norm": 0.6970045979027315,
+      "learning_rate": 1.3429935593189996e-05,
+      "loss": 0.7364,
+      "step": 778
+    },
+    {
+      "epoch": 1.2223802285097827,
+      "grad_norm": 0.6501064907259375,
+      "learning_rate": 1.3414005055589057e-05,
+      "loss": 0.7335,
+      "step": 779
+    },
+    {
+      "epoch": 1.2239493944000392,
+      "grad_norm": 0.7288492056653614,
+      "learning_rate": 1.3398064704735862e-05,
+      "loss": 0.7472,
+      "step": 780
+    },
+    {
+      "epoch": 1.2255185602902956,
+      "grad_norm": 0.63512445384459,
+      "learning_rate": 1.3382114586449533e-05,
+      "loss": 0.7205,
+      "step": 781
+    },
+    {
+      "epoch": 1.2270877261805522,
+      "grad_norm": 0.6802757181095903,
+      "learning_rate": 1.3366154746577254e-05,
+      "loss": 0.7087,
+      "step": 782
+    },
+    {
+      "epoch": 1.2286568920708085,
+      "grad_norm": 0.6614354443478233,
+      "learning_rate": 1.3350185230994157e-05,
+      "loss": 0.7131,
+      "step": 783
+    },
+    {
+      "epoch": 1.230226057961065,
+      "grad_norm": 0.6961912892855205,
+      "learning_rate": 1.3334206085603186e-05,
+      "loss": 0.7344,
+      "step": 784
+    },
+    {
+      "epoch": 1.2317952238513215,
+      "grad_norm": 0.628800560490552,
+      "learning_rate": 1.3318217356334967e-05,
+      "loss": 0.7327,
+      "step": 785
+    },
+    {
+      "epoch": 1.233364389741578,
+      "grad_norm": 0.6672480947779101,
+      "learning_rate": 1.3302219089147675e-05,
+      "loss": 0.7006,
+      "step": 786
+    },
+    {
+      "epoch": 1.2349335556318344,
+      "grad_norm": 0.6448005745755109,
+      "learning_rate": 1.3286211330026894e-05,
+      "loss": 0.7056,
+      "step": 787
+    },
+    {
+      "epoch": 1.236502721522091,
+      "grad_norm": 0.6765035508388694,
+      "learning_rate": 1.3270194124985499e-05,
+      "loss": 0.7042,
+      "step": 788
+    },
+    {
+      "epoch": 1.2380718874123473,
+      "grad_norm": 0.647612951228948,
+      "learning_rate": 1.325416752006351e-05,
+      "loss": 0.7176,
+      "step": 789
+    },
+    {
+      "epoch": 1.2396410533026039,
+      "grad_norm": 0.6339962885133088,
+      "learning_rate": 1.3238131561327976e-05,
+      "loss": 0.7378,
+      "step": 790
+    },
+    {
+      "epoch": 1.2412102191928602,
+      "grad_norm": 0.6286832535598963,
+      "learning_rate": 1.3222086294872818e-05,
+      "loss": 0.6975,
+      "step": 791
+    },
+    {
+      "epoch": 1.2427793850831168,
+      "grad_norm": 0.6162134218155843,
+      "learning_rate": 1.320603176681873e-05,
+      "loss": 0.703,
+      "step": 792
+    },
+    {
+      "epoch": 1.2443485509733732,
+      "grad_norm": 0.6364009286271917,
+      "learning_rate": 1.318996802331301e-05,
+      "loss": 0.7414,
+      "step": 793
+    },
+    {
+      "epoch": 1.2459177168636297,
+      "grad_norm": 0.6171774323053136,
+      "learning_rate": 1.3173895110529454e-05,
+      "loss": 0.7154,
+      "step": 794
+    },
+    {
+      "epoch": 1.247486882753886,
+      "grad_norm": 0.6681351003352828,
+      "learning_rate": 1.3157813074668213e-05,
+      "loss": 0.7326,
+      "step": 795
+    },
+    {
+      "epoch": 1.2490560486441427,
+      "grad_norm": 0.6922564812364559,
+      "learning_rate": 1.3141721961955658e-05,
+      "loss": 0.7361,
+      "step": 796
+    },
+    {
+      "epoch": 1.250625214534399,
+      "grad_norm": 0.6245513089701152,
+      "learning_rate": 1.3125621818644258e-05,
+      "loss": 0.7005,
+      "step": 797
+    },
+    {
+      "epoch": 1.2521943804246556,
+      "grad_norm": 0.6756554219192724,
+      "learning_rate": 1.3109512691012434e-05,
+      "loss": 0.7336,
+      "step": 798
+    },
+    {
+      "epoch": 1.253763546314912,
+      "grad_norm": 0.6316842690348985,
+      "learning_rate": 1.3093394625364433e-05,
+      "loss": 0.7289,
+      "step": 799
+    },
+    {
+      "epoch": 1.2553327122051685,
+      "grad_norm": 0.6594970458642331,
+      "learning_rate": 1.3077267668030194e-05,
+      "loss": 0.735,
+      "step": 800
+    },
+    {
+      "epoch": 1.2569018780954249,
+      "grad_norm": 0.6774121437546732,
+      "learning_rate": 1.3061131865365211e-05,
+      "loss": 0.7332,
+      "step": 801
+    },
+    {
+      "epoch": 1.2584710439856814,
+      "grad_norm": 0.672321809942326,
+      "learning_rate": 1.3044987263750409e-05,
+      "loss": 0.7121,
+      "step": 802
+    },
+    {
+      "epoch": 1.2600402098759378,
+      "grad_norm": 0.6527618468069383,
+      "learning_rate": 1.3028833909591999e-05,
+      "loss": 0.7213,
+      "step": 803
+    },
+    {
+      "epoch": 1.2616093757661944,
+      "grad_norm": 0.6822686695702428,
+      "learning_rate": 1.3012671849321351e-05,
+      "loss": 0.7206,
+      "step": 804
+    },
+    {
+      "epoch": 1.2631785416564507,
+      "grad_norm": 0.7522933567335175,
+      "learning_rate": 1.2996501129394872e-05,
+      "loss": 0.7433,
+      "step": 805
+    },
+    {
+      "epoch": 1.2647477075467073,
+      "grad_norm": 0.7042896031448598,
+      "learning_rate": 1.2980321796293838e-05,
+      "loss": 0.7248,
+      "step": 806
+    },
+    {
+      "epoch": 1.2663168734369636,
+      "grad_norm": 0.629881001262114,
+      "learning_rate": 1.2964133896524305e-05,
+      "loss": 0.7209,
+      "step": 807
+    },
+    {
+      "epoch": 1.26788603932722,
+      "grad_norm": 0.7250615513612696,
+      "learning_rate": 1.2947937476616938e-05,
+      "loss": 0.7348,
+      "step": 808
+    },
+    {
+      "epoch": 1.2694552052174766,
+      "grad_norm": 0.5960591541052541,
+      "learning_rate": 1.29317325831269e-05,
+      "loss": 0.7036,
+      "step": 809
+    },
+    {
+      "epoch": 1.2710243711077331,
+      "grad_norm": 0.6460594144376343,
+      "learning_rate": 1.2915519262633704e-05,
+      "loss": 0.7256,
+      "step": 810
+    },
+    {
+      "epoch": 1.2725935369979895,
+      "grad_norm": 0.6504432261618547,
+      "learning_rate": 1.2899297561741097e-05,
+      "loss": 0.7315,
+      "step": 811
+    },
+    {
+      "epoch": 1.2741627028882458,
+      "grad_norm": 0.6589905476272644,
+      "learning_rate": 1.2883067527076904e-05,
+      "loss": 0.7374,
+      "step": 812
+    },
+    {
+      "epoch": 1.2757318687785024,
+      "grad_norm": 0.652258833483685,
+      "learning_rate": 1.286682920529291e-05,
+      "loss": 0.6959,
+      "step": 813
+    },
+    {
+      "epoch": 1.277301034668759,
+      "grad_norm": 0.6376618945006635,
+      "learning_rate": 1.2850582643064713e-05,
+      "loss": 0.7477,
+      "step": 814
+    },
+    {
+      "epoch": 1.2788702005590153,
+      "grad_norm": 0.6288339391724042,
+      "learning_rate": 1.2834327887091611e-05,
+      "loss": 0.723,
+      "step": 815
+    },
+    {
+      "epoch": 1.2804393664492717,
+      "grad_norm": 0.6841754013444288,
+      "learning_rate": 1.2818064984096443e-05,
+      "loss": 0.734,
+      "step": 816
+    },
+    {
+      "epoch": 1.2820085323395283,
+      "grad_norm": 0.665660129580248,
+      "learning_rate": 1.280179398082547e-05,
+      "loss": 0.7168,
+      "step": 817
+    },
+    {
+      "epoch": 1.2835776982297848,
+      "grad_norm": 0.6379562045996827,
+      "learning_rate": 1.2785514924048235e-05,
+      "loss": 0.7134,
+      "step": 818
+    },
+    {
+      "epoch": 1.2851468641200412,
+      "grad_norm": 0.7173502137871494,
+      "learning_rate": 1.2769227860557431e-05,
+      "loss": 0.7375,
+      "step": 819
+    },
+    {
+      "epoch": 1.2867160300102976,
+      "grad_norm": 0.6323317821606141,
+      "learning_rate": 1.2752932837168768e-05,
+      "loss": 0.7129,
+      "step": 820
+    },
+    {
+      "epoch": 1.2882851959005541,
+      "grad_norm": 0.7095835462650433,
+      "learning_rate": 1.2736629900720832e-05,
+      "loss": 0.7189,
+      "step": 821
+    },
+    {
+      "epoch": 1.2898543617908107,
+      "grad_norm": 0.7039545759984338,
+      "learning_rate": 1.2720319098074954e-05,
+      "loss": 0.7191,
+      "step": 822
+    },
+    {
+      "epoch": 1.291423527681067,
+      "grad_norm": 0.6414711038775592,
+      "learning_rate": 1.2704000476115079e-05,
+      "loss": 0.7093,
+      "step": 823
+    },
+    {
+      "epoch": 1.2929926935713234,
+      "grad_norm": 0.6892424705413851,
+      "learning_rate": 1.268767408174763e-05,
+      "loss": 0.7222,
+      "step": 824
+    },
+    {
+      "epoch": 1.29456185946158,
+      "grad_norm": 0.6872882040827433,
+      "learning_rate": 1.2671339961901366e-05,
+      "loss": 0.691,
+      "step": 825
+    },
+    {
+      "epoch": 1.2961310253518366,
+      "grad_norm": 0.6826371199489001,
+      "learning_rate": 1.265499816352725e-05,
+      "loss": 0.7234,
+      "step": 826
+    },
+    {
+      "epoch": 1.297700191242093,
+      "grad_norm": 0.7122200020281779,
+      "learning_rate": 1.2638648733598325e-05,
+      "loss": 0.7192,
+      "step": 827
+    },
+    {
+      "epoch": 1.2992693571323493,
+      "grad_norm": 0.6604076487985651,
+      "learning_rate": 1.2622291719109559e-05,
+      "loss": 0.699,
+      "step": 828
+    },
+    {
+      "epoch": 1.3008385230226058,
+      "grad_norm": 0.6211196438803553,
+      "learning_rate": 1.2605927167077736e-05,
+      "loss": 0.7118,
+      "step": 829
+    },
+    {
+      "epoch": 1.3024076889128622,
+      "grad_norm": 0.7220367233158411,
+      "learning_rate": 1.2589555124541293e-05,
+      "loss": 0.7249,
+      "step": 830
+    },
+    {
+      "epoch": 1.3039768548031188,
+      "grad_norm": 0.6593177938665435,
+      "learning_rate": 1.25731756385602e-05,
+      "loss": 0.7245,
+      "step": 831
+    },
+    {
+      "epoch": 1.305546020693375,
+      "grad_norm": 0.7087644841287654,
+      "learning_rate": 1.255678875621583e-05,
+      "loss": 0.7469,
+      "step": 832
+    },
+    {
+      "epoch": 1.3071151865836317,
+      "grad_norm": 0.6872448238166301,
+      "learning_rate": 1.2540394524610808e-05,
+      "loss": 0.7055,
+      "step": 833
+    },
+    {
+      "epoch": 1.308684352473888,
+      "grad_norm": 0.666042649831981,
+      "learning_rate": 1.252399299086889e-05,
+      "loss": 0.7154,
+      "step": 834
+    },
+    {
+      "epoch": 1.3102535183641446,
+      "grad_norm": 0.6208951368120518,
+      "learning_rate": 1.2507584202134812e-05,
+      "loss": 0.7144,
+      "step": 835
+    },
+    {
+      "epoch": 1.311822684254401,
+      "grad_norm": 0.6673651013952011,
+      "learning_rate": 1.2491168205574175e-05,
+      "loss": 0.7128,
+      "step": 836
+    },
+    {
+      "epoch": 1.3133918501446575,
+      "grad_norm": 0.6000189443577801,
+      "learning_rate": 1.2474745048373292e-05,
+      "loss": 0.729,
+      "step": 837
+    },
+    {
+      "epoch": 1.3149610160349139,
+      "grad_norm": 0.6417302271800848,
+      "learning_rate": 1.2458314777739056e-05,
+      "loss": 0.7267,
+      "step": 838
+    },
+    {
+      "epoch": 1.3165301819251705,
+      "grad_norm": 0.6337429718201932,
+      "learning_rate": 1.2441877440898817e-05,
+      "loss": 0.7128,
+      "step": 839
+    },
+    {
+      "epoch": 1.3180993478154268,
+      "grad_norm": 0.6170874332239124,
+      "learning_rate": 1.2425433085100224e-05,
+      "loss": 0.7122,
+      "step": 840
+    },
+    {
+      "epoch": 1.3196685137056834,
+      "grad_norm": 0.678129615260254,
+      "learning_rate": 1.2408981757611112e-05,
+      "loss": 0.705,
+      "step": 841
+    },
+    {
+      "epoch": 1.3212376795959397,
+      "grad_norm": 0.6109538192264312,
+      "learning_rate": 1.2392523505719349e-05,
+      "loss": 0.7287,
+      "step": 842
+    },
+    {
+      "epoch": 1.3228068454861963,
+      "grad_norm": 0.6361898884654272,
+      "learning_rate": 1.237605837673271e-05,
+      "loss": 0.726,
+      "step": 843
+    },
+    {
+      "epoch": 1.3243760113764527,
+      "grad_norm": 0.6321164785889781,
+      "learning_rate": 1.2359586417978733e-05,
+      "loss": 0.7106,
+      "step": 844
+    },
+    {
+      "epoch": 1.3259451772667092,
+      "grad_norm": 0.689710845047404,
+      "learning_rate": 1.2343107676804593e-05,
+      "loss": 0.7328,
+      "step": 845
+    },
+    {
+      "epoch": 1.3275143431569656,
+      "grad_norm": 0.6377904917047713,
+      "learning_rate": 1.2326622200576956e-05,
+      "loss": 0.7021,
+      "step": 846
+    },
+    {
+      "epoch": 1.3290835090472222,
+      "grad_norm": 0.6342268562642017,
+      "learning_rate": 1.2310130036681847e-05,
+      "loss": 0.7269,
+      "step": 847
+    },
+    {
+      "epoch": 1.3306526749374785,
+      "grad_norm": 0.7301799069542851,
+      "learning_rate": 1.2293631232524524e-05,
+      "loss": 0.7186,
+      "step": 848
+    },
+    {
+      "epoch": 1.332221840827735,
+      "grad_norm": 0.6924306019618529,
+      "learning_rate": 1.2277125835529318e-05,
+      "loss": 0.7238,
+      "step": 849
+    },
+    {
+      "epoch": 1.3337910067179914,
+      "grad_norm": 0.66191286060955,
+      "learning_rate": 1.2260613893139517e-05,
+      "loss": 0.7336,
+      "step": 850
+    },
+    {
+      "epoch": 1.3353601726082478,
+      "grad_norm": 0.6341200858429045,
+      "learning_rate": 1.2244095452817227e-05,
+      "loss": 0.7136,
+      "step": 851
+    },
+    {
+      "epoch": 1.3369293384985044,
+      "grad_norm": 0.6305949205529846,
+      "learning_rate": 1.2227570562043223e-05,
+      "loss": 0.6972,
+      "step": 852
+    },
+    {
+      "epoch": 1.338498504388761,
+      "grad_norm": 0.7402111928148778,
+      "learning_rate": 1.221103926831683e-05,
+      "loss": 0.7116,
+      "step": 853
+    },
+    {
+      "epoch": 1.3400676702790173,
+      "grad_norm": 0.6774097888733212,
+      "learning_rate": 1.219450161915577e-05,
+      "loss": 0.7155,
+      "step": 854
+    },
+    {
+      "epoch": 1.3416368361692736,
+      "grad_norm": 0.6928756650842154,
+      "learning_rate": 1.2177957662096038e-05,
+      "loss": 0.7438,
+      "step": 855
+    },
+    {
+      "epoch": 1.3432060020595302,
+      "grad_norm": 0.6333148904267516,
+      "learning_rate": 1.2161407444691761e-05,
+      "loss": 0.7259,
+      "step": 856
+    },
+    {
+      "epoch": 1.3447751679497868,
+      "grad_norm": 0.6716744132199812,
+      "learning_rate": 1.2144851014515055e-05,
+      "loss": 0.7165,
+      "step": 857
+    },
+    {
+      "epoch": 1.3463443338400432,
+      "grad_norm": 0.6584659508564903,
+      "learning_rate": 1.2128288419155903e-05,
+      "loss": 0.7201,
+      "step": 858
+    },
+    {
+      "epoch": 1.3479134997302995,
+      "grad_norm": 0.6367110648116499,
+      "learning_rate": 1.2111719706222e-05,
+      "loss": 0.7264,
+      "step": 859
+    },
+    {
+      "epoch": 1.349482665620556,
+      "grad_norm": 0.6911286087877214,
+      "learning_rate": 1.2095144923338633e-05,
+      "loss": 0.7066,
+      "step": 860
+    },
+    {
+      "epoch": 1.3510518315108127,
+      "grad_norm": 0.6641566696877007,
+      "learning_rate": 1.2078564118148531e-05,
+      "loss": 0.7082,
+      "step": 861
+    },
+    {
+      "epoch": 1.352620997401069,
+      "grad_norm": 0.6285900612911928,
+      "learning_rate": 1.2061977338311737e-05,
+      "loss": 0.7015,
+      "step": 862
+    },
+    {
+      "epoch": 1.3541901632913254,
+      "grad_norm": 0.6706389266328956,
+      "learning_rate": 1.2045384631505466e-05,
+      "loss": 0.701,
+      "step": 863
+    },
+    {
+      "epoch": 1.355759329181582,
+      "grad_norm": 0.716743179939992,
+      "learning_rate": 1.202878604542397e-05,
+      "loss": 0.7161,
+      "step": 864
+    },
+    {
+      "epoch": 1.3573284950718385,
+      "grad_norm": 0.7172535310656336,
+      "learning_rate": 1.20121816277784e-05,
+      "loss": 0.7423,
+      "step": 865
+    },
+    {
+      "epoch": 1.3588976609620949,
+      "grad_norm": 0.6552993900588525,
+      "learning_rate": 1.1995571426296671e-05,
+      "loss": 0.7231,
+      "step": 866
+    },
+    {
+      "epoch": 1.3604668268523512,
+      "grad_norm": 0.7122718419281147,
+      "learning_rate": 1.1978955488723323e-05,
+      "loss": 0.728,
+      "step": 867
+    },
+    {
+      "epoch": 1.3620359927426078,
+      "grad_norm": 0.7508958503348475,
+      "learning_rate": 1.1962333862819379e-05,
+      "loss": 0.7379,
+      "step": 868
+    },
+    {
+      "epoch": 1.3636051586328644,
+      "grad_norm": 0.692924199197633,
+      "learning_rate": 1.1945706596362216e-05,
+      "loss": 0.7197,
+      "step": 869
+    },
+    {
+      "epoch": 1.3651743245231207,
+      "grad_norm": 0.7066239667161673,
+      "learning_rate": 1.1929073737145432e-05,
+      "loss": 0.7069,
+      "step": 870
+    },
+    {
+      "epoch": 1.366743490413377,
+      "grad_norm": 0.6595984818284824,
+      "learning_rate": 1.1912435332978684e-05,
+      "loss": 0.7082,
+      "step": 871
+    },
+    {
+      "epoch": 1.3683126563036336,
+      "grad_norm": 0.6582476309008898,
+      "learning_rate": 1.1895791431687585e-05,
+      "loss": 0.7301,
+      "step": 872
+    },
+    {
+      "epoch": 1.36988182219389,
+      "grad_norm": 0.6255889268832004,
+      "learning_rate": 1.1879142081113535e-05,
+      "loss": 0.7078,
+      "step": 873
+    },
+    {
+      "epoch": 1.3714509880841466,
+      "grad_norm": 0.6091634211001478,
+      "learning_rate": 1.1862487329113606e-05,
+      "loss": 0.6962,
+      "step": 874
+    },
+    {
+      "epoch": 1.373020153974403,
+      "grad_norm": 0.6069056316047424,
+      "learning_rate": 1.1845827223560394e-05,
+      "loss": 0.719,
+      "step": 875
+    },
+    {
+      "epoch": 1.3745893198646595,
+      "grad_norm": 0.6581162509079973,
+      "learning_rate": 1.1829161812341883e-05,
+      "loss": 0.7136,
+      "step": 876
+    },
+    {
+      "epoch": 1.3761584857549158,
+      "grad_norm": 0.7146293984341919,
+      "learning_rate": 1.1812491143361305e-05,
+      "loss": 0.6956,
+      "step": 877
+    },
+    {
+      "epoch": 1.3777276516451724,
+      "grad_norm": 0.6398066925579683,
+      "learning_rate": 1.1795815264537008e-05,
+      "loss": 0.7096,
+      "step": 878
+    },
+    {
+      "epoch": 1.3792968175354288,
+      "grad_norm": 0.7040517341531731,
+      "learning_rate": 1.1779134223802315e-05,
+      "loss": 0.7224,
+      "step": 879
+    },
+    {
+      "epoch": 1.3808659834256853,
+      "grad_norm": 0.7262898787492248,
+      "learning_rate": 1.1762448069105387e-05,
+      "loss": 0.7184,
+      "step": 880
+    },
+    {
+      "epoch": 1.3824351493159417,
+      "grad_norm": 0.6169288405090357,
+      "learning_rate": 1.1745756848409083e-05,
+      "loss": 0.7163,
+      "step": 881
+    },
+    {
+      "epoch": 1.3840043152061983,
+      "grad_norm": 0.6489901700081867,
+      "learning_rate": 1.1729060609690824e-05,
+      "loss": 0.7356,
+      "step": 882
+    },
+    {
+      "epoch": 1.3855734810964546,
+      "grad_norm": 0.7332798261806479,
+      "learning_rate": 1.171235940094245e-05,
+      "loss": 0.7118,
+      "step": 883
+    },
+    {
+      "epoch": 1.3871426469867112,
+      "grad_norm": 0.7068124427937897,
+      "learning_rate": 1.1695653270170102e-05,
+      "loss": 0.7357,
+      "step": 884
+    },
+    {
+      "epoch": 1.3887118128769675,
+      "grad_norm": 0.6551653615753867,
+      "learning_rate": 1.1678942265394055e-05,
+      "loss": 0.7343,
+      "step": 885
+    },
+    {
+      "epoch": 1.3902809787672241,
+      "grad_norm": 0.6492077536437721,
+      "learning_rate": 1.1662226434648595e-05,
+      "loss": 0.7163,
+      "step": 886
+    },
+    {
+      "epoch": 1.3918501446574805,
+      "grad_norm": 0.6076820885521024,
+      "learning_rate": 1.1645505825981884e-05,
+      "loss": 0.7355,
+      "step": 887
+    },
+    {
+      "epoch": 1.393419310547737,
+      "grad_norm": 0.6353027453712743,
+      "learning_rate": 1.162878048745582e-05,
+      "loss": 0.6992,
+      "step": 888
+    },
+    {
+      "epoch": 1.3949884764379934,
+      "grad_norm": 0.6379801521717718,
+      "learning_rate": 1.161205046714589e-05,
+      "loss": 0.719,
+      "step": 889
+    },
+    {
+      "epoch": 1.39655764232825,
+      "grad_norm": 0.655493219820094,
+      "learning_rate": 1.1595315813141041e-05,
+      "loss": 0.7052,
+      "step": 890
+    },
+    {
+      "epoch": 1.3981268082185063,
+      "grad_norm": 0.6490874447096325,
+      "learning_rate": 1.1578576573543541e-05,
+      "loss": 0.7244,
+      "step": 891
+    },
+    {
+      "epoch": 1.399695974108763,
+      "grad_norm": 0.6070219004139445,
+      "learning_rate": 1.1561832796468837e-05,
+      "loss": 0.7064,
+      "step": 892
+    },
+    {
+      "epoch": 1.4012651399990192,
+      "grad_norm": 0.5971904718538777,
+      "learning_rate": 1.1545084530045419e-05,
+      "loss": 0.7143,
+      "step": 893
+    },
+    {
+      "epoch": 1.4028343058892756,
+      "grad_norm": 0.661237907441588,
+      "learning_rate": 1.1528331822414681e-05,
+      "loss": 0.738,
+      "step": 894
+    },
+    {
+      "epoch": 1.4044034717795322,
+      "grad_norm": 0.7048163716097692,
+      "learning_rate": 1.1511574721730781e-05,
+      "loss": 0.727,
+      "step": 895
+    },
+    {
+      "epoch": 1.4059726376697887,
+      "grad_norm": 0.6543903582649592,
+      "learning_rate": 1.1494813276160511e-05,
+      "loss": 0.7091,
+      "step": 896
+    },
+    {
+      "epoch": 1.407541803560045,
+      "grad_norm": 0.6464902292006662,
+      "learning_rate": 1.1478047533883143e-05,
+      "loss": 0.7264,
+      "step": 897
+    },
+    {
+      "epoch": 1.4091109694503015,
+      "grad_norm": 0.6354389748862034,
+      "learning_rate": 1.1461277543090308e-05,
+      "loss": 0.7208,
+      "step": 898
+    },
+    {
+      "epoch": 1.410680135340558,
+      "grad_norm": 0.6426275815483544,
+      "learning_rate": 1.144450335198584e-05,
+      "loss": 0.7328,
+      "step": 899
+    },
+    {
+      "epoch": 1.4122493012308146,
+      "grad_norm": 0.7113472619113095,
+      "learning_rate": 1.1427725008785655e-05,
+      "loss": 0.7173,
+      "step": 900
+    },
+    {
+      "epoch": 1.413818467121071,
+      "grad_norm": 0.6492531266677515,
+      "learning_rate": 1.1410942561717601e-05,
+      "loss": 0.708,
+      "step": 901
+    },
+    {
+      "epoch": 1.4153876330113273,
+      "grad_norm": 0.6004987321507724,
+      "learning_rate": 1.139415605902132e-05,
+      "loss": 0.6925,
+      "step": 902
+    },
+    {
+      "epoch": 1.4169567989015839,
+      "grad_norm": 0.6544327255517137,
+      "learning_rate": 1.1377365548948114e-05,
+      "loss": 0.725,
+      "step": 903
+    },
+    {
+      "epoch": 1.4185259647918405,
+      "grad_norm": 0.6706443927063417,
+      "learning_rate": 1.1360571079760798e-05,
+      "loss": 0.714,
+      "step": 904
+    },
+    {
+      "epoch": 1.4200951306820968,
+      "grad_norm": 0.6554918234186878,
+      "learning_rate": 1.1343772699733579e-05,
+      "loss": 0.747,
+      "step": 905
+    },
+    {
+      "epoch": 1.4216642965723532,
+      "grad_norm": 0.667318349315597,
+      "learning_rate": 1.132697045715189e-05,
+      "loss": 0.7337,
+      "step": 906
+    },
+    {
+      "epoch": 1.4232334624626097,
+      "grad_norm": 0.6498718980360371,
+      "learning_rate": 1.1310164400312284e-05,
+      "loss": 0.7227,
+      "step": 907
+    },
+    {
+      "epoch": 1.4248026283528663,
+      "grad_norm": 0.6638037233928351,
+      "learning_rate": 1.1293354577522264e-05,
+      "loss": 0.7194,
+      "step": 908
+    },
+    {
+      "epoch": 1.4263717942431227,
+      "grad_norm": 0.6009365998164207,
+      "learning_rate": 1.127654103710016e-05,
+      "loss": 0.6977,
+      "step": 909
+    },
+    {
+      "epoch": 1.427940960133379,
+      "grad_norm": 0.6693527197521654,
+      "learning_rate": 1.1259723827374996e-05,
+      "loss": 0.7079,
+      "step": 910
+    },
+    {
+      "epoch": 1.4295101260236356,
+      "grad_norm": 0.6414718328304878,
+      "learning_rate": 1.1242902996686333e-05,
+      "loss": 0.7193,
+      "step": 911
+    },
+    {
+      "epoch": 1.4310792919138922,
+      "grad_norm": 0.6675790358892798,
+      "learning_rate": 1.122607859338415e-05,
+      "loss": 0.6981,
+      "step": 912
+    },
+    {
+      "epoch": 1.4326484578041485,
+      "grad_norm": 0.6441348355283296,
+      "learning_rate": 1.1209250665828684e-05,
+      "loss": 0.7118,
+      "step": 913
+    },
+    {
+      "epoch": 1.4342176236944049,
+      "grad_norm": 0.6164443619823742,
+      "learning_rate": 1.119241926239031e-05,
+      "loss": 0.6896,
+      "step": 914
+    },
+    {
+      "epoch": 1.4357867895846614,
+      "grad_norm": 0.6471133591799608,
+      "learning_rate": 1.1175584431449393e-05,
+      "loss": 0.7232,
+      "step": 915
+    },
+    {
+      "epoch": 1.4373559554749178,
+      "grad_norm": 0.6037249696523037,
+      "learning_rate": 1.1158746221396148e-05,
+      "loss": 0.7037,
+      "step": 916
+    },
+    {
+      "epoch": 1.4389251213651744,
+      "grad_norm": 0.609308940441626,
+      "learning_rate": 1.1141904680630504e-05,
+      "loss": 0.704,
+      "step": 917
+    },
+    {
+      "epoch": 1.4404942872554307,
+      "grad_norm": 0.6461029103749456,
+      "learning_rate": 1.1125059857561966e-05,
+      "loss": 0.7224,
+      "step": 918
+    },
+    {
+      "epoch": 1.4420634531456873,
+      "grad_norm": 0.7260077174455071,
+      "learning_rate": 1.110821180060947e-05,
+      "loss": 0.7468,
+      "step": 919
+    },
+    {
+      "epoch": 1.4436326190359436,
+      "grad_norm": 0.6927068323498747,
+      "learning_rate": 1.1091360558201249e-05,
+      "loss": 0.7221,
+      "step": 920
+    },
+    {
+      "epoch": 1.4452017849262002,
+      "grad_norm": 0.66661507681019,
+      "learning_rate": 1.1074506178774696e-05,
+      "loss": 0.7149,
+      "step": 921
+    },
+    {
+      "epoch": 1.4467709508164566,
+      "grad_norm": 0.6620128299023971,
+      "learning_rate": 1.1057648710776212e-05,
+      "loss": 0.6944,
+      "step": 922
+    },
+    {
+      "epoch": 1.4483401167067131,
+      "grad_norm": 0.6804221522462558,
+      "learning_rate": 1.1040788202661084e-05,
+      "loss": 0.7282,
+      "step": 923
+    },
+    {
+      "epoch": 1.4499092825969695,
+      "grad_norm": 0.6320356229533106,
+      "learning_rate": 1.1023924702893334e-05,
+      "loss": 0.7202,
+      "step": 924
+    },
+    {
+      "epoch": 1.451478448487226,
+      "grad_norm": 0.6686623512159883,
+      "learning_rate": 1.1007058259945584e-05,
+      "loss": 0.7094,
+      "step": 925
+    },
+    {
+      "epoch": 1.4530476143774824,
+      "grad_norm": 0.6887700510314242,
+      "learning_rate": 1.0990188922298918e-05,
+      "loss": 0.7413,
+      "step": 926
+    },
+    {
+      "epoch": 1.454616780267739,
+      "grad_norm": 0.6231211563761995,
+      "learning_rate": 1.0973316738442738e-05,
+      "loss": 0.7225,
+      "step": 927
+    },
+    {
+      "epoch": 1.4561859461579953,
+      "grad_norm": 0.6751436015523725,
+      "learning_rate": 1.0956441756874628e-05,
+      "loss": 0.6898,
+      "step": 928
+    },
+    {
+      "epoch": 1.457755112048252,
+      "grad_norm": 0.6625187200945359,
+      "learning_rate": 1.0939564026100213e-05,
+      "loss": 0.7264,
+      "step": 929
+    },
+    {
+      "epoch": 1.4593242779385083,
+      "grad_norm": 0.6482001119532237,
+      "learning_rate": 1.092268359463302e-05,
+      "loss": 0.7112,
+      "step": 930
+    },
+    {
+      "epoch": 1.4608934438287648,
+      "grad_norm": 0.6301893529545485,
+      "learning_rate": 1.0905800510994341e-05,
+      "loss": 0.686,
+      "step": 931
+    },
+    {
+      "epoch": 1.4624626097190212,
+      "grad_norm": 0.6796906393477904,
+      "learning_rate": 1.0888914823713092e-05,
+      "loss": 0.698,
+      "step": 932
+    },
+    {
+      "epoch": 1.4640317756092778,
+      "grad_norm": 0.670826610392774,
+      "learning_rate": 1.0872026581325667e-05,
+      "loss": 0.7136,
+      "step": 933
+    },
+    {
+      "epoch": 1.4656009414995341,
+      "grad_norm": 0.6748241181522556,
+      "learning_rate": 1.0855135832375809e-05,
+      "loss": 0.7312,
+      "step": 934
+    },
+    {
+      "epoch": 1.4671701073897907,
+      "grad_norm": 0.6810985311683945,
+      "learning_rate": 1.083824262541447e-05,
+      "loss": 0.7226,
+      "step": 935
+    },
+    {
+      "epoch": 1.468739273280047,
+      "grad_norm": 0.6504131617101103,
+      "learning_rate": 1.0821347008999658e-05,
+      "loss": 0.7232,
+      "step": 936
+    },
+    {
+      "epoch": 1.4703084391703034,
+      "grad_norm": 0.6876471796689092,
+      "learning_rate": 1.0804449031696315e-05,
+      "loss": 0.7214,
+      "step": 937
+    },
+    {
+      "epoch": 1.47187760506056,
+      "grad_norm": 0.6586583558429676,
+      "learning_rate": 1.0787548742076164e-05,
+      "loss": 0.7194,
+      "step": 938
+    },
+    {
+      "epoch": 1.4734467709508166,
+      "grad_norm": 0.6545289283894062,
+      "learning_rate": 1.0770646188717578e-05,
+      "loss": 0.731,
+      "step": 939
+    },
+    {
+      "epoch": 1.475015936841073,
+      "grad_norm": 0.605386804813954,
+      "learning_rate": 1.0753741420205431e-05,
+      "loss": 0.7006,
+      "step": 940
+    },
+    {
+      "epoch": 1.4765851027313293,
+      "grad_norm": 0.6464394500621592,
+      "learning_rate": 1.0736834485130972e-05,
+      "loss": 0.718,
+      "step": 941
+    },
+    {
+      "epoch": 1.4781542686215858,
+      "grad_norm": 0.7027032163748959,
+      "learning_rate": 1.0719925432091671e-05,
+      "loss": 0.7013,
+      "step": 942
+    },
+    {
+      "epoch": 1.4797234345118424,
+      "grad_norm": 0.6779436071932261,
+      "learning_rate": 1.0703014309691096e-05,
+      "loss": 0.7037,
+      "step": 943
+    },
+    {
+      "epoch": 1.4812926004020988,
+      "grad_norm": 0.6230572746544857,
+      "learning_rate": 1.068610116653875e-05,
+      "loss": 0.6978,
+      "step": 944
+    },
+    {
+      "epoch": 1.482861766292355,
+      "grad_norm": 0.7217539540895462,
+      "learning_rate": 1.0669186051249948e-05,
+      "loss": 0.7134,
+      "step": 945
+    },
+    {
+      "epoch": 1.4844309321826117,
+      "grad_norm": 0.7009593547944113,
+      "learning_rate": 1.0652269012445684e-05,
+      "loss": 0.7151,
+      "step": 946
+    },
+    {
+      "epoch": 1.4860000980728683,
+      "grad_norm": 0.6544138110624551,
+      "learning_rate": 1.0635350098752468e-05,
+      "loss": 0.7244,
+      "step": 947
+    },
+    {
+      "epoch": 1.4875692639631246,
+      "grad_norm": 0.652779647397337,
+      "learning_rate": 1.0618429358802209e-05,
+      "loss": 0.715,
+      "step": 948
+    },
+    {
+      "epoch": 1.489138429853381,
+      "grad_norm": 0.7527013775930431,
+      "learning_rate": 1.0601506841232057e-05,
+      "loss": 0.6921,
+      "step": 949
+    },
+    {
+      "epoch": 1.4907075957436375,
+      "grad_norm": 0.6203219744767766,
+      "learning_rate": 1.0584582594684274e-05,
+      "loss": 0.7031,
+      "step": 950
+    },
+    {
+      "epoch": 1.492276761633894,
+      "grad_norm": 0.6564685383938513,
+      "learning_rate": 1.0567656667806097e-05,
+      "loss": 0.71,
+      "step": 951
+    },
+    {
+      "epoch": 1.4938459275241505,
+      "grad_norm": 0.7000099478905847,
+      "learning_rate": 1.055072910924959e-05,
+      "loss": 0.7124,
+      "step": 952
+    },
+    {
+      "epoch": 1.4954150934144068,
+      "grad_norm": 0.7004083911366619,
+      "learning_rate": 1.0533799967671503e-05,
+      "loss": 0.7151,
+      "step": 953
+    },
+    {
+      "epoch": 1.4969842593046634,
+      "grad_norm": 0.5966368787953654,
+      "learning_rate": 1.0516869291733142e-05,
+      "loss": 0.7132,
+      "step": 954
+    },
+    {
+      "epoch": 1.49855342519492,
+      "grad_norm": 0.6635042394956364,
+      "learning_rate": 1.0499937130100222e-05,
+      "loss": 0.7156,
+      "step": 955
+    },
+    {
+      "epoch": 1.5001225910851763,
+      "grad_norm": 0.6403983044201088,
+      "learning_rate": 1.0483003531442726e-05,
+      "loss": 0.711,
+      "step": 956
+    },
+    {
+      "epoch": 1.5016917569754327,
+      "grad_norm": 0.6205113619290785,
+      "learning_rate": 1.0466068544434766e-05,
+      "loss": 0.6914,
+      "step": 957
+    },
+    {
+      "epoch": 1.5032609228656892,
+      "grad_norm": 0.6144033172763296,
+      "learning_rate": 1.0449132217754455e-05,
+      "loss": 0.7161,
+      "step": 958
+    },
+    {
+      "epoch": 1.5048300887559458,
+      "grad_norm": 0.6503831884098968,
+      "learning_rate": 1.043219460008374e-05,
+      "loss": 0.7077,
+      "step": 959
+    },
+    {
+      "epoch": 1.5063992546462022,
+      "grad_norm": 0.6559066820649389,
+      "learning_rate": 1.0415255740108299e-05,
+      "loss": 0.7126,
+      "step": 960
+    },
+    {
+      "epoch": 1.5079684205364585,
+      "grad_norm": 0.6669233492788981,
+      "learning_rate": 1.0398315686517362e-05,
+      "loss": 0.7314,
+      "step": 961
+    },
+    {
+      "epoch": 1.509537586426715,
+      "grad_norm": 0.6699956584009586,
+      "learning_rate": 1.0381374488003603e-05,
+      "loss": 0.7058,
+      "step": 962
+    },
+    {
+      "epoch": 1.5111067523169717,
+      "grad_norm": 0.6297987639550877,
+      "learning_rate": 1.036443219326298e-05,
+      "loss": 0.7384,
+      "step": 963
+    },
+    {
+      "epoch": 1.512675918207228,
+      "grad_norm": 0.6489531235629726,
+      "learning_rate": 1.0347488850994608e-05,
+      "loss": 0.6993,
+      "step": 964
+    },
+    {
+      "epoch": 1.5142450840974844,
+      "grad_norm": 0.6696723773300036,
+      "learning_rate": 1.0330544509900603e-05,
+      "loss": 0.694,
+      "step": 965
+    },
+    {
+      "epoch": 1.515814249987741,
+      "grad_norm": 0.6365754829771774,
+      "learning_rate": 1.0313599218685961e-05,
+      "loss": 0.7094,
+      "step": 966
+    },
+    {
+      "epoch": 1.5173834158779973,
+      "grad_norm": 0.6328993746213125,
+      "learning_rate": 1.029665302605841e-05,
+      "loss": 0.6914,
+      "step": 967
+    },
+    {
+      "epoch": 1.5189525817682537,
+      "grad_norm": 0.6139947594794258,
+      "learning_rate": 1.0279705980728258e-05,
+      "loss": 0.7228,
+      "step": 968
+    },
+    {
+      "epoch": 1.5205217476585102,
+      "grad_norm": 0.6390231354865114,
+      "learning_rate": 1.026275813140828e-05,
+      "loss": 0.7143,
+      "step": 969
+    },
+    {
+      "epoch": 1.5220909135487668,
+      "grad_norm": 0.627828876181501,
+      "learning_rate": 1.0245809526813547e-05,
+      "loss": 0.709,
+      "step": 970
+    },
+    {
+      "epoch": 1.5236600794390232,
+      "grad_norm": 0.6732805664396306,
+      "learning_rate": 1.0228860215661309e-05,
+      "loss": 0.7157,
+      "step": 971
+    },
+    {
+      "epoch": 1.5252292453292795,
+      "grad_norm": 0.6259352912612589,
+      "learning_rate": 1.0211910246670848e-05,
+      "loss": 0.7146,
+      "step": 972
+    },
+    {
+      "epoch": 1.526798411219536,
+      "grad_norm": 0.6191689066059438,
+      "learning_rate": 1.0194959668563332e-05,
+      "loss": 0.7007,
+      "step": 973
+    },
+    {
+      "epoch": 1.5283675771097927,
+      "grad_norm": 0.6275859426803602,
+      "learning_rate": 1.017800853006168e-05,
+      "loss": 0.734,
+      "step": 974
+    },
+    {
+      "epoch": 1.529936743000049,
+      "grad_norm": 0.6701857975513259,
+      "learning_rate": 1.0161056879890427e-05,
+      "loss": 0.7195,
+      "step": 975
+    },
+    {
+      "epoch": 1.5315059088903054,
+      "grad_norm": 0.6377874209019501,
+      "learning_rate": 1.0144104766775574e-05,
+      "loss": 0.7168,
+      "step": 976
+    },
+    {
+      "epoch": 1.533075074780562,
+      "grad_norm": 0.6397745601573495,
+      "learning_rate": 1.0127152239444453e-05,
+      "loss": 0.7086,
+      "step": 977
+    },
+    {
+      "epoch": 1.5346442406708185,
+      "grad_norm": 0.6255335784383,
+      "learning_rate": 1.0110199346625593e-05,
+      "loss": 0.6759,
+      "step": 978
+    },
+    {
+      "epoch": 1.5362134065610749,
+      "grad_norm": 0.635433391648022,
+      "learning_rate": 1.0093246137048563e-05,
+      "loss": 0.7035,
+      "step": 979
+    },
+    {
+      "epoch": 1.5377825724513312,
+      "grad_norm": 0.6904416974712441,
+      "learning_rate": 1.0076292659443852e-05,
+      "loss": 0.7104,
+      "step": 980
+    },
+    {
+      "epoch": 1.5393517383415878,
+      "grad_norm": 0.6837712528244906,
+      "learning_rate": 1.0059338962542713e-05,
+      "loss": 0.6925,
+      "step": 981
+    },
+    {
+      "epoch": 1.5409209042318444,
+      "grad_norm": 0.6494950457449515,
+      "learning_rate": 1.0042385095077032e-05,
+      "loss": 0.7164,
+      "step": 982
+    },
+    {
+      "epoch": 1.5424900701221007,
+      "grad_norm": 0.757896168999957,
+      "learning_rate": 1.0025431105779184e-05,
+      "loss": 0.7279,
+      "step": 983
+    },
+    {
+      "epoch": 1.544059236012357,
+      "grad_norm": 0.7448199748521469,
+      "learning_rate": 1.0008477043381896e-05,
+      "loss": 0.7191,
+      "step": 984
+    },
+    {
+      "epoch": 1.5456284019026136,
+      "grad_norm": 0.6136418718230442,
+      "learning_rate": 9.991522956618107e-06,
+      "loss": 0.7051,
+      "step": 985
+    },
+    {
+      "epoch": 1.5471975677928702,
+      "grad_norm": 0.6288927505949025,
+      "learning_rate": 9.974568894220821e-06,
+      "loss": 0.7127,
+      "step": 986
+    },
+    {
+      "epoch": 1.5487667336831266,
+      "grad_norm": 0.6824877652531189,
+      "learning_rate": 9.957614904922971e-06,
+      "loss": 0.7273,
+      "step": 987
+    },
+    {
+      "epoch": 1.550335899573383,
+      "grad_norm": 0.7006077310202782,
+      "learning_rate": 9.940661037457294e-06,
+      "loss": 0.7122,
+      "step": 988
+    },
+    {
+      "epoch": 1.5519050654636395,
+      "grad_norm": 0.6708701065789046,
+      "learning_rate": 9.923707340556151e-06,
+      "loss": 0.7177,
+      "step": 989
+    },
+    {
+      "epoch": 1.553474231353896,
+      "grad_norm": 0.6582307582516453,
+      "learning_rate": 9.906753862951437e-06,
+      "loss": 0.7063,
+      "step": 990
+    },
+    {
+      "epoch": 1.5550433972441524,
+      "grad_norm": 0.6347410586607902,
+      "learning_rate": 9.88980065337441e-06,
+      "loss": 0.7294,
+      "step": 991
+    },
+    {
+      "epoch": 1.5566125631344088,
+      "grad_norm": 0.6348844146933083,
+      "learning_rate": 9.872847760555547e-06,
+      "loss": 0.7165,
+      "step": 992
+    },
+    {
+      "epoch": 1.5581817290246653,
+      "grad_norm": 0.6760485941537946,
+      "learning_rate": 9.855895233224431e-06,
+      "loss": 0.7032,
+      "step": 993
+    },
+    {
+      "epoch": 1.559750894914922,
+      "grad_norm": 0.6775588684652758,
+      "learning_rate": 9.838943120109576e-06,
+      "loss": 0.7317,
+      "step": 994
+    },
+    {
+      "epoch": 1.5613200608051783,
+      "grad_norm": 0.6336931532586304,
+      "learning_rate": 9.821991469938325e-06,
+      "loss": 0.7073,
+      "step": 995
+    },
+    {
+      "epoch": 1.5628892266954346,
+      "grad_norm": 0.6405181006341011,
+      "learning_rate": 9.805040331436672e-06,
+      "loss": 0.7149,
+      "step": 996
+    },
+    {
+      "epoch": 1.5644583925856912,
+      "grad_norm": 0.651328071023763,
+      "learning_rate": 9.788089753329157e-06,
+      "loss": 0.7031,
+      "step": 997
+    },
+    {
+      "epoch": 1.5660275584759478,
+      "grad_norm": 0.6580422372108625,
+      "learning_rate": 9.771139784338693e-06,
+      "loss": 0.7031,
+      "step": 998
+    },
+    {
+      "epoch": 1.5675967243662041,
+      "grad_norm": 0.6747374973150568,
+      "learning_rate": 9.754190473186455e-06,
+      "loss": 0.7052,
+      "step": 999
+    },
+    {
+      "epoch": 1.5691658902564605,
+      "grad_norm": 0.6788063663213355,
+      "learning_rate": 9.737241868591724e-06,
+      "loss": 0.6949,
+      "step": 1000
+    },
+    {
+      "epoch": 1.570735056146717,
+      "grad_norm": 0.6514039944228376,
+      "learning_rate": 9.720294019271744e-06,
+      "loss": 0.7176,
+      "step": 1001
+    },
+    {
+      "epoch": 1.5723042220369736,
+      "grad_norm": 0.6408944598434222,
+      "learning_rate": 9.703346973941595e-06,
+      "loss": 0.6805,
+      "step": 1002
+    },
+    {
+      "epoch": 1.57387338792723,
+      "grad_norm": 0.6423383519755441,
+      "learning_rate": 9.686400781314042e-06,
+      "loss": 0.7159,
+      "step": 1003
+    },
+    {
+      "epoch": 1.5754425538174863,
+      "grad_norm": 0.6986579452497751,
+      "learning_rate": 9.669455490099402e-06,
+      "loss": 0.7017,
+      "step": 1004
+    },
+    {
+      "epoch": 1.577011719707743,
+      "grad_norm": 0.6869231881799038,
+      "learning_rate": 9.652511149005396e-06,
+      "loss": 0.7135,
+      "step": 1005
+    },
+    {
+      "epoch": 1.5785808855979995,
+      "grad_norm": 0.6365031741288624,
+      "learning_rate": 9.635567806737021e-06,
+      "loss": 0.7096,
+      "step": 1006
+    },
+    {
+      "epoch": 1.5801500514882558,
+      "grad_norm": 0.6917464393492744,
+      "learning_rate": 9.618625511996398e-06,
+      "loss": 0.7066,
+      "step": 1007
+    },
+    {
+      "epoch": 1.5817192173785122,
+      "grad_norm": 0.640456584705113,
+      "learning_rate": 9.601684313482638e-06,
+      "loss": 0.6949,
+      "step": 1008
+    },
+    {
+      "epoch": 1.5832883832687688,
+      "grad_norm": 0.6226025973147625,
+      "learning_rate": 9.584744259891706e-06,
+      "loss": 0.7048,
+      "step": 1009
+    },
+    {
+      "epoch": 1.584857549159025,
+      "grad_norm": 0.6239819004200348,
+      "learning_rate": 9.56780539991626e-06,
+      "loss": 0.7051,
+      "step": 1010
+    },
+    {
+      "epoch": 1.5864267150492815,
+      "grad_norm": 0.6370590711664889,
+      "learning_rate": 9.55086778224555e-06,
+      "loss": 0.7057,
+      "step": 1011
+    },
+    {
+      "epoch": 1.587995880939538,
+      "grad_norm": 0.6742768140779845,
+      "learning_rate": 9.533931455565236e-06,
+      "loss": 0.6935,
+      "step": 1012
+    },
+    {
+      "epoch": 1.5895650468297946,
+      "grad_norm": 0.6928710897901328,
+      "learning_rate": 9.51699646855728e-06,
+      "loss": 0.702,
+      "step": 1013
+    },
+    {
+      "epoch": 1.591134212720051,
+      "grad_norm": 0.66231702914188,
+      "learning_rate": 9.50006286989978e-06,
+      "loss": 0.7122,
+      "step": 1014
+    },
+    {
+      "epoch": 1.5927033786103073,
+      "grad_norm": 0.6895821314961985,
+      "learning_rate": 9.483130708266856e-06,
+      "loss": 0.6984,
+      "step": 1015
+    },
+    {
+      "epoch": 1.5942725445005639,
+      "grad_norm": 0.6340169173777641,
+      "learning_rate": 9.466200032328499e-06,
+      "loss": 0.704,
+      "step": 1016
+    },
+    {
+      "epoch": 1.5958417103908205,
+      "grad_norm": 0.6652887889787393,
+      "learning_rate": 9.44927089075041e-06,
+      "loss": 0.7142,
+      "step": 1017
+    },
+    {
+      "epoch": 1.5974108762810768,
+      "grad_norm": 0.6436118759130055,
+      "learning_rate": 9.432343332193907e-06,
+      "loss": 0.7178,
+      "step": 1018
+    },
+    {
+      "epoch": 1.5989800421713332,
+      "grad_norm": 0.6175284973383992,
+      "learning_rate": 9.415417405315727e-06,
+      "loss": 0.7241,
+      "step": 1019
+    },
+    {
+      "epoch": 1.6005492080615897,
+      "grad_norm": 0.6883805518240275,
+      "learning_rate": 9.39849315876795e-06,
+      "loss": 0.7085,
+      "step": 1020
+    },
+    {
+      "epoch": 1.6021183739518463,
+      "grad_norm": 0.6510785737423868,
+      "learning_rate": 9.381570641197795e-06,
+      "loss": 0.703,
+      "step": 1021
+    },
+    {
+      "epoch": 1.6036875398421027,
+      "grad_norm": 0.6316146644684786,
+      "learning_rate": 9.364649901247534e-06,
+      "loss": 0.6784,
+      "step": 1022
+    },
+    {
+      "epoch": 1.605256705732359,
+      "grad_norm": 0.631761598142362,
+      "learning_rate": 9.34773098755432e-06,
+      "loss": 0.7171,
+      "step": 1023
+    },
+    {
+      "epoch": 1.6068258716226156,
+      "grad_norm": 0.6243893090534003,
+      "learning_rate": 9.330813948750053e-06,
+      "loss": 0.7045,
+      "step": 1024
+    },
+    {
+      "epoch": 1.6083950375128722,
+      "grad_norm": 0.6762987341217479,
+      "learning_rate": 9.313898833461255e-06,
+      "loss": 0.7161,
+      "step": 1025
+    },
+    {
+      "epoch": 1.6099642034031285,
+      "grad_norm": 0.6961664714220779,
+      "learning_rate": 9.29698569030891e-06,
+      "loss": 0.6936,
+      "step": 1026
+    },
+    {
+      "epoch": 1.6115333692933849,
+      "grad_norm": 0.6543853600703728,
+      "learning_rate": 9.28007456790833e-06,
+      "loss": 0.7066,
+      "step": 1027
+    },
+    {
+      "epoch": 1.6131025351836414,
+      "grad_norm": 0.6577394657678143,
+      "learning_rate": 9.26316551486903e-06,
+      "loss": 0.7042,
+      "step": 1028
+    },
+    {
+      "epoch": 1.614671701073898,
+      "grad_norm": 0.6213845720425254,
+      "learning_rate": 9.246258579794574e-06,
+      "loss": 0.706,
+      "step": 1029
+    },
+    {
+      "epoch": 1.6162408669641544,
+      "grad_norm": 0.646385894153232,
+      "learning_rate": 9.229353811282427e-06,
+      "loss": 0.6994,
+      "step": 1030
+    },
+    {
+      "epoch": 1.6178100328544107,
+      "grad_norm": 0.6002939184814735,
+      "learning_rate": 9.212451257923841e-06,
+      "loss": 0.7147,
+      "step": 1031
+    },
+    {
+      "epoch": 1.6193791987446673,
+      "grad_norm": 0.6177556132508117,
+      "learning_rate": 9.195550968303688e-06,
+      "loss": 0.7206,
+      "step": 1032
+    },
+    {
+      "epoch": 1.6209483646349239,
+      "grad_norm": 0.6676433083358203,
+      "learning_rate": 9.178652991000342e-06,
+      "loss": 0.7135,
+      "step": 1033
+    },
+    {
+      "epoch": 1.6225175305251802,
+      "grad_norm": 0.6447717531362551,
+      "learning_rate": 9.161757374585533e-06,
+      "loss": 0.6947,
+      "step": 1034
+    },
+    {
+      "epoch": 1.6240866964154366,
+      "grad_norm": 0.6496436648371882,
+      "learning_rate": 9.144864167624191e-06,
+      "loss": 0.7062,
+      "step": 1035
+    },
+    {
+      "epoch": 1.6256558623056931,
+      "grad_norm": 0.6282422668589572,
+      "learning_rate": 9.127973418674338e-06,
+      "loss": 0.7131,
+      "step": 1036
+    },
+    {
+      "epoch": 1.6272250281959497,
+      "grad_norm": 0.6412536147560591,
+      "learning_rate": 9.111085176286912e-06,
+      "loss": 0.705,
+      "step": 1037
+    },
+    {
+      "epoch": 1.628794194086206,
+      "grad_norm": 0.6398745555621472,
+      "learning_rate": 9.094199489005664e-06,
+      "loss": 0.7072,
+      "step": 1038
+    },
+    {
+      "epoch": 1.6303633599764624,
+      "grad_norm": 0.6630610452324819,
+      "learning_rate": 9.07731640536698e-06,
+      "loss": 0.7169,
+      "step": 1039
+    },
+    {
+      "epoch": 1.631932525866719,
+      "grad_norm": 0.6095763545126665,
+      "learning_rate": 9.06043597389979e-06,
+      "loss": 0.7042,
+      "step": 1040
+    },
+    {
+      "epoch": 1.6335016917569756,
+      "grad_norm": 0.6281401607345313,
+      "learning_rate": 9.043558243125374e-06,
+      "loss": 0.7202,
+      "step": 1041
+    },
+    {
+      "epoch": 1.635070857647232,
+      "grad_norm": 0.6491471947760574,
+      "learning_rate": 9.026683261557263e-06,
+      "loss": 0.7088,
+      "step": 1042
+    },
+    {
+      "epoch": 1.6366400235374883,
+      "grad_norm": 0.6245528400850705,
+      "learning_rate": 9.009811077701083e-06,
+      "loss": 0.6997,
+      "step": 1043
+    },
+    {
+      "epoch": 1.6382091894277448,
+      "grad_norm": 0.6014793709833866,
+      "learning_rate": 8.992941740054418e-06,
+      "loss": 0.6864,
+      "step": 1044
+    },
+    {
+      "epoch": 1.6397783553180014,
+      "grad_norm": 0.6712013468046127,
+      "learning_rate": 8.97607529710667e-06,
+      "loss": 0.7091,
+      "step": 1045
+    },
+    {
+      "epoch": 1.6413475212082578,
+      "grad_norm": 0.6456142832514454,
+      "learning_rate": 8.959211797338919e-06,
+      "loss": 0.7274,
+      "step": 1046
+    },
+    {
+      "epoch": 1.6429166870985141,
+      "grad_norm": 0.631039106626239,
+      "learning_rate": 8.942351289223795e-06,
+      "loss": 0.696,
+      "step": 1047
+    },
+    {
+      "epoch": 1.6444858529887707,
+      "grad_norm": 0.6452024951206032,
+      "learning_rate": 8.925493821225309e-06,
+      "loss": 0.7155,
+      "step": 1048
+    },
+    {
+      "epoch": 1.6460550188790273,
+      "grad_norm": 0.6925349978803207,
+      "learning_rate": 8.908639441798751e-06,
+      "loss": 0.7118,
+      "step": 1049
+    },
+    {
+      "epoch": 1.6476241847692836,
+      "grad_norm": 0.623460794807272,
+      "learning_rate": 8.891788199390533e-06,
+      "loss": 0.7196,
+      "step": 1050
+    },
+    {
+      "epoch": 1.64919335065954,
+      "grad_norm": 0.6139370875582836,
+      "learning_rate": 8.874940142438036e-06,
+      "loss": 0.7025,
+      "step": 1051
+    },
+    {
+      "epoch": 1.6507625165497966,
+      "grad_norm": 0.6621554794128032,
+      "learning_rate": 8.858095319369499e-06,
+      "loss": 0.7,
+      "step": 1052
+    },
+    {
+      "epoch": 1.652331682440053,
+      "grad_norm": 0.6638135232309305,
+      "learning_rate": 8.841253778603855e-06,
+      "loss": 0.7089,
+      "step": 1053
+    },
+    {
+      "epoch": 1.6539008483303093,
+      "grad_norm": 0.6247517907078712,
+      "learning_rate": 8.824415568550612e-06,
+      "loss": 0.7008,
+      "step": 1054
+    },
+    {
+      "epoch": 1.6554700142205658,
+      "grad_norm": 0.6244272954448689,
+      "learning_rate": 8.807580737609693e-06,
+      "loss": 0.6954,
+      "step": 1055
+    },
+    {
+      "epoch": 1.6570391801108224,
+      "grad_norm": 0.5945918864496853,
+      "learning_rate": 8.790749334171323e-06,
+      "loss": 0.7039,
+      "step": 1056
+    },
+    {
+      "epoch": 1.6586083460010788,
+      "grad_norm": 0.6509265609516273,
+      "learning_rate": 8.773921406615854e-06,
+      "loss": 0.6901,
+      "step": 1057
+    },
+    {
+      "epoch": 1.6601775118913351,
+      "grad_norm": 0.6405197143222408,
+      "learning_rate": 8.757097003313665e-06,
+      "loss": 0.7107,
+      "step": 1058
+    },
+    {
+      "epoch": 1.6617466777815917,
+      "grad_norm": 0.6280967002098,
+      "learning_rate": 8.740276172625007e-06,
+      "loss": 0.6948,
+      "step": 1059
+    },
+    {
+      "epoch": 1.6633158436718483,
+      "grad_norm": 0.6928514339663628,
+      "learning_rate": 8.72345896289984e-06,
+      "loss": 0.6884,
+      "step": 1060
+    },
+    {
+      "epoch": 1.6648850095621046,
+      "grad_norm": 0.6558722594715815,
+      "learning_rate": 8.706645422477739e-06,
+      "loss": 0.6991,
+      "step": 1061
+    },
+    {
+      "epoch": 1.666454175452361,
+      "grad_norm": 0.6324092178787085,
+      "learning_rate": 8.689835599687717e-06,
+      "loss": 0.7082,
+      "step": 1062
+    },
+    {
+      "epoch": 1.6680233413426175,
+      "grad_norm": 0.5916052072948866,
+      "learning_rate": 8.67302954284811e-06,
+      "loss": 0.7047,
+      "step": 1063
+    },
+    {
+      "epoch": 1.6695925072328741,
+      "grad_norm": 0.6207266100408737,
+      "learning_rate": 8.656227300266425e-06,
+      "loss": 0.6962,
+      "step": 1064
+    },
+    {
+      "epoch": 1.6711616731231305,
+      "grad_norm": 0.6353384778587967,
+      "learning_rate": 8.639428920239205e-06,
+      "loss": 0.7207,
+      "step": 1065
+    },
+    {
+      "epoch": 1.6727308390133868,
+      "grad_norm": 0.5980011237721573,
+      "learning_rate": 8.62263445105189e-06,
+      "loss": 0.6928,
+      "step": 1066
+    },
+    {
+      "epoch": 1.6743000049036434,
+      "grad_norm": 0.6019753532644648,
+      "learning_rate": 8.60584394097868e-06,
+      "loss": 0.7066,
+      "step": 1067
+    },
+    {
+      "epoch": 1.6758691707939,
+      "grad_norm": 0.6245204809660225,
+      "learning_rate": 8.5890574382824e-06,
+      "loss": 0.7018,
+      "step": 1068
+    },
+    {
+      "epoch": 1.6774383366841563,
+      "grad_norm": 0.6188899619713829,
+      "learning_rate": 8.572274991214345e-06,
+      "loss": 0.7271,
+      "step": 1069
+    },
+    {
+      "epoch": 1.6790075025744127,
+      "grad_norm": 0.6620429411198595,
+      "learning_rate": 8.555496648014163e-06,
+      "loss": 0.691,
+      "step": 1070
+    },
+    {
+      "epoch": 1.6805766684646692,
+      "grad_norm": 0.6145848182046878,
+      "learning_rate": 8.538722456909694e-06,
+      "loss": 0.7108,
+      "step": 1071
+    },
+    {
+      "epoch": 1.6821458343549258,
+      "grad_norm": 0.6745297311424414,
+      "learning_rate": 8.521952466116862e-06,
+      "loss": 0.7336,
+      "step": 1072
+    },
+    {
+      "epoch": 1.6837150002451822,
+      "grad_norm": 0.6322868925886423,
+      "learning_rate": 8.505186723839492e-06,
+      "loss": 0.7365,
+      "step": 1073
+    },
+    {
+      "epoch": 1.6852841661354385,
+      "grad_norm": 0.6175557879197305,
+      "learning_rate": 8.488425278269224e-06,
+      "loss": 0.6984,
+      "step": 1074
+    },
+    {
+      "epoch": 1.686853332025695,
+      "grad_norm": 0.6568470238648557,
+      "learning_rate": 8.471668177585324e-06,
+      "loss": 0.7123,
+      "step": 1075
+    },
+    {
+      "epoch": 1.6884224979159517,
+      "grad_norm": 0.6592345405941343,
+      "learning_rate": 8.454915469954583e-06,
+      "loss": 0.704,
+      "step": 1076
+    },
+    {
+      "epoch": 1.689991663806208,
+      "grad_norm": 0.6791387183045869,
+      "learning_rate": 8.438167203531166e-06,
+      "loss": 0.6947,
+      "step": 1077
+    },
+    {
+      "epoch": 1.6915608296964644,
+      "grad_norm": 0.6171406367367374,
+      "learning_rate": 8.42142342645646e-06,
+      "loss": 0.7178,
+      "step": 1078
+    },
+    {
+      "epoch": 1.693129995586721,
+      "grad_norm": 0.708568725588992,
+      "learning_rate": 8.404684186858962e-06,
+      "loss": 0.6958,
+      "step": 1079
+    },
+    {
+      "epoch": 1.6946991614769775,
+      "grad_norm": 0.6341635834833368,
+      "learning_rate": 8.387949532854113e-06,
+      "loss": 0.7012,
+      "step": 1080
+    },
+    {
+      "epoch": 1.6962683273672339,
+      "grad_norm": 0.6458269723907708,
+      "learning_rate": 8.371219512544182e-06,
+      "loss": 0.6971,
+      "step": 1081
+    },
+    {
+      "epoch": 1.6978374932574902,
+      "grad_norm": 0.6702174914601319,
+      "learning_rate": 8.354494174018118e-06,
+      "loss": 0.7254,
+      "step": 1082
+    },
+    {
+      "epoch": 1.6994066591477468,
+      "grad_norm": 0.6496132767064503,
+      "learning_rate": 8.337773565351407e-06,
+      "loss": 0.705,
+      "step": 1083
+    },
+    {
+      "epoch": 1.7009758250380034,
+      "grad_norm": 0.6965140918408609,
+      "learning_rate": 8.321057734605949e-06,
+      "loss": 0.7035,
+      "step": 1084
+    },
+    {
+      "epoch": 1.7025449909282597,
+      "grad_norm": 0.6492386198040799,
+      "learning_rate": 8.3043467298299e-06,
+      "loss": 0.7107,
+      "step": 1085
+    },
+    {
+      "epoch": 1.704114156818516,
+      "grad_norm": 0.6844803310270812,
+      "learning_rate": 8.287640599057551e-06,
+      "loss": 0.7278,
+      "step": 1086
+    },
+    {
+      "epoch": 1.7056833227087727,
+      "grad_norm": 0.6601689944185404,
+      "learning_rate": 8.270939390309181e-06,
+      "loss": 0.721,
+      "step": 1087
+    },
+    {
+      "epoch": 1.7072524885990292,
+      "grad_norm": 0.6055565802709788,
+      "learning_rate": 8.254243151590922e-06,
+      "loss": 0.6885,
+      "step": 1088
+    },
+    {
+      "epoch": 1.7088216544892856,
+      "grad_norm": 0.659804664271136,
+      "learning_rate": 8.237551930894616e-06,
+      "loss": 0.699,
+      "step": 1089
+    },
+    {
+      "epoch": 1.710390820379542,
+      "grad_norm": 0.5884698439102443,
+      "learning_rate": 8.22086577619769e-06,
+      "loss": 0.7148,
+      "step": 1090
+    },
+    {
+      "epoch": 1.7119599862697985,
+      "grad_norm": 0.5971594463869768,
+      "learning_rate": 8.204184735462994e-06,
+      "loss": 0.7134,
+      "step": 1091
+    },
+    {
+      "epoch": 1.713529152160055,
+      "grad_norm": 0.6169754619874099,
+      "learning_rate": 8.187508856638697e-06,
+      "loss": 0.7109,
+      "step": 1092
+    },
+    {
+      "epoch": 1.7150983180503114,
+      "grad_norm": 0.6443043273623997,
+      "learning_rate": 8.17083818765812e-06,
+      "loss": 0.6808,
+      "step": 1093
+    },
+    {
+      "epoch": 1.7166674839405678,
+      "grad_norm": 0.6946723093457059,
+      "learning_rate": 8.154172776439608e-06,
+      "loss": 0.7127,
+      "step": 1094
+    },
+    {
+      "epoch": 1.7182366498308244,
+      "grad_norm": 0.6047874938726835,
+      "learning_rate": 8.137512670886397e-06,
+      "loss": 0.6803,
+      "step": 1095
+    },
+    {
+      "epoch": 1.7198058157210807,
+      "grad_norm": 0.6656951538199034,
+      "learning_rate": 8.120857918886467e-06,
+      "loss": 0.7093,
+      "step": 1096
+    },
+    {
+      "epoch": 1.721374981611337,
+      "grad_norm": 0.6837507066445706,
+      "learning_rate": 8.10420856831242e-06,
+      "loss": 0.6942,
+      "step": 1097
+    },
+    {
+      "epoch": 1.7229441475015936,
+      "grad_norm": 0.635602582087701,
+      "learning_rate": 8.087564667021317e-06,
+      "loss": 0.7012,
+      "step": 1098
+    },
+    {
+      "epoch": 1.7245133133918502,
+      "grad_norm": 0.599712409319767,
+      "learning_rate": 8.070926262854573e-06,
+      "loss": 0.7049,
+      "step": 1099
+    },
+    {
+      "epoch": 1.7260824792821066,
+      "grad_norm": 0.6495867025142901,
+      "learning_rate": 8.054293403637787e-06,
+      "loss": 0.6966,
+      "step": 1100
+    },
+    {
+      "epoch": 1.727651645172363,
+      "grad_norm": 0.6718493170372416,
+      "learning_rate": 8.037666137180624e-06,
+      "loss": 0.7085,
+      "step": 1101
+    },
+    {
+      "epoch": 1.7292208110626195,
+      "grad_norm": 0.6749526961002337,
+      "learning_rate": 8.021044511276682e-06,
+      "loss": 0.7028,
+      "step": 1102
+    },
+    {
+      "epoch": 1.730789976952876,
+      "grad_norm": 0.6270599084234583,
+      "learning_rate": 8.00442857370333e-06,
+      "loss": 0.7098,
+      "step": 1103
+    },
+    {
+      "epoch": 1.7323591428431324,
+      "grad_norm": 0.6648225603969561,
+      "learning_rate": 7.987818372221603e-06,
+      "loss": 0.7173,
+      "step": 1104
+    },
+    {
+      "epoch": 1.7339283087333888,
+      "grad_norm": 0.6348604444177273,
+      "learning_rate": 7.971213954576031e-06,
+      "loss": 0.7104,
+      "step": 1105
+    },
+    {
+      "epoch": 1.7354974746236453,
+      "grad_norm": 0.6001370659336149,
+      "learning_rate": 7.954615368494538e-06,
+      "loss": 0.7059,
+      "step": 1106
+    },
+    {
+      "epoch": 1.737066640513902,
+      "grad_norm": 0.6798105594418516,
+      "learning_rate": 7.938022661688265e-06,
+      "loss": 0.7145,
+      "step": 1107
+    },
+    {
+      "epoch": 1.7386358064041583,
+      "grad_norm": 0.6360471172877283,
+      "learning_rate": 7.921435881851474e-06,
+      "loss": 0.705,
+      "step": 1108
+    },
+    {
+      "epoch": 1.7402049722944146,
+      "grad_norm": 0.6201481389190848,
+      "learning_rate": 7.90485507666137e-06,
+      "loss": 0.6879,
+      "step": 1109
+    },
+    {
+      "epoch": 1.7417741381846712,
+      "grad_norm": 0.6459395460997385,
+      "learning_rate": 7.888280293778001e-06,
+      "loss": 0.703,
+      "step": 1110
+    },
+    {
+      "epoch": 1.7433433040749278,
+      "grad_norm": 0.612585709914007,
+      "learning_rate": 7.871711580844102e-06,
+      "loss": 0.7029,
+      "step": 1111
+    },
+    {
+      "epoch": 1.7449124699651841,
+      "grad_norm": 0.6470757813091548,
+      "learning_rate": 7.855148985484946e-06,
+      "loss": 0.7088,
+      "step": 1112
+    },
+    {
+      "epoch": 1.7464816358554405,
+      "grad_norm": 0.6097058940318653,
+      "learning_rate": 7.838592555308244e-06,
+      "loss": 0.7018,
+      "step": 1113
+    },
+    {
+      "epoch": 1.748050801745697,
+      "grad_norm": 0.6122460281216233,
+      "learning_rate": 7.822042337903964e-06,
+      "loss": 0.7031,
+      "step": 1114
+    },
+    {
+      "epoch": 1.7496199676359536,
+      "grad_norm": 0.5971811555292899,
+      "learning_rate": 7.805498380844236e-06,
+      "loss": 0.7043,
+      "step": 1115
+    },
+    {
+      "epoch": 1.75118913352621,
+      "grad_norm": 0.6353531859378245,
+      "learning_rate": 7.788960731683174e-06,
+      "loss": 0.7323,
+      "step": 1116
+    },
+    {
+      "epoch": 1.7527582994164663,
+      "grad_norm": 0.6195863182578605,
+      "learning_rate": 7.772429437956777e-06,
+      "loss": 0.6801,
+      "step": 1117
+    },
+    {
+      "epoch": 1.754327465306723,
+      "grad_norm": 0.6534979629661563,
+      "learning_rate": 7.755904547182778e-06,
+      "loss": 0.7307,
+      "step": 1118
+    },
+    {
+      "epoch": 1.7558966311969795,
+      "grad_norm": 0.6203916414515298,
+      "learning_rate": 7.739386106860485e-06,
+      "loss": 0.727,
+      "step": 1119
+    },
+    {
+      "epoch": 1.7574657970872358,
+      "grad_norm": 0.572405593970221,
+      "learning_rate": 7.722874164470685e-06,
+      "loss": 0.6816,
+      "step": 1120
+    },
+    {
+      "epoch": 1.7590349629774922,
+      "grad_norm": 0.6302250883720385,
+      "learning_rate": 7.70636876747548e-06,
+      "loss": 0.6967,
+      "step": 1121
+    },
+    {
+      "epoch": 1.7606041288677488,
+      "grad_norm": 0.677220696069866,
+      "learning_rate": 7.689869963318154e-06,
+      "loss": 0.7089,
+      "step": 1122
+    },
+    {
+      "epoch": 1.7621732947580053,
+      "grad_norm": 0.6541747936470931,
+      "learning_rate": 7.673377799423047e-06,
+      "loss": 0.7224,
+      "step": 1123
+    },
+    {
+      "epoch": 1.7637424606482617,
+      "grad_norm": 0.6566712143387494,
+      "learning_rate": 7.656892323195414e-06,
+      "loss": 0.6964,
+      "step": 1124
+    },
+    {
+      "epoch": 1.765311626538518,
+      "grad_norm": 0.665809611809167,
+      "learning_rate": 7.64041358202127e-06,
+      "loss": 0.7227,
+      "step": 1125
+    },
+    {
+      "epoch": 1.7668807924287746,
+      "grad_norm": 0.6281827985793729,
+      "learning_rate": 7.6239416232672904e-06,
+      "loss": 0.6997,
+      "step": 1126
+    },
+    {
+      "epoch": 1.7684499583190312,
+      "grad_norm": 0.6507539991939618,
+      "learning_rate": 7.607476494280653e-06,
+      "loss": 0.6937,
+      "step": 1127
+    },
+    {
+      "epoch": 1.7700191242092875,
+      "grad_norm": 0.6359338791494793,
+      "learning_rate": 7.591018242388888e-06,
+      "loss": 0.7067,
+      "step": 1128
+    },
+    {
+      "epoch": 1.7715882900995439,
+      "grad_norm": 0.6121130124235268,
+      "learning_rate": 7.574566914899779e-06,
+      "loss": 0.687,
+      "step": 1129
+    },
+    {
+      "epoch": 1.7731574559898005,
+      "grad_norm": 0.6585417807949921,
+      "learning_rate": 7.5581225591011864e-06,
+      "loss": 0.7131,
+      "step": 1130
+    },
+    {
+      "epoch": 1.774726621880057,
+      "grad_norm": 0.7000636376122298,
+      "learning_rate": 7.541685222260948e-06,
+      "loss": 0.6959,
+      "step": 1131
+    },
+    {
+      "epoch": 1.7762957877703134,
+      "grad_norm": 0.651308126942087,
+      "learning_rate": 7.525254951626712e-06,
+      "loss": 0.7051,
+      "step": 1132
+    },
+    {
+      "epoch": 1.7778649536605697,
+      "grad_norm": 0.6549972425198098,
+      "learning_rate": 7.5088317944258305e-06,
+      "loss": 0.6977,
+      "step": 1133
+    },
+    {
+      "epoch": 1.7794341195508263,
+      "grad_norm": 0.6300073025952618,
+      "learning_rate": 7.492415797865191e-06,
+      "loss": 0.7004,
+      "step": 1134
+    },
+    {
+      "epoch": 1.7810032854410829,
+      "grad_norm": 0.6601507019986638,
+      "learning_rate": 7.476007009131112e-06,
+      "loss": 0.6983,
+      "step": 1135
+    },
+    {
+      "epoch": 1.7825724513313392,
+      "grad_norm": 0.621486442547712,
+      "learning_rate": 7.459605475389193e-06,
+      "loss": 0.6929,
+      "step": 1136
+    },
+    {
+      "epoch": 1.7841416172215956,
+      "grad_norm": 0.6140955996488306,
+      "learning_rate": 7.44321124378417e-06,
+      "loss": 0.6872,
+      "step": 1137
+    },
+    {
+      "epoch": 1.7857107831118522,
+      "grad_norm": 0.6319486258470721,
+      "learning_rate": 7.426824361439803e-06,
+      "loss": 0.7088,
+      "step": 1138
+    },
+    {
+      "epoch": 1.7872799490021085,
+      "grad_norm": 0.6263204875215514,
+      "learning_rate": 7.41044487545871e-06,
+      "loss": 0.7126,
+      "step": 1139
+    },
+    {
+      "epoch": 1.7888491148923649,
+      "grad_norm": 0.6359878910671518,
+      "learning_rate": 7.394072832922268e-06,
+      "loss": 0.7023,
+      "step": 1140
+    },
+    {
+      "epoch": 1.7904182807826214,
+      "grad_norm": 0.6152475421394614,
+      "learning_rate": 7.3777082808904435e-06,
+      "loss": 0.7007,
+      "step": 1141
+    },
+    {
+      "epoch": 1.791987446672878,
+      "grad_norm": 0.7300175781449448,
+      "learning_rate": 7.361351266401682e-06,
+      "loss": 0.7179,
+      "step": 1142
+    },
+    {
+      "epoch": 1.7935566125631344,
+      "grad_norm": 0.6734639434083697,
+      "learning_rate": 7.345001836472754e-06,
+      "loss": 0.6954,
+      "step": 1143
+    },
+    {
+      "epoch": 1.7951257784533907,
+      "grad_norm": 0.6559915907143103,
+      "learning_rate": 7.328660038098636e-06,
+      "loss": 0.6975,
+      "step": 1144
+    },
+    {
+      "epoch": 1.7966949443436473,
+      "grad_norm": 0.6230344497628719,
+      "learning_rate": 7.312325918252373e-06,
+      "loss": 0.6896,
+      "step": 1145
+    },
+    {
+      "epoch": 1.7982641102339039,
+      "grad_norm": 0.6418743424655805,
+      "learning_rate": 7.295999523884921e-06,
+      "loss": 0.7075,
+      "step": 1146
+    },
+    {
+      "epoch": 1.7998332761241602,
+      "grad_norm": 0.6345174041260885,
+      "learning_rate": 7.279680901925051e-06,
+      "loss": 0.6982,
+      "step": 1147
+    },
+    {
+      "epoch": 1.8014024420144166,
+      "grad_norm": 0.6815099482696455,
+      "learning_rate": 7.263370099279173e-06,
+      "loss": 0.7061,
+      "step": 1148
+    },
+    {
+      "epoch": 1.8029716079046731,
+      "grad_norm": 0.6375270587693587,
+      "learning_rate": 7.247067162831237e-06,
+      "loss": 0.7279,
+      "step": 1149
+    },
+    {
+      "epoch": 1.8045407737949297,
+      "grad_norm": 0.6504743023164037,
+      "learning_rate": 7.230772139442572e-06,
+      "loss": 0.6958,
+      "step": 1150
+    },
+    {
+      "epoch": 1.806109939685186,
+      "grad_norm": 0.6605354589680494,
+      "learning_rate": 7.214485075951771e-06,
+      "loss": 0.6904,
+      "step": 1151
+    },
+    {
+      "epoch": 1.8076791055754424,
+      "grad_norm": 0.6170601615698432,
+      "learning_rate": 7.198206019174534e-06,
+      "loss": 0.6947,
+      "step": 1152
+    },
+    {
+      "epoch": 1.809248271465699,
+      "grad_norm": 0.6541446258968322,
+      "learning_rate": 7.181935015903558e-06,
+      "loss": 0.7062,
+      "step": 1153
+    },
+    {
+      "epoch": 1.8108174373559556,
+      "grad_norm": 0.6584528391904312,
+      "learning_rate": 7.165672112908393e-06,
+      "loss": 0.702,
+      "step": 1154
+    },
+    {
+      "epoch": 1.812386603246212,
+      "grad_norm": 0.6332537278885018,
+      "learning_rate": 7.149417356935289e-06,
+      "loss": 0.6919,
+      "step": 1155
+    },
+    {
+      "epoch": 1.8139557691364683,
+      "grad_norm": 0.6131041572435836,
+      "learning_rate": 7.133170794707096e-06,
+      "loss": 0.6835,
+      "step": 1156
+    },
+    {
+      "epoch": 1.8155249350267249,
+      "grad_norm": 0.6285075754323795,
+      "learning_rate": 7.116932472923098e-06,
+      "loss": 0.7236,
+      "step": 1157
+    },
+    {
+      "epoch": 1.8170941009169814,
+      "grad_norm": 0.6553522780245459,
+      "learning_rate": 7.100702438258905e-06,
+      "loss": 0.7045,
+      "step": 1158
+    },
+    {
+      "epoch": 1.8186632668072378,
+      "grad_norm": 0.6814098888024047,
+      "learning_rate": 7.084480737366297e-06,
+      "loss": 0.7048,
+      "step": 1159
+    },
+    {
+      "epoch": 1.8202324326974941,
+      "grad_norm": 0.6812839488763198,
+      "learning_rate": 7.068267416873102e-06,
+      "loss": 0.7076,
+      "step": 1160
+    },
+    {
+      "epoch": 1.8218015985877507,
+      "grad_norm": 0.6119633775778521,
+      "learning_rate": 7.052062523383065e-06,
+      "loss": 0.7157,
+      "step": 1161
+    },
+    {
+      "epoch": 1.8233707644780073,
+      "grad_norm": 0.6395066806290381,
+      "learning_rate": 7.035866103475698e-06,
+      "loss": 0.7048,
+      "step": 1162
+    },
+    {
+      "epoch": 1.8249399303682636,
+      "grad_norm": 0.6420004479602329,
+      "learning_rate": 7.019678203706164e-06,
+      "loss": 0.688,
+      "step": 1163
+    },
+    {
+      "epoch": 1.82650909625852,
+      "grad_norm": 0.6428224074749301,
+      "learning_rate": 7.0034988706051325e-06,
+      "loss": 0.7029,
+      "step": 1164
+    },
+    {
+      "epoch": 1.8280782621487766,
+      "grad_norm": 0.6180945076131441,
+      "learning_rate": 6.987328150678652e-06,
+      "loss": 0.6981,
+      "step": 1165
+    },
+    {
+      "epoch": 1.8296474280390331,
+      "grad_norm": 0.6658191389298024,
+      "learning_rate": 6.9711660904080055e-06,
+      "loss": 0.7133,
+      "step": 1166
+    },
+    {
+      "epoch": 1.8312165939292895,
+      "grad_norm": 0.6269620808352516,
+      "learning_rate": 6.955012736249598e-06,
+      "loss": 0.6915,
+      "step": 1167
+    },
+    {
+      "epoch": 1.8327857598195458,
+      "grad_norm": 0.6347435455195121,
+      "learning_rate": 6.938868134634793e-06,
+      "loss": 0.6862,
+      "step": 1168
+    },
+    {
+      "epoch": 1.8343549257098024,
+      "grad_norm": 0.68734574220207,
+      "learning_rate": 6.922732331969808e-06,
+      "loss": 0.7223,
+      "step": 1169
+    },
+    {
+      "epoch": 1.835924091600059,
+      "grad_norm": 0.7521968601568131,
+      "learning_rate": 6.90660537463557e-06,
+      "loss": 0.7278,
+      "step": 1170
+    },
+    {
+      "epoch": 1.8374932574903153,
+      "grad_norm": 0.6615864158860839,
+      "learning_rate": 6.890487308987567e-06,
+      "loss": 0.6935,
+      "step": 1171
+    },
+    {
+      "epoch": 1.8390624233805717,
+      "grad_norm": 0.6575366132447145,
+      "learning_rate": 6.8743781813557455e-06,
+      "loss": 0.699,
+      "step": 1172
+    },
+    {
+      "epoch": 1.8406315892708283,
+      "grad_norm": 0.605647784389949,
+      "learning_rate": 6.858278038044344e-06,
+      "loss": 0.7106,
+      "step": 1173
+    },
+    {
+      "epoch": 1.8422007551610848,
+      "grad_norm": 0.6151684801912656,
+      "learning_rate": 6.842186925331793e-06,
+      "loss": 0.6822,
+      "step": 1174
+    },
+    {
+      "epoch": 1.8437699210513412,
+      "grad_norm": 0.6496773321129877,
+      "learning_rate": 6.8261048894705505e-06,
+      "loss": 0.7023,
+      "step": 1175
+    },
+    {
+      "epoch": 1.8453390869415975,
+      "grad_norm": 0.6790779081064787,
+      "learning_rate": 6.810031976686994e-06,
+      "loss": 0.7009,
+      "step": 1176
+    },
+    {
+      "epoch": 1.8469082528318541,
+      "grad_norm": 0.6492562303262582,
+      "learning_rate": 6.793968233181274e-06,
+      "loss": 0.7033,
+      "step": 1177
+    },
+    {
+      "epoch": 1.8484774187221107,
+      "grad_norm": 0.6795535516752428,
+      "learning_rate": 6.777913705127181e-06,
+      "loss": 0.6974,
+      "step": 1178
+    },
+    {
+      "epoch": 1.850046584612367,
+      "grad_norm": 0.6133435325482682,
+      "learning_rate": 6.761868438672028e-06,
+      "loss": 0.6768,
+      "step": 1179
+    },
+    {
+      "epoch": 1.8516157505026234,
+      "grad_norm": 0.6734783786535161,
+      "learning_rate": 6.745832479936492e-06,
+      "loss": 0.7258,
+      "step": 1180
+    },
+    {
+      "epoch": 1.85318491639288,
+      "grad_norm": 0.6553186703528671,
+      "learning_rate": 6.729805875014506e-06,
+      "loss": 0.692,
+      "step": 1181
+    },
+    {
+      "epoch": 1.8547540822831363,
+      "grad_norm": 0.6689015345419655,
+      "learning_rate": 6.713788669973108e-06,
+      "loss": 0.7275,
+      "step": 1182
+    },
+    {
+      "epoch": 1.8563232481733927,
+      "grad_norm": 0.652414854955715,
+      "learning_rate": 6.6977809108523295e-06,
+      "loss": 0.7032,
+      "step": 1183
+    },
+    {
+      "epoch": 1.8578924140636492,
+      "grad_norm": 0.6660192902805323,
+      "learning_rate": 6.6817826436650335e-06,
+      "loss": 0.709,
+      "step": 1184
+    },
+    {
+      "epoch": 1.8594615799539058,
+      "grad_norm": 0.6294970132004326,
+      "learning_rate": 6.665793914396818e-06,
+      "loss": 0.6991,
+      "step": 1185
+    },
+    {
+      "epoch": 1.8610307458441622,
+      "grad_norm": 0.6494049490430254,
+      "learning_rate": 6.649814769005847e-06,
+      "loss": 0.7049,
+      "step": 1186
+    },
+    {
+      "epoch": 1.8625999117344185,
+      "grad_norm": 0.6450618366003338,
+      "learning_rate": 6.633845253422748e-06,
+      "loss": 0.686,
+      "step": 1187
+    },
+    {
+      "epoch": 1.864169077624675,
+      "grad_norm": 0.6273351263132932,
+      "learning_rate": 6.617885413550471e-06,
+      "loss": 0.6983,
+      "step": 1188
+    },
+    {
+      "epoch": 1.8657382435149317,
+      "grad_norm": 0.6407254618041669,
+      "learning_rate": 6.601935295264139e-06,
+      "loss": 0.6948,
+      "step": 1189
+    },
+    {
+      "epoch": 1.867307409405188,
+      "grad_norm": 0.6199466084000794,
+      "learning_rate": 6.5859949444109495e-06,
+      "loss": 0.6991,
+      "step": 1190
+    },
+    {
+      "epoch": 1.8688765752954444,
+      "grad_norm": 0.6436014472593918,
+      "learning_rate": 6.570064406810008e-06,
+      "loss": 0.7057,
+      "step": 1191
+    },
+    {
+      "epoch": 1.870445741185701,
+      "grad_norm": 0.606860907475786,
+      "learning_rate": 6.554143728252229e-06,
+      "loss": 0.7173,
+      "step": 1192
+    },
+    {
+      "epoch": 1.8720149070759575,
+      "grad_norm": 0.6066268486690809,
+      "learning_rate": 6.53823295450017e-06,
+      "loss": 0.7038,
+      "step": 1193
+    },
+    {
+      "epoch": 1.8735840729662139,
+      "grad_norm": 0.6279211466266119,
+      "learning_rate": 6.522332131287931e-06,
+      "loss": 0.6995,
+      "step": 1194
+    },
+    {
+      "epoch": 1.8751532388564702,
+      "grad_norm": 0.6780794136317954,
+      "learning_rate": 6.506441304321012e-06,
+      "loss": 0.7021,
+      "step": 1195
+    },
+    {
+      "epoch": 1.8767224047467268,
+      "grad_norm": 0.6918516064639121,
+      "learning_rate": 6.4905605192761656e-06,
+      "loss": 0.7025,
+      "step": 1196
+    },
+    {
+      "epoch": 1.8782915706369834,
+      "grad_norm": 0.6397444651462989,
+      "learning_rate": 6.474689821801295e-06,
+      "loss": 0.7016,
+      "step": 1197
+    },
+    {
+      "epoch": 1.8798607365272397,
+      "grad_norm": 0.6586939047854918,
+      "learning_rate": 6.4588292575152955e-06,
+      "loss": 0.6935,
+      "step": 1198
+    },
+    {
+      "epoch": 1.881429902417496,
+      "grad_norm": 0.659129961467629,
+      "learning_rate": 6.442978872007946e-06,
+      "loss": 0.72,
+      "step": 1199
+    },
+    {
+      "epoch": 1.8829990683077527,
+      "grad_norm": 0.6227414109356603,
+      "learning_rate": 6.42713871083976e-06,
+      "loss": 0.6982,
+      "step": 1200
+    },
+    {
+      "epoch": 1.8845682341980092,
+      "grad_norm": 0.5865974974437855,
+      "learning_rate": 6.411308819541863e-06,
+      "loss": 0.6706,
+      "step": 1201
+    },
+    {
+      "epoch": 1.8861374000882656,
+      "grad_norm": 0.6426127256430466,
+      "learning_rate": 6.395489243615861e-06,
+      "loss": 0.683,
+      "step": 1202
+    },
+    {
+      "epoch": 1.887706565978522,
+      "grad_norm": 0.7075310580622208,
+      "learning_rate": 6.379680028533714e-06,
+      "loss": 0.7047,
+      "step": 1203
+    },
+    {
+      "epoch": 1.8892757318687785,
+      "grad_norm": 0.6629674504863359,
+      "learning_rate": 6.363881219737599e-06,
+      "loss": 0.7181,
+      "step": 1204
+    },
+    {
+      "epoch": 1.890844897759035,
+      "grad_norm": 0.6551682368715769,
+      "learning_rate": 6.348092862639772e-06,
+      "loss": 0.7074,
+      "step": 1205
+    },
+    {
+      "epoch": 1.8924140636492914,
+      "grad_norm": 0.6061861217859599,
+      "learning_rate": 6.332315002622461e-06,
+      "loss": 0.7005,
+      "step": 1206
+    },
+    {
+      "epoch": 1.8939832295395478,
+      "grad_norm": 0.675041981785903,
+      "learning_rate": 6.316547685037707e-06,
+      "loss": 0.6858,
+      "step": 1207
+    },
+    {
+      "epoch": 1.8955523954298044,
+      "grad_norm": 0.6344423128781724,
+      "learning_rate": 6.300790955207262e-06,
+      "loss": 0.6825,
+      "step": 1208
+    },
+    {
+      "epoch": 1.897121561320061,
+      "grad_norm": 0.6460233777183989,
+      "learning_rate": 6.285044858422429e-06,
+      "loss": 0.6993,
+      "step": 1209
+    },
+    {
+      "epoch": 1.8986907272103173,
+      "grad_norm": 0.6365019211850873,
+      "learning_rate": 6.269309439943963e-06,
+      "loss": 0.6914,
+      "step": 1210
+    },
+    {
+      "epoch": 1.9002598931005736,
+      "grad_norm": 0.6522649543255505,
+      "learning_rate": 6.253584745001908e-06,
+      "loss": 0.7229,
+      "step": 1211
+    },
+    {
+      "epoch": 1.9018290589908302,
+      "grad_norm": 0.6055627378981961,
+      "learning_rate": 6.237870818795499e-06,
+      "loss": 0.6996,
+      "step": 1212
+    },
+    {
+      "epoch": 1.9033982248810868,
+      "grad_norm": 0.6837493058201992,
+      "learning_rate": 6.222167706493016e-06,
+      "loss": 0.6924,
+      "step": 1213
+    },
+    {
+      "epoch": 1.9049673907713431,
+      "grad_norm": 0.6552310529367559,
+      "learning_rate": 6.206475453231644e-06,
+      "loss": 0.7007,
+      "step": 1214
+    },
+    {
+      "epoch": 1.9065365566615995,
+      "grad_norm": 0.6126641002931191,
+      "learning_rate": 6.190794104117368e-06,
+      "loss": 0.6744,
+      "step": 1215
+    },
+    {
+      "epoch": 1.908105722551856,
+      "grad_norm": 0.5974081882305156,
+      "learning_rate": 6.175123704224814e-06,
+      "loss": 0.6889,
+      "step": 1216
+    },
+    {
+      "epoch": 1.9096748884421126,
+      "grad_norm": 0.6345803586892547,
+      "learning_rate": 6.1594642985971565e-06,
+      "loss": 0.6874,
+      "step": 1217
+    },
+    {
+      "epoch": 1.911244054332369,
+      "grad_norm": 0.5714376332451121,
+      "learning_rate": 6.143815932245951e-06,
+      "loss": 0.685,
+      "step": 1218
+    },
+    {
+      "epoch": 1.9128132202226253,
+      "grad_norm": 0.609582576120246,
+      "learning_rate": 6.128178650151026e-06,
+      "loss": 0.6934,
+      "step": 1219
+    },
+    {
+      "epoch": 1.914382386112882,
+      "grad_norm": 0.6239403437710609,
+      "learning_rate": 6.112552497260353e-06,
+      "loss": 0.7222,
+      "step": 1220
+    },
+    {
+      "epoch": 1.9159515520031385,
+      "grad_norm": 0.6256073955673842,
+      "learning_rate": 6.096937518489912e-06,
+      "loss": 0.7009,
+      "step": 1221
+    },
+    {
+      "epoch": 1.9175207178933948,
+      "grad_norm": 0.6455703858912576,
+      "learning_rate": 6.081333758723562e-06,
+      "loss": 0.6953,
+      "step": 1222
+    },
+    {
+      "epoch": 1.9190898837836512,
+      "grad_norm": 0.6125302149136747,
+      "learning_rate": 6.065741262812914e-06,
+      "loss": 0.6943,
+      "step": 1223
+    },
+    {
+      "epoch": 1.9206590496739078,
+      "grad_norm": 0.6061051920487148,
+      "learning_rate": 6.0501600755772074e-06,
+      "loss": 0.7012,
+      "step": 1224
+    },
+    {
+      "epoch": 1.9222282155641641,
+      "grad_norm": 0.6187039627628196,
+      "learning_rate": 6.0345902418031645e-06,
+      "loss": 0.7029,
+      "step": 1225
+    },
+    {
+      "epoch": 1.9237973814544205,
+      "grad_norm": 0.6376717568508281,
+      "learning_rate": 6.019031806244888e-06,
+      "loss": 0.7116,
+      "step": 1226
+    },
+    {
+      "epoch": 1.925366547344677,
+      "grad_norm": 0.610664608337934,
+      "learning_rate": 6.0034848136237035e-06,
+      "loss": 0.7179,
+      "step": 1227
+    },
+    {
+      "epoch": 1.9269357132349336,
+      "grad_norm": 0.6886841222426429,
+      "learning_rate": 5.987949308628054e-06,
+      "loss": 0.7282,
+      "step": 1228
+    },
+    {
+      "epoch": 1.92850487912519,
+      "grad_norm": 0.617905273538838,
+      "learning_rate": 5.9724253359133614e-06,
+      "loss": 0.688,
+      "step": 1229
+    },
+    {
+      "epoch": 1.9300740450154463,
+      "grad_norm": 0.6166033854421795,
+      "learning_rate": 5.956912940101891e-06,
+      "loss": 0.7218,
+      "step": 1230
+    },
+    {
+      "epoch": 1.931643210905703,
+      "grad_norm": 0.6449146850384282,
+      "learning_rate": 5.941412165782645e-06,
+      "loss": 0.6818,
+      "step": 1231
+    },
+    {
+      "epoch": 1.9332123767959595,
+      "grad_norm": 0.6648675935318192,
+      "learning_rate": 5.925923057511207e-06,
+      "loss": 0.7072,
+      "step": 1232
+    },
+    {
+      "epoch": 1.9347815426862158,
+      "grad_norm": 0.6225488767890067,
+      "learning_rate": 5.910445659809638e-06,
+      "loss": 0.6811,
+      "step": 1233
+    },
+    {
+      "epoch": 1.9363507085764722,
+      "grad_norm": 0.6079131353571832,
+      "learning_rate": 5.894980017166329e-06,
+      "loss": 0.6923,
+      "step": 1234
+    },
+    {
+      "epoch": 1.9379198744667288,
+      "grad_norm": 0.6054139293696174,
+      "learning_rate": 5.87952617403589e-06,
+      "loss": 0.7024,
+      "step": 1235
+    },
+    {
+      "epoch": 1.9394890403569853,
+      "grad_norm": 0.6313623365459131,
+      "learning_rate": 5.864084174839013e-06,
+      "loss": 0.6896,
+      "step": 1236
+    },
+    {
+      "epoch": 1.9410582062472417,
+      "grad_norm": 0.6724157702289305,
+      "learning_rate": 5.848654063962338e-06,
+      "loss": 0.7045,
+      "step": 1237
+    },
+    {
+      "epoch": 1.942627372137498,
+      "grad_norm": 0.6486622322371252,
+      "learning_rate": 5.833235885758346e-06,
+      "loss": 0.7116,
+      "step": 1238
+    },
+    {
+      "epoch": 1.9441965380277546,
+      "grad_norm": 0.6690736010146388,
+      "learning_rate": 5.817829684545204e-06,
+      "loss": 0.7081,
+      "step": 1239
+    },
+    {
+      "epoch": 1.9457657039180112,
+      "grad_norm": 0.6395365210454842,
+      "learning_rate": 5.80243550460667e-06,
+      "loss": 0.6985,
+      "step": 1240
+    },
+    {
+      "epoch": 1.9473348698082675,
+      "grad_norm": 0.653059451196709,
+      "learning_rate": 5.787053390191925e-06,
+      "loss": 0.6826,
+      "step": 1241
+    },
+    {
+      "epoch": 1.9489040356985239,
+      "grad_norm": 0.6438436764289824,
+      "learning_rate": 5.771683385515495e-06,
+      "loss": 0.7017,
+      "step": 1242
+    },
+    {
+      "epoch": 1.9504732015887805,
+      "grad_norm": 0.6053209673124681,
+      "learning_rate": 5.75632553475707e-06,
+      "loss": 0.6874,
+      "step": 1243
+    },
+    {
+      "epoch": 1.952042367479037,
+      "grad_norm": 0.6734140575018616,
+      "learning_rate": 5.740979882061433e-06,
+      "loss": 0.7007,
+      "step": 1244
+    },
+    {
+      "epoch": 1.9536115333692934,
+      "grad_norm": 0.650156651309369,
+      "learning_rate": 5.7256464715382754e-06,
+      "loss": 0.7005,
+      "step": 1245
+    },
+    {
+      "epoch": 1.9551806992595497,
+      "grad_norm": 0.6343794182017768,
+      "learning_rate": 5.710325347262125e-06,
+      "loss": 0.6989,
+      "step": 1246
+    },
+    {
+      "epoch": 1.9567498651498063,
+      "grad_norm": 0.6338702298199358,
+      "learning_rate": 5.695016553272182e-06,
+      "loss": 0.6771,
+      "step": 1247
+    },
+    {
+      "epoch": 1.9583190310400629,
+      "grad_norm": 0.6407917999891801,
+      "learning_rate": 5.6797201335722064e-06,
+      "loss": 0.7093,
+      "step": 1248
+    },
+    {
+      "epoch": 1.9598881969303192,
+      "grad_norm": 0.6399231661940882,
+      "learning_rate": 5.664436132130384e-06,
+      "loss": 0.6993,
+      "step": 1249
+    },
+    {
+      "epoch": 1.9614573628205756,
+      "grad_norm": 0.6184857642387768,
+      "learning_rate": 5.649164592879213e-06,
+      "loss": 0.708,
+      "step": 1250
+    },
+    {
+      "epoch": 1.9630265287108322,
+      "grad_norm": 0.6253891495439089,
+      "learning_rate": 5.6339055597153734e-06,
+      "loss": 0.7043,
+      "step": 1251
+    },
+    {
+      "epoch": 1.9645956946010887,
+      "grad_norm": 0.6237149252509561,
+      "learning_rate": 5.618659076499578e-06,
+      "loss": 0.6883,
+      "step": 1252
+    },
+    {
+      "epoch": 1.966164860491345,
+      "grad_norm": 0.6610964719405238,
+      "learning_rate": 5.603425187056496e-06,
+      "loss": 0.6828,
+      "step": 1253
+    },
+    {
+      "epoch": 1.9677340263816014,
+      "grad_norm": 0.6728073741297635,
+      "learning_rate": 5.588203935174563e-06,
+      "loss": 0.7036,
+      "step": 1254
+    },
+    {
+      "epoch": 1.969303192271858,
+      "grad_norm": 0.65038180509265,
+      "learning_rate": 5.572995364605921e-06,
+      "loss": 0.6883,
+      "step": 1255
+    },
+    {
+      "epoch": 1.9708723581621146,
+      "grad_norm": 0.6352491383991571,
+      "learning_rate": 5.55779951906624e-06,
+      "loss": 0.7105,
+      "step": 1256
+    },
+    {
+      "epoch": 1.972441524052371,
+      "grad_norm": 0.6335832091684889,
+      "learning_rate": 5.542616442234618e-06,
+      "loss": 0.7004,
+      "step": 1257
+    },
+    {
+      "epoch": 1.9740106899426273,
+      "grad_norm": 0.5831399595779609,
+      "learning_rate": 5.527446177753455e-06,
+      "loss": 0.6986,
+      "step": 1258
+    },
+    {
+      "epoch": 1.9755798558328839,
+      "grad_norm": 0.610766430097919,
+      "learning_rate": 5.512288769228318e-06,
+      "loss": 0.7071,
+      "step": 1259
+    },
+    {
+      "epoch": 1.9771490217231404,
+      "grad_norm": 0.6782468188201447,
+      "learning_rate": 5.49714426022782e-06,
+      "loss": 0.7053,
+      "step": 1260
+    },
+    {
+      "epoch": 1.9787181876133968,
+      "grad_norm": 0.6364731002267259,
+      "learning_rate": 5.482012694283503e-06,
+      "loss": 0.6756,
+      "step": 1261
+    },
+    {
+      "epoch": 1.9802873535036531,
+      "grad_norm": 0.6209452005657374,
+      "learning_rate": 5.466894114889697e-06,
+      "loss": 0.6814,
+      "step": 1262
+    },
+    {
+      "epoch": 1.9818565193939097,
+      "grad_norm": 0.6252634965080772,
+      "learning_rate": 5.451788565503406e-06,
+      "loss": 0.7067,
+      "step": 1263
+    },
+    {
+      "epoch": 1.9834256852841663,
+      "grad_norm": 0.6515761958953034,
+      "learning_rate": 5.43669608954418e-06,
+      "loss": 0.7025,
+      "step": 1264
+    },
+    {
+      "epoch": 1.9849948511744226,
+      "grad_norm": 0.6395587401705605,
+      "learning_rate": 5.421616730394e-06,
+      "loss": 0.687,
+      "step": 1265
+    },
+    {
+      "epoch": 1.986564017064679,
+      "grad_norm": 0.611002126634849,
+      "learning_rate": 5.406550531397123e-06,
+      "loss": 0.6937,
+      "step": 1266
+    },
+    {
+      "epoch": 1.9881331829549356,
+      "grad_norm": 0.6403543434666442,
+      "learning_rate": 5.391497535859999e-06,
+      "loss": 0.6971,
+      "step": 1267
+    },
+    {
+      "epoch": 1.989702348845192,
+      "grad_norm": 0.6403736195000795,
+      "learning_rate": 5.376457787051117e-06,
+      "loss": 0.6933,
+      "step": 1268
+    },
+    {
+      "epoch": 1.9912715147354483,
+      "grad_norm": 0.5968961618757637,
+      "learning_rate": 5.361431328200888e-06,
+      "loss": 0.6731,
+      "step": 1269
+    },
+    {
+      "epoch": 1.9928406806257049,
+      "grad_norm": 0.6478190412143306,
+      "learning_rate": 5.346418202501526e-06,
+      "loss": 0.6992,
+      "step": 1270
+    },
+    {
+      "epoch": 1.9944098465159614,
+      "grad_norm": 0.6327793952156039,
+      "learning_rate": 5.331418453106916e-06,
+      "loss": 0.6892,
+      "step": 1271
+    },
+    {
+      "epoch": 1.9959790124062178,
+      "grad_norm": 0.6470153466147767,
+      "learning_rate": 5.316432123132498e-06,
+      "loss": 0.684,
+      "step": 1272
+    },
+    {
+      "epoch": 1.9975481782964741,
+      "grad_norm": 0.61029701971341,
+      "learning_rate": 5.301459255655134e-06,
+      "loss": 0.702,
+      "step": 1273
+    },
+    {
+      "epoch": 1.9991173441867307,
+      "grad_norm": 0.6172624667831884,
+      "learning_rate": 5.286499893713003e-06,
+      "loss": 0.6801,
+      "step": 1274
+    },
+    {
+      "epoch": 2.0006865100769873,
+      "grad_norm": 0.6502943362915738,
+      "learning_rate": 5.271554080305439e-06,
+      "loss": 0.6862,
+      "step": 1275
+    },
+    {
+      "epoch": 2.002255675967244,
+      "grad_norm": 0.6532695834133205,
+      "learning_rate": 5.256621858392861e-06,
+      "loss": 0.645,
+      "step": 1276
+    },
+    {
+      "epoch": 2.0038248418575,
+      "grad_norm": 0.679386923860287,
+      "learning_rate": 5.241703270896589e-06,
+      "loss": 0.6094,
+      "step": 1277
+    },
+    {
+      "epoch": 2.0053940077477566,
+      "grad_norm": 0.5922442173268206,
+      "learning_rate": 5.226798360698786e-06,
+      "loss": 0.593,
+      "step": 1278
+    },
+    {
+      "epoch": 2.006963173638013,
+      "grad_norm": 0.6207292942961803,
+      "learning_rate": 5.211907170642267e-06,
+      "loss": 0.6396,
+      "step": 1279
+    },
+    {
+      "epoch": 2.0085323395282697,
+      "grad_norm": 0.6938543608650418,
+      "learning_rate": 5.1970297435304395e-06,
+      "loss": 0.6225,
+      "step": 1280
+    },
+    {
+      "epoch": 2.010101505418526,
+      "grad_norm": 0.6951419788663918,
+      "learning_rate": 5.18216612212713e-06,
+      "loss": 0.633,
+      "step": 1281
+    },
+    {
+      "epoch": 2.0116706713087824,
+      "grad_norm": 0.6685786850904167,
+      "learning_rate": 5.167316349156495e-06,
+      "loss": 0.6046,
+      "step": 1282
+    },
+    {
+      "epoch": 2.013239837199039,
+      "grad_norm": 0.6525886452405972,
+      "learning_rate": 5.1524804673028725e-06,
+      "loss": 0.636,
+      "step": 1283
+    },
+    {
+      "epoch": 2.0148090030892956,
+      "grad_norm": 0.6444216013425876,
+      "learning_rate": 5.1376585192106824e-06,
+      "loss": 0.6379,
+      "step": 1284
+    },
+    {
+      "epoch": 2.0163781689795517,
+      "grad_norm": 0.6223171786423164,
+      "learning_rate": 5.1228505474842885e-06,
+      "loss": 0.6328,
+      "step": 1285
+    },
+    {
+      "epoch": 2.0179473348698083,
+      "grad_norm": 0.5954563203202149,
+      "learning_rate": 5.10805659468788e-06,
+      "loss": 0.6072,
+      "step": 1286
+    },
+    {
+      "epoch": 2.019516500760065,
+      "grad_norm": 0.609285786041519,
+      "learning_rate": 5.093276703345351e-06,
+      "loss": 0.6329,
+      "step": 1287
+    },
+    {
+      "epoch": 2.021085666650321,
+      "grad_norm": 0.6138686877469576,
+      "learning_rate": 5.078510915940178e-06,
+      "loss": 0.6145,
+      "step": 1288
+    },
+    {
+      "epoch": 2.0226548325405775,
+      "grad_norm": 0.6359266591864806,
+      "learning_rate": 5.063759274915299e-06,
+      "loss": 0.6162,
+      "step": 1289
+    },
+    {
+      "epoch": 2.024223998430834,
+      "grad_norm": 0.5900005292856023,
+      "learning_rate": 5.0490218226729855e-06,
+      "loss": 0.6036,
+      "step": 1290
+    },
+    {
+      "epoch": 2.0257931643210907,
+      "grad_norm": 0.647771565313278,
+      "learning_rate": 5.034298601574726e-06,
+      "loss": 0.6316,
+      "step": 1291
+    },
+    {
+      "epoch": 2.027362330211347,
+      "grad_norm": 0.6180522481125589,
+      "learning_rate": 5.019589653941103e-06,
+      "loss": 0.6229,
+      "step": 1292
+    },
+    {
+      "epoch": 2.0289314961016034,
+      "grad_norm": 0.6121145229950486,
+      "learning_rate": 5.0048950220516725e-06,
+      "loss": 0.6297,
+      "step": 1293
+    },
+    {
+      "epoch": 2.03050066199186,
+      "grad_norm": 0.6516540867206794,
+      "learning_rate": 4.990214748144837e-06,
+      "loss": 0.6256,
+      "step": 1294
+    },
+    {
+      "epoch": 2.0320698278821165,
+      "grad_norm": 0.6896438757414021,
+      "learning_rate": 4.975548874417733e-06,
+      "loss": 0.639,
+      "step": 1295
+    },
+    {
+      "epoch": 2.0336389937723727,
+      "grad_norm": 0.6546359792448009,
+      "learning_rate": 4.9608974430261025e-06,
+      "loss": 0.614,
+      "step": 1296
+    },
+    {
+      "epoch": 2.0352081596626292,
+      "grad_norm": 0.6495885745428986,
+      "learning_rate": 4.946260496084175e-06,
+      "loss": 0.6188,
+      "step": 1297
+    },
+    {
+      "epoch": 2.036777325552886,
+      "grad_norm": 0.6670220750986432,
+      "learning_rate": 4.931638075664544e-06,
+      "loss": 0.6279,
+      "step": 1298
+    },
+    {
+      "epoch": 2.0383464914431424,
+      "grad_norm": 0.6466837858565543,
+      "learning_rate": 4.917030223798057e-06,
+      "loss": 0.6226,
+      "step": 1299
+    },
+    {
+      "epoch": 2.0399156573333985,
+      "grad_norm": 0.6252366671151887,
+      "learning_rate": 4.9024369824736654e-06,
+      "loss": 0.6166,
+      "step": 1300
+    },
+    {
+      "epoch": 2.041484823223655,
+      "grad_norm": 0.618641258105832,
+      "learning_rate": 4.88785839363835e-06,
+      "loss": 0.6072,
+      "step": 1301
+    },
+    {
+      "epoch": 2.0430539891139117,
+      "grad_norm": 0.6261693841939681,
+      "learning_rate": 4.873294499196946e-06,
+      "loss": 0.6112,
+      "step": 1302
+    },
+    {
+      "epoch": 2.0446231550041682,
+      "grad_norm": 0.6374330661267035,
+      "learning_rate": 4.85874534101208e-06,
+      "loss": 0.6188,
+      "step": 1303
+    },
+    {
+      "epoch": 2.0461923208944244,
+      "grad_norm": 0.6451208890553017,
+      "learning_rate": 4.844210960903998e-06,
+      "loss": 0.6146,
+      "step": 1304
+    },
+    {
+      "epoch": 2.047761486784681,
+      "grad_norm": 0.6171151222356629,
+      "learning_rate": 4.829691400650478e-06,
+      "loss": 0.6246,
+      "step": 1305
+    },
+    {
+      "epoch": 2.0493306526749375,
+      "grad_norm": 0.6209394509604264,
+      "learning_rate": 4.8151867019866985e-06,
+      "loss": 0.6135,
+      "step": 1306
+    },
+    {
+      "epoch": 2.050899818565194,
+      "grad_norm": 0.6479307989678751,
+      "learning_rate": 4.800696906605117e-06,
+      "loss": 0.6291,
+      "step": 1307
+    },
+    {
+      "epoch": 2.0524689844554502,
+      "grad_norm": 0.626833864075317,
+      "learning_rate": 4.786222056155356e-06,
+      "loss": 0.6297,
+      "step": 1308
+    },
+    {
+      "epoch": 2.054038150345707,
+      "grad_norm": 0.6884180450327059,
+      "learning_rate": 4.771762192244075e-06,
+      "loss": 0.6318,
+      "step": 1309
+    },
+    {
+      "epoch": 2.0556073162359634,
+      "grad_norm": 0.5961981176872094,
+      "learning_rate": 4.757317356434869e-06,
+      "loss": 0.6143,
+      "step": 1310
+    },
+    {
+      "epoch": 2.05717648212622,
+      "grad_norm": 0.6240117691297208,
+      "learning_rate": 4.742887590248114e-06,
+      "loss": 0.6174,
+      "step": 1311
+    },
+    {
+      "epoch": 2.058745648016476,
+      "grad_norm": 0.6225306028664694,
+      "learning_rate": 4.7284729351608985e-06,
+      "loss": 0.6437,
+      "step": 1312
+    },
+    {
+      "epoch": 2.0603148139067327,
+      "grad_norm": 0.6769715250396284,
+      "learning_rate": 4.714073432606841e-06,
+      "loss": 0.6149,
+      "step": 1313
+    },
+    {
+      "epoch": 2.0618839797969892,
+      "grad_norm": 0.6330285259582916,
+      "learning_rate": 4.699689123976039e-06,
+      "loss": 0.6218,
+      "step": 1314
+    },
+    {
+      "epoch": 2.063453145687246,
+      "grad_norm": 0.5975271077189797,
+      "learning_rate": 4.6853200506148955e-06,
+      "loss": 0.6238,
+      "step": 1315
+    },
+    {
+      "epoch": 2.065022311577502,
+      "grad_norm": 0.6462635721840223,
+      "learning_rate": 4.670966253826027e-06,
+      "loss": 0.5896,
+      "step": 1316
+    },
+    {
+      "epoch": 2.0665914774677585,
+      "grad_norm": 0.6577166325621118,
+      "learning_rate": 4.65662777486814e-06,
+      "loss": 0.6292,
+      "step": 1317
+    },
+    {
+      "epoch": 2.068160643358015,
+      "grad_norm": 0.6566413920626784,
+      "learning_rate": 4.6423046549559095e-06,
+      "loss": 0.6286,
+      "step": 1318
+    },
+    {
+      "epoch": 2.0697298092482717,
+      "grad_norm": 0.6238734664689921,
+      "learning_rate": 4.627996935259862e-06,
+      "loss": 0.6091,
+      "step": 1319
+    },
+    {
+      "epoch": 2.071298975138528,
+      "grad_norm": 0.5806528818105448,
+      "learning_rate": 4.6137046569062595e-06,
+      "loss": 0.6239,
+      "step": 1320
+    },
+    {
+      "epoch": 2.0728681410287844,
+      "grad_norm": 0.6226478025987044,
+      "learning_rate": 4.599427860976978e-06,
+      "loss": 0.6304,
+      "step": 1321
+    },
+    {
+      "epoch": 2.074437306919041,
+      "grad_norm": 0.6440113133813798,
+      "learning_rate": 4.585166588509391e-06,
+      "loss": 0.6233,
+      "step": 1322
+    },
+    {
+      "epoch": 2.0760064728092975,
+      "grad_norm": 0.6380787166210546,
+      "learning_rate": 4.570920880496247e-06,
+      "loss": 0.6083,
+      "step": 1323
+    },
+    {
+      "epoch": 2.0775756386995536,
+      "grad_norm": 0.6352201664360289,
+      "learning_rate": 4.556690777885571e-06,
+      "loss": 0.6094,
+      "step": 1324
+    },
+    {
+      "epoch": 2.07914480458981,
+      "grad_norm": 0.6340091351803824,
+      "learning_rate": 4.542476321580505e-06,
+      "loss": 0.612,
+      "step": 1325
+    },
+    {
+      "epoch": 2.080713970480067,
+      "grad_norm": 0.6379808320273905,
+      "learning_rate": 4.528277552439244e-06,
+      "loss": 0.6369,
+      "step": 1326
+    },
+    {
+      "epoch": 2.0822831363703234,
+      "grad_norm": 0.6182812260685615,
+      "learning_rate": 4.514094511274877e-06,
+      "loss": 0.6165,
+      "step": 1327
+    },
+    {
+      "epoch": 2.0838523022605795,
+      "grad_norm": 0.6373230561822038,
+      "learning_rate": 4.499927238855287e-06,
+      "loss": 0.6084,
+      "step": 1328
+    },
+    {
+      "epoch": 2.085421468150836,
+      "grad_norm": 0.6409460157675354,
+      "learning_rate": 4.485775775903033e-06,
+      "loss": 0.639,
+      "step": 1329
+    },
+    {
+      "epoch": 2.0869906340410926,
+      "grad_norm": 0.6282039128609315,
+      "learning_rate": 4.471640163095229e-06,
+      "loss": 0.6157,
+      "step": 1330
+    },
+    {
+      "epoch": 2.0885597999313488,
+      "grad_norm": 0.6137741793218672,
+      "learning_rate": 4.457520441063428e-06,
+      "loss": 0.6124,
+      "step": 1331
+    },
+    {
+      "epoch": 2.0901289658216053,
+      "grad_norm": 0.681372423604883,
+      "learning_rate": 4.443416650393509e-06,
+      "loss": 0.6013,
+      "step": 1332
+    },
+    {
+      "epoch": 2.091698131711862,
+      "grad_norm": 0.6590385412311799,
+      "learning_rate": 4.429328831625565e-06,
+      "loss": 0.6252,
+      "step": 1333
+    },
+    {
+      "epoch": 2.0932672976021185,
+      "grad_norm": 0.6197618086568621,
+      "learning_rate": 4.415257025253758e-06,
+      "loss": 0.6033,
+      "step": 1334
+    },
+    {
+      "epoch": 2.0948364634923746,
+      "grad_norm": 0.6156000017412134,
+      "learning_rate": 4.40120127172625e-06,
+      "loss": 0.5991,
+      "step": 1335
+    },
+    {
+      "epoch": 2.096405629382631,
+      "grad_norm": 0.664306039819029,
+      "learning_rate": 4.387161611445034e-06,
+      "loss": 0.6217,
+      "step": 1336
+    },
+    {
+      "epoch": 2.0979747952728878,
+      "grad_norm": 0.6611599761889236,
+      "learning_rate": 4.373138084765875e-06,
+      "loss": 0.618,
+      "step": 1337
+    },
+    {
+      "epoch": 2.0995439611631443,
+      "grad_norm": 0.6222404404322623,
+      "learning_rate": 4.359130731998127e-06,
+      "loss": 0.6288,
+      "step": 1338
+    },
+    {
+      "epoch": 2.1011131270534005,
+      "grad_norm": 0.6311429425363364,
+      "learning_rate": 4.345139593404688e-06,
+      "loss": 0.6184,
+      "step": 1339
+    },
+    {
+      "epoch": 2.102682292943657,
+      "grad_norm": 0.6445149714231355,
+      "learning_rate": 4.331164709201829e-06,
+      "loss": 0.6249,
+      "step": 1340
+    },
+    {
+      "epoch": 2.1042514588339136,
+      "grad_norm": 0.6173377175050655,
+      "learning_rate": 4.317206119559108e-06,
+      "loss": 0.6219,
+      "step": 1341
+    },
+    {
+      "epoch": 2.10582062472417,
+      "grad_norm": 0.6342165323182424,
+      "learning_rate": 4.303263864599239e-06,
+      "loss": 0.6175,
+      "step": 1342
+    },
+    {
+      "epoch": 2.1073897906144263,
+      "grad_norm": 0.6290617791208115,
+      "learning_rate": 4.289337984397988e-06,
+      "loss": 0.6151,
+      "step": 1343
+    },
+    {
+      "epoch": 2.108958956504683,
+      "grad_norm": 0.6403426983817079,
+      "learning_rate": 4.275428518984054e-06,
+      "loss": 0.6248,
+      "step": 1344
+    },
+    {
+      "epoch": 2.1105281223949395,
+      "grad_norm": 0.6560238644443025,
+      "learning_rate": 4.261535508338951e-06,
+      "loss": 0.5888,
+      "step": 1345
+    },
+    {
+      "epoch": 2.112097288285196,
+      "grad_norm": 0.6341580814633958,
+      "learning_rate": 4.2476589923968946e-06,
+      "loss": 0.6214,
+      "step": 1346
+    },
+    {
+      "epoch": 2.113666454175452,
+      "grad_norm": 0.6611016260475742,
+      "learning_rate": 4.233799011044687e-06,
+      "loss": 0.6371,
+      "step": 1347
+    },
+    {
+      "epoch": 2.1152356200657088,
+      "grad_norm": 0.6085285537758993,
+      "learning_rate": 4.219955604121611e-06,
+      "loss": 0.6333,
+      "step": 1348
+    },
+    {
+      "epoch": 2.1168047859559653,
+      "grad_norm": 0.6262253705752535,
+      "learning_rate": 4.206128811419299e-06,
+      "loss": 0.6215,
+      "step": 1349
+    },
+    {
+      "epoch": 2.118373951846222,
+      "grad_norm": 0.6286833131008006,
+      "learning_rate": 4.192318672681631e-06,
+      "loss": 0.6319,
+      "step": 1350
+    },
+    {
+      "epoch": 2.119943117736478,
+      "grad_norm": 0.6150093205982207,
+      "learning_rate": 4.178525227604615e-06,
+      "loss": 0.6107,
+      "step": 1351
+    },
+    {
+      "epoch": 2.1215122836267346,
+      "grad_norm": 0.5989661130072645,
+      "learning_rate": 4.164748515836276e-06,
+      "loss": 0.6257,
+      "step": 1352
+    },
+    {
+      "epoch": 2.123081449516991,
+      "grad_norm": 0.6149175273364935,
+      "learning_rate": 4.150988576976539e-06,
+      "loss": 0.6256,
+      "step": 1353
+    },
+    {
+      "epoch": 2.1246506154072478,
+      "grad_norm": 0.6534618762641019,
+      "learning_rate": 4.1372454505771186e-06,
+      "loss": 0.6286,
+      "step": 1354
+    },
+    {
+      "epoch": 2.126219781297504,
+      "grad_norm": 0.6149216597100641,
+      "learning_rate": 4.1235191761414025e-06,
+      "loss": 0.6326,
+      "step": 1355
+    },
+    {
+      "epoch": 2.1277889471877605,
+      "grad_norm": 0.6293541320131315,
+      "learning_rate": 4.10980979312434e-06,
+      "loss": 0.6046,
+      "step": 1356
+    },
+    {
+      "epoch": 2.129358113078017,
+      "grad_norm": 0.654792487329429,
+      "learning_rate": 4.096117340932322e-06,
+      "loss": 0.6081,
+      "step": 1357
+    },
+    {
+      "epoch": 2.1309272789682736,
+      "grad_norm": 0.6733007709475541,
+      "learning_rate": 4.082441858923088e-06,
+      "loss": 0.6375,
+      "step": 1358
+    },
+    {
+      "epoch": 2.1324964448585297,
+      "grad_norm": 0.6067073291253214,
+      "learning_rate": 4.068783386405577e-06,
+      "loss": 0.6242,
+      "step": 1359
+    },
+    {
+      "epoch": 2.1340656107487863,
+      "grad_norm": 0.6350207595746162,
+      "learning_rate": 4.055141962639858e-06,
+      "loss": 0.6089,
+      "step": 1360
+    },
+    {
+      "epoch": 2.135634776639043,
+      "grad_norm": 0.6503175818174727,
+      "learning_rate": 4.0415176268369714e-06,
+      "loss": 0.6093,
+      "step": 1361
+    },
+    {
+      "epoch": 2.1372039425292995,
+      "grad_norm": 0.6552626052070196,
+      "learning_rate": 4.027910418158861e-06,
+      "loss": 0.6177,
+      "step": 1362
+    },
+    {
+      "epoch": 2.1387731084195556,
+      "grad_norm": 0.607160411260159,
+      "learning_rate": 4.01432037571823e-06,
+      "loss": 0.594,
+      "step": 1363
+    },
+    {
+      "epoch": 2.140342274309812,
+      "grad_norm": 0.6141818558596257,
+      "learning_rate": 4.00074753857844e-06,
+      "loss": 0.6302,
+      "step": 1364
+    },
+    {
+      "epoch": 2.1419114402000687,
+      "grad_norm": 0.6539732506640329,
+      "learning_rate": 3.987191945753396e-06,
+      "loss": 0.6071,
+      "step": 1365
+    },
+    {
+      "epoch": 2.1434806060903253,
+      "grad_norm": 0.6262626578183587,
+      "learning_rate": 3.973653636207437e-06,
+      "loss": 0.6164,
+      "step": 1366
+    },
+    {
+      "epoch": 2.1450497719805814,
+      "grad_norm": 0.6445800626193894,
+      "learning_rate": 3.960132648855226e-06,
+      "loss": 0.6101,
+      "step": 1367
+    },
+    {
+      "epoch": 2.146618937870838,
+      "grad_norm": 0.6075052382718521,
+      "learning_rate": 3.946629022561627e-06,
+      "loss": 0.6229,
+      "step": 1368
+    },
+    {
+      "epoch": 2.1481881037610946,
+      "grad_norm": 0.6296286988783455,
+      "learning_rate": 3.933142796141616e-06,
+      "loss": 0.6257,
+      "step": 1369
+    },
+    {
+      "epoch": 2.149757269651351,
+      "grad_norm": 0.6530009437245898,
+      "learning_rate": 3.919674008360133e-06,
+      "loss": 0.6243,
+      "step": 1370
+    },
+    {
+      "epoch": 2.1513264355416073,
+      "grad_norm": 0.6348907597378193,
+      "learning_rate": 3.906222697932016e-06,
+      "loss": 0.6174,
+      "step": 1371
+    },
+    {
+      "epoch": 2.152895601431864,
+      "grad_norm": 0.6318441977343,
+      "learning_rate": 3.89278890352184e-06,
+      "loss": 0.6223,
+      "step": 1372
+    },
+    {
+      "epoch": 2.1544647673221204,
+      "grad_norm": 0.6489749457884836,
+      "learning_rate": 3.879372663743862e-06,
+      "loss": 0.5928,
+      "step": 1373
+    },
+    {
+      "epoch": 2.1560339332123766,
+      "grad_norm": 0.6382645515107063,
+      "learning_rate": 3.8659740171618485e-06,
+      "loss": 0.6193,
+      "step": 1374
+    },
+    {
+      "epoch": 2.157603099102633,
+      "grad_norm": 0.6608734648068885,
+      "learning_rate": 3.85259300228902e-06,
+      "loss": 0.6149,
+      "step": 1375
+    },
+    {
+      "epoch": 2.1591722649928897,
+      "grad_norm": 0.6211503702573417,
+      "learning_rate": 3.839229657587906e-06,
+      "loss": 0.6152,
+      "step": 1376
+    },
+    {
+      "epoch": 2.1607414308831463,
+      "grad_norm": 0.6476283598850665,
+      "learning_rate": 3.825884021470244e-06,
+      "loss": 0.6096,
+      "step": 1377
+    },
+    {
+      "epoch": 2.162310596773403,
+      "grad_norm": 0.626403655602433,
+      "learning_rate": 3.812556132296873e-06,
+      "loss": 0.615,
+      "step": 1378
+    },
+    {
+      "epoch": 2.163879762663659,
+      "grad_norm": 0.6153082963429655,
+      "learning_rate": 3.799246028377618e-06,
+      "loss": 0.6154,
+      "step": 1379
+    },
+    {
+      "epoch": 2.1654489285539156,
+      "grad_norm": 0.6038055156462527,
+      "learning_rate": 3.785953747971183e-06,
+      "loss": 0.6146,
+      "step": 1380
+    },
+    {
+      "epoch": 2.167018094444172,
+      "grad_norm": 0.6580520250662772,
+      "learning_rate": 3.7726793292850382e-06,
+      "loss": 0.6098,
+      "step": 1381
+    },
+    {
+      "epoch": 2.1685872603344283,
+      "grad_norm": 0.6288998309682243,
+      "learning_rate": 3.7594228104753107e-06,
+      "loss": 0.5862,
+      "step": 1382
+    },
+    {
+      "epoch": 2.170156426224685,
+      "grad_norm": 0.6313827812211942,
+      "learning_rate": 3.746184229646689e-06,
+      "loss": 0.6177,
+      "step": 1383
+    },
+    {
+      "epoch": 2.1717255921149414,
+      "grad_norm": 0.6599611497599748,
+      "learning_rate": 3.732963624852275e-06,
+      "loss": 0.608,
+      "step": 1384
+    },
+    {
+      "epoch": 2.173294758005198,
+      "grad_norm": 0.6498663474744056,
+      "learning_rate": 3.7197610340935263e-06,
+      "loss": 0.6186,
+      "step": 1385
+    },
+    {
+      "epoch": 2.174863923895454,
+      "grad_norm": 0.6838926563034045,
+      "learning_rate": 3.7065764953201066e-06,
+      "loss": 0.6212,
+      "step": 1386
+    },
+    {
+      "epoch": 2.1764330897857107,
+      "grad_norm": 0.6344757183599451,
+      "learning_rate": 3.693410046429793e-06,
+      "loss": 0.6334,
+      "step": 1387
+    },
+    {
+      "epoch": 2.1780022556759673,
+      "grad_norm": 0.6261863487618142,
+      "learning_rate": 3.680261725268364e-06,
+      "loss": 0.6151,
+      "step": 1388
+    },
+    {
+      "epoch": 2.179571421566224,
+      "grad_norm": 0.6547877002885869,
+      "learning_rate": 3.667131569629495e-06,
+      "loss": 0.6157,
+      "step": 1389
+    },
+    {
+      "epoch": 2.18114058745648,
+      "grad_norm": 0.6262514821781034,
+      "learning_rate": 3.654019617254644e-06,
+      "loss": 0.6096,
+      "step": 1390
+    },
+    {
+      "epoch": 2.1827097533467366,
+      "grad_norm": 0.6087299134232946,
+      "learning_rate": 3.6409259058329404e-06,
+      "loss": 0.6301,
+      "step": 1391
+    },
+    {
+      "epoch": 2.184278919236993,
+      "grad_norm": 0.6345458747487406,
+      "learning_rate": 3.6278504730011e-06,
+      "loss": 0.6284,
+      "step": 1392
+    },
+    {
+      "epoch": 2.1858480851272497,
+      "grad_norm": 0.6048122479797837,
+      "learning_rate": 3.614793356343269e-06,
+      "loss": 0.6218,
+      "step": 1393
+    },
+    {
+      "epoch": 2.187417251017506,
+      "grad_norm": 0.6125538089626119,
+      "learning_rate": 3.601754593390978e-06,
+      "loss": 0.61,
+      "step": 1394
+    },
+    {
+      "epoch": 2.1889864169077624,
+      "grad_norm": 0.7803536148969056,
+      "learning_rate": 3.5887342216229694e-06,
+      "loss": 0.6115,
+      "step": 1395
+    },
+    {
+      "epoch": 2.190555582798019,
+      "grad_norm": 0.6226438132889969,
+      "learning_rate": 3.5757322784651506e-06,
+      "loss": 0.6065,
+      "step": 1396
+    },
+    {
+      "epoch": 2.1921247486882756,
+      "grad_norm": 0.6216705847894933,
+      "learning_rate": 3.5627488012904333e-06,
+      "loss": 0.6003,
+      "step": 1397
+    },
+    {
+      "epoch": 2.1936939145785317,
+      "grad_norm": 0.6046800762296665,
+      "learning_rate": 3.5497838274186737e-06,
+      "loss": 0.6025,
+      "step": 1398
+    },
+    {
+      "epoch": 2.1952630804687883,
+      "grad_norm": 0.6362031206869132,
+      "learning_rate": 3.5368373941165167e-06,
+      "loss": 0.6301,
+      "step": 1399
+    },
+    {
+      "epoch": 2.196832246359045,
+      "grad_norm": 0.668758539321539,
+      "learning_rate": 3.5239095385973377e-06,
+      "loss": 0.6134,
+      "step": 1400
+    },
+    {
+      "epoch": 2.1984014122493014,
+      "grad_norm": 0.6403967472883619,
+      "learning_rate": 3.511000298021098e-06,
+      "loss": 0.6143,
+      "step": 1401
+    },
+    {
+      "epoch": 2.1999705781395575,
+      "grad_norm": 0.5970553908619288,
+      "learning_rate": 3.498109709494254e-06,
+      "loss": 0.607,
+      "step": 1402
+    },
+    {
+      "epoch": 2.201539744029814,
+      "grad_norm": 0.6227702242395404,
+      "learning_rate": 3.4852378100696505e-06,
+      "loss": 0.628,
+      "step": 1403
+    },
+    {
+      "epoch": 2.2031089099200707,
+      "grad_norm": 0.6338408588806297,
+      "learning_rate": 3.472384636746412e-06,
+      "loss": 0.6198,
+      "step": 1404
+    },
+    {
+      "epoch": 2.2046780758103273,
+      "grad_norm": 0.6376720106825551,
+      "learning_rate": 3.4595502264698356e-06,
+      "loss": 0.6078,
+      "step": 1405
+    },
+    {
+      "epoch": 2.2062472417005834,
+      "grad_norm": 0.6306285832000212,
+      "learning_rate": 3.446734616131282e-06,
+      "loss": 0.625,
+      "step": 1406
+    },
+    {
+      "epoch": 2.20781640759084,
+      "grad_norm": 0.6226892428730488,
+      "learning_rate": 3.433937842568088e-06,
+      "loss": 0.622,
+      "step": 1407
+    },
+    {
+      "epoch": 2.2093855734810965,
+      "grad_norm": 0.6207352149813079,
+      "learning_rate": 3.421159942563421e-06,
+      "loss": 0.6128,
+      "step": 1408
+    },
+    {
+      "epoch": 2.210954739371353,
+      "grad_norm": 0.6222677005956953,
+      "learning_rate": 3.4084009528462227e-06,
+      "loss": 0.6137,
+      "step": 1409
+    },
+    {
+      "epoch": 2.2125239052616092,
+      "grad_norm": 0.6186315547078428,
+      "learning_rate": 3.3956609100910643e-06,
+      "loss": 0.6094,
+      "step": 1410
+    },
+    {
+      "epoch": 2.214093071151866,
+      "grad_norm": 0.6329189874983667,
+      "learning_rate": 3.382939850918059e-06,
+      "loss": 0.6222,
+      "step": 1411
+    },
+    {
+      "epoch": 2.2156622370421224,
+      "grad_norm": 0.6344490203147958,
+      "learning_rate": 3.3702378118927537e-06,
+      "loss": 0.6111,
+      "step": 1412
+    },
+    {
+      "epoch": 2.217231402932379,
+      "grad_norm": 0.6537806725501569,
+      "learning_rate": 3.357554829526025e-06,
+      "loss": 0.6263,
+      "step": 1413
+    },
+    {
+      "epoch": 2.218800568822635,
+      "grad_norm": 0.6101179396070554,
+      "learning_rate": 3.3448909402739716e-06,
+      "loss": 0.6087,
+      "step": 1414
+    },
+    {
+      "epoch": 2.2203697347128917,
+      "grad_norm": 0.6150161152521031,
+      "learning_rate": 3.3322461805378094e-06,
+      "loss": 0.6204,
+      "step": 1415
+    },
+    {
+      "epoch": 2.2219389006031482,
+      "grad_norm": 0.6425715998201768,
+      "learning_rate": 3.3196205866637667e-06,
+      "loss": 0.6307,
+      "step": 1416
+    },
+    {
+      "epoch": 2.2235080664934044,
+      "grad_norm": 0.634070363391312,
+      "learning_rate": 3.3070141949429945e-06,
+      "loss": 0.6048,
+      "step": 1417
+    },
+    {
+      "epoch": 2.225077232383661,
+      "grad_norm": 0.6165257938318652,
+      "learning_rate": 3.2944270416114256e-06,
+      "loss": 0.6407,
+      "step": 1418
+    },
+    {
+      "epoch": 2.2266463982739175,
+      "grad_norm": 0.6323362993027953,
+      "learning_rate": 3.28185916284972e-06,
+      "loss": 0.6185,
+      "step": 1419
+    },
+    {
+      "epoch": 2.228215564164174,
+      "grad_norm": 0.6564535829246948,
+      "learning_rate": 3.269310594783106e-06,
+      "loss": 0.6008,
+      "step": 1420
+    },
+    {
+      "epoch": 2.2297847300544307,
+      "grad_norm": 0.5872342949393278,
+      "learning_rate": 3.2567813734813336e-06,
+      "loss": 0.6105,
+      "step": 1421
+    },
+    {
+      "epoch": 2.231353895944687,
+      "grad_norm": 0.6310778473204862,
+      "learning_rate": 3.244271534958523e-06,
+      "loss": 0.6205,
+      "step": 1422
+    },
+    {
+      "epoch": 2.2329230618349434,
+      "grad_norm": 0.6600569488406903,
+      "learning_rate": 3.231781115173088e-06,
+      "loss": 0.6053,
+      "step": 1423
+    },
+    {
+      "epoch": 2.2344922277252,
+      "grad_norm": 0.6288294461832861,
+      "learning_rate": 3.2193101500276226e-06,
+      "loss": 0.6173,
+      "step": 1424
+    },
+    {
+      "epoch": 2.236061393615456,
+      "grad_norm": 0.6590897159665384,
+      "learning_rate": 3.2068586753688014e-06,
+      "loss": 0.6276,
+      "step": 1425
+    },
+    {
+      "epoch": 2.2376305595057127,
+      "grad_norm": 0.6163557437757586,
+      "learning_rate": 3.1944267269872754e-06,
+      "loss": 0.6128,
+      "step": 1426
+    },
+    {
+      "epoch": 2.2391997253959692,
+      "grad_norm": 0.6144722120020086,
+      "learning_rate": 3.1820143406175642e-06,
+      "loss": 0.5974,
+      "step": 1427
+    },
+    {
+      "epoch": 2.240768891286226,
+      "grad_norm": 0.6199432490099716,
+      "learning_rate": 3.1696215519379726e-06,
+      "loss": 0.6155,
+      "step": 1428
+    },
+    {
+      "epoch": 2.242338057176482,
+      "grad_norm": 0.6419359527546662,
+      "learning_rate": 3.1572483965704503e-06,
+      "loss": 0.6304,
+      "step": 1429
+    },
+    {
+      "epoch": 2.2439072230667385,
+      "grad_norm": 0.6285531009181134,
+      "learning_rate": 3.1448949100805394e-06,
+      "loss": 0.6251,
+      "step": 1430
+    },
+    {
+      "epoch": 2.245476388956995,
+      "grad_norm": 0.6596871033902281,
+      "learning_rate": 3.1325611279772194e-06,
+      "loss": 0.6183,
+      "step": 1431
+    },
+    {
+      "epoch": 2.2470455548472517,
+      "grad_norm": 0.617400824424676,
+      "learning_rate": 3.1202470857128564e-06,
+      "loss": 0.6201,
+      "step": 1432
+    },
+    {
+      "epoch": 2.248614720737508,
+      "grad_norm": 0.6000444558853422,
+      "learning_rate": 3.107952818683052e-06,
+      "loss": 0.6211,
+      "step": 1433
+    },
+    {
+      "epoch": 2.2501838866277644,
+      "grad_norm": 0.6151810101126691,
+      "learning_rate": 3.095678362226585e-06,
+      "loss": 0.6163,
+      "step": 1434
+    },
+    {
+      "epoch": 2.251753052518021,
+      "grad_norm": 0.5995386290624433,
+      "learning_rate": 3.0834237516252817e-06,
+      "loss": 0.6149,
+      "step": 1435
+    },
+    {
+      "epoch": 2.2533222184082775,
+      "grad_norm": 0.6386655256355857,
+      "learning_rate": 3.071189022103923e-06,
+      "loss": 0.6298,
+      "step": 1436
+    },
+    {
+      "epoch": 2.2548913842985336,
+      "grad_norm": 0.6376300732459658,
+      "learning_rate": 3.058974208830147e-06,
+      "loss": 0.6257,
+      "step": 1437
+    },
+    {
+      "epoch": 2.25646055018879,
+      "grad_norm": 0.6798538813287663,
+      "learning_rate": 3.046779346914339e-06,
+      "loss": 0.6193,
+      "step": 1438
+    },
+    {
+      "epoch": 2.258029716079047,
+      "grad_norm": 0.6397045527151177,
+      "learning_rate": 3.0346044714095425e-06,
+      "loss": 0.6144,
+      "step": 1439
+    },
+    {
+      "epoch": 2.2595988819693034,
+      "grad_norm": 0.6266078173704976,
+      "learning_rate": 3.022449617311346e-06,
+      "loss": 0.6109,
+      "step": 1440
+    },
+    {
+      "epoch": 2.2611680478595595,
+      "grad_norm": 0.6278434253419796,
+      "learning_rate": 3.0103148195577914e-06,
+      "loss": 0.6369,
+      "step": 1441
+    },
+    {
+      "epoch": 2.262737213749816,
+      "grad_norm": 0.6194184838780724,
+      "learning_rate": 2.9982001130292703e-06,
+      "loss": 0.5819,
+      "step": 1442
+    },
+    {
+      "epoch": 2.2643063796400726,
+      "grad_norm": 0.6445298685317932,
+      "learning_rate": 2.98610553254842e-06,
+      "loss": 0.6071,
+      "step": 1443
+    },
+    {
+      "epoch": 2.265875545530329,
+      "grad_norm": 0.6622508987875027,
+      "learning_rate": 2.974031112880037e-06,
+      "loss": 0.62,
+      "step": 1444
+    },
+    {
+      "epoch": 2.2674447114205853,
+      "grad_norm": 0.6260982640737418,
+      "learning_rate": 2.961976888730956e-06,
+      "loss": 0.6063,
+      "step": 1445
+    },
+    {
+      "epoch": 2.269013877310842,
+      "grad_norm": 0.6213660116532819,
+      "learning_rate": 2.9499428947499696e-06,
+      "loss": 0.6168,
+      "step": 1446
+    },
+    {
+      "epoch": 2.2705830432010985,
+      "grad_norm": 0.6327257030732624,
+      "learning_rate": 2.9379291655277143e-06,
+      "loss": 0.6187,
+      "step": 1447
+    },
+    {
+      "epoch": 2.272152209091355,
+      "grad_norm": 0.6676047779512317,
+      "learning_rate": 2.925935735596582e-06,
+      "loss": 0.6096,
+      "step": 1448
+    },
+    {
+      "epoch": 2.273721374981611,
+      "grad_norm": 0.6710377516211432,
+      "learning_rate": 2.913962639430615e-06,
+      "loss": 0.6099,
+      "step": 1449
+    },
+    {
+      "epoch": 2.2752905408718678,
+      "grad_norm": 0.6312375938585953,
+      "learning_rate": 2.9020099114454046e-06,
+      "loss": 0.6217,
+      "step": 1450
+    },
+    {
+      "epoch": 2.2768597067621243,
+      "grad_norm": 0.6110392396180195,
+      "learning_rate": 2.890077585998007e-06,
+      "loss": 0.6056,
+      "step": 1451
+    },
+    {
+      "epoch": 2.2784288726523805,
+      "grad_norm": 0.6366367213227224,
+      "learning_rate": 2.878165697386812e-06,
+      "loss": 0.6295,
+      "step": 1452
+    },
+    {
+      "epoch": 2.279998038542637,
+      "grad_norm": 0.6102884849262568,
+      "learning_rate": 2.866274279851491e-06,
+      "loss": 0.6251,
+      "step": 1453
+    },
+    {
+      "epoch": 2.2815672044328936,
+      "grad_norm": 0.6522005329996032,
+      "learning_rate": 2.854403367572849e-06,
+      "loss": 0.6246,
+      "step": 1454
+    },
+    {
+      "epoch": 2.28313637032315,
+      "grad_norm": 0.6070571619116706,
+      "learning_rate": 2.8425529946727725e-06,
+      "loss": 0.6126,
+      "step": 1455
+    },
+    {
+      "epoch": 2.2847055362134068,
+      "grad_norm": 0.6341043240762326,
+      "learning_rate": 2.830723195214088e-06,
+      "loss": 0.6076,
+      "step": 1456
+    },
+    {
+      "epoch": 2.286274702103663,
+      "grad_norm": 0.6321485028521564,
+      "learning_rate": 2.818914003200508e-06,
+      "loss": 0.6263,
+      "step": 1457
+    },
+    {
+      "epoch": 2.2878438679939195,
+      "grad_norm": 0.6511982109621314,
+      "learning_rate": 2.807125452576487e-06,
+      "loss": 0.6134,
+      "step": 1458
+    },
+    {
+      "epoch": 2.289413033884176,
+      "grad_norm": 0.657722824823969,
+      "learning_rate": 2.7953575772271667e-06,
+      "loss": 0.6205,
+      "step": 1459
+    },
+    {
+      "epoch": 2.290982199774432,
+      "grad_norm": 0.6407152335124848,
+      "learning_rate": 2.7836104109782525e-06,
+      "loss": 0.6261,
+      "step": 1460
+    },
+    {
+      "epoch": 2.2925513656646888,
+      "grad_norm": 0.6464924200998755,
+      "learning_rate": 2.7718839875959213e-06,
+      "loss": 0.617,
+      "step": 1461
+    },
+    {
+      "epoch": 2.2941205315549453,
+      "grad_norm": 0.6363617759835972,
+      "learning_rate": 2.760178340786729e-06,
+      "loss": 0.6141,
+      "step": 1462
+    },
+    {
+      "epoch": 2.295689697445202,
+      "grad_norm": 0.6382021366682696,
+      "learning_rate": 2.7484935041975113e-06,
+      "loss": 0.6366,
+      "step": 1463
+    },
+    {
+      "epoch": 2.2972588633354585,
+      "grad_norm": 0.6264419892564671,
+      "learning_rate": 2.7368295114152852e-06,
+      "loss": 0.614,
+      "step": 1464
+    },
+    {
+      "epoch": 2.2988280292257146,
+      "grad_norm": 0.6195880762505518,
+      "learning_rate": 2.7251863959671533e-06,
+      "loss": 0.6219,
+      "step": 1465
+    },
+    {
+      "epoch": 2.300397195115971,
+      "grad_norm": 0.6337207342705035,
+      "learning_rate": 2.713564191320217e-06,
+      "loss": 0.6257,
+      "step": 1466
+    },
+    {
+      "epoch": 2.3019663610062278,
+      "grad_norm": 0.6401883764413624,
+      "learning_rate": 2.7019629308814544e-06,
+      "loss": 0.607,
+      "step": 1467
+    },
+    {
+      "epoch": 2.303535526896484,
+      "grad_norm": 0.6455545084912034,
+      "learning_rate": 2.6903826479976582e-06,
+      "loss": 0.6203,
+      "step": 1468
+    },
+    {
+      "epoch": 2.3051046927867405,
+      "grad_norm": 0.6319529124339137,
+      "learning_rate": 2.678823375955314e-06,
+      "loss": 0.605,
+      "step": 1469
+    },
+    {
+      "epoch": 2.306673858676997,
+      "grad_norm": 0.6734464564763516,
+      "learning_rate": 2.667285147980515e-06,
+      "loss": 0.6004,
+      "step": 1470
+    },
+    {
+      "epoch": 2.3082430245672536,
+      "grad_norm": 0.6457108215880251,
+      "learning_rate": 2.6557679972388674e-06,
+      "loss": 0.6407,
+      "step": 1471
+    },
+    {
+      "epoch": 2.3098121904575097,
+      "grad_norm": 0.623145989695181,
+      "learning_rate": 2.6442719568353894e-06,
+      "loss": 0.581,
+      "step": 1472
+    },
+    {
+      "epoch": 2.3113813563477663,
+      "grad_norm": 0.6249008620059275,
+      "learning_rate": 2.6327970598144214e-06,
+      "loss": 0.6084,
+      "step": 1473
+    },
+    {
+      "epoch": 2.312950522238023,
+      "grad_norm": 0.6205727447146092,
+      "learning_rate": 2.62134333915953e-06,
+      "loss": 0.6089,
+      "step": 1474
+    },
+    {
+      "epoch": 2.3145196881282795,
+      "grad_norm": 0.6215601762263706,
+      "learning_rate": 2.6099108277934105e-06,
+      "loss": 0.6235,
+      "step": 1475
+    },
+    {
+      "epoch": 2.3160888540185356,
+      "grad_norm": 0.6623986841030728,
+      "learning_rate": 2.598499558577794e-06,
+      "loss": 0.6189,
+      "step": 1476
+    },
+    {
+      "epoch": 2.317658019908792,
+      "grad_norm": 0.6598567447429622,
+      "learning_rate": 2.5871095643133502e-06,
+      "loss": 0.6142,
+      "step": 1477
+    },
+    {
+      "epoch": 2.3192271857990487,
+      "grad_norm": 0.626942727703376,
+      "learning_rate": 2.57574087773961e-06,
+      "loss": 0.6071,
+      "step": 1478
+    },
+    {
+      "epoch": 2.3207963516893053,
+      "grad_norm": 0.6621000130405794,
+      "learning_rate": 2.564393531534832e-06,
+      "loss": 0.6068,
+      "step": 1479
+    },
+    {
+      "epoch": 2.3223655175795614,
+      "grad_norm": 0.6578535480770981,
+      "learning_rate": 2.553067558315958e-06,
+      "loss": 0.6232,
+      "step": 1480
+    },
+    {
+      "epoch": 2.323934683469818,
+      "grad_norm": 0.6148389298447706,
+      "learning_rate": 2.5417629906384834e-06,
+      "loss": 0.603,
+      "step": 1481
+    },
+    {
+      "epoch": 2.3255038493600746,
+      "grad_norm": 0.6147817570612852,
+      "learning_rate": 2.5304798609963754e-06,
+      "loss": 0.6059,
+      "step": 1482
+    },
+    {
+      "epoch": 2.327073015250331,
+      "grad_norm": 0.6312216443998306,
+      "learning_rate": 2.5192182018219826e-06,
+      "loss": 0.6372,
+      "step": 1483
+    },
+    {
+      "epoch": 2.3286421811405873,
+      "grad_norm": 0.6433123033871498,
+      "learning_rate": 2.507978045485937e-06,
+      "loss": 0.6077,
+      "step": 1484
+    },
+    {
+      "epoch": 2.330211347030844,
+      "grad_norm": 0.5893717212409233,
+      "learning_rate": 2.496759424297062e-06,
+      "loss": 0.6144,
+      "step": 1485
+    },
+    {
+      "epoch": 2.3317805129211004,
+      "grad_norm": 0.6137350126249266,
+      "learning_rate": 2.485562370502279e-06,
+      "loss": 0.5823,
+      "step": 1486
+    },
+    {
+      "epoch": 2.333349678811357,
+      "grad_norm": 0.6341703219917879,
+      "learning_rate": 2.474386916286526e-06,
+      "loss": 0.6267,
+      "step": 1487
+    },
+    {
+      "epoch": 2.334918844701613,
+      "grad_norm": 0.6589104086276759,
+      "learning_rate": 2.4632330937726356e-06,
+      "loss": 0.6099,
+      "step": 1488
+    },
+    {
+      "epoch": 2.3364880105918697,
+      "grad_norm": 0.6442357630846699,
+      "learning_rate": 2.4521009350212854e-06,
+      "loss": 0.6229,
+      "step": 1489
+    },
+    {
+      "epoch": 2.3380571764821263,
+      "grad_norm": 0.6673942610035479,
+      "learning_rate": 2.4409904720308576e-06,
+      "loss": 0.5899,
+      "step": 1490
+    },
+    {
+      "epoch": 2.339626342372383,
+      "grad_norm": 0.6595420768429126,
+      "learning_rate": 2.4299017367373966e-06,
+      "loss": 0.5922,
+      "step": 1491
+    },
+    {
+      "epoch": 2.341195508262639,
+      "grad_norm": 0.6723612633056628,
+      "learning_rate": 2.4188347610144693e-06,
+      "loss": 0.6243,
+      "step": 1492
+    },
+    {
+      "epoch": 2.3427646741528956,
+      "grad_norm": 0.6011007231850244,
+      "learning_rate": 2.407789576673115e-06,
+      "loss": 0.6138,
+      "step": 1493
+    },
+    {
+      "epoch": 2.344333840043152,
+      "grad_norm": 0.6115044818220867,
+      "learning_rate": 2.396766215461728e-06,
+      "loss": 0.5967,
+      "step": 1494
+    },
+    {
+      "epoch": 2.3459030059334083,
+      "grad_norm": 0.6219391681432415,
+      "learning_rate": 2.385764709065973e-06,
+      "loss": 0.6129,
+      "step": 1495
+    },
+    {
+      "epoch": 2.347472171823665,
+      "grad_norm": 0.6175259072542536,
+      "learning_rate": 2.3747850891086956e-06,
+      "loss": 0.6045,
+      "step": 1496
+    },
+    {
+      "epoch": 2.3490413377139214,
+      "grad_norm": 0.6343486512456172,
+      "learning_rate": 2.363827387149833e-06,
+      "loss": 0.602,
+      "step": 1497
+    },
+    {
+      "epoch": 2.350610503604178,
+      "grad_norm": 0.6182504030823537,
+      "learning_rate": 2.3528916346863185e-06,
+      "loss": 0.6085,
+      "step": 1498
+    },
+    {
+      "epoch": 2.3521796694944346,
+      "grad_norm": 0.6289914071248726,
+      "learning_rate": 2.3419778631519963e-06,
+      "loss": 0.6081,
+      "step": 1499
+    },
+    {
+      "epoch": 2.3537488353846907,
+      "grad_norm": 0.6437968229643949,
+      "learning_rate": 2.331086103917525e-06,
+      "loss": 0.6258,
+      "step": 1500
+    },
+    {
+      "epoch": 2.3553180012749473,
+      "grad_norm": 0.6143000348105965,
+      "learning_rate": 2.320216388290293e-06,
+      "loss": 0.6119,
+      "step": 1501
+    },
+    {
+      "epoch": 2.356887167165204,
+      "grad_norm": 0.6430232985289135,
+      "learning_rate": 2.3093687475143236e-06,
+      "loss": 0.611,
+      "step": 1502
+    },
+    {
+      "epoch": 2.35845633305546,
+      "grad_norm": 0.6525804169675627,
+      "learning_rate": 2.2985432127701945e-06,
+      "loss": 0.5954,
+      "step": 1503
+    },
+    {
+      "epoch": 2.3600254989457166,
+      "grad_norm": 0.6405938021716366,
+      "learning_rate": 2.2877398151749375e-06,
+      "loss": 0.6083,
+      "step": 1504
+    },
+    {
+      "epoch": 2.361594664835973,
+      "grad_norm": 0.6301358688309666,
+      "learning_rate": 2.27695858578195e-06,
+      "loss": 0.6195,
+      "step": 1505
+    },
+    {
+      "epoch": 2.3631638307262297,
+      "grad_norm": 0.6376332562953813,
+      "learning_rate": 2.2661995555809123e-06,
+      "loss": 0.607,
+      "step": 1506
+    },
+    {
+      "epoch": 2.3647329966164863,
+      "grad_norm": 0.6528772246271954,
+      "learning_rate": 2.255462755497696e-06,
+      "loss": 0.6117,
+      "step": 1507
+    },
+    {
+      "epoch": 2.3663021625067424,
+      "grad_norm": 0.6405262043615331,
+      "learning_rate": 2.2447482163942725e-06,
+      "loss": 0.6098,
+      "step": 1508
+    },
+    {
+      "epoch": 2.367871328396999,
+      "grad_norm": 0.6157495247738175,
+      "learning_rate": 2.2340559690686247e-06,
+      "loss": 0.606,
+      "step": 1509
+    },
+    {
+      "epoch": 2.3694404942872556,
+      "grad_norm": 0.6173747553819633,
+      "learning_rate": 2.223386044254663e-06,
+      "loss": 0.6191,
+      "step": 1510
+    },
+    {
+      "epoch": 2.3710096601775117,
+      "grad_norm": 0.6495632557522085,
+      "learning_rate": 2.2127384726221303e-06,
+      "loss": 0.6174,
+      "step": 1511
+    },
+    {
+      "epoch": 2.3725788260677683,
+      "grad_norm": 0.63804910516746,
+      "learning_rate": 2.2021132847765246e-06,
+      "loss": 0.615,
+      "step": 1512
+    },
+    {
+      "epoch": 2.374147991958025,
+      "grad_norm": 0.6007113635502407,
+      "learning_rate": 2.1915105112589894e-06,
+      "loss": 0.6063,
+      "step": 1513
+    },
+    {
+      "epoch": 2.3757171578482814,
+      "grad_norm": 0.6256421483133896,
+      "learning_rate": 2.1809301825462593e-06,
+      "loss": 0.6112,
+      "step": 1514
+    },
+    {
+      "epoch": 2.3772863237385375,
+      "grad_norm": 0.6338794634542337,
+      "learning_rate": 2.170372329050533e-06,
+      "loss": 0.6238,
+      "step": 1515
+    },
+    {
+      "epoch": 2.378855489628794,
+      "grad_norm": 0.6262070628639611,
+      "learning_rate": 2.159836981119425e-06,
+      "loss": 0.6003,
+      "step": 1516
+    },
+    {
+      "epoch": 2.3804246555190507,
+      "grad_norm": 0.628382539066662,
+      "learning_rate": 2.1493241690358437e-06,
+      "loss": 0.6239,
+      "step": 1517
+    },
+    {
+      "epoch": 2.3819938214093073,
+      "grad_norm": 0.6441384619280617,
+      "learning_rate": 2.138833923017931e-06,
+      "loss": 0.6336,
+      "step": 1518
+    },
+    {
+      "epoch": 2.3835629872995634,
+      "grad_norm": 0.6203007050445565,
+      "learning_rate": 2.1283662732189624e-06,
+      "loss": 0.6076,
+      "step": 1519
+    },
+    {
+      "epoch": 2.38513215318982,
+      "grad_norm": 0.6373622299258588,
+      "learning_rate": 2.1179212497272582e-06,
+      "loss": 0.6162,
+      "step": 1520
+    },
+    {
+      "epoch": 2.3867013190800765,
+      "grad_norm": 0.6473460496058665,
+      "learning_rate": 2.107498882566107e-06,
+      "loss": 0.6309,
+      "step": 1521
+    },
+    {
+      "epoch": 2.388270484970333,
+      "grad_norm": 0.6281388779099333,
+      "learning_rate": 2.0970992016936685e-06,
+      "loss": 0.6105,
+      "step": 1522
+    },
+    {
+      "epoch": 2.3898396508605892,
+      "grad_norm": 0.612528178413823,
+      "learning_rate": 2.0867222370028985e-06,
+      "loss": 0.6147,
+      "step": 1523
+    },
+    {
+      "epoch": 2.391408816750846,
+      "grad_norm": 0.6393001306756122,
+      "learning_rate": 2.0763680183214485e-06,
+      "loss": 0.6055,
+      "step": 1524
+    },
+    {
+      "epoch": 2.3929779826411024,
+      "grad_norm": 0.6355165574692425,
+      "learning_rate": 2.0660365754116042e-06,
+      "loss": 0.5971,
+      "step": 1525
+    },
+    {
+      "epoch": 2.394547148531359,
+      "grad_norm": 0.6508164857285735,
+      "learning_rate": 2.0557279379701632e-06,
+      "loss": 0.6096,
+      "step": 1526
+    },
+    {
+      "epoch": 2.396116314421615,
+      "grad_norm": 0.6195889265363922,
+      "learning_rate": 2.045442135628389e-06,
+      "loss": 0.6171,
+      "step": 1527
+    },
+    {
+      "epoch": 2.3976854803118717,
+      "grad_norm": 0.625707846687031,
+      "learning_rate": 2.0351791979518997e-06,
+      "loss": 0.602,
+      "step": 1528
+    },
+    {
+      "epoch": 2.3992546462021282,
+      "grad_norm": 0.6405651850003998,
+      "learning_rate": 2.0249391544405916e-06,
+      "loss": 0.5988,
+      "step": 1529
+    },
+    {
+      "epoch": 2.400823812092385,
+      "grad_norm": 0.638746243332164,
+      "learning_rate": 2.0147220345285524e-06,
+      "loss": 0.5967,
+      "step": 1530
+    },
+    {
+      "epoch": 2.402392977982641,
+      "grad_norm": 0.6270345942193685,
+      "learning_rate": 2.004527867583982e-06,
+      "loss": 0.6134,
+      "step": 1531
+    },
+    {
+      "epoch": 2.4039621438728975,
+      "grad_norm": 0.6451295459806061,
+      "learning_rate": 1.9943566829091e-06,
+      "loss": 0.6204,
+      "step": 1532
+    },
+    {
+      "epoch": 2.405531309763154,
+      "grad_norm": 0.6805599696936404,
+      "learning_rate": 1.9842085097400678e-06,
+      "loss": 0.6469,
+      "step": 1533
+    },
+    {
+      "epoch": 2.4071004756534107,
+      "grad_norm": 0.6287805201611575,
+      "learning_rate": 1.9740833772469025e-06,
+      "loss": 0.6068,
+      "step": 1534
+    },
+    {
+      "epoch": 2.408669641543667,
+      "grad_norm": 0.6099299399274067,
+      "learning_rate": 1.963981314533391e-06,
+      "loss": 0.6125,
+      "step": 1535
+    },
+    {
+      "epoch": 2.4102388074339234,
+      "grad_norm": 0.6363862048066421,
+      "learning_rate": 1.953902350637006e-06,
+      "loss": 0.6009,
+      "step": 1536
+    },
+    {
+      "epoch": 2.41180797332418,
+      "grad_norm": 0.6500867083635392,
+      "learning_rate": 1.9438465145288377e-06,
+      "loss": 0.6297,
+      "step": 1537
+    },
+    {
+      "epoch": 2.413377139214436,
+      "grad_norm": 0.6247533149417938,
+      "learning_rate": 1.9338138351134773e-06,
+      "loss": 0.6208,
+      "step": 1538
+    },
+    {
+      "epoch": 2.4149463051046927,
+      "grad_norm": 0.6077696409784458,
+      "learning_rate": 1.9238043412289696e-06,
+      "loss": 0.6064,
+      "step": 1539
+    },
+    {
+      "epoch": 2.4165154709949492,
+      "grad_norm": 0.6421658176199089,
+      "learning_rate": 1.9138180616467105e-06,
+      "loss": 0.6197,
+      "step": 1540
+    },
+    {
+      "epoch": 2.418084636885206,
+      "grad_norm": 0.6416440586342028,
+      "learning_rate": 1.903855025071366e-06,
+      "loss": 0.6129,
+      "step": 1541
+    },
+    {
+      "epoch": 2.4196538027754624,
+      "grad_norm": 0.6533822232027569,
+      "learning_rate": 1.8939152601407952e-06,
+      "loss": 0.6291,
+      "step": 1542
+    },
+    {
+      "epoch": 2.4212229686657185,
+      "grad_norm": 0.6246858478025527,
+      "learning_rate": 1.8839987954259609e-06,
+      "loss": 0.6021,
+      "step": 1543
+    },
+    {
+      "epoch": 2.422792134555975,
+      "grad_norm": 0.6070615898532166,
+      "learning_rate": 1.8741056594308571e-06,
+      "loss": 0.6086,
+      "step": 1544
+    },
+    {
+      "epoch": 2.4243613004462317,
+      "grad_norm": 0.6822191373430361,
+      "learning_rate": 1.8642358805924144e-06,
+      "loss": 0.6253,
+      "step": 1545
+    },
+    {
+      "epoch": 2.425930466336488,
+      "grad_norm": 0.6150816960462416,
+      "learning_rate": 1.8543894872804368e-06,
+      "loss": 0.615,
+      "step": 1546
+    },
+    {
+      "epoch": 2.4274996322267444,
+      "grad_norm": 0.6116811488446279,
+      "learning_rate": 1.8445665077974928e-06,
+      "loss": 0.5908,
+      "step": 1547
+    },
+    {
+      "epoch": 2.429068798117001,
+      "grad_norm": 0.6111796080758666,
+      "learning_rate": 1.8347669703788683e-06,
+      "loss": 0.5947,
+      "step": 1548
+    },
+    {
+      "epoch": 2.4306379640072575,
+      "grad_norm": 0.6391360438485262,
+      "learning_rate": 1.824990903192445e-06,
+      "loss": 0.6142,
+      "step": 1549
+    },
+    {
+      "epoch": 2.432207129897514,
+      "grad_norm": 0.6370102971888556,
+      "learning_rate": 1.815238334338667e-06,
+      "loss": 0.616,
+      "step": 1550
+    },
+    {
+      "epoch": 2.43377629578777,
+      "grad_norm": 0.6249494383931662,
+      "learning_rate": 1.8055092918504103e-06,
+      "loss": 0.6042,
+      "step": 1551
+    },
+    {
+      "epoch": 2.435345461678027,
+      "grad_norm": 0.6354472986625086,
+      "learning_rate": 1.795803803692947e-06,
+      "loss": 0.582,
+      "step": 1552
+    },
+    {
+      "epoch": 2.4369146275682834,
+      "grad_norm": 0.6164983696970364,
+      "learning_rate": 1.7861218977638252e-06,
+      "loss": 0.6157,
+      "step": 1553
+    },
+    {
+      "epoch": 2.4384837934585395,
+      "grad_norm": 0.5999431542191072,
+      "learning_rate": 1.7764636018928249e-06,
+      "loss": 0.5996,
+      "step": 1554
+    },
+    {
+      "epoch": 2.440052959348796,
+      "grad_norm": 0.5997694077107606,
+      "learning_rate": 1.766828943841853e-06,
+      "loss": 0.6242,
+      "step": 1555
+    },
+    {
+      "epoch": 2.4416221252390526,
+      "grad_norm": 0.6275576192505237,
+      "learning_rate": 1.7572179513048703e-06,
+      "loss": 0.6017,
+      "step": 1556
+    },
+    {
+      "epoch": 2.443191291129309,
+      "grad_norm": 0.6269667230125978,
+      "learning_rate": 1.7476306519078168e-06,
+      "loss": 0.6242,
+      "step": 1557
+    },
+    {
+      "epoch": 2.4447604570195653,
+      "grad_norm": 0.6351257941001033,
+      "learning_rate": 1.738067073208527e-06,
+      "loss": 0.6127,
+      "step": 1558
+    },
+    {
+      "epoch": 2.446329622909822,
+      "grad_norm": 0.6254297485677375,
+      "learning_rate": 1.7285272426966526e-06,
+      "loss": 0.6381,
+      "step": 1559
+    },
+    {
+      "epoch": 2.4478987888000785,
+      "grad_norm": 0.6326320202845116,
+      "learning_rate": 1.7190111877935833e-06,
+      "loss": 0.6016,
+      "step": 1560
+    },
+    {
+      "epoch": 2.449467954690335,
+      "grad_norm": 0.6314581644694905,
+      "learning_rate": 1.7095189358523668e-06,
+      "loss": 0.6079,
+      "step": 1561
+    },
+    {
+      "epoch": 2.451037120580591,
+      "grad_norm": 0.6466999920670993,
+      "learning_rate": 1.700050514157635e-06,
+      "loss": 0.597,
+      "step": 1562
+    },
+    {
+      "epoch": 2.4526062864708478,
+      "grad_norm": 0.622701857312493,
+      "learning_rate": 1.690605949925519e-06,
+      "loss": 0.618,
+      "step": 1563
+    },
+    {
+      "epoch": 2.4541754523611043,
+      "grad_norm": 0.6647600633659189,
+      "learning_rate": 1.6811852703035737e-06,
+      "loss": 0.6397,
+      "step": 1564
+    },
+    {
+      "epoch": 2.455744618251361,
+      "grad_norm": 0.6322663886358674,
+      "learning_rate": 1.6717885023707004e-06,
+      "loss": 0.5989,
+      "step": 1565
+    },
+    {
+      "epoch": 2.457313784141617,
+      "grad_norm": 0.6387131085067256,
+      "learning_rate": 1.6624156731370666e-06,
+      "loss": 0.6211,
+      "step": 1566
+    },
+    {
+      "epoch": 2.4588829500318736,
+      "grad_norm": 0.6385964583943042,
+      "learning_rate": 1.6530668095440339e-06,
+      "loss": 0.6034,
+      "step": 1567
+    },
+    {
+      "epoch": 2.46045211592213,
+      "grad_norm": 0.6262228713544584,
+      "learning_rate": 1.643741938464073e-06,
+      "loss": 0.6188,
+      "step": 1568
+    },
+    {
+      "epoch": 2.4620212818123868,
+      "grad_norm": 0.6363780515861948,
+      "learning_rate": 1.6344410867006944e-06,
+      "loss": 0.5898,
+      "step": 1569
+    },
+    {
+      "epoch": 2.463590447702643,
+      "grad_norm": 0.6426953500638161,
+      "learning_rate": 1.6251642809883595e-06,
+      "loss": 0.6057,
+      "step": 1570
+    },
+    {
+      "epoch": 2.4651596135928995,
+      "grad_norm": 0.631158519559688,
+      "learning_rate": 1.6159115479924259e-06,
+      "loss": 0.6173,
+      "step": 1571
+    },
+    {
+      "epoch": 2.466728779483156,
+      "grad_norm": 0.5963181133175319,
+      "learning_rate": 1.606682914309038e-06,
+      "loss": 0.5858,
+      "step": 1572
+    },
+    {
+      "epoch": 2.4682979453734126,
+      "grad_norm": 0.6505244760665863,
+      "learning_rate": 1.597478406465085e-06,
+      "loss": 0.5898,
+      "step": 1573
+    },
+    {
+      "epoch": 2.4698671112636688,
+      "grad_norm": 0.6358307157040985,
+      "learning_rate": 1.5882980509180956e-06,
+      "loss": 0.6146,
+      "step": 1574
+    },
+    {
+      "epoch": 2.4714362771539253,
+      "grad_norm": 0.5903480280756508,
+      "learning_rate": 1.579141874056187e-06,
+      "loss": 0.6047,
+      "step": 1575
+    },
+    {
+      "epoch": 2.473005443044182,
+      "grad_norm": 0.6189676720446393,
+      "learning_rate": 1.5700099021979642e-06,
+      "loss": 0.6348,
+      "step": 1576
+    },
+    {
+      "epoch": 2.4745746089344385,
+      "grad_norm": 0.6443719263161586,
+      "learning_rate": 1.5609021615924702e-06,
+      "loss": 0.6315,
+      "step": 1577
+    },
+    {
+      "epoch": 2.4761437748246946,
+      "grad_norm": 0.6145981724704054,
+      "learning_rate": 1.551818678419087e-06,
+      "loss": 0.6145,
+      "step": 1578
+    },
+    {
+      "epoch": 2.477712940714951,
+      "grad_norm": 0.622934927524397,
+      "learning_rate": 1.5427594787874766e-06,
+      "loss": 0.5948,
+      "step": 1579
+    },
+    {
+      "epoch": 2.4792821066052078,
+      "grad_norm": 0.6699845892368447,
+      "learning_rate": 1.5337245887374962e-06,
+      "loss": 0.6047,
+      "step": 1580
+    },
+    {
+      "epoch": 2.480851272495464,
+      "grad_norm": 0.6375995339298633,
+      "learning_rate": 1.524714034239132e-06,
+      "loss": 0.6144,
+      "step": 1581
+    },
+    {
+      "epoch": 2.4824204383857205,
+      "grad_norm": 0.6569547672403754,
+      "learning_rate": 1.5157278411924148e-06,
+      "loss": 0.6132,
+      "step": 1582
+    },
+    {
+      "epoch": 2.483989604275977,
+      "grad_norm": 0.6029476296063775,
+      "learning_rate": 1.5067660354273527e-06,
+      "loss": 0.6208,
+      "step": 1583
+    },
+    {
+      "epoch": 2.4855587701662336,
+      "grad_norm": 0.6511040050621936,
+      "learning_rate": 1.4978286427038602e-06,
+      "loss": 0.6145,
+      "step": 1584
+    },
+    {
+      "epoch": 2.48712793605649,
+      "grad_norm": 0.618052434634993,
+      "learning_rate": 1.4889156887116652e-06,
+      "loss": 0.6018,
+      "step": 1585
+    },
+    {
+      "epoch": 2.4886971019467463,
+      "grad_norm": 0.6050298694027211,
+      "learning_rate": 1.4800271990702664e-06,
+      "loss": 0.6093,
+      "step": 1586
+    },
+    {
+      "epoch": 2.490266267837003,
+      "grad_norm": 0.6391172084775967,
+      "learning_rate": 1.4711631993288222e-06,
+      "loss": 0.6395,
+      "step": 1587
+    },
+    {
+      "epoch": 2.4918354337272595,
+      "grad_norm": 0.6054235436330725,
+      "learning_rate": 1.462323714966114e-06,
+      "loss": 0.6055,
+      "step": 1588
+    },
+    {
+      "epoch": 2.4934045996175156,
+      "grad_norm": 0.6548187603233284,
+      "learning_rate": 1.4535087713904484e-06,
+      "loss": 0.6182,
+      "step": 1589
+    },
+    {
+      "epoch": 2.494973765507772,
+      "grad_norm": 0.6196289256254138,
+      "learning_rate": 1.4447183939395925e-06,
+      "loss": 0.6051,
+      "step": 1590
+    },
+    {
+      "epoch": 2.4965429313980287,
+      "grad_norm": 0.6278881333433757,
+      "learning_rate": 1.4359526078806985e-06,
+      "loss": 0.6206,
+      "step": 1591
+    },
+    {
+      "epoch": 2.4981120972882853,
+      "grad_norm": 0.6279817779364651,
+      "learning_rate": 1.4272114384102353e-06,
+      "loss": 0.5997,
+      "step": 1592
+    },
+    {
+      "epoch": 2.499681263178542,
+      "grad_norm": 0.6403440499985011,
+      "learning_rate": 1.4184949106539137e-06,
+      "loss": 0.6276,
+      "step": 1593
+    },
+    {
+      "epoch": 2.501250429068798,
+      "grad_norm": 0.6329686166680581,
+      "learning_rate": 1.4098030496666126e-06,
+      "loss": 0.6154,
+      "step": 1594
+    },
+    {
+      "epoch": 2.5028195949590546,
+      "grad_norm": 0.6248742808250498,
+      "learning_rate": 1.4011358804323095e-06,
+      "loss": 0.6283,
+      "step": 1595
+    },
+    {
+      "epoch": 2.504388760849311,
+      "grad_norm": 0.6405945581797248,
+      "learning_rate": 1.3924934278640068e-06,
+      "loss": 0.6415,
+      "step": 1596
+    },
+    {
+      "epoch": 2.5059579267395673,
+      "grad_norm": 0.6263078722686413,
+      "learning_rate": 1.3838757168036598e-06,
+      "loss": 0.6178,
+      "step": 1597
+    },
+    {
+      "epoch": 2.507527092629824,
+      "grad_norm": 0.6853831416678203,
+      "learning_rate": 1.3752827720221128e-06,
+      "loss": 0.6047,
+      "step": 1598
+    },
+    {
+      "epoch": 2.5090962585200804,
+      "grad_norm": 0.6038787174236342,
+      "learning_rate": 1.366714618219015e-06,
+      "loss": 0.5963,
+      "step": 1599
+    },
+    {
+      "epoch": 2.510665424410337,
+      "grad_norm": 0.6462779919253984,
+      "learning_rate": 1.3581712800227576e-06,
+      "loss": 0.6056,
+      "step": 1600
+    },
+    {
+      "epoch": 2.5122345903005936,
+      "grad_norm": 0.6639922882540314,
+      "learning_rate": 1.3496527819904038e-06,
+      "loss": 0.6277,
+      "step": 1601
+    },
+    {
+      "epoch": 2.5138037561908497,
+      "grad_norm": 0.6312314509306844,
+      "learning_rate": 1.3411591486076136e-06,
+      "loss": 0.6082,
+      "step": 1602
+    },
+    {
+      "epoch": 2.5153729220811063,
+      "grad_norm": 0.60419353027876,
+      "learning_rate": 1.3326904042885757e-06,
+      "loss": 0.6174,
+      "step": 1603
+    },
+    {
+      "epoch": 2.516942087971363,
+      "grad_norm": 0.6394953202022576,
+      "learning_rate": 1.324246573375938e-06,
+      "loss": 0.6223,
+      "step": 1604
+    },
+    {
+      "epoch": 2.518511253861619,
+      "grad_norm": 0.6301197385255249,
+      "learning_rate": 1.3158276801407432e-06,
+      "loss": 0.6033,
+      "step": 1605
+    },
+    {
+      "epoch": 2.5200804197518756,
+      "grad_norm": 0.6403050882453963,
+      "learning_rate": 1.3074337487823374e-06,
+      "loss": 0.6195,
+      "step": 1606
+    },
+    {
+      "epoch": 2.521649585642132,
+      "grad_norm": 0.6483566179126462,
+      "learning_rate": 1.299064803428336e-06,
+      "loss": 0.6085,
+      "step": 1607
+    },
+    {
+      "epoch": 2.5232187515323887,
+      "grad_norm": 0.6462911527899499,
+      "learning_rate": 1.290720868134515e-06,
+      "loss": 0.613,
+      "step": 1608
+    },
+    {
+      "epoch": 2.5247879174226453,
+      "grad_norm": 0.6061301684390736,
+      "learning_rate": 1.2824019668847775e-06,
+      "loss": 0.6237,
+      "step": 1609
+    },
+    {
+      "epoch": 2.5263570833129014,
+      "grad_norm": 0.6199885041349689,
+      "learning_rate": 1.2741081235910547e-06,
+      "loss": 0.6087,
+      "step": 1610
+    },
+    {
+      "epoch": 2.527926249203158,
+      "grad_norm": 0.6228775327756475,
+      "learning_rate": 1.2658393620932652e-06,
+      "loss": 0.6263,
+      "step": 1611
+    },
+    {
+      "epoch": 2.5294954150934146,
+      "grad_norm": 0.6411144224242795,
+      "learning_rate": 1.2575957061592159e-06,
+      "loss": 0.6166,
+      "step": 1612
+    },
+    {
+      "epoch": 2.5310645809836707,
+      "grad_norm": 0.6577772393020911,
+      "learning_rate": 1.2493771794845644e-06,
+      "loss": 0.6016,
+      "step": 1613
+    },
+    {
+      "epoch": 2.5326337468739273,
+      "grad_norm": 0.6286159437770955,
+      "learning_rate": 1.2411838056927295e-06,
+      "loss": 0.6071,
+      "step": 1614
+    },
+    {
+      "epoch": 2.534202912764184,
+      "grad_norm": 0.6198771581646637,
+      "learning_rate": 1.233015608334831e-06,
+      "loss": 0.6126,
+      "step": 1615
+    },
+    {
+      "epoch": 2.53577207865444,
+      "grad_norm": 0.6128245165835714,
+      "learning_rate": 1.2248726108896225e-06,
+      "loss": 0.606,
+      "step": 1616
+    },
+    {
+      "epoch": 2.5373412445446966,
+      "grad_norm": 0.613565041753448,
+      "learning_rate": 1.216754836763422e-06,
+      "loss": 0.6153,
+      "step": 1617
+    },
+    {
+      "epoch": 2.538910410434953,
+      "grad_norm": 0.6266682143529125,
+      "learning_rate": 1.2086623092900441e-06,
+      "loss": 0.618,
+      "step": 1618
+    },
+    {
+      "epoch": 2.5404795763252097,
+      "grad_norm": 0.6043725875760491,
+      "learning_rate": 1.2005950517307374e-06,
+      "loss": 0.6002,
+      "step": 1619
+    },
+    {
+      "epoch": 2.5420487422154663,
+      "grad_norm": 0.621036861043719,
+      "learning_rate": 1.1925530872741098e-06,
+      "loss": 0.6276,
+      "step": 1620
+    },
+    {
+      "epoch": 2.5436179081057224,
+      "grad_norm": 0.5933230843164937,
+      "learning_rate": 1.1845364390360692e-06,
+      "loss": 0.6143,
+      "step": 1621
+    },
+    {
+      "epoch": 2.545187073995979,
+      "grad_norm": 0.6187328487055244,
+      "learning_rate": 1.1765451300597574e-06,
+      "loss": 0.5989,
+      "step": 1622
+    },
+    {
+      "epoch": 2.5467562398862356,
+      "grad_norm": 0.5716651604588641,
+      "learning_rate": 1.1685791833154748e-06,
+      "loss": 0.6009,
+      "step": 1623
+    },
+    {
+      "epoch": 2.5483254057764917,
+      "grad_norm": 0.6498092570434579,
+      "learning_rate": 1.160638621700624e-06,
+      "loss": 0.6073,
+      "step": 1624
+    },
+    {
+      "epoch": 2.5498945716667483,
+      "grad_norm": 0.6113170699342618,
+      "learning_rate": 1.1527234680396392e-06,
+      "loss": 0.6253,
+      "step": 1625
+    },
+    {
+      "epoch": 2.551463737557005,
+      "grad_norm": 0.6511765985576149,
+      "learning_rate": 1.1448337450839208e-06,
+      "loss": 0.5792,
+      "step": 1626
+    },
+    {
+      "epoch": 2.5530329034472614,
+      "grad_norm": 0.6469524388900086,
+      "learning_rate": 1.1369694755117733e-06,
+      "loss": 0.6375,
+      "step": 1627
+    },
+    {
+      "epoch": 2.554602069337518,
+      "grad_norm": 0.6193317460102629,
+      "learning_rate": 1.1291306819283353e-06,
+      "loss": 0.6091,
+      "step": 1628
+    },
+    {
+      "epoch": 2.556171235227774,
+      "grad_norm": 0.642537669933156,
+      "learning_rate": 1.121317386865518e-06,
+      "loss": 0.6158,
+      "step": 1629
+    },
+    {
+      "epoch": 2.5577404011180307,
+      "grad_norm": 0.6238691080621291,
+      "learning_rate": 1.1135296127819407e-06,
+      "loss": 0.5872,
+      "step": 1630
+    },
+    {
+      "epoch": 2.5593095670082873,
+      "grad_norm": 0.6123001185398639,
+      "learning_rate": 1.1057673820628612e-06,
+      "loss": 0.6096,
+      "step": 1631
+    },
+    {
+      "epoch": 2.5608787328985434,
+      "grad_norm": 0.6174704202396814,
+      "learning_rate": 1.0980307170201222e-06,
+      "loss": 0.602,
+      "step": 1632
+    },
+    {
+      "epoch": 2.5624478987888,
+      "grad_norm": 0.6409484261998208,
+      "learning_rate": 1.090319639892069e-06,
+      "loss": 0.6177,
+      "step": 1633
+    },
+    {
+      "epoch": 2.5640170646790565,
+      "grad_norm": 0.6306409839687421,
+      "learning_rate": 1.0826341728435096e-06,
+      "loss": 0.5895,
+      "step": 1634
+    },
+    {
+      "epoch": 2.565586230569313,
+      "grad_norm": 0.6200811666561175,
+      "learning_rate": 1.0749743379656263e-06,
+      "loss": 0.6165,
+      "step": 1635
+    },
+    {
+      "epoch": 2.5671553964595697,
+      "grad_norm": 0.6069260835975323,
+      "learning_rate": 1.0673401572759335e-06,
+      "loss": 0.6018,
+      "step": 1636
+    },
+    {
+      "epoch": 2.568724562349826,
+      "grad_norm": 0.6239421313429688,
+      "learning_rate": 1.0597316527182e-06,
+      "loss": 0.6117,
+      "step": 1637
+    },
+    {
+      "epoch": 2.5702937282400824,
+      "grad_norm": 0.622423182665628,
+      "learning_rate": 1.0521488461623919e-06,
+      "loss": 0.6131,
+      "step": 1638
+    },
+    {
+      "epoch": 2.571862894130339,
+      "grad_norm": 0.6491405530354314,
+      "learning_rate": 1.0445917594046073e-06,
+      "loss": 0.6131,
+      "step": 1639
+    },
+    {
+      "epoch": 2.573432060020595,
+      "grad_norm": 0.6414385864650671,
+      "learning_rate": 1.037060414167017e-06,
+      "loss": 0.6072,
+      "step": 1640
+    },
+    {
+      "epoch": 2.5750012259108517,
+      "grad_norm": 0.6192783327142664,
+      "learning_rate": 1.029554832097801e-06,
+      "loss": 0.6082,
+      "step": 1641
+    },
+    {
+      "epoch": 2.5765703918011083,
+      "grad_norm": 0.6320681366323113,
+      "learning_rate": 1.0220750347710795e-06,
+      "loss": 0.5985,
+      "step": 1642
+    },
+    {
+      "epoch": 2.578139557691365,
+      "grad_norm": 0.6267841186014629,
+      "learning_rate": 1.0146210436868675e-06,
+      "loss": 0.613,
+      "step": 1643
+    },
+    {
+      "epoch": 2.5797087235816214,
+      "grad_norm": 0.6116247762639864,
+      "learning_rate": 1.0071928802709886e-06,
+      "loss": 0.6104,
+      "step": 1644
+    },
+    {
+      "epoch": 2.5812778894718775,
+      "grad_norm": 0.6595623554979907,
+      "learning_rate": 9.997905658750428e-07,
+      "loss": 0.6047,
+      "step": 1645
+    },
+    {
+      "epoch": 2.582847055362134,
+      "grad_norm": 0.6269886873212843,
+      "learning_rate": 9.924141217763117e-07,
+      "loss": 0.6116,
+      "step": 1646
+    },
+    {
+      "epoch": 2.5844162212523907,
+      "grad_norm": 0.625318304559223,
+      "learning_rate": 9.8506356917773e-07,
+      "loss": 0.6166,
+      "step": 1647
+    },
+    {
+      "epoch": 2.585985387142647,
+      "grad_norm": 0.6089992405678766,
+      "learning_rate": 9.777389292078033e-07,
+      "loss": 0.5962,
+      "step": 1648
+    },
+    {
+      "epoch": 2.5875545530329034,
+      "grad_norm": 0.6233430244452259,
+      "learning_rate": 9.704402229205534e-07,
+      "loss": 0.5938,
+      "step": 1649
+    },
+    {
+      "epoch": 2.58912371892316,
+      "grad_norm": 0.6500889366305391,
+      "learning_rate": 9.631674712954586e-07,
+      "loss": 0.6171,
+      "step": 1650
+    },
+    {
+      "epoch": 2.5906928848134165,
+      "grad_norm": 0.6018959802186566,
+      "learning_rate": 9.559206952373933e-07,
+      "loss": 0.5982,
+      "step": 1651
+    },
+    {
+      "epoch": 2.592262050703673,
+      "grad_norm": 0.6292766046892906,
+      "learning_rate": 9.486999155765675e-07,
+      "loss": 0.6034,
+      "step": 1652
+    },
+    {
+      "epoch": 2.5938312165939292,
+      "grad_norm": 0.6567950396723267,
+      "learning_rate": 9.415051530684671e-07,
+      "loss": 0.5869,
+      "step": 1653
+    },
+    {
+      "epoch": 2.595400382484186,
+      "grad_norm": 0.6470169840291483,
+      "learning_rate": 9.343364283937928e-07,
+      "loss": 0.6256,
+      "step": 1654
+    },
+    {
+      "epoch": 2.5969695483744424,
+      "grad_norm": 0.6144759975367846,
+      "learning_rate": 9.271937621584026e-07,
+      "loss": 0.5908,
+      "step": 1655
+    },
+    {
+      "epoch": 2.5985387142646985,
+      "grad_norm": 0.6443112424048698,
+      "learning_rate": 9.200771748932513e-07,
+      "loss": 0.6117,
+      "step": 1656
+    },
+    {
+      "epoch": 2.600107880154955,
+      "grad_norm": 0.6488140315556548,
+      "learning_rate": 9.129866870543347e-07,
+      "loss": 0.5967,
+      "step": 1657
+    },
+    {
+      "epoch": 2.6016770460452117,
+      "grad_norm": 0.6194486411360511,
+      "learning_rate": 9.059223190226251e-07,
+      "loss": 0.6212,
+      "step": 1658
+    },
+    {
+      "epoch": 2.603246211935468,
+      "grad_norm": 0.6429705904587635,
+      "learning_rate": 8.988840911040165e-07,
+      "loss": 0.6019,
+      "step": 1659
+    },
+    {
+      "epoch": 2.6048153778257244,
+      "grad_norm": 0.6435054574798533,
+      "learning_rate": 8.918720235292655e-07,
+      "loss": 0.6271,
+      "step": 1660
+    },
+    {
+      "epoch": 2.606384543715981,
+      "grad_norm": 0.5956639052828546,
+      "learning_rate": 8.84886136453933e-07,
+      "loss": 0.6005,
+      "step": 1661
+    },
+    {
+      "epoch": 2.6079537096062375,
+      "grad_norm": 0.6073791304043207,
+      "learning_rate": 8.779264499583284e-07,
+      "loss": 0.6099,
+      "step": 1662
+    },
+    {
+      "epoch": 2.609522875496494,
+      "grad_norm": 0.5948979355736304,
+      "learning_rate": 8.709929840474463e-07,
+      "loss": 0.6171,
+      "step": 1663
+    },
+    {
+      "epoch": 2.61109204138675,
+      "grad_norm": 0.6541370980029577,
+      "learning_rate": 8.640857586509155e-07,
+      "loss": 0.6322,
+      "step": 1664
+    },
+    {
+      "epoch": 2.612661207277007,
+      "grad_norm": 0.6590286866274873,
+      "learning_rate": 8.572047936229355e-07,
+      "loss": 0.6011,
+      "step": 1665
+    },
+    {
+      "epoch": 2.6142303731672634,
+      "grad_norm": 0.6279250782414297,
+      "learning_rate": 8.503501087422305e-07,
+      "loss": 0.6022,
+      "step": 1666
+    },
+    {
+      "epoch": 2.6157995390575195,
+      "grad_norm": 0.6325486947752222,
+      "learning_rate": 8.435217237119719e-07,
+      "loss": 0.6369,
+      "step": 1667
+    },
+    {
+      "epoch": 2.617368704947776,
+      "grad_norm": 0.6177234534565488,
+      "learning_rate": 8.367196581597503e-07,
+      "loss": 0.5925,
+      "step": 1668
+    },
+    {
+      "epoch": 2.6189378708380326,
+      "grad_norm": 0.6165720105584769,
+      "learning_rate": 8.299439316374858e-07,
+      "loss": 0.6104,
+      "step": 1669
+    },
+    {
+      "epoch": 2.620507036728289,
+      "grad_norm": 0.5954218678375357,
+      "learning_rate": 8.231945636214068e-07,
+      "loss": 0.626,
+      "step": 1670
+    },
+    {
+      "epoch": 2.622076202618546,
+      "grad_norm": 0.6432133768174138,
+      "learning_rate": 8.164715735119599e-07,
+      "loss": 0.6144,
+      "step": 1671
+    },
+    {
+      "epoch": 2.623645368508802,
+      "grad_norm": 0.6193005560267426,
+      "learning_rate": 8.097749806337851e-07,
+      "loss": 0.5937,
+      "step": 1672
+    },
+    {
+      "epoch": 2.6252145343990585,
+      "grad_norm": 0.6006556888678798,
+      "learning_rate": 8.031048042356393e-07,
+      "loss": 0.6118,
+      "step": 1673
+    },
+    {
+      "epoch": 2.626783700289315,
+      "grad_norm": 0.6159492173114185,
+      "learning_rate": 7.964610634903458e-07,
+      "loss": 0.601,
+      "step": 1674
+    },
+    {
+      "epoch": 2.628352866179571,
+      "grad_norm": 0.6323529032694348,
+      "learning_rate": 7.898437774947442e-07,
+      "loss": 0.6012,
+      "step": 1675
+    },
+    {
+      "epoch": 2.6299220320698278,
+      "grad_norm": 0.6154230134847168,
+      "learning_rate": 7.832529652696319e-07,
+      "loss": 0.6138,
+      "step": 1676
+    },
+    {
+      "epoch": 2.6314911979600843,
+      "grad_norm": 0.6357829028911038,
+      "learning_rate": 7.766886457597089e-07,
+      "loss": 0.6154,
+      "step": 1677
+    },
+    {
+      "epoch": 2.633060363850341,
+      "grad_norm": 0.6397687770530546,
+      "learning_rate": 7.701508378335243e-07,
+      "loss": 0.6184,
+      "step": 1678
+    },
+    {
+      "epoch": 2.6346295297405975,
+      "grad_norm": 0.6347302023334771,
+      "learning_rate": 7.636395602834223e-07,
+      "loss": 0.6136,
+      "step": 1679
+    },
+    {
+      "epoch": 2.6361986956308536,
+      "grad_norm": 0.622009642155309,
+      "learning_rate": 7.571548318254874e-07,
+      "loss": 0.6069,
+      "step": 1680
+    },
+    {
+      "epoch": 2.63776786152111,
+      "grad_norm": 0.6299062315815387,
+      "learning_rate": 7.506966710994945e-07,
+      "loss": 0.6217,
+      "step": 1681
+    },
+    {
+      "epoch": 2.6393370274113668,
+      "grad_norm": 0.6285087030260604,
+      "learning_rate": 7.442650966688469e-07,
+      "loss": 0.6016,
+      "step": 1682
+    },
+    {
+      "epoch": 2.640906193301623,
+      "grad_norm": 0.6398852513724924,
+      "learning_rate": 7.378601270205299e-07,
+      "loss": 0.626,
+      "step": 1683
+    },
+    {
+      "epoch": 2.6424753591918795,
+      "grad_norm": 0.6125662679027378,
+      "learning_rate": 7.314817805650576e-07,
+      "loss": 0.6201,
+      "step": 1684
+    },
+    {
+      "epoch": 2.644044525082136,
+      "grad_norm": 0.6022790662824037,
+      "learning_rate": 7.251300756364154e-07,
+      "loss": 0.6042,
+      "step": 1685
+    },
+    {
+      "epoch": 2.6456136909723926,
+      "grad_norm": 0.6399851300869409,
+      "learning_rate": 7.188050304920113e-07,
+      "loss": 0.6154,
+      "step": 1686
+    },
+    {
+      "epoch": 2.647182856862649,
+      "grad_norm": 0.6176334154422897,
+      "learning_rate": 7.125066633126221e-07,
+      "loss": 0.6053,
+      "step": 1687
+    },
+    {
+      "epoch": 2.6487520227529053,
+      "grad_norm": 0.641212058982805,
+      "learning_rate": 7.062349922023414e-07,
+      "loss": 0.5947,
+      "step": 1688
+    },
+    {
+      "epoch": 2.650321188643162,
+      "grad_norm": 0.6223828261776106,
+      "learning_rate": 6.999900351885258e-07,
+      "loss": 0.587,
+      "step": 1689
+    },
+    {
+      "epoch": 2.6518903545334185,
+      "grad_norm": 0.6466995608169822,
+      "learning_rate": 6.937718102217461e-07,
+      "loss": 0.6303,
+      "step": 1690
+    },
+    {
+      "epoch": 2.6534595204236746,
+      "grad_norm": 0.6277732028739825,
+      "learning_rate": 6.875803351757371e-07,
+      "loss": 0.6002,
+      "step": 1691
+    },
+    {
+      "epoch": 2.655028686313931,
+      "grad_norm": 0.5966572750248873,
+      "learning_rate": 6.814156278473338e-07,
+      "loss": 0.6093,
+      "step": 1692
+    },
+    {
+      "epoch": 2.6565978522041878,
+      "grad_norm": 0.6192878936800507,
+      "learning_rate": 6.752777059564431e-07,
+      "loss": 0.6265,
+      "step": 1693
+    },
+    {
+      "epoch": 2.6581670180944443,
+      "grad_norm": 0.6059962082795073,
+      "learning_rate": 6.691665871459674e-07,
+      "loss": 0.584,
+      "step": 1694
+    },
+    {
+      "epoch": 2.659736183984701,
+      "grad_norm": 0.6269888314020751,
+      "learning_rate": 6.630822889817734e-07,
+      "loss": 0.6096,
+      "step": 1695
+    },
+    {
+      "epoch": 2.661305349874957,
+      "grad_norm": 0.6256330750742394,
+      "learning_rate": 6.570248289526326e-07,
+      "loss": 0.6146,
+      "step": 1696
+    },
+    {
+      "epoch": 2.6628745157652136,
+      "grad_norm": 0.6243811466130468,
+      "learning_rate": 6.509942244701717e-07,
+      "loss": 0.5997,
+      "step": 1697
+    },
+    {
+      "epoch": 2.66444368165547,
+      "grad_norm": 0.6144060701586297,
+      "learning_rate": 6.449904928688244e-07,
+      "loss": 0.6135,
+      "step": 1698
+    },
+    {
+      "epoch": 2.6660128475457263,
+      "grad_norm": 0.6476907715758169,
+      "learning_rate": 6.390136514057788e-07,
+      "loss": 0.6231,
+      "step": 1699
+    },
+    {
+      "epoch": 2.667582013435983,
+      "grad_norm": 0.6311623347938916,
+      "learning_rate": 6.330637172609333e-07,
+      "loss": 0.629,
+      "step": 1700
+    },
+    {
+      "epoch": 2.6691511793262395,
+      "grad_norm": 0.6046918718544069,
+      "learning_rate": 6.271407075368385e-07,
+      "loss": 0.6084,
+      "step": 1701
+    },
+    {
+      "epoch": 2.6707203452164956,
+      "grad_norm": 0.6286572909511952,
+      "learning_rate": 6.212446392586613e-07,
+      "loss": 0.6224,
+      "step": 1702
+    },
+    {
+      "epoch": 2.672289511106752,
+      "grad_norm": 0.6207655106756741,
+      "learning_rate": 6.15375529374116e-07,
+      "loss": 0.6282,
+      "step": 1703
+    },
+    {
+      "epoch": 2.6738586769970087,
+      "grad_norm": 0.6302683784481893,
+      "learning_rate": 6.095333947534421e-07,
+      "loss": 0.6027,
+      "step": 1704
+    },
+    {
+      "epoch": 2.6754278428872653,
+      "grad_norm": 0.6099747919191167,
+      "learning_rate": 6.037182521893247e-07,
+      "loss": 0.6092,
+      "step": 1705
+    },
+    {
+      "epoch": 2.676997008777522,
+      "grad_norm": 0.5981094613416341,
+      "learning_rate": 5.979301183968789e-07,
+      "loss": 0.6048,
+      "step": 1706
+    },
+    {
+      "epoch": 2.678566174667778,
+      "grad_norm": 0.6024927225313307,
+      "learning_rate": 5.921690100135713e-07,
+      "loss": 0.5849,
+      "step": 1707
+    },
+    {
+      "epoch": 2.6801353405580346,
+      "grad_norm": 0.6330637434067633,
+      "learning_rate": 5.864349435991978e-07,
+      "loss": 0.6181,
+      "step": 1708
+    },
+    {
+      "epoch": 2.681704506448291,
+      "grad_norm": 0.6458744268718086,
+      "learning_rate": 5.807279356358164e-07,
+      "loss": 0.6278,
+      "step": 1709
+    },
+    {
+      "epoch": 2.6832736723385473,
+      "grad_norm": 0.6324015374998069,
+      "learning_rate": 5.750480025277116e-07,
+      "loss": 0.6082,
+      "step": 1710
+    },
+    {
+      "epoch": 2.684842838228804,
+      "grad_norm": 0.613714779013516,
+      "learning_rate": 5.693951606013448e-07,
+      "loss": 0.6121,
+      "step": 1711
+    },
+    {
+      "epoch": 2.6864120041190604,
+      "grad_norm": 0.5997658841682052,
+      "learning_rate": 5.637694261053029e-07,
+      "loss": 0.6178,
+      "step": 1712
+    },
+    {
+      "epoch": 2.687981170009317,
+      "grad_norm": 0.6325968452961189,
+      "learning_rate": 5.581708152102583e-07,
+      "loss": 0.6083,
+      "step": 1713
+    },
+    {
+      "epoch": 2.6895503358995736,
+      "grad_norm": 0.6409704318561665,
+      "learning_rate": 5.525993440089161e-07,
+      "loss": 0.6181,
+      "step": 1714
+    },
+    {
+      "epoch": 2.6911195017898297,
+      "grad_norm": 0.6204250558165678,
+      "learning_rate": 5.470550285159704e-07,
+      "loss": 0.6077,
+      "step": 1715
+    },
+    {
+      "epoch": 2.6926886676800863,
+      "grad_norm": 0.6279969649977819,
+      "learning_rate": 5.415378846680641e-07,
+      "loss": 0.6044,
+      "step": 1716
+    },
+    {
+      "epoch": 2.694257833570343,
+      "grad_norm": 0.6280679478451149,
+      "learning_rate": 5.360479283237319e-07,
+      "loss": 0.5867,
+      "step": 1717
+    },
+    {
+      "epoch": 2.695826999460599,
+      "grad_norm": 0.6131444761744856,
+      "learning_rate": 5.305851752633628e-07,
+      "loss": 0.6166,
+      "step": 1718
+    },
+    {
+      "epoch": 2.6973961653508556,
+      "grad_norm": 0.6382530040309086,
+      "learning_rate": 5.251496411891521e-07,
+      "loss": 0.6038,
+      "step": 1719
+    },
+    {
+      "epoch": 2.698965331241112,
+      "grad_norm": 0.5920627773228943,
+      "learning_rate": 5.197413417250563e-07,
+      "loss": 0.6059,
+      "step": 1720
+    },
+    {
+      "epoch": 2.7005344971313687,
+      "grad_norm": 0.6121144084531541,
+      "learning_rate": 5.143602924167479e-07,
+      "loss": 0.618,
+      "step": 1721
+    },
+    {
+      "epoch": 2.7021036630216253,
+      "grad_norm": 0.6400788668283861,
+      "learning_rate": 5.090065087315721e-07,
+      "loss": 0.6154,
+      "step": 1722
+    },
+    {
+      "epoch": 2.7036728289118814,
+      "grad_norm": 0.5962691473983363,
+      "learning_rate": 5.036800060585023e-07,
+      "loss": 0.5994,
+      "step": 1723
+    },
+    {
+      "epoch": 2.705241994802138,
+      "grad_norm": 0.6067094750757442,
+      "learning_rate": 4.983807997080925e-07,
+      "loss": 0.6158,
+      "step": 1724
+    },
+    {
+      "epoch": 2.7068111606923946,
+      "grad_norm": 0.6131021891472516,
+      "learning_rate": 4.931089049124405e-07,
+      "loss": 0.5968,
+      "step": 1725
+    },
+    {
+      "epoch": 2.7083803265826507,
+      "grad_norm": 0.6159120657158506,
+      "learning_rate": 4.878643368251323e-07,
+      "loss": 0.6099,
+      "step": 1726
+    },
+    {
+      "epoch": 2.7099494924729073,
+      "grad_norm": 0.6219286721157657,
+      "learning_rate": 4.826471105212149e-07,
+      "loss": 0.5978,
+      "step": 1727
+    },
+    {
+      "epoch": 2.711518658363164,
+      "grad_norm": 0.6330727581857869,
+      "learning_rate": 4.774572409971334e-07,
+      "loss": 0.6049,
+      "step": 1728
+    },
+    {
+      "epoch": 2.7130878242534204,
+      "grad_norm": 0.619731754175007,
+      "learning_rate": 4.722947431707092e-07,
+      "loss": 0.6092,
+      "step": 1729
+    },
+    {
+      "epoch": 2.714656990143677,
+      "grad_norm": 0.6392807152782485,
+      "learning_rate": 4.671596318810745e-07,
+      "loss": 0.609,
+      "step": 1730
+    },
+    {
+      "epoch": 2.716226156033933,
+      "grad_norm": 0.626448411328319,
+      "learning_rate": 4.620519218886543e-07,
+      "loss": 0.6038,
+      "step": 1731
+    },
+    {
+      "epoch": 2.7177953219241897,
+      "grad_norm": 0.6215557330430416,
+      "learning_rate": 4.5697162787509686e-07,
+      "loss": 0.6106,
+      "step": 1732
+    },
+    {
+      "epoch": 2.7193644878144463,
+      "grad_norm": 0.6246367760947891,
+      "learning_rate": 4.5191876444325986e-07,
+      "loss": 0.6396,
+      "step": 1733
+    },
+    {
+      "epoch": 2.7209336537047024,
+      "grad_norm": 0.6139685110655894,
+      "learning_rate": 4.468933461171432e-07,
+      "loss": 0.6217,
+      "step": 1734
+    },
+    {
+      "epoch": 2.722502819594959,
+      "grad_norm": 0.6287472961922276,
+      "learning_rate": 4.418953873418641e-07,
+      "loss": 0.5999,
+      "step": 1735
+    },
+    {
+      "epoch": 2.7240719854852156,
+      "grad_norm": 0.6360286353553899,
+      "learning_rate": 4.3692490248360755e-07,
+      "loss": 0.6041,
+      "step": 1736
+    },
+    {
+      "epoch": 2.725641151375472,
+      "grad_norm": 0.6111515665662038,
+      "learning_rate": 4.319819058295893e-07,
+      "loss": 0.6013,
+      "step": 1737
+    },
+    {
+      "epoch": 2.7272103172657287,
+      "grad_norm": 0.6299515696538398,
+      "learning_rate": 4.2706641158800855e-07,
+      "loss": 0.592,
+      "step": 1738
+    },
+    {
+      "epoch": 2.728779483155985,
+      "grad_norm": 0.6285255481204092,
+      "learning_rate": 4.221784338880153e-07,
+      "loss": 0.6216,
+      "step": 1739
+    },
+    {
+      "epoch": 2.7303486490462414,
+      "grad_norm": 0.6274256875476529,
+      "learning_rate": 4.173179867796662e-07,
+      "loss": 0.6168,
+      "step": 1740
+    },
+    {
+      "epoch": 2.731917814936498,
+      "grad_norm": 0.6145593914482529,
+      "learning_rate": 4.124850842338779e-07,
+      "loss": 0.6231,
+      "step": 1741
+    },
+    {
+      "epoch": 2.733486980826754,
+      "grad_norm": 0.6315720550867571,
+      "learning_rate": 4.076797401423993e-07,
+      "loss": 0.6019,
+      "step": 1742
+    },
+    {
+      "epoch": 2.7350561467170107,
+      "grad_norm": 0.6394659496581799,
+      "learning_rate": 4.0290196831776264e-07,
+      "loss": 0.6104,
+      "step": 1743
+    },
+    {
+      "epoch": 2.7366253126072673,
+      "grad_norm": 0.6239458047902254,
+      "learning_rate": 3.981517824932446e-07,
+      "loss": 0.6227,
+      "step": 1744
+    },
+    {
+      "epoch": 2.7381944784975234,
+      "grad_norm": 0.6229992986733072,
+      "learning_rate": 3.9342919632283093e-07,
+      "loss": 0.6048,
+      "step": 1745
+    },
+    {
+      "epoch": 2.73976364438778,
+      "grad_norm": 0.6087428019577692,
+      "learning_rate": 3.8873422338117084e-07,
+      "loss": 0.5854,
+      "step": 1746
+    },
+    {
+      "epoch": 2.7413328102780365,
+      "grad_norm": 0.626181564902417,
+      "learning_rate": 3.8406687716354695e-07,
+      "loss": 0.608,
+      "step": 1747
+    },
+    {
+      "epoch": 2.742901976168293,
+      "grad_norm": 0.6232293317789762,
+      "learning_rate": 3.7942717108582663e-07,
+      "loss": 0.6225,
+      "step": 1748
+    },
+    {
+      "epoch": 2.7444711420585497,
+      "grad_norm": 0.6174512725503123,
+      "learning_rate": 3.748151184844295e-07,
+      "loss": 0.6034,
+      "step": 1749
+    },
+    {
+      "epoch": 2.746040307948806,
+      "grad_norm": 0.6118531305616566,
+      "learning_rate": 3.702307326162924e-07,
+      "loss": 0.6006,
+      "step": 1750
+    },
+    {
+      "epoch": 2.7476094738390624,
+      "grad_norm": 0.6303333813449055,
+      "learning_rate": 3.656740266588177e-07,
+      "loss": 0.6116,
+      "step": 1751
+    },
+    {
+      "epoch": 2.749178639729319,
+      "grad_norm": 0.6771241557171276,
+      "learning_rate": 3.611450137098538e-07,
+      "loss": 0.6087,
+      "step": 1752
+    },
+    {
+      "epoch": 2.750747805619575,
+      "grad_norm": 0.6108654732292905,
+      "learning_rate": 3.5664370678763825e-07,
+      "loss": 0.6211,
+      "step": 1753
+    },
+    {
+      "epoch": 2.7523169715098317,
+      "grad_norm": 0.6290060081596168,
+      "learning_rate": 3.5217011883077914e-07,
+      "loss": 0.6128,
+      "step": 1754
+    },
+    {
+      "epoch": 2.7538861374000883,
+      "grad_norm": 0.6159020657547403,
+      "learning_rate": 3.477242626982047e-07,
+      "loss": 0.599,
+      "step": 1755
+    },
+    {
+      "epoch": 2.755455303290345,
+      "grad_norm": 0.6422097275912046,
+      "learning_rate": 3.4330615116912933e-07,
+      "loss": 0.6112,
+      "step": 1756
+    },
+    {
+      "epoch": 2.7570244691806014,
+      "grad_norm": 0.6126211203416192,
+      "learning_rate": 3.389157969430201e-07,
+      "loss": 0.6183,
+      "step": 1757
+    },
+    {
+      "epoch": 2.7585936350708575,
+      "grad_norm": 0.6239291110253911,
+      "learning_rate": 3.345532126395579e-07,
+      "loss": 0.6077,
+      "step": 1758
+    },
+    {
+      "epoch": 2.760162800961114,
+      "grad_norm": 0.6342786200286286,
+      "learning_rate": 3.302184107986017e-07,
+      "loss": 0.5963,
+      "step": 1759
+    },
+    {
+      "epoch": 2.7617319668513707,
+      "grad_norm": 0.6327285715488448,
+      "learning_rate": 3.259114038801503e-07,
+      "loss": 0.6047,
+      "step": 1760
+    },
+    {
+      "epoch": 2.763301132741627,
+      "grad_norm": 0.6525724232968162,
+      "learning_rate": 3.216322042643127e-07,
+      "loss": 0.6084,
+      "step": 1761
+    },
+    {
+      "epoch": 2.7648702986318834,
+      "grad_norm": 0.6364815901300408,
+      "learning_rate": 3.17380824251261e-07,
+      "loss": 0.6012,
+      "step": 1762
+    },
+    {
+      "epoch": 2.76643946452214,
+      "grad_norm": 0.6288438646347232,
+      "learning_rate": 3.1315727606121226e-07,
+      "loss": 0.6187,
+      "step": 1763
+    },
+    {
+      "epoch": 2.7680086304123965,
+      "grad_norm": 0.6273627975177132,
+      "learning_rate": 3.0896157183437305e-07,
+      "loss": 0.6077,
+      "step": 1764
+    },
+    {
+      "epoch": 2.769577796302653,
+      "grad_norm": 0.6366765864678001,
+      "learning_rate": 3.0479372363092506e-07,
+      "loss": 0.6104,
+      "step": 1765
+    },
+    {
+      "epoch": 2.7711469621929092,
+      "grad_norm": 0.636397196659369,
+      "learning_rate": 3.0065374343096974e-07,
+      "loss": 0.6117,
+      "step": 1766
+    },
+    {
+      "epoch": 2.772716128083166,
+      "grad_norm": 0.6398011152595193,
+      "learning_rate": 2.965416431345147e-07,
+      "loss": 0.5909,
+      "step": 1767
+    },
+    {
+      "epoch": 2.7742852939734224,
+      "grad_norm": 0.6283700991353168,
+      "learning_rate": 2.9245743456142283e-07,
+      "loss": 0.5978,
+      "step": 1768
+    },
+    {
+      "epoch": 2.7758544598636785,
+      "grad_norm": 0.6160353193768956,
+      "learning_rate": 2.8840112945138553e-07,
+      "loss": 0.6246,
+      "step": 1769
+    },
+    {
+      "epoch": 2.777423625753935,
+      "grad_norm": 0.6381385874899927,
+      "learning_rate": 2.8437273946388954e-07,
+      "loss": 0.6193,
+      "step": 1770
+    },
+    {
+      "epoch": 2.7789927916441917,
+      "grad_norm": 0.6257187563897191,
+      "learning_rate": 2.803722761781824e-07,
+      "loss": 0.5951,
+      "step": 1771
+    },
+    {
+      "epoch": 2.7805619575344482,
+      "grad_norm": 0.6291003830444258,
+      "learning_rate": 2.763997510932381e-07,
+      "loss": 0.6173,
+      "step": 1772
+    },
+    {
+      "epoch": 2.782131123424705,
+      "grad_norm": 0.6183145230722933,
+      "learning_rate": 2.7245517562772385e-07,
+      "loss": 0.6232,
+      "step": 1773
+    },
+    {
+      "epoch": 2.783700289314961,
+      "grad_norm": 0.6048824669502935,
+      "learning_rate": 2.685385611199698e-07,
+      "loss": 0.6141,
+      "step": 1774
+    },
+    {
+      "epoch": 2.7852694552052175,
+      "grad_norm": 0.6155741411631986,
+      "learning_rate": 2.646499188279328e-07,
+      "loss": 0.5949,
+      "step": 1775
+    },
+    {
+      "epoch": 2.786838621095474,
+      "grad_norm": 0.5866504981490761,
+      "learning_rate": 2.607892599291695e-07,
+      "loss": 0.6109,
+      "step": 1776
+    },
+    {
+      "epoch": 2.78840778698573,
+      "grad_norm": 0.6342916046998135,
+      "learning_rate": 2.5695659552079866e-07,
+      "loss": 0.5994,
+      "step": 1777
+    },
+    {
+      "epoch": 2.789976952875987,
+      "grad_norm": 0.6202593091452479,
+      "learning_rate": 2.5315193661947013e-07,
+      "loss": 0.5905,
+      "step": 1778
+    },
+    {
+      "epoch": 2.7915461187662434,
+      "grad_norm": 0.6458404780592726,
+      "learning_rate": 2.493752941613348e-07,
+      "loss": 0.6195,
+      "step": 1779
+    },
+    {
+      "epoch": 2.7931152846565,
+      "grad_norm": 0.6359881774462259,
+      "learning_rate": 2.4562667900201677e-07,
+      "loss": 0.6165,
+      "step": 1780
+    },
+    {
+      "epoch": 2.7946844505467565,
+      "grad_norm": 0.6034287806264477,
+      "learning_rate": 2.4190610191657246e-07,
+      "loss": 0.6085,
+      "step": 1781
+    },
+    {
+      "epoch": 2.7962536164370126,
+      "grad_norm": 0.598533459476681,
+      "learning_rate": 2.3821357359946838e-07,
+      "loss": 0.586,
+      "step": 1782
+    },
+    {
+      "epoch": 2.797822782327269,
+      "grad_norm": 0.6337616449867283,
+      "learning_rate": 2.345491046645465e-07,
+      "loss": 0.6072,
+      "step": 1783
+    },
+    {
+      "epoch": 2.799391948217526,
+      "grad_norm": 0.6180899901002155,
+      "learning_rate": 2.309127056449989e-07,
+      "loss": 0.6302,
+      "step": 1784
+    },
+    {
+      "epoch": 2.800961114107782,
+      "grad_norm": 0.6274120475368292,
+      "learning_rate": 2.2730438699332446e-07,
+      "loss": 0.6182,
+      "step": 1785
+    },
+    {
+      "epoch": 2.8025302799980385,
+      "grad_norm": 0.6286903227810913,
+      "learning_rate": 2.237241590813155e-07,
+      "loss": 0.6119,
+      "step": 1786
+    },
+    {
+      "epoch": 2.804099445888295,
+      "grad_norm": 0.6200394016947913,
+      "learning_rate": 2.2017203220001559e-07,
+      "loss": 0.6252,
+      "step": 1787
+    },
+    {
+      "epoch": 2.805668611778551,
+      "grad_norm": 0.6095346503938988,
+      "learning_rate": 2.1664801655969735e-07,
+      "loss": 0.6071,
+      "step": 1788
+    },
+    {
+      "epoch": 2.8072377776688078,
+      "grad_norm": 0.6070171687559665,
+      "learning_rate": 2.1315212228982584e-07,
+      "loss": 0.5908,
+      "step": 1789
+    },
+    {
+      "epoch": 2.8088069435590644,
+      "grad_norm": 0.610167775406647,
+      "learning_rate": 2.0968435943904075e-07,
+      "loss": 0.6047,
+      "step": 1790
+    },
+    {
+      "epoch": 2.810376109449321,
+      "grad_norm": 0.6092962332697563,
+      "learning_rate": 2.062447379751109e-07,
+      "loss": 0.6007,
+      "step": 1791
+    },
+    {
+      "epoch": 2.8119452753395775,
+      "grad_norm": 0.6069155785582789,
+      "learning_rate": 2.028332677849254e-07,
+      "loss": 0.588,
+      "step": 1792
+    },
+    {
+      "epoch": 2.8135144412298336,
+      "grad_norm": 0.6195372476122729,
+      "learning_rate": 1.9944995867444806e-07,
+      "loss": 0.6263,
+      "step": 1793
+    },
+    {
+      "epoch": 2.81508360712009,
+      "grad_norm": 0.6258172521903729,
+      "learning_rate": 1.9609482036869853e-07,
+      "loss": 0.6113,
+      "step": 1794
+    },
+    {
+      "epoch": 2.8166527730103468,
+      "grad_norm": 0.5957456998484894,
+      "learning_rate": 1.927678625117224e-07,
+      "loss": 0.6173,
+      "step": 1795
+    },
+    {
+      "epoch": 2.818221938900603,
+      "grad_norm": 0.6013420772952011,
+      "learning_rate": 1.8946909466656115e-07,
+      "loss": 0.6294,
+      "step": 1796
+    },
+    {
+      "epoch": 2.8197911047908595,
+      "grad_norm": 0.6178391550460209,
+      "learning_rate": 1.8619852631523105e-07,
+      "loss": 0.6075,
+      "step": 1797
+    },
+    {
+      "epoch": 2.821360270681116,
+      "grad_norm": 0.6115085507289373,
+      "learning_rate": 1.8295616685868434e-07,
+      "loss": 0.61,
+      "step": 1798
+    },
+    {
+      "epoch": 2.8229294365713726,
+      "grad_norm": 0.6332524030633829,
+      "learning_rate": 1.797420256167981e-07,
+      "loss": 0.6111,
+      "step": 1799
+    },
+    {
+      "epoch": 2.824498602461629,
+      "grad_norm": 0.6189568380201572,
+      "learning_rate": 1.7655611182832767e-07,
+      "loss": 0.617,
+      "step": 1800
+    },
+    {
+      "epoch": 2.8260677683518853,
+      "grad_norm": 0.592681618893148,
+      "learning_rate": 1.7339843465090322e-07,
+      "loss": 0.6154,
+      "step": 1801
+    },
+    {
+      "epoch": 2.827636934242142,
+      "grad_norm": 0.6030740318403597,
+      "learning_rate": 1.7026900316098217e-07,
+      "loss": 0.6328,
+      "step": 1802
+    },
+    {
+      "epoch": 2.8292061001323985,
+      "grad_norm": 0.6496354626918339,
+      "learning_rate": 1.6716782635383677e-07,
+      "loss": 0.6122,
+      "step": 1803
+    },
+    {
+      "epoch": 2.8307752660226546,
+      "grad_norm": 0.6192955883853107,
+      "learning_rate": 1.640949131435221e-07,
+      "loss": 0.6251,
+      "step": 1804
+    },
+    {
+      "epoch": 2.832344431912911,
+      "grad_norm": 0.6324870043465912,
+      "learning_rate": 1.6105027236285265e-07,
+      "loss": 0.6234,
+      "step": 1805
+    },
+    {
+      "epoch": 2.8339135978031678,
+      "grad_norm": 0.613572627637444,
+      "learning_rate": 1.5803391276337565e-07,
+      "loss": 0.6172,
+      "step": 1806
+    },
+    {
+      "epoch": 2.8354827636934243,
+      "grad_norm": 0.6008663021488783,
+      "learning_rate": 1.5504584301534677e-07,
+      "loss": 0.6136,
+      "step": 1807
+    },
+    {
+      "epoch": 2.837051929583681,
+      "grad_norm": 0.6357660273456727,
+      "learning_rate": 1.520860717077055e-07,
+      "loss": 0.6086,
+      "step": 1808
+    },
+    {
+      "epoch": 2.838621095473937,
+      "grad_norm": 0.6096253514157608,
+      "learning_rate": 1.49154607348051e-07,
+      "loss": 0.605,
+      "step": 1809
+    },
+    {
+      "epoch": 2.8401902613641936,
+      "grad_norm": 0.6379925629048546,
+      "learning_rate": 1.462514583626129e-07,
+      "loss": 0.6072,
+      "step": 1810
+    },
+    {
+      "epoch": 2.84175942725445,
+      "grad_norm": 0.6145059421094945,
+      "learning_rate": 1.4337663309623718e-07,
+      "loss": 0.6267,
+      "step": 1811
+    },
+    {
+      "epoch": 2.8433285931447063,
+      "grad_norm": 0.6422167553048911,
+      "learning_rate": 1.405301398123482e-07,
+      "loss": 0.5996,
+      "step": 1812
+    },
+    {
+      "epoch": 2.844897759034963,
+      "grad_norm": 0.6628217797141939,
+      "learning_rate": 1.3771198669293885e-07,
+      "loss": 0.589,
+      "step": 1813
+    },
+    {
+      "epoch": 2.8464669249252195,
+      "grad_norm": 0.6025837806882417,
+      "learning_rate": 1.3492218183853712e-07,
+      "loss": 0.5896,
+      "step": 1814
+    },
+    {
+      "epoch": 2.848036090815476,
+      "grad_norm": 0.6214085884461172,
+      "learning_rate": 1.3216073326818736e-07,
+      "loss": 0.6026,
+      "step": 1815
+    },
+    {
+      "epoch": 2.8496052567057326,
+      "grad_norm": 0.6445313920113861,
+      "learning_rate": 1.2942764891942793e-07,
+      "loss": 0.6132,
+      "step": 1816
+    },
+    {
+      "epoch": 2.8511744225959887,
+      "grad_norm": 0.6343902514011032,
+      "learning_rate": 1.2672293664826364e-07,
+      "loss": 0.6308,
+      "step": 1817
+    },
+    {
+      "epoch": 2.8527435884862453,
+      "grad_norm": 0.6610384011800102,
+      "learning_rate": 1.2404660422914993e-07,
+      "loss": 0.609,
+      "step": 1818
+    },
+    {
+      "epoch": 2.854312754376502,
+      "grad_norm": 0.6061586499692299,
+      "learning_rate": 1.213986593549621e-07,
+      "loss": 0.6004,
+      "step": 1819
+    },
+    {
+      "epoch": 2.855881920266758,
+      "grad_norm": 0.6077126550131754,
+      "learning_rate": 1.1877910963698502e-07,
+      "loss": 0.6152,
+      "step": 1820
+    },
+    {
+      "epoch": 2.8574510861570146,
+      "grad_norm": 0.6243076938643307,
+      "learning_rate": 1.1618796260487675e-07,
+      "loss": 0.6134,
+      "step": 1821
+    },
+    {
+      "epoch": 2.859020252047271,
+      "grad_norm": 0.6551016168182041,
+      "learning_rate": 1.1362522570665945e-07,
+      "loss": 0.6198,
+      "step": 1822
+    },
+    {
+      "epoch": 2.8605894179375277,
+      "grad_norm": 0.6170006393898833,
+      "learning_rate": 1.1109090630868958e-07,
+      "loss": 0.6149,
+      "step": 1823
+    },
+    {
+      "epoch": 2.8621585838277843,
+      "grad_norm": 0.6025650371690818,
+      "learning_rate": 1.0858501169564439e-07,
+      "loss": 0.5793,
+      "step": 1824
+    },
+    {
+      "epoch": 2.8637277497180404,
+      "grad_norm": 0.6009150271064042,
+      "learning_rate": 1.0610754907049104e-07,
+      "loss": 0.6229,
+      "step": 1825
+    },
+    {
+      "epoch": 2.865296915608297,
+      "grad_norm": 0.633054890173234,
+      "learning_rate": 1.0365852555447642e-07,
+      "loss": 0.6026,
+      "step": 1826
+    },
+    {
+      "epoch": 2.8668660814985536,
+      "grad_norm": 0.6073046160185441,
+      "learning_rate": 1.0123794818709732e-07,
+      "loss": 0.6193,
+      "step": 1827
+    },
+    {
+      "epoch": 2.8684352473888097,
+      "grad_norm": 0.6893878610570311,
+      "learning_rate": 9.884582392608921e-08,
+      "loss": 0.6158,
+      "step": 1828
+    },
+    {
+      "epoch": 2.8700044132790663,
+      "grad_norm": 0.6158337050229648,
+      "learning_rate": 9.64821596473986e-08,
+      "loss": 0.6226,
+      "step": 1829
+    },
+    {
+      "epoch": 2.871573579169323,
+      "grad_norm": 0.6057275738681348,
+      "learning_rate": 9.414696214516628e-08,
+      "loss": 0.6116,
+      "step": 1830
+    },
+    {
+      "epoch": 2.873142745059579,
+      "grad_norm": 0.6296795526642787,
+      "learning_rate": 9.184023813170961e-08,
+      "loss": 0.6148,
+      "step": 1831
+    },
+    {
+      "epoch": 2.8747119109498356,
+      "grad_norm": 0.6371941823339383,
+      "learning_rate": 8.956199423750033e-08,
+      "loss": 0.6054,
+      "step": 1832
+    },
+    {
+      "epoch": 2.876281076840092,
+      "grad_norm": 0.6246056754726999,
+      "learning_rate": 8.731223701114566e-08,
+      "loss": 0.6216,
+      "step": 1833
+    },
+    {
+      "epoch": 2.8778502427303487,
+      "grad_norm": 0.6218270704544585,
+      "learning_rate": 8.509097291937273e-08,
+      "loss": 0.6105,
+      "step": 1834
+    },
+    {
+      "epoch": 2.8794194086206053,
+      "grad_norm": 0.6028587942745154,
+      "learning_rate": 8.289820834700757e-08,
+      "loss": 0.6316,
+      "step": 1835
+    },
+    {
+      "epoch": 2.8809885745108614,
+      "grad_norm": 0.6197897768450511,
+      "learning_rate": 8.073394959695391e-08,
+      "loss": 0.6104,
+      "step": 1836
+    },
+    {
+      "epoch": 2.882557740401118,
+      "grad_norm": 0.6056141265481063,
+      "learning_rate": 7.859820289018105e-08,
+      "loss": 0.6059,
+      "step": 1837
+    },
+    {
+      "epoch": 2.8841269062913746,
+      "grad_norm": 0.6237582952875913,
+      "learning_rate": 7.649097436570164e-08,
+      "loss": 0.6204,
+      "step": 1838
+    },
+    {
+      "epoch": 2.8856960721816307,
+      "grad_norm": 0.6137871564411005,
+      "learning_rate": 7.441227008055607e-08,
+      "loss": 0.599,
+      "step": 1839
+    },
+    {
+      "epoch": 2.8872652380718873,
+      "grad_norm": 0.5809785939935747,
+      "learning_rate": 7.23620960097926e-08,
+      "loss": 0.6171,
+      "step": 1840
+    },
+    {
+      "epoch": 2.888834403962144,
+      "grad_norm": 0.6063946492156929,
+      "learning_rate": 7.034045804645506e-08,
+      "loss": 0.6123,
+      "step": 1841
+    },
+    {
+      "epoch": 2.8904035698524004,
+      "grad_norm": 0.6276637285570226,
+      "learning_rate": 6.834736200155956e-08,
+      "loss": 0.6131,
+      "step": 1842
+    },
+    {
+      "epoch": 2.891972735742657,
+      "grad_norm": 0.6190183279985443,
+      "learning_rate": 6.638281360408339e-08,
+      "loss": 0.619,
+      "step": 1843
+    },
+    {
+      "epoch": 2.893541901632913,
+      "grad_norm": 0.6238441748545265,
+      "learning_rate": 6.444681850094392e-08,
+      "loss": 0.6276,
+      "step": 1844
+    },
+    {
+      "epoch": 2.8951110675231697,
+      "grad_norm": 0.6379794182036369,
+      "learning_rate": 6.253938225698863e-08,
+      "loss": 0.6142,
+      "step": 1845
+    },
+    {
+      "epoch": 2.8966802334134263,
+      "grad_norm": 0.591226045286818,
+      "learning_rate": 6.066051035496956e-08,
+      "loss": 0.6072,
+      "step": 1846
+    },
+    {
+      "epoch": 2.8982493993036824,
+      "grad_norm": 0.6287830824690123,
+      "learning_rate": 5.881020819553773e-08,
+      "loss": 0.6041,
+      "step": 1847
+    },
+    {
+      "epoch": 2.899818565193939,
+      "grad_norm": 0.6297126090441673,
+      "learning_rate": 5.6988481097219886e-08,
+      "loss": 0.6114,
+      "step": 1848
+    },
+    {
+      "epoch": 2.9013877310841956,
+      "grad_norm": 0.629255305646127,
+      "learning_rate": 5.519533429640955e-08,
+      "loss": 0.6163,
+      "step": 1849
+    },
+    {
+      "epoch": 2.902956896974452,
+      "grad_norm": 0.6510614775787666,
+      "learning_rate": 5.343077294734378e-08,
+      "loss": 0.6217,
+      "step": 1850
+    },
+    {
+      "epoch": 2.9045260628647087,
+      "grad_norm": 0.6135214725232454,
+      "learning_rate": 5.1694802122099763e-08,
+      "loss": 0.6136,
+      "step": 1851
+    },
+    {
+      "epoch": 2.906095228754965,
+      "grad_norm": 0.6162932803993775,
+      "learning_rate": 4.9987426810569337e-08,
+      "loss": 0.609,
+      "step": 1852
+    },
+    {
+      "epoch": 2.9076643946452214,
+      "grad_norm": 0.6182593240308235,
+      "learning_rate": 4.8308651920448976e-08,
+      "loss": 0.5812,
+      "step": 1853
+    },
+    {
+      "epoch": 2.909233560535478,
+      "grad_norm": 0.6251815502912391,
+      "learning_rate": 4.665848227722869e-08,
+      "loss": 0.6243,
+      "step": 1854
+    },
+    {
+      "epoch": 2.910802726425734,
+      "grad_norm": 0.6280364432992998,
+      "learning_rate": 4.503692262417203e-08,
+      "loss": 0.6202,
+      "step": 1855
+    },
+    {
+      "epoch": 2.9123718923159907,
+      "grad_norm": 0.6042553815150471,
+      "learning_rate": 4.344397762230834e-08,
+      "loss": 0.6007,
+      "step": 1856
+    },
+    {
+      "epoch": 2.9139410582062473,
+      "grad_norm": 0.6062804696233999,
+      "learning_rate": 4.1879651850414984e-08,
+      "loss": 0.6215,
+      "step": 1857
+    },
+    {
+      "epoch": 2.915510224096504,
+      "grad_norm": 0.6265930786009388,
+      "learning_rate": 4.0343949805006224e-08,
+      "loss": 0.6395,
+      "step": 1858
+    },
+    {
+      "epoch": 2.9170793899867604,
+      "grad_norm": 0.6463660972716352,
+      "learning_rate": 3.8836875900321036e-08,
+      "loss": 0.6017,
+      "step": 1859
+    },
+    {
+      "epoch": 2.9186485558770165,
+      "grad_norm": 0.630618534169613,
+      "learning_rate": 3.735843446830867e-08,
+      "loss": 0.6174,
+      "step": 1860
+    },
+    {
+      "epoch": 2.920217721767273,
+      "grad_norm": 0.6170787683735347,
+      "learning_rate": 3.590862975861642e-08,
+      "loss": 0.6032,
+      "step": 1861
+    },
+    {
+      "epoch": 2.9217868876575297,
+      "grad_norm": 0.6143039913473612,
+      "learning_rate": 3.448746593857744e-08,
+      "loss": 0.6056,
+      "step": 1862
+    },
+    {
+      "epoch": 2.923356053547786,
+      "grad_norm": 0.5986342660526142,
+      "learning_rate": 3.309494709320182e-08,
+      "loss": 0.6214,
+      "step": 1863
+    },
+    {
+      "epoch": 2.9249252194380424,
+      "grad_norm": 0.6093712718987271,
+      "learning_rate": 3.173107722515889e-08,
+      "loss": 0.6182,
+      "step": 1864
+    },
+    {
+      "epoch": 2.926494385328299,
+      "grad_norm": 0.6178238828402622,
+      "learning_rate": 3.0395860254772705e-08,
+      "loss": 0.6155,
+      "step": 1865
+    },
+    {
+      "epoch": 2.9280635512185555,
+      "grad_norm": 0.6460856927157047,
+      "learning_rate": 2.9089300020002097e-08,
+      "loss": 0.6138,
+      "step": 1866
+    },
+    {
+      "epoch": 2.929632717108812,
+      "grad_norm": 0.6146017431779306,
+      "learning_rate": 2.781140027643958e-08,
+      "loss": 0.6066,
+      "step": 1867
+    },
+    {
+      "epoch": 2.9312018829990683,
+      "grad_norm": 0.5992576952826651,
+      "learning_rate": 2.656216469729356e-08,
+      "loss": 0.618,
+      "step": 1868
+    },
+    {
+      "epoch": 2.932771048889325,
+      "grad_norm": 0.5966088207618054,
+      "learning_rate": 2.5341596873379447e-08,
+      "loss": 0.6129,
+      "step": 1869
+    },
+    {
+      "epoch": 2.9343402147795814,
+      "grad_norm": 0.6067118519460458,
+      "learning_rate": 2.4149700313110813e-08,
+      "loss": 0.6266,
+      "step": 1870
+    },
+    {
+      "epoch": 2.9359093806698375,
+      "grad_norm": 0.5883316843022128,
+      "learning_rate": 2.2986478442486026e-08,
+      "loss": 0.6089,
+      "step": 1871
+    },
+    {
+      "epoch": 2.937478546560094,
+      "grad_norm": 0.6295074553287695,
+      "learning_rate": 2.1851934605086057e-08,
+      "loss": 0.6315,
+      "step": 1872
+    },
+    {
+      "epoch": 2.9390477124503507,
+      "grad_norm": 0.6330491159810617,
+      "learning_rate": 2.074607206205226e-08,
+      "loss": 0.6011,
+      "step": 1873
+    },
+    {
+      "epoch": 2.940616878340607,
+      "grad_norm": 0.5985870659231455,
+      "learning_rate": 1.96688939920886e-08,
+      "loss": 0.5912,
+      "step": 1874
+    },
+    {
+      "epoch": 2.9421860442308634,
+      "grad_norm": 0.6338601801216073,
+      "learning_rate": 1.8620403491448334e-08,
+      "loss": 0.6227,
+      "step": 1875
+    },
+    {
+      "epoch": 2.94375521012112,
+      "grad_norm": 0.6058036703571001,
+      "learning_rate": 1.7600603573921794e-08,
+      "loss": 0.5991,
+      "step": 1876
+    },
+    {
+      "epoch": 2.9453243760113765,
+      "grad_norm": 0.6286878897897321,
+      "learning_rate": 1.6609497170834154e-08,
+      "loss": 0.6258,
+      "step": 1877
+    },
+    {
+      "epoch": 2.946893541901633,
+      "grad_norm": 0.6157360265928736,
+      "learning_rate": 1.5647087131031023e-08,
+      "loss": 0.6049,
+      "step": 1878
+    },
+    {
+      "epoch": 2.9484627077918892,
+      "grad_norm": 0.6250367911619527,
+      "learning_rate": 1.4713376220875097e-08,
+      "loss": 0.6107,
+      "step": 1879
+    },
+    {
+      "epoch": 2.950031873682146,
+      "grad_norm": 0.5996011159704814,
+      "learning_rate": 1.3808367124233945e-08,
+      "loss": 0.6071,
+      "step": 1880
+    },
+    {
+      "epoch": 2.9516010395724024,
+      "grad_norm": 0.5991967501581563,
+      "learning_rate": 1.2932062442475579e-08,
+      "loss": 0.5944,
+      "step": 1881
+    },
+    {
+      "epoch": 2.9531702054626585,
+      "grad_norm": 0.6147068311522826,
+      "learning_rate": 1.2084464694459563e-08,
+      "loss": 0.6068,
+      "step": 1882
+    },
+    {
+      "epoch": 2.954739371352915,
+      "grad_norm": 0.644616917490021,
+      "learning_rate": 1.1265576316530358e-08,
+      "loss": 0.6178,
+      "step": 1883
+    },
+    {
+      "epoch": 2.9563085372431717,
+      "grad_norm": 0.6060120438513794,
+      "learning_rate": 1.0475399662508434e-08,
+      "loss": 0.602,
+      "step": 1884
+    },
+    {
+      "epoch": 2.9578777031334282,
+      "grad_norm": 0.6169294407291479,
+      "learning_rate": 9.713937003685836e-09,
+      "loss": 0.6237,
+      "step": 1885
+    },
+    {
+      "epoch": 2.959446869023685,
+      "grad_norm": 0.6351522926910512,
+      "learning_rate": 8.981190528819517e-09,
+      "loss": 0.6144,
+      "step": 1886
+    },
+    {
+      "epoch": 2.961016034913941,
+      "grad_norm": 0.6495294100031579,
+      "learning_rate": 8.277162344123568e-09,
+      "loss": 0.6382,
+      "step": 1887
+    },
+    {
+      "epoch": 2.9625852008041975,
+      "grad_norm": 0.5984936307497688,
+      "learning_rate": 7.60185447326478e-09,
+      "loss": 0.6145,
+      "step": 1888
+    },
+    {
+      "epoch": 2.964154366694454,
+      "grad_norm": 0.6159161135481825,
+      "learning_rate": 6.955268857353759e-09,
+      "loss": 0.6023,
+      "step": 1889
+    },
+    {
+      "epoch": 2.96572353258471,
+      "grad_norm": 0.6281340357840098,
+      "learning_rate": 6.3374073549438144e-09,
+      "loss": 0.6166,
+      "step": 1890
+    },
+    {
+      "epoch": 2.967292698474967,
+      "grad_norm": 0.6119302841491187,
+      "learning_rate": 5.748271742023192e-09,
+      "loss": 0.6272,
+      "step": 1891
+    },
+    {
+      "epoch": 2.9688618643652234,
+      "grad_norm": 0.62284830258427,
+      "learning_rate": 5.187863712007302e-09,
+      "loss": 0.617,
+      "step": 1892
+    },
+    {
+      "epoch": 2.97043103025548,
+      "grad_norm": 0.6240913216520743,
+      "learning_rate": 4.6561848757420425e-09,
+      "loss": 0.6063,
+      "step": 1893
+    },
+    {
+      "epoch": 2.9720001961457365,
+      "grad_norm": 0.6264882659448358,
+      "learning_rate": 4.153236761488266e-09,
+      "loss": 0.6168,
+      "step": 1894
+    },
+    {
+      "epoch": 2.9735693620359926,
+      "grad_norm": 0.6255129673312394,
+      "learning_rate": 3.679020814926215e-09,
+      "loss": 0.6165,
+      "step": 1895
+    },
+    {
+      "epoch": 2.975138527926249,
+      "grad_norm": 0.6262445792960478,
+      "learning_rate": 3.2335383991477507e-09,
+      "loss": 0.6019,
+      "step": 1896
+    },
+    {
+      "epoch": 2.976707693816506,
+      "grad_norm": 0.6187188850763499,
+      "learning_rate": 2.8167907946519134e-09,
+      "loss": 0.623,
+      "step": 1897
+    },
+    {
+      "epoch": 2.978276859706762,
+      "grad_norm": 0.6117463895884275,
+      "learning_rate": 2.428779199341591e-09,
+      "loss": 0.6031,
+      "step": 1898
+    },
+    {
+      "epoch": 2.9798460255970185,
+      "grad_norm": 0.6168498754899621,
+      "learning_rate": 2.069504728522409e-09,
+      "loss": 0.5916,
+      "step": 1899
+    },
+    {
+      "epoch": 2.981415191487275,
+      "grad_norm": 0.6015360856359104,
+      "learning_rate": 1.7389684148960694e-09,
+      "loss": 0.5989,
+      "step": 1900
+    },
+    {
+      "epoch": 2.9829843573775316,
+      "grad_norm": 0.6235934671090169,
+      "learning_rate": 1.4371712085603506e-09,
+      "loss": 0.613,
+      "step": 1901
+    },
+    {
+      "epoch": 2.984553523267788,
+      "grad_norm": 0.6006278001550592,
+      "learning_rate": 1.164113977003556e-09,
+      "loss": 0.6167,
+      "step": 1902
+    },
+    {
+      "epoch": 2.9861226891580444,
+      "grad_norm": 0.6120152357372275,
+      "learning_rate": 9.197975051045138e-10,
+      "loss": 0.6139,
+      "step": 1903
+    },
+    {
+      "epoch": 2.987691855048301,
+      "grad_norm": 0.6062239494003836,
+      "learning_rate": 7.042224951292475e-10,
+      "loss": 0.622,
+      "step": 1904
+    },
+    {
+      "epoch": 2.9892610209385575,
+      "grad_norm": 0.6041678467085977,
+      "learning_rate": 5.173895667287543e-10,
+      "loss": 0.6111,
+      "step": 1905
+    },
+    {
+      "epoch": 2.9908301868288136,
+      "grad_norm": 0.6160400943913001,
+      "learning_rate": 3.5929925693789504e-10,
+      "loss": 0.6239,
+      "step": 1906
+    },
+    {
+      "epoch": 2.99239935271907,
+      "grad_norm": 0.6337867043172234,
+      "learning_rate": 2.2995202017317418e-10,
+      "loss": 0.6093,
+      "step": 1907
+    },
+    {
+      "epoch": 2.993968518609327,
+      "grad_norm": 0.5992238579034956,
+      "learning_rate": 1.2934822823051917e-10,
+      "loss": 0.62,
+      "step": 1908
+    },
+    {
+      "epoch": 2.9955376844995834,
+      "grad_norm": 0.6147206067789769,
+      "learning_rate": 5.748817028750075e-11,
+      "loss": 0.6138,
+      "step": 1909
+    },
+    {
+      "epoch": 2.99710685038984,
+      "grad_norm": 0.6291299009729636,
+      "learning_rate": 1.4372052900002432e-11,
+      "loss": 0.603,
+      "step": 1910
+    },
+    {
+      "epoch": 2.998676016280096,
+      "grad_norm": 0.5934030572394862,
+      "learning_rate": 0.0,
+      "loss": 0.5879,
+      "step": 1911
+    },
+    {
+      "epoch": 2.998676016280096,
+      "step": 1911,
+      "total_flos": 7.847567120190669e+16,
+      "train_loss": 0.7266823927623822,
+      "train_runtime": 148650.1651,
+      "train_samples_per_second": 26.339,
+      "train_steps_per_second": 0.013
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1911,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "total_flos": 7.847567120190669e+16,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}