diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,9597 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.99835255354201,
+  "eval_steps": 500,
+  "global_step": 1365,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.002196595277320154,
+      "grad_norm": 0.09368424117565155,
+      "learning_rate": 0.0,
+      "loss": 0.8395,
+      "step": 1
+    },
+    {
+      "epoch": 0.004393190554640308,
+      "grad_norm": 0.07957754284143448,
+      "learning_rate": 2.18978102189781e-06,
+      "loss": 1.0597,
+      "step": 2
+    },
+    {
+      "epoch": 0.006589785831960461,
+      "grad_norm": 0.07214013487100601,
+      "learning_rate": 4.37956204379562e-06,
+      "loss": 0.8589,
+      "step": 3
+    },
+    {
+      "epoch": 0.008786381109280615,
+      "grad_norm": 0.05852081626653671,
+      "learning_rate": 6.56934306569343e-06,
+      "loss": 0.7461,
+      "step": 4
+    },
+    {
+      "epoch": 0.010982976386600769,
+      "grad_norm": 0.0548330582678318,
+      "learning_rate": 8.75912408759124e-06,
+      "loss": 0.8293,
+      "step": 5
+    },
+    {
+      "epoch": 0.013179571663920923,
+      "grad_norm": 0.05844108760356903,
+      "learning_rate": 1.0948905109489049e-05,
+      "loss": 0.6642,
+      "step": 6
+    },
+    {
+      "epoch": 0.015376166941241077,
+      "grad_norm": 0.05149471014738083,
+      "learning_rate": 1.313868613138686e-05,
+      "loss": 0.6719,
+      "step": 7
+    },
+    {
+      "epoch": 0.01757276221856123,
+      "grad_norm": 0.05950519070029259,
+      "learning_rate": 1.532846715328467e-05,
+      "loss": 0.7239,
+      "step": 8
+    },
+    {
+      "epoch": 0.019769357495881382,
+      "grad_norm": 0.0998830646276474,
+      "learning_rate": 1.751824817518248e-05,
+      "loss": 1.0669,
+      "step": 9
+    },
+    {
+      "epoch": 0.021965952773201538,
+      "grad_norm": 0.06528295576572418,
+      "learning_rate": 1.9708029197080292e-05,
+      "loss": 0.741,
+      "step": 10
+    },
+    {
+      "epoch": 0.02416254805052169,
+      "grad_norm": 0.0863526314496994,
+      "learning_rate": 2.1897810218978098e-05,
+      "loss": 0.9336,
+      "step": 11
+    },
+    {
+      "epoch": 0.026359143327841845,
+      "grad_norm": 0.051527634263038635,
+      "learning_rate": 2.408759124087591e-05,
+      "loss": 0.8121,
+      "step": 12
+    },
+    {
+      "epoch": 0.028555738605161998,
+      "grad_norm": 0.06828700751066208,
+      "learning_rate": 2.627737226277372e-05,
+      "loss": 0.7897,
+      "step": 13
+    },
+    {
+      "epoch": 0.030752333882482153,
+      "grad_norm": 0.05828464776277542,
+      "learning_rate": 2.846715328467153e-05,
+      "loss": 0.8097,
+      "step": 14
+    },
+    {
+      "epoch": 0.032948929159802305,
+      "grad_norm": 0.0687652975320816,
+      "learning_rate": 3.065693430656934e-05,
+      "loss": 0.8632,
+      "step": 15
+    },
+    {
+      "epoch": 0.03514552443712246,
+      "grad_norm": 0.05913132429122925,
+      "learning_rate": 3.284671532846715e-05,
+      "loss": 0.6898,
+      "step": 16
+    },
+    {
+      "epoch": 0.037342119714442616,
+      "grad_norm": 0.07138373702764511,
+      "learning_rate": 3.503649635036496e-05,
+      "loss": 0.8091,
+      "step": 17
+    },
+    {
+      "epoch": 0.039538714991762765,
+      "grad_norm": 0.07203706353902817,
+      "learning_rate": 3.722627737226277e-05,
+      "loss": 0.8459,
+      "step": 18
+    },
+    {
+      "epoch": 0.04173531026908292,
+      "grad_norm": 0.05224800482392311,
+      "learning_rate": 3.9416058394160584e-05,
+      "loss": 0.9044,
+      "step": 19
+    },
+    {
+      "epoch": 0.043931905546403076,
+      "grad_norm": 0.06632044166326523,
+      "learning_rate": 4.160583941605839e-05,
+      "loss": 0.8141,
+      "step": 20
+    },
+    {
+      "epoch": 0.04612850082372323,
+      "grad_norm": 0.06928423047065735,
+      "learning_rate": 4.3795620437956196e-05,
+      "loss": 0.8945,
+      "step": 21
+    },
+    {
+      "epoch": 0.04832509610104338,
+      "grad_norm": 0.058030448853969574,
+      "learning_rate": 4.5985401459854016e-05,
+      "loss": 0.7738,
+      "step": 22
+    },
+    {
+      "epoch": 0.050521691378363535,
+      "grad_norm": 0.09114016592502594,
+      "learning_rate": 4.817518248175182e-05,
+      "loss": 0.9901,
+      "step": 23
+    },
+    {
+      "epoch": 0.05271828665568369,
+      "grad_norm": 0.06744793802499771,
+      "learning_rate": 5.036496350364963e-05,
+      "loss": 0.8745,
+      "step": 24
+    },
+    {
+      "epoch": 0.054914881933003847,
+      "grad_norm": 0.067961685359478,
+      "learning_rate": 5.255474452554744e-05,
+      "loss": 0.7961,
+      "step": 25
+    },
+    {
+      "epoch": 0.057111477210323995,
+      "grad_norm": 0.0889795646071434,
+      "learning_rate": 5.4744525547445253e-05,
+      "loss": 0.7916,
+      "step": 26
+    },
+    {
+      "epoch": 0.05930807248764415,
+      "grad_norm": 0.10735788196325302,
+      "learning_rate": 5.693430656934306e-05,
+      "loss": 0.8373,
+      "step": 27
+    },
+    {
+      "epoch": 0.061504667764964306,
+      "grad_norm": 0.06766892969608307,
+      "learning_rate": 5.912408759124087e-05,
+      "loss": 0.79,
+      "step": 28
+    },
+    {
+      "epoch": 0.06370126304228446,
+      "grad_norm": 0.054694995284080505,
+      "learning_rate": 6.131386861313868e-05,
+      "loss": 0.7589,
+      "step": 29
+    },
+    {
+      "epoch": 0.06589785831960461,
+      "grad_norm": 0.06533095985651016,
+      "learning_rate": 6.350364963503648e-05,
+      "loss": 0.7947,
+      "step": 30
+    },
+    {
+      "epoch": 0.06809445359692477,
+      "grad_norm": 0.07474099099636078,
+      "learning_rate": 6.56934306569343e-05,
+      "loss": 0.7711,
+      "step": 31
+    },
+    {
+      "epoch": 0.07029104887424492,
+      "grad_norm": 0.10328605026006699,
+      "learning_rate": 6.788321167883211e-05,
+      "loss": 0.9575,
+      "step": 32
+    },
+    {
+      "epoch": 0.07248764415156507,
+      "grad_norm": 0.0813312754034996,
+      "learning_rate": 7.007299270072992e-05,
+      "loss": 1.1005,
+      "step": 33
+    },
+    {
+      "epoch": 0.07468423942888523,
+      "grad_norm": 0.07534842193126678,
+      "learning_rate": 7.226277372262774e-05,
+      "loss": 0.8917,
+      "step": 34
+    },
+    {
+      "epoch": 0.07688083470620538,
+      "grad_norm": 0.06850400567054749,
+      "learning_rate": 7.445255474452554e-05,
+      "loss": 0.8311,
+      "step": 35
+    },
+    {
+      "epoch": 0.07907742998352553,
+      "grad_norm": 0.08918626606464386,
+      "learning_rate": 7.664233576642336e-05,
+      "loss": 0.9399,
+      "step": 36
+    },
+    {
+      "epoch": 0.08127402526084569,
+      "grad_norm": 0.07072232663631439,
+      "learning_rate": 7.883211678832117e-05,
+      "loss": 0.8374,
+      "step": 37
+    },
+    {
+      "epoch": 0.08347062053816584,
+      "grad_norm": 0.08290290087461472,
+      "learning_rate": 8.102189781021897e-05,
+      "loss": 0.9685,
+      "step": 38
+    },
+    {
+      "epoch": 0.085667215815486,
+      "grad_norm": 0.07387691736221313,
+      "learning_rate": 8.321167883211678e-05,
+      "loss": 0.9722,
+      "step": 39
+    },
+    {
+      "epoch": 0.08786381109280615,
+      "grad_norm": 0.09207072108983994,
+      "learning_rate": 8.540145985401459e-05,
+      "loss": 0.7773,
+      "step": 40
+    },
+    {
+      "epoch": 0.0900604063701263,
+      "grad_norm": 0.06684845685958862,
+      "learning_rate": 8.759124087591239e-05,
+      "loss": 0.8774,
+      "step": 41
+    },
+    {
+      "epoch": 0.09225700164744646,
+      "grad_norm": 0.09227459877729416,
+      "learning_rate": 8.978102189781021e-05,
+      "loss": 0.9656,
+      "step": 42
+    },
+    {
+      "epoch": 0.09445359692476661,
+      "grad_norm": 0.08900696784257889,
+      "learning_rate": 9.197080291970803e-05,
+      "loss": 0.8676,
+      "step": 43
+    },
+    {
+      "epoch": 0.09665019220208676,
+      "grad_norm": 0.07540471106767654,
+      "learning_rate": 9.416058394160584e-05,
+      "loss": 0.9037,
+      "step": 44
+    },
+    {
+      "epoch": 0.09884678747940692,
+      "grad_norm": 0.06883317977190018,
+      "learning_rate": 9.635036496350364e-05,
+      "loss": 0.9332,
+      "step": 45
+    },
+    {
+      "epoch": 0.10104338275672707,
+      "grad_norm": 0.06095288321375847,
+      "learning_rate": 9.854014598540145e-05,
+      "loss": 0.7869,
+      "step": 46
+    },
+    {
+      "epoch": 0.10323997803404723,
+      "grad_norm": 0.05872150510549545,
+      "learning_rate": 0.00010072992700729926,
+      "loss": 0.8301,
+      "step": 47
+    },
+    {
+      "epoch": 0.10543657331136738,
+      "grad_norm": 0.04909675195813179,
+      "learning_rate": 0.00010291970802919708,
+      "loss": 0.8333,
+      "step": 48
+    },
+    {
+      "epoch": 0.10763316858868753,
+      "grad_norm": 0.07996238768100739,
+      "learning_rate": 0.00010510948905109488,
+      "loss": 0.9812,
+      "step": 49
+    },
+    {
+      "epoch": 0.10982976386600769,
+      "grad_norm": 0.07555586099624634,
+      "learning_rate": 0.00010729927007299269,
+      "loss": 0.9272,
+      "step": 50
+    },
+    {
+      "epoch": 0.11202635914332784,
+      "grad_norm": 0.09978742152452469,
+      "learning_rate": 0.00010948905109489051,
+      "loss": 0.9312,
+      "step": 51
+    },
+    {
+      "epoch": 0.11422295442064799,
+      "grad_norm": 0.06448879837989807,
+      "learning_rate": 0.00011167883211678831,
+      "loss": 0.816,
+      "step": 52
+    },
+    {
+      "epoch": 0.11641954969796815,
+      "grad_norm": 0.09131381660699844,
+      "learning_rate": 0.00011386861313868612,
+      "loss": 0.8732,
+      "step": 53
+    },
+    {
+      "epoch": 0.1186161449752883,
+      "grad_norm": 0.07666397094726562,
+      "learning_rate": 0.00011605839416058394,
+      "loss": 0.6953,
+      "step": 54
+    },
+    {
+      "epoch": 0.12081274025260846,
+      "grad_norm": 0.056689534336328506,
+      "learning_rate": 0.00011824817518248174,
+      "loss": 0.7901,
+      "step": 55
+    },
+    {
+      "epoch": 0.12300933552992861,
+      "grad_norm": 0.10934755951166153,
+      "learning_rate": 0.00012043795620437955,
+      "loss": 1.0675,
+      "step": 56
+    },
+    {
+      "epoch": 0.12520593080724876,
+      "grad_norm": 0.061348967254161835,
+      "learning_rate": 0.00012262773722627736,
+      "loss": 0.7428,
+      "step": 57
+    },
+    {
+      "epoch": 0.12740252608456892,
+      "grad_norm": 0.07868394255638123,
+      "learning_rate": 0.00012481751824817516,
+      "loss": 0.959,
+      "step": 58
+    },
+    {
+      "epoch": 0.12959912136188906,
+      "grad_norm": 0.05839238315820694,
+      "learning_rate": 0.00012700729927007297,
+      "loss": 0.715,
+      "step": 59
+    },
+    {
+      "epoch": 0.13179571663920922,
+      "grad_norm": 0.05481383949518204,
+      "learning_rate": 0.00012919708029197077,
+      "loss": 0.7748,
+      "step": 60
+    },
+    {
+      "epoch": 0.13399231191652938,
+      "grad_norm": 0.06998932361602783,
+      "learning_rate": 0.0001313868613138686,
+      "loss": 0.868,
+      "step": 61
+    },
+    {
+      "epoch": 0.13618890719384955,
+      "grad_norm": 0.06081048399209976,
+      "learning_rate": 0.00013357664233576641,
+      "loss": 0.734,
+      "step": 62
+    },
+    {
+      "epoch": 0.13838550247116968,
+      "grad_norm": 0.07183390855789185,
+      "learning_rate": 0.00013576642335766422,
+      "loss": 0.7546,
+      "step": 63
+    },
+    {
+      "epoch": 0.14058209774848984,
+      "grad_norm": 0.07835814356803894,
+      "learning_rate": 0.00013795620437956203,
+      "loss": 0.8754,
+      "step": 64
+    },
+    {
+      "epoch": 0.14277869302581,
+      "grad_norm": 0.04541606828570366,
+      "learning_rate": 0.00014014598540145983,
+      "loss": 0.7274,
+      "step": 65
+    },
+    {
+      "epoch": 0.14497528830313014,
+      "grad_norm": 0.050496987998485565,
+      "learning_rate": 0.00014233576642335764,
+      "loss": 0.6557,
+      "step": 66
+    },
+    {
+      "epoch": 0.1471718835804503,
+      "grad_norm": 0.06741216033697128,
+      "learning_rate": 0.00014452554744525547,
+      "loss": 0.7853,
+      "step": 67
+    },
+    {
+      "epoch": 0.14936847885777046,
+      "grad_norm": 0.05799397826194763,
+      "learning_rate": 0.00014671532846715328,
+      "loss": 0.6912,
+      "step": 68
+    },
+    {
+      "epoch": 0.1515650741350906,
+      "grad_norm": 0.08100557327270508,
+      "learning_rate": 0.00014890510948905108,
+      "loss": 1.0204,
+      "step": 69
+    },
+    {
+      "epoch": 0.15376166941241076,
+      "grad_norm": 0.08090441673994064,
+      "learning_rate": 0.00015109489051094892,
+      "loss": 0.7844,
+      "step": 70
+    },
+    {
+      "epoch": 0.15595826468973092,
+      "grad_norm": 0.062162335962057114,
+      "learning_rate": 0.00015328467153284672,
+      "loss": 0.8829,
+      "step": 71
+    },
+    {
+      "epoch": 0.15815485996705106,
+      "grad_norm": 0.06271839141845703,
+      "learning_rate": 0.00015547445255474453,
+      "loss": 0.7453,
+      "step": 72
+    },
+    {
+      "epoch": 0.16035145524437122,
+      "grad_norm": 0.09264097362756729,
+      "learning_rate": 0.00015766423357664234,
+      "loss": 0.9994,
+      "step": 73
+    },
+    {
+      "epoch": 0.16254805052169138,
+      "grad_norm": 0.062283921986818314,
+      "learning_rate": 0.00015985401459854014,
+      "loss": 0.8463,
+      "step": 74
+    },
+    {
+      "epoch": 0.16474464579901152,
+      "grad_norm": 0.06915900856256485,
+      "learning_rate": 0.00016204379562043795,
+      "loss": 0.7171,
+      "step": 75
+    },
+    {
+      "epoch": 0.16694124107633168,
+      "grad_norm": 0.06716243922710419,
+      "learning_rate": 0.00016423357664233575,
+      "loss": 0.8937,
+      "step": 76
+    },
+    {
+      "epoch": 0.16913783635365184,
+      "grad_norm": 0.06265877187252045,
+      "learning_rate": 0.00016642335766423356,
+      "loss": 0.8112,
+      "step": 77
+    },
+    {
+      "epoch": 0.171334431630972,
+      "grad_norm": 0.06956397742033005,
+      "learning_rate": 0.00016861313868613137,
+      "loss": 0.6043,
+      "step": 78
+    },
+    {
+      "epoch": 0.17353102690829214,
+      "grad_norm": 0.0620533749461174,
+      "learning_rate": 0.00017080291970802917,
+      "loss": 0.8352,
+      "step": 79
+    },
+    {
+      "epoch": 0.1757276221856123,
+      "grad_norm": 0.07454898208379745,
+      "learning_rate": 0.00017299270072992698,
+      "loss": 0.8129,
+      "step": 80
+    },
+    {
+      "epoch": 0.17792421746293247,
+      "grad_norm": 0.05988597124814987,
+      "learning_rate": 0.00017518248175182478,
+      "loss": 0.7515,
+      "step": 81
+    },
+    {
+      "epoch": 0.1801208127402526,
+      "grad_norm": 0.06786584854125977,
+      "learning_rate": 0.00017737226277372262,
+      "loss": 0.7505,
+      "step": 82
+    },
+    {
+      "epoch": 0.18231740801757276,
+      "grad_norm": 0.049385055899620056,
+      "learning_rate": 0.00017956204379562042,
+      "loss": 0.6942,
+      "step": 83
+    },
+    {
+      "epoch": 0.18451400329489293,
+      "grad_norm": 0.09457117319107056,
+      "learning_rate": 0.00018175182481751826,
+      "loss": 0.8961,
+      "step": 84
+    },
+    {
+      "epoch": 0.18671059857221306,
+      "grad_norm": 0.0817558765411377,
+      "learning_rate": 0.00018394160583941606,
+      "loss": 0.9277,
+      "step": 85
+    },
+    {
+      "epoch": 0.18890719384953322,
+      "grad_norm": 0.07502376288175583,
+      "learning_rate": 0.00018613138686131387,
+      "loss": 1.0283,
+      "step": 86
+    },
+    {
+      "epoch": 0.19110378912685339,
+      "grad_norm": 0.05120819807052612,
+      "learning_rate": 0.00018832116788321167,
+      "loss": 0.8361,
+      "step": 87
+    },
+    {
+      "epoch": 0.19330038440417352,
+      "grad_norm": 0.07280557602643967,
+      "learning_rate": 0.00019051094890510948,
+      "loss": 0.9357,
+      "step": 88
+    },
+    {
+      "epoch": 0.19549697968149368,
+      "grad_norm": 0.07645975053310394,
+      "learning_rate": 0.0001927007299270073,
+      "loss": 0.7198,
+      "step": 89
+    },
+    {
+      "epoch": 0.19769357495881384,
+      "grad_norm": 0.08283960074186325,
+      "learning_rate": 0.0001948905109489051,
+      "loss": 1.0339,
+      "step": 90
+    },
+    {
+      "epoch": 0.19989017023613398,
+      "grad_norm": 0.04628637805581093,
+      "learning_rate": 0.0001970802919708029,
+      "loss": 0.7246,
+      "step": 91
+    },
+    {
+      "epoch": 0.20208676551345414,
+      "grad_norm": 0.047697197645902634,
+      "learning_rate": 0.0001992700729927007,
+      "loss": 0.7441,
+      "step": 92
+    },
+    {
+      "epoch": 0.2042833607907743,
+      "grad_norm": 0.07364491373300552,
+      "learning_rate": 0.0002014598540145985,
+      "loss": 0.7784,
+      "step": 93
+    },
+    {
+      "epoch": 0.20647995606809447,
+      "grad_norm": 0.0743069127202034,
+      "learning_rate": 0.00020364963503649632,
+      "loss": 0.9773,
+      "step": 94
+    },
+    {
+      "epoch": 0.2086765513454146,
+      "grad_norm": 0.059089288115501404,
+      "learning_rate": 0.00020583941605839415,
+      "loss": 0.7419,
+      "step": 95
+    },
+    {
+      "epoch": 0.21087314662273476,
+      "grad_norm": 0.07568493485450745,
+      "learning_rate": 0.00020802919708029196,
+      "loss": 0.8705,
+      "step": 96
+    },
+    {
+      "epoch": 0.21306974190005493,
+      "grad_norm": 0.07957746833562851,
+      "learning_rate": 0.00021021897810218976,
+      "loss": 0.9486,
+      "step": 97
+    },
+    {
+      "epoch": 0.21526633717737506,
+      "grad_norm": 0.07052905112504959,
+      "learning_rate": 0.00021240875912408757,
+      "loss": 0.832,
+      "step": 98
+    },
+    {
+      "epoch": 0.21746293245469522,
+      "grad_norm": 0.06128270551562309,
+      "learning_rate": 0.00021459854014598537,
+      "loss": 0.7649,
+      "step": 99
+    },
+    {
+      "epoch": 0.21965952773201539,
+      "grad_norm": 0.10580066591501236,
+      "learning_rate": 0.0002167883211678832,
+      "loss": 0.9327,
+      "step": 100
+    },
+    {
+      "epoch": 0.22185612300933552,
+      "grad_norm": 0.08737997710704803,
+      "learning_rate": 0.00021897810218978101,
+      "loss": 0.9517,
+      "step": 101
+    },
+    {
+      "epoch": 0.22405271828665568,
+      "grad_norm": 0.07992607355117798,
+      "learning_rate": 0.00022116788321167882,
+      "loss": 0.8463,
+      "step": 102
+    },
+    {
+      "epoch": 0.22624931356397585,
+      "grad_norm": 0.05659569054841995,
+      "learning_rate": 0.00022335766423357663,
+      "loss": 0.8145,
+      "step": 103
+    },
+    {
+      "epoch": 0.22844590884129598,
+      "grad_norm": 0.06634163856506348,
+      "learning_rate": 0.00022554744525547443,
+      "loss": 0.7103,
+      "step": 104
+    },
+    {
+      "epoch": 0.23064250411861614,
+      "grad_norm": 0.05726408213376999,
+      "learning_rate": 0.00022773722627737224,
+      "loss": 0.8268,
+      "step": 105
+    },
+    {
+      "epoch": 0.2328390993959363,
+      "grad_norm": 0.061387598514556885,
+      "learning_rate": 0.00022992700729927004,
+      "loss": 0.8989,
+      "step": 106
+    },
+    {
+      "epoch": 0.23503569467325644,
+      "grad_norm": 0.10654595494270325,
+      "learning_rate": 0.00023211678832116788,
+      "loss": 0.8958,
+      "step": 107
+    },
+    {
+      "epoch": 0.2372322899505766,
+      "grad_norm": 0.05953432247042656,
+      "learning_rate": 0.00023430656934306568,
+      "loss": 0.6935,
+      "step": 108
+    },
+    {
+      "epoch": 0.23942888522789676,
+      "grad_norm": 0.07555431127548218,
+      "learning_rate": 0.0002364963503649635,
+      "loss": 0.7281,
+      "step": 109
+    },
+    {
+      "epoch": 0.24162548050521693,
+      "grad_norm": 0.06416428089141846,
+      "learning_rate": 0.0002386861313868613,
+      "loss": 0.7588,
+      "step": 110
+    },
+    {
+      "epoch": 0.24382207578253706,
+      "grad_norm": 0.08092894405126572,
+      "learning_rate": 0.0002408759124087591,
+      "loss": 0.8226,
+      "step": 111
+    },
+    {
+      "epoch": 0.24601867105985722,
+      "grad_norm": 0.08186564594507217,
+      "learning_rate": 0.0002430656934306569,
+      "loss": 0.8446,
+      "step": 112
+    },
+    {
+      "epoch": 0.2482152663371774,
+      "grad_norm": 0.07057618349790573,
+      "learning_rate": 0.0002452554744525547,
+      "loss": 0.7437,
+      "step": 113
+    },
+    {
+      "epoch": 0.2504118616144975,
+      "grad_norm": 0.0712006688117981,
+      "learning_rate": 0.0002474452554744525,
+      "loss": 0.72,
+      "step": 114
+    },
+    {
+      "epoch": 0.25260845689181766,
+      "grad_norm": 0.0903545469045639,
+      "learning_rate": 0.0002496350364963503,
+      "loss": 0.8596,
+      "step": 115
+    },
+    {
+      "epoch": 0.25480505216913785,
+      "grad_norm": 0.06417638808488846,
+      "learning_rate": 0.00025182481751824813,
+      "loss": 0.9108,
+      "step": 116
+    },
+    {
+      "epoch": 0.257001647446458,
+      "grad_norm": 0.07485811412334442,
+      "learning_rate": 0.00025401459854014594,
+      "loss": 0.8721,
+      "step": 117
+    },
+    {
+      "epoch": 0.2591982427237781,
+      "grad_norm": 0.051193512976169586,
+      "learning_rate": 0.00025620437956204374,
+      "loss": 0.7606,
+      "step": 118
+    },
+    {
+      "epoch": 0.2613948380010983,
+      "grad_norm": 0.06474370509386063,
+      "learning_rate": 0.00025839416058394155,
+      "loss": 0.7512,
+      "step": 119
+    },
+    {
+      "epoch": 0.26359143327841844,
+      "grad_norm": 0.05385642126202583,
+      "learning_rate": 0.0002605839416058394,
+      "loss": 0.6462,
+      "step": 120
+    },
+    {
+      "epoch": 0.26578802855573863,
+      "grad_norm": 0.05074644461274147,
+      "learning_rate": 0.0002627737226277372,
+      "loss": 0.696,
+      "step": 121
+    },
+    {
+      "epoch": 0.26798462383305877,
+      "grad_norm": 0.0700865164399147,
+      "learning_rate": 0.000264963503649635,
+      "loss": 0.741,
+      "step": 122
+    },
+    {
+      "epoch": 0.2701812191103789,
+      "grad_norm": 0.08031459152698517,
+      "learning_rate": 0.00026715328467153283,
+      "loss": 0.7789,
+      "step": 123
+    },
+    {
+      "epoch": 0.2723778143876991,
+      "grad_norm": 0.06968266516923904,
+      "learning_rate": 0.00026934306569343063,
+      "loss": 0.6849,
+      "step": 124
+    },
+    {
+      "epoch": 0.2745744096650192,
+      "grad_norm": 0.07033291459083557,
+      "learning_rate": 0.00027153284671532844,
+      "loss": 0.8538,
+      "step": 125
+    },
+    {
+      "epoch": 0.27677100494233936,
+      "grad_norm": 0.0653030201792717,
+      "learning_rate": 0.00027372262773722625,
+      "loss": 0.605,
+      "step": 126
+    },
+    {
+      "epoch": 0.27896760021965955,
+      "grad_norm": 0.07047812640666962,
+      "learning_rate": 0.00027591240875912405,
+      "loss": 0.7744,
+      "step": 127
+    },
+    {
+      "epoch": 0.2811641954969797,
+      "grad_norm": 0.07086916267871857,
+      "learning_rate": 0.00027810218978102186,
+      "loss": 0.9742,
+      "step": 128
+    },
+    {
+      "epoch": 0.2833607907742998,
+      "grad_norm": 0.05834396556019783,
+      "learning_rate": 0.00028029197080291966,
+      "loss": 0.8322,
+      "step": 129
+    },
+    {
+      "epoch": 0.28555738605162,
+      "grad_norm": 0.07212558388710022,
+      "learning_rate": 0.00028248175182481747,
+      "loss": 0.719,
+      "step": 130
+    },
+    {
+      "epoch": 0.28775398132894014,
+      "grad_norm": 0.057706017047166824,
+      "learning_rate": 0.0002846715328467153,
+      "loss": 0.7827,
+      "step": 131
+    },
+    {
+      "epoch": 0.2899505766062603,
+      "grad_norm": 0.06932689994573593,
+      "learning_rate": 0.00028686131386861314,
+      "loss": 0.6932,
+      "step": 132
+    },
+    {
+      "epoch": 0.29214717188358047,
+      "grad_norm": 0.08389901369810104,
+      "learning_rate": 0.00028905109489051094,
+      "loss": 0.7219,
+      "step": 133
+    },
+    {
+      "epoch": 0.2943437671609006,
+      "grad_norm": 0.08421478420495987,
+      "learning_rate": 0.00029124087591240875,
+      "loss": 0.7718,
+      "step": 134
+    },
+    {
+      "epoch": 0.29654036243822074,
+      "grad_norm": 0.0458688959479332,
+      "learning_rate": 0.00029343065693430656,
+      "loss": 0.764,
+      "step": 135
+    },
+    {
+      "epoch": 0.29873695771554093,
+      "grad_norm": 0.09606773406267166,
+      "learning_rate": 0.00029562043795620436,
+      "loss": 0.8812,
+      "step": 136
+    },
+    {
+      "epoch": 0.30093355299286106,
+      "grad_norm": 0.06875104457139969,
+      "learning_rate": 0.00029781021897810217,
+      "loss": 0.6979,
+      "step": 137
+    },
+    {
+      "epoch": 0.3031301482701812,
+      "grad_norm": 0.05998622253537178,
+      "learning_rate": 0.0003,
+      "loss": 0.7853,
+      "step": 138
+    },
+    {
+      "epoch": 0.3053267435475014,
+      "grad_norm": 0.058912474662065506,
+      "learning_rate": 0.00029975570032573286,
+      "loss": 0.6941,
+      "step": 139
+    },
+    {
+      "epoch": 0.3075233388248215,
+      "grad_norm": 0.05363420024514198,
+      "learning_rate": 0.00029951140065146575,
+      "loss": 0.7914,
+      "step": 140
+    },
+    {
+      "epoch": 0.30971993410214166,
+      "grad_norm": 0.21214614808559418,
+      "learning_rate": 0.0002992671009771987,
+      "loss": 0.7579,
+      "step": 141
+    },
+    {
+      "epoch": 0.31191652937946185,
+      "grad_norm": 0.08337908983230591,
+      "learning_rate": 0.0002990228013029316,
+      "loss": 0.851,
+      "step": 142
+    },
+    {
+      "epoch": 0.314113124656782,
+      "grad_norm": 0.07102375477552414,
+      "learning_rate": 0.0002987785016286645,
+      "loss": 0.8352,
+      "step": 143
+    },
+    {
+      "epoch": 0.3163097199341021,
+      "grad_norm": 0.09091459214687347,
+      "learning_rate": 0.00029853420195439736,
+      "loss": 0.9523,
+      "step": 144
+    },
+    {
+      "epoch": 0.3185063152114223,
+      "grad_norm": 0.061527032405138016,
+      "learning_rate": 0.00029828990228013025,
+      "loss": 0.7559,
+      "step": 145
+    },
+    {
+      "epoch": 0.32070291048874244,
+      "grad_norm": 0.061387546360492706,
+      "learning_rate": 0.00029804560260586314,
+      "loss": 0.8038,
+      "step": 146
+    },
+    {
+      "epoch": 0.3228995057660626,
+      "grad_norm": 0.05838935077190399,
+      "learning_rate": 0.0002978013029315961,
+      "loss": 0.6139,
+      "step": 147
+    },
+    {
+      "epoch": 0.32509610104338277,
+      "grad_norm": 0.049171656370162964,
+      "learning_rate": 0.000297557003257329,
+      "loss": 0.6492,
+      "step": 148
+    },
+    {
+      "epoch": 0.3272926963207029,
+      "grad_norm": 0.07386723905801773,
+      "learning_rate": 0.00029731270358306186,
+      "loss": 0.7368,
+      "step": 149
+    },
+    {
+      "epoch": 0.32948929159802304,
+      "grad_norm": 0.06298866122961044,
+      "learning_rate": 0.00029706840390879475,
+      "loss": 0.9454,
+      "step": 150
+    },
+    {
+      "epoch": 0.3316858868753432,
+      "grad_norm": 0.05815809220075607,
+      "learning_rate": 0.00029682410423452764,
+      "loss": 0.8447,
+      "step": 151
+    },
+    {
+      "epoch": 0.33388248215266336,
+      "grad_norm": 0.07850918173789978,
+      "learning_rate": 0.0002965798045602606,
+      "loss": 0.9031,
+      "step": 152
+    },
+    {
+      "epoch": 0.33607907742998355,
+      "grad_norm": 0.06274339556694031,
+      "learning_rate": 0.0002963355048859935,
+      "loss": 0.6847,
+      "step": 153
+    },
+    {
+      "epoch": 0.3382756727073037,
+      "grad_norm": 0.07347492128610611,
+      "learning_rate": 0.00029609120521172636,
+      "loss": 0.7598,
+      "step": 154
+    },
+    {
+      "epoch": 0.3404722679846238,
+      "grad_norm": 0.06832234561443329,
+      "learning_rate": 0.00029584690553745925,
+      "loss": 0.6799,
+      "step": 155
+    },
+    {
+      "epoch": 0.342668863261944,
+      "grad_norm": 0.053192079067230225,
+      "learning_rate": 0.00029560260586319214,
+      "loss": 0.6877,
+      "step": 156
+    },
+    {
+      "epoch": 0.34486545853926415,
+      "grad_norm": 0.11681066453456879,
+      "learning_rate": 0.00029535830618892503,
+      "loss": 0.7697,
+      "step": 157
+    },
+    {
+      "epoch": 0.3470620538165843,
+      "grad_norm": 0.08324015140533447,
+      "learning_rate": 0.000295114006514658,
+      "loss": 0.751,
+      "step": 158
+    },
+    {
+      "epoch": 0.34925864909390447,
+      "grad_norm": 0.062357187271118164,
+      "learning_rate": 0.00029486970684039086,
+      "loss": 0.8613,
+      "step": 159
+    },
+    {
+      "epoch": 0.3514552443712246,
+      "grad_norm": 0.07192494720220566,
+      "learning_rate": 0.00029462540716612375,
+      "loss": 0.7714,
+      "step": 160
+    },
+    {
+      "epoch": 0.35365183964854474,
+      "grad_norm": 0.08102371543645859,
+      "learning_rate": 0.00029438110749185664,
+      "loss": 0.7397,
+      "step": 161
+    },
+    {
+      "epoch": 0.35584843492586493,
+      "grad_norm": 0.05772097036242485,
+      "learning_rate": 0.00029413680781758953,
+      "loss": 0.6198,
+      "step": 162
+    },
+    {
+      "epoch": 0.35804503020318507,
+      "grad_norm": 0.05155602842569351,
+      "learning_rate": 0.0002938925081433224,
+      "loss": 0.5073,
+      "step": 163
+    },
+    {
+      "epoch": 0.3602416254805052,
+      "grad_norm": 0.08115146309137344,
+      "learning_rate": 0.00029364820846905536,
+      "loss": 0.6907,
+      "step": 164
+    },
+    {
+      "epoch": 0.3624382207578254,
+      "grad_norm": 0.07688108831644058,
+      "learning_rate": 0.00029340390879478825,
+      "loss": 0.8097,
+      "step": 165
+    },
+    {
+      "epoch": 0.3646348160351455,
+      "grad_norm": 0.09109821170568466,
+      "learning_rate": 0.00029315960912052114,
+      "loss": 0.7904,
+      "step": 166
+    },
+    {
+      "epoch": 0.36683141131246566,
+      "grad_norm": 0.06498458981513977,
+      "learning_rate": 0.00029291530944625403,
+      "loss": 0.6746,
+      "step": 167
+    },
+    {
+      "epoch": 0.36902800658978585,
+      "grad_norm": 0.08671940118074417,
+      "learning_rate": 0.0002926710097719869,
+      "loss": 0.8064,
+      "step": 168
+    },
+    {
+      "epoch": 0.371224601867106,
+      "grad_norm": 0.06002287194132805,
+      "learning_rate": 0.00029242671009771986,
+      "loss": 0.7684,
+      "step": 169
+    },
+    {
+      "epoch": 0.3734211971444261,
+      "grad_norm": 0.07022416591644287,
+      "learning_rate": 0.00029218241042345275,
+      "loss": 0.834,
+      "step": 170
+    },
+    {
+      "epoch": 0.3756177924217463,
+      "grad_norm": 0.07985968142747879,
+      "learning_rate": 0.00029193811074918564,
+      "loss": 0.7098,
+      "step": 171
+    },
+    {
+      "epoch": 0.37781438769906645,
+      "grad_norm": 0.06475462019443512,
+      "learning_rate": 0.00029169381107491853,
+      "loss": 0.6133,
+      "step": 172
+    },
+    {
+      "epoch": 0.3800109829763866,
+      "grad_norm": 0.060454897582530975,
+      "learning_rate": 0.0002914495114006514,
+      "loss": 0.7347,
+      "step": 173
+    },
+    {
+      "epoch": 0.38220757825370677,
+      "grad_norm": 0.07031078636646271,
+      "learning_rate": 0.0002912052117263843,
+      "loss": 0.95,
+      "step": 174
+    },
+    {
+      "epoch": 0.3844041735310269,
+      "grad_norm": 0.05603813752532005,
+      "learning_rate": 0.00029096091205211725,
+      "loss": 0.5946,
+      "step": 175
+    },
+    {
+      "epoch": 0.38660076880834704,
+      "grad_norm": 0.046303559094667435,
+      "learning_rate": 0.00029071661237785014,
+      "loss": 0.7976,
+      "step": 176
+    },
+    {
+      "epoch": 0.38879736408566723,
+      "grad_norm": 0.06395114958286285,
+      "learning_rate": 0.00029047231270358303,
+      "loss": 0.4958,
+      "step": 177
+    },
+    {
+      "epoch": 0.39099395936298736,
+      "grad_norm": 0.08462050557136536,
+      "learning_rate": 0.0002902280130293159,
+      "loss": 0.7312,
+      "step": 178
+    },
+    {
+      "epoch": 0.3931905546403075,
+      "grad_norm": 0.0730491355061531,
+      "learning_rate": 0.0002899837133550488,
+      "loss": 0.7501,
+      "step": 179
+    },
+    {
+      "epoch": 0.3953871499176277,
+      "grad_norm": 0.0624474361538887,
+      "learning_rate": 0.00028973941368078175,
+      "loss": 0.6107,
+      "step": 180
+    },
+    {
+      "epoch": 0.3975837451949478,
+      "grad_norm": 0.08671862632036209,
+      "learning_rate": 0.00028949511400651464,
+      "loss": 0.5783,
+      "step": 181
+    },
+    {
+      "epoch": 0.39978034047226796,
+      "grad_norm": 0.0546480156481266,
+      "learning_rate": 0.00028925081433224753,
+      "loss": 0.5778,
+      "step": 182
+    },
+    {
+      "epoch": 0.40197693574958815,
+      "grad_norm": 0.0740947350859642,
+      "learning_rate": 0.0002890065146579804,
+      "loss": 0.7729,
+      "step": 183
+    },
+    {
+      "epoch": 0.4041735310269083,
+      "grad_norm": 0.06049893796443939,
+      "learning_rate": 0.0002887622149837133,
+      "loss": 0.9998,
+      "step": 184
+    },
+    {
+      "epoch": 0.4063701263042284,
+      "grad_norm": 0.056841589510440826,
+      "learning_rate": 0.0002885179153094462,
+      "loss": 0.7461,
+      "step": 185
+    },
+    {
+      "epoch": 0.4085667215815486,
+      "grad_norm": 0.06030123680830002,
+      "learning_rate": 0.00028827361563517914,
+      "loss": 0.7373,
+      "step": 186
+    },
+    {
+      "epoch": 0.41076331685886874,
+      "grad_norm": 0.08607760071754456,
+      "learning_rate": 0.00028802931596091203,
+      "loss": 0.5802,
+      "step": 187
+    },
+    {
+      "epoch": 0.41295991213618893,
+      "grad_norm": 0.0810663178563118,
+      "learning_rate": 0.0002877850162866449,
+      "loss": 0.8243,
+      "step": 188
+    },
+    {
+      "epoch": 0.41515650741350907,
+      "grad_norm": 0.09073084592819214,
+      "learning_rate": 0.0002875407166123778,
+      "loss": 0.8053,
+      "step": 189
+    },
+    {
+      "epoch": 0.4173531026908292,
+      "grad_norm": 0.058536745607852936,
+      "learning_rate": 0.0002872964169381107,
+      "loss": 0.6984,
+      "step": 190
+    },
+    {
+      "epoch": 0.4195496979681494,
+      "grad_norm": 0.09180819988250732,
+      "learning_rate": 0.00028705211726384364,
+      "loss": 0.6705,
+      "step": 191
+    },
+    {
+      "epoch": 0.42174629324546953,
+      "grad_norm": 0.059555936604738235,
+      "learning_rate": 0.00028680781758957653,
+      "loss": 0.803,
+      "step": 192
+    },
+    {
+      "epoch": 0.42394288852278966,
+      "grad_norm": 0.0732739195227623,
+      "learning_rate": 0.0002865635179153094,
+      "loss": 0.7823,
+      "step": 193
+    },
+    {
+      "epoch": 0.42613948380010985,
+      "grad_norm": 0.059682246297597885,
+      "learning_rate": 0.0002863192182410423,
+      "loss": 0.7723,
+      "step": 194
+    },
+    {
+      "epoch": 0.42833607907743,
+      "grad_norm": 0.07256857305765152,
+      "learning_rate": 0.0002860749185667752,
+      "loss": 0.8889,
+      "step": 195
+    },
+    {
+      "epoch": 0.4305326743547501,
+      "grad_norm": 0.06505439430475235,
+      "learning_rate": 0.0002858306188925081,
+      "loss": 1.0675,
+      "step": 196
+    },
+    {
+      "epoch": 0.4327292696320703,
+      "grad_norm": 0.05685146152973175,
+      "learning_rate": 0.00028558631921824103,
+      "loss": 0.6376,
+      "step": 197
+    },
+    {
+      "epoch": 0.43492586490939045,
+      "grad_norm": 0.08345243334770203,
+      "learning_rate": 0.0002853420195439739,
+      "loss": 0.6554,
+      "step": 198
+    },
+    {
+      "epoch": 0.4371224601867106,
+      "grad_norm": 0.0606047622859478,
+      "learning_rate": 0.0002850977198697068,
+      "loss": 0.6857,
+      "step": 199
+    },
+    {
+      "epoch": 0.43931905546403077,
+      "grad_norm": 0.05236055701971054,
+      "learning_rate": 0.0002848534201954397,
+      "loss": 0.7361,
+      "step": 200
+    },
+    {
+      "epoch": 0.4415156507413509,
+      "grad_norm": 0.06410698592662811,
+      "learning_rate": 0.0002846091205211726,
+      "loss": 0.6465,
+      "step": 201
+    },
+    {
+      "epoch": 0.44371224601867104,
+      "grad_norm": 0.04402509704232216,
+      "learning_rate": 0.0002843648208469055,
+      "loss": 0.6407,
+      "step": 202
+    },
+    {
+      "epoch": 0.44590884129599123,
+      "grad_norm": 0.0492163822054863,
+      "learning_rate": 0.0002841205211726384,
+      "loss": 0.697,
+      "step": 203
+    },
+    {
+      "epoch": 0.44810543657331137,
+      "grad_norm": 0.09770838916301727,
+      "learning_rate": 0.0002838762214983713,
+      "loss": 0.7488,
+      "step": 204
+    },
+    {
+      "epoch": 0.4503020318506315,
+      "grad_norm": 0.08673590421676636,
+      "learning_rate": 0.0002836319218241042,
+      "loss": 0.8844,
+      "step": 205
+    },
+    {
+      "epoch": 0.4524986271279517,
+      "grad_norm": 0.0633588433265686,
+      "learning_rate": 0.0002833876221498371,
+      "loss": 0.8417,
+      "step": 206
+    },
+    {
+      "epoch": 0.4546952224052718,
+      "grad_norm": 0.04995807260274887,
+      "learning_rate": 0.00028314332247557,
+      "loss": 0.7281,
+      "step": 207
+    },
+    {
+      "epoch": 0.45689181768259196,
+      "grad_norm": 0.05731228366494179,
+      "learning_rate": 0.0002828990228013029,
+      "loss": 0.6367,
+      "step": 208
+    },
+    {
+      "epoch": 0.45908841295991215,
+      "grad_norm": 0.11119459569454193,
+      "learning_rate": 0.0002826547231270358,
+      "loss": 0.7086,
+      "step": 209
+    },
+    {
+      "epoch": 0.4612850082372323,
+      "grad_norm": 0.06479672342538834,
+      "learning_rate": 0.0002824104234527687,
+      "loss": 0.8242,
+      "step": 210
+    },
+    {
+      "epoch": 0.4634816035145524,
+      "grad_norm": 0.05451243370771408,
+      "learning_rate": 0.0002821661237785016,
+      "loss": 0.7898,
+      "step": 211
+    },
+    {
+      "epoch": 0.4656781987918726,
+      "grad_norm": 0.05641823261976242,
+      "learning_rate": 0.0002819218241042345,
+      "loss": 0.7529,
+      "step": 212
+    },
+    {
+      "epoch": 0.46787479406919275,
+      "grad_norm": 0.05920799449086189,
+      "learning_rate": 0.00028167752442996737,
+      "loss": 0.5424,
+      "step": 213
+    },
+    {
+      "epoch": 0.4700713893465129,
+      "grad_norm": 0.03466494008898735,
+      "learning_rate": 0.0002814332247557003,
+      "loss": 0.6346,
+      "step": 214
+    },
+    {
+      "epoch": 0.47226798462383307,
+      "grad_norm": 0.04757579788565636,
+      "learning_rate": 0.0002811889250814332,
+      "loss": 0.7383,
+      "step": 215
+    },
+    {
+      "epoch": 0.4744645799011532,
+      "grad_norm": 0.06769683212041855,
+      "learning_rate": 0.0002809446254071661,
+      "loss": 0.7402,
+      "step": 216
+    },
+    {
+      "epoch": 0.47666117517847334,
+      "grad_norm": 0.047434259206056595,
+      "learning_rate": 0.000280700325732899,
+      "loss": 0.5883,
+      "step": 217
+    },
+    {
+      "epoch": 0.47885777045579353,
+      "grad_norm": 0.05035366117954254,
+      "learning_rate": 0.00028045602605863187,
+      "loss": 0.777,
+      "step": 218
+    },
+    {
+      "epoch": 0.48105436573311366,
+      "grad_norm": 0.06778255850076675,
+      "learning_rate": 0.0002802117263843648,
+      "loss": 0.6835,
+      "step": 219
+    },
+    {
+      "epoch": 0.48325096101043385,
+      "grad_norm": 0.04769022762775421,
+      "learning_rate": 0.0002799674267100977,
+      "loss": 0.7615,
+      "step": 220
+    },
+    {
+      "epoch": 0.485447556287754,
+      "grad_norm": 0.06975305825471878,
+      "learning_rate": 0.0002797231270358306,
+      "loss": 0.6741,
+      "step": 221
+    },
+    {
+      "epoch": 0.4876441515650741,
+      "grad_norm": 0.0725470632314682,
+      "learning_rate": 0.0002794788273615635,
+      "loss": 0.626,
+      "step": 222
+    },
+    {
+      "epoch": 0.4898407468423943,
+      "grad_norm": 0.055425189435482025,
+      "learning_rate": 0.00027923452768729637,
+      "loss": 0.7386,
+      "step": 223
+    },
+    {
+      "epoch": 0.49203734211971445,
+      "grad_norm": 0.046605244278907776,
+      "learning_rate": 0.00027899022801302926,
+      "loss": 0.7661,
+      "step": 224
+    },
+    {
+      "epoch": 0.4942339373970346,
+      "grad_norm": 0.05119828134775162,
+      "learning_rate": 0.0002787459283387622,
+      "loss": 0.5156,
+      "step": 225
+    },
+    {
+      "epoch": 0.4964305326743548,
+      "grad_norm": 0.051588162779808044,
+      "learning_rate": 0.0002785016286644951,
+      "loss": 0.9542,
+      "step": 226
+    },
+    {
+      "epoch": 0.4986271279516749,
+      "grad_norm": 0.1783459633588791,
+      "learning_rate": 0.000278257328990228,
+      "loss": 0.7026,
+      "step": 227
+    },
+    {
+      "epoch": 0.500823723228995,
+      "grad_norm": 0.06174377724528313,
+      "learning_rate": 0.00027801302931596087,
+      "loss": 0.6692,
+      "step": 228
+    },
+    {
+      "epoch": 0.5030203185063152,
+      "grad_norm": 0.051825955510139465,
+      "learning_rate": 0.00027776872964169376,
+      "loss": 0.7019,
+      "step": 229
+    },
+    {
+      "epoch": 0.5052169137836353,
+      "grad_norm": 0.05067208409309387,
+      "learning_rate": 0.0002775244299674267,
+      "loss": 0.7867,
+      "step": 230
+    },
+    {
+      "epoch": 0.5074135090609555,
+      "grad_norm": 0.06491252779960632,
+      "learning_rate": 0.0002772801302931596,
+      "loss": 0.7593,
+      "step": 231
+    },
+    {
+      "epoch": 0.5096101043382757,
+      "grad_norm": 0.052263930439949036,
+      "learning_rate": 0.0002770358306188925,
+      "loss": 0.6484,
+      "step": 232
+    },
+    {
+      "epoch": 0.5118066996155958,
+      "grad_norm": 0.059053026139736176,
+      "learning_rate": 0.00027679153094462537,
+      "loss": 0.7726,
+      "step": 233
+    },
+    {
+      "epoch": 0.514003294892916,
+      "grad_norm": 0.05262107402086258,
+      "learning_rate": 0.00027654723127035826,
+      "loss": 0.7085,
+      "step": 234
+    },
+    {
+      "epoch": 0.5161998901702362,
+      "grad_norm": 0.06159406527876854,
+      "learning_rate": 0.00027630293159609115,
+      "loss": 0.7029,
+      "step": 235
+    },
+    {
+      "epoch": 0.5183964854475562,
+      "grad_norm": 0.04124659299850464,
+      "learning_rate": 0.0002760586319218241,
+      "loss": 0.6227,
+      "step": 236
+    },
+    {
+      "epoch": 0.5205930807248764,
+      "grad_norm": 0.07333070039749146,
+      "learning_rate": 0.000275814332247557,
+      "loss": 0.9546,
+      "step": 237
+    },
+    {
+      "epoch": 0.5227896760021966,
+      "grad_norm": 0.0633094385266304,
+      "learning_rate": 0.00027557003257328987,
+      "loss": 0.7767,
+      "step": 238
+    },
+    {
+      "epoch": 0.5249862712795168,
+      "grad_norm": 0.05801127851009369,
+      "learning_rate": 0.00027532573289902276,
+      "loss": 0.6928,
+      "step": 239
+    },
+    {
+      "epoch": 0.5271828665568369,
+      "grad_norm": 0.057393934577703476,
+      "learning_rate": 0.00027508143322475565,
+      "loss": 0.7498,
+      "step": 240
+    },
+    {
+      "epoch": 0.5293794618341571,
+      "grad_norm": 0.049782704561948776,
+      "learning_rate": 0.00027483713355048854,
+      "loss": 0.7324,
+      "step": 241
+    },
+    {
+      "epoch": 0.5315760571114773,
+      "grad_norm": 0.04805954173207283,
+      "learning_rate": 0.0002745928338762215,
+      "loss": 0.6652,
+      "step": 242
+    },
+    {
+      "epoch": 0.5337726523887973,
+      "grad_norm": 0.04546598717570305,
+      "learning_rate": 0.00027434853420195437,
+      "loss": 0.6298,
+      "step": 243
+    },
+    {
+      "epoch": 0.5359692476661175,
+      "grad_norm": 0.06261162459850311,
+      "learning_rate": 0.00027410423452768726,
+      "loss": 0.8132,
+      "step": 244
+    },
+    {
+      "epoch": 0.5381658429434377,
+      "grad_norm": 0.05043161287903786,
+      "learning_rate": 0.00027385993485342015,
+      "loss": 0.6995,
+      "step": 245
+    },
+    {
+      "epoch": 0.5403624382207578,
+      "grad_norm": 0.058040980249643326,
+      "learning_rate": 0.00027361563517915304,
+      "loss": 0.7067,
+      "step": 246
+    },
+    {
+      "epoch": 0.542559033498078,
+      "grad_norm": 0.06532083451747894,
+      "learning_rate": 0.000273371335504886,
+      "loss": 0.6584,
+      "step": 247
+    },
+    {
+      "epoch": 0.5447556287753982,
+      "grad_norm": 0.07981614023447037,
+      "learning_rate": 0.00027312703583061887,
+      "loss": 0.7891,
+      "step": 248
+    },
+    {
+      "epoch": 0.5469522240527183,
+      "grad_norm": 0.04965173825621605,
+      "learning_rate": 0.00027288273615635176,
+      "loss": 0.6484,
+      "step": 249
+    },
+    {
+      "epoch": 0.5491488193300385,
+      "grad_norm": 0.060680437833070755,
+      "learning_rate": 0.00027263843648208465,
+      "loss": 0.741,
+      "step": 250
+    },
+    {
+      "epoch": 0.5513454146073586,
+      "grad_norm": 0.062146201729774475,
+      "learning_rate": 0.00027239413680781754,
+      "loss": 0.7158,
+      "step": 251
+    },
+    {
+      "epoch": 0.5535420098846787,
+      "grad_norm": 0.058960285037755966,
+      "learning_rate": 0.0002721498371335504,
+      "loss": 0.7222,
+      "step": 252
+    },
+    {
+      "epoch": 0.5557386051619989,
+      "grad_norm": 0.04842915013432503,
+      "learning_rate": 0.00027190553745928337,
+      "loss": 0.8026,
+      "step": 253
+    },
+    {
+      "epoch": 0.5579352004393191,
+      "grad_norm": 0.05300576239824295,
+      "learning_rate": 0.00027166123778501626,
+      "loss": 0.6973,
+      "step": 254
+    },
+    {
+      "epoch": 0.5601317957166392,
+      "grad_norm": 0.05646580457687378,
+      "learning_rate": 0.00027141693811074915,
+      "loss": 0.7424,
+      "step": 255
+    },
+    {
+      "epoch": 0.5623283909939594,
+      "grad_norm": 0.04713377356529236,
+      "learning_rate": 0.00027117263843648204,
+      "loss": 0.6688,
+      "step": 256
+    },
+    {
+      "epoch": 0.5645249862712796,
+      "grad_norm": 0.04357854649424553,
+      "learning_rate": 0.0002709283387622149,
+      "loss": 0.6035,
+      "step": 257
+    },
+    {
+      "epoch": 0.5667215815485996,
+      "grad_norm": 0.05619659647345543,
+      "learning_rate": 0.00027068403908794787,
+      "loss": 0.834,
+      "step": 258
+    },
+    {
+      "epoch": 0.5689181768259198,
+      "grad_norm": 0.05279466509819031,
+      "learning_rate": 0.00027043973941368076,
+      "loss": 0.6179,
+      "step": 259
+    },
+    {
+      "epoch": 0.57111477210324,
+      "grad_norm": 0.05597279593348503,
+      "learning_rate": 0.00027019543973941365,
+      "loss": 0.6988,
+      "step": 260
+    },
+    {
+      "epoch": 0.5733113673805601,
+      "grad_norm": 0.050016503781080246,
+      "learning_rate": 0.00026995114006514654,
+      "loss": 0.7449,
+      "step": 261
+    },
+    {
+      "epoch": 0.5755079626578803,
+      "grad_norm": 0.04297545179724693,
+      "learning_rate": 0.00026970684039087943,
+      "loss": 0.7036,
+      "step": 262
+    },
+    {
+      "epoch": 0.5777045579352005,
+      "grad_norm": 0.045138273388147354,
+      "learning_rate": 0.0002694625407166123,
+      "loss": 0.6087,
+      "step": 263
+    },
+    {
+      "epoch": 0.5799011532125206,
+      "grad_norm": 0.06470554322004318,
+      "learning_rate": 0.00026921824104234526,
+      "loss": 0.7049,
+      "step": 264
+    },
+    {
+      "epoch": 0.5820977484898407,
+      "grad_norm": 0.05990125611424446,
+      "learning_rate": 0.00026897394136807815,
+      "loss": 0.9789,
+      "step": 265
+    },
+    {
+      "epoch": 0.5842943437671609,
+      "grad_norm": 0.04458112642168999,
+      "learning_rate": 0.00026872964169381104,
+      "loss": 0.7714,
+      "step": 266
+    },
+    {
+      "epoch": 0.586490939044481,
+      "grad_norm": 0.04603700712323189,
+      "learning_rate": 0.00026848534201954393,
+      "loss": 0.7234,
+      "step": 267
+    },
+    {
+      "epoch": 0.5886875343218012,
+      "grad_norm": 0.05222811922430992,
+      "learning_rate": 0.0002682410423452768,
+      "loss": 0.7531,
+      "step": 268
+    },
+    {
+      "epoch": 0.5908841295991214,
+      "grad_norm": 0.0743446946144104,
+      "learning_rate": 0.00026799674267100976,
+      "loss": 0.8359,
+      "step": 269
+    },
+    {
+      "epoch": 0.5930807248764415,
+      "grad_norm": 0.04982232674956322,
+      "learning_rate": 0.00026775244299674265,
+      "loss": 0.7521,
+      "step": 270
+    },
+    {
+      "epoch": 0.5952773201537617,
+      "grad_norm": 0.05457405373454094,
+      "learning_rate": 0.00026750814332247554,
+      "loss": 0.6456,
+      "step": 271
+    },
+    {
+      "epoch": 0.5974739154310819,
+      "grad_norm": 0.05579550936818123,
+      "learning_rate": 0.00026726384364820843,
+      "loss": 0.7309,
+      "step": 272
+    },
+    {
+      "epoch": 0.5996705107084019,
+      "grad_norm": 0.06359129399061203,
+      "learning_rate": 0.0002670195439739413,
+      "loss": 0.6133,
+      "step": 273
+    },
+    {
+      "epoch": 0.6018671059857221,
+      "grad_norm": 0.07442571967840195,
+      "learning_rate": 0.0002667752442996742,
+      "loss": 0.6771,
+      "step": 274
+    },
+    {
+      "epoch": 0.6040637012630423,
+      "grad_norm": 0.06577091664075851,
+      "learning_rate": 0.00026653094462540715,
+      "loss": 0.7365,
+      "step": 275
+    },
+    {
+      "epoch": 0.6062602965403624,
+      "grad_norm": 0.051220279186964035,
+      "learning_rate": 0.00026628664495114004,
+      "loss": 0.5611,
+      "step": 276
+    },
+    {
+      "epoch": 0.6084568918176826,
+      "grad_norm": 0.045341070741415024,
+      "learning_rate": 0.00026604234527687293,
+      "loss": 0.6786,
+      "step": 277
+    },
+    {
+      "epoch": 0.6106534870950028,
+      "grad_norm": 0.0620180107653141,
+      "learning_rate": 0.0002657980456026058,
+      "loss": 0.8132,
+      "step": 278
+    },
+    {
+      "epoch": 0.6128500823723229,
+      "grad_norm": 0.06206509843468666,
+      "learning_rate": 0.0002655537459283387,
+      "loss": 0.6962,
+      "step": 279
+    },
+    {
+      "epoch": 0.615046677649643,
+      "grad_norm": 0.05237254872918129,
+      "learning_rate": 0.00026530944625407165,
+      "loss": 0.6417,
+      "step": 280
+    },
+    {
+      "epoch": 0.6172432729269632,
+      "grad_norm": 0.04775477945804596,
+      "learning_rate": 0.00026506514657980454,
+      "loss": 0.5697,
+      "step": 281
+    },
+    {
+      "epoch": 0.6194398682042833,
+      "grad_norm": 0.05023486912250519,
+      "learning_rate": 0.00026482084690553743,
+      "loss": 0.6309,
+      "step": 282
+    },
+    {
+      "epoch": 0.6216364634816035,
+      "grad_norm": 0.06396158784627914,
+      "learning_rate": 0.0002645765472312703,
+      "loss": 0.7762,
+      "step": 283
+    },
+    {
+      "epoch": 0.6238330587589237,
+      "grad_norm": 0.04276019707322121,
+      "learning_rate": 0.0002643322475570032,
+      "loss": 0.6183,
+      "step": 284
+    },
+    {
+      "epoch": 0.6260296540362438,
+      "grad_norm": 0.045517537742853165,
+      "learning_rate": 0.0002640879478827361,
+      "loss": 0.6534,
+      "step": 285
+    },
+    {
+      "epoch": 0.628226249313564,
+      "grad_norm": 0.04261459410190582,
+      "learning_rate": 0.00026384364820846904,
+      "loss": 0.6309,
+      "step": 286
+    },
+    {
+      "epoch": 0.6304228445908842,
+      "grad_norm": 0.04310956224799156,
+      "learning_rate": 0.00026359934853420193,
+      "loss": 0.7227,
+      "step": 287
+    },
+    {
+      "epoch": 0.6326194398682042,
+      "grad_norm": 0.04422944039106369,
+      "learning_rate": 0.0002633550488599348,
+      "loss": 0.6375,
+      "step": 288
+    },
+    {
+      "epoch": 0.6348160351455244,
+      "grad_norm": 0.04686834290623665,
+      "learning_rate": 0.0002631107491856677,
+      "loss": 0.5817,
+      "step": 289
+    },
+    {
+      "epoch": 0.6370126304228446,
+      "grad_norm": 0.05435696616768837,
+      "learning_rate": 0.0002628664495114006,
+      "loss": 0.6136,
+      "step": 290
+    },
+    {
+      "epoch": 0.6392092257001647,
+      "grad_norm": 0.07915763556957245,
+      "learning_rate": 0.0002626221498371335,
+      "loss": 0.7798,
+      "step": 291
+    },
+    {
+      "epoch": 0.6414058209774849,
+      "grad_norm": 0.06726890057325363,
+      "learning_rate": 0.00026237785016286643,
+      "loss": 0.7894,
+      "step": 292
+    },
+    {
+      "epoch": 0.6436024162548051,
+      "grad_norm": 0.06199405714869499,
+      "learning_rate": 0.0002621335504885993,
+      "loss": 0.7938,
+      "step": 293
+    },
+    {
+      "epoch": 0.6457990115321252,
+      "grad_norm": 0.06117792800068855,
+      "learning_rate": 0.00026188925081433226,
+      "loss": 0.7667,
+      "step": 294
+    },
+    {
+      "epoch": 0.6479956068094453,
+      "grad_norm": 0.050212424248456955,
+      "learning_rate": 0.00026164495114006515,
+      "loss": 0.6959,
+      "step": 295
+    },
+    {
+      "epoch": 0.6501922020867655,
+      "grad_norm": 0.05087039992213249,
+      "learning_rate": 0.00026140065146579804,
+      "loss": 0.696,
+      "step": 296
+    },
+    {
+      "epoch": 0.6523887973640856,
+      "grad_norm": 0.04351802170276642,
+      "learning_rate": 0.00026115635179153093,
+      "loss": 0.6119,
+      "step": 297
+    },
+    {
+      "epoch": 0.6545853926414058,
+      "grad_norm": 0.05898391455411911,
+      "learning_rate": 0.0002609120521172638,
+      "loss": 0.6155,
+      "step": 298
+    },
+    {
+      "epoch": 0.656781987918726,
+      "grad_norm": 0.03949552774429321,
+      "learning_rate": 0.00026066775244299676,
+      "loss": 0.7009,
+      "step": 299
+    },
+    {
+      "epoch": 0.6589785831960461,
+      "grad_norm": 0.0561998188495636,
+      "learning_rate": 0.00026042345276872965,
+      "loss": 0.6573,
+      "step": 300
+    },
+    {
+      "epoch": 0.6611751784733663,
+      "grad_norm": 0.07027997076511383,
+      "learning_rate": 0.00026017915309446254,
+      "loss": 0.7692,
+      "step": 301
+    },
+    {
+      "epoch": 0.6633717737506865,
+      "grad_norm": 0.04690250754356384,
+      "learning_rate": 0.00025993485342019543,
+      "loss": 0.6603,
+      "step": 302
+    },
+    {
+      "epoch": 0.6655683690280065,
+      "grad_norm": 0.07027994841337204,
+      "learning_rate": 0.0002596905537459283,
+      "loss": 0.5999,
+      "step": 303
+    },
+    {
+      "epoch": 0.6677649643053267,
+      "grad_norm": 0.06607215851545334,
+      "learning_rate": 0.0002594462540716612,
+      "loss": 0.7808,
+      "step": 304
+    },
+    {
+      "epoch": 0.6699615595826469,
+      "grad_norm": 0.061052385717630386,
+      "learning_rate": 0.00025920195439739415,
+      "loss": 0.6839,
+      "step": 305
+    },
+    {
+      "epoch": 0.6721581548599671,
+      "grad_norm": 0.057753391563892365,
+      "learning_rate": 0.00025895765472312704,
+      "loss": 0.7209,
+      "step": 306
+    },
+    {
+      "epoch": 0.6743547501372872,
+      "grad_norm": 0.060185305774211884,
+      "learning_rate": 0.00025871335504885993,
+      "loss": 0.7376,
+      "step": 307
+    },
+    {
+      "epoch": 0.6765513454146074,
+      "grad_norm": 0.05048409849405289,
+      "learning_rate": 0.0002584690553745928,
+      "loss": 0.7646,
+      "step": 308
+    },
+    {
+      "epoch": 0.6787479406919276,
+      "grad_norm": 0.05279074236750603,
+      "learning_rate": 0.0002582247557003257,
+      "loss": 0.6964,
+      "step": 309
+    },
+    {
+      "epoch": 0.6809445359692476,
+      "grad_norm": 0.04628496244549751,
+      "learning_rate": 0.00025798045602605865,
+      "loss": 0.7263,
+      "step": 310
+    },
+    {
+      "epoch": 0.6831411312465678,
+      "grad_norm": 0.06292164325714111,
+      "learning_rate": 0.00025773615635179154,
+      "loss": 0.5775,
+      "step": 311
+    },
+    {
+      "epoch": 0.685337726523888,
+      "grad_norm": 0.05216914042830467,
+      "learning_rate": 0.00025749185667752443,
+      "loss": 0.7599,
+      "step": 312
+    },
+    {
+      "epoch": 0.6875343218012081,
+      "grad_norm": 0.0711362361907959,
+      "learning_rate": 0.0002572475570032573,
+      "loss": 0.6705,
+      "step": 313
+    },
+    {
+      "epoch": 0.6897309170785283,
+      "grad_norm": 0.03843238577246666,
+      "learning_rate": 0.0002570032573289902,
+      "loss": 0.4722,
+      "step": 314
+    },
+    {
+      "epoch": 0.6919275123558485,
+      "grad_norm": 0.05418127775192261,
+      "learning_rate": 0.0002567589576547231,
+      "loss": 0.5147,
+      "step": 315
+    },
+    {
+      "epoch": 0.6941241076331686,
+      "grad_norm": 0.061103999614715576,
+      "learning_rate": 0.00025651465798045604,
+      "loss": 0.761,
+      "step": 316
+    },
+    {
+      "epoch": 0.6963207029104888,
+      "grad_norm": 0.050811514258384705,
+      "learning_rate": 0.00025627035830618893,
+      "loss": 0.8575,
+      "step": 317
+    },
+    {
+      "epoch": 0.6985172981878089,
+      "grad_norm": 0.05052623152732849,
+      "learning_rate": 0.0002560260586319218,
+      "loss": 0.7842,
+      "step": 318
+    },
+    {
+      "epoch": 0.700713893465129,
+      "grad_norm": 0.05087927356362343,
+      "learning_rate": 0.0002557817589576547,
+      "loss": 0.6584,
+      "step": 319
+    },
+    {
+      "epoch": 0.7029104887424492,
+      "grad_norm": 0.041777510195970535,
+      "learning_rate": 0.0002555374592833876,
+      "loss": 0.6491,
+      "step": 320
+    },
+    {
+      "epoch": 0.7051070840197694,
+      "grad_norm": 0.045631300657987595,
+      "learning_rate": 0.0002552931596091205,
+      "loss": 0.7189,
+      "step": 321
+    },
+    {
+      "epoch": 0.7073036792970895,
+      "grad_norm": 0.04235495626926422,
+      "learning_rate": 0.00025504885993485343,
+      "loss": 0.7269,
+      "step": 322
+    },
+    {
+      "epoch": 0.7095002745744097,
+      "grad_norm": 0.04950536787509918,
+      "learning_rate": 0.0002548045602605863,
+      "loss": 0.5914,
+      "step": 323
+    },
+    {
+      "epoch": 0.7116968698517299,
+      "grad_norm": 0.05852117761969566,
+      "learning_rate": 0.0002545602605863192,
+      "loss": 0.7652,
+      "step": 324
+    },
+    {
+      "epoch": 0.7138934651290499,
+      "grad_norm": 0.0719272643327713,
+      "learning_rate": 0.0002543159609120521,
+      "loss": 0.7394,
+      "step": 325
+    },
+    {
+      "epoch": 0.7160900604063701,
+      "grad_norm": 0.0593254491686821,
+      "learning_rate": 0.000254071661237785,
+      "loss": 0.6966,
+      "step": 326
+    },
+    {
+      "epoch": 0.7182866556836903,
+      "grad_norm": 0.04738825559616089,
+      "learning_rate": 0.00025382736156351793,
+      "loss": 0.6779,
+      "step": 327
+    },
+    {
+      "epoch": 0.7204832509610104,
+      "grad_norm": 0.05696015805006027,
+      "learning_rate": 0.0002535830618892508,
+      "loss": 0.7888,
+      "step": 328
+    },
+    {
+      "epoch": 0.7226798462383306,
+      "grad_norm": 0.03948872163891792,
+      "learning_rate": 0.0002533387622149837,
+      "loss": 0.6655,
+      "step": 329
+    },
+    {
+      "epoch": 0.7248764415156508,
+      "grad_norm": 0.05530393868684769,
+      "learning_rate": 0.0002530944625407166,
+      "loss": 0.6375,
+      "step": 330
+    },
+    {
+      "epoch": 0.7270730367929709,
+      "grad_norm": 0.06435840576887131,
+      "learning_rate": 0.0002528501628664495,
+      "loss": 0.8043,
+      "step": 331
+    },
+    {
+      "epoch": 0.729269632070291,
+      "grad_norm": 0.04414074495434761,
+      "learning_rate": 0.0002526058631921824,
+      "loss": 0.6234,
+      "step": 332
+    },
+    {
+      "epoch": 0.7314662273476112,
+      "grad_norm": 0.04935838282108307,
+      "learning_rate": 0.0002523615635179153,
+      "loss": 0.7207,
+      "step": 333
+    },
+    {
+      "epoch": 0.7336628226249313,
+      "grad_norm": 0.04916159436106682,
+      "learning_rate": 0.0002521172638436482,
+      "loss": 0.7327,
+      "step": 334
+    },
+    {
+      "epoch": 0.7358594179022515,
+      "grad_norm": 0.0506322905421257,
+      "learning_rate": 0.0002518729641693811,
+      "loss": 0.705,
+      "step": 335
+    },
+    {
+      "epoch": 0.7380560131795717,
+      "grad_norm": 0.05323386564850807,
+      "learning_rate": 0.000251628664495114,
+      "loss": 0.6962,
+      "step": 336
+    },
+    {
+      "epoch": 0.7402526084568918,
+      "grad_norm": 0.042582739144563675,
+      "learning_rate": 0.0002513843648208469,
+      "loss": 0.8309,
+      "step": 337
+    },
+    {
+      "epoch": 0.742449203734212,
+      "grad_norm": 0.05412566661834717,
+      "learning_rate": 0.0002511400651465798,
+      "loss": 0.6626,
+      "step": 338
+    },
+    {
+      "epoch": 0.7446457990115322,
+      "grad_norm": 0.057325392961502075,
+      "learning_rate": 0.0002508957654723127,
+      "loss": 0.7543,
+      "step": 339
+    },
+    {
+      "epoch": 0.7468423942888522,
+      "grad_norm": 0.054078057408332825,
+      "learning_rate": 0.0002506514657980456,
+      "loss": 0.7228,
+      "step": 340
+    },
+    {
+      "epoch": 0.7490389895661724,
+      "grad_norm": 0.05624162778258324,
+      "learning_rate": 0.0002504071661237785,
+      "loss": 0.65,
+      "step": 341
+    },
+    {
+      "epoch": 0.7512355848434926,
+      "grad_norm": 0.04718158766627312,
+      "learning_rate": 0.0002501628664495114,
+      "loss": 0.5851,
+      "step": 342
+    },
+    {
+      "epoch": 0.7534321801208127,
+      "grad_norm": 0.044785093516111374,
+      "learning_rate": 0.00024991856677524427,
+      "loss": 0.653,
+      "step": 343
+    },
+    {
+      "epoch": 0.7556287753981329,
+      "grad_norm": 0.052535783499479294,
+      "learning_rate": 0.0002496742671009772,
+      "loss": 0.6678,
+      "step": 344
+    },
+    {
+      "epoch": 0.7578253706754531,
+      "grad_norm": 0.04284857586026192,
+      "learning_rate": 0.0002494299674267101,
+      "loss": 0.577,
+      "step": 345
+    },
+    {
+      "epoch": 0.7600219659527732,
+      "grad_norm": 0.07952070236206055,
+      "learning_rate": 0.000249185667752443,
+      "loss": 0.7346,
+      "step": 346
+    },
+    {
+      "epoch": 0.7622185612300933,
+      "grad_norm": 0.0500759556889534,
+      "learning_rate": 0.0002489413680781759,
+      "loss": 0.6877,
+      "step": 347
+    },
+    {
+      "epoch": 0.7644151565074135,
+      "grad_norm": 0.05391065031290054,
+      "learning_rate": 0.00024869706840390877,
+      "loss": 0.6815,
+      "step": 348
+    },
+    {
+      "epoch": 0.7666117517847336,
+      "grad_norm": 0.041693173348903656,
+      "learning_rate": 0.0002484527687296417,
+      "loss": 0.5043,
+      "step": 349
+    },
+    {
+      "epoch": 0.7688083470620538,
+      "grad_norm": 0.055352237075567245,
+      "learning_rate": 0.0002482084690553746,
+      "loss": 0.8104,
+      "step": 350
+    },
+    {
+      "epoch": 0.771004942339374,
+      "grad_norm": 0.059511568397283554,
+      "learning_rate": 0.0002479641693811075,
+      "loss": 0.9011,
+      "step": 351
+    },
+    {
+      "epoch": 0.7732015376166941,
+      "grad_norm": 0.06932779401540756,
+      "learning_rate": 0.0002477198697068404,
+      "loss": 0.6964,
+      "step": 352
+    },
+    {
+      "epoch": 0.7753981328940143,
+      "grad_norm": 0.045885760337114334,
+      "learning_rate": 0.00024747557003257327,
+      "loss": 0.5526,
+      "step": 353
+    },
+    {
+      "epoch": 0.7775947281713345,
+      "grad_norm": 0.06347452104091644,
+      "learning_rate": 0.00024723127035830616,
+      "loss": 0.6988,
+      "step": 354
+    },
+    {
+      "epoch": 0.7797913234486545,
+      "grad_norm": 0.060996223241090775,
+      "learning_rate": 0.0002469869706840391,
+      "loss": 0.518,
+      "step": 355
+    },
+    {
+      "epoch": 0.7819879187259747,
+      "grad_norm": 0.05093398317694664,
+      "learning_rate": 0.000246742671009772,
+      "loss": 0.5963,
+      "step": 356
+    },
+    {
+      "epoch": 0.7841845140032949,
+      "grad_norm": 0.04339175298810005,
+      "learning_rate": 0.0002464983713355049,
+      "loss": 0.7216,
+      "step": 357
+    },
+    {
+      "epoch": 0.786381109280615,
+      "grad_norm": 0.04919297620654106,
+      "learning_rate": 0.00024625407166123777,
+      "loss": 0.6406,
+      "step": 358
+    },
+    {
+      "epoch": 0.7885777045579352,
+      "grad_norm": 0.08795499056577682,
+      "learning_rate": 0.00024600977198697066,
+      "loss": 0.6478,
+      "step": 359
+    },
+    {
+      "epoch": 0.7907742998352554,
+      "grad_norm": 0.04211808741092682,
+      "learning_rate": 0.00024576547231270354,
+      "loss": 0.607,
+      "step": 360
+    },
+    {
+      "epoch": 0.7929708951125755,
+      "grad_norm": 0.05462827533483505,
+      "learning_rate": 0.0002455211726384365,
+      "loss": 0.7166,
+      "step": 361
+    },
+    {
+      "epoch": 0.7951674903898956,
+      "grad_norm": 0.0573272630572319,
+      "learning_rate": 0.0002452768729641694,
+      "loss": 0.7341,
+      "step": 362
+    },
+    {
+      "epoch": 0.7973640856672158,
+      "grad_norm": 0.04874618351459503,
+      "learning_rate": 0.00024503257328990227,
+      "loss": 0.5878,
+      "step": 363
+    },
+    {
+      "epoch": 0.7995606809445359,
+      "grad_norm": 0.056756071746349335,
+      "learning_rate": 0.00024478827361563516,
+      "loss": 0.6283,
+      "step": 364
+    },
+    {
+      "epoch": 0.8017572762218561,
+      "grad_norm": 0.050058893859386444,
+      "learning_rate": 0.00024454397394136804,
+      "loss": 0.5111,
+      "step": 365
+    },
+    {
+      "epoch": 0.8039538714991763,
+      "grad_norm": 0.0445009246468544,
+      "learning_rate": 0.000244299674267101,
+      "loss": 0.7481,
+      "step": 366
+    },
+    {
+      "epoch": 0.8061504667764964,
+      "grad_norm": 0.09190984815359116,
+      "learning_rate": 0.00024405537459283385,
+      "loss": 0.744,
+      "step": 367
+    },
+    {
+      "epoch": 0.8083470620538166,
+      "grad_norm": 0.05386214330792427,
+      "learning_rate": 0.00024381107491856677,
+      "loss": 0.5524,
+      "step": 368
+    },
+    {
+      "epoch": 0.8105436573311368,
+      "grad_norm": 0.03755839914083481,
+      "learning_rate": 0.00024356677524429966,
+      "loss": 0.5951,
+      "step": 369
+    },
+    {
+      "epoch": 0.8127402526084568,
+      "grad_norm": 0.04785095900297165,
+      "learning_rate": 0.00024332247557003257,
+      "loss": 0.5013,
+      "step": 370
+    },
+    {
+      "epoch": 0.814936847885777,
+      "grad_norm": 0.035032473504543304,
+      "learning_rate": 0.00024307817589576546,
+      "loss": 0.7563,
+      "step": 371
+    },
+    {
+      "epoch": 0.8171334431630972,
+      "grad_norm": 0.05640648305416107,
+      "learning_rate": 0.00024283387622149835,
+      "loss": 0.6958,
+      "step": 372
+    },
+    {
+      "epoch": 0.8193300384404174,
+      "grad_norm": 0.03728983923792839,
+      "learning_rate": 0.00024258957654723127,
+      "loss": 0.6545,
+      "step": 373
+    },
+    {
+      "epoch": 0.8215266337177375,
+      "grad_norm": 0.07797161489725113,
+      "learning_rate": 0.00024234527687296416,
+      "loss": 0.6847,
+      "step": 374
+    },
+    {
+      "epoch": 0.8237232289950577,
+      "grad_norm": 0.055005770176649094,
+      "learning_rate": 0.00024210097719869705,
+      "loss": 0.7758,
+      "step": 375
+    },
+    {
+      "epoch": 0.8259198242723779,
+      "grad_norm": 0.05566948279738426,
+      "learning_rate": 0.00024185667752442996,
+      "loss": 0.754,
+      "step": 376
+    },
+    {
+      "epoch": 0.828116419549698,
+      "grad_norm": 0.05472628399729729,
+      "learning_rate": 0.00024161237785016285,
+      "loss": 0.688,
+      "step": 377
+    },
+    {
+      "epoch": 0.8303130148270181,
+      "grad_norm": 0.04331599548459053,
+      "learning_rate": 0.00024136807817589574,
+      "loss": 0.646,
+      "step": 378
+    },
+    {
+      "epoch": 0.8325096101043383,
+      "grad_norm": 0.04198076203465462,
+      "learning_rate": 0.00024112377850162866,
+      "loss": 0.5098,
+      "step": 379
+    },
+    {
+      "epoch": 0.8347062053816584,
+      "grad_norm": 0.05673768371343613,
+      "learning_rate": 0.00024087947882736155,
+      "loss": 0.7292,
+      "step": 380
+    },
+    {
+      "epoch": 0.8369028006589786,
+      "grad_norm": 0.06165642663836479,
+      "learning_rate": 0.00024063517915309446,
+      "loss": 0.6463,
+      "step": 381
+    },
+    {
+      "epoch": 0.8390993959362988,
+      "grad_norm": 0.05126790329813957,
+      "learning_rate": 0.00024039087947882735,
+      "loss": 0.7427,
+      "step": 382
+    },
+    {
+      "epoch": 0.8412959912136189,
+      "grad_norm": 0.04800880700349808,
+      "learning_rate": 0.00024014657980456024,
+      "loss": 0.7385,
+      "step": 383
+    },
+    {
+      "epoch": 0.8434925864909391,
+      "grad_norm": 0.06537099927663803,
+      "learning_rate": 0.00023990228013029316,
+      "loss": 0.7598,
+      "step": 384
+    },
+    {
+      "epoch": 0.8456891817682592,
+      "grad_norm": 0.0711362361907959,
+      "learning_rate": 0.00023965798045602605,
+      "loss": 0.8168,
+      "step": 385
+    },
+    {
+      "epoch": 0.8478857770455793,
+      "grad_norm": 0.06028452143073082,
+      "learning_rate": 0.00023941368078175893,
+      "loss": 0.7449,
+      "step": 386
+    },
+    {
+      "epoch": 0.8500823723228995,
+      "grad_norm": 0.04362437129020691,
+      "learning_rate": 0.00023916938110749185,
+      "loss": 0.6074,
+      "step": 387
+    },
+    {
+      "epoch": 0.8522789676002197,
+      "grad_norm": 0.05339638888835907,
+      "learning_rate": 0.00023892508143322474,
+      "loss": 0.7208,
+      "step": 388
+    },
+    {
+      "epoch": 0.8544755628775398,
+      "grad_norm": 0.049732714891433716,
+      "learning_rate": 0.00023868078175895763,
+      "loss": 0.5567,
+      "step": 389
+    },
+    {
+      "epoch": 0.85667215815486,
+      "grad_norm": 0.050275735557079315,
+      "learning_rate": 0.00023843648208469055,
+      "loss": 0.5994,
+      "step": 390
+    },
+    {
+      "epoch": 0.8588687534321802,
+      "grad_norm": 0.05760905146598816,
+      "learning_rate": 0.00023819218241042343,
+      "loss": 0.7006,
+      "step": 391
+    },
+    {
+      "epoch": 0.8610653487095002,
+      "grad_norm": 0.06789691001176834,
+      "learning_rate": 0.00023794788273615632,
+      "loss": 0.8992,
+      "step": 392
+    },
+    {
+      "epoch": 0.8632619439868204,
+      "grad_norm": 0.06151839718222618,
+      "learning_rate": 0.00023770358306188924,
+      "loss": 0.9705,
+      "step": 393
+    },
+    {
+      "epoch": 0.8654585392641406,
+      "grad_norm": 0.055463626980781555,
+      "learning_rate": 0.00023745928338762213,
+      "loss": 0.6691,
+      "step": 394
+    },
+    {
+      "epoch": 0.8676551345414607,
+      "grad_norm": 0.04577269032597542,
+      "learning_rate": 0.00023721498371335505,
+      "loss": 0.5623,
+      "step": 395
+    },
+    {
+      "epoch": 0.8698517298187809,
+      "grad_norm": 0.04549311101436615,
+      "learning_rate": 0.00023697068403908794,
+      "loss": 0.6417,
+      "step": 396
+    },
+    {
+      "epoch": 0.8720483250961011,
+      "grad_norm": 0.04944036155939102,
+      "learning_rate": 0.00023672638436482082,
+      "loss": 0.6516,
+      "step": 397
+    },
+    {
+      "epoch": 0.8742449203734212,
+      "grad_norm": 0.042285725474357605,
+      "learning_rate": 0.00023648208469055374,
+      "loss": 0.5873,
+      "step": 398
+    },
+    {
+      "epoch": 0.8764415156507414,
+      "grad_norm": 0.058092810213565826,
+      "learning_rate": 0.00023623778501628663,
+      "loss": 0.7076,
+      "step": 399
+    },
+    {
+      "epoch": 0.8786381109280615,
+      "grad_norm": 0.05264095962047577,
+      "learning_rate": 0.00023599348534201952,
+      "loss": 0.7239,
+      "step": 400
+    },
+    {
+      "epoch": 0.8808347062053816,
+      "grad_norm": 0.07236024737358093,
+      "learning_rate": 0.00023574918566775244,
+      "loss": 0.5782,
+      "step": 401
+    },
+    {
+      "epoch": 0.8830313014827018,
+      "grad_norm": 0.03647659346461296,
+      "learning_rate": 0.00023550488599348532,
+      "loss": 0.6416,
+      "step": 402
+    },
+    {
+      "epoch": 0.885227896760022,
+      "grad_norm": 0.05197187140583992,
+      "learning_rate": 0.00023526058631921821,
+      "loss": 0.8022,
+      "step": 403
+    },
+    {
+      "epoch": 0.8874244920373421,
+      "grad_norm": 0.03680606186389923,
+      "learning_rate": 0.00023501628664495113,
+      "loss": 0.796,
+      "step": 404
+    },
+    {
+      "epoch": 0.8896210873146623,
+      "grad_norm": 0.04213280603289604,
+      "learning_rate": 0.00023477198697068402,
+      "loss": 0.5616,
+      "step": 405
+    },
+    {
+      "epoch": 0.8918176825919825,
+      "grad_norm": 0.04371798038482666,
+      "learning_rate": 0.0002345276872964169,
+      "loss": 0.5991,
+      "step": 406
+    },
+    {
+      "epoch": 0.8940142778693025,
+      "grad_norm": 0.053443361073732376,
+      "learning_rate": 0.00023428338762214982,
+      "loss": 0.544,
+      "step": 407
+    },
+    {
+      "epoch": 0.8962108731466227,
+      "grad_norm": 0.04580046981573105,
+      "learning_rate": 0.00023403908794788271,
+      "loss": 0.5709,
+      "step": 408
+    },
+    {
+      "epoch": 0.8984074684239429,
+      "grad_norm": 0.0343971811234951,
+      "learning_rate": 0.00023379478827361563,
+      "loss": 0.6029,
+      "step": 409
+    },
+    {
+      "epoch": 0.900604063701263,
+      "grad_norm": 0.062083858996629715,
+      "learning_rate": 0.00023355048859934852,
+      "loss": 0.7742,
+      "step": 410
+    },
+    {
+      "epoch": 0.9028006589785832,
+      "grad_norm": 0.040063194930553436,
+      "learning_rate": 0.0002333061889250814,
+      "loss": 0.5643,
+      "step": 411
+    },
+    {
+      "epoch": 0.9049972542559034,
+      "grad_norm": 0.06296666711568832,
+      "learning_rate": 0.00023306188925081432,
+      "loss": 0.6406,
+      "step": 412
+    },
+    {
+      "epoch": 0.9071938495332235,
+      "grad_norm": 0.04029727727174759,
+      "learning_rate": 0.00023281758957654721,
+      "loss": 0.6608,
+      "step": 413
+    },
+    {
+      "epoch": 0.9093904448105437,
+      "grad_norm": 0.055288396775722504,
+      "learning_rate": 0.0002325732899022801,
+      "loss": 0.7396,
+      "step": 414
+    },
+    {
+      "epoch": 0.9115870400878638,
+      "grad_norm": 0.044744208455085754,
+      "learning_rate": 0.00023232899022801302,
+      "loss": 0.646,
+      "step": 415
+    },
+    {
+      "epoch": 0.9137836353651839,
+      "grad_norm": 0.06047806143760681,
+      "learning_rate": 0.0002320846905537459,
+      "loss": 0.8133,
+      "step": 416
+    },
+    {
+      "epoch": 0.9159802306425041,
+      "grad_norm": 0.054290421307086945,
+      "learning_rate": 0.0002318403908794788,
+      "loss": 0.6989,
+      "step": 417
+    },
+    {
+      "epoch": 0.9181768259198243,
+      "grad_norm": 0.05836010351777077,
+      "learning_rate": 0.00023159609120521171,
+      "loss": 0.5014,
+      "step": 418
+    },
+    {
+      "epoch": 0.9203734211971444,
+      "grad_norm": 0.05172126740217209,
+      "learning_rate": 0.0002313517915309446,
+      "loss": 0.5723,
+      "step": 419
+    },
+    {
+      "epoch": 0.9225700164744646,
+      "grad_norm": 0.03858590126037598,
+      "learning_rate": 0.00023110749185667752,
+      "loss": 0.6683,
+      "step": 420
+    },
+    {
+      "epoch": 0.9247666117517848,
+      "grad_norm": 0.05288545414805412,
+      "learning_rate": 0.0002308631921824104,
+      "loss": 0.6083,
+      "step": 421
+    },
+    {
+      "epoch": 0.9269632070291048,
+      "grad_norm": 0.049517158418893814,
+      "learning_rate": 0.0002306188925081433,
+      "loss": 0.6561,
+      "step": 422
+    },
+    {
+      "epoch": 0.929159802306425,
+      "grad_norm": 0.0585198700428009,
+      "learning_rate": 0.00023037459283387621,
+      "loss": 0.8331,
+      "step": 423
+    },
+    {
+      "epoch": 0.9313563975837452,
+      "grad_norm": 0.05595822259783745,
+      "learning_rate": 0.0002301302931596091,
+      "loss": 0.6627,
+      "step": 424
+    },
+    {
+      "epoch": 0.9335529928610653,
+      "grad_norm": 0.03902159631252289,
+      "learning_rate": 0.000229885993485342,
+      "loss": 0.5872,
+      "step": 425
+    },
+    {
+      "epoch": 0.9357495881383855,
+      "grad_norm": 0.04444313421845436,
+      "learning_rate": 0.0002296416938110749,
+      "loss": 0.6126,
+      "step": 426
+    },
+    {
+      "epoch": 0.9379461834157057,
+      "grad_norm": 0.04075470194220543,
+      "learning_rate": 0.0002293973941368078,
+      "loss": 0.6372,
+      "step": 427
+    },
+    {
+      "epoch": 0.9401427786930258,
+      "grad_norm": 0.04287320002913475,
+      "learning_rate": 0.0002291530944625407,
+      "loss": 0.5355,
+      "step": 428
+    },
+    {
+      "epoch": 0.942339373970346,
+      "grad_norm": 0.04004767909646034,
+      "learning_rate": 0.0002289087947882736,
+      "loss": 0.476,
+      "step": 429
+    },
+    {
+      "epoch": 0.9445359692476661,
+      "grad_norm": 0.04870041459798813,
+      "learning_rate": 0.0002286644951140065,
+      "loss": 0.7218,
+      "step": 430
+    },
+    {
+      "epoch": 0.9467325645249862,
+      "grad_norm": 0.05161641538143158,
+      "learning_rate": 0.00022842019543973938,
+      "loss": 0.5477,
+      "step": 431
+    },
+    {
+      "epoch": 0.9489291598023064,
+      "grad_norm": 0.09497353434562683,
+      "learning_rate": 0.0002281758957654723,
+      "loss": 0.8361,
+      "step": 432
+    },
+    {
+      "epoch": 0.9511257550796266,
+      "grad_norm": 0.04529595747590065,
+      "learning_rate": 0.0002279315960912052,
+      "loss": 0.6591,
+      "step": 433
+    },
+    {
+      "epoch": 0.9533223503569467,
+      "grad_norm": 0.049864016473293304,
+      "learning_rate": 0.0002276872964169381,
+      "loss": 0.561,
+      "step": 434
+    },
+    {
+      "epoch": 0.9555189456342669,
+      "grad_norm": 0.03673512116074562,
+      "learning_rate": 0.000227442996742671,
+      "loss": 0.5714,
+      "step": 435
+    },
+    {
+      "epoch": 0.9577155409115871,
+      "grad_norm": 0.04806803911924362,
+      "learning_rate": 0.00022719869706840388,
+      "loss": 0.5068,
+      "step": 436
+    },
+    {
+      "epoch": 0.9599121361889071,
+      "grad_norm": 0.046189140528440475,
+      "learning_rate": 0.0002269543973941368,
+      "loss": 0.6344,
+      "step": 437
+    },
+    {
+      "epoch": 0.9621087314662273,
+      "grad_norm": 0.042612023651599884,
+      "learning_rate": 0.0002267100977198697,
+      "loss": 0.5453,
+      "step": 438
+    },
+    {
+      "epoch": 0.9643053267435475,
+      "grad_norm": 0.04492894560098648,
+      "learning_rate": 0.00022646579804560258,
+      "loss": 0.6854,
+      "step": 439
+    },
+    {
+      "epoch": 0.9665019220208677,
+      "grad_norm": 0.05652786046266556,
+      "learning_rate": 0.0002262214983713355,
+      "loss": 0.7428,
+      "step": 440
+    },
+    {
+      "epoch": 0.9686985172981878,
+      "grad_norm": 0.054750144481658936,
+      "learning_rate": 0.00022597719869706838,
+      "loss": 0.6135,
+      "step": 441
+    },
+    {
+      "epoch": 0.970895112575508,
+      "grad_norm": 0.04848687723278999,
+      "learning_rate": 0.00022573289902280127,
+      "loss": 0.5865,
+      "step": 442
+    },
+    {
+      "epoch": 0.9730917078528282,
+      "grad_norm": 0.03940533474087715,
+      "learning_rate": 0.0002254885993485342,
+      "loss": 0.6836,
+      "step": 443
+    },
+    {
+      "epoch": 0.9752883031301482,
+      "grad_norm": 0.04875492677092552,
+      "learning_rate": 0.00022524429967426708,
+      "loss": 0.7485,
+      "step": 444
+    },
+    {
+      "epoch": 0.9774848984074684,
+      "grad_norm": 0.06128212809562683,
+      "learning_rate": 0.000225,
+      "loss": 0.73,
+      "step": 445
+    },
+    {
+      "epoch": 0.9796814936847886,
+      "grad_norm": 0.09782739728689194,
+      "learning_rate": 0.00022475570032573288,
+      "loss": 0.7955,
+      "step": 446
+    },
+    {
+      "epoch": 0.9818780889621087,
+      "grad_norm": 0.06957724690437317,
+      "learning_rate": 0.00022451140065146577,
+      "loss": 0.7571,
+      "step": 447
+    },
+    {
+      "epoch": 0.9840746842394289,
+      "grad_norm": 0.04065481945872307,
+      "learning_rate": 0.0002242671009771987,
+      "loss": 0.6567,
+      "step": 448
+    },
+    {
+      "epoch": 0.9862712795167491,
+      "grad_norm": 0.04748554527759552,
+      "learning_rate": 0.00022402280130293158,
+      "loss": 0.7358,
+      "step": 449
+    },
+    {
+      "epoch": 0.9884678747940692,
+      "grad_norm": 0.04544506594538689,
+      "learning_rate": 0.00022377850162866447,
+      "loss": 0.6632,
+      "step": 450
+    },
+    {
+      "epoch": 0.9906644700713894,
+      "grad_norm": 0.046616628766059875,
+      "learning_rate": 0.00022353420195439738,
+      "loss": 0.608,
+      "step": 451
+    },
+    {
+      "epoch": 0.9928610653487095,
+      "grad_norm": 0.04831715673208237,
+      "learning_rate": 0.00022328990228013027,
+      "loss": 0.6751,
+      "step": 452
+    },
+    {
+      "epoch": 0.9950576606260296,
+      "grad_norm": 0.04835525527596474,
+      "learning_rate": 0.00022304560260586316,
+      "loss": 0.7125,
+      "step": 453
+    },
+    {
+      "epoch": 0.9972542559033498,
+      "grad_norm": 0.043945323675870895,
+      "learning_rate": 0.00022280130293159608,
+      "loss": 0.5494,
+      "step": 454
+    },
+    {
+      "epoch": 0.99945085118067,
+      "grad_norm": 0.04118410125374794,
+      "learning_rate": 0.00022255700325732897,
+      "loss": 0.6313,
+      "step": 455
+    },
+    {
+      "epoch": 1.00164744645799,
+      "grad_norm": 0.05268587917089462,
+      "learning_rate": 0.00022231270358306186,
+      "loss": 0.7393,
+      "step": 456
+    },
+    {
+      "epoch": 1.0038440417353103,
+      "grad_norm": 0.04583257809281349,
+      "learning_rate": 0.00022206840390879477,
+      "loss": 0.7311,
+      "step": 457
+    },
+    {
+      "epoch": 1.0060406370126305,
+      "grad_norm": 0.041914235800504684,
+      "learning_rate": 0.00022182410423452766,
+      "loss": 0.6592,
+      "step": 458
+    },
+    {
+      "epoch": 1.0082372322899507,
+      "grad_norm": 0.05415317043662071,
+      "learning_rate": 0.00022157980456026058,
+      "loss": 0.5943,
+      "step": 459
+    },
+    {
+      "epoch": 1.0104338275672706,
+      "grad_norm": 0.039053451269865036,
+      "learning_rate": 0.00022133550488599347,
+      "loss": 0.5909,
+      "step": 460
+    },
+    {
+      "epoch": 1.0126304228445908,
+      "grad_norm": 0.04172314330935478,
+      "learning_rate": 0.00022109120521172636,
+      "loss": 0.529,
+      "step": 461
+    },
+    {
+      "epoch": 1.014827018121911,
+      "grad_norm": 0.059966228902339935,
+      "learning_rate": 0.00022084690553745927,
+      "loss": 0.8445,
+      "step": 462
+    },
+    {
+      "epoch": 1.0170236133992312,
+      "grad_norm": 0.054427627474069595,
+      "learning_rate": 0.00022060260586319216,
+      "loss": 0.7625,
+      "step": 463
+    },
+    {
+      "epoch": 1.0192202086765514,
+      "grad_norm": 0.05113132670521736,
+      "learning_rate": 0.00022035830618892505,
+      "loss": 0.6079,
+      "step": 464
+    },
+    {
+      "epoch": 1.0214168039538716,
+      "grad_norm": 0.05766759812831879,
+      "learning_rate": 0.00022011400651465797,
+      "loss": 0.7953,
+      "step": 465
+    },
+    {
+      "epoch": 1.0236133992311915,
+      "grad_norm": 0.055021774023771286,
+      "learning_rate": 0.00021986970684039086,
+      "loss": 0.8009,
+      "step": 466
+    },
+    {
+      "epoch": 1.0258099945085117,
+      "grad_norm": 0.049860063940286636,
+      "learning_rate": 0.00021962540716612375,
+      "loss": 0.683,
+      "step": 467
+    },
+    {
+      "epoch": 1.028006589785832,
+      "grad_norm": 0.059873066842556,
+      "learning_rate": 0.00021938110749185666,
+      "loss": 0.6204,
+      "step": 468
+    },
+    {
+      "epoch": 1.0302031850631521,
+      "grad_norm": 0.04648579657077789,
+      "learning_rate": 0.00021913680781758955,
+      "loss": 0.7137,
+      "step": 469
+    },
+    {
+      "epoch": 1.0323997803404723,
+      "grad_norm": 0.05671359598636627,
+      "learning_rate": 0.00021889250814332244,
+      "loss": 0.7367,
+      "step": 470
+    },
+    {
+      "epoch": 1.0345963756177925,
+      "grad_norm": 0.051784005016088486,
+      "learning_rate": 0.00021864820846905536,
+      "loss": 0.5914,
+      "step": 471
+    },
+    {
+      "epoch": 1.0367929708951125,
+      "grad_norm": 0.05361558869481087,
+      "learning_rate": 0.00021840390879478825,
+      "loss": 0.6544,
+      "step": 472
+    },
+    {
+      "epoch": 1.0389895661724327,
+      "grad_norm": 0.05350130423903465,
+      "learning_rate": 0.00021815960912052116,
+      "loss": 0.8281,
+      "step": 473
+    },
+    {
+      "epoch": 1.0411861614497528,
+      "grad_norm": 0.046200599521398544,
+      "learning_rate": 0.00021791530944625405,
+      "loss": 0.6194,
+      "step": 474
+    },
+    {
+      "epoch": 1.043382756727073,
+      "grad_norm": 0.08131860941648483,
+      "learning_rate": 0.00021767100977198694,
+      "loss": 0.6572,
+      "step": 475
+    },
+    {
+      "epoch": 1.0455793520043932,
+      "grad_norm": 0.050514254719018936,
+      "learning_rate": 0.00021742671009771986,
+      "loss": 0.7339,
+      "step": 476
+    },
+    {
+      "epoch": 1.0477759472817134,
+      "grad_norm": 0.05298462510108948,
+      "learning_rate": 0.00021718241042345275,
+      "loss": 0.5557,
+      "step": 477
+    },
+    {
+      "epoch": 1.0499725425590336,
+      "grad_norm": 0.05039222911000252,
+      "learning_rate": 0.00021693811074918564,
+      "loss": 0.7133,
+      "step": 478
+    },
+    {
+      "epoch": 1.0521691378363536,
+      "grad_norm": 0.05472245067358017,
+      "learning_rate": 0.00021669381107491855,
+      "loss": 0.7076,
+      "step": 479
+    },
+    {
+      "epoch": 1.0543657331136738,
+      "grad_norm": 0.039077069610357285,
+      "learning_rate": 0.00021644951140065144,
+      "loss": 0.5951,
+      "step": 480
+    },
+    {
+      "epoch": 1.056562328390994,
+      "grad_norm": 0.03200119361281395,
+      "learning_rate": 0.00021620521172638433,
+      "loss": 0.5825,
+      "step": 481
+    },
+    {
+      "epoch": 1.0587589236683141,
+      "grad_norm": 0.04200253635644913,
+      "learning_rate": 0.00021596091205211725,
+      "loss": 0.5881,
+      "step": 482
+    },
+    {
+      "epoch": 1.0609555189456343,
+      "grad_norm": 0.0591256357729435,
+      "learning_rate": 0.00021571661237785014,
+      "loss": 0.647,
+      "step": 483
+    },
+    {
+      "epoch": 1.0631521142229543,
+      "grad_norm": 0.04347528517246246,
+      "learning_rate": 0.00021547231270358305,
+      "loss": 0.671,
+      "step": 484
+    },
+    {
+      "epoch": 1.0653487095002745,
+      "grad_norm": 0.04377702996134758,
+      "learning_rate": 0.00021522801302931594,
+      "loss": 0.6533,
+      "step": 485
+    },
+    {
+      "epoch": 1.0675453047775947,
+      "grad_norm": 0.05368449166417122,
+      "learning_rate": 0.00021498371335504883,
+      "loss": 0.5239,
+      "step": 486
+    },
+    {
+      "epoch": 1.0697419000549149,
+      "grad_norm": 0.04010557755827904,
+      "learning_rate": 0.00021473941368078175,
+      "loss": 0.6293,
+      "step": 487
+    },
+    {
+      "epoch": 1.071938495332235,
+      "grad_norm": 0.04686107859015465,
+      "learning_rate": 0.00021449511400651464,
+      "loss": 0.7006,
+      "step": 488
+    },
+    {
+      "epoch": 1.0741350906095553,
+      "grad_norm": 0.03812983259558678,
+      "learning_rate": 0.00021425081433224753,
+      "loss": 0.6003,
+      "step": 489
+    },
+    {
+      "epoch": 1.0763316858868754,
+      "grad_norm": 0.035127971321344376,
+      "learning_rate": 0.00021400651465798044,
+      "loss": 0.5701,
+      "step": 490
+    },
+    {
+      "epoch": 1.0785282811641954,
+      "grad_norm": 0.07218277454376221,
+      "learning_rate": 0.00021376221498371333,
+      "loss": 0.8555,
+      "step": 491
+    },
+    {
+      "epoch": 1.0807248764415156,
+      "grad_norm": 0.046801965683698654,
+      "learning_rate": 0.00021351791530944622,
+      "loss": 0.5919,
+      "step": 492
+    },
+    {
+      "epoch": 1.0829214717188358,
+      "grad_norm": 0.06883776932954788,
+      "learning_rate": 0.00021327361563517914,
+      "loss": 0.7253,
+      "step": 493
+    },
+    {
+      "epoch": 1.085118066996156,
+      "grad_norm": 0.03979892656207085,
+      "learning_rate": 0.00021302931596091203,
+      "loss": 0.5356,
+      "step": 494
+    },
+    {
+      "epoch": 1.0873146622734762,
+      "grad_norm": 0.07111027836799622,
+      "learning_rate": 0.00021278501628664491,
+      "loss": 0.6938,
+      "step": 495
+    },
+    {
+      "epoch": 1.0895112575507964,
+      "grad_norm": 0.040993962436914444,
+      "learning_rate": 0.00021254071661237783,
+      "loss": 0.6093,
+      "step": 496
+    },
+    {
+      "epoch": 1.0917078528281163,
+      "grad_norm": 0.033425960689783096,
+      "learning_rate": 0.00021229641693811072,
+      "loss": 0.5784,
+      "step": 497
+    },
+    {
+      "epoch": 1.0939044481054365,
+      "grad_norm": 0.058829471468925476,
+      "learning_rate": 0.00021205211726384364,
+      "loss": 0.6227,
+      "step": 498
+    },
+    {
+      "epoch": 1.0961010433827567,
+      "grad_norm": 0.04688352346420288,
+      "learning_rate": 0.00021180781758957653,
+      "loss": 0.6884,
+      "step": 499
+    },
+    {
+      "epoch": 1.098297638660077,
+      "grad_norm": 0.059555862098932266,
+      "learning_rate": 0.00021156351791530941,
+      "loss": 0.7715,
+      "step": 500
+    },
+    {
+      "epoch": 1.100494233937397,
+      "grad_norm": 0.0509280227124691,
+      "learning_rate": 0.00021131921824104233,
+      "loss": 0.7527,
+      "step": 501
+    },
+    {
+      "epoch": 1.1026908292147173,
+      "grad_norm": 0.036750249564647675,
+      "learning_rate": 0.00021107491856677522,
+      "loss": 0.5671,
+      "step": 502
+    },
+    {
+      "epoch": 1.1048874244920373,
+      "grad_norm": 0.07052327692508698,
+      "learning_rate": 0.0002108306188925081,
+      "loss": 0.7258,
+      "step": 503
+    },
+    {
+      "epoch": 1.1070840197693574,
+      "grad_norm": 0.0397505946457386,
+      "learning_rate": 0.00021058631921824103,
+      "loss": 0.5837,
+      "step": 504
+    },
+    {
+      "epoch": 1.1092806150466776,
+      "grad_norm": 0.053040631115436554,
+      "learning_rate": 0.00021034201954397392,
+      "loss": 0.8447,
+      "step": 505
+    },
+    {
+      "epoch": 1.1114772103239978,
+      "grad_norm": 0.04555618017911911,
+      "learning_rate": 0.0002100977198697068,
+      "loss": 0.6716,
+      "step": 506
+    },
+    {
+      "epoch": 1.113673805601318,
+      "grad_norm": 0.04914266616106033,
+      "learning_rate": 0.00020985342019543972,
+      "loss": 0.6719,
+      "step": 507
+    },
+    {
+      "epoch": 1.1158704008786382,
+      "grad_norm": 0.05914847552776337,
+      "learning_rate": 0.0002096091205211726,
+      "loss": 0.738,
+      "step": 508
+    },
+    {
+      "epoch": 1.1180669961559582,
+      "grad_norm": 0.0362103171646595,
+      "learning_rate": 0.0002093648208469055,
+      "loss": 0.6914,
+      "step": 509
+    },
+    {
+      "epoch": 1.1202635914332784,
+      "grad_norm": 0.04875648021697998,
+      "learning_rate": 0.00020912052117263842,
+      "loss": 0.7716,
+      "step": 510
+    },
+    {
+      "epoch": 1.1224601867105986,
+      "grad_norm": 0.052181925624608994,
+      "learning_rate": 0.0002088762214983713,
+      "loss": 0.6453,
+      "step": 511
+    },
+    {
+      "epoch": 1.1246567819879187,
+      "grad_norm": 0.054814066737890244,
+      "learning_rate": 0.00020863192182410422,
+      "loss": 0.6659,
+      "step": 512
+    },
+    {
+      "epoch": 1.126853377265239,
+      "grad_norm": 0.047833219170570374,
+      "learning_rate": 0.0002083876221498371,
+      "loss": 0.7771,
+      "step": 513
+    },
+    {
+      "epoch": 1.1290499725425591,
+      "grad_norm": 0.056477759033441544,
+      "learning_rate": 0.00020814332247557,
+      "loss": 0.6761,
+      "step": 514
+    },
+    {
+      "epoch": 1.131246567819879,
+      "grad_norm": 0.050696369260549545,
+      "learning_rate": 0.00020789902280130292,
+      "loss": 0.6711,
+      "step": 515
+    },
+    {
+      "epoch": 1.1334431630971993,
+      "grad_norm": 0.05367648974061012,
+      "learning_rate": 0.0002076547231270358,
+      "loss": 0.6757,
+      "step": 516
+    },
+    {
+      "epoch": 1.1356397583745195,
+      "grad_norm": 0.05610479786992073,
+      "learning_rate": 0.0002074104234527687,
+      "loss": 0.6412,
+      "step": 517
+    },
+    {
+      "epoch": 1.1378363536518397,
+      "grad_norm": 0.0620594285428524,
+      "learning_rate": 0.0002071661237785016,
+      "loss": 0.519,
+      "step": 518
+    },
+    {
+      "epoch": 1.1400329489291599,
+      "grad_norm": 0.05160127580165863,
+      "learning_rate": 0.0002069218241042345,
+      "loss": 0.4844,
+      "step": 519
+    },
+    {
+      "epoch": 1.14222954420648,
+      "grad_norm": 0.04619447514414787,
+      "learning_rate": 0.0002066775244299674,
+      "loss": 0.6709,
+      "step": 520
+    },
+    {
+      "epoch": 1.1444261394838002,
+      "grad_norm": 0.055884506553411484,
+      "learning_rate": 0.0002064332247557003,
+      "loss": 0.5976,
+      "step": 521
+    },
+    {
+      "epoch": 1.1466227347611202,
+      "grad_norm": 0.050843894481658936,
+      "learning_rate": 0.0002061889250814332,
+      "loss": 0.6171,
+      "step": 522
+    },
+    {
+      "epoch": 1.1488193300384404,
+      "grad_norm": 0.04294108971953392,
+      "learning_rate": 0.0002059446254071661,
+      "loss": 0.5099,
+      "step": 523
+    },
+    {
+      "epoch": 1.1510159253157606,
+      "grad_norm": 0.04971957579255104,
+      "learning_rate": 0.000205700325732899,
+      "loss": 0.7442,
+      "step": 524
+    },
+    {
+      "epoch": 1.1532125205930808,
+      "grad_norm": 0.037918634712696075,
+      "learning_rate": 0.0002054560260586319,
+      "loss": 0.6345,
+      "step": 525
+    },
+    {
+      "epoch": 1.155409115870401,
+      "grad_norm": 0.0499645359814167,
+      "learning_rate": 0.0002052117263843648,
+      "loss": 0.8493,
+      "step": 526
+    },
+    {
+      "epoch": 1.157605711147721,
+      "grad_norm": 0.04419689625501633,
+      "learning_rate": 0.0002049674267100977,
+      "loss": 0.5747,
+      "step": 527
+    },
+    {
+      "epoch": 1.1598023064250411,
+      "grad_norm": 0.06275136768817902,
+      "learning_rate": 0.00020472312703583058,
+      "loss": 0.7256,
+      "step": 528
+    },
+    {
+      "epoch": 1.1619989017023613,
+      "grad_norm": 0.05085553973913193,
+      "learning_rate": 0.0002044788273615635,
+      "loss": 0.5588,
+      "step": 529
+    },
+    {
+      "epoch": 1.1641954969796815,
+      "grad_norm": 0.04484010860323906,
+      "learning_rate": 0.0002042345276872964,
+      "loss": 0.4919,
+      "step": 530
+    },
+    {
+      "epoch": 1.1663920922570017,
+      "grad_norm": 0.044836804270744324,
+      "learning_rate": 0.00020399022801302928,
+      "loss": 0.5655,
+      "step": 531
+    },
+    {
+      "epoch": 1.1685886875343219,
+      "grad_norm": 0.04288703575730324,
+      "learning_rate": 0.0002037459283387622,
+      "loss": 0.7729,
+      "step": 532
+    },
+    {
+      "epoch": 1.170785282811642,
+      "grad_norm": 0.04978484287858009,
+      "learning_rate": 0.00020350162866449508,
+      "loss": 0.6874,
+      "step": 533
+    },
+    {
+      "epoch": 1.172981878088962,
+      "grad_norm": 0.061364252120256424,
+      "learning_rate": 0.00020325732899022797,
+      "loss": 0.7798,
+      "step": 534
+    },
+    {
+      "epoch": 1.1751784733662822,
+      "grad_norm": 0.04522540792822838,
+      "learning_rate": 0.0002030130293159609,
+      "loss": 0.6139,
+      "step": 535
+    },
+    {
+      "epoch": 1.1773750686436024,
+      "grad_norm": 0.0677039623260498,
+      "learning_rate": 0.00020276872964169378,
+      "loss": 0.6335,
+      "step": 536
+    },
+    {
+      "epoch": 1.1795716639209226,
+      "grad_norm": 0.04621490091085434,
+      "learning_rate": 0.0002025244299674267,
+      "loss": 0.6156,
+      "step": 537
+    },
+    {
+      "epoch": 1.1817682591982428,
+      "grad_norm": 0.04518371820449829,
+      "learning_rate": 0.00020228013029315958,
+      "loss": 0.5472,
+      "step": 538
+    },
+    {
+      "epoch": 1.1839648544755628,
+      "grad_norm": 0.0515042245388031,
+      "learning_rate": 0.00020203583061889247,
+      "loss": 0.8595,
+      "step": 539
+    },
+    {
+      "epoch": 1.186161449752883,
+      "grad_norm": 0.0516064427793026,
+      "learning_rate": 0.0002017915309446254,
+      "loss": 0.595,
+      "step": 540
+    },
+    {
+      "epoch": 1.1883580450302031,
+      "grad_norm": 0.049294911324977875,
+      "learning_rate": 0.00020154723127035828,
+      "loss": 0.6437,
+      "step": 541
+    },
+    {
+      "epoch": 1.1905546403075233,
+      "grad_norm": 0.03713231161236763,
+      "learning_rate": 0.00020130293159609117,
+      "loss": 0.5496,
+      "step": 542
+    },
+    {
+      "epoch": 1.1927512355848435,
+      "grad_norm": 0.04774945229291916,
+      "learning_rate": 0.00020105863192182408,
+      "loss": 0.7191,
+      "step": 543
+    },
+    {
+      "epoch": 1.1949478308621637,
+      "grad_norm": 0.04245544224977493,
+      "learning_rate": 0.00020081433224755697,
+      "loss": 0.6713,
+      "step": 544
+    },
+    {
+      "epoch": 1.197144426139484,
+      "grad_norm": 0.05695553123950958,
+      "learning_rate": 0.00020057003257328986,
+      "loss": 0.7036,
+      "step": 545
+    },
+    {
+      "epoch": 1.1993410214168039,
+      "grad_norm": 0.043208975344896317,
+      "learning_rate": 0.00020032573289902278,
+      "loss": 0.6773,
+      "step": 546
+    },
+    {
+      "epoch": 1.201537616694124,
+      "grad_norm": 0.061872243881225586,
+      "learning_rate": 0.00020008143322475567,
+      "loss": 0.9214,
+      "step": 547
+    },
+    {
+      "epoch": 1.2037342119714443,
+      "grad_norm": 0.03670836240053177,
+      "learning_rate": 0.00019983713355048856,
+      "loss": 0.6221,
+      "step": 548
+    },
+    {
+      "epoch": 1.2059308072487644,
+      "grad_norm": 0.05437465012073517,
+      "learning_rate": 0.00019959283387622147,
+      "loss": 0.8771,
+      "step": 549
+    },
+    {
+      "epoch": 1.2081274025260846,
+      "grad_norm": 0.07304202020168304,
+      "learning_rate": 0.00019934853420195436,
+      "loss": 0.7336,
+      "step": 550
+    },
+    {
+      "epoch": 1.2103239978034046,
+      "grad_norm": 0.05822945386171341,
+      "learning_rate": 0.00019910423452768728,
+      "loss": 0.6365,
+      "step": 551
+    },
+    {
+      "epoch": 1.2125205930807248,
+      "grad_norm": 0.051955416798591614,
+      "learning_rate": 0.00019885993485342017,
+      "loss": 0.7081,
+      "step": 552
+    },
+    {
+      "epoch": 1.214717188358045,
+      "grad_norm": 0.04791713133454323,
+      "learning_rate": 0.00019861563517915306,
+      "loss": 0.7363,
+      "step": 553
+    },
+    {
+      "epoch": 1.2169137836353652,
+      "grad_norm": 0.038635943084955215,
+      "learning_rate": 0.00019837133550488597,
+      "loss": 0.677,
+      "step": 554
+    },
+    {
+      "epoch": 1.2191103789126854,
+      "grad_norm": 0.07108104974031448,
+      "learning_rate": 0.00019812703583061886,
+      "loss": 0.7667,
+      "step": 555
+    },
+    {
+      "epoch": 1.2213069741900056,
+      "grad_norm": 0.050478167831897736,
+      "learning_rate": 0.00019788273615635175,
+      "loss": 0.5538,
+      "step": 556
+    },
+    {
+      "epoch": 1.2235035694673257,
+      "grad_norm": 0.050831057131290436,
+      "learning_rate": 0.00019763843648208467,
+      "loss": 0.6641,
+      "step": 557
+    },
+    {
+      "epoch": 1.2257001647446457,
+      "grad_norm": 0.058054275810718536,
+      "learning_rate": 0.00019739413680781756,
+      "loss": 0.6325,
+      "step": 558
+    },
+    {
+      "epoch": 1.227896760021966,
+      "grad_norm": 0.085708387196064,
+      "learning_rate": 0.00019714983713355045,
+      "loss": 0.8618,
+      "step": 559
+    },
+    {
+      "epoch": 1.230093355299286,
+      "grad_norm": 0.04360990226268768,
+      "learning_rate": 0.00019690553745928336,
+      "loss": 0.7238,
+      "step": 560
+    },
+    {
+      "epoch": 1.2322899505766063,
+      "grad_norm": 0.07026379555463791,
+      "learning_rate": 0.00019666123778501625,
+      "loss": 0.6265,
+      "step": 561
+    },
+    {
+      "epoch": 1.2344865458539265,
+      "grad_norm": 0.05597195774316788,
+      "learning_rate": 0.00019641693811074917,
+      "loss": 0.6385,
+      "step": 562
+    },
+    {
+      "epoch": 1.2366831411312464,
+      "grad_norm": 0.06764718890190125,
+      "learning_rate": 0.00019617263843648206,
+      "loss": 0.8007,
+      "step": 563
+    },
+    {
+      "epoch": 1.2388797364085666,
+      "grad_norm": 0.05328952148556709,
+      "learning_rate": 0.00019592833876221495,
+      "loss": 0.5902,
+      "step": 564
+    },
+    {
+      "epoch": 1.2410763316858868,
+      "grad_norm": 0.045794982463121414,
+      "learning_rate": 0.00019568403908794786,
+      "loss": 0.6765,
+      "step": 565
+    },
+    {
+      "epoch": 1.243272926963207,
+      "grad_norm": 0.09876284003257751,
+      "learning_rate": 0.00019543973941368075,
+      "loss": 0.6767,
+      "step": 566
+    },
+    {
+      "epoch": 1.2454695222405272,
+      "grad_norm": 0.04325791075825691,
+      "learning_rate": 0.00019519543973941364,
+      "loss": 0.6679,
+      "step": 567
+    },
+    {
+      "epoch": 1.2476661175178474,
+      "grad_norm": 0.04940588399767876,
+      "learning_rate": 0.00019495114006514656,
+      "loss": 0.6811,
+      "step": 568
+    },
+    {
+      "epoch": 1.2498627127951676,
+      "grad_norm": 0.04802437499165535,
+      "learning_rate": 0.00019470684039087945,
+      "loss": 0.5913,
+      "step": 569
+    },
+    {
+      "epoch": 1.2520593080724876,
+      "grad_norm": 0.04493922367691994,
+      "learning_rate": 0.00019446254071661234,
+      "loss": 0.472,
+      "step": 570
+    },
+    {
+      "epoch": 1.2542559033498077,
+      "grad_norm": 0.05015527084469795,
+      "learning_rate": 0.00019421824104234525,
+      "loss": 0.791,
+      "step": 571
+    },
+    {
+      "epoch": 1.256452498627128,
+      "grad_norm": 0.061855364590883255,
+      "learning_rate": 0.00019397394136807814,
+      "loss": 0.6994,
+      "step": 572
+    },
+    {
+      "epoch": 1.2586490939044481,
+      "grad_norm": 0.05905802920460701,
+      "learning_rate": 0.00019372964169381103,
+      "loss": 0.6168,
+      "step": 573
+    },
+    {
+      "epoch": 1.2608456891817683,
+      "grad_norm": 0.04185137152671814,
+      "learning_rate": 0.00019348534201954395,
+      "loss": 0.7272,
+      "step": 574
+    },
+    {
+      "epoch": 1.2630422844590883,
+      "grad_norm": 0.04353105649352074,
+      "learning_rate": 0.00019324104234527684,
+      "loss": 0.7509,
+      "step": 575
+    },
+    {
+      "epoch": 1.2652388797364087,
+      "grad_norm": 0.0427667535841465,
+      "learning_rate": 0.00019299674267100975,
+      "loss": 0.6093,
+      "step": 576
+    },
+    {
+      "epoch": 1.2674354750137287,
+      "grad_norm": 0.05642359331250191,
+      "learning_rate": 0.00019275244299674264,
+      "loss": 0.8424,
+      "step": 577
+    },
+    {
+      "epoch": 1.2696320702910489,
+      "grad_norm": 0.04383740574121475,
+      "learning_rate": 0.00019250814332247553,
+      "loss": 0.4776,
+      "step": 578
+    },
+    {
+      "epoch": 1.271828665568369,
+      "grad_norm": 0.048313695937395096,
+      "learning_rate": 0.00019226384364820845,
+      "loss": 0.5038,
+      "step": 579
+    },
+    {
+      "epoch": 1.2740252608456892,
+      "grad_norm": 0.038794662803411484,
+      "learning_rate": 0.00019201954397394134,
+      "loss": 0.5157,
+      "step": 580
+    },
+    {
+      "epoch": 1.2762218561230094,
+      "grad_norm": 0.04513511061668396,
+      "learning_rate": 0.00019177524429967423,
+      "loss": 0.678,
+      "step": 581
+    },
+    {
+      "epoch": 1.2784184514003294,
+      "grad_norm": 0.05818026885390282,
+      "learning_rate": 0.00019153094462540714,
+      "loss": 0.6184,
+      "step": 582
+    },
+    {
+      "epoch": 1.2806150466776496,
+      "grad_norm": 0.06140175834298134,
+      "learning_rate": 0.00019128664495114003,
+      "loss": 0.5967,
+      "step": 583
+    },
+    {
+      "epoch": 1.2828116419549698,
+      "grad_norm": 0.0904388502240181,
+      "learning_rate": 0.00019104234527687292,
+      "loss": 0.5606,
+      "step": 584
+    },
+    {
+      "epoch": 1.28500823723229,
+      "grad_norm": 0.06066805124282837,
+      "learning_rate": 0.00019079804560260584,
+      "loss": 0.7383,
+      "step": 585
+    },
+    {
+      "epoch": 1.2872048325096102,
+      "grad_norm": 0.04117051884531975,
+      "learning_rate": 0.00019055374592833873,
+      "loss": 0.6536,
+      "step": 586
+    },
+    {
+      "epoch": 1.2894014277869301,
+      "grad_norm": 0.04647667706012726,
+      "learning_rate": 0.00019030944625407164,
+      "loss": 0.6158,
+      "step": 587
+    },
+    {
+      "epoch": 1.2915980230642505,
+      "grad_norm": 0.05287046730518341,
+      "learning_rate": 0.00019006514657980453,
+      "loss": 0.6545,
+      "step": 588
+    },
+    {
+      "epoch": 1.2937946183415705,
+      "grad_norm": 0.05430927500128746,
+      "learning_rate": 0.00018982084690553742,
+      "loss": 0.6284,
+      "step": 589
+    },
+    {
+      "epoch": 1.2959912136188907,
+      "grad_norm": 0.04469500109553337,
+      "learning_rate": 0.00018957654723127034,
+      "loss": 0.6741,
+      "step": 590
+    },
+    {
+      "epoch": 1.2981878088962109,
+      "grad_norm": 0.030649427324533463,
+      "learning_rate": 0.00018933224755700323,
+      "loss": 0.5046,
+      "step": 591
+    },
+    {
+      "epoch": 1.300384404173531,
+      "grad_norm": 0.046733759343624115,
+      "learning_rate": 0.00018908794788273612,
+      "loss": 0.6558,
+      "step": 592
+    },
+    {
+      "epoch": 1.3025809994508513,
+      "grad_norm": 0.040193550288677216,
+      "learning_rate": 0.00018884364820846903,
+      "loss": 0.569,
+      "step": 593
+    },
+    {
+      "epoch": 1.3047775947281712,
+      "grad_norm": 0.03978565335273743,
+      "learning_rate": 0.00018859934853420192,
+      "loss": 0.6768,
+      "step": 594
+    },
+    {
+      "epoch": 1.3069741900054914,
+      "grad_norm": 0.04483840614557266,
+      "learning_rate": 0.0001883550488599348,
+      "loss": 0.7312,
+      "step": 595
+    },
+    {
+      "epoch": 1.3091707852828116,
+      "grad_norm": 0.0462617352604866,
+      "learning_rate": 0.00018811074918566773,
+      "loss": 0.5968,
+      "step": 596
+    },
+    {
+      "epoch": 1.3113673805601318,
+      "grad_norm": 0.048313964158296585,
+      "learning_rate": 0.00018786644951140062,
+      "loss": 0.6074,
+      "step": 597
+    },
+    {
+      "epoch": 1.313563975837452,
+      "grad_norm": 0.04513964429497719,
+      "learning_rate": 0.0001876221498371335,
+      "loss": 0.5612,
+      "step": 598
+    },
+    {
+      "epoch": 1.3157605711147722,
+      "grad_norm": 0.04602311924099922,
+      "learning_rate": 0.00018737785016286645,
+      "loss": 0.6196,
+      "step": 599
+    },
+    {
+      "epoch": 1.3179571663920924,
+      "grad_norm": 0.03924334794282913,
+      "learning_rate": 0.00018713355048859934,
+      "loss": 0.7697,
+      "step": 600
+    },
+    {
+      "epoch": 1.3201537616694123,
+      "grad_norm": 0.046174440532922745,
+      "learning_rate": 0.00018688925081433225,
+      "loss": 0.6409,
+      "step": 601
+    },
+    {
+      "epoch": 1.3223503569467325,
+      "grad_norm": 0.04037371277809143,
+      "learning_rate": 0.00018664495114006514,
+      "loss": 0.6609,
+      "step": 602
+    },
+    {
+      "epoch": 1.3245469522240527,
+      "grad_norm": 0.042371075600385666,
+      "learning_rate": 0.00018640065146579803,
+      "loss": 0.6816,
+      "step": 603
+    },
+    {
+      "epoch": 1.326743547501373,
+      "grad_norm": 0.04608074575662613,
+      "learning_rate": 0.00018615635179153095,
+      "loss": 0.6682,
+      "step": 604
+    },
+    {
+      "epoch": 1.328940142778693,
+      "grad_norm": 0.046598054468631744,
+      "learning_rate": 0.00018591205211726384,
+      "loss": 0.7451,
+      "step": 605
+    },
+    {
+      "epoch": 1.331136738056013,
+      "grad_norm": 0.0409809872508049,
+      "learning_rate": 0.00018566775244299675,
+      "loss": 0.5852,
+      "step": 606
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 0.051450323313474655,
+      "learning_rate": 0.00018542345276872964,
+      "loss": 0.5106,
+      "step": 607
+    },
+    {
+      "epoch": 1.3355299286106534,
+      "grad_norm": 0.043831150978803635,
+      "learning_rate": 0.00018517915309446253,
+      "loss": 0.6222,
+      "step": 608
+    },
+    {
+      "epoch": 1.3377265238879736,
+      "grad_norm": 0.0562346987426281,
+      "learning_rate": 0.00018493485342019545,
+      "loss": 0.6442,
+      "step": 609
+    },
+    {
+      "epoch": 1.3399231191652938,
+      "grad_norm": 0.04941529780626297,
+      "learning_rate": 0.00018469055374592834,
+      "loss": 0.6927,
+      "step": 610
+    },
+    {
+      "epoch": 1.342119714442614,
+      "grad_norm": 0.051082007586956024,
+      "learning_rate": 0.00018444625407166123,
+      "loss": 0.7017,
+      "step": 611
+    },
+    {
+      "epoch": 1.3443163097199342,
+      "grad_norm": 0.046207696199417114,
+      "learning_rate": 0.00018420195439739414,
+      "loss": 0.5908,
+      "step": 612
+    },
+    {
+      "epoch": 1.3465129049972542,
+      "grad_norm": 0.039629098027944565,
+      "learning_rate": 0.00018395765472312703,
+      "loss": 0.7068,
+      "step": 613
+    },
+    {
+      "epoch": 1.3487095002745744,
+      "grad_norm": 0.045384980738162994,
+      "learning_rate": 0.00018371335504885992,
+      "loss": 0.5844,
+      "step": 614
+    },
+    {
+      "epoch": 1.3509060955518946,
+      "grad_norm": 0.05367980897426605,
+      "learning_rate": 0.00018346905537459284,
+      "loss": 0.7222,
+      "step": 615
+    },
+    {
+      "epoch": 1.3531026908292147,
+      "grad_norm": 0.061644140630960464,
+      "learning_rate": 0.00018322475570032573,
+      "loss": 0.7469,
+      "step": 616
+    },
+    {
+      "epoch": 1.355299286106535,
+      "grad_norm": 0.04570198804140091,
+      "learning_rate": 0.00018298045602605864,
+      "loss": 0.6605,
+      "step": 617
+    },
+    {
+      "epoch": 1.357495881383855,
+      "grad_norm": 0.05299101769924164,
+      "learning_rate": 0.00018273615635179153,
+      "loss": 0.8565,
+      "step": 618
+    },
+    {
+      "epoch": 1.359692476661175,
+      "grad_norm": 0.043971553444862366,
+      "learning_rate": 0.00018249185667752442,
+      "loss": 0.509,
+      "step": 619
+    },
+    {
+      "epoch": 1.3618890719384953,
+      "grad_norm": 0.04689047113060951,
+      "learning_rate": 0.00018224755700325734,
+      "loss": 0.5978,
+      "step": 620
+    },
+    {
+      "epoch": 1.3640856672158155,
+      "grad_norm": 0.0547180213034153,
+      "learning_rate": 0.00018200325732899023,
+      "loss": 0.8353,
+      "step": 621
+    },
+    {
+      "epoch": 1.3662822624931357,
+      "grad_norm": 0.045855265110731125,
+      "learning_rate": 0.00018175895765472312,
+      "loss": 0.6055,
+      "step": 622
+    },
+    {
+      "epoch": 1.3684788577704559,
+      "grad_norm": 0.043083783239126205,
+      "learning_rate": 0.00018151465798045603,
+      "loss": 0.5349,
+      "step": 623
+    },
+    {
+      "epoch": 1.370675453047776,
+      "grad_norm": 0.056552454829216,
+      "learning_rate": 0.00018127035830618892,
+      "loss": 0.6284,
+      "step": 624
+    },
+    {
+      "epoch": 1.372872048325096,
+      "grad_norm": 0.05099363625049591,
+      "learning_rate": 0.0001810260586319218,
+      "loss": 0.7383,
+      "step": 625
+    },
+    {
+      "epoch": 1.3750686436024162,
+      "grad_norm": 0.039746589958667755,
+      "learning_rate": 0.00018078175895765473,
+      "loss": 0.5563,
+      "step": 626
+    },
+    {
+      "epoch": 1.3772652388797364,
+      "grad_norm": 0.03771736100316048,
+      "learning_rate": 0.00018053745928338762,
+      "loss": 0.6183,
+      "step": 627
+    },
+    {
+      "epoch": 1.3794618341570566,
+      "grad_norm": 0.06937503814697266,
+      "learning_rate": 0.0001802931596091205,
+      "loss": 0.4919,
+      "step": 628
+    },
+    {
+      "epoch": 1.3816584294343768,
+      "grad_norm": 0.04858769476413727,
+      "learning_rate": 0.00018004885993485342,
+      "loss": 0.6382,
+      "step": 629
+    },
+    {
+      "epoch": 1.3838550247116967,
+      "grad_norm": 0.04866888001561165,
+      "learning_rate": 0.0001798045602605863,
+      "loss": 0.6089,
+      "step": 630
+    },
+    {
+      "epoch": 1.3860516199890172,
+      "grad_norm": 0.04268645867705345,
+      "learning_rate": 0.00017956026058631923,
+      "loss": 0.63,
+      "step": 631
+    },
+    {
+      "epoch": 1.3882482152663371,
+      "grad_norm": 0.04429534077644348,
+      "learning_rate": 0.00017931596091205212,
+      "loss": 0.7758,
+      "step": 632
+    },
+    {
+      "epoch": 1.3904448105436573,
+      "grad_norm": 0.04046213626861572,
+      "learning_rate": 0.000179071661237785,
+      "loss": 0.6882,
+      "step": 633
+    },
+    {
+      "epoch": 1.3926414058209775,
+      "grad_norm": 0.049879636615514755,
+      "learning_rate": 0.00017882736156351792,
+      "loss": 0.5664,
+      "step": 634
+    },
+    {
+      "epoch": 1.3948380010982977,
+      "grad_norm": 0.051655709743499756,
+      "learning_rate": 0.0001785830618892508,
+      "loss": 0.7156,
+      "step": 635
+    },
+    {
+      "epoch": 1.3970345963756179,
+      "grad_norm": 0.03926423192024231,
+      "learning_rate": 0.0001783387622149837,
+      "loss": 0.7299,
+      "step": 636
+    },
+    {
+      "epoch": 1.3992311916529379,
+      "grad_norm": 0.059483956545591354,
+      "learning_rate": 0.00017809446254071662,
+      "loss": 0.5872,
+      "step": 637
+    },
+    {
+      "epoch": 1.401427786930258,
+      "grad_norm": 0.04855528101325035,
+      "learning_rate": 0.0001778501628664495,
+      "loss": 0.5601,
+      "step": 638
+    },
+    {
+      "epoch": 1.4036243822075782,
+      "grad_norm": 0.048911917954683304,
+      "learning_rate": 0.0001776058631921824,
+      "loss": 0.5874,
+      "step": 639
+    },
+    {
+      "epoch": 1.4058209774848984,
+      "grad_norm": 0.04301845282316208,
+      "learning_rate": 0.0001773615635179153,
+      "loss": 0.8704,
+      "step": 640
+    },
+    {
+      "epoch": 1.4080175727622186,
+      "grad_norm": 0.043345287442207336,
+      "learning_rate": 0.0001771172638436482,
+      "loss": 0.6319,
+      "step": 641
+    },
+    {
+      "epoch": 1.4102141680395386,
+      "grad_norm": 0.04333937540650368,
+      "learning_rate": 0.0001768729641693811,
+      "loss": 0.7228,
+      "step": 642
+    },
+    {
+      "epoch": 1.412410763316859,
+      "grad_norm": 0.20310437679290771,
+      "learning_rate": 0.000176628664495114,
+      "loss": 0.5866,
+      "step": 643
+    },
+    {
+      "epoch": 1.414607358594179,
+      "grad_norm": 0.04585960507392883,
+      "learning_rate": 0.0001763843648208469,
+      "loss": 0.5515,
+      "step": 644
+    },
+    {
+      "epoch": 1.4168039538714992,
+      "grad_norm": 0.06021832302212715,
+      "learning_rate": 0.0001761400651465798,
+      "loss": 0.575,
+      "step": 645
+    },
+    {
+      "epoch": 1.4190005491488193,
+      "grad_norm": 0.04396074265241623,
+      "learning_rate": 0.0001758957654723127,
+      "loss": 0.5602,
+      "step": 646
+    },
+    {
+      "epoch": 1.4211971444261395,
+      "grad_norm": 0.05046122521162033,
+      "learning_rate": 0.0001756514657980456,
+      "loss": 0.6205,
+      "step": 647
+    },
+    {
+      "epoch": 1.4233937397034597,
+      "grad_norm": 0.045648425817489624,
+      "learning_rate": 0.0001754071661237785,
+      "loss": 0.5896,
+      "step": 648
+    },
+    {
+      "epoch": 1.4255903349807797,
+      "grad_norm": 0.06430377811193466,
+      "learning_rate": 0.0001751628664495114,
+      "loss": 0.7242,
+      "step": 649
+    },
+    {
+      "epoch": 1.4277869302580999,
+      "grad_norm": 0.04144120216369629,
+      "learning_rate": 0.00017491856677524429,
+      "loss": 0.5391,
+      "step": 650
+    },
+    {
+      "epoch": 1.42998352553542,
+      "grad_norm": 0.037751659750938416,
+      "learning_rate": 0.0001746742671009772,
+      "loss": 0.6035,
+      "step": 651
+    },
+    {
+      "epoch": 1.4321801208127403,
+      "grad_norm": 0.04347790405154228,
+      "learning_rate": 0.0001744299674267101,
+      "loss": 0.5921,
+      "step": 652
+    },
+    {
+      "epoch": 1.4343767160900605,
+      "grad_norm": 0.04341914877295494,
+      "learning_rate": 0.00017418566775244298,
+      "loss": 0.6021,
+      "step": 653
+    },
+    {
+      "epoch": 1.4365733113673804,
+      "grad_norm": 0.04380672052502632,
+      "learning_rate": 0.0001739413680781759,
+      "loss": 0.57,
+      "step": 654
+    },
+    {
+      "epoch": 1.4387699066447008,
+      "grad_norm": 0.05512508749961853,
+      "learning_rate": 0.00017369706840390879,
+      "loss": 0.6191,
+      "step": 655
+    },
+    {
+      "epoch": 1.4409665019220208,
+      "grad_norm": 0.05672155320644379,
+      "learning_rate": 0.0001734527687296417,
+      "loss": 0.7185,
+      "step": 656
+    },
+    {
+      "epoch": 1.443163097199341,
+      "grad_norm": 0.04579592123627663,
+      "learning_rate": 0.0001732084690553746,
+      "loss": 0.7705,
+      "step": 657
+    },
+    {
+      "epoch": 1.4453596924766612,
+      "grad_norm": 0.04407493770122528,
+      "learning_rate": 0.00017296416938110748,
+      "loss": 0.7518,
+      "step": 658
+    },
+    {
+      "epoch": 1.4475562877539814,
+      "grad_norm": 0.05515037477016449,
+      "learning_rate": 0.0001727198697068404,
+      "loss": 0.8036,
+      "step": 659
+    },
+    {
+      "epoch": 1.4497528830313016,
+      "grad_norm": 0.04245726764202118,
+      "learning_rate": 0.00017247557003257329,
+      "loss": 0.5768,
+      "step": 660
+    },
+    {
+      "epoch": 1.4519494783086215,
+      "grad_norm": 0.03963717445731163,
+      "learning_rate": 0.00017223127035830618,
+      "loss": 0.5828,
+      "step": 661
+    },
+    {
+      "epoch": 1.4541460735859417,
+      "grad_norm": 0.05516964569687843,
+      "learning_rate": 0.0001719869706840391,
+      "loss": 0.6889,
+      "step": 662
+    },
+    {
+      "epoch": 1.456342668863262,
+      "grad_norm": 0.0535304956138134,
+      "learning_rate": 0.00017174267100977198,
+      "loss": 0.6561,
+      "step": 663
+    },
+    {
+      "epoch": 1.458539264140582,
+      "grad_norm": 0.03565557673573494,
+      "learning_rate": 0.00017149837133550487,
+      "loss": 0.5585,
+      "step": 664
+    },
+    {
+      "epoch": 1.4607358594179023,
+      "grad_norm": 0.0661967322230339,
+      "learning_rate": 0.00017125407166123779,
+      "loss": 0.8141,
+      "step": 665
+    },
+    {
+      "epoch": 1.4629324546952225,
+      "grad_norm": 0.060851022601127625,
+      "learning_rate": 0.00017100977198697068,
+      "loss": 0.7993,
+      "step": 666
+    },
+    {
+      "epoch": 1.4651290499725427,
+      "grad_norm": 0.055130865424871445,
+      "learning_rate": 0.00017076547231270356,
+      "loss": 0.6111,
+      "step": 667
+    },
+    {
+      "epoch": 1.4673256452498626,
+      "grad_norm": 0.044011928141117096,
+      "learning_rate": 0.00017052117263843648,
+      "loss": 0.7623,
+      "step": 668
+    },
+    {
+      "epoch": 1.4695222405271828,
+      "grad_norm": 0.04537820816040039,
+      "learning_rate": 0.00017027687296416937,
+      "loss": 0.6055,
+      "step": 669
+    },
+    {
+      "epoch": 1.471718835804503,
+      "grad_norm": 0.045265693217515945,
+      "learning_rate": 0.00017003257328990229,
+      "loss": 0.6416,
+      "step": 670
+    },
+    {
+      "epoch": 1.4739154310818232,
+      "grad_norm": 0.07514453679323196,
+      "learning_rate": 0.00016978827361563518,
+      "loss": 0.7886,
+      "step": 671
+    },
+    {
+      "epoch": 1.4761120263591434,
+      "grad_norm": 0.041896089911460876,
+      "learning_rate": 0.00016954397394136806,
+      "loss": 0.6698,
+      "step": 672
+    },
+    {
+      "epoch": 1.4783086216364634,
+      "grad_norm": 0.050291791558265686,
+      "learning_rate": 0.00016929967426710098,
+      "loss": 0.6528,
+      "step": 673
+    },
+    {
+      "epoch": 1.4805052169137836,
+      "grad_norm": 0.037543244659900665,
+      "learning_rate": 0.00016905537459283387,
+      "loss": 0.6816,
+      "step": 674
+    },
+    {
+      "epoch": 1.4827018121911038,
+      "grad_norm": 0.06479065120220184,
+      "learning_rate": 0.00016881107491856676,
+      "loss": 0.5854,
+      "step": 675
+    },
+    {
+      "epoch": 1.484898407468424,
+      "grad_norm": 0.0435757115483284,
+      "learning_rate": 0.00016856677524429968,
+      "loss": 0.6497,
+      "step": 676
+    },
+    {
+      "epoch": 1.4870950027457441,
+      "grad_norm": 0.04248249530792236,
+      "learning_rate": 0.00016832247557003257,
+      "loss": 0.6321,
+      "step": 677
+    },
+    {
+      "epoch": 1.4892915980230643,
+      "grad_norm": 0.056113965809345245,
+      "learning_rate": 0.00016807817589576545,
+      "loss": 0.7081,
+      "step": 678
+    },
+    {
+      "epoch": 1.4914881933003845,
+      "grad_norm": 0.04325173795223236,
+      "learning_rate": 0.00016783387622149837,
+      "loss": 0.6985,
+      "step": 679
+    },
+    {
+      "epoch": 1.4936847885777045,
+      "grad_norm": 0.10305589437484741,
+      "learning_rate": 0.00016758957654723126,
+      "loss": 0.8552,
+      "step": 680
+    },
+    {
+      "epoch": 1.4958813838550247,
+      "grad_norm": 0.04724079370498657,
+      "learning_rate": 0.00016734527687296415,
+      "loss": 0.6303,
+      "step": 681
+    },
+    {
+      "epoch": 1.4980779791323449,
+      "grad_norm": 0.05535902827978134,
+      "learning_rate": 0.00016710097719869707,
+      "loss": 0.8582,
+      "step": 682
+    },
+    {
+      "epoch": 1.500274574409665,
+      "grad_norm": 0.04309092089533806,
+      "learning_rate": 0.00016685667752442995,
+      "loss": 0.4383,
+      "step": 683
+    },
+    {
+      "epoch": 1.5024711696869852,
+      "grad_norm": 0.049390096217393875,
+      "learning_rate": 0.00016661237785016287,
+      "loss": 0.5833,
+      "step": 684
+    },
+    {
+      "epoch": 1.5046677649643052,
+      "grad_norm": 0.0520150363445282,
+      "learning_rate": 0.00016636807817589576,
+      "loss": 0.7291,
+      "step": 685
+    },
+    {
+      "epoch": 1.5068643602416256,
+      "grad_norm": 0.07185570150613785,
+      "learning_rate": 0.00016612377850162865,
+      "loss": 0.8213,
+      "step": 686
+    },
+    {
+      "epoch": 1.5090609555189456,
+      "grad_norm": 0.06040317192673683,
+      "learning_rate": 0.00016587947882736157,
+      "loss": 0.6683,
+      "step": 687
+    },
+    {
+      "epoch": 1.5112575507962658,
+      "grad_norm": 0.050129249691963196,
+      "learning_rate": 0.00016563517915309445,
+      "loss": 0.7417,
+      "step": 688
+    },
+    {
+      "epoch": 1.513454146073586,
+      "grad_norm": 0.057126522064208984,
+      "learning_rate": 0.00016539087947882734,
+      "loss": 0.5871,
+      "step": 689
+    },
+    {
+      "epoch": 1.515650741350906,
+      "grad_norm": 0.037670109421014786,
+      "learning_rate": 0.00016514657980456026,
+      "loss": 0.6136,
+      "step": 690
+    },
+    {
+      "epoch": 1.5178473366282264,
+      "grad_norm": 0.048299022018909454,
+      "learning_rate": 0.00016490228013029315,
+      "loss": 0.6504,
+      "step": 691
+    },
+    {
+      "epoch": 1.5200439319055463,
+      "grad_norm": 0.04665425419807434,
+      "learning_rate": 0.00016465798045602604,
+      "loss": 0.6824,
+      "step": 692
+    },
+    {
+      "epoch": 1.5222405271828665,
+      "grad_norm": 0.052132029086351395,
+      "learning_rate": 0.00016441368078175895,
+      "loss": 0.678,
+      "step": 693
+    },
+    {
+      "epoch": 1.5244371224601867,
+      "grad_norm": 0.05157755687832832,
+      "learning_rate": 0.00016416938110749184,
+      "loss": 0.5837,
+      "step": 694
+    },
+    {
+      "epoch": 1.526633717737507,
+      "grad_norm": 0.04299292340874672,
+      "learning_rate": 0.00016392508143322476,
+      "loss": 0.6819,
+      "step": 695
+    },
+    {
+      "epoch": 1.528830313014827,
+      "grad_norm": 0.04690634086728096,
+      "learning_rate": 0.00016368078175895765,
+      "loss": 0.6123,
+      "step": 696
+    },
+    {
+      "epoch": 1.531026908292147,
+      "grad_norm": 0.04270879551768303,
+      "learning_rate": 0.00016343648208469054,
+      "loss": 0.6294,
+      "step": 697
+    },
+    {
+      "epoch": 1.5332235035694675,
+      "grad_norm": 0.038513366132974625,
+      "learning_rate": 0.00016319218241042346,
+      "loss": 0.6058,
+      "step": 698
+    },
+    {
+      "epoch": 1.5354200988467874,
+      "grad_norm": 0.05122312158346176,
+      "learning_rate": 0.00016294788273615634,
+      "loss": 0.6818,
+      "step": 699
+    },
+    {
+      "epoch": 1.5376166941241076,
+      "grad_norm": 0.038340240716934204,
+      "learning_rate": 0.00016270358306188923,
+      "loss": 0.7551,
+      "step": 700
+    },
+    {
+      "epoch": 1.5398132894014278,
+      "grad_norm": 0.04914093762636185,
+      "learning_rate": 0.00016245928338762215,
+      "loss": 0.5806,
+      "step": 701
+    },
+    {
+      "epoch": 1.5420098846787478,
+      "grad_norm": 0.04841599613428116,
+      "learning_rate": 0.00016221498371335504,
+      "loss": 0.6435,
+      "step": 702
+    },
+    {
+      "epoch": 1.5442064799560682,
+      "grad_norm": 0.03922514617443085,
+      "learning_rate": 0.00016197068403908793,
+      "loss": 0.5564,
+      "step": 703
+    },
+    {
+      "epoch": 1.5464030752333882,
+      "grad_norm": 0.06409287452697754,
+      "learning_rate": 0.00016172638436482084,
+      "loss": 0.5631,
+      "step": 704
+    },
+    {
+      "epoch": 1.5485996705107083,
+      "grad_norm": 0.07203608751296997,
+      "learning_rate": 0.00016148208469055373,
+      "loss": 0.6606,
+      "step": 705
+    },
+    {
+      "epoch": 1.5507962657880285,
+      "grad_norm": 0.048426203429698944,
+      "learning_rate": 0.00016123778501628662,
+      "loss": 0.6252,
+      "step": 706
+    },
+    {
+      "epoch": 1.5529928610653487,
+      "grad_norm": 0.054609306156635284,
+      "learning_rate": 0.00016099348534201954,
+      "loss": 0.6719,
+      "step": 707
+    },
+    {
+      "epoch": 1.555189456342669,
+      "grad_norm": 0.08309051394462585,
+      "learning_rate": 0.00016074918566775243,
+      "loss": 0.7236,
+      "step": 708
+    },
+    {
+      "epoch": 1.5573860516199889,
+      "grad_norm": 0.06068706884980202,
+      "learning_rate": 0.00016050488599348534,
+      "loss": 0.7964,
+      "step": 709
+    },
+    {
+      "epoch": 1.5595826468973093,
+      "grad_norm": 0.05096975341439247,
+      "learning_rate": 0.00016026058631921823,
+      "loss": 0.4562,
+      "step": 710
+    },
+    {
+      "epoch": 1.5617792421746293,
+      "grad_norm": 0.04532795771956444,
+      "learning_rate": 0.00016001628664495112,
+      "loss": 0.7656,
+      "step": 711
+    },
+    {
+      "epoch": 1.5639758374519495,
+      "grad_norm": 0.046231381595134735,
+      "learning_rate": 0.00015977198697068404,
+      "loss": 0.5014,
+      "step": 712
+    },
+    {
+      "epoch": 1.5661724327292696,
+      "grad_norm": 0.04625224322080612,
+      "learning_rate": 0.00015952768729641693,
+      "loss": 0.5836,
+      "step": 713
+    },
+    {
+      "epoch": 1.5683690280065898,
+      "grad_norm": 0.03278486803174019,
+      "learning_rate": 0.00015928338762214982,
+      "loss": 0.5181,
+      "step": 714
+    },
+    {
+      "epoch": 1.57056562328391,
+      "grad_norm": 0.0495121143758297,
+      "learning_rate": 0.00015903908794788273,
+      "loss": 0.6985,
+      "step": 715
+    },
+    {
+      "epoch": 1.57276221856123,
+      "grad_norm": 0.06934545189142227,
+      "learning_rate": 0.00015879478827361562,
+      "loss": 0.8638,
+      "step": 716
+    },
+    {
+      "epoch": 1.5749588138385504,
+      "grad_norm": 0.04187803342938423,
+      "learning_rate": 0.0001585504885993485,
+      "loss": 0.6577,
+      "step": 717
+    },
+    {
+      "epoch": 1.5771554091158704,
+      "grad_norm": 0.07038550078868866,
+      "learning_rate": 0.00015830618892508143,
+      "loss": 0.593,
+      "step": 718
+    },
+    {
+      "epoch": 1.5793520043931906,
+      "grad_norm": 0.05784309282898903,
+      "learning_rate": 0.00015806188925081432,
+      "loss": 0.6606,
+      "step": 719
+    },
+    {
+      "epoch": 1.5815485996705108,
+      "grad_norm": 0.050657592713832855,
+      "learning_rate": 0.00015781758957654723,
+      "loss": 0.5663,
+      "step": 720
+    },
+    {
+      "epoch": 1.5837451949478307,
+      "grad_norm": 0.055596090853214264,
+      "learning_rate": 0.00015757328990228012,
+      "loss": 0.7576,
+      "step": 721
+    },
+    {
+      "epoch": 1.5859417902251511,
+      "grad_norm": 0.06155412271618843,
+      "learning_rate": 0.000157328990228013,
+      "loss": 0.8413,
+      "step": 722
+    },
+    {
+      "epoch": 1.588138385502471,
+      "grad_norm": 0.04519607871770859,
+      "learning_rate": 0.00015708469055374593,
+      "loss": 0.6418,
+      "step": 723
+    },
+    {
+      "epoch": 1.5903349807797913,
+      "grad_norm": 0.049259744584560394,
+      "learning_rate": 0.00015684039087947882,
+      "loss": 0.5145,
+      "step": 724
+    },
+    {
+      "epoch": 1.5925315760571115,
+      "grad_norm": 0.04998054727911949,
+      "learning_rate": 0.0001565960912052117,
+      "loss": 0.7043,
+      "step": 725
+    },
+    {
+      "epoch": 1.5947281713344317,
+      "grad_norm": 0.0548371784389019,
+      "learning_rate": 0.00015635179153094462,
+      "loss": 0.6797,
+      "step": 726
+    },
+    {
+      "epoch": 1.5969247666117519,
+      "grad_norm": 0.048441048711538315,
+      "learning_rate": 0.0001561074918566775,
+      "loss": 0.7109,
+      "step": 727
+    },
+    {
+      "epoch": 1.5991213618890718,
+      "grad_norm": 0.04135994240641594,
+      "learning_rate": 0.0001558631921824104,
+      "loss": 0.7135,
+      "step": 728
+    },
+    {
+      "epoch": 1.6013179571663922,
+      "grad_norm": 0.06847979873418808,
+      "learning_rate": 0.00015561889250814332,
+      "loss": 0.6801,
+      "step": 729
+    },
+    {
+      "epoch": 1.6035145524437122,
+      "grad_norm": 0.050964195281267166,
+      "learning_rate": 0.0001553745928338762,
+      "loss": 0.6747,
+      "step": 730
+    },
+    {
+      "epoch": 1.6057111477210324,
+      "grad_norm": 0.04272550716996193,
+      "learning_rate": 0.0001551302931596091,
+      "loss": 0.4606,
+      "step": 731
+    },
+    {
+      "epoch": 1.6079077429983526,
+      "grad_norm": 0.05093299224972725,
+      "learning_rate": 0.000154885993485342,
+      "loss": 0.6401,
+      "step": 732
+    },
+    {
+      "epoch": 1.6101043382756726,
+      "grad_norm": 0.07038379460573196,
+      "learning_rate": 0.0001546416938110749,
+      "loss": 0.8044,
+      "step": 733
+    },
+    {
+      "epoch": 1.612300933552993,
+      "grad_norm": 0.04339474439620972,
+      "learning_rate": 0.00015439739413680782,
+      "loss": 0.6374,
+      "step": 734
+    },
+    {
+      "epoch": 1.614497528830313,
+      "grad_norm": 0.08015977591276169,
+      "learning_rate": 0.0001541530944625407,
+      "loss": 0.7327,
+      "step": 735
+    },
+    {
+      "epoch": 1.6166941241076331,
+      "grad_norm": 0.04767515882849693,
+      "learning_rate": 0.0001539087947882736,
+      "loss": 0.6442,
+      "step": 736
+    },
+    {
+      "epoch": 1.6188907193849533,
+      "grad_norm": 0.04954354092478752,
+      "learning_rate": 0.0001536644951140065,
+      "loss": 0.7242,
+      "step": 737
+    },
+    {
+      "epoch": 1.6210873146622735,
+      "grad_norm": 0.06409866362810135,
+      "learning_rate": 0.0001534201954397394,
+      "loss": 0.6139,
+      "step": 738
+    },
+    {
+      "epoch": 1.6232839099395937,
+      "grad_norm": 0.04829477518796921,
+      "learning_rate": 0.0001531758957654723,
+      "loss": 0.6117,
+      "step": 739
+    },
+    {
+      "epoch": 1.6254805052169137,
+      "grad_norm": 0.05684004724025726,
+      "learning_rate": 0.0001529315960912052,
+      "loss": 0.6686,
+      "step": 740
+    },
+    {
+      "epoch": 1.627677100494234,
+      "grad_norm": 0.040473535656929016,
+      "learning_rate": 0.0001526872964169381,
+      "loss": 0.7073,
+      "step": 741
+    },
+    {
+      "epoch": 1.629873695771554,
+      "grad_norm": 0.05740467086434364,
+      "learning_rate": 0.00015244299674267099,
+      "loss": 0.6241,
+      "step": 742
+    },
+    {
+      "epoch": 1.6320702910488742,
+      "grad_norm": 0.06075328588485718,
+      "learning_rate": 0.0001521986970684039,
+      "loss": 0.587,
+      "step": 743
+    },
+    {
+      "epoch": 1.6342668863261944,
+      "grad_norm": 0.05512019619345665,
+      "learning_rate": 0.0001519543973941368,
+      "loss": 0.503,
+      "step": 744
+    },
+    {
+      "epoch": 1.6364634816035144,
+      "grad_norm": 0.06742412596940994,
+      "learning_rate": 0.00015171009771986968,
+      "loss": 0.6021,
+      "step": 745
+    },
+    {
+      "epoch": 1.6386600768808348,
+      "grad_norm": 0.0417785570025444,
+      "learning_rate": 0.0001514657980456026,
+      "loss": 0.4437,
+      "step": 746
+    },
+    {
+      "epoch": 1.6408566721581548,
+      "grad_norm": 0.08356238156557083,
+      "learning_rate": 0.0001512214983713355,
+      "loss": 0.6415,
+      "step": 747
+    },
+    {
+      "epoch": 1.643053267435475,
+      "grad_norm": 0.04881569743156433,
+      "learning_rate": 0.0001509771986970684,
+      "loss": 0.4988,
+      "step": 748
+    },
+    {
+      "epoch": 1.6452498627127952,
+      "grad_norm": 0.047157756984233856,
+      "learning_rate": 0.0001507328990228013,
+      "loss": 0.502,
+      "step": 749
+    },
+    {
+      "epoch": 1.6474464579901154,
+      "grad_norm": 0.05139964818954468,
+      "learning_rate": 0.00015048859934853418,
+      "loss": 0.591,
+      "step": 750
+    },
+    {
+      "epoch": 1.6496430532674355,
+      "grad_norm": 0.0414644293487072,
+      "learning_rate": 0.0001502442996742671,
+      "loss": 0.6039,
+      "step": 751
+    },
+    {
+      "epoch": 1.6518396485447555,
+      "grad_norm": 0.06458733975887299,
+      "learning_rate": 0.00015,
+      "loss": 0.732,
+      "step": 752
+    },
+    {
+      "epoch": 1.654036243822076,
+      "grad_norm": 0.07901707291603088,
+      "learning_rate": 0.00014975570032573288,
+      "loss": 0.9098,
+      "step": 753
+    },
+    {
+      "epoch": 1.656232839099396,
+      "grad_norm": 0.047911275178194046,
+      "learning_rate": 0.0001495114006514658,
+      "loss": 0.5262,
+      "step": 754
+    },
+    {
+      "epoch": 1.658429434376716,
+      "grad_norm": 0.05376958101987839,
+      "learning_rate": 0.00014926710097719868,
+      "loss": 0.6248,
+      "step": 755
+    },
+    {
+      "epoch": 1.6606260296540363,
+      "grad_norm": 0.03677194193005562,
+      "learning_rate": 0.00014902280130293157,
+      "loss": 0.6117,
+      "step": 756
+    },
+    {
+      "epoch": 1.6628226249313562,
+      "grad_norm": 0.05050716921687126,
+      "learning_rate": 0.0001487785016286645,
+      "loss": 0.7372,
+      "step": 757
+    },
+    {
+      "epoch": 1.6650192202086767,
+      "grad_norm": 0.050005607306957245,
+      "learning_rate": 0.00014853420195439738,
+      "loss": 0.7221,
+      "step": 758
+    },
+    {
+      "epoch": 1.6672158154859966,
+      "grad_norm": 0.04903872683644295,
+      "learning_rate": 0.0001482899022801303,
+      "loss": 0.7553,
+      "step": 759
+    },
+    {
+      "epoch": 1.6694124107633168,
+      "grad_norm": 0.06483814120292664,
+      "learning_rate": 0.00014804560260586318,
+      "loss": 0.5945,
+      "step": 760
+    },
+    {
+      "epoch": 1.671609006040637,
+      "grad_norm": 0.04501137509942055,
+      "learning_rate": 0.00014780130293159607,
+      "loss": 0.6599,
+      "step": 761
+    },
+    {
+      "epoch": 1.6738056013179572,
+      "grad_norm": 0.04700218141078949,
+      "learning_rate": 0.000147557003257329,
+      "loss": 0.6844,
+      "step": 762
+    },
+    {
+      "epoch": 1.6760021965952774,
+      "grad_norm": 0.04598642885684967,
+      "learning_rate": 0.00014731270358306188,
+      "loss": 0.662,
+      "step": 763
+    },
+    {
+      "epoch": 1.6781987918725974,
+      "grad_norm": 0.0409390851855278,
+      "learning_rate": 0.00014706840390879477,
+      "loss": 0.7319,
+      "step": 764
+    },
+    {
+      "epoch": 1.6803953871499178,
+      "grad_norm": 0.06088387221097946,
+      "learning_rate": 0.00014682410423452768,
+      "loss": 0.6724,
+      "step": 765
+    },
+    {
+      "epoch": 1.6825919824272377,
+      "grad_norm": 0.04854418337345123,
+      "learning_rate": 0.00014657980456026057,
+      "loss": 0.6068,
+      "step": 766
+    },
+    {
+      "epoch": 1.684788577704558,
+      "grad_norm": 0.049988895654678345,
+      "learning_rate": 0.00014633550488599346,
+      "loss": 0.6674,
+      "step": 767
+    },
+    {
+      "epoch": 1.6869851729818781,
+      "grad_norm": 0.07750127464532852,
+      "learning_rate": 0.00014609120521172638,
+      "loss": 0.6047,
+      "step": 768
+    },
+    {
+      "epoch": 1.689181768259198,
+      "grad_norm": 0.04291775822639465,
+      "learning_rate": 0.00014584690553745927,
+      "loss": 0.6315,
+      "step": 769
+    },
+    {
+      "epoch": 1.6913783635365185,
+      "grad_norm": 0.04422176256775856,
+      "learning_rate": 0.00014560260586319216,
+      "loss": 0.6874,
+      "step": 770
+    },
+    {
+      "epoch": 1.6935749588138385,
+      "grad_norm": 0.049668315798044205,
+      "learning_rate": 0.00014535830618892507,
+      "loss": 0.8245,
+      "step": 771
+    },
+    {
+      "epoch": 1.6957715540911587,
+      "grad_norm": 0.051815249025821686,
+      "learning_rate": 0.00014511400651465796,
+      "loss": 0.7336,
+      "step": 772
+    },
+    {
+      "epoch": 1.6979681493684788,
+      "grad_norm": 0.04109729453921318,
+      "learning_rate": 0.00014486970684039088,
+      "loss": 0.6031,
+      "step": 773
+    },
+    {
+      "epoch": 1.700164744645799,
+      "grad_norm": 0.0370013527572155,
+      "learning_rate": 0.00014462540716612377,
+      "loss": 0.5536,
+      "step": 774
+    },
+    {
+      "epoch": 1.7023613399231192,
+      "grad_norm": 0.052517328411340714,
+      "learning_rate": 0.00014438110749185666,
+      "loss": 0.6791,
+      "step": 775
+    },
+    {
+      "epoch": 1.7045579352004392,
+      "grad_norm": 0.04123876616358757,
+      "learning_rate": 0.00014413680781758957,
+      "loss": 0.4564,
+      "step": 776
+    },
+    {
+      "epoch": 1.7067545304777596,
+      "grad_norm": 0.049101028591394424,
+      "learning_rate": 0.00014389250814332246,
+      "loss": 0.73,
+      "step": 777
+    },
+    {
+      "epoch": 1.7089511257550796,
+      "grad_norm": 0.0485854297876358,
+      "learning_rate": 0.00014364820846905535,
+      "loss": 0.729,
+      "step": 778
+    },
+    {
+      "epoch": 1.7111477210323998,
+      "grad_norm": 0.0474361851811409,
+      "learning_rate": 0.00014340390879478827,
+      "loss": 0.6365,
+      "step": 779
+    },
+    {
+      "epoch": 1.71334431630972,
+      "grad_norm": 0.05722922831773758,
+      "learning_rate": 0.00014315960912052116,
+      "loss": 0.6634,
+      "step": 780
+    },
+    {
+      "epoch": 1.7155409115870401,
+      "grad_norm": 0.05596005171537399,
+      "learning_rate": 0.00014291530944625404,
+      "loss": 0.6379,
+      "step": 781
+    },
+    {
+      "epoch": 1.7177375068643603,
+      "grad_norm": 0.06430881470441818,
+      "learning_rate": 0.00014267100977198696,
+      "loss": 0.5975,
+      "step": 782
+    },
+    {
+      "epoch": 1.7199341021416803,
+      "grad_norm": 0.041797831654548645,
+      "learning_rate": 0.00014242671009771985,
+      "loss": 0.6402,
+      "step": 783
+    },
+    {
+      "epoch": 1.7221306974190007,
+      "grad_norm": 0.04201621934771538,
+      "learning_rate": 0.00014218241042345274,
+      "loss": 0.6083,
+      "step": 784
+    },
+    {
+      "epoch": 1.7243272926963207,
+      "grad_norm": 0.0454474575817585,
+      "learning_rate": 0.00014193811074918566,
+      "loss": 0.856,
+      "step": 785
+    },
+    {
+      "epoch": 1.7265238879736409,
+      "grad_norm": 0.03413139283657074,
+      "learning_rate": 0.00014169381107491854,
+      "loss": 0.5574,
+      "step": 786
+    },
+    {
+      "epoch": 1.728720483250961,
+      "grad_norm": 0.05276301130652428,
+      "learning_rate": 0.00014144951140065146,
+      "loss": 0.5608,
+      "step": 787
+    },
+    {
+      "epoch": 1.730917078528281,
+      "grad_norm": 0.04063792899250984,
+      "learning_rate": 0.00014120521172638435,
+      "loss": 0.648,
+      "step": 788
+    },
+    {
+      "epoch": 1.7331136738056014,
+      "grad_norm": 0.04240501672029495,
+      "learning_rate": 0.00014096091205211724,
+      "loss": 0.6826,
+      "step": 789
+    },
+    {
+      "epoch": 1.7353102690829214,
+      "grad_norm": 0.0462966114282608,
+      "learning_rate": 0.00014071661237785016,
+      "loss": 0.8253,
+      "step": 790
+    },
+    {
+      "epoch": 1.7375068643602416,
+      "grad_norm": 0.06632352620363235,
+      "learning_rate": 0.00014047231270358305,
+      "loss": 0.6522,
+      "step": 791
+    },
+    {
+      "epoch": 1.7397034596375618,
+      "grad_norm": 0.0381113663315773,
+      "learning_rate": 0.00014022801302931593,
+      "loss": 0.5937,
+      "step": 792
+    },
+    {
+      "epoch": 1.741900054914882,
+      "grad_norm": 0.056441642343997955,
+      "learning_rate": 0.00013998371335504885,
+      "loss": 0.61,
+      "step": 793
+    },
+    {
+      "epoch": 1.7440966501922022,
+      "grad_norm": 0.0543392077088356,
+      "learning_rate": 0.00013973941368078174,
+      "loss": 0.7435,
+      "step": 794
+    },
+    {
+      "epoch": 1.7462932454695221,
+      "grad_norm": 0.03827011212706566,
+      "learning_rate": 0.00013949511400651463,
+      "loss": 0.5941,
+      "step": 795
+    },
+    {
+      "epoch": 1.7484898407468425,
+      "grad_norm": 0.060338567942380905,
+      "learning_rate": 0.00013925081433224755,
+      "loss": 0.582,
+      "step": 796
+    },
+    {
+      "epoch": 1.7506864360241625,
+      "grad_norm": 0.03799464553594589,
+      "learning_rate": 0.00013900651465798043,
+      "loss": 0.6215,
+      "step": 797
+    },
+    {
+      "epoch": 1.7528830313014827,
+      "grad_norm": 0.043356601148843765,
+      "learning_rate": 0.00013876221498371335,
+      "loss": 0.6715,
+      "step": 798
+    },
+    {
+      "epoch": 1.755079626578803,
+      "grad_norm": 0.08416961133480072,
+      "learning_rate": 0.00013851791530944624,
+      "loss": 0.7502,
+      "step": 799
+    },
+    {
+      "epoch": 1.7572762218561229,
+      "grad_norm": 0.05539776757359505,
+      "learning_rate": 0.00013827361563517913,
+      "loss": 0.8666,
+      "step": 800
+    },
+    {
+      "epoch": 1.7594728171334433,
+      "grad_norm": 0.05246887728571892,
+      "learning_rate": 0.00013802931596091205,
+      "loss": 0.7886,
+      "step": 801
+    },
+    {
+      "epoch": 1.7616694124107632,
+      "grad_norm": 0.07970348745584488,
+      "learning_rate": 0.00013778501628664493,
+      "loss": 0.8519,
+      "step": 802
+    },
+    {
+      "epoch": 1.7638660076880834,
+      "grad_norm": 0.07714508473873138,
+      "learning_rate": 0.00013754071661237782,
+      "loss": 0.9883,
+      "step": 803
+    },
+    {
+      "epoch": 1.7660626029654036,
+      "grad_norm": 0.055044207721948624,
+      "learning_rate": 0.00013729641693811074,
+      "loss": 0.8199,
+      "step": 804
+    },
+    {
+      "epoch": 1.7682591982427238,
+      "grad_norm": 0.05030933395028114,
+      "learning_rate": 0.00013705211726384363,
+      "loss": 0.6976,
+      "step": 805
+    },
+    {
+      "epoch": 1.770455793520044,
+      "grad_norm": 0.048857420682907104,
+      "learning_rate": 0.00013680781758957652,
+      "loss": 0.6373,
+      "step": 806
+    },
+    {
+      "epoch": 1.772652388797364,
+      "grad_norm": 0.05215909704566002,
+      "learning_rate": 0.00013656351791530943,
+      "loss": 0.6757,
+      "step": 807
+    },
+    {
+      "epoch": 1.7748489840746844,
+      "grad_norm": 0.07715169340372086,
+      "learning_rate": 0.00013631921824104232,
+      "loss": 0.752,
+      "step": 808
+    },
+    {
+      "epoch": 1.7770455793520044,
+      "grad_norm": 0.058291271328926086,
+      "learning_rate": 0.0001360749185667752,
+      "loss": 0.7419,
+      "step": 809
+    },
+    {
+      "epoch": 1.7792421746293245,
+      "grad_norm": 0.05078558251261711,
+      "learning_rate": 0.00013583061889250813,
+      "loss": 0.4623,
+      "step": 810
+    },
+    {
+      "epoch": 1.7814387699066447,
+      "grad_norm": 0.0427047535777092,
+      "learning_rate": 0.00013558631921824102,
+      "loss": 0.536,
+      "step": 811
+    },
+    {
+      "epoch": 1.7836353651839647,
+      "grad_norm": 0.055373404175043106,
+      "learning_rate": 0.00013534201954397394,
+      "loss": 0.6415,
+      "step": 812
+    },
+    {
+      "epoch": 1.7858319604612851,
+      "grad_norm": 0.05837016552686691,
+      "learning_rate": 0.00013509771986970682,
+      "loss": 0.7953,
+      "step": 813
+    },
+    {
+      "epoch": 1.788028555738605,
+      "grad_norm": 0.05480782315135002,
+      "learning_rate": 0.00013485342019543971,
+      "loss": 0.6892,
+      "step": 814
+    },
+    {
+      "epoch": 1.7902251510159253,
+      "grad_norm": 0.04352695494890213,
+      "learning_rate": 0.00013460912052117263,
+      "loss": 0.6963,
+      "step": 815
+    },
+    {
+      "epoch": 1.7924217462932455,
+      "grad_norm": 0.048882078379392624,
+      "learning_rate": 0.00013436482084690552,
+      "loss": 0.6821,
+      "step": 816
+    },
+    {
+      "epoch": 1.7946183415705657,
+      "grad_norm": 0.04737727344036102,
+      "learning_rate": 0.0001341205211726384,
+      "loss": 0.5867,
+      "step": 817
+    },
+    {
+      "epoch": 1.7968149368478858,
+      "grad_norm": 0.048038333654403687,
+      "learning_rate": 0.00013387622149837132,
+      "loss": 0.6541,
+      "step": 818
+    },
+    {
+      "epoch": 1.7990115321252058,
+      "grad_norm": 0.04460786655545235,
+      "learning_rate": 0.00013363192182410421,
+      "loss": 0.6313,
+      "step": 819
+    },
+    {
+      "epoch": 1.8012081274025262,
+      "grad_norm": 0.05005588009953499,
+      "learning_rate": 0.0001333876221498371,
+      "loss": 0.6801,
+      "step": 820
+    },
+    {
+      "epoch": 1.8034047226798462,
+      "grad_norm": 0.04483172670006752,
+      "learning_rate": 0.00013314332247557002,
+      "loss": 0.7252,
+      "step": 821
+    },
+    {
+      "epoch": 1.8056013179571664,
+      "grad_norm": 0.05668759346008301,
+      "learning_rate": 0.0001328990228013029,
+      "loss": 0.7627,
+      "step": 822
+    },
+    {
+      "epoch": 1.8077979132344866,
+      "grad_norm": 0.04755188897252083,
+      "learning_rate": 0.00013265472312703582,
+      "loss": 0.6951,
+      "step": 823
+    },
+    {
+      "epoch": 1.8099945085118065,
+      "grad_norm": 0.04183987155556679,
+      "learning_rate": 0.00013241042345276871,
+      "loss": 0.6022,
+      "step": 824
+    },
+    {
+      "epoch": 1.812191103789127,
+      "grad_norm": 0.0463097020983696,
+      "learning_rate": 0.0001321661237785016,
+      "loss": 0.6062,
+      "step": 825
+    },
+    {
+      "epoch": 1.814387699066447,
+      "grad_norm": 0.04682566225528717,
+      "learning_rate": 0.00013192182410423452,
+      "loss": 0.4859,
+      "step": 826
+    },
+    {
+      "epoch": 1.8165842943437671,
+      "grad_norm": 0.05168147012591362,
+      "learning_rate": 0.0001316775244299674,
+      "loss": 0.7072,
+      "step": 827
+    },
+    {
+      "epoch": 1.8187808896210873,
+      "grad_norm": 0.04258091375231743,
+      "learning_rate": 0.0001314332247557003,
+      "loss": 0.6083,
+      "step": 828
+    },
+    {
+      "epoch": 1.8209774848984075,
+      "grad_norm": 0.04315861314535141,
+      "learning_rate": 0.00013118892508143321,
+      "loss": 0.5703,
+      "step": 829
+    },
+    {
+      "epoch": 1.8231740801757277,
+      "grad_norm": 0.05403318628668785,
+      "learning_rate": 0.00013094462540716613,
+      "loss": 0.5154,
+      "step": 830
+    },
+    {
+      "epoch": 1.8253706754530477,
+      "grad_norm": 0.03748399391770363,
+      "learning_rate": 0.00013070032573289902,
+      "loss": 0.4465,
+      "step": 831
+    },
+    {
+      "epoch": 1.827567270730368,
+      "grad_norm": 0.04764214903116226,
+      "learning_rate": 0.0001304560260586319,
+      "loss": 0.6787,
+      "step": 832
+    },
+    {
+      "epoch": 1.829763866007688,
+      "grad_norm": 0.0467853844165802,
+      "learning_rate": 0.00013021172638436483,
+      "loss": 0.6927,
+      "step": 833
+    },
+    {
+      "epoch": 1.8319604612850082,
+      "grad_norm": 0.049767203629016876,
+      "learning_rate": 0.00012996742671009771,
+      "loss": 0.7413,
+      "step": 834
+    },
+    {
+      "epoch": 1.8341570565623284,
+      "grad_norm": 0.049734946340322495,
+      "learning_rate": 0.0001297231270358306,
+      "loss": 0.7474,
+      "step": 835
+    },
+    {
+      "epoch": 1.8363536518396484,
+      "grad_norm": 0.05754285678267479,
+      "learning_rate": 0.00012947882736156352,
+      "loss": 0.5436,
+      "step": 836
+    },
+    {
+      "epoch": 1.8385502471169688,
+      "grad_norm": 0.08427241444587708,
+      "learning_rate": 0.0001292345276872964,
+      "loss": 0.6116,
+      "step": 837
+    },
+    {
+      "epoch": 1.8407468423942888,
+      "grad_norm": 0.051138222217559814,
+      "learning_rate": 0.00012899022801302933,
+      "loss": 0.6005,
+      "step": 838
+    },
+    {
+      "epoch": 1.842943437671609,
+      "grad_norm": 0.0441889688372612,
+      "learning_rate": 0.00012874592833876221,
+      "loss": 0.5388,
+      "step": 839
+    },
+    {
+      "epoch": 1.8451400329489291,
+      "grad_norm": 0.057013627141714096,
+      "learning_rate": 0.0001285016286644951,
+      "loss": 0.7047,
+      "step": 840
+    },
+    {
+      "epoch": 1.8473366282262493,
+      "grad_norm": 0.04341554641723633,
+      "learning_rate": 0.00012825732899022802,
+      "loss": 0.636,
+      "step": 841
+    },
+    {
+      "epoch": 1.8495332235035695,
+      "grad_norm": 0.05924821272492409,
+      "learning_rate": 0.0001280130293159609,
+      "loss": 0.6177,
+      "step": 842
+    },
+    {
+      "epoch": 1.8517298187808895,
+      "grad_norm": 0.05318041145801544,
+      "learning_rate": 0.0001277687296416938,
+      "loss": 0.8203,
+      "step": 843
+    },
+    {
+      "epoch": 1.85392641405821,
+      "grad_norm": 0.062347158789634705,
+      "learning_rate": 0.00012752442996742671,
+      "loss": 0.7046,
+      "step": 844
+    },
+    {
+      "epoch": 1.8561230093355299,
+      "grad_norm": 0.053158268332481384,
+      "learning_rate": 0.0001272801302931596,
+      "loss": 0.6758,
+      "step": 845
+    },
+    {
+      "epoch": 1.85831960461285,
+      "grad_norm": 0.04612778499722481,
+      "learning_rate": 0.0001270358306188925,
+      "loss": 0.5712,
+      "step": 846
+    },
+    {
+      "epoch": 1.8605161998901703,
+      "grad_norm": 0.05728829279541969,
+      "learning_rate": 0.0001267915309446254,
+      "loss": 0.7674,
+      "step": 847
+    },
+    {
+      "epoch": 1.8627127951674904,
+      "grad_norm": 0.046414975076913834,
+      "learning_rate": 0.0001265472312703583,
+      "loss": 0.5906,
+      "step": 848
+    },
+    {
+      "epoch": 1.8649093904448106,
+      "grad_norm": 0.04038268327713013,
+      "learning_rate": 0.0001263029315960912,
+      "loss": 0.5586,
+      "step": 849
+    },
+    {
+      "epoch": 1.8671059857221306,
+      "grad_norm": 0.03641192987561226,
+      "learning_rate": 0.0001260586319218241,
+      "loss": 0.6012,
+      "step": 850
+    },
+    {
+      "epoch": 1.869302580999451,
+      "grad_norm": 0.03718649223446846,
+      "learning_rate": 0.000125814332247557,
+      "loss": 0.5135,
+      "step": 851
+    },
+    {
+      "epoch": 1.871499176276771,
+      "grad_norm": 0.03840146213769913,
+      "learning_rate": 0.0001255700325732899,
+      "loss": 0.6627,
+      "step": 852
+    },
+    {
+      "epoch": 1.8736957715540912,
+      "grad_norm": 0.044655878096818924,
+      "learning_rate": 0.0001253257328990228,
+      "loss": 0.6727,
+      "step": 853
+    },
+    {
+      "epoch": 1.8758923668314114,
+      "grad_norm": 0.05583556368947029,
+      "learning_rate": 0.0001250814332247557,
+      "loss": 0.8199,
+      "step": 854
+    },
+    {
+      "epoch": 1.8780889621087313,
+      "grad_norm": 0.03336760401725769,
+      "learning_rate": 0.0001248371335504886,
+      "loss": 0.5403,
+      "step": 855
+    },
+    {
+      "epoch": 1.8802855573860517,
+      "grad_norm": 0.05149347707629204,
+      "learning_rate": 0.0001245928338762215,
+      "loss": 0.7947,
+      "step": 856
+    },
+    {
+      "epoch": 1.8824821526633717,
+      "grad_norm": 0.04355181008577347,
+      "learning_rate": 0.00012434853420195438,
+      "loss": 0.6353,
+      "step": 857
+    },
+    {
+      "epoch": 1.884678747940692,
+      "grad_norm": 0.04656535014510155,
+      "learning_rate": 0.0001241042345276873,
+      "loss": 0.7081,
+      "step": 858
+    },
+    {
+      "epoch": 1.886875343218012,
+      "grad_norm": 0.04054791107773781,
+      "learning_rate": 0.0001238599348534202,
+      "loss": 0.592,
+      "step": 859
+    },
+    {
+      "epoch": 1.8890719384953323,
+      "grad_norm": 0.034135375171899796,
+      "learning_rate": 0.00012361563517915308,
+      "loss": 0.5938,
+      "step": 860
+    },
+    {
+      "epoch": 1.8912685337726525,
+      "grad_norm": 0.075068399310112,
+      "learning_rate": 0.000123371335504886,
+      "loss": 0.8108,
+      "step": 861
+    },
+    {
+      "epoch": 1.8934651290499724,
+      "grad_norm": 0.04905351996421814,
+      "learning_rate": 0.00012312703583061888,
+      "loss": 0.657,
+      "step": 862
+    },
+    {
+      "epoch": 1.8956617243272929,
+      "grad_norm": 0.047971729189157486,
+      "learning_rate": 0.00012288273615635177,
+      "loss": 0.8336,
+      "step": 863
+    },
+    {
+      "epoch": 1.8978583196046128,
+      "grad_norm": 0.04546576365828514,
+      "learning_rate": 0.0001226384364820847,
+      "loss": 0.5656,
+      "step": 864
+    },
+    {
+      "epoch": 1.900054914881933,
+      "grad_norm": 0.048822496086359024,
+      "learning_rate": 0.00012239413680781758,
+      "loss": 0.7217,
+      "step": 865
+    },
+    {
+      "epoch": 1.9022515101592532,
+      "grad_norm": 0.0437794029712677,
+      "learning_rate": 0.0001221498371335505,
+      "loss": 0.6082,
+      "step": 866
+    },
+    {
+      "epoch": 1.9044481054365732,
+      "grad_norm": 0.05672450736165047,
+      "learning_rate": 0.00012190553745928338,
+      "loss": 0.6227,
+      "step": 867
+    },
+    {
+      "epoch": 1.9066447007138936,
+      "grad_norm": 0.07971169799566269,
+      "learning_rate": 0.00012166123778501629,
+      "loss": 0.8374,
+      "step": 868
+    },
+    {
+      "epoch": 1.9088412959912135,
+      "grad_norm": 0.04980779439210892,
+      "learning_rate": 0.00012141693811074918,
+      "loss": 0.7397,
+      "step": 869
+    },
+    {
+      "epoch": 1.9110378912685337,
+      "grad_norm": 0.03777102008461952,
+      "learning_rate": 0.00012117263843648208,
+      "loss": 0.6381,
+      "step": 870
+    },
+    {
+      "epoch": 1.913234486545854,
+      "grad_norm": 0.049683138728141785,
+      "learning_rate": 0.00012092833876221498,
+      "loss": 0.6401,
+      "step": 871
+    },
+    {
+      "epoch": 1.9154310818231741,
+      "grad_norm": 0.03963594511151314,
+      "learning_rate": 0.00012068403908794787,
+      "loss": 0.7263,
+      "step": 872
+    },
+    {
+      "epoch": 1.9176276771004943,
+      "grad_norm": 0.05527025833725929,
+      "learning_rate": 0.00012043973941368077,
+      "loss": 0.7111,
+      "step": 873
+    },
+    {
+      "epoch": 1.9198242723778143,
+      "grad_norm": 0.04393410310149193,
+      "learning_rate": 0.00012019543973941368,
+      "loss": 0.8593,
+      "step": 874
+    },
+    {
+      "epoch": 1.9220208676551347,
+      "grad_norm": 0.062434881925582886,
+      "learning_rate": 0.00011995114006514658,
+      "loss": 0.7676,
+      "step": 875
+    },
+    {
+      "epoch": 1.9242174629324547,
+      "grad_norm": 0.07280800491571426,
+      "learning_rate": 0.00011970684039087947,
+      "loss": 0.7656,
+      "step": 876
+    },
+    {
+      "epoch": 1.9264140582097748,
+      "grad_norm": 0.05514119192957878,
+      "learning_rate": 0.00011946254071661237,
+      "loss": 0.5903,
+      "step": 877
+    },
+    {
+      "epoch": 1.928610653487095,
+      "grad_norm": 0.06919682025909424,
+      "learning_rate": 0.00011921824104234527,
+      "loss": 0.72,
+      "step": 878
+    },
+    {
+      "epoch": 1.930807248764415,
+      "grad_norm": 0.04627091437578201,
+      "learning_rate": 0.00011897394136807816,
+      "loss": 0.6976,
+      "step": 879
+    },
+    {
+      "epoch": 1.9330038440417354,
+      "grad_norm": 0.05157488211989403,
+      "learning_rate": 0.00011872964169381106,
+      "loss": 0.5897,
+      "step": 880
+    },
+    {
+      "epoch": 1.9352004393190554,
+      "grad_norm": 0.05058778077363968,
+      "learning_rate": 0.00011848534201954397,
+      "loss": 0.7003,
+      "step": 881
+    },
+    {
+      "epoch": 1.9373970345963756,
+      "grad_norm": 0.03946829214692116,
+      "learning_rate": 0.00011824104234527687,
+      "loss": 0.5076,
+      "step": 882
+    },
+    {
+      "epoch": 1.9395936298736958,
+      "grad_norm": 0.04601827263832092,
+      "learning_rate": 0.00011799674267100976,
+      "loss": 0.7444,
+      "step": 883
+    },
+    {
+      "epoch": 1.941790225151016,
+      "grad_norm": 0.05405599623918533,
+      "learning_rate": 0.00011775244299674266,
+      "loss": 0.7075,
+      "step": 884
+    },
+    {
+      "epoch": 1.9439868204283361,
+      "grad_norm": 0.04909859970211983,
+      "learning_rate": 0.00011750814332247557,
+      "loss": 1.0181,
+      "step": 885
+    },
+    {
+      "epoch": 1.9461834157056561,
+      "grad_norm": 0.055644210427999496,
+      "learning_rate": 0.00011726384364820845,
+      "loss": 0.7143,
+      "step": 886
+    },
+    {
+      "epoch": 1.9483800109829765,
+      "grad_norm": 0.033494554460048676,
+      "learning_rate": 0.00011701954397394136,
+      "loss": 0.5905,
+      "step": 887
+    },
+    {
+      "epoch": 1.9505766062602965,
+      "grad_norm": 0.048672039061784744,
+      "learning_rate": 0.00011677524429967426,
+      "loss": 0.5846,
+      "step": 888
+    },
+    {
+      "epoch": 1.9527732015376167,
+      "grad_norm": 0.052615948021411896,
+      "learning_rate": 0.00011653094462540716,
+      "loss": 0.7174,
+      "step": 889
+    },
+    {
+      "epoch": 1.9549697968149369,
+      "grad_norm": 0.04827093333005905,
+      "learning_rate": 0.00011628664495114005,
+      "loss": 0.6304,
+      "step": 890
+    },
+    {
+      "epoch": 1.9571663920922568,
+      "grad_norm": 0.042970217764377594,
+      "learning_rate": 0.00011604234527687295,
+      "loss": 0.7678,
+      "step": 891
+    },
+    {
+      "epoch": 1.9593629873695773,
+      "grad_norm": 0.04997050389647484,
+      "learning_rate": 0.00011579804560260586,
+      "loss": 0.6019,
+      "step": 892
+    },
+    {
+      "epoch": 1.9615595826468972,
+      "grad_norm": 0.04948752373456955,
+      "learning_rate": 0.00011555374592833876,
+      "loss": 0.6425,
+      "step": 893
+    },
+    {
+      "epoch": 1.9637561779242174,
+      "grad_norm": 0.04462786018848419,
+      "learning_rate": 0.00011530944625407165,
+      "loss": 0.7068,
+      "step": 894
+    },
+    {
+      "epoch": 1.9659527732015376,
+      "grad_norm": 0.051252443343400955,
+      "learning_rate": 0.00011506514657980455,
+      "loss": 0.5193,
+      "step": 895
+    },
+    {
+      "epoch": 1.9681493684788578,
+      "grad_norm": 0.045937035232782364,
+      "learning_rate": 0.00011482084690553745,
+      "loss": 0.7832,
+      "step": 896
+    },
+    {
+      "epoch": 1.970345963756178,
+      "grad_norm": 0.04210161417722702,
+      "learning_rate": 0.00011457654723127034,
+      "loss": 0.5735,
+      "step": 897
+    },
+    {
+      "epoch": 1.972542559033498,
+      "grad_norm": 0.04671672359108925,
+      "learning_rate": 0.00011433224755700325,
+      "loss": 0.7834,
+      "step": 898
+    },
+    {
+      "epoch": 1.9747391543108184,
+      "grad_norm": 0.06232306733727455,
+      "learning_rate": 0.00011408794788273615,
+      "loss": 0.7376,
+      "step": 899
+    },
+    {
+      "epoch": 1.9769357495881383,
+      "grad_norm": 0.05481863394379616,
+      "learning_rate": 0.00011384364820846905,
+      "loss": 0.5515,
+      "step": 900
+    },
+    {
+      "epoch": 1.9791323448654585,
+      "grad_norm": 0.04274000972509384,
+      "learning_rate": 0.00011359934853420194,
+      "loss": 0.7249,
+      "step": 901
+    },
+    {
+      "epoch": 1.9813289401427787,
+      "grad_norm": 0.05033233016729355,
+      "learning_rate": 0.00011335504885993484,
+      "loss": 0.7522,
+      "step": 902
+    },
+    {
+      "epoch": 1.9835255354200987,
+      "grad_norm": 0.056208908557891846,
+      "learning_rate": 0.00011311074918566775,
+      "loss": 0.6917,
+      "step": 903
+    },
+    {
+      "epoch": 1.985722130697419,
+      "grad_norm": 0.039569608867168427,
+      "learning_rate": 0.00011286644951140064,
+      "loss": 0.5706,
+      "step": 904
+    },
+    {
+      "epoch": 1.987918725974739,
+      "grad_norm": 0.061526212841272354,
+      "learning_rate": 0.00011262214983713354,
+      "loss": 0.5911,
+      "step": 905
+    },
+    {
+      "epoch": 1.9901153212520593,
+      "grad_norm": 0.04074925556778908,
+      "learning_rate": 0.00011237785016286644,
+      "loss": 0.7588,
+      "step": 906
+    },
+    {
+      "epoch": 1.9923119165293794,
+      "grad_norm": 0.0786409005522728,
+      "learning_rate": 0.00011213355048859934,
+      "loss": 0.5221,
+      "step": 907
+    },
+    {
+      "epoch": 1.9945085118066996,
+      "grad_norm": 0.03439033031463623,
+      "learning_rate": 0.00011188925081433223,
+      "loss": 0.5648,
+      "step": 908
+    },
+    {
+      "epoch": 1.9967051070840198,
+      "grad_norm": 0.03860178589820862,
+      "learning_rate": 0.00011164495114006514,
+      "loss": 0.4791,
+      "step": 909
+    },
+    {
+      "epoch": 1.9989017023613398,
+      "grad_norm": 0.09322784096002579,
+      "learning_rate": 0.00011140065146579804,
+      "loss": 0.6842,
+      "step": 910
+    },
+    {
+      "epoch": 2.00109829763866,
+      "grad_norm": 0.05441444739699364,
+      "learning_rate": 0.00011115635179153093,
+      "loss": 0.774,
+      "step": 911
+    },
+    {
+      "epoch": 2.00329489291598,
+      "grad_norm": 0.052353233098983765,
+      "learning_rate": 0.00011091205211726383,
+      "loss": 0.6595,
+      "step": 912
+    },
+    {
+      "epoch": 2.0054914881933006,
+      "grad_norm": 0.046645645052194595,
+      "learning_rate": 0.00011066775244299673,
+      "loss": 0.5674,
+      "step": 913
+    },
+    {
+      "epoch": 2.0076880834706206,
+      "grad_norm": 0.052380114793777466,
+      "learning_rate": 0.00011042345276872964,
+      "loss": 0.6832,
+      "step": 914
+    },
+    {
+      "epoch": 2.0098846787479405,
+      "grad_norm": 0.05334731563925743,
+      "learning_rate": 0.00011017915309446253,
+      "loss": 0.6386,
+      "step": 915
+    },
+    {
+      "epoch": 2.012081274025261,
+      "grad_norm": 0.04087246209383011,
+      "learning_rate": 0.00010993485342019543,
+      "loss": 0.7986,
+      "step": 916
+    },
+    {
+      "epoch": 2.014277869302581,
+      "grad_norm": 0.05971505120396614,
+      "learning_rate": 0.00010969055374592833,
+      "loss": 0.8546,
+      "step": 917
+    },
+    {
+      "epoch": 2.0164744645799013,
+      "grad_norm": 0.04933095723390579,
+      "learning_rate": 0.00010944625407166122,
+      "loss": 0.7736,
+      "step": 918
+    },
+    {
+      "epoch": 2.0186710598572213,
+      "grad_norm": 0.046272121369838715,
+      "learning_rate": 0.00010920195439739412,
+      "loss": 0.629,
+      "step": 919
+    },
+    {
+      "epoch": 2.0208676551345413,
+      "grad_norm": 0.04796759784221649,
+      "learning_rate": 0.00010895765472312703,
+      "loss": 0.6354,
+      "step": 920
+    },
+    {
+      "epoch": 2.0230642504118617,
+      "grad_norm": 0.05017026141285896,
+      "learning_rate": 0.00010871335504885993,
+      "loss": 0.5821,
+      "step": 921
+    },
+    {
+      "epoch": 2.0252608456891816,
+      "grad_norm": 0.04263904318213463,
+      "learning_rate": 0.00010846905537459282,
+      "loss": 0.6971,
+      "step": 922
+    },
+    {
+      "epoch": 2.027457440966502,
+      "grad_norm": 0.05482470244169235,
+      "learning_rate": 0.00010822475570032572,
+      "loss": 0.6048,
+      "step": 923
+    },
+    {
+      "epoch": 2.029654036243822,
+      "grad_norm": 0.05933058261871338,
+      "learning_rate": 0.00010798045602605862,
+      "loss": 0.8025,
+      "step": 924
+    },
+    {
+      "epoch": 2.0318506315211424,
+      "grad_norm": 0.057402413338422775,
+      "learning_rate": 0.00010773615635179153,
+      "loss": 0.6536,
+      "step": 925
+    },
+    {
+      "epoch": 2.0340472267984624,
+      "grad_norm": 0.04043892025947571,
+      "learning_rate": 0.00010749185667752442,
+      "loss": 0.6329,
+      "step": 926
+    },
+    {
+      "epoch": 2.0362438220757824,
+      "grad_norm": 0.04894665256142616,
+      "learning_rate": 0.00010724755700325732,
+      "loss": 0.7953,
+      "step": 927
+    },
+    {
+      "epoch": 2.0384404173531028,
+      "grad_norm": 0.04023706912994385,
+      "learning_rate": 0.00010700325732899022,
+      "loss": 0.6657,
+      "step": 928
+    },
+    {
+      "epoch": 2.0406370126304227,
+      "grad_norm": 0.04153961315751076,
+      "learning_rate": 0.00010675895765472311,
+      "loss": 0.5859,
+      "step": 929
+    },
+    {
+      "epoch": 2.042833607907743,
+      "grad_norm": 0.034479089081287384,
+      "learning_rate": 0.00010651465798045601,
+      "loss": 0.5664,
+      "step": 930
+    },
+    {
+      "epoch": 2.045030203185063,
+      "grad_norm": 0.04258917272090912,
+      "learning_rate": 0.00010627035830618892,
+      "loss": 0.6099,
+      "step": 931
+    },
+    {
+      "epoch": 2.047226798462383,
+      "grad_norm": 0.06230498105287552,
+      "learning_rate": 0.00010602605863192182,
+      "loss": 0.6836,
+      "step": 932
+    },
+    {
+      "epoch": 2.0494233937397035,
+      "grad_norm": 0.0505065955221653,
+      "learning_rate": 0.00010578175895765471,
+      "loss": 0.699,
+      "step": 933
+    },
+    {
+      "epoch": 2.0516199890170235,
+      "grad_norm": 0.04756687209010124,
+      "learning_rate": 0.00010553745928338761,
+      "loss": 0.5818,
+      "step": 934
+    },
+    {
+      "epoch": 2.053816584294344,
+      "grad_norm": 0.03504083678126335,
+      "learning_rate": 0.00010529315960912051,
+      "loss": 0.5922,
+      "step": 935
+    },
+    {
+      "epoch": 2.056013179571664,
+      "grad_norm": 0.04338408634066582,
+      "learning_rate": 0.0001050488599348534,
+      "loss": 0.7937,
+      "step": 936
+    },
+    {
+      "epoch": 2.0582097748489843,
+      "grad_norm": 0.04064236208796501,
+      "learning_rate": 0.0001048045602605863,
+      "loss": 0.7191,
+      "step": 937
+    },
+    {
+      "epoch": 2.0604063701263042,
+      "grad_norm": 0.053273994475603104,
+      "learning_rate": 0.00010456026058631921,
+      "loss": 0.4936,
+      "step": 938
+    },
+    {
+      "epoch": 2.062602965403624,
+      "grad_norm": 0.06294192373752594,
+      "learning_rate": 0.00010431596091205211,
+      "loss": 0.8157,
+      "step": 939
+    },
+    {
+      "epoch": 2.0647995606809446,
+      "grad_norm": 0.04619294032454491,
+      "learning_rate": 0.000104071661237785,
+      "loss": 0.6172,
+      "step": 940
+    },
+    {
+      "epoch": 2.0669961559582646,
+      "grad_norm": 0.03779340907931328,
+      "learning_rate": 0.0001038273615635179,
+      "loss": 0.6405,
+      "step": 941
+    },
+    {
+      "epoch": 2.069192751235585,
+      "grad_norm": 0.04682890698313713,
+      "learning_rate": 0.0001035830618892508,
+      "loss": 0.5335,
+      "step": 942
+    },
+    {
+      "epoch": 2.071389346512905,
+      "grad_norm": 0.037470750510692596,
+      "learning_rate": 0.0001033387622149837,
+      "loss": 0.6709,
+      "step": 943
+    },
+    {
+      "epoch": 2.073585941790225,
+      "grad_norm": 0.06082337349653244,
+      "learning_rate": 0.0001030944625407166,
+      "loss": 0.6056,
+      "step": 944
+    },
+    {
+      "epoch": 2.0757825370675453,
+      "grad_norm": 0.03778183087706566,
+      "learning_rate": 0.0001028501628664495,
+      "loss": 0.6584,
+      "step": 945
+    },
+    {
+      "epoch": 2.0779791323448653,
+      "grad_norm": 0.05191833898425102,
+      "learning_rate": 0.0001026058631921824,
+      "loss": 0.6648,
+      "step": 946
+    },
+    {
+      "epoch": 2.0801757276221857,
+      "grad_norm": 0.04200424626469612,
+      "learning_rate": 0.00010236156351791529,
+      "loss": 0.5183,
+      "step": 947
+    },
+    {
+      "epoch": 2.0823723228995057,
+      "grad_norm": 0.04820658266544342,
+      "learning_rate": 0.0001021172638436482,
+      "loss": 0.5925,
+      "step": 948
+    },
+    {
+      "epoch": 2.084568918176826,
+      "grad_norm": 0.0433504655957222,
+      "learning_rate": 0.0001018729641693811,
+      "loss": 0.8134,
+      "step": 949
+    },
+    {
+      "epoch": 2.086765513454146,
+      "grad_norm": 0.03969017416238785,
+      "learning_rate": 0.00010162866449511399,
+      "loss": 0.6694,
+      "step": 950
+    },
+    {
+      "epoch": 2.088962108731466,
+      "grad_norm": 0.04646764323115349,
+      "learning_rate": 0.00010138436482084689,
+      "loss": 0.6878,
+      "step": 951
+    },
+    {
+      "epoch": 2.0911587040087865,
+      "grad_norm": 0.06577397137880325,
+      "learning_rate": 0.00010114006514657979,
+      "loss": 0.7297,
+      "step": 952
+    },
+    {
+      "epoch": 2.0933552992861064,
+      "grad_norm": 0.05729677528142929,
+      "learning_rate": 0.0001008957654723127,
+      "loss": 0.733,
+      "step": 953
+    },
+    {
+      "epoch": 2.095551894563427,
+      "grad_norm": 0.03979307413101196,
+      "learning_rate": 0.00010065146579804558,
+      "loss": 0.5778,
+      "step": 954
+    },
+    {
+      "epoch": 2.097748489840747,
+      "grad_norm": 0.05198509618639946,
+      "learning_rate": 0.00010040716612377849,
+      "loss": 0.731,
+      "step": 955
+    },
+    {
+      "epoch": 2.099945085118067,
+      "grad_norm": 0.05052348971366882,
+      "learning_rate": 0.00010016286644951139,
+      "loss": 0.6816,
+      "step": 956
+    },
+    {
+      "epoch": 2.102141680395387,
+      "grad_norm": 0.046905145049095154,
+      "learning_rate": 9.991856677524428e-05,
+      "loss": 0.6744,
+      "step": 957
+    },
+    {
+      "epoch": 2.104338275672707,
+      "grad_norm": 0.04074231535196304,
+      "learning_rate": 9.967426710097718e-05,
+      "loss": 0.6129,
+      "step": 958
+    },
+    {
+      "epoch": 2.1065348709500276,
+      "grad_norm": 0.049627695232629776,
+      "learning_rate": 9.942996742671008e-05,
+      "loss": 0.6302,
+      "step": 959
+    },
+    {
+      "epoch": 2.1087314662273475,
+      "grad_norm": 0.04932795837521553,
+      "learning_rate": 9.918566775244299e-05,
+      "loss": 0.6256,
+      "step": 960
+    },
+    {
+      "epoch": 2.110928061504668,
+      "grad_norm": 0.05283183604478836,
+      "learning_rate": 9.894136807817588e-05,
+      "loss": 0.7748,
+      "step": 961
+    },
+    {
+      "epoch": 2.113124656781988,
+      "grad_norm": 0.05655486509203911,
+      "learning_rate": 9.869706840390878e-05,
+      "loss": 0.6449,
+      "step": 962
+    },
+    {
+      "epoch": 2.115321252059308,
+      "grad_norm": 0.056285202503204346,
+      "learning_rate": 9.845276872964168e-05,
+      "loss": 0.5045,
+      "step": 963
+    },
+    {
+      "epoch": 2.1175178473366283,
+      "grad_norm": 0.0356547012925148,
+      "learning_rate": 9.820846905537458e-05,
+      "loss": 0.6095,
+      "step": 964
+    },
+    {
+      "epoch": 2.1197144426139483,
+      "grad_norm": 0.042278558015823364,
+      "learning_rate": 9.796416938110747e-05,
+      "loss": 0.593,
+      "step": 965
+    },
+    {
+      "epoch": 2.1219110378912687,
+      "grad_norm": 0.06479320675134659,
+      "learning_rate": 9.771986970684038e-05,
+      "loss": 0.6385,
+      "step": 966
+    },
+    {
+      "epoch": 2.1241076331685886,
+      "grad_norm": 0.042625319212675095,
+      "learning_rate": 9.747557003257328e-05,
+      "loss": 0.6546,
+      "step": 967
+    },
+    {
+      "epoch": 2.1263042284459086,
+      "grad_norm": 0.04128136858344078,
+      "learning_rate": 9.723127035830617e-05,
+      "loss": 0.6223,
+      "step": 968
+    },
+    {
+      "epoch": 2.128500823723229,
+      "grad_norm": 0.06777093559503555,
+      "learning_rate": 9.698697068403907e-05,
+      "loss": 0.6113,
+      "step": 969
+    },
+    {
+      "epoch": 2.130697419000549,
+      "grad_norm": 0.0486929826438427,
+      "learning_rate": 9.674267100977197e-05,
+      "loss": 0.7408,
+      "step": 970
+    },
+    {
+      "epoch": 2.1328940142778694,
+      "grad_norm": 0.06708229333162308,
+      "learning_rate": 9.649837133550488e-05,
+      "loss": 0.665,
+      "step": 971
+    },
+    {
+      "epoch": 2.1350906095551894,
+      "grad_norm": 0.060830358415842056,
+      "learning_rate": 9.625407166123777e-05,
+      "loss": 0.642,
+      "step": 972
+    },
+    {
+      "epoch": 2.13728720483251,
+      "grad_norm": 0.04374934732913971,
+      "learning_rate": 9.600977198697067e-05,
+      "loss": 0.6245,
+      "step": 973
+    },
+    {
+      "epoch": 2.1394838001098297,
+      "grad_norm": 0.049657490104436874,
+      "learning_rate": 9.576547231270357e-05,
+      "loss": 0.7092,
+      "step": 974
+    },
+    {
+      "epoch": 2.1416803953871497,
+      "grad_norm": 0.04418911039829254,
+      "learning_rate": 9.552117263843646e-05,
+      "loss": 0.7128,
+      "step": 975
+    },
+    {
+      "epoch": 2.14387699066447,
+      "grad_norm": 0.05693337321281433,
+      "learning_rate": 9.527687296416936e-05,
+      "loss": 0.6567,
+      "step": 976
+    },
+    {
+      "epoch": 2.14607358594179,
+      "grad_norm": 0.03899364918470383,
+      "learning_rate": 9.503257328990227e-05,
+      "loss": 0.6489,
+      "step": 977
+    },
+    {
+      "epoch": 2.1482701812191105,
+      "grad_norm": 0.05260545760393143,
+      "learning_rate": 9.478827361563517e-05,
+      "loss": 0.6311,
+      "step": 978
+    },
+    {
+      "epoch": 2.1504667764964305,
+      "grad_norm": 0.05104517191648483,
+      "learning_rate": 9.454397394136806e-05,
+      "loss": 0.6754,
+      "step": 979
+    },
+    {
+      "epoch": 2.152663371773751,
+      "grad_norm": 0.056579869240522385,
+      "learning_rate": 9.429967426710096e-05,
+      "loss": 0.742,
+      "step": 980
+    },
+    {
+      "epoch": 2.154859967051071,
+      "grad_norm": 0.04211939126253128,
+      "learning_rate": 9.405537459283386e-05,
+      "loss": 0.4709,
+      "step": 981
+    },
+    {
+      "epoch": 2.157056562328391,
+      "grad_norm": 0.04625248908996582,
+      "learning_rate": 9.381107491856675e-05,
+      "loss": 0.5968,
+      "step": 982
+    },
+    {
+      "epoch": 2.1592531576057112,
+      "grad_norm": 0.057066936045885086,
+      "learning_rate": 9.356677524429967e-05,
+      "loss": 0.7691,
+      "step": 983
+    },
+    {
+      "epoch": 2.161449752883031,
+      "grad_norm": 0.05318013206124306,
+      "learning_rate": 9.332247557003257e-05,
+      "loss": 0.6044,
+      "step": 984
+    },
+    {
+      "epoch": 2.1636463481603516,
+      "grad_norm": 0.05898617208003998,
+      "learning_rate": 9.307817589576547e-05,
+      "loss": 0.5647,
+      "step": 985
+    },
+    {
+      "epoch": 2.1658429434376716,
+      "grad_norm": 0.061303168535232544,
+      "learning_rate": 9.283387622149838e-05,
+      "loss": 0.5899,
+      "step": 986
+    },
+    {
+      "epoch": 2.168039538714992,
+      "grad_norm": 0.04799562692642212,
+      "learning_rate": 9.258957654723127e-05,
+      "loss": 0.7572,
+      "step": 987
+    },
+    {
+      "epoch": 2.170236133992312,
+      "grad_norm": 0.044705580919981,
+      "learning_rate": 9.234527687296417e-05,
+      "loss": 0.5564,
+      "step": 988
+    },
+    {
+      "epoch": 2.172432729269632,
+      "grad_norm": 0.041354063898324966,
+      "learning_rate": 9.210097719869707e-05,
+      "loss": 0.6036,
+      "step": 989
+    },
+    {
+      "epoch": 2.1746293245469523,
+      "grad_norm": 0.05093163624405861,
+      "learning_rate": 9.185667752442996e-05,
+      "loss": 0.8037,
+      "step": 990
+    },
+    {
+      "epoch": 2.1768259198242723,
+      "grad_norm": 0.049296095967292786,
+      "learning_rate": 9.161237785016286e-05,
+      "loss": 0.6168,
+      "step": 991
+    },
+    {
+      "epoch": 2.1790225151015927,
+      "grad_norm": 0.036712076514959335,
+      "learning_rate": 9.136807817589577e-05,
+      "loss": 0.5757,
+      "step": 992
+    },
+    {
+      "epoch": 2.1812191103789127,
+      "grad_norm": 0.061949945986270905,
+      "learning_rate": 9.112377850162867e-05,
+      "loss": 0.7439,
+      "step": 993
+    },
+    {
+      "epoch": 2.1834157056562327,
+      "grad_norm": 0.04836397245526314,
+      "learning_rate": 9.087947882736156e-05,
+      "loss": 0.5687,
+      "step": 994
+    },
+    {
+      "epoch": 2.185612300933553,
+      "grad_norm": 0.03990459069609642,
+      "learning_rate": 9.063517915309446e-05,
+      "loss": 0.5616,
+      "step": 995
+    },
+    {
+      "epoch": 2.187808896210873,
+      "grad_norm": 0.05062219500541687,
+      "learning_rate": 9.039087947882736e-05,
+      "loss": 0.7688,
+      "step": 996
+    },
+    {
+      "epoch": 2.1900054914881935,
+      "grad_norm": 0.049748651683330536,
+      "learning_rate": 9.014657980456025e-05,
+      "loss": 0.6998,
+      "step": 997
+    },
+    {
+      "epoch": 2.1922020867655134,
+      "grad_norm": 0.04505657032132149,
+      "learning_rate": 8.990228013029316e-05,
+      "loss": 0.6697,
+      "step": 998
+    },
+    {
+      "epoch": 2.1943986820428334,
+      "grad_norm": 0.051343776285648346,
+      "learning_rate": 8.965798045602606e-05,
+      "loss": 0.806,
+      "step": 999
+    },
+    {
+      "epoch": 2.196595277320154,
+      "grad_norm": 0.0413266085088253,
+      "learning_rate": 8.941368078175896e-05,
+      "loss": 0.5986,
+      "step": 1000
+    },
+    {
+      "epoch": 2.1987918725974738,
+      "grad_norm": 0.048408783972263336,
+      "learning_rate": 8.916938110749185e-05,
+      "loss": 0.478,
+      "step": 1001
+    },
+    {
+      "epoch": 2.200988467874794,
+      "grad_norm": 0.03760487586259842,
+      "learning_rate": 8.892508143322475e-05,
+      "loss": 0.5888,
+      "step": 1002
+    },
+    {
+      "epoch": 2.203185063152114,
+      "grad_norm": 0.04939005896449089,
+      "learning_rate": 8.868078175895766e-05,
+      "loss": 0.6409,
+      "step": 1003
+    },
+    {
+      "epoch": 2.2053816584294346,
+      "grad_norm": 0.04825667291879654,
+      "learning_rate": 8.843648208469055e-05,
+      "loss": 0.5696,
+      "step": 1004
+    },
+    {
+      "epoch": 2.2075782537067545,
+      "grad_norm": 0.04127410426735878,
+      "learning_rate": 8.819218241042345e-05,
+      "loss": 0.6175,
+      "step": 1005
+    },
+    {
+      "epoch": 2.2097748489840745,
+      "grad_norm": 0.05631517246365547,
+      "learning_rate": 8.794788273615635e-05,
+      "loss": 0.6802,
+      "step": 1006
+    },
+    {
+      "epoch": 2.211971444261395,
+      "grad_norm": 0.04190594330430031,
+      "learning_rate": 8.770358306188925e-05,
+      "loss": 0.5481,
+      "step": 1007
+    },
+    {
+      "epoch": 2.214168039538715,
+      "grad_norm": 0.05055314674973488,
+      "learning_rate": 8.745928338762214e-05,
+      "loss": 0.7913,
+      "step": 1008
+    },
+    {
+      "epoch": 2.2163646348160353,
+      "grad_norm": 0.04413871094584465,
+      "learning_rate": 8.721498371335505e-05,
+      "loss": 0.683,
+      "step": 1009
+    },
+    {
+      "epoch": 2.2185612300933553,
+      "grad_norm": 0.06312219798564911,
+      "learning_rate": 8.697068403908795e-05,
+      "loss": 0.5454,
+      "step": 1010
+    },
+    {
+      "epoch": 2.2207578253706757,
+      "grad_norm": 0.03939354047179222,
+      "learning_rate": 8.672638436482085e-05,
+      "loss": 0.6548,
+      "step": 1011
+    },
+    {
+      "epoch": 2.2229544206479956,
+      "grad_norm": 0.04720534384250641,
+      "learning_rate": 8.648208469055374e-05,
+      "loss": 0.869,
+      "step": 1012
+    },
+    {
+      "epoch": 2.2251510159253156,
+      "grad_norm": 0.09054285287857056,
+      "learning_rate": 8.623778501628664e-05,
+      "loss": 0.6499,
+      "step": 1013
+    },
+    {
+      "epoch": 2.227347611202636,
+      "grad_norm": 0.06740359216928482,
+      "learning_rate": 8.599348534201955e-05,
+      "loss": 0.7657,
+      "step": 1014
+    },
+    {
+      "epoch": 2.229544206479956,
+      "grad_norm": 0.08037208765745163,
+      "learning_rate": 8.574918566775243e-05,
+      "loss": 0.825,
+      "step": 1015
+    },
+    {
+      "epoch": 2.2317408017572764,
+      "grad_norm": 0.0381237156689167,
+      "learning_rate": 8.550488599348534e-05,
+      "loss": 0.6602,
+      "step": 1016
+    },
+    {
+      "epoch": 2.2339373970345964,
+      "grad_norm": 0.05984542518854141,
+      "learning_rate": 8.526058631921824e-05,
+      "loss": 0.7109,
+      "step": 1017
+    },
+    {
+      "epoch": 2.2361339923119163,
+      "grad_norm": 0.057318367063999176,
+      "learning_rate": 8.501628664495114e-05,
+      "loss": 0.7673,
+      "step": 1018
+    },
+    {
+      "epoch": 2.2383305875892368,
+      "grad_norm": 0.048353392630815506,
+      "learning_rate": 8.477198697068403e-05,
+      "loss": 0.731,
+      "step": 1019
+    },
+    {
+      "epoch": 2.2405271828665567,
+      "grad_norm": 0.09073536843061447,
+      "learning_rate": 8.452768729641694e-05,
+      "loss": 0.7223,
+      "step": 1020
+    },
+    {
+      "epoch": 2.242723778143877,
+      "grad_norm": 0.0407768115401268,
+      "learning_rate": 8.428338762214984e-05,
+      "loss": 0.5907,
+      "step": 1021
+    },
+    {
+      "epoch": 2.244920373421197,
+      "grad_norm": 0.04868381470441818,
+      "learning_rate": 8.403908794788273e-05,
+      "loss": 0.5471,
+      "step": 1022
+    },
+    {
+      "epoch": 2.247116968698517,
+      "grad_norm": 0.056227799504995346,
+      "learning_rate": 8.379478827361563e-05,
+      "loss": 0.7733,
+      "step": 1023
+    },
+    {
+      "epoch": 2.2493135639758375,
+      "grad_norm": 0.05046963691711426,
+      "learning_rate": 8.355048859934853e-05,
+      "loss": 0.6275,
+      "step": 1024
+    },
+    {
+      "epoch": 2.2515101592531575,
+      "grad_norm": 0.03843815252184868,
+      "learning_rate": 8.330618892508144e-05,
+      "loss": 0.6212,
+      "step": 1025
+    },
+    {
+      "epoch": 2.253706754530478,
+      "grad_norm": 0.04150962457060814,
+      "learning_rate": 8.306188925081432e-05,
+      "loss": 0.5651,
+      "step": 1026
+    },
+    {
+      "epoch": 2.255903349807798,
+      "grad_norm": 0.05324546620249748,
+      "learning_rate": 8.281758957654723e-05,
+      "loss": 0.6938,
+      "step": 1027
+    },
+    {
+      "epoch": 2.2580999450851182,
+      "grad_norm": 0.039021965116262436,
+      "learning_rate": 8.257328990228013e-05,
+      "loss": 0.725,
+      "step": 1028
+    },
+    {
+      "epoch": 2.260296540362438,
+      "grad_norm": 0.060293395072221756,
+      "learning_rate": 8.232899022801302e-05,
+      "loss": 0.7034,
+      "step": 1029
+    },
+    {
+      "epoch": 2.262493135639758,
+      "grad_norm": 0.046492014080286026,
+      "learning_rate": 8.208469055374592e-05,
+      "loss": 0.5136,
+      "step": 1030
+    },
+    {
+      "epoch": 2.2646897309170786,
+      "grad_norm": 0.04280600696802139,
+      "learning_rate": 8.184039087947882e-05,
+      "loss": 0.5654,
+      "step": 1031
+    },
+    {
+      "epoch": 2.2668863261943986,
+      "grad_norm": 0.05168813094496727,
+      "learning_rate": 8.159609120521173e-05,
+      "loss": 0.5706,
+      "step": 1032
+    },
+    {
+      "epoch": 2.269082921471719,
+      "grad_norm": 0.06857836991548538,
+      "learning_rate": 8.135179153094462e-05,
+      "loss": 0.8093,
+      "step": 1033
+    },
+    {
+      "epoch": 2.271279516749039,
+      "grad_norm": 0.0616997666656971,
+      "learning_rate": 8.110749185667752e-05,
+      "loss": 0.6479,
+      "step": 1034
+    },
+    {
+      "epoch": 2.2734761120263594,
+      "grad_norm": 0.04081263393163681,
+      "learning_rate": 8.086319218241042e-05,
+      "loss": 0.5847,
+      "step": 1035
+    },
+    {
+      "epoch": 2.2756727073036793,
+      "grad_norm": 0.0425679050385952,
+      "learning_rate": 8.061889250814331e-05,
+      "loss": 0.6171,
+      "step": 1036
+    },
+    {
+      "epoch": 2.2778693025809993,
+      "grad_norm": 0.036269091069698334,
+      "learning_rate": 8.037459283387621e-05,
+      "loss": 0.5724,
+      "step": 1037
+    },
+    {
+      "epoch": 2.2800658978583197,
+      "grad_norm": 0.04767118766903877,
+      "learning_rate": 8.013029315960912e-05,
+      "loss": 0.6088,
+      "step": 1038
+    },
+    {
+      "epoch": 2.2822624931356397,
+      "grad_norm": 0.04134832322597504,
+      "learning_rate": 7.988599348534202e-05,
+      "loss": 0.5544,
+      "step": 1039
+    },
+    {
+      "epoch": 2.28445908841296,
+      "grad_norm": 0.05227472260594368,
+      "learning_rate": 7.964169381107491e-05,
+      "loss": 0.6697,
+      "step": 1040
+    },
+    {
+      "epoch": 2.28665568369028,
+      "grad_norm": 0.05427214875817299,
+      "learning_rate": 7.939739413680781e-05,
+      "loss": 0.6783,
+      "step": 1041
+    },
+    {
+      "epoch": 2.2888522789676005,
+      "grad_norm": 0.05169343948364258,
+      "learning_rate": 7.915309446254071e-05,
+      "loss": 0.7326,
+      "step": 1042
+    },
+    {
+      "epoch": 2.2910488742449204,
+      "grad_norm": 0.03582863137125969,
+      "learning_rate": 7.890879478827362e-05,
+      "loss": 0.5292,
+      "step": 1043
+    },
+    {
+      "epoch": 2.2932454695222404,
+      "grad_norm": 0.04561784490942955,
+      "learning_rate": 7.86644951140065e-05,
+      "loss": 0.6495,
+      "step": 1044
+    },
+    {
+      "epoch": 2.295442064799561,
+      "grad_norm": 0.04176398739218712,
+      "learning_rate": 7.842019543973941e-05,
+      "loss": 0.7175,
+      "step": 1045
+    },
+    {
+      "epoch": 2.297638660076881,
+      "grad_norm": 0.04040032997727394,
+      "learning_rate": 7.817589576547231e-05,
+      "loss": 0.6138,
+      "step": 1046
+    },
+    {
+      "epoch": 2.2998352553542007,
+      "grad_norm": 0.06742952764034271,
+      "learning_rate": 7.79315960912052e-05,
+      "loss": 0.547,
+      "step": 1047
+    },
+    {
+      "epoch": 2.302031850631521,
+      "grad_norm": 0.04996305704116821,
+      "learning_rate": 7.76872964169381e-05,
+      "loss": 0.6176,
+      "step": 1048
+    },
+    {
+      "epoch": 2.304228445908841,
+      "grad_norm": 0.04592761769890785,
+      "learning_rate": 7.7442996742671e-05,
+      "loss": 0.6208,
+      "step": 1049
+    },
+    {
+      "epoch": 2.3064250411861615,
+      "grad_norm": 0.05607963353395462,
+      "learning_rate": 7.719869706840391e-05,
+      "loss": 0.675,
+      "step": 1050
+    },
+    {
+      "epoch": 2.3086216364634815,
+      "grad_norm": 0.04420773312449455,
+      "learning_rate": 7.69543973941368e-05,
+      "loss": 0.5119,
+      "step": 1051
+    },
+    {
+      "epoch": 2.310818231740802,
+      "grad_norm": 0.044873543083667755,
+      "learning_rate": 7.67100977198697e-05,
+      "loss": 0.6016,
+      "step": 1052
+    },
+    {
+      "epoch": 2.313014827018122,
+      "grad_norm": 0.0465567484498024,
+      "learning_rate": 7.64657980456026e-05,
+      "loss": 0.686,
+      "step": 1053
+    },
+    {
+      "epoch": 2.315211422295442,
+      "grad_norm": 0.04192805662751198,
+      "learning_rate": 7.622149837133549e-05,
+      "loss": 0.6144,
+      "step": 1054
+    },
+    {
+      "epoch": 2.3174080175727623,
+      "grad_norm": 0.04011636599898338,
+      "learning_rate": 7.59771986970684e-05,
+      "loss": 0.8029,
+      "step": 1055
+    },
+    {
+      "epoch": 2.3196046128500822,
+      "grad_norm": 0.04277574643492699,
+      "learning_rate": 7.57328990228013e-05,
+      "loss": 0.6306,
+      "step": 1056
+    },
+    {
+      "epoch": 2.3218012081274026,
+      "grad_norm": 0.05560114234685898,
+      "learning_rate": 7.54885993485342e-05,
+      "loss": 0.8514,
+      "step": 1057
+    },
+    {
+      "epoch": 2.3239978034047226,
+      "grad_norm": 0.037569232285022736,
+      "learning_rate": 7.524429967426709e-05,
+      "loss": 0.469,
+      "step": 1058
+    },
+    {
+      "epoch": 2.326194398682043,
+      "grad_norm": 0.045851316303014755,
+      "learning_rate": 7.5e-05,
+      "loss": 0.5967,
+      "step": 1059
+    },
+    {
+      "epoch": 2.328390993959363,
+      "grad_norm": 0.03776060789823532,
+      "learning_rate": 7.47557003257329e-05,
+      "loss": 0.5763,
+      "step": 1060
+    },
+    {
+      "epoch": 2.330587589236683,
+      "grad_norm": 0.04432753473520279,
+      "learning_rate": 7.451140065146579e-05,
+      "loss": 0.6155,
+      "step": 1061
+    },
+    {
+      "epoch": 2.3327841845140034,
+      "grad_norm": 0.04885265603661537,
+      "learning_rate": 7.426710097719869e-05,
+      "loss": 0.7005,
+      "step": 1062
+    },
+    {
+      "epoch": 2.3349807797913233,
+      "grad_norm": 0.04530097171664238,
+      "learning_rate": 7.402280130293159e-05,
+      "loss": 0.7549,
+      "step": 1063
+    },
+    {
+      "epoch": 2.3371773750686438,
+      "grad_norm": 0.04386676102876663,
+      "learning_rate": 7.37785016286645e-05,
+      "loss": 0.6721,
+      "step": 1064
+    },
+    {
+      "epoch": 2.3393739703459637,
+      "grad_norm": 0.04051361232995987,
+      "learning_rate": 7.353420195439738e-05,
+      "loss": 0.6354,
+      "step": 1065
+    },
+    {
+      "epoch": 2.341570565623284,
+      "grad_norm": 0.05379435792565346,
+      "learning_rate": 7.328990228013029e-05,
+      "loss": 0.7204,
+      "step": 1066
+    },
+    {
+      "epoch": 2.343767160900604,
+      "grad_norm": 0.03002171590924263,
+      "learning_rate": 7.304560260586319e-05,
+      "loss": 0.5271,
+      "step": 1067
+    },
+    {
+      "epoch": 2.345963756177924,
+      "grad_norm": 0.03606308996677399,
+      "learning_rate": 7.280130293159608e-05,
+      "loss": 0.6648,
+      "step": 1068
+    },
+    {
+      "epoch": 2.3481603514552445,
+      "grad_norm": 0.058911386877298355,
+      "learning_rate": 7.255700325732898e-05,
+      "loss": 0.7517,
+      "step": 1069
+    },
+    {
+      "epoch": 2.3503569467325645,
+      "grad_norm": 0.04218167066574097,
+      "learning_rate": 7.231270358306188e-05,
+      "loss": 0.5927,
+      "step": 1070
+    },
+    {
+      "epoch": 2.352553542009885,
+      "grad_norm": 0.04816456139087677,
+      "learning_rate": 7.206840390879479e-05,
+      "loss": 0.6595,
+      "step": 1071
+    },
+    {
+      "epoch": 2.354750137287205,
+      "grad_norm": 0.04421921819448471,
+      "learning_rate": 7.182410423452768e-05,
+      "loss": 0.5484,
+      "step": 1072
+    },
+    {
+      "epoch": 2.356946732564525,
+      "grad_norm": 0.04146098718047142,
+      "learning_rate": 7.157980456026058e-05,
+      "loss": 0.5741,
+      "step": 1073
+    },
+    {
+      "epoch": 2.359143327841845,
+      "grad_norm": 0.045608628541231155,
+      "learning_rate": 7.133550488599348e-05,
+      "loss": 0.6229,
+      "step": 1074
+    },
+    {
+      "epoch": 2.361339923119165,
+      "grad_norm": 0.04145771637558937,
+      "learning_rate": 7.109120521172637e-05,
+      "loss": 0.74,
+      "step": 1075
+    },
+    {
+      "epoch": 2.3635365183964856,
+      "grad_norm": 0.05360572785139084,
+      "learning_rate": 7.084690553745927e-05,
+      "loss": 0.7087,
+      "step": 1076
+    },
+    {
+      "epoch": 2.3657331136738056,
+      "grad_norm": 0.051699042320251465,
+      "learning_rate": 7.060260586319218e-05,
+      "loss": 0.6353,
+      "step": 1077
+    },
+    {
+      "epoch": 2.3679297089511255,
+      "grad_norm": 0.05747654661536217,
+      "learning_rate": 7.035830618892508e-05,
+      "loss": 0.7254,
+      "step": 1078
+    },
+    {
+      "epoch": 2.370126304228446,
+      "grad_norm": 0.04842827096581459,
+      "learning_rate": 7.011400651465797e-05,
+      "loss": 0.5057,
+      "step": 1079
+    },
+    {
+      "epoch": 2.372322899505766,
+      "grad_norm": 0.056867651641368866,
+      "learning_rate": 6.986970684039087e-05,
+      "loss": 0.6912,
+      "step": 1080
+    },
+    {
+      "epoch": 2.3745194947830863,
+      "grad_norm": 0.06867987662553787,
+      "learning_rate": 6.962540716612377e-05,
+      "loss": 0.9034,
+      "step": 1081
+    },
+    {
+      "epoch": 2.3767160900604063,
+      "grad_norm": 0.044970154762268066,
+      "learning_rate": 6.938110749185668e-05,
+      "loss": 0.5275,
+      "step": 1082
+    },
+    {
+      "epoch": 2.3789126853377267,
+      "grad_norm": 0.04874912649393082,
+      "learning_rate": 6.913680781758956e-05,
+      "loss": 0.6088,
+      "step": 1083
+    },
+    {
+      "epoch": 2.3811092806150467,
+      "grad_norm": 0.06273435056209564,
+      "learning_rate": 6.889250814332247e-05,
+      "loss": 0.7871,
+      "step": 1084
+    },
+    {
+      "epoch": 2.3833058758923666,
+      "grad_norm": 0.06409198045730591,
+      "learning_rate": 6.864820846905537e-05,
+      "loss": 0.8235,
+      "step": 1085
+    },
+    {
+      "epoch": 2.385502471169687,
+      "grad_norm": 0.04517259821295738,
+      "learning_rate": 6.840390879478826e-05,
+      "loss": 0.5531,
+      "step": 1086
+    },
+    {
+      "epoch": 2.387699066447007,
+      "grad_norm": 0.06406451761722565,
+      "learning_rate": 6.815960912052116e-05,
+      "loss": 0.8643,
+      "step": 1087
+    },
+    {
+      "epoch": 2.3898956617243274,
+      "grad_norm": 0.06429564207792282,
+      "learning_rate": 6.791530944625406e-05,
+      "loss": 0.5985,
+      "step": 1088
+    },
+    {
+      "epoch": 2.3920922570016474,
+      "grad_norm": 0.058752164244651794,
+      "learning_rate": 6.767100977198697e-05,
+      "loss": 0.7691,
+      "step": 1089
+    },
+    {
+      "epoch": 2.394288852278968,
+      "grad_norm": 0.04136451706290245,
+      "learning_rate": 6.742671009771986e-05,
+      "loss": 0.614,
+      "step": 1090
+    },
+    {
+      "epoch": 2.396485447556288,
+      "grad_norm": 0.06292608380317688,
+      "learning_rate": 6.718241042345276e-05,
+      "loss": 0.6292,
+      "step": 1091
+    },
+    {
+      "epoch": 2.3986820428336078,
+      "grad_norm": 0.06885907799005508,
+      "learning_rate": 6.693811074918566e-05,
+      "loss": 0.6684,
+      "step": 1092
+    },
+    {
+      "epoch": 2.400878638110928,
+      "grad_norm": 0.04652557522058487,
+      "learning_rate": 6.669381107491855e-05,
+      "loss": 0.6725,
+      "step": 1093
+    },
+    {
+      "epoch": 2.403075233388248,
+      "grad_norm": 0.03902400657534599,
+      "learning_rate": 6.644951140065145e-05,
+      "loss": 0.6752,
+      "step": 1094
+    },
+    {
+      "epoch": 2.4052718286655685,
+      "grad_norm": 0.05414823442697525,
+      "learning_rate": 6.620521172638436e-05,
+      "loss": 0.603,
+      "step": 1095
+    },
+    {
+      "epoch": 2.4074684239428885,
+      "grad_norm": 0.04311055690050125,
+      "learning_rate": 6.596091205211726e-05,
+      "loss": 0.5174,
+      "step": 1096
+    },
+    {
+      "epoch": 2.409665019220209,
+      "grad_norm": 0.05905736982822418,
+      "learning_rate": 6.571661237785015e-05,
+      "loss": 0.7347,
+      "step": 1097
+    },
+    {
+      "epoch": 2.411861614497529,
+      "grad_norm": 0.04053014516830444,
+      "learning_rate": 6.547231270358307e-05,
+      "loss": 0.6997,
+      "step": 1098
+    },
+    {
+      "epoch": 2.414058209774849,
+      "grad_norm": 0.053303156048059464,
+      "learning_rate": 6.522801302931595e-05,
+      "loss": 0.7643,
+      "step": 1099
+    },
+    {
+      "epoch": 2.4162548050521693,
+      "grad_norm": 0.047551583498716354,
+      "learning_rate": 6.498371335504886e-05,
+      "loss": 0.6913,
+      "step": 1100
+    },
+    {
+      "epoch": 2.4184514003294892,
+      "grad_norm": 0.04742944985628128,
+      "learning_rate": 6.473941368078176e-05,
+      "loss": 0.5594,
+      "step": 1101
+    },
+    {
+      "epoch": 2.420647995606809,
+      "grad_norm": 0.04457717761397362,
+      "learning_rate": 6.449511400651466e-05,
+      "loss": 0.7067,
+      "step": 1102
+    },
+    {
+      "epoch": 2.4228445908841296,
+      "grad_norm": 0.05029049888253212,
+      "learning_rate": 6.425081433224755e-05,
+      "loss": 0.6744,
+      "step": 1103
+    },
+    {
+      "epoch": 2.4250411861614496,
+      "grad_norm": 0.057819634675979614,
+      "learning_rate": 6.400651465798045e-05,
+      "loss": 0.7409,
+      "step": 1104
+    },
+    {
+      "epoch": 2.42723778143877,
+      "grad_norm": 0.06167594715952873,
+      "learning_rate": 6.376221498371336e-05,
+      "loss": 0.6952,
+      "step": 1105
+    },
+    {
+      "epoch": 2.42943437671609,
+      "grad_norm": 0.04718726500868797,
+      "learning_rate": 6.351791530944625e-05,
+      "loss": 0.5742,
+      "step": 1106
+    },
+    {
+      "epoch": 2.4316309719934104,
+      "grad_norm": 0.045384399592876434,
+      "learning_rate": 6.327361563517915e-05,
+      "loss": 0.7108,
+      "step": 1107
+    },
+    {
+      "epoch": 2.4338275672707304,
+      "grad_norm": 0.048894964158535004,
+      "learning_rate": 6.302931596091205e-05,
+      "loss": 0.7631,
+      "step": 1108
+    },
+    {
+      "epoch": 2.4360241625480503,
+      "grad_norm": 0.0457146130502224,
+      "learning_rate": 6.278501628664495e-05,
+      "loss": 0.6816,
+      "step": 1109
+    },
+    {
+      "epoch": 2.4382207578253707,
+      "grad_norm": 0.04110351949930191,
+      "learning_rate": 6.254071661237784e-05,
+      "loss": 0.676,
+      "step": 1110
+    },
+    {
+      "epoch": 2.4404173531026907,
+      "grad_norm": 0.04073873534798622,
+      "learning_rate": 6.229641693811075e-05,
+      "loss": 0.6479,
+      "step": 1111
+    },
+    {
+      "epoch": 2.442613948380011,
+      "grad_norm": 0.054913751780986786,
+      "learning_rate": 6.205211726384365e-05,
+      "loss": 0.4896,
+      "step": 1112
+    },
+    {
+      "epoch": 2.444810543657331,
+      "grad_norm": 0.05325063318014145,
+      "learning_rate": 6.180781758957654e-05,
+      "loss": 0.8076,
+      "step": 1113
+    },
+    {
+      "epoch": 2.4470071389346515,
+      "grad_norm": 0.05473218858242035,
+      "learning_rate": 6.156351791530944e-05,
+      "loss": 0.8352,
+      "step": 1114
+    },
+    {
+      "epoch": 2.4492037342119715,
+      "grad_norm": 0.046617619693279266,
+      "learning_rate": 6.131921824104234e-05,
+      "loss": 0.575,
+      "step": 1115
+    },
+    {
+      "epoch": 2.4514003294892914,
+      "grad_norm": 0.0457211434841156,
+      "learning_rate": 6.107491856677525e-05,
+      "loss": 0.5843,
+      "step": 1116
+    },
+    {
+      "epoch": 2.453596924766612,
+      "grad_norm": 0.05672460421919823,
+      "learning_rate": 6.083061889250814e-05,
+      "loss": 0.6063,
+      "step": 1117
+    },
+    {
+      "epoch": 2.455793520043932,
+      "grad_norm": 0.04907252639532089,
+      "learning_rate": 6.058631921824104e-05,
+      "loss": 0.5895,
+      "step": 1118
+    },
+    {
+      "epoch": 2.4579901153212522,
+      "grad_norm": 0.04526342824101448,
+      "learning_rate": 6.0342019543973935e-05,
+      "loss": 0.7618,
+      "step": 1119
+    },
+    {
+      "epoch": 2.460186710598572,
+      "grad_norm": 0.045307356864213943,
+      "learning_rate": 6.009771986970684e-05,
+      "loss": 0.6353,
+      "step": 1120
+    },
+    {
+      "epoch": 2.4623833058758926,
+      "grad_norm": 0.05209842324256897,
+      "learning_rate": 5.9853420195439734e-05,
+      "loss": 0.6152,
+      "step": 1121
+    },
+    {
+      "epoch": 2.4645799011532126,
+      "grad_norm": 0.06889799982309341,
+      "learning_rate": 5.9609120521172636e-05,
+      "loss": 1.0135,
+      "step": 1122
+    },
+    {
+      "epoch": 2.4667764964305325,
+      "grad_norm": 0.05798589438199997,
+      "learning_rate": 5.936482084690553e-05,
+      "loss": 0.6518,
+      "step": 1123
+    },
+    {
+      "epoch": 2.468973091707853,
+      "grad_norm": 0.04932552948594093,
+      "learning_rate": 5.9120521172638435e-05,
+      "loss": 0.6739,
+      "step": 1124
+    },
+    {
+      "epoch": 2.471169686985173,
+      "grad_norm": 0.04319072514772415,
+      "learning_rate": 5.887622149837133e-05,
+      "loss": 0.5889,
+      "step": 1125
+    },
+    {
+      "epoch": 2.473366282262493,
+      "grad_norm": 0.054561588913202286,
+      "learning_rate": 5.863192182410423e-05,
+      "loss": 0.7459,
+      "step": 1126
+    },
+    {
+      "epoch": 2.4755628775398133,
+      "grad_norm": 0.05109608545899391,
+      "learning_rate": 5.838762214983713e-05,
+      "loss": 0.5898,
+      "step": 1127
+    },
+    {
+      "epoch": 2.4777594728171333,
+      "grad_norm": 0.030667340382933617,
+      "learning_rate": 5.8143322475570026e-05,
+      "loss": 0.6168,
+      "step": 1128
+    },
+    {
+      "epoch": 2.4799560680944537,
+      "grad_norm": 0.050101760774850845,
+      "learning_rate": 5.789902280130293e-05,
+      "loss": 0.8355,
+      "step": 1129
+    },
+    {
+      "epoch": 2.4821526633717736,
+      "grad_norm": 0.049108896404504776,
+      "learning_rate": 5.7654723127035825e-05,
+      "loss": 0.6122,
+      "step": 1130
+    },
+    {
+      "epoch": 2.484349258649094,
+      "grad_norm": 0.0410931296646595,
+      "learning_rate": 5.741042345276873e-05,
+      "loss": 0.6199,
+      "step": 1131
+    },
+    {
+      "epoch": 2.486545853926414,
+      "grad_norm": 0.05107707902789116,
+      "learning_rate": 5.716612377850162e-05,
+      "loss": 0.692,
+      "step": 1132
+    },
+    {
+      "epoch": 2.488742449203734,
+      "grad_norm": 0.05445995554327965,
+      "learning_rate": 5.6921824104234526e-05,
+      "loss": 0.5391,
+      "step": 1133
+    },
+    {
+      "epoch": 2.4909390444810544,
+      "grad_norm": 0.05041612312197685,
+      "learning_rate": 5.667752442996742e-05,
+      "loss": 0.6833,
+      "step": 1134
+    },
+    {
+      "epoch": 2.4931356397583744,
+      "grad_norm": 0.05433562770485878,
+      "learning_rate": 5.643322475570032e-05,
+      "loss": 0.6777,
+      "step": 1135
+    },
+    {
+      "epoch": 2.495332235035695,
+      "grad_norm": 0.045349642634391785,
+      "learning_rate": 5.618892508143322e-05,
+      "loss": 0.5513,
+      "step": 1136
+    },
+    {
+      "epoch": 2.4975288303130148,
+      "grad_norm": 0.04080528765916824,
+      "learning_rate": 5.594462540716612e-05,
+      "loss": 0.4565,
+      "step": 1137
+    },
+    {
+      "epoch": 2.499725425590335,
+      "grad_norm": 0.05346320569515228,
+      "learning_rate": 5.570032573289902e-05,
+      "loss": 0.6752,
+      "step": 1138
+    },
+    {
+      "epoch": 2.501922020867655,
+      "grad_norm": 0.04871304705739021,
+      "learning_rate": 5.5456026058631915e-05,
+      "loss": 0.7841,
+      "step": 1139
+    },
+    {
+      "epoch": 2.504118616144975,
+      "grad_norm": 0.08751735091209412,
+      "learning_rate": 5.521172638436482e-05,
+      "loss": 0.5952,
+      "step": 1140
+    },
+    {
+      "epoch": 2.5063152114222955,
+      "grad_norm": 0.033952705562114716,
+      "learning_rate": 5.4967426710097714e-05,
+      "loss": 0.5781,
+      "step": 1141
+    },
+    {
+      "epoch": 2.5085118066996155,
+      "grad_norm": 0.04253567382693291,
+      "learning_rate": 5.472312703583061e-05,
+      "loss": 0.7076,
+      "step": 1142
+    },
+    {
+      "epoch": 2.510708401976936,
+      "grad_norm": 0.04273783415555954,
+      "learning_rate": 5.447882736156351e-05,
+      "loss": 0.7619,
+      "step": 1143
+    },
+    {
+      "epoch": 2.512904997254256,
+      "grad_norm": 0.041224155575037,
+      "learning_rate": 5.423452768729641e-05,
+      "loss": 0.6162,
+      "step": 1144
+    },
+    {
+      "epoch": 2.5151015925315763,
+      "grad_norm": 0.04299287870526314,
+      "learning_rate": 5.399022801302931e-05,
+      "loss": 0.6113,
+      "step": 1145
+    },
+    {
+      "epoch": 2.5172981878088962,
+      "grad_norm": 0.06844445317983627,
+      "learning_rate": 5.374592833876221e-05,
+      "loss": 0.6811,
+      "step": 1146
+    },
+    {
+      "epoch": 2.519494783086216,
+      "grad_norm": 0.04922829195857048,
+      "learning_rate": 5.350162866449511e-05,
+      "loss": 0.6346,
+      "step": 1147
+    },
+    {
+      "epoch": 2.5216913783635366,
+      "grad_norm": 0.04479237645864487,
+      "learning_rate": 5.3257328990228006e-05,
+      "loss": 0.5057,
+      "step": 1148
+    },
+    {
+      "epoch": 2.5238879736408566,
+      "grad_norm": 0.05080297216773033,
+      "learning_rate": 5.301302931596091e-05,
+      "loss": 0.694,
+      "step": 1149
+    },
+    {
+      "epoch": 2.5260845689181766,
+      "grad_norm": 0.04080790653824806,
+      "learning_rate": 5.2768729641693805e-05,
+      "loss": 0.8815,
+      "step": 1150
+    },
+    {
+      "epoch": 2.528281164195497,
+      "grad_norm": 0.05466169863939285,
+      "learning_rate": 5.25244299674267e-05,
+      "loss": 0.663,
+      "step": 1151
+    },
+    {
+      "epoch": 2.5304777594728174,
+      "grad_norm": 0.05815236642956734,
+      "learning_rate": 5.2280130293159604e-05,
+      "loss": 0.6391,
+      "step": 1152
+    },
+    {
+      "epoch": 2.5326743547501374,
+      "grad_norm": 0.043016865849494934,
+      "learning_rate": 5.20358306188925e-05,
+      "loss": 0.6724,
+      "step": 1153
+    },
+    {
+      "epoch": 2.5348709500274573,
+      "grad_norm": 0.03767896816134453,
+      "learning_rate": 5.17915309446254e-05,
+      "loss": 0.6149,
+      "step": 1154
+    },
+    {
+      "epoch": 2.5370675453047777,
+      "grad_norm": 0.04087197780609131,
+      "learning_rate": 5.15472312703583e-05,
+      "loss": 0.5995,
+      "step": 1155
+    },
+    {
+      "epoch": 2.5392641405820977,
+      "grad_norm": 0.055968958884477615,
+      "learning_rate": 5.13029315960912e-05,
+      "loss": 0.7466,
+      "step": 1156
+    },
+    {
+      "epoch": 2.5414607358594177,
+      "grad_norm": 0.0782134160399437,
+      "learning_rate": 5.10586319218241e-05,
+      "loss": 0.7928,
+      "step": 1157
+    },
+    {
+      "epoch": 2.543657331136738,
+      "grad_norm": 0.04652741923928261,
+      "learning_rate": 5.081433224755699e-05,
+      "loss": 0.5977,
+      "step": 1158
+    },
+    {
+      "epoch": 2.545853926414058,
+      "grad_norm": 0.04434973746538162,
+      "learning_rate": 5.0570032573289896e-05,
+      "loss": 0.5646,
+      "step": 1159
+    },
+    {
+      "epoch": 2.5480505216913785,
+      "grad_norm": 0.042396001517772675,
+      "learning_rate": 5.032573289902279e-05,
+      "loss": 0.625,
+      "step": 1160
+    },
+    {
+      "epoch": 2.5502471169686984,
+      "grad_norm": 0.04982843995094299,
+      "learning_rate": 5.0081433224755695e-05,
+      "loss": 0.5383,
+      "step": 1161
+    },
+    {
+      "epoch": 2.552443712246019,
+      "grad_norm": 0.04226350039243698,
+      "learning_rate": 4.983713355048859e-05,
+      "loss": 0.683,
+      "step": 1162
+    },
+    {
+      "epoch": 2.554640307523339,
+      "grad_norm": 0.05881736800074577,
+      "learning_rate": 4.9592833876221493e-05,
+      "loss": 0.5796,
+      "step": 1163
+    },
+    {
+      "epoch": 2.556836902800659,
+      "grad_norm": 0.04545532539486885,
+      "learning_rate": 4.934853420195439e-05,
+      "loss": 0.7102,
+      "step": 1164
+    },
+    {
+      "epoch": 2.559033498077979,
+      "grad_norm": 0.04276625066995621,
+      "learning_rate": 4.910423452768729e-05,
+      "loss": 0.6821,
+      "step": 1165
+    },
+    {
+      "epoch": 2.561230093355299,
+      "grad_norm": 0.04249290004372597,
+      "learning_rate": 4.885993485342019e-05,
+      "loss": 0.6215,
+      "step": 1166
+    },
+    {
+      "epoch": 2.5634266886326196,
+      "grad_norm": 0.04267478361725807,
+      "learning_rate": 4.8615635179153084e-05,
+      "loss": 0.5601,
+      "step": 1167
+    },
+    {
+      "epoch": 2.5656232839099395,
+      "grad_norm": 0.05035392940044403,
+      "learning_rate": 4.837133550488599e-05,
+      "loss": 0.6485,
+      "step": 1168
+    },
+    {
+      "epoch": 2.56781987918726,
+      "grad_norm": 0.04392477497458458,
+      "learning_rate": 4.812703583061888e-05,
+      "loss": 0.6361,
+      "step": 1169
+    },
+    {
+      "epoch": 2.57001647446458,
+      "grad_norm": 0.044216789305210114,
+      "learning_rate": 4.7882736156351786e-05,
+      "loss": 0.6166,
+      "step": 1170
+    },
+    {
+      "epoch": 2.5722130697419,
+      "grad_norm": 0.04420490562915802,
+      "learning_rate": 4.763843648208468e-05,
+      "loss": 0.5684,
+      "step": 1171
+    },
+    {
+      "epoch": 2.5744096650192203,
+      "grad_norm": 0.09682516753673553,
+      "learning_rate": 4.7394136807817584e-05,
+      "loss": 0.6564,
+      "step": 1172
+    },
+    {
+      "epoch": 2.5766062602965403,
+      "grad_norm": 0.04140635207295418,
+      "learning_rate": 4.714983713355048e-05,
+      "loss": 0.6094,
+      "step": 1173
+    },
+    {
+      "epoch": 2.5788028555738602,
+      "grad_norm": 0.04643336683511734,
+      "learning_rate": 4.6905537459283376e-05,
+      "loss": 0.7015,
+      "step": 1174
+    },
+    {
+      "epoch": 2.5809994508511807,
+      "grad_norm": 0.05290674418210983,
+      "learning_rate": 4.6661237785016286e-05,
+      "loss": 0.6969,
+      "step": 1175
+    },
+    {
+      "epoch": 2.583196046128501,
+      "grad_norm": 0.047984324395656586,
+      "learning_rate": 4.641693811074919e-05,
+      "loss": 0.5882,
+      "step": 1176
+    },
+    {
+      "epoch": 2.585392641405821,
+      "grad_norm": 0.055153653025627136,
+      "learning_rate": 4.6172638436482085e-05,
+      "loss": 0.8228,
+      "step": 1177
+    },
+    {
+      "epoch": 2.587589236683141,
+      "grad_norm": 0.049381084740161896,
+      "learning_rate": 4.592833876221498e-05,
+      "loss": 0.596,
+      "step": 1178
+    },
+    {
+      "epoch": 2.5897858319604614,
+      "grad_norm": 0.07250449061393738,
+      "learning_rate": 4.568403908794788e-05,
+      "loss": 0.7352,
+      "step": 1179
+    },
+    {
+      "epoch": 2.5919824272377814,
+      "grad_norm": 0.04308245703577995,
+      "learning_rate": 4.543973941368078e-05,
+      "loss": 0.571,
+      "step": 1180
+    },
+    {
+      "epoch": 2.5941790225151014,
+      "grad_norm": 0.05505221337080002,
+      "learning_rate": 4.519543973941368e-05,
+      "loss": 0.6903,
+      "step": 1181
+    },
+    {
+      "epoch": 2.5963756177924218,
+      "grad_norm": 0.06684863567352295,
+      "learning_rate": 4.495114006514658e-05,
+      "loss": 0.7423,
+      "step": 1182
+    },
+    {
+      "epoch": 2.598572213069742,
+      "grad_norm": 0.04111980274319649,
+      "learning_rate": 4.470684039087948e-05,
+      "loss": 0.6431,
+      "step": 1183
+    },
+    {
+      "epoch": 2.600768808347062,
+      "grad_norm": 0.04753178358078003,
+      "learning_rate": 4.446254071661238e-05,
+      "loss": 0.6756,
+      "step": 1184
+    },
+    {
+      "epoch": 2.602965403624382,
+      "grad_norm": 0.043900731950998306,
+      "learning_rate": 4.421824104234527e-05,
+      "loss": 0.6985,
+      "step": 1185
+    },
+    {
+      "epoch": 2.6051619989017025,
+      "grad_norm": 0.03854718059301376,
+      "learning_rate": 4.3973941368078175e-05,
+      "loss": 0.7089,
+      "step": 1186
+    },
+    {
+      "epoch": 2.6073585941790225,
+      "grad_norm": 0.0341448150575161,
+      "learning_rate": 4.372964169381107e-05,
+      "loss": 0.5652,
+      "step": 1187
+    },
+    {
+      "epoch": 2.6095551894563425,
+      "grad_norm": 0.04500808194279671,
+      "learning_rate": 4.3485342019543974e-05,
+      "loss": 0.6157,
+      "step": 1188
+    },
+    {
+      "epoch": 2.611751784733663,
+      "grad_norm": 0.050450973212718964,
+      "learning_rate": 4.324104234527687e-05,
+      "loss": 0.5892,
+      "step": 1189
+    },
+    {
+      "epoch": 2.613948380010983,
+      "grad_norm": 0.04859516769647598,
+      "learning_rate": 4.299674267100977e-05,
+      "loss": 0.6381,
+      "step": 1190
+    },
+    {
+      "epoch": 2.6161449752883033,
+      "grad_norm": 0.041921671479940414,
+      "learning_rate": 4.275244299674267e-05,
+      "loss": 0.6333,
+      "step": 1191
+    },
+    {
+      "epoch": 2.618341570565623,
+      "grad_norm": 0.05008811876177788,
+      "learning_rate": 4.250814332247557e-05,
+      "loss": 0.5237,
+      "step": 1192
+    },
+    {
+      "epoch": 2.6205381658429436,
+      "grad_norm": 0.10139350593090057,
+      "learning_rate": 4.226384364820847e-05,
+      "loss": 1.0228,
+      "step": 1193
+    },
+    {
+      "epoch": 2.6227347611202636,
+      "grad_norm": 0.06467008590698242,
+      "learning_rate": 4.2019543973941364e-05,
+      "loss": 0.7578,
+      "step": 1194
+    },
+    {
+      "epoch": 2.6249313563975836,
+      "grad_norm": 0.030447738245129585,
+      "learning_rate": 4.1775244299674266e-05,
+      "loss": 0.4413,
+      "step": 1195
+    },
+    {
+      "epoch": 2.627127951674904,
+      "grad_norm": 0.05087581276893616,
+      "learning_rate": 4.153094462540716e-05,
+      "loss": 0.6297,
+      "step": 1196
+    },
+    {
+      "epoch": 2.629324546952224,
+      "grad_norm": 0.11560789495706558,
+      "learning_rate": 4.1286644951140065e-05,
+      "loss": 0.6206,
+      "step": 1197
+    },
+    {
+      "epoch": 2.6315211422295444,
+      "grad_norm": 0.04665865749120712,
+      "learning_rate": 4.104234527687296e-05,
+      "loss": 0.6543,
+      "step": 1198
+    },
+    {
+      "epoch": 2.6337177375068643,
+      "grad_norm": 0.04100664332509041,
+      "learning_rate": 4.0798045602605864e-05,
+      "loss": 0.5507,
+      "step": 1199
+    },
+    {
+      "epoch": 2.6359143327841847,
+      "grad_norm": 0.05304735153913498,
+      "learning_rate": 4.055374592833876e-05,
+      "loss": 0.6102,
+      "step": 1200
+    },
+    {
+      "epoch": 2.6381109280615047,
+      "grad_norm": 0.04621303454041481,
+      "learning_rate": 4.0309446254071656e-05,
+      "loss": 0.5275,
+      "step": 1201
+    },
+    {
+      "epoch": 2.6403075233388247,
+      "grad_norm": 0.04916003346443176,
+      "learning_rate": 4.006514657980456e-05,
+      "loss": 0.5175,
+      "step": 1202
+    },
+    {
+      "epoch": 2.642504118616145,
+      "grad_norm": 0.056979499757289886,
+      "learning_rate": 3.9820846905537454e-05,
+      "loss": 0.8422,
+      "step": 1203
+    },
+    {
+      "epoch": 2.644700713893465,
+      "grad_norm": 0.05574516952037811,
+      "learning_rate": 3.957654723127036e-05,
+      "loss": 0.635,
+      "step": 1204
+    },
+    {
+      "epoch": 2.646897309170785,
+      "grad_norm": 0.04193108156323433,
+      "learning_rate": 3.933224755700325e-05,
+      "loss": 0.6862,
+      "step": 1205
+    },
+    {
+      "epoch": 2.6490939044481054,
+      "grad_norm": 0.04623300954699516,
+      "learning_rate": 3.9087947882736156e-05,
+      "loss": 0.7349,
+      "step": 1206
+    },
+    {
+      "epoch": 2.651290499725426,
+      "grad_norm": 0.06209337338805199,
+      "learning_rate": 3.884364820846905e-05,
+      "loss": 0.7061,
+      "step": 1207
+    },
+    {
+      "epoch": 2.653487095002746,
+      "grad_norm": 0.05206606909632683,
+      "learning_rate": 3.8599348534201955e-05,
+      "loss": 0.6362,
+      "step": 1208
+    },
+    {
+      "epoch": 2.655683690280066,
+      "grad_norm": 0.06646497547626495,
+      "learning_rate": 3.835504885993485e-05,
+      "loss": 0.5787,
+      "step": 1209
+    },
+    {
+      "epoch": 2.657880285557386,
+      "grad_norm": 0.04821066930890083,
+      "learning_rate": 3.8110749185667747e-05,
+      "loss": 0.8365,
+      "step": 1210
+    },
+    {
+      "epoch": 2.660076880834706,
+      "grad_norm": 0.04882679879665375,
+      "learning_rate": 3.786644951140065e-05,
+      "loss": 0.527,
+      "step": 1211
+    },
+    {
+      "epoch": 2.662273476112026,
+      "grad_norm": 0.05172707512974739,
+      "learning_rate": 3.7622149837133545e-05,
+      "loss": 0.593,
+      "step": 1212
+    },
+    {
+      "epoch": 2.6644700713893466,
+      "grad_norm": 0.04725964367389679,
+      "learning_rate": 3.737785016286645e-05,
+      "loss": 0.7141,
+      "step": 1213
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 0.0511508584022522,
+      "learning_rate": 3.7133550488599344e-05,
+      "loss": 0.5239,
+      "step": 1214
+    },
+    {
+      "epoch": 2.668863261943987,
+      "grad_norm": 0.045084815472364426,
+      "learning_rate": 3.688925081433225e-05,
+      "loss": 0.6523,
+      "step": 1215
+    },
+    {
+      "epoch": 2.671059857221307,
+      "grad_norm": 0.04477803781628609,
+      "learning_rate": 3.664495114006514e-05,
+      "loss": 0.6102,
+      "step": 1216
+    },
+    {
+      "epoch": 2.6732564524986273,
+      "grad_norm": 0.03817501664161682,
+      "learning_rate": 3.640065146579804e-05,
+      "loss": 0.6925,
+      "step": 1217
+    },
+    {
+      "epoch": 2.6754530477759473,
+      "grad_norm": 0.04120345041155815,
+      "learning_rate": 3.615635179153094e-05,
+      "loss": 0.6503,
+      "step": 1218
+    },
+    {
+      "epoch": 2.6776496430532672,
+      "grad_norm": 0.05116845294833183,
+      "learning_rate": 3.591205211726384e-05,
+      "loss": 0.6252,
+      "step": 1219
+    },
+    {
+      "epoch": 2.6798462383305877,
+      "grad_norm": 0.048078861087560654,
+      "learning_rate": 3.566775244299674e-05,
+      "loss": 0.6434,
+      "step": 1220
+    },
+    {
+      "epoch": 2.6820428336079076,
+      "grad_norm": 0.050224531441926956,
+      "learning_rate": 3.5423452768729636e-05,
+      "loss": 0.7836,
+      "step": 1221
+    },
+    {
+      "epoch": 2.684239428885228,
+      "grad_norm": 0.04820210486650467,
+      "learning_rate": 3.517915309446254e-05,
+      "loss": 0.4624,
+      "step": 1222
+    },
+    {
+      "epoch": 2.686436024162548,
+      "grad_norm": 0.04261908680200577,
+      "learning_rate": 3.4934853420195435e-05,
+      "loss": 0.6753,
+      "step": 1223
+    },
+    {
+      "epoch": 2.6886326194398684,
+      "grad_norm": 0.05452783405780792,
+      "learning_rate": 3.469055374592834e-05,
+      "loss": 0.7181,
+      "step": 1224
+    },
+    {
+      "epoch": 2.6908292147171884,
+      "grad_norm": 0.038788363337516785,
+      "learning_rate": 3.4446254071661234e-05,
+      "loss": 0.683,
+      "step": 1225
+    },
+    {
+      "epoch": 2.6930258099945084,
+      "grad_norm": 0.0443214513361454,
+      "learning_rate": 3.420195439739413e-05,
+      "loss": 0.5505,
+      "step": 1226
+    },
+    {
+      "epoch": 2.6952224052718288,
+      "grad_norm": 0.07653029263019562,
+      "learning_rate": 3.395765472312703e-05,
+      "loss": 0.7052,
+      "step": 1227
+    },
+    {
+      "epoch": 2.6974190005491487,
+      "grad_norm": 0.03736657649278641,
+      "learning_rate": 3.371335504885993e-05,
+      "loss": 0.5925,
+      "step": 1228
+    },
+    {
+      "epoch": 2.6996155958264687,
+      "grad_norm": 0.049569420516490936,
+      "learning_rate": 3.346905537459283e-05,
+      "loss": 0.7129,
+      "step": 1229
+    },
+    {
+      "epoch": 2.701812191103789,
+      "grad_norm": 0.04310569539666176,
+      "learning_rate": 3.322475570032573e-05,
+      "loss": 0.5297,
+      "step": 1230
+    },
+    {
+      "epoch": 2.7040087863811095,
+      "grad_norm": 0.032269321382045746,
+      "learning_rate": 3.298045602605863e-05,
+      "loss": 0.5152,
+      "step": 1231
+    },
+    {
+      "epoch": 2.7062053816584295,
+      "grad_norm": 0.04561099037528038,
+      "learning_rate": 3.273615635179153e-05,
+      "loss": 0.7281,
+      "step": 1232
+    },
+    {
+      "epoch": 2.7084019769357495,
+      "grad_norm": 0.045137908309698105,
+      "learning_rate": 3.249185667752443e-05,
+      "loss": 0.6058,
+      "step": 1233
+    },
+    {
+      "epoch": 2.71059857221307,
+      "grad_norm": 0.04589609429240227,
+      "learning_rate": 3.224755700325733e-05,
+      "loss": 0.662,
+      "step": 1234
+    },
+    {
+      "epoch": 2.71279516749039,
+      "grad_norm": 0.03858232870697975,
+      "learning_rate": 3.200325732899023e-05,
+      "loss": 0.745,
+      "step": 1235
+    },
+    {
+      "epoch": 2.71499176276771,
+      "grad_norm": 0.06444483250379562,
+      "learning_rate": 3.175895765472312e-05,
+      "loss": 0.7367,
+      "step": 1236
+    },
+    {
+      "epoch": 2.7171883580450302,
+      "grad_norm": 0.05342705175280571,
+      "learning_rate": 3.1514657980456026e-05,
+      "loss": 0.6469,
+      "step": 1237
+    },
+    {
+      "epoch": 2.71938495332235,
+      "grad_norm": 0.05334605276584625,
+      "learning_rate": 3.127035830618892e-05,
+      "loss": 0.6761,
+      "step": 1238
+    },
+    {
+      "epoch": 2.7215815485996706,
+      "grad_norm": 0.06340347975492477,
+      "learning_rate": 3.1026058631921825e-05,
+      "loss": 0.7222,
+      "step": 1239
+    },
+    {
+      "epoch": 2.7237781438769906,
+      "grad_norm": 0.03992515802383423,
+      "learning_rate": 3.078175895765472e-05,
+      "loss": 0.6016,
+      "step": 1240
+    },
+    {
+      "epoch": 2.725974739154311,
+      "grad_norm": 0.04782949388027191,
+      "learning_rate": 3.0537459283387624e-05,
+      "loss": 0.5176,
+      "step": 1241
+    },
+    {
+      "epoch": 2.728171334431631,
+      "grad_norm": 0.04865286499261856,
+      "learning_rate": 3.029315960912052e-05,
+      "loss": 0.5885,
+      "step": 1242
+    },
+    {
+      "epoch": 2.730367929708951,
+      "grad_norm": 0.04451761022210121,
+      "learning_rate": 3.004885993485342e-05,
+      "loss": 0.7112,
+      "step": 1243
+    },
+    {
+      "epoch": 2.7325645249862713,
+      "grad_norm": 0.057043615728616714,
+      "learning_rate": 2.9804560260586318e-05,
+      "loss": 0.6835,
+      "step": 1244
+    },
+    {
+      "epoch": 2.7347611202635913,
+      "grad_norm": 0.0446697361767292,
+      "learning_rate": 2.9560260586319218e-05,
+      "loss": 0.6343,
+      "step": 1245
+    },
+    {
+      "epoch": 2.7369577155409117,
+      "grad_norm": 0.05795365571975708,
+      "learning_rate": 2.9315960912052114e-05,
+      "loss": 0.6624,
+      "step": 1246
+    },
+    {
+      "epoch": 2.7391543108182317,
+      "grad_norm": 0.05948972329497337,
+      "learning_rate": 2.9071661237785013e-05,
+      "loss": 0.8657,
+      "step": 1247
+    },
+    {
+      "epoch": 2.741350906095552,
+      "grad_norm": 0.045655228197574615,
+      "learning_rate": 2.8827361563517912e-05,
+      "loss": 0.623,
+      "step": 1248
+    },
+    {
+      "epoch": 2.743547501372872,
+      "grad_norm": 0.056424789130687714,
+      "learning_rate": 2.858306188925081e-05,
+      "loss": 0.762,
+      "step": 1249
+    },
+    {
+      "epoch": 2.745744096650192,
+      "grad_norm": 0.04328913614153862,
+      "learning_rate": 2.833876221498371e-05,
+      "loss": 0.6743,
+      "step": 1250
+    },
+    {
+      "epoch": 2.7479406919275124,
+      "grad_norm": 0.043471962213516235,
+      "learning_rate": 2.809446254071661e-05,
+      "loss": 0.7574,
+      "step": 1251
+    },
+    {
+      "epoch": 2.7501372872048324,
+      "grad_norm": 0.057528410106897354,
+      "learning_rate": 2.785016286644951e-05,
+      "loss": 0.6993,
+      "step": 1252
+    },
+    {
+      "epoch": 2.752333882482153,
+      "grad_norm": 0.046321526169776917,
+      "learning_rate": 2.760586319218241e-05,
+      "loss": 0.5547,
+      "step": 1253
+    },
+    {
+      "epoch": 2.754530477759473,
+      "grad_norm": 0.042450498789548874,
+      "learning_rate": 2.7361563517915305e-05,
+      "loss": 0.5814,
+      "step": 1254
+    },
+    {
+      "epoch": 2.756727073036793,
+      "grad_norm": 0.06893017143011093,
+      "learning_rate": 2.7117263843648204e-05,
+      "loss": 0.7708,
+      "step": 1255
+    },
+    {
+      "epoch": 2.758923668314113,
+      "grad_norm": 0.04471489042043686,
+      "learning_rate": 2.6872964169381104e-05,
+      "loss": 0.4695,
+      "step": 1256
+    },
+    {
+      "epoch": 2.761120263591433,
+      "grad_norm": 0.051368217915296555,
+      "learning_rate": 2.6628664495114003e-05,
+      "loss": 0.6192,
+      "step": 1257
+    },
+    {
+      "epoch": 2.7633168588687536,
+      "grad_norm": 0.04400815814733505,
+      "learning_rate": 2.6384364820846903e-05,
+      "loss": 0.5535,
+      "step": 1258
+    },
+    {
+      "epoch": 2.7655134541460735,
+      "grad_norm": 0.05980153754353523,
+      "learning_rate": 2.6140065146579802e-05,
+      "loss": 0.7952,
+      "step": 1259
+    },
+    {
+      "epoch": 2.7677100494233935,
+      "grad_norm": 0.04607835412025452,
+      "learning_rate": 2.58957654723127e-05,
+      "loss": 0.6893,
+      "step": 1260
+    },
+    {
+      "epoch": 2.769906644700714,
+      "grad_norm": 0.05787874385714531,
+      "learning_rate": 2.56514657980456e-05,
+      "loss": 0.6042,
+      "step": 1261
+    },
+    {
+      "epoch": 2.7721032399780343,
+      "grad_norm": 0.05453767254948616,
+      "learning_rate": 2.5407166123778497e-05,
+      "loss": 0.6614,
+      "step": 1262
+    },
+    {
+      "epoch": 2.7742998352553543,
+      "grad_norm": 0.05665965378284454,
+      "learning_rate": 2.5162866449511396e-05,
+      "loss": 0.7342,
+      "step": 1263
+    },
+    {
+      "epoch": 2.7764964305326743,
+      "grad_norm": 0.03880994766950607,
+      "learning_rate": 2.4918566775244295e-05,
+      "loss": 0.7974,
+      "step": 1264
+    },
+    {
+      "epoch": 2.7786930258099947,
+      "grad_norm": 0.039176661521196365,
+      "learning_rate": 2.4674267100977195e-05,
+      "loss": 0.7382,
+      "step": 1265
+    },
+    {
+      "epoch": 2.7808896210873146,
+      "grad_norm": 0.036031998693943024,
+      "learning_rate": 2.4429967426710094e-05,
+      "loss": 0.5444,
+      "step": 1266
+    },
+    {
+      "epoch": 2.7830862163646346,
+      "grad_norm": 0.04688276723027229,
+      "learning_rate": 2.4185667752442993e-05,
+      "loss": 0.5756,
+      "step": 1267
+    },
+    {
+      "epoch": 2.785282811641955,
+      "grad_norm": 0.04207608476281166,
+      "learning_rate": 2.3941368078175893e-05,
+      "loss": 0.6617,
+      "step": 1268
+    },
+    {
+      "epoch": 2.787479406919275,
+      "grad_norm": 0.053132861852645874,
+      "learning_rate": 2.3697068403908792e-05,
+      "loss": 0.5698,
+      "step": 1269
+    },
+    {
+      "epoch": 2.7896760021965954,
+      "grad_norm": 0.04987029731273651,
+      "learning_rate": 2.3452768729641688e-05,
+      "loss": 0.6068,
+      "step": 1270
+    },
+    {
+      "epoch": 2.7918725974739154,
+      "grad_norm": 0.04091858118772507,
+      "learning_rate": 2.3208469055374594e-05,
+      "loss": 0.5744,
+      "step": 1271
+    },
+    {
+      "epoch": 2.7940691927512358,
+      "grad_norm": 0.04796380549669266,
+      "learning_rate": 2.296416938110749e-05,
+      "loss": 0.6778,
+      "step": 1272
+    },
+    {
+      "epoch": 2.7962657880285557,
+      "grad_norm": 0.05361456796526909,
+      "learning_rate": 2.271986970684039e-05,
+      "loss": 0.6834,
+      "step": 1273
+    },
+    {
+      "epoch": 2.7984623833058757,
+      "grad_norm": 0.04959200695157051,
+      "learning_rate": 2.247557003257329e-05,
+      "loss": 0.8341,
+      "step": 1274
+    },
+    {
+      "epoch": 2.800658978583196,
+      "grad_norm": 0.0428910106420517,
+      "learning_rate": 2.223127035830619e-05,
+      "loss": 0.6286,
+      "step": 1275
+    },
+    {
+      "epoch": 2.802855573860516,
+      "grad_norm": 0.05645201355218887,
+      "learning_rate": 2.1986970684039088e-05,
+      "loss": 0.8027,
+      "step": 1276
+    },
+    {
+      "epoch": 2.8050521691378365,
+      "grad_norm": 0.059787213802337646,
+      "learning_rate": 2.1742671009771987e-05,
+      "loss": 0.8865,
+      "step": 1277
+    },
+    {
+      "epoch": 2.8072487644151565,
+      "grad_norm": 0.04704824462532997,
+      "learning_rate": 2.1498371335504886e-05,
+      "loss": 0.731,
+      "step": 1278
+    },
+    {
+      "epoch": 2.809445359692477,
+      "grad_norm": 0.04552409425377846,
+      "learning_rate": 2.1254071661237786e-05,
+      "loss": 0.6338,
+      "step": 1279
+    },
+    {
+      "epoch": 2.811641954969797,
+      "grad_norm": 0.04034719616174698,
+      "learning_rate": 2.1009771986970682e-05,
+      "loss": 0.6228,
+      "step": 1280
+    },
+    {
+      "epoch": 2.813838550247117,
+      "grad_norm": 0.05566655471920967,
+      "learning_rate": 2.076547231270358e-05,
+      "loss": 0.6912,
+      "step": 1281
+    },
+    {
+      "epoch": 2.8160351455244372,
+      "grad_norm": 0.05668025091290474,
+      "learning_rate": 2.052117263843648e-05,
+      "loss": 0.7036,
+      "step": 1282
+    },
+    {
+      "epoch": 2.818231740801757,
+      "grad_norm": 0.04943126440048218,
+      "learning_rate": 2.027687296416938e-05,
+      "loss": 0.7858,
+      "step": 1283
+    },
+    {
+      "epoch": 2.820428336079077,
+      "grad_norm": 0.03997279703617096,
+      "learning_rate": 2.003257328990228e-05,
+      "loss": 0.7082,
+      "step": 1284
+    },
+    {
+      "epoch": 2.8226249313563976,
+      "grad_norm": 0.050963182002305984,
+      "learning_rate": 1.978827361563518e-05,
+      "loss": 0.6874,
+      "step": 1285
+    },
+    {
+      "epoch": 2.824821526633718,
+      "grad_norm": 0.051822926849126816,
+      "learning_rate": 1.9543973941368078e-05,
+      "loss": 0.8515,
+      "step": 1286
+    },
+    {
+      "epoch": 2.827018121911038,
+      "grad_norm": 0.03817344456911087,
+      "learning_rate": 1.9299674267100977e-05,
+      "loss": 0.5715,
+      "step": 1287
+    },
+    {
+      "epoch": 2.829214717188358,
+      "grad_norm": 0.05846104770898819,
+      "learning_rate": 1.9055374592833873e-05,
+      "loss": 0.6425,
+      "step": 1288
+    },
+    {
+      "epoch": 2.8314113124656783,
+      "grad_norm": 0.04724184796214104,
+      "learning_rate": 1.8811074918566773e-05,
+      "loss": 0.6654,
+      "step": 1289
+    },
+    {
+      "epoch": 2.8336079077429983,
+      "grad_norm": 0.05438625440001488,
+      "learning_rate": 1.8566775244299672e-05,
+      "loss": 0.7263,
+      "step": 1290
+    },
+    {
+      "epoch": 2.8358045030203183,
+      "grad_norm": 0.0520639531314373,
+      "learning_rate": 1.832247557003257e-05,
+      "loss": 0.5943,
+      "step": 1291
+    },
+    {
+      "epoch": 2.8380010982976387,
+      "grad_norm": 0.043778836727142334,
+      "learning_rate": 1.807817589576547e-05,
+      "loss": 0.6249,
+      "step": 1292
+    },
+    {
+      "epoch": 2.8401976935749587,
+      "grad_norm": 0.04170997813344002,
+      "learning_rate": 1.783387622149837e-05,
+      "loss": 0.4842,
+      "step": 1293
+    },
+    {
+      "epoch": 2.842394288852279,
+      "grad_norm": 0.048355937004089355,
+      "learning_rate": 1.758957654723127e-05,
+      "loss": 0.7141,
+      "step": 1294
+    },
+    {
+      "epoch": 2.844590884129599,
+      "grad_norm": 0.05793530493974686,
+      "learning_rate": 1.734527687296417e-05,
+      "loss": 0.7181,
+      "step": 1295
+    },
+    {
+      "epoch": 2.8467874794069195,
+      "grad_norm": 0.049781594425439835,
+      "learning_rate": 1.7100977198697065e-05,
+      "loss": 0.7524,
+      "step": 1296
+    },
+    {
+      "epoch": 2.8489840746842394,
+      "grad_norm": 0.034993454813957214,
+      "learning_rate": 1.6856677524429964e-05,
+      "loss": 0.5538,
+      "step": 1297
+    },
+    {
+      "epoch": 2.8511806699615594,
+      "grad_norm": 0.05010677129030228,
+      "learning_rate": 1.6612377850162864e-05,
+      "loss": 0.7309,
+      "step": 1298
+    },
+    {
+      "epoch": 2.85337726523888,
+      "grad_norm": 0.05319371074438095,
+      "learning_rate": 1.6368078175895766e-05,
+      "loss": 0.662,
+      "step": 1299
+    },
+    {
+      "epoch": 2.8555738605161998,
+      "grad_norm": 0.06000995635986328,
+      "learning_rate": 1.6123778501628666e-05,
+      "loss": 0.8406,
+      "step": 1300
+    },
+    {
+      "epoch": 2.85777045579352,
+      "grad_norm": 0.03820439800620079,
+      "learning_rate": 1.587947882736156e-05,
+      "loss": 0.6687,
+      "step": 1301
+    },
+    {
+      "epoch": 2.85996705107084,
+      "grad_norm": 0.05080217868089676,
+      "learning_rate": 1.563517915309446e-05,
+      "loss": 0.5845,
+      "step": 1302
+    },
+    {
+      "epoch": 2.8621636463481606,
+      "grad_norm": 0.04689435660839081,
+      "learning_rate": 1.539087947882736e-05,
+      "loss": 0.8441,
+      "step": 1303
+    },
+    {
+      "epoch": 2.8643602416254805,
+      "grad_norm": 0.05374585837125778,
+      "learning_rate": 1.514657980456026e-05,
+      "loss": 0.5995,
+      "step": 1304
+    },
+    {
+      "epoch": 2.8665568369028005,
+      "grad_norm": 0.044032514095306396,
+      "learning_rate": 1.4902280130293159e-05,
+      "loss": 0.6525,
+      "step": 1305
+    },
+    {
+      "epoch": 2.868753432180121,
+      "grad_norm": 0.041226308792829514,
+      "learning_rate": 1.4657980456026057e-05,
+      "loss": 0.6631,
+      "step": 1306
+    },
+    {
+      "epoch": 2.870950027457441,
+      "grad_norm": 0.046438008546829224,
+      "learning_rate": 1.4413680781758956e-05,
+      "loss": 0.711,
+      "step": 1307
+    },
+    {
+      "epoch": 2.873146622734761,
+      "grad_norm": 0.04149239510297775,
+      "learning_rate": 1.4169381107491856e-05,
+      "loss": 0.5172,
+      "step": 1308
+    },
+    {
+      "epoch": 2.8753432180120813,
+      "grad_norm": 0.04777297005057335,
+      "learning_rate": 1.3925081433224755e-05,
+      "loss": 0.6341,
+      "step": 1309
+    },
+    {
+      "epoch": 2.8775398132894017,
+      "grad_norm": 0.05444321781396866,
+      "learning_rate": 1.3680781758957653e-05,
+      "loss": 0.7063,
+      "step": 1310
+    },
+    {
+      "epoch": 2.8797364085667216,
+      "grad_norm": 0.05217491462826729,
+      "learning_rate": 1.3436482084690552e-05,
+      "loss": 0.6434,
+      "step": 1311
+    },
+    {
+      "epoch": 2.8819330038440416,
+      "grad_norm": 0.045649029314517975,
+      "learning_rate": 1.3192182410423451e-05,
+      "loss": 0.6995,
+      "step": 1312
+    },
+    {
+      "epoch": 2.884129599121362,
+      "grad_norm": 0.059278231114149094,
+      "learning_rate": 1.294788273615635e-05,
+      "loss": 0.7023,
+      "step": 1313
+    },
+    {
+      "epoch": 2.886326194398682,
+      "grad_norm": 0.0433654710650444,
+      "learning_rate": 1.2703583061889248e-05,
+      "loss": 0.675,
+      "step": 1314
+    },
+    {
+      "epoch": 2.888522789676002,
+      "grad_norm": 0.05471406131982803,
+      "learning_rate": 1.2459283387622148e-05,
+      "loss": 0.6003,
+      "step": 1315
+    },
+    {
+      "epoch": 2.8907193849533224,
+      "grad_norm": 0.05296127498149872,
+      "learning_rate": 1.2214983713355047e-05,
+      "loss": 0.686,
+      "step": 1316
+    },
+    {
+      "epoch": 2.892915980230643,
+      "grad_norm": 0.05806552991271019,
+      "learning_rate": 1.1970684039087946e-05,
+      "loss": 0.5582,
+      "step": 1317
+    },
+    {
+      "epoch": 2.8951125755079627,
+      "grad_norm": 0.054739389568567276,
+      "learning_rate": 1.1726384364820844e-05,
+      "loss": 0.5321,
+      "step": 1318
+    },
+    {
+      "epoch": 2.8973091707852827,
+      "grad_norm": 0.040236715227365494,
+      "learning_rate": 1.1482084690553745e-05,
+      "loss": 0.5278,
+      "step": 1319
+    },
+    {
+      "epoch": 2.899505766062603,
+      "grad_norm": 0.05634508281946182,
+      "learning_rate": 1.1237785016286644e-05,
+      "loss": 0.6725,
+      "step": 1320
+    },
+    {
+      "epoch": 2.901702361339923,
+      "grad_norm": 0.04822323098778725,
+      "learning_rate": 1.0993485342019544e-05,
+      "loss": 0.4929,
+      "step": 1321
+    },
+    {
+      "epoch": 2.903898956617243,
+      "grad_norm": 0.06421532481908798,
+      "learning_rate": 1.0749185667752443e-05,
+      "loss": 0.7244,
+      "step": 1322
+    },
+    {
+      "epoch": 2.9060955518945635,
+      "grad_norm": 0.05510374531149864,
+      "learning_rate": 1.0504885993485341e-05,
+      "loss": 0.6058,
+      "step": 1323
+    },
+    {
+      "epoch": 2.9082921471718834,
+      "grad_norm": 0.05010756105184555,
+      "learning_rate": 1.026058631921824e-05,
+      "loss": 0.6489,
+      "step": 1324
+    },
+    {
+      "epoch": 2.910488742449204,
+      "grad_norm": 0.05712300166487694,
+      "learning_rate": 1.001628664495114e-05,
+      "loss": 0.5218,
+      "step": 1325
+    },
+    {
+      "epoch": 2.912685337726524,
+      "grad_norm": 0.05162881687283516,
+      "learning_rate": 9.771986970684039e-06,
+      "loss": 0.6576,
+      "step": 1326
+    },
+    {
+      "epoch": 2.9148819330038442,
+      "grad_norm": 0.04074005410075188,
+      "learning_rate": 9.527687296416937e-06,
+      "loss": 0.6519,
+      "step": 1327
+    },
+    {
+      "epoch": 2.917078528281164,
+      "grad_norm": 0.040444131940603256,
+      "learning_rate": 9.283387622149836e-06,
+      "loss": 0.4823,
+      "step": 1328
+    },
+    {
+      "epoch": 2.919275123558484,
+      "grad_norm": 0.04682661220431328,
+      "learning_rate": 9.039087947882735e-06,
+      "loss": 0.4372,
+      "step": 1329
+    },
+    {
+      "epoch": 2.9214717188358046,
+      "grad_norm": 0.05561075732111931,
+      "learning_rate": 8.794788273615635e-06,
+      "loss": 0.5898,
+      "step": 1330
+    },
+    {
+      "epoch": 2.9236683141131246,
+      "grad_norm": 0.04917816445231438,
+      "learning_rate": 8.550488599348532e-06,
+      "loss": 0.756,
+      "step": 1331
+    },
+    {
+      "epoch": 2.925864909390445,
+      "grad_norm": 0.035087183117866516,
+      "learning_rate": 8.306188925081432e-06,
+      "loss": 0.367,
+      "step": 1332
+    },
+    {
+      "epoch": 2.928061504667765,
+      "grad_norm": 0.07068229466676712,
+      "learning_rate": 8.061889250814333e-06,
+      "loss": 0.6114,
+      "step": 1333
+    },
+    {
+      "epoch": 2.9302580999450853,
+      "grad_norm": 0.03798624500632286,
+      "learning_rate": 7.81758957654723e-06,
+      "loss": 0.5505,
+      "step": 1334
+    },
+    {
+      "epoch": 2.9324546952224053,
+      "grad_norm": 0.04883285611867905,
+      "learning_rate": 7.57328990228013e-06,
+      "loss": 0.5005,
+      "step": 1335
+    },
+    {
+      "epoch": 2.9346512904997253,
+      "grad_norm": 0.044976841658353806,
+      "learning_rate": 7.328990228013028e-06,
+      "loss": 0.5555,
+      "step": 1336
+    },
+    {
+      "epoch": 2.9368478857770457,
+      "grad_norm": 0.03776836767792702,
+      "learning_rate": 7.084690553745928e-06,
+      "loss": 0.5256,
+      "step": 1337
+    },
+    {
+      "epoch": 2.9390444810543657,
+      "grad_norm": 0.042885709553956985,
+      "learning_rate": 6.840390879478826e-06,
+      "loss": 0.5995,
+      "step": 1338
+    },
+    {
+      "epoch": 2.9412410763316856,
+      "grad_norm": 0.040374260395765305,
+      "learning_rate": 6.596091205211726e-06,
+      "loss": 0.6082,
+      "step": 1339
+    },
+    {
+      "epoch": 2.943437671609006,
+      "grad_norm": 0.04279405623674393,
+      "learning_rate": 6.351791530944624e-06,
+      "loss": 0.4615,
+      "step": 1340
+    },
+    {
+      "epoch": 2.9456342668863265,
+      "grad_norm": 0.039576370269060135,
+      "learning_rate": 6.1074918566775235e-06,
+      "loss": 0.6118,
+      "step": 1341
+    },
+    {
+      "epoch": 2.9478308621636464,
+      "grad_norm": 0.04402873292565346,
+      "learning_rate": 5.863192182410422e-06,
+      "loss": 0.681,
+      "step": 1342
+    },
+    {
+      "epoch": 2.9500274574409664,
+      "grad_norm": 0.044643841683864594,
+      "learning_rate": 5.618892508143322e-06,
+      "loss": 0.7758,
+      "step": 1343
+    },
+    {
+      "epoch": 2.952224052718287,
+      "grad_norm": 0.05211080610752106,
+      "learning_rate": 5.374592833876222e-06,
+      "loss": 0.7354,
+      "step": 1344
+    },
+    {
+      "epoch": 2.9544206479956068,
+      "grad_norm": 0.046959247440099716,
+      "learning_rate": 5.13029315960912e-06,
+      "loss": 0.6511,
+      "step": 1345
+    },
+    {
+      "epoch": 2.9566172432729267,
+      "grad_norm": 0.05532398819923401,
+      "learning_rate": 4.8859934853420195e-06,
+      "loss": 0.5952,
+      "step": 1346
+    },
+    {
+      "epoch": 2.958813838550247,
+      "grad_norm": 0.04401236027479172,
+      "learning_rate": 4.641693811074918e-06,
+      "loss": 0.526,
+      "step": 1347
+    },
+    {
+      "epoch": 2.961010433827567,
+      "grad_norm": 0.054430510848760605,
+      "learning_rate": 4.397394136807817e-06,
+      "loss": 0.5618,
+      "step": 1348
+    },
+    {
+      "epoch": 2.9632070291048875,
+      "grad_norm": 0.06004808843135834,
+      "learning_rate": 4.153094462540716e-06,
+      "loss": 0.8459,
+      "step": 1349
+    },
+    {
+      "epoch": 2.9654036243822075,
+      "grad_norm": 0.05838881433010101,
+      "learning_rate": 3.908794788273615e-06,
+      "loss": 0.5954,
+      "step": 1350
+    },
+    {
+      "epoch": 2.967600219659528,
+      "grad_norm": 0.04087888449430466,
+      "learning_rate": 3.664495114006514e-06,
+      "loss": 0.749,
+      "step": 1351
+    },
+    {
+      "epoch": 2.969796814936848,
+      "grad_norm": 0.05262356251478195,
+      "learning_rate": 3.420195439739413e-06,
+      "loss": 0.6423,
+      "step": 1352
+    },
+    {
+      "epoch": 2.971993410214168,
+      "grad_norm": 0.04079955816268921,
+      "learning_rate": 3.175895765472312e-06,
+      "loss": 0.6893,
+      "step": 1353
+    },
+    {
+      "epoch": 2.9741900054914883,
+      "grad_norm": 0.049717098474502563,
+      "learning_rate": 2.931596091205211e-06,
+      "loss": 0.7561,
+      "step": 1354
+    },
+    {
+      "epoch": 2.9763866007688082,
+      "grad_norm": 0.042594753205776215,
+      "learning_rate": 2.687296416938111e-06,
+      "loss": 0.5603,
+      "step": 1355
+    },
+    {
+      "epoch": 2.9785831960461286,
+      "grad_norm": 0.0829310268163681,
+      "learning_rate": 2.4429967426710097e-06,
+      "loss": 0.6953,
+      "step": 1356
+    },
+    {
+      "epoch": 2.9807797913234486,
+      "grad_norm": 0.05175120010972023,
+      "learning_rate": 2.1986970684039087e-06,
+      "loss": 0.6679,
+      "step": 1357
+    },
+    {
+      "epoch": 2.982976386600769,
+      "grad_norm": 0.05891898646950722,
+      "learning_rate": 1.9543973941368076e-06,
+      "loss": 0.5024,
+      "step": 1358
+    },
+    {
+      "epoch": 2.985172981878089,
+      "grad_norm": 0.045325834304094315,
+      "learning_rate": 1.7100977198697066e-06,
+      "loss": 0.7896,
+      "step": 1359
+    },
+    {
+      "epoch": 2.987369577155409,
+      "grad_norm": 0.04623181000351906,
+      "learning_rate": 1.4657980456026055e-06,
+      "loss": 0.5756,
+      "step": 1360
+    },
+    {
+      "epoch": 2.9895661724327294,
+      "grad_norm": 0.041907262057065964,
+      "learning_rate": 1.2214983713355049e-06,
+      "loss": 0.6516,
+      "step": 1361
+    },
+    {
+      "epoch": 2.9917627677100493,
+      "grad_norm": 0.07447605580091476,
+      "learning_rate": 9.771986970684038e-07,
+      "loss": 0.7468,
+      "step": 1362
+    },
+    {
+      "epoch": 2.9939593629873693,
+      "grad_norm": 0.05689909681677818,
+      "learning_rate": 7.328990228013028e-07,
+      "loss": 0.6006,
+      "step": 1363
+    },
+    {
+      "epoch": 2.9961559582646897,
+      "grad_norm": 0.04323657602071762,
+      "learning_rate": 4.885993485342019e-07,
+      "loss": 0.6885,
+      "step": 1364
+    },
+    {
+      "epoch": 2.99835255354201,
+      "grad_norm": 0.04761827364563942,
+      "learning_rate": 2.4429967426710095e-07,
+      "loss": 0.7291,
+      "step": 1365
+    },
+    {
+      "epoch": 2.99835255354201,
+      "step": 1365,
+      "total_flos": 1.0038997179244216e+18,
+      "train_loss": 0.6840552445296403,
+      "train_runtime": 5408.1966,
+      "train_samples_per_second": 4.04,
+      "train_steps_per_second": 0.252
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1365,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.0038997179244216e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}