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": 0.19471493363380432,
+      "learning_rate": 0.0,
+      "loss": 1.0369,
+      "step": 1
+    },
+    {
+      "epoch": 0.004393190554640308,
+      "grad_norm": 0.15064771473407745,
+      "learning_rate": 2.18978102189781e-06,
+      "loss": 1.2693,
+      "step": 2
+    },
+    {
+      "epoch": 0.006589785831960461,
+      "grad_norm": 0.20301605761051178,
+      "learning_rate": 4.37956204379562e-06,
+      "loss": 1.0009,
+      "step": 3
+    },
+    {
+      "epoch": 0.008786381109280615,
+      "grad_norm": 0.12252406030893326,
+      "learning_rate": 6.56934306569343e-06,
+      "loss": 0.9181,
+      "step": 4
+    },
+    {
+      "epoch": 0.010982976386600769,
+      "grad_norm": 0.11816470324993134,
+      "learning_rate": 8.75912408759124e-06,
+      "loss": 0.9776,
+      "step": 5
+    },
+    {
+      "epoch": 0.013179571663920923,
+      "grad_norm": 0.14238254725933075,
+      "learning_rate": 1.0948905109489049e-05,
+      "loss": 0.8766,
+      "step": 6
+    },
+    {
+      "epoch": 0.015376166941241077,
+      "grad_norm": 0.17792834341526031,
+      "learning_rate": 1.313868613138686e-05,
+      "loss": 0.8325,
+      "step": 7
+    },
+    {
+      "epoch": 0.01757276221856123,
+      "grad_norm": 0.15998893976211548,
+      "learning_rate": 1.532846715328467e-05,
+      "loss": 0.8443,
+      "step": 8
+    },
+    {
+      "epoch": 0.019769357495881382,
+      "grad_norm": 0.3409581482410431,
+      "learning_rate": 1.751824817518248e-05,
+      "loss": 1.2307,
+      "step": 9
+    },
+    {
+      "epoch": 0.021965952773201538,
+      "grad_norm": 0.1347101777791977,
+      "learning_rate": 1.9708029197080292e-05,
+      "loss": 0.8755,
+      "step": 10
+    },
+    {
+      "epoch": 0.02416254805052169,
+      "grad_norm": 0.22755971550941467,
+      "learning_rate": 2.1897810218978098e-05,
+      "loss": 1.0875,
+      "step": 11
+    },
+    {
+      "epoch": 0.026359143327841845,
+      "grad_norm": 0.1632407158613205,
+      "learning_rate": 2.408759124087591e-05,
+      "loss": 0.9505,
+      "step": 12
+    },
+    {
+      "epoch": 0.028555738605161998,
+      "grad_norm": 0.16671378910541534,
+      "learning_rate": 2.627737226277372e-05,
+      "loss": 0.9337,
+      "step": 13
+    },
+    {
+      "epoch": 0.030752333882482153,
+      "grad_norm": 0.15404529869556427,
+      "learning_rate": 2.846715328467153e-05,
+      "loss": 0.9832,
+      "step": 14
+    },
+    {
+      "epoch": 0.032948929159802305,
+      "grad_norm": 0.12389346957206726,
+      "learning_rate": 3.065693430656934e-05,
+      "loss": 1.0623,
+      "step": 15
+    },
+    {
+      "epoch": 0.03514552443712246,
+      "grad_norm": 0.13049551844596863,
+      "learning_rate": 3.284671532846715e-05,
+      "loss": 0.8008,
+      "step": 16
+    },
+    {
+      "epoch": 0.037342119714442616,
+      "grad_norm": 0.1575450897216797,
+      "learning_rate": 3.503649635036496e-05,
+      "loss": 0.9346,
+      "step": 17
+    },
+    {
+      "epoch": 0.039538714991762765,
+      "grad_norm": 0.1810751110315323,
+      "learning_rate": 3.722627737226277e-05,
+      "loss": 1.0535,
+      "step": 18
+    },
+    {
+      "epoch": 0.04173531026908292,
+      "grad_norm": 0.14116878807544708,
+      "learning_rate": 3.9416058394160584e-05,
+      "loss": 1.0588,
+      "step": 19
+    },
+    {
+      "epoch": 0.043931905546403076,
+      "grad_norm": 0.14644716680049896,
+      "learning_rate": 4.160583941605839e-05,
+      "loss": 0.9568,
+      "step": 20
+    },
+    {
+      "epoch": 0.04612850082372323,
+      "grad_norm": 0.14529097080230713,
+      "learning_rate": 4.3795620437956196e-05,
+      "loss": 1.0274,
+      "step": 21
+    },
+    {
+      "epoch": 0.04832509610104338,
+      "grad_norm": 0.18500636518001556,
+      "learning_rate": 4.5985401459854016e-05,
+      "loss": 0.9344,
+      "step": 22
+    },
+    {
+      "epoch": 0.050521691378363535,
+      "grad_norm": 0.19902238249778748,
+      "learning_rate": 4.817518248175182e-05,
+      "loss": 1.1168,
+      "step": 23
+    },
+    {
+      "epoch": 0.05271828665568369,
+      "grad_norm": 0.17660082876682281,
+      "learning_rate": 5.036496350364963e-05,
+      "loss": 0.9788,
+      "step": 24
+    },
+    {
+      "epoch": 0.054914881933003847,
+      "grad_norm": 0.16928720474243164,
+      "learning_rate": 5.255474452554744e-05,
+      "loss": 0.9694,
+      "step": 25
+    },
+    {
+      "epoch": 0.057111477210323995,
+      "grad_norm": 0.2330097258090973,
+      "learning_rate": 5.4744525547445253e-05,
+      "loss": 0.9648,
+      "step": 26
+    },
+    {
+      "epoch": 0.05930807248764415,
+      "grad_norm": 0.18641650676727295,
+      "learning_rate": 5.693430656934306e-05,
+      "loss": 1.0088,
+      "step": 27
+    },
+    {
+      "epoch": 0.061504667764964306,
+      "grad_norm": 0.14449548721313477,
+      "learning_rate": 5.912408759124087e-05,
+      "loss": 0.9229,
+      "step": 28
+    },
+    {
+      "epoch": 0.06370126304228446,
+      "grad_norm": 0.1297532469034195,
+      "learning_rate": 6.131386861313868e-05,
+      "loss": 0.8736,
+      "step": 29
+    },
+    {
+      "epoch": 0.06589785831960461,
+      "grad_norm": 0.15038494765758514,
+      "learning_rate": 6.350364963503648e-05,
+      "loss": 0.9305,
+      "step": 30
+    },
+    {
+      "epoch": 0.06809445359692477,
+      "grad_norm": 0.17437539994716644,
+      "learning_rate": 6.56934306569343e-05,
+      "loss": 0.9413,
+      "step": 31
+    },
+    {
+      "epoch": 0.07029104887424492,
+      "grad_norm": 0.1912246197462082,
+      "learning_rate": 6.788321167883211e-05,
+      "loss": 1.2184,
+      "step": 32
+    },
+    {
+      "epoch": 0.07248764415156507,
+      "grad_norm": 0.15581797063350677,
+      "learning_rate": 7.007299270072992e-05,
+      "loss": 1.3594,
+      "step": 33
+    },
+    {
+      "epoch": 0.07468423942888523,
+      "grad_norm": 0.18623195588588715,
+      "learning_rate": 7.226277372262774e-05,
+      "loss": 1.0484,
+      "step": 34
+    },
+    {
+      "epoch": 0.07688083470620538,
+      "grad_norm": 0.1680649071931839,
+      "learning_rate": 7.445255474452554e-05,
+      "loss": 0.9843,
+      "step": 35
+    },
+    {
+      "epoch": 0.07907742998352553,
+      "grad_norm": 0.16884709894657135,
+      "learning_rate": 7.664233576642336e-05,
+      "loss": 1.1218,
+      "step": 36
+    },
+    {
+      "epoch": 0.08127402526084569,
+      "grad_norm": 0.12885382771492004,
+      "learning_rate": 7.883211678832117e-05,
+      "loss": 1.0374,
+      "step": 37
+    },
+    {
+      "epoch": 0.08347062053816584,
+      "grad_norm": 0.1769813746213913,
+      "learning_rate": 8.102189781021897e-05,
+      "loss": 1.1467,
+      "step": 38
+    },
+    {
+      "epoch": 0.085667215815486,
+      "grad_norm": 0.16934935748577118,
+      "learning_rate": 8.321167883211678e-05,
+      "loss": 1.1699,
+      "step": 39
+    },
+    {
+      "epoch": 0.08786381109280615,
+      "grad_norm": 0.1536577045917511,
+      "learning_rate": 8.540145985401459e-05,
+      "loss": 0.8951,
+      "step": 40
+    },
+    {
+      "epoch": 0.0900604063701263,
+      "grad_norm": 0.19077053666114807,
+      "learning_rate": 8.759124087591239e-05,
+      "loss": 1.0106,
+      "step": 41
+    },
+    {
+      "epoch": 0.09225700164744646,
+      "grad_norm": 0.18641969561576843,
+      "learning_rate": 8.978102189781021e-05,
+      "loss": 1.1178,
+      "step": 42
+    },
+    {
+      "epoch": 0.09445359692476661,
+      "grad_norm": 0.19781143963336945,
+      "learning_rate": 9.197080291970803e-05,
+      "loss": 1.0591,
+      "step": 43
+    },
+    {
+      "epoch": 0.09665019220208676,
+      "grad_norm": 0.20191045105457306,
+      "learning_rate": 9.416058394160584e-05,
+      "loss": 1.1224,
+      "step": 44
+    },
+    {
+      "epoch": 0.09884678747940692,
+      "grad_norm": 0.1581476479768753,
+      "learning_rate": 9.635036496350364e-05,
+      "loss": 1.0658,
+      "step": 45
+    },
+    {
+      "epoch": 0.10104338275672707,
+      "grad_norm": 0.1604839414358139,
+      "learning_rate": 9.854014598540145e-05,
+      "loss": 0.9564,
+      "step": 46
+    },
+    {
+      "epoch": 0.10323997803404723,
+      "grad_norm": 0.1293153613805771,
+      "learning_rate": 0.00010072992700729926,
+      "loss": 0.9737,
+      "step": 47
+    },
+    {
+      "epoch": 0.10543657331136738,
+      "grad_norm": 0.19610147178173065,
+      "learning_rate": 0.00010291970802919708,
+      "loss": 1.0529,
+      "step": 48
+    },
+    {
+      "epoch": 0.10763316858868753,
+      "grad_norm": 0.2109193503856659,
+      "learning_rate": 0.00010510948905109488,
+      "loss": 1.1856,
+      "step": 49
+    },
+    {
+      "epoch": 0.10982976386600769,
+      "grad_norm": 0.21855053305625916,
+      "learning_rate": 0.00010729927007299269,
+      "loss": 1.0799,
+      "step": 50
+    },
+    {
+      "epoch": 0.11202635914332784,
+      "grad_norm": 0.1673671007156372,
+      "learning_rate": 0.00010948905109489051,
+      "loss": 1.0506,
+      "step": 51
+    },
+    {
+      "epoch": 0.11422295442064799,
+      "grad_norm": 0.2702706456184387,
+      "learning_rate": 0.00011167883211678831,
+      "loss": 0.9496,
+      "step": 52
+    },
+    {
+      "epoch": 0.11641954969796815,
+      "grad_norm": 0.1377696841955185,
+      "learning_rate": 0.00011386861313868612,
+      "loss": 1.0976,
+      "step": 53
+    },
+    {
+      "epoch": 0.1186161449752883,
+      "grad_norm": 0.1439305990934372,
+      "learning_rate": 0.00011605839416058394,
+      "loss": 0.8003,
+      "step": 54
+    },
+    {
+      "epoch": 0.12081274025260846,
+      "grad_norm": 0.1528768092393875,
+      "learning_rate": 0.00011824817518248174,
+      "loss": 0.96,
+      "step": 55
+    },
+    {
+      "epoch": 0.12300933552992861,
+      "grad_norm": 0.20019182562828064,
+      "learning_rate": 0.00012043795620437955,
+      "loss": 1.2285,
+      "step": 56
+    },
+    {
+      "epoch": 0.12520593080724876,
+      "grad_norm": 0.13644832372665405,
+      "learning_rate": 0.00012262773722627736,
+      "loss": 0.8608,
+      "step": 57
+    },
+    {
+      "epoch": 0.12740252608456892,
+      "grad_norm": 0.3876163065433502,
+      "learning_rate": 0.00012481751824817516,
+      "loss": 1.1582,
+      "step": 58
+    },
+    {
+      "epoch": 0.12959912136188906,
+      "grad_norm": 0.5714845657348633,
+      "learning_rate": 0.00012700729927007297,
+      "loss": 0.8656,
+      "step": 59
+    },
+    {
+      "epoch": 0.13179571663920922,
+      "grad_norm": 0.1504809409379959,
+      "learning_rate": 0.00012919708029197077,
+      "loss": 0.9313,
+      "step": 60
+    },
+    {
+      "epoch": 0.13399231191652938,
+      "grad_norm": 0.1797901839017868,
+      "learning_rate": 0.0001313868613138686,
+      "loss": 1.004,
+      "step": 61
+    },
+    {
+      "epoch": 0.13618890719384955,
+      "grad_norm": 0.1749698966741562,
+      "learning_rate": 0.00013357664233576641,
+      "loss": 0.8709,
+      "step": 62
+    },
+    {
+      "epoch": 0.13838550247116968,
+      "grad_norm": 0.14003662765026093,
+      "learning_rate": 0.00013576642335766422,
+      "loss": 0.9077,
+      "step": 63
+    },
+    {
+      "epoch": 0.14058209774848984,
+      "grad_norm": 0.1614685207605362,
+      "learning_rate": 0.00013795620437956203,
+      "loss": 1.0631,
+      "step": 64
+    },
+    {
+      "epoch": 0.14277869302581,
+      "grad_norm": 0.1968277543783188,
+      "learning_rate": 0.00014014598540145983,
+      "loss": 0.8721,
+      "step": 65
+    },
+    {
+      "epoch": 0.14497528830313014,
+      "grad_norm": 0.10564833134412766,
+      "learning_rate": 0.00014233576642335764,
+      "loss": 0.7734,
+      "step": 66
+    },
+    {
+      "epoch": 0.1471718835804503,
+      "grad_norm": 0.12768442928791046,
+      "learning_rate": 0.00014452554744525547,
+      "loss": 0.9881,
+      "step": 67
+    },
+    {
+      "epoch": 0.14936847885777046,
+      "grad_norm": 0.1462821066379547,
+      "learning_rate": 0.00014671532846715328,
+      "loss": 0.7884,
+      "step": 68
+    },
+    {
+      "epoch": 0.1515650741350906,
+      "grad_norm": 0.1584155112504959,
+      "learning_rate": 0.00014890510948905108,
+      "loss": 1.0954,
+      "step": 69
+    },
+    {
+      "epoch": 0.15376166941241076,
+      "grad_norm": 0.15821506083011627,
+      "learning_rate": 0.00015109489051094892,
+      "loss": 0.8921,
+      "step": 70
+    },
+    {
+      "epoch": 0.15595826468973092,
+      "grad_norm": 0.12775461375713348,
+      "learning_rate": 0.00015328467153284672,
+      "loss": 1.0497,
+      "step": 71
+    },
+    {
+      "epoch": 0.15815485996705106,
+      "grad_norm": 0.12319904565811157,
+      "learning_rate": 0.00015547445255474453,
+      "loss": 0.9112,
+      "step": 72
+    },
+    {
+      "epoch": 0.16035145524437122,
+      "grad_norm": 0.16026116907596588,
+      "learning_rate": 0.00015766423357664234,
+      "loss": 1.1613,
+      "step": 73
+    },
+    {
+      "epoch": 0.16254805052169138,
+      "grad_norm": 0.13379302620887756,
+      "learning_rate": 0.00015985401459854014,
+      "loss": 1.0376,
+      "step": 74
+    },
+    {
+      "epoch": 0.16474464579901152,
+      "grad_norm": 0.13290898501873016,
+      "learning_rate": 0.00016204379562043795,
+      "loss": 0.8284,
+      "step": 75
+    },
+    {
+      "epoch": 0.16694124107633168,
+      "grad_norm": 0.15173152089118958,
+      "learning_rate": 0.00016423357664233575,
+      "loss": 0.9723,
+      "step": 76
+    },
+    {
+      "epoch": 0.16913783635365184,
+      "grad_norm": 0.12692321836948395,
+      "learning_rate": 0.00016642335766423356,
+      "loss": 0.906,
+      "step": 77
+    },
+    {
+      "epoch": 0.171334431630972,
+      "grad_norm": 0.15603457391262054,
+      "learning_rate": 0.00016861313868613137,
+      "loss": 0.7096,
+      "step": 78
+    },
+    {
+      "epoch": 0.17353102690829214,
+      "grad_norm": 0.14791829884052277,
+      "learning_rate": 0.00017080291970802917,
+      "loss": 0.9471,
+      "step": 79
+    },
+    {
+      "epoch": 0.1757276221856123,
+      "grad_norm": 0.11918247491121292,
+      "learning_rate": 0.00017299270072992698,
+      "loss": 0.9394,
+      "step": 80
+    },
+    {
+      "epoch": 0.17792421746293247,
+      "grad_norm": 0.20789018273353577,
+      "learning_rate": 0.00017518248175182478,
+      "loss": 0.8983,
+      "step": 81
+    },
+    {
+      "epoch": 0.1801208127402526,
+      "grad_norm": 0.15847823023796082,
+      "learning_rate": 0.00017737226277372262,
+      "loss": 0.9278,
+      "step": 82
+    },
+    {
+      "epoch": 0.18231740801757276,
+      "grad_norm": 0.1210375651717186,
+      "learning_rate": 0.00017956204379562042,
+      "loss": 0.8585,
+      "step": 83
+    },
+    {
+      "epoch": 0.18451400329489293,
+      "grad_norm": 0.15506596863269806,
+      "learning_rate": 0.00018175182481751826,
+      "loss": 0.9768,
+      "step": 84
+    },
+    {
+      "epoch": 0.18671059857221306,
+      "grad_norm": 0.20856642723083496,
+      "learning_rate": 0.00018394160583941606,
+      "loss": 1.117,
+      "step": 85
+    },
+    {
+      "epoch": 0.18890719384953322,
+      "grad_norm": 0.27022668719291687,
+      "learning_rate": 0.00018613138686131387,
+      "loss": 1.2502,
+      "step": 86
+    },
+    {
+      "epoch": 0.19110378912685339,
+      "grad_norm": 0.13160812854766846,
+      "learning_rate": 0.00018832116788321167,
+      "loss": 0.9517,
+      "step": 87
+    },
+    {
+      "epoch": 0.19330038440417352,
+      "grad_norm": 0.13891175389289856,
+      "learning_rate": 0.00019051094890510948,
+      "loss": 1.0114,
+      "step": 88
+    },
+    {
+      "epoch": 0.19549697968149368,
+      "grad_norm": 0.15862490236759186,
+      "learning_rate": 0.0001927007299270073,
+      "loss": 0.8157,
+      "step": 89
+    },
+    {
+      "epoch": 0.19769357495881384,
+      "grad_norm": 0.1573576182126999,
+      "learning_rate": 0.0001948905109489051,
+      "loss": 1.1404,
+      "step": 90
+    },
+    {
+      "epoch": 0.19989017023613398,
+      "grad_norm": 0.09982264041900635,
+      "learning_rate": 0.0001970802919708029,
+      "loss": 0.8252,
+      "step": 91
+    },
+    {
+      "epoch": 0.20208676551345414,
+      "grad_norm": 0.1547119915485382,
+      "learning_rate": 0.0001992700729927007,
+      "loss": 0.8475,
+      "step": 92
+    },
+    {
+      "epoch": 0.2042833607907743,
+      "grad_norm": 0.1394474059343338,
+      "learning_rate": 0.0002014598540145985,
+      "loss": 0.9184,
+      "step": 93
+    },
+    {
+      "epoch": 0.20647995606809447,
+      "grad_norm": 0.13905766606330872,
+      "learning_rate": 0.00020364963503649632,
+      "loss": 1.1519,
+      "step": 94
+    },
+    {
+      "epoch": 0.2086765513454146,
+      "grad_norm": 0.11851614713668823,
+      "learning_rate": 0.00020583941605839415,
+      "loss": 0.8575,
+      "step": 95
+    },
+    {
+      "epoch": 0.21087314662273476,
+      "grad_norm": 0.14083677530288696,
+      "learning_rate": 0.00020802919708029196,
+      "loss": 1.0386,
+      "step": 96
+    },
+    {
+      "epoch": 0.21306974190005493,
+      "grad_norm": 0.1396334171295166,
+      "learning_rate": 0.00021021897810218976,
+      "loss": 1.0113,
+      "step": 97
+    },
+    {
+      "epoch": 0.21526633717737506,
+      "grad_norm": 0.13279949128627777,
+      "learning_rate": 0.00021240875912408757,
+      "loss": 0.8821,
+      "step": 98
+    },
+    {
+      "epoch": 0.21746293245469522,
+      "grad_norm": 0.13948281109333038,
+      "learning_rate": 0.00021459854014598537,
+      "loss": 0.9735,
+      "step": 99
+    },
+    {
+      "epoch": 0.21965952773201539,
+      "grad_norm": 0.15915612876415253,
+      "learning_rate": 0.0002167883211678832,
+      "loss": 1.0638,
+      "step": 100
+    },
+    {
+      "epoch": 0.22185612300933552,
+      "grad_norm": 0.24225056171417236,
+      "learning_rate": 0.00021897810218978101,
+      "loss": 1.067,
+      "step": 101
+    },
+    {
+      "epoch": 0.22405271828665568,
+      "grad_norm": 0.17314741015434265,
+      "learning_rate": 0.00022116788321167882,
+      "loss": 0.965,
+      "step": 102
+    },
+    {
+      "epoch": 0.22624931356397585,
+      "grad_norm": 0.10907436907291412,
+      "learning_rate": 0.00022335766423357663,
+      "loss": 0.9314,
+      "step": 103
+    },
+    {
+      "epoch": 0.22844590884129598,
+      "grad_norm": 0.14774860441684723,
+      "learning_rate": 0.00022554744525547443,
+      "loss": 0.8288,
+      "step": 104
+    },
+    {
+      "epoch": 0.23064250411861614,
+      "grad_norm": 0.14873632788658142,
+      "learning_rate": 0.00022773722627737224,
+      "loss": 0.9543,
+      "step": 105
+    },
+    {
+      "epoch": 0.2328390993959363,
+      "grad_norm": 0.13154087960720062,
+      "learning_rate": 0.00022992700729927004,
+      "loss": 1.03,
+      "step": 106
+    },
+    {
+      "epoch": 0.23503569467325644,
+      "grad_norm": 0.22245362401008606,
+      "learning_rate": 0.00023211678832116788,
+      "loss": 1.0415,
+      "step": 107
+    },
+    {
+      "epoch": 0.2372322899505766,
+      "grad_norm": 0.12790049612522125,
+      "learning_rate": 0.00023430656934306568,
+      "loss": 0.8201,
+      "step": 108
+    },
+    {
+      "epoch": 0.23942888522789676,
+      "grad_norm": 0.14593057334423065,
+      "learning_rate": 0.0002364963503649635,
+      "loss": 0.8345,
+      "step": 109
+    },
+    {
+      "epoch": 0.24162548050521693,
+      "grad_norm": 0.16340065002441406,
+      "learning_rate": 0.0002386861313868613,
+      "loss": 0.8684,
+      "step": 110
+    },
+    {
+      "epoch": 0.24382207578253706,
+      "grad_norm": 0.15796829760074615,
+      "learning_rate": 0.0002408759124087591,
+      "loss": 0.9408,
+      "step": 111
+    },
+    {
+      "epoch": 0.24601867105985722,
+      "grad_norm": 0.15153130888938904,
+      "learning_rate": 0.0002430656934306569,
+      "loss": 0.9379,
+      "step": 112
+    },
+    {
+      "epoch": 0.2482152663371774,
+      "grad_norm": 0.11751651018857956,
+      "learning_rate": 0.0002452554744525547,
+      "loss": 0.8432,
+      "step": 113
+    },
+    {
+      "epoch": 0.2504118616144975,
+      "grad_norm": 0.13884197175502777,
+      "learning_rate": 0.0002474452554744525,
+      "loss": 0.8301,
+      "step": 114
+    },
+    {
+      "epoch": 0.25260845689181766,
+      "grad_norm": 0.14576764404773712,
+      "learning_rate": 0.0002496350364963503,
+      "loss": 0.9244,
+      "step": 115
+    },
+    {
+      "epoch": 0.25480505216913785,
+      "grad_norm": 0.28019213676452637,
+      "learning_rate": 0.00025182481751824813,
+      "loss": 1.0155,
+      "step": 116
+    },
+    {
+      "epoch": 0.257001647446458,
+      "grad_norm": 0.15313687920570374,
+      "learning_rate": 0.00025401459854014594,
+      "loss": 0.9448,
+      "step": 117
+    },
+    {
+      "epoch": 0.2591982427237781,
+      "grad_norm": 0.11438549309968948,
+      "learning_rate": 0.00025620437956204374,
+      "loss": 0.8633,
+      "step": 118
+    },
+    {
+      "epoch": 0.2613948380010983,
+      "grad_norm": 0.12334372848272324,
+      "learning_rate": 0.00025839416058394155,
+      "loss": 0.85,
+      "step": 119
+    },
+    {
+      "epoch": 0.26359143327841844,
+      "grad_norm": 0.10154549032449722,
+      "learning_rate": 0.0002605839416058394,
+      "loss": 0.7322,
+      "step": 120
+    },
+    {
+      "epoch": 0.26578802855573863,
+      "grad_norm": 0.10096869617700577,
+      "learning_rate": 0.0002627737226277372,
+      "loss": 0.872,
+      "step": 121
+    },
+    {
+      "epoch": 0.26798462383305877,
+      "grad_norm": 0.14952614903450012,
+      "learning_rate": 0.000264963503649635,
+      "loss": 0.7973,
+      "step": 122
+    },
+    {
+      "epoch": 0.2701812191103789,
+      "grad_norm": 0.13486889004707336,
+      "learning_rate": 0.00026715328467153283,
+      "loss": 0.8104,
+      "step": 123
+    },
+    {
+      "epoch": 0.2723778143876991,
+      "grad_norm": 0.1295621544122696,
+      "learning_rate": 0.00026934306569343063,
+      "loss": 0.7799,
+      "step": 124
+    },
+    {
+      "epoch": 0.2745744096650192,
+      "grad_norm": 0.1486479938030243,
+      "learning_rate": 0.00027153284671532844,
+      "loss": 0.9493,
+      "step": 125
+    },
+    {
+      "epoch": 0.27677100494233936,
+      "grad_norm": 0.42850103974342346,
+      "learning_rate": 0.00027372262773722625,
+      "loss": 0.7009,
+      "step": 126
+    },
+    {
+      "epoch": 0.27896760021965955,
+      "grad_norm": 0.12433044612407684,
+      "learning_rate": 0.00027591240875912405,
+      "loss": 0.8463,
+      "step": 127
+    },
+    {
+      "epoch": 0.2811641954969797,
+      "grad_norm": 0.16575849056243896,
+      "learning_rate": 0.00027810218978102186,
+      "loss": 1.144,
+      "step": 128
+    },
+    {
+      "epoch": 0.2833607907742998,
+      "grad_norm": 0.13898643851280212,
+      "learning_rate": 0.00028029197080291966,
+      "loss": 0.9783,
+      "step": 129
+    },
+    {
+      "epoch": 0.28555738605162,
+      "grad_norm": 0.1803688257932663,
+      "learning_rate": 0.00028248175182481747,
+      "loss": 0.8282,
+      "step": 130
+    },
+    {
+      "epoch": 0.28775398132894014,
+      "grad_norm": 0.10425428301095963,
+      "learning_rate": 0.0002846715328467153,
+      "loss": 0.8593,
+      "step": 131
+    },
+    {
+      "epoch": 0.2899505766062603,
+      "grad_norm": 0.20991037786006927,
+      "learning_rate": 0.00028686131386861314,
+      "loss": 0.81,
+      "step": 132
+    },
+    {
+      "epoch": 0.29214717188358047,
+      "grad_norm": 0.15783126652240753,
+      "learning_rate": 0.00028905109489051094,
+      "loss": 0.7979,
+      "step": 133
+    },
+    {
+      "epoch": 0.2943437671609006,
+      "grad_norm": 0.13156543672084808,
+      "learning_rate": 0.00029124087591240875,
+      "loss": 0.871,
+      "step": 134
+    },
+    {
+      "epoch": 0.29654036243822074,
+      "grad_norm": 0.15906956791877747,
+      "learning_rate": 0.00029343065693430656,
+      "loss": 0.8432,
+      "step": 135
+    },
+    {
+      "epoch": 0.29873695771554093,
+      "grad_norm": 0.14525553584098816,
+      "learning_rate": 0.00029562043795620436,
+      "loss": 0.9821,
+      "step": 136
+    },
+    {
+      "epoch": 0.30093355299286106,
+      "grad_norm": 0.15281268954277039,
+      "learning_rate": 0.00029781021897810217,
+      "loss": 0.8557,
+      "step": 137
+    },
+    {
+      "epoch": 0.3031301482701812,
+      "grad_norm": 0.13318589329719543,
+      "learning_rate": 0.0003,
+      "loss": 0.8863,
+      "step": 138
+    },
+    {
+      "epoch": 0.3053267435475014,
+      "grad_norm": 0.1681632250547409,
+      "learning_rate": 0.00029975570032573286,
+      "loss": 0.8525,
+      "step": 139
+    },
+    {
+      "epoch": 0.3075233388248215,
+      "grad_norm": 0.17074619233608246,
+      "learning_rate": 0.00029951140065146575,
+      "loss": 0.8874,
+      "step": 140
+    },
+    {
+      "epoch": 0.30971993410214166,
+      "grad_norm": 0.15967530012130737,
+      "learning_rate": 0.0002992671009771987,
+      "loss": 0.9021,
+      "step": 141
+    },
+    {
+      "epoch": 0.31191652937946185,
+      "grad_norm": 0.14229127764701843,
+      "learning_rate": 0.0002990228013029316,
+      "loss": 0.9625,
+      "step": 142
+    },
+    {
+      "epoch": 0.314113124656782,
+      "grad_norm": 0.15182703733444214,
+      "learning_rate": 0.0002987785016286645,
+      "loss": 0.8692,
+      "step": 143
+    },
+    {
+      "epoch": 0.3163097199341021,
+      "grad_norm": 0.14878253638744354,
+      "learning_rate": 0.00029853420195439736,
+      "loss": 0.9817,
+      "step": 144
+    },
+    {
+      "epoch": 0.3185063152114223,
+      "grad_norm": 0.1443973034620285,
+      "learning_rate": 0.00029828990228013025,
+      "loss": 0.9054,
+      "step": 145
+    },
+    {
+      "epoch": 0.32070291048874244,
+      "grad_norm": 0.13824306428432465,
+      "learning_rate": 0.00029804560260586314,
+      "loss": 0.8709,
+      "step": 146
+    },
+    {
+      "epoch": 0.3228995057660626,
+      "grad_norm": 0.11518386751413345,
+      "learning_rate": 0.0002978013029315961,
+      "loss": 0.714,
+      "step": 147
+    },
+    {
+      "epoch": 0.32509610104338277,
+      "grad_norm": 0.10005150735378265,
+      "learning_rate": 0.000297557003257329,
+      "loss": 0.7404,
+      "step": 148
+    },
+    {
+      "epoch": 0.3272926963207029,
+      "grad_norm": 0.13207948207855225,
+      "learning_rate": 0.00029731270358306186,
+      "loss": 0.8135,
+      "step": 149
+    },
+    {
+      "epoch": 0.32948929159802304,
+      "grad_norm": 0.2112632840871811,
+      "learning_rate": 0.00029706840390879475,
+      "loss": 1.1552,
+      "step": 150
+    },
+    {
+      "epoch": 0.3316858868753432,
+      "grad_norm": 0.1803838610649109,
+      "learning_rate": 0.00029682410423452764,
+      "loss": 0.9471,
+      "step": 151
+    },
+    {
+      "epoch": 0.33388248215266336,
+      "grad_norm": 0.144296333193779,
+      "learning_rate": 0.0002965798045602606,
+      "loss": 0.9815,
+      "step": 152
+    },
+    {
+      "epoch": 0.33607907742998355,
+      "grad_norm": 0.1251664161682129,
+      "learning_rate": 0.0002963355048859935,
+      "loss": 0.7871,
+      "step": 153
+    },
+    {
+      "epoch": 0.3382756727073037,
+      "grad_norm": 0.1101875975728035,
+      "learning_rate": 0.00029609120521172636,
+      "loss": 0.8135,
+      "step": 154
+    },
+    {
+      "epoch": 0.3404722679846238,
+      "grad_norm": 0.120697982609272,
+      "learning_rate": 0.00029584690553745925,
+      "loss": 0.7602,
+      "step": 155
+    },
+    {
+      "epoch": 0.342668863261944,
+      "grad_norm": 0.13179142773151398,
+      "learning_rate": 0.00029560260586319214,
+      "loss": 0.795,
+      "step": 156
+    },
+    {
+      "epoch": 0.34486545853926415,
+      "grad_norm": 0.6565778255462646,
+      "learning_rate": 0.00029535830618892503,
+      "loss": 1.0316,
+      "step": 157
+    },
+    {
+      "epoch": 0.3470620538165843,
+      "grad_norm": 0.12122897803783417,
+      "learning_rate": 0.000295114006514658,
+      "loss": 0.8384,
+      "step": 158
+    },
+    {
+      "epoch": 0.34925864909390447,
+      "grad_norm": 0.1196598932147026,
+      "learning_rate": 0.00029486970684039086,
+      "loss": 0.9366,
+      "step": 159
+    },
+    {
+      "epoch": 0.3514552443712246,
+      "grad_norm": 0.15032830834388733,
+      "learning_rate": 0.00029462540716612375,
+      "loss": 0.8415,
+      "step": 160
+    },
+    {
+      "epoch": 0.35365183964854474,
+      "grad_norm": 0.12760332226753235,
+      "learning_rate": 0.00029438110749185664,
+      "loss": 0.8085,
+      "step": 161
+    },
+    {
+      "epoch": 0.35584843492586493,
+      "grad_norm": 0.11488742381334305,
+      "learning_rate": 0.00029413680781758953,
+      "loss": 0.6981,
+      "step": 162
+    },
+    {
+      "epoch": 0.35804503020318507,
+      "grad_norm": 0.10087805986404419,
+      "learning_rate": 0.0002938925081433224,
+      "loss": 0.604,
+      "step": 163
+    },
+    {
+      "epoch": 0.3602416254805052,
+      "grad_norm": 0.16445618867874146,
+      "learning_rate": 0.00029364820846905536,
+      "loss": 0.8211,
+      "step": 164
+    },
+    {
+      "epoch": 0.3624382207578254,
+      "grad_norm": 0.12089920043945312,
+      "learning_rate": 0.00029340390879478825,
+      "loss": 0.9139,
+      "step": 165
+    },
+    {
+      "epoch": 0.3646348160351455,
+      "grad_norm": 0.16482849419116974,
+      "learning_rate": 0.00029315960912052114,
+      "loss": 0.8737,
+      "step": 166
+    },
+    {
+      "epoch": 0.36683141131246566,
+      "grad_norm": 0.13793058693408966,
+      "learning_rate": 0.00029291530944625403,
+      "loss": 0.7935,
+      "step": 167
+    },
+    {
+      "epoch": 0.36902800658978585,
+      "grad_norm": 0.1376744508743286,
+      "learning_rate": 0.0002926710097719869,
+      "loss": 0.8653,
+      "step": 168
+    },
+    {
+      "epoch": 0.371224601867106,
+      "grad_norm": 0.12906500697135925,
+      "learning_rate": 0.00029242671009771986,
+      "loss": 0.8869,
+      "step": 169
+    },
+    {
+      "epoch": 0.3734211971444261,
+      "grad_norm": 0.21045434474945068,
+      "learning_rate": 0.00029218241042345275,
+      "loss": 0.9773,
+      "step": 170
+    },
+    {
+      "epoch": 0.3756177924217463,
+      "grad_norm": 0.14139027893543243,
+      "learning_rate": 0.00029193811074918564,
+      "loss": 0.8006,
+      "step": 171
+    },
+    {
+      "epoch": 0.37781438769906645,
+      "grad_norm": 0.11151039600372314,
+      "learning_rate": 0.00029169381107491853,
+      "loss": 0.6979,
+      "step": 172
+    },
+    {
+      "epoch": 0.3800109829763866,
+      "grad_norm": 0.16388607025146484,
+      "learning_rate": 0.0002914495114006514,
+      "loss": 0.8441,
+      "step": 173
+    },
+    {
+      "epoch": 0.38220757825370677,
+      "grad_norm": 0.12790323793888092,
+      "learning_rate": 0.0002912052117263843,
+      "loss": 1.1248,
+      "step": 174
+    },
+    {
+      "epoch": 0.3844041735310269,
+      "grad_norm": 0.11524307727813721,
+      "learning_rate": 0.00029096091205211725,
+      "loss": 0.6624,
+      "step": 175
+    },
+    {
+      "epoch": 0.38660076880834704,
+      "grad_norm": 0.09693888574838638,
+      "learning_rate": 0.00029071661237785014,
+      "loss": 0.9627,
+      "step": 176
+    },
+    {
+      "epoch": 0.38879736408566723,
+      "grad_norm": 0.11282742023468018,
+      "learning_rate": 0.00029047231270358303,
+      "loss": 0.5847,
+      "step": 177
+    },
+    {
+      "epoch": 0.39099395936298736,
+      "grad_norm": 0.10943374782800674,
+      "learning_rate": 0.0002902280130293159,
+      "loss": 0.8329,
+      "step": 178
+    },
+    {
+      "epoch": 0.3931905546403075,
+      "grad_norm": 0.13954195380210876,
+      "learning_rate": 0.0002899837133550488,
+      "loss": 0.8689,
+      "step": 179
+    },
+    {
+      "epoch": 0.3953871499176277,
+      "grad_norm": 0.14839622378349304,
+      "learning_rate": 0.00028973941368078175,
+      "loss": 0.7292,
+      "step": 180
+    },
+    {
+      "epoch": 0.3975837451949478,
+      "grad_norm": 0.1340896189212799,
+      "learning_rate": 0.00028949511400651464,
+      "loss": 0.7136,
+      "step": 181
+    },
+    {
+      "epoch": 0.39978034047226796,
+      "grad_norm": 0.13768185675144196,
+      "learning_rate": 0.00028925081433224753,
+      "loss": 0.6376,
+      "step": 182
+    },
+    {
+      "epoch": 0.40197693574958815,
+      "grad_norm": 0.5511095523834229,
+      "learning_rate": 0.0002890065146579804,
+      "loss": 0.8696,
+      "step": 183
+    },
+    {
+      "epoch": 0.4041735310269083,
+      "grad_norm": 0.14440785348415375,
+      "learning_rate": 0.0002887622149837133,
+      "loss": 1.0853,
+      "step": 184
+    },
+    {
+      "epoch": 0.4063701263042284,
+      "grad_norm": 0.11086300760507584,
+      "learning_rate": 0.0002885179153094462,
+      "loss": 0.8789,
+      "step": 185
+    },
+    {
+      "epoch": 0.4085667215815486,
+      "grad_norm": 0.10120414197444916,
+      "learning_rate": 0.00028827361563517914,
+      "loss": 0.7875,
+      "step": 186
+    },
+    {
+      "epoch": 0.41076331685886874,
+      "grad_norm": 0.10476330667734146,
+      "learning_rate": 0.00028802931596091203,
+      "loss": 0.6277,
+      "step": 187
+    },
+    {
+      "epoch": 0.41295991213618893,
+      "grad_norm": 0.13569848239421844,
+      "learning_rate": 0.0002877850162866449,
+      "loss": 0.8633,
+      "step": 188
+    },
+    {
+      "epoch": 0.41515650741350907,
+      "grad_norm": 0.16974738240242004,
+      "learning_rate": 0.0002875407166123778,
+      "loss": 0.9285,
+      "step": 189
+    },
+    {
+      "epoch": 0.4173531026908292,
+      "grad_norm": 0.09820197522640228,
+      "learning_rate": 0.0002872964169381107,
+      "loss": 0.7368,
+      "step": 190
+    },
+    {
+      "epoch": 0.4195496979681494,
+      "grad_norm": 0.12278851121664047,
+      "learning_rate": 0.00028705211726384364,
+      "loss": 0.7624,
+      "step": 191
+    },
+    {
+      "epoch": 0.42174629324546953,
+      "grad_norm": 0.10428282618522644,
+      "learning_rate": 0.00028680781758957653,
+      "loss": 0.869,
+      "step": 192
+    },
+    {
+      "epoch": 0.42394288852278966,
+      "grad_norm": 0.13109053671360016,
+      "learning_rate": 0.0002865635179153094,
+      "loss": 0.8601,
+      "step": 193
+    },
+    {
+      "epoch": 0.42613948380010985,
+      "grad_norm": 0.14107254147529602,
+      "learning_rate": 0.0002863192182410423,
+      "loss": 0.8699,
+      "step": 194
+    },
+    {
+      "epoch": 0.42833607907743,
+      "grad_norm": 0.11819636076688766,
+      "learning_rate": 0.0002860749185667752,
+      "loss": 1.0829,
+      "step": 195
+    },
+    {
+      "epoch": 0.4305326743547501,
+      "grad_norm": 0.16183224320411682,
+      "learning_rate": 0.0002858306188925081,
+      "loss": 1.2516,
+      "step": 196
+    },
+    {
+      "epoch": 0.4327292696320703,
+      "grad_norm": 0.126215860247612,
+      "learning_rate": 0.00028558631921824103,
+      "loss": 0.7477,
+      "step": 197
+    },
+    {
+      "epoch": 0.43492586490939045,
+      "grad_norm": 0.10160335898399353,
+      "learning_rate": 0.0002853420195439739,
+      "loss": 0.7476,
+      "step": 198
+    },
+    {
+      "epoch": 0.4371224601867106,
+      "grad_norm": 0.10714973509311676,
+      "learning_rate": 0.0002850977198697068,
+      "loss": 0.789,
+      "step": 199
+    },
+    {
+      "epoch": 0.43931905546403077,
+      "grad_norm": 0.11634092032909393,
+      "learning_rate": 0.0002848534201954397,
+      "loss": 0.8834,
+      "step": 200
+    },
+    {
+      "epoch": 0.4415156507413509,
+      "grad_norm": 0.11050505936145782,
+      "learning_rate": 0.0002846091205211726,
+      "loss": 0.6684,
+      "step": 201
+    },
+    {
+      "epoch": 0.44371224601867104,
+      "grad_norm": 0.3374190628528595,
+      "learning_rate": 0.0002843648208469055,
+      "loss": 0.7041,
+      "step": 202
+    },
+    {
+      "epoch": 0.44590884129599123,
+      "grad_norm": 0.08308324962854385,
+      "learning_rate": 0.0002841205211726384,
+      "loss": 0.7965,
+      "step": 203
+    },
+    {
+      "epoch": 0.44810543657331137,
+      "grad_norm": 0.13838446140289307,
+      "learning_rate": 0.0002838762214983713,
+      "loss": 0.9431,
+      "step": 204
+    },
+    {
+      "epoch": 0.4503020318506315,
+      "grad_norm": 0.4903043508529663,
+      "learning_rate": 0.0002836319218241042,
+      "loss": 0.9939,
+      "step": 205
+    },
+    {
+      "epoch": 0.4524986271279517,
+      "grad_norm": 0.17547567188739777,
+      "learning_rate": 0.0002833876221498371,
+      "loss": 0.9722,
+      "step": 206
+    },
+    {
+      "epoch": 0.4546952224052718,
+      "grad_norm": 0.08944044262170792,
+      "learning_rate": 0.00028314332247557,
+      "loss": 0.8053,
+      "step": 207
+    },
+    {
+      "epoch": 0.45689181768259196,
+      "grad_norm": 0.4859505891799927,
+      "learning_rate": 0.0002828990228013029,
+      "loss": 0.7568,
+      "step": 208
+    },
+    {
+      "epoch": 0.45908841295991215,
+      "grad_norm": 0.16345354914665222,
+      "learning_rate": 0.0002826547231270358,
+      "loss": 0.8577,
+      "step": 209
+    },
+    {
+      "epoch": 0.4612850082372323,
+      "grad_norm": 0.11055205762386322,
+      "learning_rate": 0.0002824104234527687,
+      "loss": 0.8993,
+      "step": 210
+    },
+    {
+      "epoch": 0.4634816035145524,
+      "grad_norm": 0.11288679391145706,
+      "learning_rate": 0.0002821661237785016,
+      "loss": 0.9589,
+      "step": 211
+    },
+    {
+      "epoch": 0.4656781987918726,
+      "grad_norm": 0.13978932797908783,
+      "learning_rate": 0.0002819218241042345,
+      "loss": 0.8578,
+      "step": 212
+    },
+    {
+      "epoch": 0.46787479406919275,
+      "grad_norm": 0.08493883907794952,
+      "learning_rate": 0.00028167752442996737,
+      "loss": 0.5956,
+      "step": 213
+    },
+    {
+      "epoch": 0.4700713893465129,
+      "grad_norm": 0.09494692087173462,
+      "learning_rate": 0.0002814332247557003,
+      "loss": 0.7156,
+      "step": 214
+    },
+    {
+      "epoch": 0.47226798462383307,
+      "grad_norm": 0.1031997799873352,
+      "learning_rate": 0.0002811889250814332,
+      "loss": 0.8704,
+      "step": 215
+    },
+    {
+      "epoch": 0.4744645799011532,
+      "grad_norm": 0.13378386199474335,
+      "learning_rate": 0.0002809446254071661,
+      "loss": 0.7907,
+      "step": 216
+    },
+    {
+      "epoch": 0.47666117517847334,
+      "grad_norm": 0.13447315990924835,
+      "learning_rate": 0.000280700325732899,
+      "loss": 0.7159,
+      "step": 217
+    },
+    {
+      "epoch": 0.47885777045579353,
+      "grad_norm": 0.7438109517097473,
+      "learning_rate": 0.00028045602605863187,
+      "loss": 0.8813,
+      "step": 218
+    },
+    {
+      "epoch": 0.48105436573311366,
+      "grad_norm": 0.14037659764289856,
+      "learning_rate": 0.0002802117263843648,
+      "loss": 0.8172,
+      "step": 219
+    },
+    {
+      "epoch": 0.48325096101043385,
+      "grad_norm": 0.13234059512615204,
+      "learning_rate": 0.0002799674267100977,
+      "loss": 0.8477,
+      "step": 220
+    },
+    {
+      "epoch": 0.485447556287754,
+      "grad_norm": 0.13452236354351044,
+      "learning_rate": 0.0002797231270358306,
+      "loss": 0.7402,
+      "step": 221
+    },
+    {
+      "epoch": 0.4876441515650741,
+      "grad_norm": 0.13823570311069489,
+      "learning_rate": 0.0002794788273615635,
+      "loss": 0.705,
+      "step": 222
+    },
+    {
+      "epoch": 0.4898407468423943,
+      "grad_norm": 0.11288873851299286,
+      "learning_rate": 0.00027923452768729637,
+      "loss": 0.8026,
+      "step": 223
+    },
+    {
+      "epoch": 0.49203734211971445,
+      "grad_norm": 0.09462586045265198,
+      "learning_rate": 0.00027899022801302926,
+      "loss": 0.8354,
+      "step": 224
+    },
+    {
+      "epoch": 0.4942339373970346,
+      "grad_norm": 0.13376912474632263,
+      "learning_rate": 0.0002787459283387622,
+      "loss": 0.59,
+      "step": 225
+    },
+    {
+      "epoch": 0.4964305326743548,
+      "grad_norm": 0.09549413621425629,
+      "learning_rate": 0.0002785016286644951,
+      "loss": 1.0635,
+      "step": 226
+    },
+    {
+      "epoch": 0.4986271279516749,
+      "grad_norm": 0.10206422954797745,
+      "learning_rate": 0.000278257328990228,
+      "loss": 0.7693,
+      "step": 227
+    },
+    {
+      "epoch": 0.500823723228995,
+      "grad_norm": 0.11209911853075027,
+      "learning_rate": 0.00027801302931596087,
+      "loss": 0.7473,
+      "step": 228
+    },
+    {
+      "epoch": 0.5030203185063152,
+      "grad_norm": 0.32212138175964355,
+      "learning_rate": 0.00027776872964169376,
+      "loss": 0.7883,
+      "step": 229
+    },
+    {
+      "epoch": 0.5052169137836353,
+      "grad_norm": 0.11934421956539154,
+      "learning_rate": 0.0002775244299674267,
+      "loss": 0.8517,
+      "step": 230
+    },
+    {
+      "epoch": 0.5074135090609555,
+      "grad_norm": 0.12222785502672195,
+      "learning_rate": 0.0002772801302931596,
+      "loss": 0.8325,
+      "step": 231
+    },
+    {
+      "epoch": 0.5096101043382757,
+      "grad_norm": 0.10029341280460358,
+      "learning_rate": 0.0002770358306188925,
+      "loss": 0.7232,
+      "step": 232
+    },
+    {
+      "epoch": 0.5118066996155958,
+      "grad_norm": 0.11559610068798065,
+      "learning_rate": 0.00027679153094462537,
+      "loss": 0.869,
+      "step": 233
+    },
+    {
+      "epoch": 0.514003294892916,
+      "grad_norm": 0.12236785143613815,
+      "learning_rate": 0.00027654723127035826,
+      "loss": 0.8273,
+      "step": 234
+    },
+    {
+      "epoch": 0.5161998901702362,
+      "grad_norm": 0.13172945380210876,
+      "learning_rate": 0.00027630293159609115,
+      "loss": 0.8506,
+      "step": 235
+    },
+    {
+      "epoch": 0.5183964854475562,
+      "grad_norm": 0.23144572973251343,
+      "learning_rate": 0.0002760586319218241,
+      "loss": 0.715,
+      "step": 236
+    },
+    {
+      "epoch": 0.5205930807248764,
+      "grad_norm": 0.11784598231315613,
+      "learning_rate": 0.000275814332247557,
+      "loss": 1.0066,
+      "step": 237
+    },
+    {
+      "epoch": 0.5227896760021966,
+      "grad_norm": 0.15287131071090698,
+      "learning_rate": 0.00027557003257328987,
+      "loss": 0.8546,
+      "step": 238
+    },
+    {
+      "epoch": 0.5249862712795168,
+      "grad_norm": 0.11131344735622406,
+      "learning_rate": 0.00027532573289902276,
+      "loss": 0.7439,
+      "step": 239
+    },
+    {
+      "epoch": 0.5271828665568369,
+      "grad_norm": 0.12520942091941833,
+      "learning_rate": 0.00027508143322475565,
+      "loss": 0.8582,
+      "step": 240
+    },
+    {
+      "epoch": 0.5293794618341571,
+      "grad_norm": 0.13572880625724792,
+      "learning_rate": 0.00027483713355048854,
+      "loss": 0.8296,
+      "step": 241
+    },
+    {
+      "epoch": 0.5315760571114773,
+      "grad_norm": 0.1295921355485916,
+      "learning_rate": 0.0002745928338762215,
+      "loss": 0.7278,
+      "step": 242
+    },
+    {
+      "epoch": 0.5337726523887973,
+      "grad_norm": 0.07753497362136841,
+      "learning_rate": 0.00027434853420195437,
+      "loss": 0.708,
+      "step": 243
+    },
+    {
+      "epoch": 0.5359692476661175,
+      "grad_norm": 0.11066402494907379,
+      "learning_rate": 0.00027410423452768726,
+      "loss": 0.9077,
+      "step": 244
+    },
+    {
+      "epoch": 0.5381658429434377,
+      "grad_norm": 0.09430704265832901,
+      "learning_rate": 0.00027385993485342015,
+      "loss": 0.8044,
+      "step": 245
+    },
+    {
+      "epoch": 0.5403624382207578,
+      "grad_norm": 0.11148939281702042,
+      "learning_rate": 0.00027361563517915304,
+      "loss": 0.8466,
+      "step": 246
+    },
+    {
+      "epoch": 0.542559033498078,
+      "grad_norm": 0.43565067648887634,
+      "learning_rate": 0.000273371335504886,
+      "loss": 0.728,
+      "step": 247
+    },
+    {
+      "epoch": 0.5447556287753982,
+      "grad_norm": 0.14453081786632538,
+      "learning_rate": 0.00027312703583061887,
+      "loss": 0.8315,
+      "step": 248
+    },
+    {
+      "epoch": 0.5469522240527183,
+      "grad_norm": 1.5286325216293335,
+      "learning_rate": 0.00027288273615635176,
+      "loss": 0.7095,
+      "step": 249
+    },
+    {
+      "epoch": 0.5491488193300385,
+      "grad_norm": 0.10582248866558075,
+      "learning_rate": 0.00027263843648208465,
+      "loss": 0.794,
+      "step": 250
+    },
+    {
+      "epoch": 0.5513454146073586,
+      "grad_norm": 0.13800182938575745,
+      "learning_rate": 0.00027239413680781754,
+      "loss": 0.8104,
+      "step": 251
+    },
+    {
+      "epoch": 0.5535420098846787,
+      "grad_norm": 0.1428312361240387,
+      "learning_rate": 0.0002721498371335504,
+      "loss": 0.8254,
+      "step": 252
+    },
+    {
+      "epoch": 0.5557386051619989,
+      "grad_norm": 0.09881719201803207,
+      "learning_rate": 0.00027190553745928337,
+      "loss": 0.8586,
+      "step": 253
+    },
+    {
+      "epoch": 0.5579352004393191,
+      "grad_norm": 0.09296563267707825,
+      "learning_rate": 0.00027166123778501626,
+      "loss": 0.7562,
+      "step": 254
+    },
+    {
+      "epoch": 0.5601317957166392,
+      "grad_norm": 0.15508748590946198,
+      "learning_rate": 0.00027141693811074915,
+      "loss": 0.8589,
+      "step": 255
+    },
+    {
+      "epoch": 0.5623283909939594,
+      "grad_norm": 0.12908507883548737,
+      "learning_rate": 0.00027117263843648204,
+      "loss": 0.7475,
+      "step": 256
+    },
+    {
+      "epoch": 0.5645249862712796,
+      "grad_norm": 0.07709969580173492,
+      "learning_rate": 0.0002709283387622149,
+      "loss": 0.6687,
+      "step": 257
+    },
+    {
+      "epoch": 0.5667215815485996,
+      "grad_norm": 0.12212585657835007,
+      "learning_rate": 0.00027068403908794787,
+      "loss": 0.9472,
+      "step": 258
+    },
+    {
+      "epoch": 0.5689181768259198,
+      "grad_norm": 0.11735911667346954,
+      "learning_rate": 0.00027043973941368076,
+      "loss": 0.7185,
+      "step": 259
+    },
+    {
+      "epoch": 0.57111477210324,
+      "grad_norm": 0.10713691264390945,
+      "learning_rate": 0.00027019543973941365,
+      "loss": 0.7765,
+      "step": 260
+    },
+    {
+      "epoch": 0.5733113673805601,
+      "grad_norm": 0.08410372585058212,
+      "learning_rate": 0.00026995114006514654,
+      "loss": 0.822,
+      "step": 261
+    },
+    {
+      "epoch": 0.5755079626578803,
+      "grad_norm": 0.08931750804185867,
+      "learning_rate": 0.00026970684039087943,
+      "loss": 0.8033,
+      "step": 262
+    },
+    {
+      "epoch": 0.5777045579352005,
+      "grad_norm": 0.1014622375369072,
+      "learning_rate": 0.0002694625407166123,
+      "loss": 0.6585,
+      "step": 263
+    },
+    {
+      "epoch": 0.5799011532125206,
+      "grad_norm": 0.14549599587917328,
+      "learning_rate": 0.00026921824104234526,
+      "loss": 0.8551,
+      "step": 264
+    },
+    {
+      "epoch": 0.5820977484898407,
+      "grad_norm": 0.12771829962730408,
+      "learning_rate": 0.00026897394136807815,
+      "loss": 1.1118,
+      "step": 265
+    },
+    {
+      "epoch": 0.5842943437671609,
+      "grad_norm": 0.13730259239673615,
+      "learning_rate": 0.00026872964169381104,
+      "loss": 0.8927,
+      "step": 266
+    },
+    {
+      "epoch": 0.586490939044481,
+      "grad_norm": 0.10761181265115738,
+      "learning_rate": 0.00026848534201954393,
+      "loss": 0.8308,
+      "step": 267
+    },
+    {
+      "epoch": 0.5886875343218012,
+      "grad_norm": 0.10117297619581223,
+      "learning_rate": 0.0002682410423452768,
+      "loss": 0.875,
+      "step": 268
+    },
+    {
+      "epoch": 0.5908841295991214,
+      "grad_norm": 0.1135297417640686,
+      "learning_rate": 0.00026799674267100976,
+      "loss": 0.9189,
+      "step": 269
+    },
+    {
+      "epoch": 0.5930807248764415,
+      "grad_norm": 0.19212596118450165,
+      "learning_rate": 0.00026775244299674265,
+      "loss": 0.8824,
+      "step": 270
+    },
+    {
+      "epoch": 0.5952773201537617,
+      "grad_norm": 0.09628119319677353,
+      "learning_rate": 0.00026750814332247554,
+      "loss": 0.7027,
+      "step": 271
+    },
+    {
+      "epoch": 0.5974739154310819,
+      "grad_norm": 0.10682559013366699,
+      "learning_rate": 0.00026726384364820843,
+      "loss": 0.766,
+      "step": 272
+    },
+    {
+      "epoch": 0.5996705107084019,
+      "grad_norm": 0.1086261123418808,
+      "learning_rate": 0.0002670195439739413,
+      "loss": 0.6936,
+      "step": 273
+    },
+    {
+      "epoch": 0.6018671059857221,
+      "grad_norm": 0.11289719492197037,
+      "learning_rate": 0.0002667752442996742,
+      "loss": 0.7446,
+      "step": 274
+    },
+    {
+      "epoch": 0.6040637012630423,
+      "grad_norm": 0.12450888752937317,
+      "learning_rate": 0.00026653094462540715,
+      "loss": 0.8523,
+      "step": 275
+    },
+    {
+      "epoch": 0.6062602965403624,
+      "grad_norm": 0.10102368146181107,
+      "learning_rate": 0.00026628664495114004,
+      "loss": 0.5832,
+      "step": 276
+    },
+    {
+      "epoch": 0.6084568918176826,
+      "grad_norm": 0.08920449018478394,
+      "learning_rate": 0.00026604234527687293,
+      "loss": 0.7854,
+      "step": 277
+    },
+    {
+      "epoch": 0.6106534870950028,
+      "grad_norm": 0.12864069640636444,
+      "learning_rate": 0.0002657980456026058,
+      "loss": 0.8786,
+      "step": 278
+    },
+    {
+      "epoch": 0.6128500823723229,
+      "grad_norm": 0.17306895554065704,
+      "learning_rate": 0.0002655537459283387,
+      "loss": 0.796,
+      "step": 279
+    },
+    {
+      "epoch": 0.615046677649643,
+      "grad_norm": 0.10710766911506653,
+      "learning_rate": 0.00026530944625407165,
+      "loss": 0.692,
+      "step": 280
+    },
+    {
+      "epoch": 0.6172432729269632,
+      "grad_norm": 0.13178376853466034,
+      "learning_rate": 0.00026506514657980454,
+      "loss": 0.6778,
+      "step": 281
+    },
+    {
+      "epoch": 0.6194398682042833,
+      "grad_norm": 0.11179011315107346,
+      "learning_rate": 0.00026482084690553743,
+      "loss": 0.7454,
+      "step": 282
+    },
+    {
+      "epoch": 0.6216364634816035,
+      "grad_norm": 0.12323471158742905,
+      "learning_rate": 0.0002645765472312703,
+      "loss": 0.8803,
+      "step": 283
+    },
+    {
+      "epoch": 0.6238330587589237,
+      "grad_norm": 0.0822269394993782,
+      "learning_rate": 0.0002643322475570032,
+      "loss": 0.7196,
+      "step": 284
+    },
+    {
+      "epoch": 0.6260296540362438,
+      "grad_norm": 0.09452176094055176,
+      "learning_rate": 0.0002640879478827361,
+      "loss": 0.7619,
+      "step": 285
+    },
+    {
+      "epoch": 0.628226249313564,
+      "grad_norm": 0.11845304071903229,
+      "learning_rate": 0.00026384364820846904,
+      "loss": 0.7242,
+      "step": 286
+    },
+    {
+      "epoch": 0.6304228445908842,
+      "grad_norm": 0.10230634361505508,
+      "learning_rate": 0.00026359934853420193,
+      "loss": 0.8355,
+      "step": 287
+    },
+    {
+      "epoch": 0.6326194398682042,
+      "grad_norm": 0.0954984724521637,
+      "learning_rate": 0.0002633550488599348,
+      "loss": 0.6923,
+      "step": 288
+    },
+    {
+      "epoch": 0.6348160351455244,
+      "grad_norm": 0.0816633403301239,
+      "learning_rate": 0.0002631107491856677,
+      "loss": 0.6349,
+      "step": 289
+    },
+    {
+      "epoch": 0.6370126304228446,
+      "grad_norm": 0.1043485477566719,
+      "learning_rate": 0.0002628664495114006,
+      "loss": 0.6992,
+      "step": 290
+    },
+    {
+      "epoch": 0.6392092257001647,
+      "grad_norm": 0.14996330440044403,
+      "learning_rate": 0.0002626221498371335,
+      "loss": 0.9154,
+      "step": 291
+    },
+    {
+      "epoch": 0.6414058209774849,
+      "grad_norm": 0.1418648511171341,
+      "learning_rate": 0.00026237785016286643,
+      "loss": 0.8549,
+      "step": 292
+    },
+    {
+      "epoch": 0.6436024162548051,
+      "grad_norm": 0.10362272709608078,
+      "learning_rate": 0.0002621335504885993,
+      "loss": 0.8914,
+      "step": 293
+    },
+    {
+      "epoch": 0.6457990115321252,
+      "grad_norm": 0.17726963758468628,
+      "learning_rate": 0.00026188925081433226,
+      "loss": 0.8061,
+      "step": 294
+    },
+    {
+      "epoch": 0.6479956068094453,
+      "grad_norm": 0.10512883961200714,
+      "learning_rate": 0.00026164495114006515,
+      "loss": 0.8057,
+      "step": 295
+    },
+    {
+      "epoch": 0.6501922020867655,
+      "grad_norm": 0.13285692036151886,
+      "learning_rate": 0.00026140065146579804,
+      "loss": 0.7931,
+      "step": 296
+    },
+    {
+      "epoch": 0.6523887973640856,
+      "grad_norm": 0.0977858453989029,
+      "learning_rate": 0.00026115635179153093,
+      "loss": 0.6894,
+      "step": 297
+    },
+    {
+      "epoch": 0.6545853926414058,
+      "grad_norm": 0.14479300379753113,
+      "learning_rate": 0.0002609120521172638,
+      "loss": 0.7102,
+      "step": 298
+    },
+    {
+      "epoch": 0.656781987918726,
+      "grad_norm": 0.08681437373161316,
+      "learning_rate": 0.00026066775244299676,
+      "loss": 0.8018,
+      "step": 299
+    },
+    {
+      "epoch": 0.6589785831960461,
+      "grad_norm": 0.13528816401958466,
+      "learning_rate": 0.00026042345276872965,
+      "loss": 0.7743,
+      "step": 300
+    },
+    {
+      "epoch": 0.6611751784733663,
+      "grad_norm": 0.17040468752384186,
+      "learning_rate": 0.00026017915309446254,
+      "loss": 0.8788,
+      "step": 301
+    },
+    {
+      "epoch": 0.6633717737506865,
+      "grad_norm": 0.08778038620948792,
+      "learning_rate": 0.00025993485342019543,
+      "loss": 0.7427,
+      "step": 302
+    },
+    {
+      "epoch": 0.6655683690280065,
+      "grad_norm": 0.19405275583267212,
+      "learning_rate": 0.0002596905537459283,
+      "loss": 0.6809,
+      "step": 303
+    },
+    {
+      "epoch": 0.6677649643053267,
+      "grad_norm": 0.12351009249687195,
+      "learning_rate": 0.0002594462540716612,
+      "loss": 0.8558,
+      "step": 304
+    },
+    {
+      "epoch": 0.6699615595826469,
+      "grad_norm": 0.1889837384223938,
+      "learning_rate": 0.00025920195439739415,
+      "loss": 0.7992,
+      "step": 305
+    },
+    {
+      "epoch": 0.6721581548599671,
+      "grad_norm": 0.15696054697036743,
+      "learning_rate": 0.00025895765472312704,
+      "loss": 0.8302,
+      "step": 306
+    },
+    {
+      "epoch": 0.6743547501372872,
+      "grad_norm": 0.1058712974190712,
+      "learning_rate": 0.00025871335504885993,
+      "loss": 0.8165,
+      "step": 307
+    },
+    {
+      "epoch": 0.6765513454146074,
+      "grad_norm": 0.09586802870035172,
+      "learning_rate": 0.0002584690553745928,
+      "loss": 0.8117,
+      "step": 308
+    },
+    {
+      "epoch": 0.6787479406919276,
+      "grad_norm": 0.10028457641601562,
+      "learning_rate": 0.0002582247557003257,
+      "loss": 0.8345,
+      "step": 309
+    },
+    {
+      "epoch": 0.6809445359692476,
+      "grad_norm": 0.08676378428936005,
+      "learning_rate": 0.00025798045602605865,
+      "loss": 0.831,
+      "step": 310
+    },
+    {
+      "epoch": 0.6831411312465678,
+      "grad_norm": 0.12139932811260223,
+      "learning_rate": 0.00025773615635179154,
+      "loss": 0.6406,
+      "step": 311
+    },
+    {
+      "epoch": 0.685337726523888,
+      "grad_norm": 0.12566658854484558,
+      "learning_rate": 0.00025749185667752443,
+      "loss": 0.8502,
+      "step": 312
+    },
+    {
+      "epoch": 0.6875343218012081,
+      "grad_norm": 0.19190534949302673,
+      "learning_rate": 0.0002572475570032573,
+      "loss": 0.7829,
+      "step": 313
+    },
+    {
+      "epoch": 0.6897309170785283,
+      "grad_norm": 0.10565797239542007,
+      "learning_rate": 0.0002570032573289902,
+      "loss": 0.5257,
+      "step": 314
+    },
+    {
+      "epoch": 0.6919275123558485,
+      "grad_norm": 0.08355598896741867,
+      "learning_rate": 0.0002567589576547231,
+      "loss": 0.5888,
+      "step": 315
+    },
+    {
+      "epoch": 0.6941241076331686,
+      "grad_norm": 0.0873156264424324,
+      "learning_rate": 0.00025651465798045604,
+      "loss": 0.8022,
+      "step": 316
+    },
+    {
+      "epoch": 0.6963207029104888,
+      "grad_norm": 0.09536822140216827,
+      "learning_rate": 0.00025627035830618893,
+      "loss": 0.9403,
+      "step": 317
+    },
+    {
+      "epoch": 0.6985172981878089,
+      "grad_norm": 0.12649884819984436,
+      "learning_rate": 0.0002560260586319218,
+      "loss": 0.8829,
+      "step": 318
+    },
+    {
+      "epoch": 0.700713893465129,
+      "grad_norm": 0.12788763642311096,
+      "learning_rate": 0.0002557817589576547,
+      "loss": 0.7079,
+      "step": 319
+    },
+    {
+      "epoch": 0.7029104887424492,
+      "grad_norm": 0.07442867755889893,
+      "learning_rate": 0.0002555374592833876,
+      "loss": 0.7214,
+      "step": 320
+    },
+    {
+      "epoch": 0.7051070840197694,
+      "grad_norm": 0.0738946720957756,
+      "learning_rate": 0.0002552931596091205,
+      "loss": 0.7843,
+      "step": 321
+    },
+    {
+      "epoch": 0.7073036792970895,
+      "grad_norm": 0.09450596570968628,
+      "learning_rate": 0.00025504885993485343,
+      "loss": 0.8126,
+      "step": 322
+    },
+    {
+      "epoch": 0.7095002745744097,
+      "grad_norm": 0.10171910375356674,
+      "learning_rate": 0.0002548045602605863,
+      "loss": 0.6814,
+      "step": 323
+    },
+    {
+      "epoch": 0.7116968698517299,
+      "grad_norm": 0.12972281873226166,
+      "learning_rate": 0.0002545602605863192,
+      "loss": 0.8633,
+      "step": 324
+    },
+    {
+      "epoch": 0.7138934651290499,
+      "grad_norm": 0.1559072732925415,
+      "learning_rate": 0.0002543159609120521,
+      "loss": 0.8262,
+      "step": 325
+    },
+    {
+      "epoch": 0.7160900604063701,
+      "grad_norm": 0.12674300372600555,
+      "learning_rate": 0.000254071661237785,
+      "loss": 0.7681,
+      "step": 326
+    },
+    {
+      "epoch": 0.7182866556836903,
+      "grad_norm": 0.13792023062705994,
+      "learning_rate": 0.00025382736156351793,
+      "loss": 0.7844,
+      "step": 327
+    },
+    {
+      "epoch": 0.7204832509610104,
+      "grad_norm": 0.10216391831636429,
+      "learning_rate": 0.0002535830618892508,
+      "loss": 0.8948,
+      "step": 328
+    },
+    {
+      "epoch": 0.7226798462383306,
+      "grad_norm": 0.11408744752407074,
+      "learning_rate": 0.0002533387622149837,
+      "loss": 0.7671,
+      "step": 329
+    },
+    {
+      "epoch": 0.7248764415156508,
+      "grad_norm": 0.12805496156215668,
+      "learning_rate": 0.0002530944625407166,
+      "loss": 0.7524,
+      "step": 330
+    },
+    {
+      "epoch": 0.7270730367929709,
+      "grad_norm": 0.27026745676994324,
+      "learning_rate": 0.0002528501628664495,
+      "loss": 0.9059,
+      "step": 331
+    },
+    {
+      "epoch": 0.729269632070291,
+      "grad_norm": 0.12689338624477386,
+      "learning_rate": 0.0002526058631921824,
+      "loss": 0.7747,
+      "step": 332
+    },
+    {
+      "epoch": 0.7314662273476112,
+      "grad_norm": 0.07850451022386551,
+      "learning_rate": 0.0002523615635179153,
+      "loss": 0.8135,
+      "step": 333
+    },
+    {
+      "epoch": 0.7336628226249313,
+      "grad_norm": 0.12431463599205017,
+      "learning_rate": 0.0002521172638436482,
+      "loss": 0.882,
+      "step": 334
+    },
+    {
+      "epoch": 0.7358594179022515,
+      "grad_norm": 0.10434425622224808,
+      "learning_rate": 0.0002518729641693811,
+      "loss": 0.7908,
+      "step": 335
+    },
+    {
+      "epoch": 0.7380560131795717,
+      "grad_norm": 0.15843713283538818,
+      "learning_rate": 0.000251628664495114,
+      "loss": 0.7676,
+      "step": 336
+    },
+    {
+      "epoch": 0.7402526084568918,
+      "grad_norm": 0.12522131204605103,
+      "learning_rate": 0.0002513843648208469,
+      "loss": 0.9121,
+      "step": 337
+    },
+    {
+      "epoch": 0.742449203734212,
+      "grad_norm": 0.14041388034820557,
+      "learning_rate": 0.0002511400651465798,
+      "loss": 0.7543,
+      "step": 338
+    },
+    {
+      "epoch": 0.7446457990115322,
+      "grad_norm": 0.10586488246917725,
+      "learning_rate": 0.0002508957654723127,
+      "loss": 0.8384,
+      "step": 339
+    },
+    {
+      "epoch": 0.7468423942888522,
+      "grad_norm": 0.12159724533557892,
+      "learning_rate": 0.0002506514657980456,
+      "loss": 0.8208,
+      "step": 340
+    },
+    {
+      "epoch": 0.7490389895661724,
+      "grad_norm": 0.10033271461725235,
+      "learning_rate": 0.0002504071661237785,
+      "loss": 0.7285,
+      "step": 341
+    },
+    {
+      "epoch": 0.7512355848434926,
+      "grad_norm": 0.09773661196231842,
+      "learning_rate": 0.0002501628664495114,
+      "loss": 0.6712,
+      "step": 342
+    },
+    {
+      "epoch": 0.7534321801208127,
+      "grad_norm": 0.07319659739732742,
+      "learning_rate": 0.00024991856677524427,
+      "loss": 0.7162,
+      "step": 343
+    },
+    {
+      "epoch": 0.7556287753981329,
+      "grad_norm": 0.10162502527236938,
+      "learning_rate": 0.0002496742671009772,
+      "loss": 0.7543,
+      "step": 344
+    },
+    {
+      "epoch": 0.7578253706754531,
+      "grad_norm": 0.08760935068130493,
+      "learning_rate": 0.0002494299674267101,
+      "loss": 0.67,
+      "step": 345
+    },
+    {
+      "epoch": 0.7600219659527732,
+      "grad_norm": 0.11542056500911713,
+      "learning_rate": 0.000249185667752443,
+      "loss": 0.7698,
+      "step": 346
+    },
+    {
+      "epoch": 0.7622185612300933,
+      "grad_norm": 0.08908765763044357,
+      "learning_rate": 0.0002489413680781759,
+      "loss": 0.7915,
+      "step": 347
+    },
+    {
+      "epoch": 0.7644151565074135,
+      "grad_norm": 0.13499994575977325,
+      "learning_rate": 0.00024869706840390877,
+      "loss": 0.8985,
+      "step": 348
+    },
+    {
+      "epoch": 0.7666117517847336,
+      "grad_norm": 0.225164994597435,
+      "learning_rate": 0.0002484527687296417,
+      "loss": 0.5757,
+      "step": 349
+    },
+    {
+      "epoch": 0.7688083470620538,
+      "grad_norm": 0.10587360709905624,
+      "learning_rate": 0.0002482084690553746,
+      "loss": 0.9471,
+      "step": 350
+    },
+    {
+      "epoch": 0.771004942339374,
+      "grad_norm": 0.1371883898973465,
+      "learning_rate": 0.0002479641693811075,
+      "loss": 0.9362,
+      "step": 351
+    },
+    {
+      "epoch": 0.7732015376166941,
+      "grad_norm": 0.14055342972278595,
+      "learning_rate": 0.0002477198697068404,
+      "loss": 0.8156,
+      "step": 352
+    },
+    {
+      "epoch": 0.7753981328940143,
+      "grad_norm": 0.08085603266954422,
+      "learning_rate": 0.00024747557003257327,
+      "loss": 0.6305,
+      "step": 353
+    },
+    {
+      "epoch": 0.7775947281713345,
+      "grad_norm": 0.07912547141313553,
+      "learning_rate": 0.00024723127035830616,
+      "loss": 0.8295,
+      "step": 354
+    },
+    {
+      "epoch": 0.7797913234486545,
+      "grad_norm": 0.12284508347511292,
+      "learning_rate": 0.0002469869706840391,
+      "loss": 0.5643,
+      "step": 355
+    },
+    {
+      "epoch": 0.7819879187259747,
+      "grad_norm": 0.08401278406381607,
+      "learning_rate": 0.000246742671009772,
+      "loss": 0.6622,
+      "step": 356
+    },
+    {
+      "epoch": 0.7841845140032949,
+      "grad_norm": 0.07165351510047913,
+      "learning_rate": 0.0002464983713355049,
+      "loss": 0.7716,
+      "step": 357
+    },
+    {
+      "epoch": 0.786381109280615,
+      "grad_norm": 0.10170119255781174,
+      "learning_rate": 0.00024625407166123777,
+      "loss": 0.7416,
+      "step": 358
+    },
+    {
+      "epoch": 0.7885777045579352,
+      "grad_norm": 0.14219675958156586,
+      "learning_rate": 0.00024600977198697066,
+      "loss": 0.6682,
+      "step": 359
+    },
+    {
+      "epoch": 0.7907742998352554,
+      "grad_norm": 0.08274268358945847,
+      "learning_rate": 0.00024576547231270354,
+      "loss": 0.6844,
+      "step": 360
+    },
+    {
+      "epoch": 0.7929708951125755,
+      "grad_norm": 0.08963588625192642,
+      "learning_rate": 0.0002455211726384365,
+      "loss": 0.7972,
+      "step": 361
+    },
+    {
+      "epoch": 0.7951674903898956,
+      "grad_norm": 0.10314128547906876,
+      "learning_rate": 0.0002452768729641694,
+      "loss": 0.8114,
+      "step": 362
+    },
+    {
+      "epoch": 0.7973640856672158,
+      "grad_norm": 0.08634983748197556,
+      "learning_rate": 0.00024503257328990227,
+      "loss": 0.6335,
+      "step": 363
+    },
+    {
+      "epoch": 0.7995606809445359,
+      "grad_norm": 0.11043889820575714,
+      "learning_rate": 0.00024478827361563516,
+      "loss": 0.7429,
+      "step": 364
+    },
+    {
+      "epoch": 0.8017572762218561,
+      "grad_norm": 0.14205126464366913,
+      "learning_rate": 0.00024454397394136804,
+      "loss": 0.5711,
+      "step": 365
+    },
+    {
+      "epoch": 0.8039538714991763,
+      "grad_norm": 0.09631361067295074,
+      "learning_rate": 0.000244299674267101,
+      "loss": 0.8278,
+      "step": 366
+    },
+    {
+      "epoch": 0.8061504667764964,
+      "grad_norm": 0.1846916675567627,
+      "learning_rate": 0.00024405537459283385,
+      "loss": 0.781,
+      "step": 367
+    },
+    {
+      "epoch": 0.8083470620538166,
+      "grad_norm": 0.08213674277067184,
+      "learning_rate": 0.00024381107491856677,
+      "loss": 0.6438,
+      "step": 368
+    },
+    {
+      "epoch": 0.8105436573311368,
+      "grad_norm": 0.07400628179311752,
+      "learning_rate": 0.00024356677524429966,
+      "loss": 0.6704,
+      "step": 369
+    },
+    {
+      "epoch": 0.8127402526084568,
+      "grad_norm": 0.08337616175413132,
+      "learning_rate": 0.00024332247557003257,
+      "loss": 0.5662,
+      "step": 370
+    },
+    {
+      "epoch": 0.814936847885777,
+      "grad_norm": 0.07169657200574875,
+      "learning_rate": 0.00024307817589576546,
+      "loss": 0.8197,
+      "step": 371
+    },
+    {
+      "epoch": 0.8171334431630972,
+      "grad_norm": 0.1379915177822113,
+      "learning_rate": 0.00024283387622149835,
+      "loss": 0.7962,
+      "step": 372
+    },
+    {
+      "epoch": 0.8193300384404174,
+      "grad_norm": 0.08688277751207352,
+      "learning_rate": 0.00024258957654723127,
+      "loss": 0.6965,
+      "step": 373
+    },
+    {
+      "epoch": 0.8215266337177375,
+      "grad_norm": 0.1228555291891098,
+      "learning_rate": 0.00024234527687296416,
+      "loss": 0.7557,
+      "step": 374
+    },
+    {
+      "epoch": 0.8237232289950577,
+      "grad_norm": 0.10932645201683044,
+      "learning_rate": 0.00024210097719869705,
+      "loss": 0.8649,
+      "step": 375
+    },
+    {
+      "epoch": 0.8259198242723779,
+      "grad_norm": 0.22051644325256348,
+      "learning_rate": 0.00024185667752442996,
+      "loss": 0.8258,
+      "step": 376
+    },
+    {
+      "epoch": 0.828116419549698,
+      "grad_norm": 0.09149803966283798,
+      "learning_rate": 0.00024161237785016285,
+      "loss": 0.7787,
+      "step": 377
+    },
+    {
+      "epoch": 0.8303130148270181,
+      "grad_norm": 0.0913648009300232,
+      "learning_rate": 0.00024136807817589574,
+      "loss": 0.7377,
+      "step": 378
+    },
+    {
+      "epoch": 0.8325096101043383,
+      "grad_norm": 0.06680730730295181,
+      "learning_rate": 0.00024112377850162866,
+      "loss": 0.5654,
+      "step": 379
+    },
+    {
+      "epoch": 0.8347062053816584,
+      "grad_norm": 0.10964051634073257,
+      "learning_rate": 0.00024087947882736155,
+      "loss": 0.812,
+      "step": 380
+    },
+    {
+      "epoch": 0.8369028006589786,
+      "grad_norm": 0.08345823734998703,
+      "learning_rate": 0.00024063517915309446,
+      "loss": 0.7189,
+      "step": 381
+    },
+    {
+      "epoch": 0.8390993959362988,
+      "grad_norm": 0.09597508609294891,
+      "learning_rate": 0.00024039087947882735,
+      "loss": 0.821,
+      "step": 382
+    },
+    {
+      "epoch": 0.8412959912136189,
+      "grad_norm": 0.19232520461082458,
+      "learning_rate": 0.00024014657980456024,
+      "loss": 0.8221,
+      "step": 383
+    },
+    {
+      "epoch": 0.8434925864909391,
+      "grad_norm": 0.15436063706874847,
+      "learning_rate": 0.00023990228013029316,
+      "loss": 0.8571,
+      "step": 384
+    },
+    {
+      "epoch": 0.8456891817682592,
+      "grad_norm": 0.13297942280769348,
+      "learning_rate": 0.00023965798045602605,
+      "loss": 0.9594,
+      "step": 385
+    },
+    {
+      "epoch": 0.8478857770455793,
+      "grad_norm": 0.11841193586587906,
+      "learning_rate": 0.00023941368078175893,
+      "loss": 0.8862,
+      "step": 386
+    },
+    {
+      "epoch": 0.8500823723228995,
+      "grad_norm": 0.11999719589948654,
+      "learning_rate": 0.00023916938110749185,
+      "loss": 0.7066,
+      "step": 387
+    },
+    {
+      "epoch": 0.8522789676002197,
+      "grad_norm": 0.2062544971704483,
+      "learning_rate": 0.00023892508143322474,
+      "loss": 0.813,
+      "step": 388
+    },
+    {
+      "epoch": 0.8544755628775398,
+      "grad_norm": 0.07809589803218842,
+      "learning_rate": 0.00023868078175895763,
+      "loss": 0.5947,
+      "step": 389
+    },
+    {
+      "epoch": 0.85667215815486,
+      "grad_norm": 0.1531992256641388,
+      "learning_rate": 0.00023843648208469055,
+      "loss": 0.7036,
+      "step": 390
+    },
+    {
+      "epoch": 0.8588687534321802,
+      "grad_norm": 0.09320369362831116,
+      "learning_rate": 0.00023819218241042343,
+      "loss": 0.7691,
+      "step": 391
+    },
+    {
+      "epoch": 0.8610653487095002,
+      "grad_norm": 0.08531806617975235,
+      "learning_rate": 0.00023794788273615632,
+      "loss": 1.0065,
+      "step": 392
+    },
+    {
+      "epoch": 0.8632619439868204,
+      "grad_norm": 0.13786737620830536,
+      "learning_rate": 0.00023770358306188924,
+      "loss": 1.0711,
+      "step": 393
+    },
+    {
+      "epoch": 0.8654585392641406,
+      "grad_norm": 0.08890886604785919,
+      "learning_rate": 0.00023745928338762213,
+      "loss": 0.7496,
+      "step": 394
+    },
+    {
+      "epoch": 0.8676551345414607,
+      "grad_norm": 0.08928348869085312,
+      "learning_rate": 0.00023721498371335505,
+      "loss": 0.7125,
+      "step": 395
+    },
+    {
+      "epoch": 0.8698517298187809,
+      "grad_norm": 0.10368994623422623,
+      "learning_rate": 0.00023697068403908794,
+      "loss": 0.7742,
+      "step": 396
+    },
+    {
+      "epoch": 0.8720483250961011,
+      "grad_norm": 0.08876747637987137,
+      "learning_rate": 0.00023672638436482082,
+      "loss": 0.7258,
+      "step": 397
+    },
+    {
+      "epoch": 0.8742449203734212,
+      "grad_norm": 0.07644807547330856,
+      "learning_rate": 0.00023648208469055374,
+      "loss": 0.671,
+      "step": 398
+    },
+    {
+      "epoch": 0.8764415156507414,
+      "grad_norm": 0.12549491226673126,
+      "learning_rate": 0.00023623778501628663,
+      "loss": 0.7845,
+      "step": 399
+    },
+    {
+      "epoch": 0.8786381109280615,
+      "grad_norm": 0.12591737508773804,
+      "learning_rate": 0.00023599348534201952,
+      "loss": 0.8131,
+      "step": 400
+    },
+    {
+      "epoch": 0.8808347062053816,
+      "grad_norm": 0.12803559005260468,
+      "learning_rate": 0.00023574918566775244,
+      "loss": 0.6434,
+      "step": 401
+    },
+    {
+      "epoch": 0.8830313014827018,
+      "grad_norm": 0.07376369833946228,
+      "learning_rate": 0.00023550488599348532,
+      "loss": 0.7297,
+      "step": 402
+    },
+    {
+      "epoch": 0.885227896760022,
+      "grad_norm": 0.10205817222595215,
+      "learning_rate": 0.00023526058631921821,
+      "loss": 0.9146,
+      "step": 403
+    },
+    {
+      "epoch": 0.8874244920373421,
+      "grad_norm": 0.06458426266908646,
+      "learning_rate": 0.00023501628664495113,
+      "loss": 0.9088,
+      "step": 404
+    },
+    {
+      "epoch": 0.8896210873146623,
+      "grad_norm": 0.08247632533311844,
+      "learning_rate": 0.00023477198697068402,
+      "loss": 0.6376,
+      "step": 405
+    },
+    {
+      "epoch": 0.8918176825919825,
+      "grad_norm": 0.08773287385702133,
+      "learning_rate": 0.0002345276872964169,
+      "loss": 0.7095,
+      "step": 406
+    },
+    {
+      "epoch": 0.8940142778693025,
+      "grad_norm": 0.08690772950649261,
+      "learning_rate": 0.00023428338762214982,
+      "loss": 0.5956,
+      "step": 407
+    },
+    {
+      "epoch": 0.8962108731466227,
+      "grad_norm": 0.08630208671092987,
+      "learning_rate": 0.00023403908794788271,
+      "loss": 0.6475,
+      "step": 408
+    },
+    {
+      "epoch": 0.8984074684239429,
+      "grad_norm": 0.11124998331069946,
+      "learning_rate": 0.00023379478827361563,
+      "loss": 0.6913,
+      "step": 409
+    },
+    {
+      "epoch": 0.900604063701263,
+      "grad_norm": 0.11166873574256897,
+      "learning_rate": 0.00023355048859934852,
+      "loss": 0.8632,
+      "step": 410
+    },
+    {
+      "epoch": 0.9028006589785832,
+      "grad_norm": 0.07521384954452515,
+      "learning_rate": 0.0002333061889250814,
+      "loss": 0.6463,
+      "step": 411
+    },
+    {
+      "epoch": 0.9049972542559034,
+      "grad_norm": 0.10911514610052109,
+      "learning_rate": 0.00023306188925081432,
+      "loss": 0.6734,
+      "step": 412
+    },
+    {
+      "epoch": 0.9071938495332235,
+      "grad_norm": 0.6635527610778809,
+      "learning_rate": 0.00023281758957654721,
+      "loss": 0.736,
+      "step": 413
+    },
+    {
+      "epoch": 0.9093904448105437,
+      "grad_norm": 0.11321001499891281,
+      "learning_rate": 0.0002325732899022801,
+      "loss": 0.8237,
+      "step": 414
+    },
+    {
+      "epoch": 0.9115870400878638,
+      "grad_norm": 0.07390578836202621,
+      "learning_rate": 0.00023232899022801302,
+      "loss": 0.696,
+      "step": 415
+    },
+    {
+      "epoch": 0.9137836353651839,
+      "grad_norm": 0.09732091426849365,
+      "learning_rate": 0.0002320846905537459,
+      "loss": 0.8936,
+      "step": 416
+    },
+    {
+      "epoch": 0.9159802306425041,
+      "grad_norm": 0.12673120200634003,
+      "learning_rate": 0.0002318403908794788,
+      "loss": 0.7672,
+      "step": 417
+    },
+    {
+      "epoch": 0.9181768259198243,
+      "grad_norm": 0.1106930822134018,
+      "learning_rate": 0.00023159609120521171,
+      "loss": 0.5485,
+      "step": 418
+    },
+    {
+      "epoch": 0.9203734211971444,
+      "grad_norm": 0.11515068262815475,
+      "learning_rate": 0.0002313517915309446,
+      "loss": 0.6609,
+      "step": 419
+    },
+    {
+      "epoch": 0.9225700164744646,
+      "grad_norm": 0.09618733823299408,
+      "learning_rate": 0.00023110749185667752,
+      "loss": 0.7737,
+      "step": 420
+    },
+    {
+      "epoch": 0.9247666117517848,
+      "grad_norm": 0.10351020842790604,
+      "learning_rate": 0.0002308631921824104,
+      "loss": 0.7252,
+      "step": 421
+    },
+    {
+      "epoch": 0.9269632070291048,
+      "grad_norm": 0.09867105633020401,
+      "learning_rate": 0.0002306188925081433,
+      "loss": 0.7189,
+      "step": 422
+    },
+    {
+      "epoch": 0.929159802306425,
+      "grad_norm": 0.09880604594945908,
+      "learning_rate": 0.00023037459283387621,
+      "loss": 0.9311,
+      "step": 423
+    },
+    {
+      "epoch": 0.9313563975837452,
+      "grad_norm": 0.40088096261024475,
+      "learning_rate": 0.0002301302931596091,
+      "loss": 0.7273,
+      "step": 424
+    },
+    {
+      "epoch": 0.9335529928610653,
+      "grad_norm": 0.12686477601528168,
+      "learning_rate": 0.000229885993485342,
+      "loss": 0.7122,
+      "step": 425
+    },
+    {
+      "epoch": 0.9357495881383855,
+      "grad_norm": 0.0960114374756813,
+      "learning_rate": 0.0002296416938110749,
+      "loss": 0.6633,
+      "step": 426
+    },
+    {
+      "epoch": 0.9379461834157057,
+      "grad_norm": 0.07050749659538269,
+      "learning_rate": 0.0002293973941368078,
+      "loss": 0.7086,
+      "step": 427
+    },
+    {
+      "epoch": 0.9401427786930258,
+      "grad_norm": 0.09293071925640106,
+      "learning_rate": 0.0002291530944625407,
+      "loss": 0.5904,
+      "step": 428
+    },
+    {
+      "epoch": 0.942339373970346,
+      "grad_norm": 0.10552413016557693,
+      "learning_rate": 0.0002289087947882736,
+      "loss": 0.5512,
+      "step": 429
+    },
+    {
+      "epoch": 0.9445359692476661,
+      "grad_norm": 0.10490048676729202,
+      "learning_rate": 0.0002286644951140065,
+      "loss": 0.7964,
+      "step": 430
+    },
+    {
+      "epoch": 0.9467325645249862,
+      "grad_norm": 0.10114052146673203,
+      "learning_rate": 0.00022842019543973938,
+      "loss": 0.6181,
+      "step": 431
+    },
+    {
+      "epoch": 0.9489291598023064,
+      "grad_norm": 0.11705774813890457,
+      "learning_rate": 0.0002281758957654723,
+      "loss": 0.8922,
+      "step": 432
+    },
+    {
+      "epoch": 0.9511257550796266,
+      "grad_norm": 0.08251733332872391,
+      "learning_rate": 0.0002279315960912052,
+      "loss": 0.785,
+      "step": 433
+    },
+    {
+      "epoch": 0.9533223503569467,
+      "grad_norm": 0.15024127066135406,
+      "learning_rate": 0.0002276872964169381,
+      "loss": 0.6289,
+      "step": 434
+    },
+    {
+      "epoch": 0.9555189456342669,
+      "grad_norm": 0.0716300904750824,
+      "learning_rate": 0.000227442996742671,
+      "loss": 0.6322,
+      "step": 435
+    },
+    {
+      "epoch": 0.9577155409115871,
+      "grad_norm": 0.10501925647258759,
+      "learning_rate": 0.00022719869706840388,
+      "loss": 0.6301,
+      "step": 436
+    },
+    {
+      "epoch": 0.9599121361889071,
+      "grad_norm": 0.09028838574886322,
+      "learning_rate": 0.0002269543973941368,
+      "loss": 0.7388,
+      "step": 437
+    },
+    {
+      "epoch": 0.9621087314662273,
+      "grad_norm": 0.0888151153922081,
+      "learning_rate": 0.0002267100977198697,
+      "loss": 0.6651,
+      "step": 438
+    },
+    {
+      "epoch": 0.9643053267435475,
+      "grad_norm": 0.17989221215248108,
+      "learning_rate": 0.00022646579804560258,
+      "loss": 0.8906,
+      "step": 439
+    },
+    {
+      "epoch": 0.9665019220208677,
+      "grad_norm": 0.08109401166439056,
+      "learning_rate": 0.0002262214983713355,
+      "loss": 0.8134,
+      "step": 440
+    },
+    {
+      "epoch": 0.9686985172981878,
+      "grad_norm": 0.15070408582687378,
+      "learning_rate": 0.00022597719869706838,
+      "loss": 0.6957,
+      "step": 441
+    },
+    {
+      "epoch": 0.970895112575508,
+      "grad_norm": 0.11629512161016464,
+      "learning_rate": 0.00022573289902280127,
+      "loss": 0.6416,
+      "step": 442
+    },
+    {
+      "epoch": 0.9730917078528282,
+      "grad_norm": 0.0793461725115776,
+      "learning_rate": 0.0002254885993485342,
+      "loss": 0.7378,
+      "step": 443
+    },
+    {
+      "epoch": 0.9752883031301482,
+      "grad_norm": 0.08339542895555496,
+      "learning_rate": 0.00022524429967426708,
+      "loss": 0.831,
+      "step": 444
+    },
+    {
+      "epoch": 0.9774848984074684,
+      "grad_norm": 0.11858515441417694,
+      "learning_rate": 0.000225,
+      "loss": 0.7758,
+      "step": 445
+    },
+    {
+      "epoch": 0.9796814936847886,
+      "grad_norm": 0.15611110627651215,
+      "learning_rate": 0.00022475570032573288,
+      "loss": 0.9304,
+      "step": 446
+    },
+    {
+      "epoch": 0.9818780889621087,
+      "grad_norm": 0.11646619439125061,
+      "learning_rate": 0.00022451140065146577,
+      "loss": 0.8659,
+      "step": 447
+    },
+    {
+      "epoch": 0.9840746842394289,
+      "grad_norm": 0.161260724067688,
+      "learning_rate": 0.0002242671009771987,
+      "loss": 0.7008,
+      "step": 448
+    },
+    {
+      "epoch": 0.9862712795167491,
+      "grad_norm": 0.079967200756073,
+      "learning_rate": 0.00022402280130293158,
+      "loss": 0.8043,
+      "step": 449
+    },
+    {
+      "epoch": 0.9884678747940692,
+      "grad_norm": 0.08696119487285614,
+      "learning_rate": 0.00022377850162866447,
+      "loss": 0.7339,
+      "step": 450
+    },
+    {
+      "epoch": 0.9906644700713894,
+      "grad_norm": 0.09183841198682785,
+      "learning_rate": 0.00022353420195439738,
+      "loss": 0.719,
+      "step": 451
+    },
+    {
+      "epoch": 0.9928610653487095,
+      "grad_norm": 0.09511847794055939,
+      "learning_rate": 0.00022328990228013027,
+      "loss": 0.7483,
+      "step": 452
+    },
+    {
+      "epoch": 0.9950576606260296,
+      "grad_norm": 0.07402326911687851,
+      "learning_rate": 0.00022304560260586316,
+      "loss": 0.8264,
+      "step": 453
+    },
+    {
+      "epoch": 0.9972542559033498,
+      "grad_norm": 0.14026030898094177,
+      "learning_rate": 0.00022280130293159608,
+      "loss": 0.6206,
+      "step": 454
+    },
+    {
+      "epoch": 0.99945085118067,
+      "grad_norm": 0.06916913390159607,
+      "learning_rate": 0.00022255700325732897,
+      "loss": 0.7358,
+      "step": 455
+    },
+    {
+      "epoch": 1.00164744645799,
+      "grad_norm": 0.1173158809542656,
+      "learning_rate": 0.00022231270358306186,
+      "loss": 0.8494,
+      "step": 456
+    },
+    {
+      "epoch": 1.0038440417353103,
+      "grad_norm": 0.06738580018281937,
+      "learning_rate": 0.00022206840390879477,
+      "loss": 0.8041,
+      "step": 457
+    },
+    {
+      "epoch": 1.0060406370126305,
+      "grad_norm": 0.09703294932842255,
+      "learning_rate": 0.00022182410423452766,
+      "loss": 0.749,
+      "step": 458
+    },
+    {
+      "epoch": 1.0082372322899507,
+      "grad_norm": 0.10583224892616272,
+      "learning_rate": 0.00022157980456026058,
+      "loss": 0.6707,
+      "step": 459
+    },
+    {
+      "epoch": 1.0104338275672706,
+      "grad_norm": 0.09841059148311615,
+      "learning_rate": 0.00022133550488599347,
+      "loss": 0.6502,
+      "step": 460
+    },
+    {
+      "epoch": 1.0126304228445908,
+      "grad_norm": 0.1150907352566719,
+      "learning_rate": 0.00022109120521172636,
+      "loss": 0.5928,
+      "step": 461
+    },
+    {
+      "epoch": 1.014827018121911,
+      "grad_norm": 0.09026665985584259,
+      "learning_rate": 0.00022084690553745927,
+      "loss": 1.0271,
+      "step": 462
+    },
+    {
+      "epoch": 1.0170236133992312,
+      "grad_norm": 0.07217419147491455,
+      "learning_rate": 0.00022060260586319216,
+      "loss": 0.8507,
+      "step": 463
+    },
+    {
+      "epoch": 1.0192202086765514,
+      "grad_norm": 0.09241258352994919,
+      "learning_rate": 0.00022035830618892505,
+      "loss": 0.6878,
+      "step": 464
+    },
+    {
+      "epoch": 1.0214168039538716,
+      "grad_norm": 0.10729347169399261,
+      "learning_rate": 0.00022011400651465797,
+      "loss": 0.8906,
+      "step": 465
+    },
+    {
+      "epoch": 1.0236133992311915,
+      "grad_norm": 0.09274546802043915,
+      "learning_rate": 0.00021986970684039086,
+      "loss": 0.906,
+      "step": 466
+    },
+    {
+      "epoch": 1.0258099945085117,
+      "grad_norm": 0.10703285038471222,
+      "learning_rate": 0.00021962540716612375,
+      "loss": 0.7339,
+      "step": 467
+    },
+    {
+      "epoch": 1.028006589785832,
+      "grad_norm": 0.09155743569135666,
+      "learning_rate": 0.00021938110749185666,
+      "loss": 0.6894,
+      "step": 468
+    },
+    {
+      "epoch": 1.0302031850631521,
+      "grad_norm": 0.17693020403385162,
+      "learning_rate": 0.00021913680781758955,
+      "loss": 0.8182,
+      "step": 469
+    },
+    {
+      "epoch": 1.0323997803404723,
+      "grad_norm": 0.224022775888443,
+      "learning_rate": 0.00021889250814332244,
+      "loss": 0.8177,
+      "step": 470
+    },
+    {
+      "epoch": 1.0345963756177925,
+      "grad_norm": 0.1012631431221962,
+      "learning_rate": 0.00021864820846905536,
+      "loss": 0.6436,
+      "step": 471
+    },
+    {
+      "epoch": 1.0367929708951125,
+      "grad_norm": 0.14975659549236298,
+      "learning_rate": 0.00021840390879478825,
+      "loss": 0.7442,
+      "step": 472
+    },
+    {
+      "epoch": 1.0389895661724327,
+      "grad_norm": 0.10164142400026321,
+      "learning_rate": 0.00021815960912052116,
+      "loss": 0.9261,
+      "step": 473
+    },
+    {
+      "epoch": 1.0411861614497528,
+      "grad_norm": 0.1029333770275116,
+      "learning_rate": 0.00021791530944625405,
+      "loss": 0.724,
+      "step": 474
+    },
+    {
+      "epoch": 1.043382756727073,
+      "grad_norm": 0.09123466908931732,
+      "learning_rate": 0.00021767100977198694,
+      "loss": 0.7254,
+      "step": 475
+    },
+    {
+      "epoch": 1.0455793520043932,
+      "grad_norm": 0.19603733718395233,
+      "learning_rate": 0.00021742671009771986,
+      "loss": 0.8906,
+      "step": 476
+    },
+    {
+      "epoch": 1.0477759472817134,
+      "grad_norm": 0.0775158554315567,
+      "learning_rate": 0.00021718241042345275,
+      "loss": 0.6329,
+      "step": 477
+    },
+    {
+      "epoch": 1.0499725425590336,
+      "grad_norm": 0.11025606840848923,
+      "learning_rate": 0.00021693811074918564,
+      "loss": 0.7964,
+      "step": 478
+    },
+    {
+      "epoch": 1.0521691378363536,
+      "grad_norm": 0.11824944615364075,
+      "learning_rate": 0.00021669381107491855,
+      "loss": 0.7901,
+      "step": 479
+    },
+    {
+      "epoch": 1.0543657331136738,
+      "grad_norm": 0.10872384160757065,
+      "learning_rate": 0.00021644951140065144,
+      "loss": 0.6682,
+      "step": 480
+    },
+    {
+      "epoch": 1.056562328390994,
+      "grad_norm": 0.06965261697769165,
+      "learning_rate": 0.00021620521172638433,
+      "loss": 0.6652,
+      "step": 481
+    },
+    {
+      "epoch": 1.0587589236683141,
+      "grad_norm": 0.21465571224689484,
+      "learning_rate": 0.00021596091205211725,
+      "loss": 0.6674,
+      "step": 482
+    },
+    {
+      "epoch": 1.0609555189456343,
+      "grad_norm": 0.11482791602611542,
+      "learning_rate": 0.00021571661237785014,
+      "loss": 0.6866,
+      "step": 483
+    },
+    {
+      "epoch": 1.0631521142229543,
+      "grad_norm": 0.08799201995134354,
+      "learning_rate": 0.00021547231270358305,
+      "loss": 0.7586,
+      "step": 484
+    },
+    {
+      "epoch": 1.0653487095002745,
+      "grad_norm": 0.07813264429569244,
+      "learning_rate": 0.00021522801302931594,
+      "loss": 0.7433,
+      "step": 485
+    },
+    {
+      "epoch": 1.0675453047775947,
+      "grad_norm": 0.15133236348628998,
+      "learning_rate": 0.00021498371335504883,
+      "loss": 0.6218,
+      "step": 486
+    },
+    {
+      "epoch": 1.0697419000549149,
+      "grad_norm": 0.06786596029996872,
+      "learning_rate": 0.00021473941368078175,
+      "loss": 0.7092,
+      "step": 487
+    },
+    {
+      "epoch": 1.071938495332235,
+      "grad_norm": 0.10003423690795898,
+      "learning_rate": 0.00021449511400651464,
+      "loss": 0.805,
+      "step": 488
+    },
+    {
+      "epoch": 1.0741350906095553,
+      "grad_norm": 0.09297288954257965,
+      "learning_rate": 0.00021425081433224753,
+      "loss": 0.6626,
+      "step": 489
+    },
+    {
+      "epoch": 1.0763316858868754,
+      "grad_norm": 0.06513010710477829,
+      "learning_rate": 0.00021400651465798044,
+      "loss": 0.6241,
+      "step": 490
+    },
+    {
+      "epoch": 1.0785282811641954,
+      "grad_norm": 0.18739236891269684,
+      "learning_rate": 0.00021376221498371333,
+      "loss": 0.9058,
+      "step": 491
+    },
+    {
+      "epoch": 1.0807248764415156,
+      "grad_norm": 0.08604872226715088,
+      "learning_rate": 0.00021351791530944622,
+      "loss": 0.668,
+      "step": 492
+    },
+    {
+      "epoch": 1.0829214717188358,
+      "grad_norm": 0.09682892262935638,
+      "learning_rate": 0.00021327361563517914,
+      "loss": 0.8214,
+      "step": 493
+    },
+    {
+      "epoch": 1.085118066996156,
+      "grad_norm": 0.10096686333417892,
+      "learning_rate": 0.00021302931596091203,
+      "loss": 0.6197,
+      "step": 494
+    },
+    {
+      "epoch": 1.0873146622734762,
+      "grad_norm": 0.10254428535699844,
+      "learning_rate": 0.00021278501628664491,
+      "loss": 0.7271,
+      "step": 495
+    },
+    {
+      "epoch": 1.0895112575507964,
+      "grad_norm": 0.1715797781944275,
+      "learning_rate": 0.00021254071661237783,
+      "loss": 0.6713,
+      "step": 496
+    },
+    {
+      "epoch": 1.0917078528281163,
+      "grad_norm": 0.1246112510561943,
+      "learning_rate": 0.00021229641693811072,
+      "loss": 0.6515,
+      "step": 497
+    },
+    {
+      "epoch": 1.0939044481054365,
+      "grad_norm": 0.09628249704837799,
+      "learning_rate": 0.00021205211726384364,
+      "loss": 0.7221,
+      "step": 498
+    },
+    {
+      "epoch": 1.0961010433827567,
+      "grad_norm": 0.11169828474521637,
+      "learning_rate": 0.00021180781758957653,
+      "loss": 0.7702,
+      "step": 499
+    },
+    {
+      "epoch": 1.098297638660077,
+      "grad_norm": 0.10380955785512924,
+      "learning_rate": 0.00021156351791530941,
+      "loss": 0.8514,
+      "step": 500
+    },
+    {
+      "epoch": 1.100494233937397,
+      "grad_norm": 0.07934263348579407,
+      "learning_rate": 0.00021131921824104233,
+      "loss": 0.8462,
+      "step": 501
+    },
+    {
+      "epoch": 1.1026908292147173,
+      "grad_norm": 0.07908351719379425,
+      "learning_rate": 0.00021107491856677522,
+      "loss": 0.6308,
+      "step": 502
+    },
+    {
+      "epoch": 1.1048874244920373,
+      "grad_norm": 0.12000582367181778,
+      "learning_rate": 0.0002108306188925081,
+      "loss": 0.7749,
+      "step": 503
+    },
+    {
+      "epoch": 1.1070840197693574,
+      "grad_norm": 0.08000916242599487,
+      "learning_rate": 0.00021058631921824103,
+      "loss": 0.645,
+      "step": 504
+    },
+    {
+      "epoch": 1.1092806150466776,
+      "grad_norm": 0.09414348751306534,
+      "learning_rate": 0.00021034201954397392,
+      "loss": 0.934,
+      "step": 505
+    },
+    {
+      "epoch": 1.1114772103239978,
+      "grad_norm": 0.08176867663860321,
+      "learning_rate": 0.0002100977198697068,
+      "loss": 0.7556,
+      "step": 506
+    },
+    {
+      "epoch": 1.113673805601318,
+      "grad_norm": 0.08055311441421509,
+      "learning_rate": 0.00020985342019543972,
+      "loss": 0.7511,
+      "step": 507
+    },
+    {
+      "epoch": 1.1158704008786382,
+      "grad_norm": 0.1098596453666687,
+      "learning_rate": 0.0002096091205211726,
+      "loss": 0.8387,
+      "step": 508
+    },
+    {
+      "epoch": 1.1180669961559582,
+      "grad_norm": 0.09436143189668655,
+      "learning_rate": 0.0002093648208469055,
+      "loss": 0.7534,
+      "step": 509
+    },
+    {
+      "epoch": 1.1202635914332784,
+      "grad_norm": 0.0898435041308403,
+      "learning_rate": 0.00020912052117263842,
+      "loss": 0.8341,
+      "step": 510
+    },
+    {
+      "epoch": 1.1224601867105986,
+      "grad_norm": 0.09144418686628342,
+      "learning_rate": 0.0002088762214983713,
+      "loss": 0.7164,
+      "step": 511
+    },
+    {
+      "epoch": 1.1246567819879187,
+      "grad_norm": 0.09670098125934601,
+      "learning_rate": 0.00020863192182410422,
+      "loss": 0.7507,
+      "step": 512
+    },
+    {
+      "epoch": 1.126853377265239,
+      "grad_norm": 0.1280570924282074,
+      "learning_rate": 0.0002083876221498371,
+      "loss": 0.9009,
+      "step": 513
+    },
+    {
+      "epoch": 1.1290499725425591,
+      "grad_norm": 0.09358836710453033,
+      "learning_rate": 0.00020814332247557,
+      "loss": 0.7357,
+      "step": 514
+    },
+    {
+      "epoch": 1.131246567819879,
+      "grad_norm": 0.10839324444532394,
+      "learning_rate": 0.00020789902280130292,
+      "loss": 0.7149,
+      "step": 515
+    },
+    {
+      "epoch": 1.1334431630971993,
+      "grad_norm": 0.13648030161857605,
+      "learning_rate": 0.0002076547231270358,
+      "loss": 0.7784,
+      "step": 516
+    },
+    {
+      "epoch": 1.1356397583745195,
+      "grad_norm": 0.11029171198606491,
+      "learning_rate": 0.0002074104234527687,
+      "loss": 0.7888,
+      "step": 517
+    },
+    {
+      "epoch": 1.1378363536518397,
+      "grad_norm": 0.10812164098024368,
+      "learning_rate": 0.0002071661237785016,
+      "loss": 0.6103,
+      "step": 518
+    },
+    {
+      "epoch": 1.1400329489291599,
+      "grad_norm": 0.07254088670015335,
+      "learning_rate": 0.0002069218241042345,
+      "loss": 0.5106,
+      "step": 519
+    },
+    {
+      "epoch": 1.14222954420648,
+      "grad_norm": 0.10908333212137222,
+      "learning_rate": 0.0002066775244299674,
+      "loss": 0.7937,
+      "step": 520
+    },
+    {
+      "epoch": 1.1444261394838002,
+      "grad_norm": 0.07164821773767471,
+      "learning_rate": 0.0002064332247557003,
+      "loss": 0.6897,
+      "step": 521
+    },
+    {
+      "epoch": 1.1466227347611202,
+      "grad_norm": 0.09308282285928726,
+      "learning_rate": 0.0002061889250814332,
+      "loss": 0.6949,
+      "step": 522
+    },
+    {
+      "epoch": 1.1488193300384404,
+      "grad_norm": 0.07656940072774887,
+      "learning_rate": 0.0002059446254071661,
+      "loss": 0.5819,
+      "step": 523
+    },
+    {
+      "epoch": 1.1510159253157606,
+      "grad_norm": 0.08686637133359909,
+      "learning_rate": 0.000205700325732899,
+      "loss": 0.8069,
+      "step": 524
+    },
+    {
+      "epoch": 1.1532125205930808,
+      "grad_norm": 0.10951102524995804,
+      "learning_rate": 0.0002054560260586319,
+      "loss": 0.7268,
+      "step": 525
+    },
+    {
+      "epoch": 1.155409115870401,
+      "grad_norm": 0.11923227459192276,
+      "learning_rate": 0.0002052117263843648,
+      "loss": 0.9337,
+      "step": 526
+    },
+    {
+      "epoch": 1.157605711147721,
+      "grad_norm": 0.08200602978467941,
+      "learning_rate": 0.0002049674267100977,
+      "loss": 0.6573,
+      "step": 527
+    },
+    {
+      "epoch": 1.1598023064250411,
+      "grad_norm": 0.09406640380620956,
+      "learning_rate": 0.00020472312703583058,
+      "loss": 0.8256,
+      "step": 528
+    },
+    {
+      "epoch": 1.1619989017023613,
+      "grad_norm": 0.08187665045261383,
+      "learning_rate": 0.0002044788273615635,
+      "loss": 0.6452,
+      "step": 529
+    },
+    {
+      "epoch": 1.1641954969796815,
+      "grad_norm": 0.11410734802484512,
+      "learning_rate": 0.0002042345276872964,
+      "loss": 0.5283,
+      "step": 530
+    },
+    {
+      "epoch": 1.1663920922570017,
+      "grad_norm": 0.1198214516043663,
+      "learning_rate": 0.00020399022801302928,
+      "loss": 0.6738,
+      "step": 531
+    },
+    {
+      "epoch": 1.1685886875343219,
+      "grad_norm": 0.10346844792366028,
+      "learning_rate": 0.0002037459283387622,
+      "loss": 0.8139,
+      "step": 532
+    },
+    {
+      "epoch": 1.170785282811642,
+      "grad_norm": 0.11368418484926224,
+      "learning_rate": 0.00020350162866449508,
+      "loss": 0.8118,
+      "step": 533
+    },
+    {
+      "epoch": 1.172981878088962,
+      "grad_norm": 0.1324061155319214,
+      "learning_rate": 0.00020325732899022797,
+      "loss": 0.8095,
+      "step": 534
+    },
+    {
+      "epoch": 1.1751784733662822,
+      "grad_norm": 0.09915420413017273,
+      "learning_rate": 0.0002030130293159609,
+      "loss": 0.6938,
+      "step": 535
+    },
+    {
+      "epoch": 1.1773750686436024,
+      "grad_norm": 0.082662433385849,
+      "learning_rate": 0.00020276872964169378,
+      "loss": 0.6748,
+      "step": 536
+    },
+    {
+      "epoch": 1.1795716639209226,
+      "grad_norm": 0.07669646292924881,
+      "learning_rate": 0.0002025244299674267,
+      "loss": 0.6725,
+      "step": 537
+    },
+    {
+      "epoch": 1.1817682591982428,
+      "grad_norm": 0.10580630600452423,
+      "learning_rate": 0.00020228013029315958,
+      "loss": 0.642,
+      "step": 538
+    },
+    {
+      "epoch": 1.1839648544755628,
+      "grad_norm": 0.14615952968597412,
+      "learning_rate": 0.00020203583061889247,
+      "loss": 0.9657,
+      "step": 539
+    },
+    {
+      "epoch": 1.186161449752883,
+      "grad_norm": 0.11909019947052002,
+      "learning_rate": 0.0002017915309446254,
+      "loss": 0.67,
+      "step": 540
+    },
+    {
+      "epoch": 1.1883580450302031,
+      "grad_norm": 0.07478068023920059,
+      "learning_rate": 0.00020154723127035828,
+      "loss": 0.7181,
+      "step": 541
+    },
+    {
+      "epoch": 1.1905546403075233,
+      "grad_norm": 0.08618001639842987,
+      "learning_rate": 0.00020130293159609117,
+      "loss": 0.6361,
+      "step": 542
+    },
+    {
+      "epoch": 1.1927512355848435,
+      "grad_norm": 0.07554775476455688,
+      "learning_rate": 0.00020105863192182408,
+      "loss": 0.7775,
+      "step": 543
+    },
+    {
+      "epoch": 1.1949478308621637,
+      "grad_norm": 0.09046067297458649,
+      "learning_rate": 0.00020081433224755697,
+      "loss": 0.7257,
+      "step": 544
+    },
+    {
+      "epoch": 1.197144426139484,
+      "grad_norm": 0.08012351393699646,
+      "learning_rate": 0.00020057003257328986,
+      "loss": 0.749,
+      "step": 545
+    },
+    {
+      "epoch": 1.1993410214168039,
+      "grad_norm": 0.07655482739210129,
+      "learning_rate": 0.00020032573289902278,
+      "loss": 0.7419,
+      "step": 546
+    },
+    {
+      "epoch": 1.201537616694124,
+      "grad_norm": 0.12186406552791595,
+      "learning_rate": 0.00020008143322475567,
+      "loss": 1.0671,
+      "step": 547
+    },
+    {
+      "epoch": 1.2037342119714443,
+      "grad_norm": 0.18736830353736877,
+      "learning_rate": 0.00019983713355048856,
+      "loss": 0.7201,
+      "step": 548
+    },
+    {
+      "epoch": 1.2059308072487644,
+      "grad_norm": 0.2233382612466812,
+      "learning_rate": 0.00019959283387622147,
+      "loss": 0.9729,
+      "step": 549
+    },
+    {
+      "epoch": 1.2081274025260846,
+      "grad_norm": 0.11753692477941513,
+      "learning_rate": 0.00019934853420195436,
+      "loss": 0.814,
+      "step": 550
+    },
+    {
+      "epoch": 1.2103239978034046,
+      "grad_norm": 0.08789008110761642,
+      "learning_rate": 0.00019910423452768728,
+      "loss": 0.6891,
+      "step": 551
+    },
+    {
+      "epoch": 1.2125205930807248,
+      "grad_norm": 0.12267331033945084,
+      "learning_rate": 0.00019885993485342017,
+      "loss": 0.8415,
+      "step": 552
+    },
+    {
+      "epoch": 1.214717188358045,
+      "grad_norm": 0.0826958492398262,
+      "learning_rate": 0.00019861563517915306,
+      "loss": 0.802,
+      "step": 553
+    },
+    {
+      "epoch": 1.2169137836353652,
+      "grad_norm": 0.09174685180187225,
+      "learning_rate": 0.00019837133550488597,
+      "loss": 0.7568,
+      "step": 554
+    },
+    {
+      "epoch": 1.2191103789126854,
+      "grad_norm": 0.14253146946430206,
+      "learning_rate": 0.00019812703583061886,
+      "loss": 0.8523,
+      "step": 555
+    },
+    {
+      "epoch": 1.2213069741900056,
+      "grad_norm": 0.12727683782577515,
+      "learning_rate": 0.00019788273615635175,
+      "loss": 0.5839,
+      "step": 556
+    },
+    {
+      "epoch": 1.2235035694673257,
+      "grad_norm": 0.09347337484359741,
+      "learning_rate": 0.00019763843648208467,
+      "loss": 0.7573,
+      "step": 557
+    },
+    {
+      "epoch": 1.2257001647446457,
+      "grad_norm": 0.09123852849006653,
+      "learning_rate": 0.00019739413680781756,
+      "loss": 0.688,
+      "step": 558
+    },
+    {
+      "epoch": 1.227896760021966,
+      "grad_norm": 0.17540232837200165,
+      "learning_rate": 0.00019714983713355045,
+      "loss": 1.0125,
+      "step": 559
+    },
+    {
+      "epoch": 1.230093355299286,
+      "grad_norm": 0.07990846782922745,
+      "learning_rate": 0.00019690553745928336,
+      "loss": 0.8395,
+      "step": 560
+    },
+    {
+      "epoch": 1.2322899505766063,
+      "grad_norm": 0.10395050793886185,
+      "learning_rate": 0.00019666123778501625,
+      "loss": 0.7199,
+      "step": 561
+    },
+    {
+      "epoch": 1.2344865458539265,
+      "grad_norm": 0.10653544962406158,
+      "learning_rate": 0.00019641693811074917,
+      "loss": 0.7153,
+      "step": 562
+    },
+    {
+      "epoch": 1.2366831411312464,
+      "grad_norm": 0.1086992546916008,
+      "learning_rate": 0.00019617263843648206,
+      "loss": 0.8854,
+      "step": 563
+    },
+    {
+      "epoch": 1.2388797364085666,
+      "grad_norm": 0.09712345898151398,
+      "learning_rate": 0.00019592833876221495,
+      "loss": 0.6908,
+      "step": 564
+    },
+    {
+      "epoch": 1.2410763316858868,
+      "grad_norm": 0.08922946453094482,
+      "learning_rate": 0.00019568403908794786,
+      "loss": 0.8062,
+      "step": 565
+    },
+    {
+      "epoch": 1.243272926963207,
+      "grad_norm": 0.12401293218135834,
+      "learning_rate": 0.00019543973941368075,
+      "loss": 0.7281,
+      "step": 566
+    },
+    {
+      "epoch": 1.2454695222405272,
+      "grad_norm": 0.08772603422403336,
+      "learning_rate": 0.00019519543973941364,
+      "loss": 0.7321,
+      "step": 567
+    },
+    {
+      "epoch": 1.2476661175178474,
+      "grad_norm": 0.09601080417633057,
+      "learning_rate": 0.00019495114006514656,
+      "loss": 0.7767,
+      "step": 568
+    },
+    {
+      "epoch": 1.2498627127951676,
+      "grad_norm": 0.1105591282248497,
+      "learning_rate": 0.00019470684039087945,
+      "loss": 0.6292,
+      "step": 569
+    },
+    {
+      "epoch": 1.2520593080724876,
+      "grad_norm": 0.10841323435306549,
+      "learning_rate": 0.00019446254071661234,
+      "loss": 0.5114,
+      "step": 570
+    },
+    {
+      "epoch": 1.2542559033498077,
+      "grad_norm": 0.0920185074210167,
+      "learning_rate": 0.00019421824104234525,
+      "loss": 0.8534,
+      "step": 571
+    },
+    {
+      "epoch": 1.256452498627128,
+      "grad_norm": 0.09660661220550537,
+      "learning_rate": 0.00019397394136807814,
+      "loss": 0.7696,
+      "step": 572
+    },
+    {
+      "epoch": 1.2586490939044481,
+      "grad_norm": 0.13920946419239044,
+      "learning_rate": 0.00019372964169381103,
+      "loss": 0.7683,
+      "step": 573
+    },
+    {
+      "epoch": 1.2608456891817683,
+      "grad_norm": 0.10734532028436661,
+      "learning_rate": 0.00019348534201954395,
+      "loss": 0.8063,
+      "step": 574
+    },
+    {
+      "epoch": 1.2630422844590883,
+      "grad_norm": 0.07169698923826218,
+      "learning_rate": 0.00019324104234527684,
+      "loss": 0.8048,
+      "step": 575
+    },
+    {
+      "epoch": 1.2652388797364087,
+      "grad_norm": 0.09060622751712799,
+      "learning_rate": 0.00019299674267100975,
+      "loss": 0.6837,
+      "step": 576
+    },
+    {
+      "epoch": 1.2674354750137287,
+      "grad_norm": 0.1037139892578125,
+      "learning_rate": 0.00019275244299674264,
+      "loss": 0.951,
+      "step": 577
+    },
+    {
+      "epoch": 1.2696320702910489,
+      "grad_norm": 0.1443866491317749,
+      "learning_rate": 0.00019250814332247553,
+      "loss": 0.5558,
+      "step": 578
+    },
+    {
+      "epoch": 1.271828665568369,
+      "grad_norm": 0.11959812045097351,
+      "learning_rate": 0.00019226384364820845,
+      "loss": 0.563,
+      "step": 579
+    },
+    {
+      "epoch": 1.2740252608456892,
+      "grad_norm": 0.06265049427747726,
+      "learning_rate": 0.00019201954397394134,
+      "loss": 0.582,
+      "step": 580
+    },
+    {
+      "epoch": 1.2762218561230094,
+      "grad_norm": 0.08764317631721497,
+      "learning_rate": 0.00019177524429967423,
+      "loss": 0.7644,
+      "step": 581
+    },
+    {
+      "epoch": 1.2784184514003294,
+      "grad_norm": 0.100165955722332,
+      "learning_rate": 0.00019153094462540714,
+      "loss": 0.6459,
+      "step": 582
+    },
+    {
+      "epoch": 1.2806150466776496,
+      "grad_norm": 0.12988172471523285,
+      "learning_rate": 0.00019128664495114003,
+      "loss": 0.6632,
+      "step": 583
+    },
+    {
+      "epoch": 1.2828116419549698,
+      "grad_norm": 0.17034487426280975,
+      "learning_rate": 0.00019104234527687292,
+      "loss": 0.707,
+      "step": 584
+    },
+    {
+      "epoch": 1.28500823723229,
+      "grad_norm": 0.12978188693523407,
+      "learning_rate": 0.00019079804560260584,
+      "loss": 0.8236,
+      "step": 585
+    },
+    {
+      "epoch": 1.2872048325096102,
+      "grad_norm": 0.07136764377355576,
+      "learning_rate": 0.00019055374592833873,
+      "loss": 0.7143,
+      "step": 586
+    },
+    {
+      "epoch": 1.2894014277869301,
+      "grad_norm": 0.08375712484121323,
+      "learning_rate": 0.00019030944625407164,
+      "loss": 0.7322,
+      "step": 587
+    },
+    {
+      "epoch": 1.2915980230642505,
+      "grad_norm": 0.08253293484449387,
+      "learning_rate": 0.00019006514657980453,
+      "loss": 0.7547,
+      "step": 588
+    },
+    {
+      "epoch": 1.2937946183415705,
+      "grad_norm": 0.0760529488325119,
+      "learning_rate": 0.00018982084690553742,
+      "loss": 0.6998,
+      "step": 589
+    },
+    {
+      "epoch": 1.2959912136188907,
+      "grad_norm": 0.1125057116150856,
+      "learning_rate": 0.00018957654723127034,
+      "loss": 0.7554,
+      "step": 590
+    },
+    {
+      "epoch": 1.2981878088962109,
+      "grad_norm": 0.07319024205207825,
+      "learning_rate": 0.00018933224755700323,
+      "loss": 0.547,
+      "step": 591
+    },
+    {
+      "epoch": 1.300384404173531,
+      "grad_norm": 0.08325421065092087,
+      "learning_rate": 0.00018908794788273612,
+      "loss": 0.723,
+      "step": 592
+    },
+    {
+      "epoch": 1.3025809994508513,
+      "grad_norm": 0.07649052143096924,
+      "learning_rate": 0.00018884364820846903,
+      "loss": 0.6448,
+      "step": 593
+    },
+    {
+      "epoch": 1.3047775947281712,
+      "grad_norm": 0.1915763020515442,
+      "learning_rate": 0.00018859934853420192,
+      "loss": 0.7429,
+      "step": 594
+    },
+    {
+      "epoch": 1.3069741900054914,
+      "grad_norm": 0.11058753728866577,
+      "learning_rate": 0.0001883550488599348,
+      "loss": 0.8478,
+      "step": 595
+    },
+    {
+      "epoch": 1.3091707852828116,
+      "grad_norm": 0.08878122270107269,
+      "learning_rate": 0.00018811074918566773,
+      "loss": 0.6502,
+      "step": 596
+    },
+    {
+      "epoch": 1.3113673805601318,
+      "grad_norm": 0.11363759636878967,
+      "learning_rate": 0.00018786644951140062,
+      "loss": 0.7106,
+      "step": 597
+    },
+    {
+      "epoch": 1.313563975837452,
+      "grad_norm": 0.07276875525712967,
+      "learning_rate": 0.0001876221498371335,
+      "loss": 0.6222,
+      "step": 598
+    },
+    {
+      "epoch": 1.3157605711147722,
+      "grad_norm": 0.07559020817279816,
+      "learning_rate": 0.00018737785016286645,
+      "loss": 0.6906,
+      "step": 599
+    },
+    {
+      "epoch": 1.3179571663920924,
+      "grad_norm": 0.07954646646976471,
+      "learning_rate": 0.00018713355048859934,
+      "loss": 0.8948,
+      "step": 600
+    },
+    {
+      "epoch": 1.3201537616694123,
+      "grad_norm": 0.1203920915722847,
+      "learning_rate": 0.00018688925081433225,
+      "loss": 0.7428,
+      "step": 601
+    },
+    {
+      "epoch": 1.3223503569467325,
+      "grad_norm": 0.0738171711564064,
+      "learning_rate": 0.00018664495114006514,
+      "loss": 0.7197,
+      "step": 602
+    },
+    {
+      "epoch": 1.3245469522240527,
+      "grad_norm": 0.08350927382707596,
+      "learning_rate": 0.00018640065146579803,
+      "loss": 0.8155,
+      "step": 603
+    },
+    {
+      "epoch": 1.326743547501373,
+      "grad_norm": 0.08537956327199936,
+      "learning_rate": 0.00018615635179153095,
+      "loss": 0.7541,
+      "step": 604
+    },
+    {
+      "epoch": 1.328940142778693,
+      "grad_norm": 0.08641252666711807,
+      "learning_rate": 0.00018591205211726384,
+      "loss": 0.8637,
+      "step": 605
+    },
+    {
+      "epoch": 1.331136738056013,
+      "grad_norm": 0.09421125799417496,
+      "learning_rate": 0.00018566775244299675,
+      "loss": 0.6457,
+      "step": 606
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 0.08459457010030746,
+      "learning_rate": 0.00018542345276872964,
+      "loss": 0.5529,
+      "step": 607
+    },
+    {
+      "epoch": 1.3355299286106534,
+      "grad_norm": 0.08040441572666168,
+      "learning_rate": 0.00018517915309446253,
+      "loss": 0.7007,
+      "step": 608
+    },
+    {
+      "epoch": 1.3377265238879736,
+      "grad_norm": 0.09265803545713425,
+      "learning_rate": 0.00018493485342019545,
+      "loss": 0.7615,
+      "step": 609
+    },
+    {
+      "epoch": 1.3399231191652938,
+      "grad_norm": 0.09262501448392868,
+      "learning_rate": 0.00018469055374592834,
+      "loss": 0.7986,
+      "step": 610
+    },
+    {
+      "epoch": 1.342119714442614,
+      "grad_norm": 0.07920756191015244,
+      "learning_rate": 0.00018444625407166123,
+      "loss": 0.7836,
+      "step": 611
+    },
+    {
+      "epoch": 1.3443163097199342,
+      "grad_norm": 0.07290450483560562,
+      "learning_rate": 0.00018420195439739414,
+      "loss": 0.6704,
+      "step": 612
+    },
+    {
+      "epoch": 1.3465129049972542,
+      "grad_norm": 0.4022713303565979,
+      "learning_rate": 0.00018395765472312703,
+      "loss": 0.792,
+      "step": 613
+    },
+    {
+      "epoch": 1.3487095002745744,
+      "grad_norm": 0.10242049396038055,
+      "learning_rate": 0.00018371335504885992,
+      "loss": 0.6693,
+      "step": 614
+    },
+    {
+      "epoch": 1.3509060955518946,
+      "grad_norm": 0.09896315634250641,
+      "learning_rate": 0.00018346905537459284,
+      "loss": 0.848,
+      "step": 615
+    },
+    {
+      "epoch": 1.3531026908292147,
+      "grad_norm": 0.11688041687011719,
+      "learning_rate": 0.00018322475570032573,
+      "loss": 0.8191,
+      "step": 616
+    },
+    {
+      "epoch": 1.355299286106535,
+      "grad_norm": 0.10265535116195679,
+      "learning_rate": 0.00018298045602605864,
+      "loss": 0.7496,
+      "step": 617
+    },
+    {
+      "epoch": 1.357495881383855,
+      "grad_norm": 0.10426696389913559,
+      "learning_rate": 0.00018273615635179153,
+      "loss": 0.976,
+      "step": 618
+    },
+    {
+      "epoch": 1.359692476661175,
+      "grad_norm": 0.0805555060505867,
+      "learning_rate": 0.00018249185667752442,
+      "loss": 0.5824,
+      "step": 619
+    },
+    {
+      "epoch": 1.3618890719384953,
+      "grad_norm": 0.08423875272274017,
+      "learning_rate": 0.00018224755700325734,
+      "loss": 0.6644,
+      "step": 620
+    },
+    {
+      "epoch": 1.3640856672158155,
+      "grad_norm": 0.09075197577476501,
+      "learning_rate": 0.00018200325732899023,
+      "loss": 0.8964,
+      "step": 621
+    },
+    {
+      "epoch": 1.3662822624931357,
+      "grad_norm": 0.08253109455108643,
+      "learning_rate": 0.00018175895765472312,
+      "loss": 0.707,
+      "step": 622
+    },
+    {
+      "epoch": 1.3684788577704559,
+      "grad_norm": 0.10257839411497116,
+      "learning_rate": 0.00018151465798045603,
+      "loss": 0.5983,
+      "step": 623
+    },
+    {
+      "epoch": 1.370675453047776,
+      "grad_norm": 0.13562428951263428,
+      "learning_rate": 0.00018127035830618892,
+      "loss": 0.7565,
+      "step": 624
+    },
+    {
+      "epoch": 1.372872048325096,
+      "grad_norm": 0.09330661594867706,
+      "learning_rate": 0.0001810260586319218,
+      "loss": 0.7985,
+      "step": 625
+    },
+    {
+      "epoch": 1.3750686436024162,
+      "grad_norm": 0.08755582571029663,
+      "learning_rate": 0.00018078175895765473,
+      "loss": 0.6307,
+      "step": 626
+    },
+    {
+      "epoch": 1.3772652388797364,
+      "grad_norm": 0.07036561518907547,
+      "learning_rate": 0.00018053745928338762,
+      "loss": 0.6643,
+      "step": 627
+    },
+    {
+      "epoch": 1.3794618341570566,
+      "grad_norm": 0.13484588265419006,
+      "learning_rate": 0.0001802931596091205,
+      "loss": 0.5747,
+      "step": 628
+    },
+    {
+      "epoch": 1.3816584294343768,
+      "grad_norm": 0.1814517229795456,
+      "learning_rate": 0.00018004885993485342,
+      "loss": 0.717,
+      "step": 629
+    },
+    {
+      "epoch": 1.3838550247116967,
+      "grad_norm": 0.28708839416503906,
+      "learning_rate": 0.0001798045602605863,
+      "loss": 0.709,
+      "step": 630
+    },
+    {
+      "epoch": 1.3860516199890172,
+      "grad_norm": 0.09198810905218124,
+      "learning_rate": 0.00017956026058631923,
+      "loss": 0.693,
+      "step": 631
+    },
+    {
+      "epoch": 1.3882482152663371,
+      "grad_norm": 0.09249893575906754,
+      "learning_rate": 0.00017931596091205212,
+      "loss": 0.8568,
+      "step": 632
+    },
+    {
+      "epoch": 1.3904448105436573,
+      "grad_norm": 0.0689750537276268,
+      "learning_rate": 0.000179071661237785,
+      "loss": 0.7537,
+      "step": 633
+    },
+    {
+      "epoch": 1.3926414058209775,
+      "grad_norm": 0.07518107444047928,
+      "learning_rate": 0.00017882736156351792,
+      "loss": 0.6492,
+      "step": 634
+    },
+    {
+      "epoch": 1.3948380010982977,
+      "grad_norm": 0.07328426837921143,
+      "learning_rate": 0.0001785830618892508,
+      "loss": 0.7392,
+      "step": 635
+    },
+    {
+      "epoch": 1.3970345963756179,
+      "grad_norm": 0.09382381290197372,
+      "learning_rate": 0.0001783387622149837,
+      "loss": 0.8653,
+      "step": 636
+    },
+    {
+      "epoch": 1.3992311916529379,
+      "grad_norm": 0.10559618473052979,
+      "learning_rate": 0.00017809446254071662,
+      "loss": 0.6867,
+      "step": 637
+    },
+    {
+      "epoch": 1.401427786930258,
+      "grad_norm": 0.08409029990434647,
+      "learning_rate": 0.0001778501628664495,
+      "loss": 0.6649,
+      "step": 638
+    },
+    {
+      "epoch": 1.4036243822075782,
+      "grad_norm": 0.14046898484230042,
+      "learning_rate": 0.0001776058631921824,
+      "loss": 0.6858,
+      "step": 639
+    },
+    {
+      "epoch": 1.4058209774848984,
+      "grad_norm": 0.10538306832313538,
+      "learning_rate": 0.0001773615635179153,
+      "loss": 0.9439,
+      "step": 640
+    },
+    {
+      "epoch": 1.4080175727622186,
+      "grad_norm": 0.08711986243724823,
+      "learning_rate": 0.0001771172638436482,
+      "loss": 0.7046,
+      "step": 641
+    },
+    {
+      "epoch": 1.4102141680395386,
+      "grad_norm": 0.9647109508514404,
+      "learning_rate": 0.0001768729641693811,
+      "loss": 0.8507,
+      "step": 642
+    },
+    {
+      "epoch": 1.412410763316859,
+      "grad_norm": 0.20237646996974945,
+      "learning_rate": 0.000176628664495114,
+      "loss": 0.6649,
+      "step": 643
+    },
+    {
+      "epoch": 1.414607358594179,
+      "grad_norm": 0.15469181537628174,
+      "learning_rate": 0.0001763843648208469,
+      "loss": 0.567,
+      "step": 644
+    },
+    {
+      "epoch": 1.4168039538714992,
+      "grad_norm": 0.10986842960119247,
+      "learning_rate": 0.0001761400651465798,
+      "loss": 0.653,
+      "step": 645
+    },
+    {
+      "epoch": 1.4190005491488193,
+      "grad_norm": 0.07824039459228516,
+      "learning_rate": 0.0001758957654723127,
+      "loss": 0.6444,
+      "step": 646
+    },
+    {
+      "epoch": 1.4211971444261395,
+      "grad_norm": 0.07870502024888992,
+      "learning_rate": 0.0001756514657980456,
+      "loss": 0.7054,
+      "step": 647
+    },
+    {
+      "epoch": 1.4233937397034597,
+      "grad_norm": 0.14770908653736115,
+      "learning_rate": 0.0001754071661237785,
+      "loss": 0.6814,
+      "step": 648
+    },
+    {
+      "epoch": 1.4255903349807797,
+      "grad_norm": 0.10939688980579376,
+      "learning_rate": 0.0001751628664495114,
+      "loss": 0.7833,
+      "step": 649
+    },
+    {
+      "epoch": 1.4277869302580999,
+      "grad_norm": 0.07083068042993546,
+      "learning_rate": 0.00017491856677524429,
+      "loss": 0.6025,
+      "step": 650
+    },
+    {
+      "epoch": 1.42998352553542,
+      "grad_norm": 0.08645574748516083,
+      "learning_rate": 0.0001746742671009772,
+      "loss": 0.6746,
+      "step": 651
+    },
+    {
+      "epoch": 1.4321801208127403,
+      "grad_norm": 0.07458820939064026,
+      "learning_rate": 0.0001744299674267101,
+      "loss": 0.6725,
+      "step": 652
+    },
+    {
+      "epoch": 1.4343767160900605,
+      "grad_norm": 0.09103960543870926,
+      "learning_rate": 0.00017418566775244298,
+      "loss": 0.727,
+      "step": 653
+    },
+    {
+      "epoch": 1.4365733113673804,
+      "grad_norm": 0.07630818337202072,
+      "learning_rate": 0.0001739413680781759,
+      "loss": 0.6626,
+      "step": 654
+    },
+    {
+      "epoch": 1.4387699066447008,
+      "grad_norm": 0.07506604492664337,
+      "learning_rate": 0.00017369706840390879,
+      "loss": 0.713,
+      "step": 655
+    },
+    {
+      "epoch": 1.4409665019220208,
+      "grad_norm": 0.08960114419460297,
+      "learning_rate": 0.0001734527687296417,
+      "loss": 0.7648,
+      "step": 656
+    },
+    {
+      "epoch": 1.443163097199341,
+      "grad_norm": 0.0818580612540245,
+      "learning_rate": 0.0001732084690553746,
+      "loss": 0.8596,
+      "step": 657
+    },
+    {
+      "epoch": 1.4453596924766612,
+      "grad_norm": 0.08464060723781586,
+      "learning_rate": 0.00017296416938110748,
+      "loss": 0.8378,
+      "step": 658
+    },
+    {
+      "epoch": 1.4475562877539814,
+      "grad_norm": 0.21754679083824158,
+      "learning_rate": 0.0001727198697068404,
+      "loss": 0.8394,
+      "step": 659
+    },
+    {
+      "epoch": 1.4497528830313016,
+      "grad_norm": 0.07114458084106445,
+      "learning_rate": 0.00017247557003257329,
+      "loss": 0.648,
+      "step": 660
+    },
+    {
+      "epoch": 1.4519494783086215,
+      "grad_norm": 0.08992248773574829,
+      "learning_rate": 0.00017223127035830618,
+      "loss": 0.6419,
+      "step": 661
+    },
+    {
+      "epoch": 1.4541460735859417,
+      "grad_norm": 0.21482007205486298,
+      "learning_rate": 0.0001719869706840391,
+      "loss": 0.7515,
+      "step": 662
+    },
+    {
+      "epoch": 1.456342668863262,
+      "grad_norm": 0.07746273279190063,
+      "learning_rate": 0.00017174267100977198,
+      "loss": 0.7467,
+      "step": 663
+    },
+    {
+      "epoch": 1.458539264140582,
+      "grad_norm": 0.10815945267677307,
+      "learning_rate": 0.00017149837133550487,
+      "loss": 0.6446,
+      "step": 664
+    },
+    {
+      "epoch": 1.4607358594179023,
+      "grad_norm": 0.08945509046316147,
+      "learning_rate": 0.00017125407166123779,
+      "loss": 0.8931,
+      "step": 665
+    },
+    {
+      "epoch": 1.4629324546952225,
+      "grad_norm": 0.11326180398464203,
+      "learning_rate": 0.00017100977198697068,
+      "loss": 0.9186,
+      "step": 666
+    },
+    {
+      "epoch": 1.4651290499725427,
+      "grad_norm": 0.07652183622121811,
+      "learning_rate": 0.00017076547231270356,
+      "loss": 0.6587,
+      "step": 667
+    },
+    {
+      "epoch": 1.4673256452498626,
+      "grad_norm": 0.06988921016454697,
+      "learning_rate": 0.00017052117263843648,
+      "loss": 0.8556,
+      "step": 668
+    },
+    {
+      "epoch": 1.4695222405271828,
+      "grad_norm": 0.09760496765375137,
+      "learning_rate": 0.00017027687296416937,
+      "loss": 0.7058,
+      "step": 669
+    },
+    {
+      "epoch": 1.471718835804503,
+      "grad_norm": 0.0829065665602684,
+      "learning_rate": 0.00017003257328990229,
+      "loss": 0.7083,
+      "step": 670
+    },
+    {
+      "epoch": 1.4739154310818232,
+      "grad_norm": 0.11744392663240433,
+      "learning_rate": 0.00016978827361563518,
+      "loss": 0.8903,
+      "step": 671
+    },
+    {
+      "epoch": 1.4761120263591434,
+      "grad_norm": 0.09920722246170044,
+      "learning_rate": 0.00016954397394136806,
+      "loss": 0.7584,
+      "step": 672
+    },
+    {
+      "epoch": 1.4783086216364634,
+      "grad_norm": 0.11864949017763138,
+      "learning_rate": 0.00016929967426710098,
+      "loss": 0.6979,
+      "step": 673
+    },
+    {
+      "epoch": 1.4805052169137836,
+      "grad_norm": 0.07455518841743469,
+      "learning_rate": 0.00016905537459283387,
+      "loss": 0.7666,
+      "step": 674
+    },
+    {
+      "epoch": 1.4827018121911038,
+      "grad_norm": 0.09010827541351318,
+      "learning_rate": 0.00016881107491856676,
+      "loss": 0.6141,
+      "step": 675
+    },
+    {
+      "epoch": 1.484898407468424,
+      "grad_norm": 0.07416039705276489,
+      "learning_rate": 0.00016856677524429968,
+      "loss": 0.725,
+      "step": 676
+    },
+    {
+      "epoch": 1.4870950027457441,
+      "grad_norm": 0.07443492859601974,
+      "learning_rate": 0.00016832247557003257,
+      "loss": 0.6858,
+      "step": 677
+    },
+    {
+      "epoch": 1.4892915980230643,
+      "grad_norm": 0.16061677038669586,
+      "learning_rate": 0.00016807817589576545,
+      "loss": 0.7833,
+      "step": 678
+    },
+    {
+      "epoch": 1.4914881933003845,
+      "grad_norm": 0.07762904465198517,
+      "learning_rate": 0.00016783387622149837,
+      "loss": 0.7674,
+      "step": 679
+    },
+    {
+      "epoch": 1.4936847885777045,
+      "grad_norm": 0.21594302356243134,
+      "learning_rate": 0.00016758957654723126,
+      "loss": 0.9779,
+      "step": 680
+    },
+    {
+      "epoch": 1.4958813838550247,
+      "grad_norm": 0.08476382493972778,
+      "learning_rate": 0.00016734527687296415,
+      "loss": 0.7278,
+      "step": 681
+    },
+    {
+      "epoch": 1.4980779791323449,
+      "grad_norm": 0.08337073028087616,
+      "learning_rate": 0.00016710097719869707,
+      "loss": 0.9273,
+      "step": 682
+    },
+    {
+      "epoch": 1.500274574409665,
+      "grad_norm": 0.1177864745259285,
+      "learning_rate": 0.00016685667752442995,
+      "loss": 0.5299,
+      "step": 683
+    },
+    {
+      "epoch": 1.5024711696869852,
+      "grad_norm": 0.07652102410793304,
+      "learning_rate": 0.00016661237785016287,
+      "loss": 0.6531,
+      "step": 684
+    },
+    {
+      "epoch": 1.5046677649643052,
+      "grad_norm": 0.09166745096445084,
+      "learning_rate": 0.00016636807817589576,
+      "loss": 0.8363,
+      "step": 685
+    },
+    {
+      "epoch": 1.5068643602416256,
+      "grad_norm": 0.102321095764637,
+      "learning_rate": 0.00016612377850162865,
+      "loss": 0.8824,
+      "step": 686
+    },
+    {
+      "epoch": 1.5090609555189456,
+      "grad_norm": 0.1683252602815628,
+      "learning_rate": 0.00016587947882736157,
+      "loss": 0.8136,
+      "step": 687
+    },
+    {
+      "epoch": 1.5112575507962658,
+      "grad_norm": 0.08670135587453842,
+      "learning_rate": 0.00016563517915309445,
+      "loss": 0.8029,
+      "step": 688
+    },
+    {
+      "epoch": 1.513454146073586,
+      "grad_norm": 0.07823559641838074,
+      "learning_rate": 0.00016539087947882734,
+      "loss": 0.6649,
+      "step": 689
+    },
+    {
+      "epoch": 1.515650741350906,
+      "grad_norm": 0.2524609863758087,
+      "learning_rate": 0.00016514657980456026,
+      "loss": 0.6903,
+      "step": 690
+    },
+    {
+      "epoch": 1.5178473366282264,
+      "grad_norm": 0.09681102633476257,
+      "learning_rate": 0.00016490228013029315,
+      "loss": 0.7249,
+      "step": 691
+    },
+    {
+      "epoch": 1.5200439319055463,
+      "grad_norm": 0.09815948456525803,
+      "learning_rate": 0.00016465798045602604,
+      "loss": 0.7694,
+      "step": 692
+    },
+    {
+      "epoch": 1.5222405271828665,
+      "grad_norm": 0.10188299417495728,
+      "learning_rate": 0.00016441368078175895,
+      "loss": 0.7534,
+      "step": 693
+    },
+    {
+      "epoch": 1.5244371224601867,
+      "grad_norm": 0.09090428799390793,
+      "learning_rate": 0.00016416938110749184,
+      "loss": 0.6444,
+      "step": 694
+    },
+    {
+      "epoch": 1.526633717737507,
+      "grad_norm": 0.07603637129068375,
+      "learning_rate": 0.00016392508143322476,
+      "loss": 0.7279,
+      "step": 695
+    },
+    {
+      "epoch": 1.528830313014827,
+      "grad_norm": 0.07567336410284042,
+      "learning_rate": 0.00016368078175895765,
+      "loss": 0.637,
+      "step": 696
+    },
+    {
+      "epoch": 1.531026908292147,
+      "grad_norm": 0.07270662486553192,
+      "learning_rate": 0.00016343648208469054,
+      "loss": 0.691,
+      "step": 697
+    },
+    {
+      "epoch": 1.5332235035694675,
+      "grad_norm": 0.07661709189414978,
+      "learning_rate": 0.00016319218241042346,
+      "loss": 0.654,
+      "step": 698
+    },
+    {
+      "epoch": 1.5354200988467874,
+      "grad_norm": 0.12478196620941162,
+      "learning_rate": 0.00016294788273615634,
+      "loss": 0.7383,
+      "step": 699
+    },
+    {
+      "epoch": 1.5376166941241076,
+      "grad_norm": 0.07682304084300995,
+      "learning_rate": 0.00016270358306188923,
+      "loss": 0.8559,
+      "step": 700
+    },
+    {
+      "epoch": 1.5398132894014278,
+      "grad_norm": 0.11485746502876282,
+      "learning_rate": 0.00016245928338762215,
+      "loss": 0.6817,
+      "step": 701
+    },
+    {
+      "epoch": 1.5420098846787478,
+      "grad_norm": 0.11213982850313187,
+      "learning_rate": 0.00016221498371335504,
+      "loss": 0.6974,
+      "step": 702
+    },
+    {
+      "epoch": 1.5442064799560682,
+      "grad_norm": 0.0833524689078331,
+      "learning_rate": 0.00016197068403908793,
+      "loss": 0.6304,
+      "step": 703
+    },
+    {
+      "epoch": 1.5464030752333882,
+      "grad_norm": 0.07857480645179749,
+      "learning_rate": 0.00016172638436482084,
+      "loss": 0.6384,
+      "step": 704
+    },
+    {
+      "epoch": 1.5485996705107083,
+      "grad_norm": 0.0922989547252655,
+      "learning_rate": 0.00016148208469055373,
+      "loss": 0.7245,
+      "step": 705
+    },
+    {
+      "epoch": 1.5507962657880285,
+      "grad_norm": 0.10281957685947418,
+      "learning_rate": 0.00016123778501628662,
+      "loss": 0.7421,
+      "step": 706
+    },
+    {
+      "epoch": 1.5529928610653487,
+      "grad_norm": 0.10956420004367828,
+      "learning_rate": 0.00016099348534201954,
+      "loss": 0.7539,
+      "step": 707
+    },
+    {
+      "epoch": 1.555189456342669,
+      "grad_norm": 0.14065328240394592,
+      "learning_rate": 0.00016074918566775243,
+      "loss": 0.8351,
+      "step": 708
+    },
+    {
+      "epoch": 1.5573860516199889,
+      "grad_norm": 0.0996328666806221,
+      "learning_rate": 0.00016050488599348534,
+      "loss": 0.9108,
+      "step": 709
+    },
+    {
+      "epoch": 1.5595826468973093,
+      "grad_norm": 0.11864620447158813,
+      "learning_rate": 0.00016026058631921823,
+      "loss": 0.4986,
+      "step": 710
+    },
+    {
+      "epoch": 1.5617792421746293,
+      "grad_norm": 0.08180121332406998,
+      "learning_rate": 0.00016001628664495112,
+      "loss": 0.8248,
+      "step": 711
+    },
+    {
+      "epoch": 1.5639758374519495,
+      "grad_norm": 0.08939424902200699,
+      "learning_rate": 0.00015977198697068404,
+      "loss": 0.5868,
+      "step": 712
+    },
+    {
+      "epoch": 1.5661724327292696,
+      "grad_norm": 0.09274241328239441,
+      "learning_rate": 0.00015952768729641693,
+      "loss": 0.6402,
+      "step": 713
+    },
+    {
+      "epoch": 1.5683690280065898,
+      "grad_norm": 0.058022402226924896,
+      "learning_rate": 0.00015928338762214982,
+      "loss": 0.5863,
+      "step": 714
+    },
+    {
+      "epoch": 1.57056562328391,
+      "grad_norm": 0.10023228079080582,
+      "learning_rate": 0.00015903908794788273,
+      "loss": 0.7613,
+      "step": 715
+    },
+    {
+      "epoch": 1.57276221856123,
+      "grad_norm": 0.13345441222190857,
+      "learning_rate": 0.00015879478827361562,
+      "loss": 0.948,
+      "step": 716
+    },
+    {
+      "epoch": 1.5749588138385504,
+      "grad_norm": 0.06687305122613907,
+      "learning_rate": 0.0001585504885993485,
+      "loss": 0.7116,
+      "step": 717
+    },
+    {
+      "epoch": 1.5771554091158704,
+      "grad_norm": 0.11990214884281158,
+      "learning_rate": 0.00015830618892508143,
+      "loss": 0.6393,
+      "step": 718
+    },
+    {
+      "epoch": 1.5793520043931906,
+      "grad_norm": 0.10609102994203568,
+      "learning_rate": 0.00015806188925081432,
+      "loss": 0.7194,
+      "step": 719
+    },
+    {
+      "epoch": 1.5815485996705108,
+      "grad_norm": 0.0786457285284996,
+      "learning_rate": 0.00015781758957654723,
+      "loss": 0.6428,
+      "step": 720
+    },
+    {
+      "epoch": 1.5837451949478307,
+      "grad_norm": 0.08620281517505646,
+      "learning_rate": 0.00015757328990228012,
+      "loss": 0.8658,
+      "step": 721
+    },
+    {
+      "epoch": 1.5859417902251511,
+      "grad_norm": 0.09571443498134613,
+      "learning_rate": 0.000157328990228013,
+      "loss": 0.928,
+      "step": 722
+    },
+    {
+      "epoch": 1.588138385502471,
+      "grad_norm": 0.07305927574634552,
+      "learning_rate": 0.00015708469055374593,
+      "loss": 0.7061,
+      "step": 723
+    },
+    {
+      "epoch": 1.5903349807797913,
+      "grad_norm": 0.12097378820180893,
+      "learning_rate": 0.00015684039087947882,
+      "loss": 0.5605,
+      "step": 724
+    },
+    {
+      "epoch": 1.5925315760571115,
+      "grad_norm": 0.07542566955089569,
+      "learning_rate": 0.0001565960912052117,
+      "loss": 0.768,
+      "step": 725
+    },
+    {
+      "epoch": 1.5947281713344317,
+      "grad_norm": 0.1035517156124115,
+      "learning_rate": 0.00015635179153094462,
+      "loss": 0.7219,
+      "step": 726
+    },
+    {
+      "epoch": 1.5969247666117519,
+      "grad_norm": 0.07891443371772766,
+      "learning_rate": 0.0001561074918566775,
+      "loss": 0.7857,
+      "step": 727
+    },
+    {
+      "epoch": 1.5991213618890718,
+      "grad_norm": 0.08524596691131592,
+      "learning_rate": 0.0001558631921824104,
+      "loss": 0.7636,
+      "step": 728
+    },
+    {
+      "epoch": 1.6013179571663922,
+      "grad_norm": 0.13341744244098663,
+      "learning_rate": 0.00015561889250814332,
+      "loss": 0.7329,
+      "step": 729
+    },
+    {
+      "epoch": 1.6035145524437122,
+      "grad_norm": 0.10522423684597015,
+      "learning_rate": 0.0001553745928338762,
+      "loss": 0.7602,
+      "step": 730
+    },
+    {
+      "epoch": 1.6057111477210324,
+      "grad_norm": 0.06145809590816498,
+      "learning_rate": 0.0001551302931596091,
+      "loss": 0.474,
+      "step": 731
+    },
+    {
+      "epoch": 1.6079077429983526,
+      "grad_norm": 0.08298227936029434,
+      "learning_rate": 0.000154885993485342,
+      "loss": 0.7149,
+      "step": 732
+    },
+    {
+      "epoch": 1.6101043382756726,
+      "grad_norm": 0.09817589074373245,
+      "learning_rate": 0.0001546416938110749,
+      "loss": 0.9164,
+      "step": 733
+    },
+    {
+      "epoch": 1.612300933552993,
+      "grad_norm": 0.07774459570646286,
+      "learning_rate": 0.00015439739413680782,
+      "loss": 0.7634,
+      "step": 734
+    },
+    {
+      "epoch": 1.614497528830313,
+      "grad_norm": 0.10114621371030807,
+      "learning_rate": 0.0001541530944625407,
+      "loss": 0.7938,
+      "step": 735
+    },
+    {
+      "epoch": 1.6166941241076331,
+      "grad_norm": 0.09011275321245193,
+      "learning_rate": 0.0001539087947882736,
+      "loss": 0.758,
+      "step": 736
+    },
+    {
+      "epoch": 1.6188907193849533,
+      "grad_norm": 0.08204670995473862,
+      "learning_rate": 0.0001536644951140065,
+      "loss": 0.7762,
+      "step": 737
+    },
+    {
+      "epoch": 1.6210873146622735,
+      "grad_norm": 0.12722116708755493,
+      "learning_rate": 0.0001534201954397394,
+      "loss": 0.7504,
+      "step": 738
+    },
+    {
+      "epoch": 1.6232839099395937,
+      "grad_norm": 0.08731167763471603,
+      "learning_rate": 0.0001531758957654723,
+      "loss": 0.7225,
+      "step": 739
+    },
+    {
+      "epoch": 1.6254805052169137,
+      "grad_norm": 0.086002878844738,
+      "learning_rate": 0.0001529315960912052,
+      "loss": 0.7417,
+      "step": 740
+    },
+    {
+      "epoch": 1.627677100494234,
+      "grad_norm": 0.08874348551034927,
+      "learning_rate": 0.0001526872964169381,
+      "loss": 0.7965,
+      "step": 741
+    },
+    {
+      "epoch": 1.629873695771554,
+      "grad_norm": 0.0923686996102333,
+      "learning_rate": 0.00015244299674267099,
+      "loss": 0.6964,
+      "step": 742
+    },
+    {
+      "epoch": 1.6320702910488742,
+      "grad_norm": 0.14827051758766174,
+      "learning_rate": 0.0001521986970684039,
+      "loss": 0.6615,
+      "step": 743
+    },
+    {
+      "epoch": 1.6342668863261944,
+      "grad_norm": 0.09289976209402084,
+      "learning_rate": 0.0001519543973941368,
+      "loss": 0.5839,
+      "step": 744
+    },
+    {
+      "epoch": 1.6364634816035144,
+      "grad_norm": 0.13095685839653015,
+      "learning_rate": 0.00015171009771986968,
+      "loss": 0.6602,
+      "step": 745
+    },
+    {
+      "epoch": 1.6386600768808348,
+      "grad_norm": 0.1233493909239769,
+      "learning_rate": 0.0001514657980456026,
+      "loss": 0.5198,
+      "step": 746
+    },
+    {
+      "epoch": 1.6408566721581548,
+      "grad_norm": 0.1225695013999939,
+      "learning_rate": 0.0001512214983713355,
+      "loss": 0.711,
+      "step": 747
+    },
+    {
+      "epoch": 1.643053267435475,
+      "grad_norm": 0.1031402051448822,
+      "learning_rate": 0.0001509771986970684,
+      "loss": 0.5741,
+      "step": 748
+    },
+    {
+      "epoch": 1.6452498627127952,
+      "grad_norm": 0.13425299525260925,
+      "learning_rate": 0.0001507328990228013,
+      "loss": 0.5971,
+      "step": 749
+    },
+    {
+      "epoch": 1.6474464579901154,
+      "grad_norm": 0.08660752326250076,
+      "learning_rate": 0.00015048859934853418,
+      "loss": 0.6597,
+      "step": 750
+    },
+    {
+      "epoch": 1.6496430532674355,
+      "grad_norm": 0.08104998618364334,
+      "learning_rate": 0.0001502442996742671,
+      "loss": 0.6907,
+      "step": 751
+    },
+    {
+      "epoch": 1.6518396485447555,
+      "grad_norm": 0.1011468768119812,
+      "learning_rate": 0.00015,
+      "loss": 0.8155,
+      "step": 752
+    },
+    {
+      "epoch": 1.654036243822076,
+      "grad_norm": 0.15687088668346405,
+      "learning_rate": 0.00014975570032573288,
+      "loss": 0.9908,
+      "step": 753
+    },
+    {
+      "epoch": 1.656232839099396,
+      "grad_norm": 0.10400700569152832,
+      "learning_rate": 0.0001495114006514658,
+      "loss": 0.6036,
+      "step": 754
+    },
+    {
+      "epoch": 1.658429434376716,
+      "grad_norm": 0.15937060117721558,
+      "learning_rate": 0.00014926710097719868,
+      "loss": 0.6572,
+      "step": 755
+    },
+    {
+      "epoch": 1.6606260296540363,
+      "grad_norm": 0.11421637237071991,
+      "learning_rate": 0.00014902280130293157,
+      "loss": 0.6941,
+      "step": 756
+    },
+    {
+      "epoch": 1.6628226249313562,
+      "grad_norm": 0.07699894905090332,
+      "learning_rate": 0.0001487785016286645,
+      "loss": 0.7786,
+      "step": 757
+    },
+    {
+      "epoch": 1.6650192202086767,
+      "grad_norm": 0.10346461832523346,
+      "learning_rate": 0.00014853420195439738,
+      "loss": 0.776,
+      "step": 758
+    },
+    {
+      "epoch": 1.6672158154859966,
+      "grad_norm": 0.08529362082481384,
+      "learning_rate": 0.0001482899022801303,
+      "loss": 0.8314,
+      "step": 759
+    },
+    {
+      "epoch": 1.6694124107633168,
+      "grad_norm": 0.10815656930208206,
+      "learning_rate": 0.00014804560260586318,
+      "loss": 0.7199,
+      "step": 760
+    },
+    {
+      "epoch": 1.671609006040637,
+      "grad_norm": 0.07066084444522858,
+      "learning_rate": 0.00014780130293159607,
+      "loss": 0.72,
+      "step": 761
+    },
+    {
+      "epoch": 1.6738056013179572,
+      "grad_norm": 0.09182927757501602,
+      "learning_rate": 0.000147557003257329,
+      "loss": 0.7723,
+      "step": 762
+    },
+    {
+      "epoch": 1.6760021965952774,
+      "grad_norm": 0.09432081133127213,
+      "learning_rate": 0.00014731270358306188,
+      "loss": 0.7546,
+      "step": 763
+    },
+    {
+      "epoch": 1.6781987918725974,
+      "grad_norm": 0.0820528045296669,
+      "learning_rate": 0.00014706840390879477,
+      "loss": 0.8123,
+      "step": 764
+    },
+    {
+      "epoch": 1.6803953871499178,
+      "grad_norm": 0.10416916757822037,
+      "learning_rate": 0.00014682410423452768,
+      "loss": 0.7616,
+      "step": 765
+    },
+    {
+      "epoch": 1.6825919824272377,
+      "grad_norm": 0.08981508761644363,
+      "learning_rate": 0.00014657980456026057,
+      "loss": 0.6376,
+      "step": 766
+    },
+    {
+      "epoch": 1.684788577704558,
+      "grad_norm": 0.12721554934978485,
+      "learning_rate": 0.00014633550488599346,
+      "loss": 0.7292,
+      "step": 767
+    },
+    {
+      "epoch": 1.6869851729818781,
+      "grad_norm": 0.0896964967250824,
+      "learning_rate": 0.00014609120521172638,
+      "loss": 0.6565,
+      "step": 768
+    },
+    {
+      "epoch": 1.689181768259198,
+      "grad_norm": 0.1305578500032425,
+      "learning_rate": 0.00014584690553745927,
+      "loss": 0.7175,
+      "step": 769
+    },
+    {
+      "epoch": 1.6913783635365185,
+      "grad_norm": 0.08318842947483063,
+      "learning_rate": 0.00014560260586319216,
+      "loss": 0.8087,
+      "step": 770
+    },
+    {
+      "epoch": 1.6935749588138385,
+      "grad_norm": 0.1729842573404312,
+      "learning_rate": 0.00014535830618892507,
+      "loss": 0.9475,
+      "step": 771
+    },
+    {
+      "epoch": 1.6957715540911587,
+      "grad_norm": 0.08786248415708542,
+      "learning_rate": 0.00014511400651465796,
+      "loss": 0.8516,
+      "step": 772
+    },
+    {
+      "epoch": 1.6979681493684788,
+      "grad_norm": 0.1401490867137909,
+      "learning_rate": 0.00014486970684039088,
+      "loss": 0.6591,
+      "step": 773
+    },
+    {
+      "epoch": 1.700164744645799,
+      "grad_norm": 0.06962237507104874,
+      "learning_rate": 0.00014462540716612377,
+      "loss": 0.6165,
+      "step": 774
+    },
+    {
+      "epoch": 1.7023613399231192,
+      "grad_norm": 0.13167256116867065,
+      "learning_rate": 0.00014438110749185666,
+      "loss": 0.729,
+      "step": 775
+    },
+    {
+      "epoch": 1.7045579352004392,
+      "grad_norm": 0.07879288494586945,
+      "learning_rate": 0.00014413680781758957,
+      "loss": 0.5357,
+      "step": 776
+    },
+    {
+      "epoch": 1.7067545304777596,
+      "grad_norm": 0.09663818776607513,
+      "learning_rate": 0.00014389250814332246,
+      "loss": 0.7877,
+      "step": 777
+    },
+    {
+      "epoch": 1.7089511257550796,
+      "grad_norm": 0.09674274176359177,
+      "learning_rate": 0.00014364820846905535,
+      "loss": 0.812,
+      "step": 778
+    },
+    {
+      "epoch": 1.7111477210323998,
+      "grad_norm": 0.07634049654006958,
+      "learning_rate": 0.00014340390879478827,
+      "loss": 0.6867,
+      "step": 779
+    },
+    {
+      "epoch": 1.71334431630972,
+      "grad_norm": 0.16174709796905518,
+      "learning_rate": 0.00014315960912052116,
+      "loss": 0.8002,
+      "step": 780
+    },
+    {
+      "epoch": 1.7155409115870401,
+      "grad_norm": 0.1004948541522026,
+      "learning_rate": 0.00014291530944625404,
+      "loss": 0.7432,
+      "step": 781
+    },
+    {
+      "epoch": 1.7177375068643603,
+      "grad_norm": 0.1745937168598175,
+      "learning_rate": 0.00014267100977198696,
+      "loss": 0.659,
+      "step": 782
+    },
+    {
+      "epoch": 1.7199341021416803,
+      "grad_norm": 0.09289510548114777,
+      "learning_rate": 0.00014242671009771985,
+      "loss": 0.7597,
+      "step": 783
+    },
+    {
+      "epoch": 1.7221306974190007,
+      "grad_norm": 0.07728186249732971,
+      "learning_rate": 0.00014218241042345274,
+      "loss": 0.675,
+      "step": 784
+    },
+    {
+      "epoch": 1.7243272926963207,
+      "grad_norm": 0.10437046736478806,
+      "learning_rate": 0.00014193811074918566,
+      "loss": 0.9501,
+      "step": 785
+    },
+    {
+      "epoch": 1.7265238879736409,
+      "grad_norm": 0.06784020364284515,
+      "learning_rate": 0.00014169381107491854,
+      "loss": 0.6228,
+      "step": 786
+    },
+    {
+      "epoch": 1.728720483250961,
+      "grad_norm": 0.09907460957765579,
+      "learning_rate": 0.00014144951140065146,
+      "loss": 0.6573,
+      "step": 787
+    },
+    {
+      "epoch": 1.730917078528281,
+      "grad_norm": 0.06900086998939514,
+      "learning_rate": 0.00014120521172638435,
+      "loss": 0.698,
+      "step": 788
+    },
+    {
+      "epoch": 1.7331136738056014,
+      "grad_norm": 0.08838704973459244,
+      "learning_rate": 0.00014096091205211724,
+      "loss": 0.7555,
+      "step": 789
+    },
+    {
+      "epoch": 1.7353102690829214,
+      "grad_norm": 0.08264937251806259,
+      "learning_rate": 0.00014071661237785016,
+      "loss": 0.9209,
+      "step": 790
+    },
+    {
+      "epoch": 1.7375068643602416,
+      "grad_norm": 0.12355677783489227,
+      "learning_rate": 0.00014047231270358305,
+      "loss": 0.717,
+      "step": 791
+    },
+    {
+      "epoch": 1.7397034596375618,
+      "grad_norm": 0.06342349946498871,
+      "learning_rate": 0.00014022801302931593,
+      "loss": 0.6568,
+      "step": 792
+    },
+    {
+      "epoch": 1.741900054914882,
+      "grad_norm": 0.12166967242956161,
+      "learning_rate": 0.00013998371335504885,
+      "loss": 0.7172,
+      "step": 793
+    },
+    {
+      "epoch": 1.7440966501922022,
+      "grad_norm": 0.11651445925235748,
+      "learning_rate": 0.00013973941368078174,
+      "loss": 0.8057,
+      "step": 794
+    },
+    {
+      "epoch": 1.7462932454695221,
+      "grad_norm": 0.061137691140174866,
+      "learning_rate": 0.00013949511400651463,
+      "loss": 0.6628,
+      "step": 795
+    },
+    {
+      "epoch": 1.7484898407468425,
+      "grad_norm": 0.10012196749448776,
+      "learning_rate": 0.00013925081433224755,
+      "loss": 0.6482,
+      "step": 796
+    },
+    {
+      "epoch": 1.7506864360241625,
+      "grad_norm": 0.11898797005414963,
+      "learning_rate": 0.00013900651465798043,
+      "loss": 0.8104,
+      "step": 797
+    },
+    {
+      "epoch": 1.7528830313014827,
+      "grad_norm": 0.08395113050937653,
+      "learning_rate": 0.00013876221498371335,
+      "loss": 0.7232,
+      "step": 798
+    },
+    {
+      "epoch": 1.755079626578803,
+      "grad_norm": 0.12334415316581726,
+      "learning_rate": 0.00013851791530944624,
+      "loss": 0.8471,
+      "step": 799
+    },
+    {
+      "epoch": 1.7572762218561229,
+      "grad_norm": 0.09327132999897003,
+      "learning_rate": 0.00013827361563517913,
+      "loss": 0.8857,
+      "step": 800
+    },
+    {
+      "epoch": 1.7594728171334433,
+      "grad_norm": 0.0815991535782814,
+      "learning_rate": 0.00013802931596091205,
+      "loss": 0.8742,
+      "step": 801
+    },
+    {
+      "epoch": 1.7616694124107632,
+      "grad_norm": 0.08992715179920197,
+      "learning_rate": 0.00013778501628664493,
+      "loss": 0.9308,
+      "step": 802
+    },
+    {
+      "epoch": 1.7638660076880834,
+      "grad_norm": 0.11995363980531693,
+      "learning_rate": 0.00013754071661237782,
+      "loss": 1.0942,
+      "step": 803
+    },
+    {
+      "epoch": 1.7660626029654036,
+      "grad_norm": 0.12277841567993164,
+      "learning_rate": 0.00013729641693811074,
+      "loss": 0.9253,
+      "step": 804
+    },
+    {
+      "epoch": 1.7682591982427238,
+      "grad_norm": 0.12889382243156433,
+      "learning_rate": 0.00013705211726384363,
+      "loss": 0.7771,
+      "step": 805
+    },
+    {
+      "epoch": 1.770455793520044,
+      "grad_norm": 0.09973587840795517,
+      "learning_rate": 0.00013680781758957652,
+      "loss": 0.7146,
+      "step": 806
+    },
+    {
+      "epoch": 1.772652388797364,
+      "grad_norm": 0.07475234568119049,
+      "learning_rate": 0.00013656351791530943,
+      "loss": 0.7175,
+      "step": 807
+    },
+    {
+      "epoch": 1.7748489840746844,
+      "grad_norm": 0.1101953461766243,
+      "learning_rate": 0.00013631921824104232,
+      "loss": 0.8171,
+      "step": 808
+    },
+    {
+      "epoch": 1.7770455793520044,
+      "grad_norm": 0.11260540038347244,
+      "learning_rate": 0.0001360749185667752,
+      "loss": 0.8467,
+      "step": 809
+    },
+    {
+      "epoch": 1.7792421746293245,
+      "grad_norm": 0.11126644164323807,
+      "learning_rate": 0.00013583061889250813,
+      "loss": 0.5738,
+      "step": 810
+    },
+    {
+      "epoch": 1.7814387699066447,
+      "grad_norm": 0.09789043664932251,
+      "learning_rate": 0.00013558631921824102,
+      "loss": 0.6047,
+      "step": 811
+    },
+    {
+      "epoch": 1.7836353651839647,
+      "grad_norm": 0.10954687744379044,
+      "learning_rate": 0.00013534201954397394,
+      "loss": 0.713,
+      "step": 812
+    },
+    {
+      "epoch": 1.7858319604612851,
+      "grad_norm": 0.0999966561794281,
+      "learning_rate": 0.00013509771986970682,
+      "loss": 0.8543,
+      "step": 813
+    },
+    {
+      "epoch": 1.788028555738605,
+      "grad_norm": 0.09154272079467773,
+      "learning_rate": 0.00013485342019543971,
+      "loss": 0.7545,
+      "step": 814
+    },
+    {
+      "epoch": 1.7902251510159253,
+      "grad_norm": 0.09075820446014404,
+      "learning_rate": 0.00013460912052117263,
+      "loss": 0.8355,
+      "step": 815
+    },
+    {
+      "epoch": 1.7924217462932455,
+      "grad_norm": 0.09776823222637177,
+      "learning_rate": 0.00013436482084690552,
+      "loss": 0.7569,
+      "step": 816
+    },
+    {
+      "epoch": 1.7946183415705657,
+      "grad_norm": 0.07666546106338501,
+      "learning_rate": 0.0001341205211726384,
+      "loss": 0.6404,
+      "step": 817
+    },
+    {
+      "epoch": 1.7968149368478858,
+      "grad_norm": 0.0846526250243187,
+      "learning_rate": 0.00013387622149837132,
+      "loss": 0.7689,
+      "step": 818
+    },
+    {
+      "epoch": 1.7990115321252058,
+      "grad_norm": 0.08634993433952332,
+      "learning_rate": 0.00013363192182410421,
+      "loss": 0.7446,
+      "step": 819
+    },
+    {
+      "epoch": 1.8012081274025262,
+      "grad_norm": 0.07952354103326797,
+      "learning_rate": 0.0001333876221498371,
+      "loss": 0.7626,
+      "step": 820
+    },
+    {
+      "epoch": 1.8034047226798462,
+      "grad_norm": 0.07732746750116348,
+      "learning_rate": 0.00013314332247557002,
+      "loss": 0.7788,
+      "step": 821
+    },
+    {
+      "epoch": 1.8056013179571664,
+      "grad_norm": 0.10547160357236862,
+      "learning_rate": 0.0001328990228013029,
+      "loss": 0.8406,
+      "step": 822
+    },
+    {
+      "epoch": 1.8077979132344866,
+      "grad_norm": 0.08614319562911987,
+      "learning_rate": 0.00013265472312703582,
+      "loss": 0.7517,
+      "step": 823
+    },
+    {
+      "epoch": 1.8099945085118065,
+      "grad_norm": 0.07975021004676819,
+      "learning_rate": 0.00013241042345276871,
+      "loss": 0.6858,
+      "step": 824
+    },
+    {
+      "epoch": 1.812191103789127,
+      "grad_norm": 0.10631661862134933,
+      "learning_rate": 0.0001321661237785016,
+      "loss": 0.6367,
+      "step": 825
+    },
+    {
+      "epoch": 1.814387699066447,
+      "grad_norm": 0.10156068950891495,
+      "learning_rate": 0.00013192182410423452,
+      "loss": 0.5834,
+      "step": 826
+    },
+    {
+      "epoch": 1.8165842943437671,
+      "grad_norm": 0.06984843313694,
+      "learning_rate": 0.0001316775244299674,
+      "loss": 0.7966,
+      "step": 827
+    },
+    {
+      "epoch": 1.8187808896210873,
+      "grad_norm": 0.08283521980047226,
+      "learning_rate": 0.0001314332247557003,
+      "loss": 0.6566,
+      "step": 828
+    },
+    {
+      "epoch": 1.8209774848984075,
+      "grad_norm": 0.0947842001914978,
+      "learning_rate": 0.00013118892508143321,
+      "loss": 0.6876,
+      "step": 829
+    },
+    {
+      "epoch": 1.8231740801757277,
+      "grad_norm": 0.09557327628135681,
+      "learning_rate": 0.00013094462540716613,
+      "loss": 0.5553,
+      "step": 830
+    },
+    {
+      "epoch": 1.8253706754530477,
+      "grad_norm": 0.4152759313583374,
+      "learning_rate": 0.00013070032573289902,
+      "loss": 0.4997,
+      "step": 831
+    },
+    {
+      "epoch": 1.827567270730368,
+      "grad_norm": 0.09817035496234894,
+      "learning_rate": 0.0001304560260586319,
+      "loss": 0.754,
+      "step": 832
+    },
+    {
+      "epoch": 1.829763866007688,
+      "grad_norm": 0.10244613885879517,
+      "learning_rate": 0.00013021172638436483,
+      "loss": 0.8197,
+      "step": 833
+    },
+    {
+      "epoch": 1.8319604612850082,
+      "grad_norm": 0.08157302439212799,
+      "learning_rate": 0.00012996742671009771,
+      "loss": 0.856,
+      "step": 834
+    },
+    {
+      "epoch": 1.8341570565623284,
+      "grad_norm": 0.07550210505723953,
+      "learning_rate": 0.0001297231270358306,
+      "loss": 0.8157,
+      "step": 835
+    },
+    {
+      "epoch": 1.8363536518396484,
+      "grad_norm": 0.09909326583147049,
+      "learning_rate": 0.00012947882736156352,
+      "loss": 0.593,
+      "step": 836
+    },
+    {
+      "epoch": 1.8385502471169688,
+      "grad_norm": 0.5807304382324219,
+      "learning_rate": 0.0001292345276872964,
+      "loss": 0.7137,
+      "step": 837
+    },
+    {
+      "epoch": 1.8407468423942888,
+      "grad_norm": 0.09779902547597885,
+      "learning_rate": 0.00012899022801302933,
+      "loss": 0.7139,
+      "step": 838
+    },
+    {
+      "epoch": 1.842943437671609,
+      "grad_norm": 0.12162318080663681,
+      "learning_rate": 0.00012874592833876221,
+      "loss": 0.6422,
+      "step": 839
+    },
+    {
+      "epoch": 1.8451400329489291,
+      "grad_norm": 0.10663260519504547,
+      "learning_rate": 0.0001285016286644951,
+      "loss": 0.7997,
+      "step": 840
+    },
+    {
+      "epoch": 1.8473366282262493,
+      "grad_norm": 0.0878245085477829,
+      "learning_rate": 0.00012825732899022802,
+      "loss": 0.6906,
+      "step": 841
+    },
+    {
+      "epoch": 1.8495332235035695,
+      "grad_norm": 0.09591302275657654,
+      "learning_rate": 0.0001280130293159609,
+      "loss": 0.7297,
+      "step": 842
+    },
+    {
+      "epoch": 1.8517298187808895,
+      "grad_norm": 0.17532062530517578,
+      "learning_rate": 0.0001277687296416938,
+      "loss": 0.9073,
+      "step": 843
+    },
+    {
+      "epoch": 1.85392641405821,
+      "grad_norm": 0.07593477517366409,
+      "learning_rate": 0.00012752442996742671,
+      "loss": 0.7249,
+      "step": 844
+    },
+    {
+      "epoch": 1.8561230093355299,
+      "grad_norm": 0.06457901000976562,
+      "learning_rate": 0.0001272801302931596,
+      "loss": 0.6956,
+      "step": 845
+    },
+    {
+      "epoch": 1.85831960461285,
+      "grad_norm": 0.08696170896291733,
+      "learning_rate": 0.0001270358306188925,
+      "loss": 0.6639,
+      "step": 846
+    },
+    {
+      "epoch": 1.8605161998901703,
+      "grad_norm": 0.08451491594314575,
+      "learning_rate": 0.0001267915309446254,
+      "loss": 0.879,
+      "step": 847
+    },
+    {
+      "epoch": 1.8627127951674904,
+      "grad_norm": 0.18589206039905548,
+      "learning_rate": 0.0001265472312703583,
+      "loss": 0.6506,
+      "step": 848
+    },
+    {
+      "epoch": 1.8649093904448106,
+      "grad_norm": 0.07592837512493134,
+      "learning_rate": 0.0001263029315960912,
+      "loss": 0.6281,
+      "step": 849
+    },
+    {
+      "epoch": 1.8671059857221306,
+      "grad_norm": 0.05951695144176483,
+      "learning_rate": 0.0001260586319218241,
+      "loss": 0.6377,
+      "step": 850
+    },
+    {
+      "epoch": 1.869302580999451,
+      "grad_norm": 0.08469297736883163,
+      "learning_rate": 0.000125814332247557,
+      "loss": 0.5978,
+      "step": 851
+    },
+    {
+      "epoch": 1.871499176276771,
+      "grad_norm": 0.09282078593969345,
+      "learning_rate": 0.0001255700325732899,
+      "loss": 0.7491,
+      "step": 852
+    },
+    {
+      "epoch": 1.8736957715540912,
+      "grad_norm": 0.1367119550704956,
+      "learning_rate": 0.0001253257328990228,
+      "loss": 0.8698,
+      "step": 853
+    },
+    {
+      "epoch": 1.8758923668314114,
+      "grad_norm": 0.12527549266815186,
+      "learning_rate": 0.0001250814332247557,
+      "loss": 0.8675,
+      "step": 854
+    },
+    {
+      "epoch": 1.8780889621087313,
+      "grad_norm": 0.09542479366064072,
+      "learning_rate": 0.0001248371335504886,
+      "loss": 0.6638,
+      "step": 855
+    },
+    {
+      "epoch": 1.8802855573860517,
+      "grad_norm": 0.08716510981321335,
+      "learning_rate": 0.0001245928338762215,
+      "loss": 0.8795,
+      "step": 856
+    },
+    {
+      "epoch": 1.8824821526633717,
+      "grad_norm": 0.08640987426042557,
+      "learning_rate": 0.00012434853420195438,
+      "loss": 0.7034,
+      "step": 857
+    },
+    {
+      "epoch": 1.884678747940692,
+      "grad_norm": 0.08230648189783096,
+      "learning_rate": 0.0001241042345276873,
+      "loss": 0.7732,
+      "step": 858
+    },
+    {
+      "epoch": 1.886875343218012,
+      "grad_norm": 0.0751597061753273,
+      "learning_rate": 0.0001238599348534202,
+      "loss": 0.6392,
+      "step": 859
+    },
+    {
+      "epoch": 1.8890719384953323,
+      "grad_norm": 0.06654468923807144,
+      "learning_rate": 0.00012361563517915308,
+      "loss": 0.6358,
+      "step": 860
+    },
+    {
+      "epoch": 1.8912685337726525,
+      "grad_norm": 0.08097635954618454,
+      "learning_rate": 0.000123371335504886,
+      "loss": 0.8546,
+      "step": 861
+    },
+    {
+      "epoch": 1.8934651290499724,
+      "grad_norm": 0.07630196213722229,
+      "learning_rate": 0.00012312703583061888,
+      "loss": 0.7371,
+      "step": 862
+    },
+    {
+      "epoch": 1.8956617243272929,
+      "grad_norm": 0.09249162673950195,
+      "learning_rate": 0.00012288273615635177,
+      "loss": 0.9853,
+      "step": 863
+    },
+    {
+      "epoch": 1.8978583196046128,
+      "grad_norm": 0.09021256119012833,
+      "learning_rate": 0.0001226384364820847,
+      "loss": 0.6011,
+      "step": 864
+    },
+    {
+      "epoch": 1.900054914881933,
+      "grad_norm": 0.08402383327484131,
+      "learning_rate": 0.00012239413680781758,
+      "loss": 0.7914,
+      "step": 865
+    },
+    {
+      "epoch": 1.9022515101592532,
+      "grad_norm": 0.09796092659235,
+      "learning_rate": 0.0001221498371335505,
+      "loss": 0.6726,
+      "step": 866
+    },
+    {
+      "epoch": 1.9044481054365732,
+      "grad_norm": 0.08002438396215439,
+      "learning_rate": 0.00012190553745928338,
+      "loss": 0.6634,
+      "step": 867
+    },
+    {
+      "epoch": 1.9066447007138936,
+      "grad_norm": 0.10942327976226807,
+      "learning_rate": 0.00012166123778501629,
+      "loss": 0.9135,
+      "step": 868
+    },
+    {
+      "epoch": 1.9088412959912135,
+      "grad_norm": 0.09269405901432037,
+      "learning_rate": 0.00012141693811074918,
+      "loss": 0.8432,
+      "step": 869
+    },
+    {
+      "epoch": 1.9110378912685337,
+      "grad_norm": 0.06608925014734268,
+      "learning_rate": 0.00012117263843648208,
+      "loss": 0.6858,
+      "step": 870
+    },
+    {
+      "epoch": 1.913234486545854,
+      "grad_norm": 0.130803644657135,
+      "learning_rate": 0.00012092833876221498,
+      "loss": 0.6935,
+      "step": 871
+    },
+    {
+      "epoch": 1.9154310818231741,
+      "grad_norm": 0.07922960072755814,
+      "learning_rate": 0.00012068403908794787,
+      "loss": 0.801,
+      "step": 872
+    },
+    {
+      "epoch": 1.9176276771004943,
+      "grad_norm": 0.1251368373632431,
+      "learning_rate": 0.00012043973941368077,
+      "loss": 0.8661,
+      "step": 873
+    },
+    {
+      "epoch": 1.9198242723778143,
+      "grad_norm": 0.07598315179347992,
+      "learning_rate": 0.00012019543973941368,
+      "loss": 0.9698,
+      "step": 874
+    },
+    {
+      "epoch": 1.9220208676551347,
+      "grad_norm": 0.09361431002616882,
+      "learning_rate": 0.00011995114006514658,
+      "loss": 0.8881,
+      "step": 875
+    },
+    {
+      "epoch": 1.9242174629324547,
+      "grad_norm": 0.14713425934314728,
+      "learning_rate": 0.00011970684039087947,
+      "loss": 0.8943,
+      "step": 876
+    },
+    {
+      "epoch": 1.9264140582097748,
+      "grad_norm": 0.07065768539905548,
+      "learning_rate": 0.00011946254071661237,
+      "loss": 0.6442,
+      "step": 877
+    },
+    {
+      "epoch": 1.928610653487095,
+      "grad_norm": 0.1198662742972374,
+      "learning_rate": 0.00011921824104234527,
+      "loss": 0.8335,
+      "step": 878
+    },
+    {
+      "epoch": 1.930807248764415,
+      "grad_norm": 0.07292979955673218,
+      "learning_rate": 0.00011897394136807816,
+      "loss": 0.7475,
+      "step": 879
+    },
+    {
+      "epoch": 1.9330038440417354,
+      "grad_norm": 0.09775891900062561,
+      "learning_rate": 0.00011872964169381106,
+      "loss": 0.6854,
+      "step": 880
+    },
+    {
+      "epoch": 1.9352004393190554,
+      "grad_norm": 0.07953374832868576,
+      "learning_rate": 0.00011848534201954397,
+      "loss": 0.7478,
+      "step": 881
+    },
+    {
+      "epoch": 1.9373970345963756,
+      "grad_norm": 0.06859882175922394,
+      "learning_rate": 0.00011824104234527687,
+      "loss": 0.5737,
+      "step": 882
+    },
+    {
+      "epoch": 1.9395936298736958,
+      "grad_norm": 0.11728372424840927,
+      "learning_rate": 0.00011799674267100976,
+      "loss": 0.8214,
+      "step": 883
+    },
+    {
+      "epoch": 1.941790225151016,
+      "grad_norm": 0.10215143114328384,
+      "learning_rate": 0.00011775244299674266,
+      "loss": 0.7691,
+      "step": 884
+    },
+    {
+      "epoch": 1.9439868204283361,
+      "grad_norm": 0.07876396179199219,
+      "learning_rate": 0.00011750814332247557,
+      "loss": 1.1291,
+      "step": 885
+    },
+    {
+      "epoch": 1.9461834157056561,
+      "grad_norm": 0.11178891360759735,
+      "learning_rate": 0.00011726384364820845,
+      "loss": 0.7877,
+      "step": 886
+    },
+    {
+      "epoch": 1.9483800109829765,
+      "grad_norm": 0.1061316654086113,
+      "learning_rate": 0.00011701954397394136,
+      "loss": 0.6924,
+      "step": 887
+    },
+    {
+      "epoch": 1.9505766062602965,
+      "grad_norm": 0.10667377710342407,
+      "learning_rate": 0.00011677524429967426,
+      "loss": 0.6308,
+      "step": 888
+    },
+    {
+      "epoch": 1.9527732015376167,
+      "grad_norm": 0.08948090672492981,
+      "learning_rate": 0.00011653094462540716,
+      "loss": 0.7446,
+      "step": 889
+    },
+    {
+      "epoch": 1.9549697968149369,
+      "grad_norm": 0.07279379665851593,
+      "learning_rate": 0.00011628664495114005,
+      "loss": 0.7455,
+      "step": 890
+    },
+    {
+      "epoch": 1.9571663920922568,
+      "grad_norm": 0.07213126122951508,
+      "learning_rate": 0.00011604234527687295,
+      "loss": 0.8635,
+      "step": 891
+    },
+    {
+      "epoch": 1.9593629873695773,
+      "grad_norm": 0.14409837126731873,
+      "learning_rate": 0.00011579804560260586,
+      "loss": 0.646,
+      "step": 892
+    },
+    {
+      "epoch": 1.9615595826468972,
+      "grad_norm": 0.08503877371549606,
+      "learning_rate": 0.00011555374592833876,
+      "loss": 0.6759,
+      "step": 893
+    },
+    {
+      "epoch": 1.9637561779242174,
+      "grad_norm": 0.169550821185112,
+      "learning_rate": 0.00011530944625407165,
+      "loss": 0.8217,
+      "step": 894
+    },
+    {
+      "epoch": 1.9659527732015376,
+      "grad_norm": 0.10099968314170837,
+      "learning_rate": 0.00011506514657980455,
+      "loss": 0.6071,
+      "step": 895
+    },
+    {
+      "epoch": 1.9681493684788578,
+      "grad_norm": 0.08358711749315262,
+      "learning_rate": 0.00011482084690553745,
+      "loss": 0.8094,
+      "step": 896
+    },
+    {
+      "epoch": 1.970345963756178,
+      "grad_norm": 0.06357203423976898,
+      "learning_rate": 0.00011457654723127034,
+      "loss": 0.6156,
+      "step": 897
+    },
+    {
+      "epoch": 1.972542559033498,
+      "grad_norm": 0.0759328156709671,
+      "learning_rate": 0.00011433224755700325,
+      "loss": 0.8597,
+      "step": 898
+    },
+    {
+      "epoch": 1.9747391543108184,
+      "grad_norm": 0.1413506269454956,
+      "learning_rate": 0.00011408794788273615,
+      "loss": 0.8088,
+      "step": 899
+    },
+    {
+      "epoch": 1.9769357495881383,
+      "grad_norm": 0.06626413017511368,
+      "learning_rate": 0.00011384364820846905,
+      "loss": 0.6115,
+      "step": 900
+    },
+    {
+      "epoch": 1.9791323448654585,
+      "grad_norm": 0.07184301316738129,
+      "learning_rate": 0.00011359934853420194,
+      "loss": 0.8042,
+      "step": 901
+    },
+    {
+      "epoch": 1.9813289401427787,
+      "grad_norm": 0.14539746940135956,
+      "learning_rate": 0.00011335504885993484,
+      "loss": 0.8348,
+      "step": 902
+    },
+    {
+      "epoch": 1.9835255354200987,
+      "grad_norm": 0.06488694250583649,
+      "learning_rate": 0.00011311074918566775,
+      "loss": 0.7489,
+      "step": 903
+    },
+    {
+      "epoch": 1.985722130697419,
+      "grad_norm": 0.0733429417014122,
+      "learning_rate": 0.00011286644951140064,
+      "loss": 0.6248,
+      "step": 904
+    },
+    {
+      "epoch": 1.987918725974739,
+      "grad_norm": 0.10581281036138535,
+      "learning_rate": 0.00011262214983713354,
+      "loss": 0.6217,
+      "step": 905
+    },
+    {
+      "epoch": 1.9901153212520593,
+      "grad_norm": 0.11003848910331726,
+      "learning_rate": 0.00011237785016286644,
+      "loss": 0.8393,
+      "step": 906
+    },
+    {
+      "epoch": 1.9923119165293794,
+      "grad_norm": 0.11129264533519745,
+      "learning_rate": 0.00011213355048859934,
+      "loss": 0.6455,
+      "step": 907
+    },
+    {
+      "epoch": 1.9945085118066996,
+      "grad_norm": 0.14178551733493805,
+      "learning_rate": 0.00011188925081433223,
+      "loss": 0.6235,
+      "step": 908
+    },
+    {
+      "epoch": 1.9967051070840198,
+      "grad_norm": 0.5554675459861755,
+      "learning_rate": 0.00011164495114006514,
+      "loss": 0.5604,
+      "step": 909
+    },
+    {
+      "epoch": 1.9989017023613398,
+      "grad_norm": 0.1929876059293747,
+      "learning_rate": 0.00011140065146579804,
+      "loss": 0.7504,
+      "step": 910
+    },
+    {
+      "epoch": 2.00109829763866,
+      "grad_norm": 0.07121291756629944,
+      "learning_rate": 0.00011115635179153093,
+      "loss": 0.8188,
+      "step": 911
+    },
+    {
+      "epoch": 2.00329489291598,
+      "grad_norm": 0.13832546770572662,
+      "learning_rate": 0.00011091205211726383,
+      "loss": 0.6794,
+      "step": 912
+    },
+    {
+      "epoch": 2.0054914881933006,
+      "grad_norm": 0.11828719079494476,
+      "learning_rate": 0.00011066775244299673,
+      "loss": 0.6005,
+      "step": 913
+    },
+    {
+      "epoch": 2.0076880834706206,
+      "grad_norm": 0.10825496912002563,
+      "learning_rate": 0.00011042345276872964,
+      "loss": 0.7374,
+      "step": 914
+    },
+    {
+      "epoch": 2.0098846787479405,
+      "grad_norm": 0.09058716893196106,
+      "learning_rate": 0.00011017915309446253,
+      "loss": 0.7058,
+      "step": 915
+    },
+    {
+      "epoch": 2.012081274025261,
+      "grad_norm": 0.09288596361875534,
+      "learning_rate": 0.00010993485342019543,
+      "loss": 0.8974,
+      "step": 916
+    },
+    {
+      "epoch": 2.014277869302581,
+      "grad_norm": 0.10922420769929886,
+      "learning_rate": 0.00010969055374592833,
+      "loss": 0.9562,
+      "step": 917
+    },
+    {
+      "epoch": 2.0164744645799013,
+      "grad_norm": 0.07443702965974808,
+      "learning_rate": 0.00010944625407166122,
+      "loss": 0.8458,
+      "step": 918
+    },
+    {
+      "epoch": 2.0186710598572213,
+      "grad_norm": 0.07446782290935516,
+      "learning_rate": 0.00010920195439739412,
+      "loss": 0.6571,
+      "step": 919
+    },
+    {
+      "epoch": 2.0208676551345413,
+      "grad_norm": 0.07167340815067291,
+      "learning_rate": 0.00010895765472312703,
+      "loss": 0.6747,
+      "step": 920
+    },
+    {
+      "epoch": 2.0230642504118617,
+      "grad_norm": 0.14842204749584198,
+      "learning_rate": 0.00010871335504885993,
+      "loss": 0.6627,
+      "step": 921
+    },
+    {
+      "epoch": 2.0252608456891816,
+      "grad_norm": 0.08280177414417267,
+      "learning_rate": 0.00010846905537459282,
+      "loss": 0.7555,
+      "step": 922
+    },
+    {
+      "epoch": 2.027457440966502,
+      "grad_norm": 0.11193451285362244,
+      "learning_rate": 0.00010822475570032572,
+      "loss": 0.7075,
+      "step": 923
+    },
+    {
+      "epoch": 2.029654036243822,
+      "grad_norm": 0.1014566496014595,
+      "learning_rate": 0.00010798045602605862,
+      "loss": 0.8572,
+      "step": 924
+    },
+    {
+      "epoch": 2.0318506315211424,
+      "grad_norm": 0.07931726425886154,
+      "learning_rate": 0.00010773615635179153,
+      "loss": 0.7187,
+      "step": 925
+    },
+    {
+      "epoch": 2.0340472267984624,
+      "grad_norm": 0.10226915031671524,
+      "learning_rate": 0.00010749185667752442,
+      "loss": 0.7189,
+      "step": 926
+    },
+    {
+      "epoch": 2.0362438220757824,
+      "grad_norm": 0.09880972653627396,
+      "learning_rate": 0.00010724755700325732,
+      "loss": 0.8974,
+      "step": 927
+    },
+    {
+      "epoch": 2.0384404173531028,
+      "grad_norm": 0.08638790994882584,
+      "learning_rate": 0.00010700325732899022,
+      "loss": 0.721,
+      "step": 928
+    },
+    {
+      "epoch": 2.0406370126304227,
+      "grad_norm": 0.08493978530168533,
+      "learning_rate": 0.00010675895765472311,
+      "loss": 0.6681,
+      "step": 929
+    },
+    {
+      "epoch": 2.042833607907743,
+      "grad_norm": 0.07533694058656693,
+      "learning_rate": 0.00010651465798045601,
+      "loss": 0.6076,
+      "step": 930
+    },
+    {
+      "epoch": 2.045030203185063,
+      "grad_norm": 0.08745653182268143,
+      "learning_rate": 0.00010627035830618892,
+      "loss": 0.6719,
+      "step": 931
+    },
+    {
+      "epoch": 2.047226798462383,
+      "grad_norm": 0.10198769718408585,
+      "learning_rate": 0.00010602605863192182,
+      "loss": 0.7786,
+      "step": 932
+    },
+    {
+      "epoch": 2.0494233937397035,
+      "grad_norm": 0.08564738929271698,
+      "learning_rate": 0.00010578175895765471,
+      "loss": 0.7438,
+      "step": 933
+    },
+    {
+      "epoch": 2.0516199890170235,
+      "grad_norm": 0.13399513065814972,
+      "learning_rate": 0.00010553745928338761,
+      "loss": 0.6533,
+      "step": 934
+    },
+    {
+      "epoch": 2.053816584294344,
+      "grad_norm": 0.06405504792928696,
+      "learning_rate": 0.00010529315960912051,
+      "loss": 0.662,
+      "step": 935
+    },
+    {
+      "epoch": 2.056013179571664,
+      "grad_norm": 0.08755522221326828,
+      "learning_rate": 0.0001050488599348534,
+      "loss": 0.8811,
+      "step": 936
+    },
+    {
+      "epoch": 2.0582097748489843,
+      "grad_norm": 0.07031279802322388,
+      "learning_rate": 0.0001048045602605863,
+      "loss": 0.7913,
+      "step": 937
+    },
+    {
+      "epoch": 2.0604063701263042,
+      "grad_norm": 0.08004021644592285,
+      "learning_rate": 0.00010456026058631921,
+      "loss": 0.5375,
+      "step": 938
+    },
+    {
+      "epoch": 2.062602965403624,
+      "grad_norm": 0.09203504770994186,
+      "learning_rate": 0.00010431596091205211,
+      "loss": 0.8948,
+      "step": 939
+    },
+    {
+      "epoch": 2.0647995606809446,
+      "grad_norm": 0.0798863023519516,
+      "learning_rate": 0.000104071661237785,
+      "loss": 0.7004,
+      "step": 940
+    },
+    {
+      "epoch": 2.0669961559582646,
+      "grad_norm": 0.10599152743816376,
+      "learning_rate": 0.0001038273615635179,
+      "loss": 0.7618,
+      "step": 941
+    },
+    {
+      "epoch": 2.069192751235585,
+      "grad_norm": 0.0862710252404213,
+      "learning_rate": 0.0001035830618892508,
+      "loss": 0.6054,
+      "step": 942
+    },
+    {
+      "epoch": 2.071389346512905,
+      "grad_norm": 0.07118751853704453,
+      "learning_rate": 0.0001033387622149837,
+      "loss": 0.7302,
+      "step": 943
+    },
+    {
+      "epoch": 2.073585941790225,
+      "grad_norm": 0.09174022078514099,
+      "learning_rate": 0.0001030944625407166,
+      "loss": 0.6519,
+      "step": 944
+    },
+    {
+      "epoch": 2.0757825370675453,
+      "grad_norm": 0.07677482813596725,
+      "learning_rate": 0.0001028501628664495,
+      "loss": 0.707,
+      "step": 945
+    },
+    {
+      "epoch": 2.0779791323448653,
+      "grad_norm": 0.0877048596739769,
+      "learning_rate": 0.0001026058631921824,
+      "loss": 0.748,
+      "step": 946
+    },
+    {
+      "epoch": 2.0801757276221857,
+      "grad_norm": 0.07210325449705124,
+      "learning_rate": 0.00010236156351791529,
+      "loss": 0.5893,
+      "step": 947
+    },
+    {
+      "epoch": 2.0823723228995057,
+      "grad_norm": 0.06885118037462234,
+      "learning_rate": 0.0001021172638436482,
+      "loss": 0.6423,
+      "step": 948
+    },
+    {
+      "epoch": 2.084568918176826,
+      "grad_norm": 0.2756057381629944,
+      "learning_rate": 0.0001018729641693811,
+      "loss": 0.8973,
+      "step": 949
+    },
+    {
+      "epoch": 2.086765513454146,
+      "grad_norm": 0.07163773477077484,
+      "learning_rate": 0.00010162866449511399,
+      "loss": 0.732,
+      "step": 950
+    },
+    {
+      "epoch": 2.088962108731466,
+      "grad_norm": 0.09421386569738388,
+      "learning_rate": 0.00010138436482084689,
+      "loss": 0.7445,
+      "step": 951
+    },
+    {
+      "epoch": 2.0911587040087865,
+      "grad_norm": 0.1686144918203354,
+      "learning_rate": 0.00010114006514657979,
+      "loss": 0.8277,
+      "step": 952
+    },
+    {
+      "epoch": 2.0933552992861064,
+      "grad_norm": 0.16856926679611206,
+      "learning_rate": 0.0001008957654723127,
+      "loss": 0.832,
+      "step": 953
+    },
+    {
+      "epoch": 2.095551894563427,
+      "grad_norm": 0.08712480217218399,
+      "learning_rate": 0.00010065146579804558,
+      "loss": 0.6419,
+      "step": 954
+    },
+    {
+      "epoch": 2.097748489840747,
+      "grad_norm": 0.07930180430412292,
+      "learning_rate": 0.00010040716612377849,
+      "loss": 0.8413,
+      "step": 955
+    },
+    {
+      "epoch": 2.099945085118067,
+      "grad_norm": 0.09859949350357056,
+      "learning_rate": 0.00010016286644951139,
+      "loss": 0.7405,
+      "step": 956
+    },
+    {
+      "epoch": 2.102141680395387,
+      "grad_norm": 0.1905672550201416,
+      "learning_rate": 9.991856677524428e-05,
+      "loss": 0.7999,
+      "step": 957
+    },
+    {
+      "epoch": 2.104338275672707,
+      "grad_norm": 0.07940427213907242,
+      "learning_rate": 9.967426710097718e-05,
+      "loss": 0.6854,
+      "step": 958
+    },
+    {
+      "epoch": 2.1065348709500276,
+      "grad_norm": 0.10216717422008514,
+      "learning_rate": 9.942996742671008e-05,
+      "loss": 0.6894,
+      "step": 959
+    },
+    {
+      "epoch": 2.1087314662273475,
+      "grad_norm": 0.07466311007738113,
+      "learning_rate": 9.918566775244299e-05,
+      "loss": 0.7271,
+      "step": 960
+    },
+    {
+      "epoch": 2.110928061504668,
+      "grad_norm": 0.09368406981229782,
+      "learning_rate": 9.894136807817588e-05,
+      "loss": 0.8195,
+      "step": 961
+    },
+    {
+      "epoch": 2.113124656781988,
+      "grad_norm": 0.09201403707265854,
+      "learning_rate": 9.869706840390878e-05,
+      "loss": 0.6912,
+      "step": 962
+    },
+    {
+      "epoch": 2.115321252059308,
+      "grad_norm": 0.13242104649543762,
+      "learning_rate": 9.845276872964168e-05,
+      "loss": 0.6201,
+      "step": 963
+    },
+    {
+      "epoch": 2.1175178473366283,
+      "grad_norm": 0.07200752198696136,
+      "learning_rate": 9.820846905537458e-05,
+      "loss": 0.6795,
+      "step": 964
+    },
+    {
+      "epoch": 2.1197144426139483,
+      "grad_norm": 0.07441911846399307,
+      "learning_rate": 9.796416938110747e-05,
+      "loss": 0.6565,
+      "step": 965
+    },
+    {
+      "epoch": 2.1219110378912687,
+      "grad_norm": 0.11978007107973099,
+      "learning_rate": 9.771986970684038e-05,
+      "loss": 0.6715,
+      "step": 966
+    },
+    {
+      "epoch": 2.1241076331685886,
+      "grad_norm": 0.07549604028463364,
+      "learning_rate": 9.747557003257328e-05,
+      "loss": 0.6961,
+      "step": 967
+    },
+    {
+      "epoch": 2.1263042284459086,
+      "grad_norm": 0.08318581432104111,
+      "learning_rate": 9.723127035830617e-05,
+      "loss": 0.668,
+      "step": 968
+    },
+    {
+      "epoch": 2.128500823723229,
+      "grad_norm": 0.1600196808576584,
+      "learning_rate": 9.698697068403907e-05,
+      "loss": 0.6855,
+      "step": 969
+    },
+    {
+      "epoch": 2.130697419000549,
+      "grad_norm": 0.0858573243021965,
+      "learning_rate": 9.674267100977197e-05,
+      "loss": 0.8137,
+      "step": 970
+    },
+    {
+      "epoch": 2.1328940142778694,
+      "grad_norm": 0.11876931041479111,
+      "learning_rate": 9.649837133550488e-05,
+      "loss": 0.7526,
+      "step": 971
+    },
+    {
+      "epoch": 2.1350906095551894,
+      "grad_norm": 0.10300009697675705,
+      "learning_rate": 9.625407166123777e-05,
+      "loss": 0.8017,
+      "step": 972
+    },
+    {
+      "epoch": 2.13728720483251,
+      "grad_norm": 0.08675804734230042,
+      "learning_rate": 9.600977198697067e-05,
+      "loss": 0.7161,
+      "step": 973
+    },
+    {
+      "epoch": 2.1394838001098297,
+      "grad_norm": 0.08652209490537643,
+      "learning_rate": 9.576547231270357e-05,
+      "loss": 0.7863,
+      "step": 974
+    },
+    {
+      "epoch": 2.1416803953871497,
+      "grad_norm": 0.08803403377532959,
+      "learning_rate": 9.552117263843646e-05,
+      "loss": 0.7885,
+      "step": 975
+    },
+    {
+      "epoch": 2.14387699066447,
+      "grad_norm": 0.08228699862957001,
+      "learning_rate": 9.527687296416936e-05,
+      "loss": 0.7349,
+      "step": 976
+    },
+    {
+      "epoch": 2.14607358594179,
+      "grad_norm": 0.06837975233793259,
+      "learning_rate": 9.503257328990227e-05,
+      "loss": 0.7276,
+      "step": 977
+    },
+    {
+      "epoch": 2.1482701812191105,
+      "grad_norm": 0.07937322556972504,
+      "learning_rate": 9.478827361563517e-05,
+      "loss": 0.6754,
+      "step": 978
+    },
+    {
+      "epoch": 2.1504667764964305,
+      "grad_norm": 0.07611627131700516,
+      "learning_rate": 9.454397394136806e-05,
+      "loss": 0.7182,
+      "step": 979
+    },
+    {
+      "epoch": 2.152663371773751,
+      "grad_norm": 0.09503597021102905,
+      "learning_rate": 9.429967426710096e-05,
+      "loss": 0.7906,
+      "step": 980
+    },
+    {
+      "epoch": 2.154859967051071,
+      "grad_norm": 0.26675093173980713,
+      "learning_rate": 9.405537459283386e-05,
+      "loss": 0.5422,
+      "step": 981
+    },
+    {
+      "epoch": 2.157056562328391,
+      "grad_norm": 0.07366194576025009,
+      "learning_rate": 9.381107491856675e-05,
+      "loss": 0.6434,
+      "step": 982
+    },
+    {
+      "epoch": 2.1592531576057112,
+      "grad_norm": 0.09013357758522034,
+      "learning_rate": 9.356677524429967e-05,
+      "loss": 0.8871,
+      "step": 983
+    },
+    {
+      "epoch": 2.161449752883031,
+      "grad_norm": 0.09186428785324097,
+      "learning_rate": 9.332247557003257e-05,
+      "loss": 0.6643,
+      "step": 984
+    },
+    {
+      "epoch": 2.1636463481603516,
+      "grad_norm": 0.07400774210691452,
+      "learning_rate": 9.307817589576547e-05,
+      "loss": 0.6424,
+      "step": 985
+    },
+    {
+      "epoch": 2.1658429434376716,
+      "grad_norm": 0.09549415856599808,
+      "learning_rate": 9.283387622149838e-05,
+      "loss": 0.6556,
+      "step": 986
+    },
+    {
+      "epoch": 2.168039538714992,
+      "grad_norm": 0.10859312117099762,
+      "learning_rate": 9.258957654723127e-05,
+      "loss": 0.8118,
+      "step": 987
+    },
+    {
+      "epoch": 2.170236133992312,
+      "grad_norm": 0.09639792144298553,
+      "learning_rate": 9.234527687296417e-05,
+      "loss": 0.6293,
+      "step": 988
+    },
+    {
+      "epoch": 2.172432729269632,
+      "grad_norm": 0.08275002986192703,
+      "learning_rate": 9.210097719869707e-05,
+      "loss": 0.6972,
+      "step": 989
+    },
+    {
+      "epoch": 2.1746293245469523,
+      "grad_norm": 0.09591136127710342,
+      "learning_rate": 9.185667752442996e-05,
+      "loss": 0.8958,
+      "step": 990
+    },
+    {
+      "epoch": 2.1768259198242723,
+      "grad_norm": 0.15456514060497284,
+      "learning_rate": 9.161237785016286e-05,
+      "loss": 0.693,
+      "step": 991
+    },
+    {
+      "epoch": 2.1790225151015927,
+      "grad_norm": 0.06113564223051071,
+      "learning_rate": 9.136807817589577e-05,
+      "loss": 0.6052,
+      "step": 992
+    },
+    {
+      "epoch": 2.1812191103789127,
+      "grad_norm": 0.203088641166687,
+      "learning_rate": 9.112377850162867e-05,
+      "loss": 0.8198,
+      "step": 993
+    },
+    {
+      "epoch": 2.1834157056562327,
+      "grad_norm": 0.07402162998914719,
+      "learning_rate": 9.087947882736156e-05,
+      "loss": 0.6501,
+      "step": 994
+    },
+    {
+      "epoch": 2.185612300933553,
+      "grad_norm": 0.08272642642259598,
+      "learning_rate": 9.063517915309446e-05,
+      "loss": 0.674,
+      "step": 995
+    },
+    {
+      "epoch": 2.187808896210873,
+      "grad_norm": 0.12218045443296432,
+      "learning_rate": 9.039087947882736e-05,
+      "loss": 0.8288,
+      "step": 996
+    },
+    {
+      "epoch": 2.1900054914881935,
+      "grad_norm": 0.08559254556894302,
+      "learning_rate": 9.014657980456025e-05,
+      "loss": 0.773,
+      "step": 997
+    },
+    {
+      "epoch": 2.1922020867655134,
+      "grad_norm": 0.11552423983812332,
+      "learning_rate": 8.990228013029316e-05,
+      "loss": 0.7435,
+      "step": 998
+    },
+    {
+      "epoch": 2.1943986820428334,
+      "grad_norm": 0.09281519800424576,
+      "learning_rate": 8.965798045602606e-05,
+      "loss": 0.8644,
+      "step": 999
+    },
+    {
+      "epoch": 2.196595277320154,
+      "grad_norm": 0.0705346167087555,
+      "learning_rate": 8.941368078175896e-05,
+      "loss": 0.655,
+      "step": 1000
+    },
+    {
+      "epoch": 2.1987918725974738,
+      "grad_norm": 0.10760047286748886,
+      "learning_rate": 8.916938110749185e-05,
+      "loss": 0.5835,
+      "step": 1001
+    },
+    {
+      "epoch": 2.200988467874794,
+      "grad_norm": 0.0796201229095459,
+      "learning_rate": 8.892508143322475e-05,
+      "loss": 0.6553,
+      "step": 1002
+    },
+    {
+      "epoch": 2.203185063152114,
+      "grad_norm": 0.09482817351818085,
+      "learning_rate": 8.868078175895766e-05,
+      "loss": 0.721,
+      "step": 1003
+    },
+    {
+      "epoch": 2.2053816584294346,
+      "grad_norm": 0.10091966390609741,
+      "learning_rate": 8.843648208469055e-05,
+      "loss": 0.6347,
+      "step": 1004
+    },
+    {
+      "epoch": 2.2075782537067545,
+      "grad_norm": 0.1878889501094818,
+      "learning_rate": 8.819218241042345e-05,
+      "loss": 0.653,
+      "step": 1005
+    },
+    {
+      "epoch": 2.2097748489840745,
+      "grad_norm": 0.10120739787817001,
+      "learning_rate": 8.794788273615635e-05,
+      "loss": 0.7566,
+      "step": 1006
+    },
+    {
+      "epoch": 2.211971444261395,
+      "grad_norm": 0.09095592796802521,
+      "learning_rate": 8.770358306188925e-05,
+      "loss": 0.6207,
+      "step": 1007
+    },
+    {
+      "epoch": 2.214168039538715,
+      "grad_norm": 0.10327545553445816,
+      "learning_rate": 8.745928338762214e-05,
+      "loss": 0.8814,
+      "step": 1008
+    },
+    {
+      "epoch": 2.2163646348160353,
+      "grad_norm": 0.06494822353124619,
+      "learning_rate": 8.721498371335505e-05,
+      "loss": 0.721,
+      "step": 1009
+    },
+    {
+      "epoch": 2.2185612300933553,
+      "grad_norm": 0.1915939748287201,
+      "learning_rate": 8.697068403908795e-05,
+      "loss": 0.7485,
+      "step": 1010
+    },
+    {
+      "epoch": 2.2207578253706757,
+      "grad_norm": 0.08327042311429977,
+      "learning_rate": 8.672638436482085e-05,
+      "loss": 0.737,
+      "step": 1011
+    },
+    {
+      "epoch": 2.2229544206479956,
+      "grad_norm": 0.10128361731767654,
+      "learning_rate": 8.648208469055374e-05,
+      "loss": 0.9448,
+      "step": 1012
+    },
+    {
+      "epoch": 2.2251510159253156,
+      "grad_norm": 0.1288832277059555,
+      "learning_rate": 8.623778501628664e-05,
+      "loss": 0.7745,
+      "step": 1013
+    },
+    {
+      "epoch": 2.227347611202636,
+      "grad_norm": 0.10600150376558304,
+      "learning_rate": 8.599348534201955e-05,
+      "loss": 0.879,
+      "step": 1014
+    },
+    {
+      "epoch": 2.229544206479956,
+      "grad_norm": 0.12065941095352173,
+      "learning_rate": 8.574918566775243e-05,
+      "loss": 0.8864,
+      "step": 1015
+    },
+    {
+      "epoch": 2.2317408017572764,
+      "grad_norm": 0.07569331675767899,
+      "learning_rate": 8.550488599348534e-05,
+      "loss": 0.7631,
+      "step": 1016
+    },
+    {
+      "epoch": 2.2339373970345964,
+      "grad_norm": 0.10565028339624405,
+      "learning_rate": 8.526058631921824e-05,
+      "loss": 0.8103,
+      "step": 1017
+    },
+    {
+      "epoch": 2.2361339923119163,
+      "grad_norm": 0.09105173498392105,
+      "learning_rate": 8.501628664495114e-05,
+      "loss": 0.8912,
+      "step": 1018
+    },
+    {
+      "epoch": 2.2383305875892368,
+      "grad_norm": 0.12357274442911148,
+      "learning_rate": 8.477198697068403e-05,
+      "loss": 0.8259,
+      "step": 1019
+    },
+    {
+      "epoch": 2.2405271828665567,
+      "grad_norm": 0.1457071751356125,
+      "learning_rate": 8.452768729641694e-05,
+      "loss": 0.7956,
+      "step": 1020
+    },
+    {
+      "epoch": 2.242723778143877,
+      "grad_norm": 0.2081819474697113,
+      "learning_rate": 8.428338762214984e-05,
+      "loss": 0.6719,
+      "step": 1021
+    },
+    {
+      "epoch": 2.244920373421197,
+      "grad_norm": 0.17431728541851044,
+      "learning_rate": 8.403908794788273e-05,
+      "loss": 0.5949,
+      "step": 1022
+    },
+    {
+      "epoch": 2.247116968698517,
+      "grad_norm": 0.11054105311632156,
+      "learning_rate": 8.379478827361563e-05,
+      "loss": 0.7999,
+      "step": 1023
+    },
+    {
+      "epoch": 2.2493135639758375,
+      "grad_norm": 0.1071142703294754,
+      "learning_rate": 8.355048859934853e-05,
+      "loss": 0.6982,
+      "step": 1024
+    },
+    {
+      "epoch": 2.2515101592531575,
+      "grad_norm": 0.07387082278728485,
+      "learning_rate": 8.330618892508144e-05,
+      "loss": 0.6706,
+      "step": 1025
+    },
+    {
+      "epoch": 2.253706754530478,
+      "grad_norm": 0.08180315792560577,
+      "learning_rate": 8.306188925081432e-05,
+      "loss": 0.6461,
+      "step": 1026
+    },
+    {
+      "epoch": 2.255903349807798,
+      "grad_norm": 0.11591280996799469,
+      "learning_rate": 8.281758957654723e-05,
+      "loss": 0.7579,
+      "step": 1027
+    },
+    {
+      "epoch": 2.2580999450851182,
+      "grad_norm": 0.06702913343906403,
+      "learning_rate": 8.257328990228013e-05,
+      "loss": 0.8214,
+      "step": 1028
+    },
+    {
+      "epoch": 2.260296540362438,
+      "grad_norm": 0.11246707290410995,
+      "learning_rate": 8.232899022801302e-05,
+      "loss": 0.8046,
+      "step": 1029
+    },
+    {
+      "epoch": 2.262493135639758,
+      "grad_norm": 0.11286265403032303,
+      "learning_rate": 8.208469055374592e-05,
+      "loss": 0.6319,
+      "step": 1030
+    },
+    {
+      "epoch": 2.2646897309170786,
+      "grad_norm": 0.06917431205511093,
+      "learning_rate": 8.184039087947882e-05,
+      "loss": 0.6005,
+      "step": 1031
+    },
+    {
+      "epoch": 2.2668863261943986,
+      "grad_norm": 0.1122150719165802,
+      "learning_rate": 8.159609120521173e-05,
+      "loss": 0.6358,
+      "step": 1032
+    },
+    {
+      "epoch": 2.269082921471719,
+      "grad_norm": 0.10798628628253937,
+      "learning_rate": 8.135179153094462e-05,
+      "loss": 0.9118,
+      "step": 1033
+    },
+    {
+      "epoch": 2.271279516749039,
+      "grad_norm": 0.08513644337654114,
+      "learning_rate": 8.110749185667752e-05,
+      "loss": 0.6699,
+      "step": 1034
+    },
+    {
+      "epoch": 2.2734761120263594,
+      "grad_norm": 0.07311940938234329,
+      "learning_rate": 8.086319218241042e-05,
+      "loss": 0.6237,
+      "step": 1035
+    },
+    {
+      "epoch": 2.2756727073036793,
+      "grad_norm": 0.07969298958778381,
+      "learning_rate": 8.061889250814331e-05,
+      "loss": 0.6725,
+      "step": 1036
+    },
+    {
+      "epoch": 2.2778693025809993,
+      "grad_norm": 0.07553510367870331,
+      "learning_rate": 8.037459283387621e-05,
+      "loss": 0.6514,
+      "step": 1037
+    },
+    {
+      "epoch": 2.2800658978583197,
+      "grad_norm": 0.07244187593460083,
+      "learning_rate": 8.013029315960912e-05,
+      "loss": 0.6585,
+      "step": 1038
+    },
+    {
+      "epoch": 2.2822624931356397,
+      "grad_norm": 0.10740841925144196,
+      "learning_rate": 7.988599348534202e-05,
+      "loss": 0.6629,
+      "step": 1039
+    },
+    {
+      "epoch": 2.28445908841296,
+      "grad_norm": 0.09354966133832932,
+      "learning_rate": 7.964169381107491e-05,
+      "loss": 0.7355,
+      "step": 1040
+    },
+    {
+      "epoch": 2.28665568369028,
+      "grad_norm": 0.11982624977827072,
+      "learning_rate": 7.939739413680781e-05,
+      "loss": 0.7245,
+      "step": 1041
+    },
+    {
+      "epoch": 2.2888522789676005,
+      "grad_norm": 0.09988179802894592,
+      "learning_rate": 7.915309446254071e-05,
+      "loss": 0.8289,
+      "step": 1042
+    },
+    {
+      "epoch": 2.2910488742449204,
+      "grad_norm": 0.07293007522821426,
+      "learning_rate": 7.890879478827362e-05,
+      "loss": 0.623,
+      "step": 1043
+    },
+    {
+      "epoch": 2.2932454695222404,
+      "grad_norm": 0.11974676698446274,
+      "learning_rate": 7.86644951140065e-05,
+      "loss": 0.7355,
+      "step": 1044
+    },
+    {
+      "epoch": 2.295442064799561,
+      "grad_norm": 0.07204985618591309,
+      "learning_rate": 7.842019543973941e-05,
+      "loss": 0.779,
+      "step": 1045
+    },
+    {
+      "epoch": 2.297638660076881,
+      "grad_norm": 0.08225905150175095,
+      "learning_rate": 7.817589576547231e-05,
+      "loss": 0.6771,
+      "step": 1046
+    },
+    {
+      "epoch": 2.2998352553542007,
+      "grad_norm": 0.09456238895654678,
+      "learning_rate": 7.79315960912052e-05,
+      "loss": 0.5962,
+      "step": 1047
+    },
+    {
+      "epoch": 2.302031850631521,
+      "grad_norm": 0.09710314869880676,
+      "learning_rate": 7.76872964169381e-05,
+      "loss": 0.7233,
+      "step": 1048
+    },
+    {
+      "epoch": 2.304228445908841,
+      "grad_norm": 0.1432577669620514,
+      "learning_rate": 7.7442996742671e-05,
+      "loss": 0.7091,
+      "step": 1049
+    },
+    {
+      "epoch": 2.3064250411861615,
+      "grad_norm": 0.07711708545684814,
+      "learning_rate": 7.719869706840391e-05,
+      "loss": 0.7367,
+      "step": 1050
+    },
+    {
+      "epoch": 2.3086216364634815,
+      "grad_norm": 0.10340727120637894,
+      "learning_rate": 7.69543973941368e-05,
+      "loss": 0.5496,
+      "step": 1051
+    },
+    {
+      "epoch": 2.310818231740802,
+      "grad_norm": 0.09252232313156128,
+      "learning_rate": 7.67100977198697e-05,
+      "loss": 0.6343,
+      "step": 1052
+    },
+    {
+      "epoch": 2.313014827018122,
+      "grad_norm": 0.10135426372289658,
+      "learning_rate": 7.64657980456026e-05,
+      "loss": 0.7759,
+      "step": 1053
+    },
+    {
+      "epoch": 2.315211422295442,
+      "grad_norm": 0.08097632974386215,
+      "learning_rate": 7.622149837133549e-05,
+      "loss": 0.6978,
+      "step": 1054
+    },
+    {
+      "epoch": 2.3174080175727623,
+      "grad_norm": 0.09443609416484833,
+      "learning_rate": 7.59771986970684e-05,
+      "loss": 0.918,
+      "step": 1055
+    },
+    {
+      "epoch": 2.3196046128500822,
+      "grad_norm": 0.08654357492923737,
+      "learning_rate": 7.57328990228013e-05,
+      "loss": 0.7045,
+      "step": 1056
+    },
+    {
+      "epoch": 2.3218012081274026,
+      "grad_norm": 0.11668606847524643,
+      "learning_rate": 7.54885993485342e-05,
+      "loss": 0.8875,
+      "step": 1057
+    },
+    {
+      "epoch": 2.3239978034047226,
+      "grad_norm": 0.06344422698020935,
+      "learning_rate": 7.524429967426709e-05,
+      "loss": 0.5193,
+      "step": 1058
+    },
+    {
+      "epoch": 2.326194398682043,
+      "grad_norm": 0.0994749441742897,
+      "learning_rate": 7.5e-05,
+      "loss": 0.7232,
+      "step": 1059
+    },
+    {
+      "epoch": 2.328390993959363,
+      "grad_norm": 0.0724123939871788,
+      "learning_rate": 7.47557003257329e-05,
+      "loss": 0.6661,
+      "step": 1060
+    },
+    {
+      "epoch": 2.330587589236683,
+      "grad_norm": 0.10248494893312454,
+      "learning_rate": 7.451140065146579e-05,
+      "loss": 0.6864,
+      "step": 1061
+    },
+    {
+      "epoch": 2.3327841845140034,
+      "grad_norm": 0.07849308848381042,
+      "learning_rate": 7.426710097719869e-05,
+      "loss": 0.7426,
+      "step": 1062
+    },
+    {
+      "epoch": 2.3349807797913233,
+      "grad_norm": 0.08100052922964096,
+      "learning_rate": 7.402280130293159e-05,
+      "loss": 0.8027,
+      "step": 1063
+    },
+    {
+      "epoch": 2.3371773750686438,
+      "grad_norm": 0.15193840861320496,
+      "learning_rate": 7.37785016286645e-05,
+      "loss": 0.7585,
+      "step": 1064
+    },
+    {
+      "epoch": 2.3393739703459637,
+      "grad_norm": 0.07644088566303253,
+      "learning_rate": 7.353420195439738e-05,
+      "loss": 0.7188,
+      "step": 1065
+    },
+    {
+      "epoch": 2.341570565623284,
+      "grad_norm": 0.08135520666837692,
+      "learning_rate": 7.328990228013029e-05,
+      "loss": 0.8297,
+      "step": 1066
+    },
+    {
+      "epoch": 2.343767160900604,
+      "grad_norm": 0.08497358113527298,
+      "learning_rate": 7.304560260586319e-05,
+      "loss": 0.59,
+      "step": 1067
+    },
+    {
+      "epoch": 2.345963756177924,
+      "grad_norm": 0.09644828736782074,
+      "learning_rate": 7.280130293159608e-05,
+      "loss": 0.7635,
+      "step": 1068
+    },
+    {
+      "epoch": 2.3481603514552445,
+      "grad_norm": 0.09850244224071503,
+      "learning_rate": 7.255700325732898e-05,
+      "loss": 0.8943,
+      "step": 1069
+    },
+    {
+      "epoch": 2.3503569467325645,
+      "grad_norm": 0.0912790596485138,
+      "learning_rate": 7.231270358306188e-05,
+      "loss": 0.6558,
+      "step": 1070
+    },
+    {
+      "epoch": 2.352553542009885,
+      "grad_norm": 0.1028975173830986,
+      "learning_rate": 7.206840390879479e-05,
+      "loss": 0.747,
+      "step": 1071
+    },
+    {
+      "epoch": 2.354750137287205,
+      "grad_norm": 0.08312724530696869,
+      "learning_rate": 7.182410423452768e-05,
+      "loss": 0.5572,
+      "step": 1072
+    },
+    {
+      "epoch": 2.356946732564525,
+      "grad_norm": 0.071965791285038,
+      "learning_rate": 7.157980456026058e-05,
+      "loss": 0.6061,
+      "step": 1073
+    },
+    {
+      "epoch": 2.359143327841845,
+      "grad_norm": 0.06766308099031448,
+      "learning_rate": 7.133550488599348e-05,
+      "loss": 0.6535,
+      "step": 1074
+    },
+    {
+      "epoch": 2.361339923119165,
+      "grad_norm": 0.06982740759849548,
+      "learning_rate": 7.109120521172637e-05,
+      "loss": 0.7883,
+      "step": 1075
+    },
+    {
+      "epoch": 2.3635365183964856,
+      "grad_norm": 0.11612609773874283,
+      "learning_rate": 7.084690553745927e-05,
+      "loss": 0.7679,
+      "step": 1076
+    },
+    {
+      "epoch": 2.3657331136738056,
+      "grad_norm": 0.1121784970164299,
+      "learning_rate": 7.060260586319218e-05,
+      "loss": 0.6992,
+      "step": 1077
+    },
+    {
+      "epoch": 2.3679297089511255,
+      "grad_norm": 0.1298820525407791,
+      "learning_rate": 7.035830618892508e-05,
+      "loss": 0.7869,
+      "step": 1078
+    },
+    {
+      "epoch": 2.370126304228446,
+      "grad_norm": 0.24360968172550201,
+      "learning_rate": 7.011400651465797e-05,
+      "loss": 0.5484,
+      "step": 1079
+    },
+    {
+      "epoch": 2.372322899505766,
+      "grad_norm": 0.1092473566532135,
+      "learning_rate": 6.986970684039087e-05,
+      "loss": 0.7417,
+      "step": 1080
+    },
+    {
+      "epoch": 2.3745194947830863,
+      "grad_norm": 0.10417677462100983,
+      "learning_rate": 6.962540716612377e-05,
+      "loss": 0.973,
+      "step": 1081
+    },
+    {
+      "epoch": 2.3767160900604063,
+      "grad_norm": 0.10667172819375992,
+      "learning_rate": 6.938110749185668e-05,
+      "loss": 0.6229,
+      "step": 1082
+    },
+    {
+      "epoch": 2.3789126853377267,
+      "grad_norm": 0.07256430387496948,
+      "learning_rate": 6.913680781758956e-05,
+      "loss": 0.6805,
+      "step": 1083
+    },
+    {
+      "epoch": 2.3811092806150467,
+      "grad_norm": 0.11583739519119263,
+      "learning_rate": 6.889250814332247e-05,
+      "loss": 0.8689,
+      "step": 1084
+    },
+    {
+      "epoch": 2.3833058758923666,
+      "grad_norm": 0.11317083239555359,
+      "learning_rate": 6.864820846905537e-05,
+      "loss": 0.9263,
+      "step": 1085
+    },
+    {
+      "epoch": 2.385502471169687,
+      "grad_norm": 0.08305750042200089,
+      "learning_rate": 6.840390879478826e-05,
+      "loss": 0.6201,
+      "step": 1086
+    },
+    {
+      "epoch": 2.387699066447007,
+      "grad_norm": 0.10185112059116364,
+      "learning_rate": 6.815960912052116e-05,
+      "loss": 0.9691,
+      "step": 1087
+    },
+    {
+      "epoch": 2.3898956617243274,
+      "grad_norm": 0.12586726248264313,
+      "learning_rate": 6.791530944625406e-05,
+      "loss": 0.6908,
+      "step": 1088
+    },
+    {
+      "epoch": 2.3920922570016474,
+      "grad_norm": 0.1868927925825119,
+      "learning_rate": 6.767100977198697e-05,
+      "loss": 0.8442,
+      "step": 1089
+    },
+    {
+      "epoch": 2.394288852278968,
+      "grad_norm": 0.08846687525510788,
+      "learning_rate": 6.742671009771986e-05,
+      "loss": 0.6895,
+      "step": 1090
+    },
+    {
+      "epoch": 2.396485447556288,
+      "grad_norm": 0.0687156692147255,
+      "learning_rate": 6.718241042345276e-05,
+      "loss": 0.6808,
+      "step": 1091
+    },
+    {
+      "epoch": 2.3986820428336078,
+      "grad_norm": 0.07705960422754288,
+      "learning_rate": 6.693811074918566e-05,
+      "loss": 0.7112,
+      "step": 1092
+    },
+    {
+      "epoch": 2.400878638110928,
+      "grad_norm": 0.07999630272388458,
+      "learning_rate": 6.669381107491855e-05,
+      "loss": 0.7709,
+      "step": 1093
+    },
+    {
+      "epoch": 2.403075233388248,
+      "grad_norm": 0.06903533637523651,
+      "learning_rate": 6.644951140065145e-05,
+      "loss": 0.7791,
+      "step": 1094
+    },
+    {
+      "epoch": 2.4052718286655685,
+      "grad_norm": 0.0748233050107956,
+      "learning_rate": 6.620521172638436e-05,
+      "loss": 0.6487,
+      "step": 1095
+    },
+    {
+      "epoch": 2.4074684239428885,
+      "grad_norm": 0.061351366341114044,
+      "learning_rate": 6.596091205211726e-05,
+      "loss": 0.5872,
+      "step": 1096
+    },
+    {
+      "epoch": 2.409665019220209,
+      "grad_norm": 0.1859298199415207,
+      "learning_rate": 6.571661237785015e-05,
+      "loss": 0.8135,
+      "step": 1097
+    },
+    {
+      "epoch": 2.411861614497529,
+      "grad_norm": 0.060790885239839554,
+      "learning_rate": 6.547231270358307e-05,
+      "loss": 0.7532,
+      "step": 1098
+    },
+    {
+      "epoch": 2.414058209774849,
+      "grad_norm": 0.09259055554866791,
+      "learning_rate": 6.522801302931595e-05,
+      "loss": 0.8146,
+      "step": 1099
+    },
+    {
+      "epoch": 2.4162548050521693,
+      "grad_norm": 0.07617384195327759,
+      "learning_rate": 6.498371335504886e-05,
+      "loss": 0.7443,
+      "step": 1100
+    },
+    {
+      "epoch": 2.4184514003294892,
+      "grad_norm": 0.08016043901443481,
+      "learning_rate": 6.473941368078176e-05,
+      "loss": 0.6453,
+      "step": 1101
+    },
+    {
+      "epoch": 2.420647995606809,
+      "grad_norm": 0.0935506820678711,
+      "learning_rate": 6.449511400651466e-05,
+      "loss": 0.788,
+      "step": 1102
+    },
+    {
+      "epoch": 2.4228445908841296,
+      "grad_norm": 0.09312254190444946,
+      "learning_rate": 6.425081433224755e-05,
+      "loss": 0.729,
+      "step": 1103
+    },
+    {
+      "epoch": 2.4250411861614496,
+      "grad_norm": 0.09432965517044067,
+      "learning_rate": 6.400651465798045e-05,
+      "loss": 0.7913,
+      "step": 1104
+    },
+    {
+      "epoch": 2.42723778143877,
+      "grad_norm": 0.13785262405872345,
+      "learning_rate": 6.376221498371336e-05,
+      "loss": 0.7906,
+      "step": 1105
+    },
+    {
+      "epoch": 2.42943437671609,
+      "grad_norm": 0.07341387867927551,
+      "learning_rate": 6.351791530944625e-05,
+      "loss": 0.6767,
+      "step": 1106
+    },
+    {
+      "epoch": 2.4316309719934104,
+      "grad_norm": 0.16046030819416046,
+      "learning_rate": 6.327361563517915e-05,
+      "loss": 0.7861,
+      "step": 1107
+    },
+    {
+      "epoch": 2.4338275672707304,
+      "grad_norm": 0.08474069088697433,
+      "learning_rate": 6.302931596091205e-05,
+      "loss": 0.8283,
+      "step": 1108
+    },
+    {
+      "epoch": 2.4360241625480503,
+      "grad_norm": 0.12598200142383575,
+      "learning_rate": 6.278501628664495e-05,
+      "loss": 0.7289,
+      "step": 1109
+    },
+    {
+      "epoch": 2.4382207578253707,
+      "grad_norm": 0.08919648826122284,
+      "learning_rate": 6.254071661237784e-05,
+      "loss": 0.7775,
+      "step": 1110
+    },
+    {
+      "epoch": 2.4404173531026907,
+      "grad_norm": 0.08033835142850876,
+      "learning_rate": 6.229641693811075e-05,
+      "loss": 0.6731,
+      "step": 1111
+    },
+    {
+      "epoch": 2.442613948380011,
+      "grad_norm": 0.09380998462438583,
+      "learning_rate": 6.205211726384365e-05,
+      "loss": 0.5055,
+      "step": 1112
+    },
+    {
+      "epoch": 2.444810543657331,
+      "grad_norm": 0.1517176777124405,
+      "learning_rate": 6.180781758957654e-05,
+      "loss": 0.9296,
+      "step": 1113
+    },
+    {
+      "epoch": 2.4470071389346515,
+      "grad_norm": 0.15381579101085663,
+      "learning_rate": 6.156351791530944e-05,
+      "loss": 0.9188,
+      "step": 1114
+    },
+    {
+      "epoch": 2.4492037342119715,
+      "grad_norm": 0.06494362652301788,
+      "learning_rate": 6.131921824104234e-05,
+      "loss": 0.6749,
+      "step": 1115
+    },
+    {
+      "epoch": 2.4514003294892914,
+      "grad_norm": 0.07005719095468521,
+      "learning_rate": 6.107491856677525e-05,
+      "loss": 0.6475,
+      "step": 1116
+    },
+    {
+      "epoch": 2.453596924766612,
+      "grad_norm": 1.1737658977508545,
+      "learning_rate": 6.083061889250814e-05,
+      "loss": 0.6943,
+      "step": 1117
+    },
+    {
+      "epoch": 2.455793520043932,
+      "grad_norm": 0.06896524131298065,
+      "learning_rate": 6.058631921824104e-05,
+      "loss": 0.7299,
+      "step": 1118
+    },
+    {
+      "epoch": 2.4579901153212522,
+      "grad_norm": 0.066778264939785,
+      "learning_rate": 6.0342019543973935e-05,
+      "loss": 0.7815,
+      "step": 1119
+    },
+    {
+      "epoch": 2.460186710598572,
+      "grad_norm": 0.06755838543176651,
+      "learning_rate": 6.009771986970684e-05,
+      "loss": 0.69,
+      "step": 1120
+    },
+    {
+      "epoch": 2.4623833058758926,
+      "grad_norm": 0.07871478796005249,
+      "learning_rate": 5.9853420195439734e-05,
+      "loss": 0.7077,
+      "step": 1121
+    },
+    {
+      "epoch": 2.4645799011532126,
+      "grad_norm": 0.12408053874969482,
+      "learning_rate": 5.9609120521172636e-05,
+      "loss": 1.117,
+      "step": 1122
+    },
+    {
+      "epoch": 2.4667764964305325,
+      "grad_norm": 0.10239767283201218,
+      "learning_rate": 5.936482084690553e-05,
+      "loss": 0.7647,
+      "step": 1123
+    },
+    {
+      "epoch": 2.468973091707853,
+      "grad_norm": 0.09658041596412659,
+      "learning_rate": 5.9120521172638435e-05,
+      "loss": 0.7569,
+      "step": 1124
+    },
+    {
+      "epoch": 2.471169686985173,
+      "grad_norm": 0.0856122151017189,
+      "learning_rate": 5.887622149837133e-05,
+      "loss": 0.6062,
+      "step": 1125
+    },
+    {
+      "epoch": 2.473366282262493,
+      "grad_norm": 0.09949222952127457,
+      "learning_rate": 5.863192182410423e-05,
+      "loss": 0.8446,
+      "step": 1126
+    },
+    {
+      "epoch": 2.4755628775398133,
+      "grad_norm": 0.14583195745944977,
+      "learning_rate": 5.838762214983713e-05,
+      "loss": 0.6832,
+      "step": 1127
+    },
+    {
+      "epoch": 2.4777594728171333,
+      "grad_norm": 0.059801578521728516,
+      "learning_rate": 5.8143322475570026e-05,
+      "loss": 0.7064,
+      "step": 1128
+    },
+    {
+      "epoch": 2.4799560680944537,
+      "grad_norm": 0.07657524198293686,
+      "learning_rate": 5.789902280130293e-05,
+      "loss": 0.8749,
+      "step": 1129
+    },
+    {
+      "epoch": 2.4821526633717736,
+      "grad_norm": 0.07773780822753906,
+      "learning_rate": 5.7654723127035825e-05,
+      "loss": 0.7055,
+      "step": 1130
+    },
+    {
+      "epoch": 2.484349258649094,
+      "grad_norm": 0.06075937673449516,
+      "learning_rate": 5.741042345276873e-05,
+      "loss": 0.7001,
+      "step": 1131
+    },
+    {
+      "epoch": 2.486545853926414,
+      "grad_norm": 0.08369509875774384,
+      "learning_rate": 5.716612377850162e-05,
+      "loss": 0.7417,
+      "step": 1132
+    },
+    {
+      "epoch": 2.488742449203734,
+      "grad_norm": 0.07662319391965866,
+      "learning_rate": 5.6921824104234526e-05,
+      "loss": 0.6176,
+      "step": 1133
+    },
+    {
+      "epoch": 2.4909390444810544,
+      "grad_norm": 0.08347583562135696,
+      "learning_rate": 5.667752442996742e-05,
+      "loss": 0.7381,
+      "step": 1134
+    },
+    {
+      "epoch": 2.4931356397583744,
+      "grad_norm": 0.09336914867162704,
+      "learning_rate": 5.643322475570032e-05,
+      "loss": 0.7521,
+      "step": 1135
+    },
+    {
+      "epoch": 2.495332235035695,
+      "grad_norm": 0.08895058184862137,
+      "learning_rate": 5.618892508143322e-05,
+      "loss": 0.629,
+      "step": 1136
+    },
+    {
+      "epoch": 2.4975288303130148,
+      "grad_norm": 0.06815007328987122,
+      "learning_rate": 5.594462540716612e-05,
+      "loss": 0.5191,
+      "step": 1137
+    },
+    {
+      "epoch": 2.499725425590335,
+      "grad_norm": 0.0947193056344986,
+      "learning_rate": 5.570032573289902e-05,
+      "loss": 0.7442,
+      "step": 1138
+    },
+    {
+      "epoch": 2.501922020867655,
+      "grad_norm": 0.10160721838474274,
+      "learning_rate": 5.5456026058631915e-05,
+      "loss": 0.8665,
+      "step": 1139
+    },
+    {
+      "epoch": 2.504118616144975,
+      "grad_norm": 0.12746530771255493,
+      "learning_rate": 5.521172638436482e-05,
+      "loss": 0.6749,
+      "step": 1140
+    },
+    {
+      "epoch": 2.5063152114222955,
+      "grad_norm": 0.05705619230866432,
+      "learning_rate": 5.4967426710097714e-05,
+      "loss": 0.6109,
+      "step": 1141
+    },
+    {
+      "epoch": 2.5085118066996155,
+      "grad_norm": 0.07044143229722977,
+      "learning_rate": 5.472312703583061e-05,
+      "loss": 0.7611,
+      "step": 1142
+    },
+    {
+      "epoch": 2.510708401976936,
+      "grad_norm": 0.08106867969036102,
+      "learning_rate": 5.447882736156351e-05,
+      "loss": 0.8467,
+      "step": 1143
+    },
+    {
+      "epoch": 2.512904997254256,
+      "grad_norm": 0.06889979541301727,
+      "learning_rate": 5.423452768729641e-05,
+      "loss": 0.6965,
+      "step": 1144
+    },
+    {
+      "epoch": 2.5151015925315763,
+      "grad_norm": 0.08472936600446701,
+      "learning_rate": 5.399022801302931e-05,
+      "loss": 0.6908,
+      "step": 1145
+    },
+    {
+      "epoch": 2.5172981878088962,
+      "grad_norm": 0.09621775895357132,
+      "learning_rate": 5.374592833876221e-05,
+      "loss": 0.7638,
+      "step": 1146
+    },
+    {
+      "epoch": 2.519494783086216,
+      "grad_norm": 0.0974590927362442,
+      "learning_rate": 5.350162866449511e-05,
+      "loss": 0.6906,
+      "step": 1147
+    },
+    {
+      "epoch": 2.5216913783635366,
+      "grad_norm": 0.10275783389806747,
+      "learning_rate": 5.3257328990228006e-05,
+      "loss": 0.6232,
+      "step": 1148
+    },
+    {
+      "epoch": 2.5238879736408566,
+      "grad_norm": 0.0663813129067421,
+      "learning_rate": 5.301302931596091e-05,
+      "loss": 0.7696,
+      "step": 1149
+    },
+    {
+      "epoch": 2.5260845689181766,
+      "grad_norm": 0.07898733019828796,
+      "learning_rate": 5.2768729641693805e-05,
+      "loss": 0.9672,
+      "step": 1150
+    },
+    {
+      "epoch": 2.528281164195497,
+      "grad_norm": 0.0817953422665596,
+      "learning_rate": 5.25244299674267e-05,
+      "loss": 0.7536,
+      "step": 1151
+    },
+    {
+      "epoch": 2.5304777594728174,
+      "grad_norm": 0.09350170940160751,
+      "learning_rate": 5.2280130293159604e-05,
+      "loss": 0.6729,
+      "step": 1152
+    },
+    {
+      "epoch": 2.5326743547501374,
+      "grad_norm": 0.07741988450288773,
+      "learning_rate": 5.20358306188925e-05,
+      "loss": 0.7353,
+      "step": 1153
+    },
+    {
+      "epoch": 2.5348709500274573,
+      "grad_norm": 0.06809122860431671,
+      "learning_rate": 5.17915309446254e-05,
+      "loss": 0.679,
+      "step": 1154
+    },
+    {
+      "epoch": 2.5370675453047777,
+      "grad_norm": 0.06133216619491577,
+      "learning_rate": 5.15472312703583e-05,
+      "loss": 0.6661,
+      "step": 1155
+    },
+    {
+      "epoch": 2.5392641405820977,
+      "grad_norm": 0.2147095799446106,
+      "learning_rate": 5.13029315960912e-05,
+      "loss": 0.8382,
+      "step": 1156
+    },
+    {
+      "epoch": 2.5414607358594177,
+      "grad_norm": 0.13015307486057281,
+      "learning_rate": 5.10586319218241e-05,
+      "loss": 0.8105,
+      "step": 1157
+    },
+    {
+      "epoch": 2.543657331136738,
+      "grad_norm": 0.09758636355400085,
+      "learning_rate": 5.081433224755699e-05,
+      "loss": 0.6436,
+      "step": 1158
+    },
+    {
+      "epoch": 2.545853926414058,
+      "grad_norm": 0.08221545815467834,
+      "learning_rate": 5.0570032573289896e-05,
+      "loss": 0.6255,
+      "step": 1159
+    },
+    {
+      "epoch": 2.5480505216913785,
+      "grad_norm": 0.09081722795963287,
+      "learning_rate": 5.032573289902279e-05,
+      "loss": 0.7054,
+      "step": 1160
+    },
+    {
+      "epoch": 2.5502471169686984,
+      "grad_norm": 0.07875044643878937,
+      "learning_rate": 5.0081433224755695e-05,
+      "loss": 0.61,
+      "step": 1161
+    },
+    {
+      "epoch": 2.552443712246019,
+      "grad_norm": 0.12010793387889862,
+      "learning_rate": 4.983713355048859e-05,
+      "loss": 0.7709,
+      "step": 1162
+    },
+    {
+      "epoch": 2.554640307523339,
+      "grad_norm": 0.07873586565256119,
+      "learning_rate": 4.9592833876221493e-05,
+      "loss": 0.5984,
+      "step": 1163
+    },
+    {
+      "epoch": 2.556836902800659,
+      "grad_norm": 0.08544448018074036,
+      "learning_rate": 4.934853420195439e-05,
+      "loss": 0.7688,
+      "step": 1164
+    },
+    {
+      "epoch": 2.559033498077979,
+      "grad_norm": 0.0724640041589737,
+      "learning_rate": 4.910423452768729e-05,
+      "loss": 0.7725,
+      "step": 1165
+    },
+    {
+      "epoch": 2.561230093355299,
+      "grad_norm": 0.07725770026445389,
+      "learning_rate": 4.885993485342019e-05,
+      "loss": 0.6659,
+      "step": 1166
+    },
+    {
+      "epoch": 2.5634266886326196,
+      "grad_norm": 0.10827425867319107,
+      "learning_rate": 4.8615635179153084e-05,
+      "loss": 0.645,
+      "step": 1167
+    },
+    {
+      "epoch": 2.5656232839099395,
+      "grad_norm": 0.15838967263698578,
+      "learning_rate": 4.837133550488599e-05,
+      "loss": 0.7593,
+      "step": 1168
+    },
+    {
+      "epoch": 2.56781987918726,
+      "grad_norm": 0.11101209372282028,
+      "learning_rate": 4.812703583061888e-05,
+      "loss": 0.7147,
+      "step": 1169
+    },
+    {
+      "epoch": 2.57001647446458,
+      "grad_norm": 0.0748819038271904,
+      "learning_rate": 4.7882736156351786e-05,
+      "loss": 0.6548,
+      "step": 1170
+    },
+    {
+      "epoch": 2.5722130697419,
+      "grad_norm": 0.07367917150259018,
+      "learning_rate": 4.763843648208468e-05,
+      "loss": 0.6068,
+      "step": 1171
+    },
+    {
+      "epoch": 2.5744096650192203,
+      "grad_norm": 0.11548689752817154,
+      "learning_rate": 4.7394136807817584e-05,
+      "loss": 0.7362,
+      "step": 1172
+    },
+    {
+      "epoch": 2.5766062602965403,
+      "grad_norm": 0.07397594302892685,
+      "learning_rate": 4.714983713355048e-05,
+      "loss": 0.663,
+      "step": 1173
+    },
+    {
+      "epoch": 2.5788028555738602,
+      "grad_norm": 0.0910998284816742,
+      "learning_rate": 4.6905537459283376e-05,
+      "loss": 0.7571,
+      "step": 1174
+    },
+    {
+      "epoch": 2.5809994508511807,
+      "grad_norm": 0.0982595831155777,
+      "learning_rate": 4.6661237785016286e-05,
+      "loss": 0.7551,
+      "step": 1175
+    },
+    {
+      "epoch": 2.583196046128501,
+      "grad_norm": 0.08838562667369843,
+      "learning_rate": 4.641693811074919e-05,
+      "loss": 0.6734,
+      "step": 1176
+    },
+    {
+      "epoch": 2.585392641405821,
+      "grad_norm": 0.07994061708450317,
+      "learning_rate": 4.6172638436482085e-05,
+      "loss": 0.854,
+      "step": 1177
+    },
+    {
+      "epoch": 2.587589236683141,
+      "grad_norm": 0.08448197692632675,
+      "learning_rate": 4.592833876221498e-05,
+      "loss": 0.6463,
+      "step": 1178
+    },
+    {
+      "epoch": 2.5897858319604614,
+      "grad_norm": 0.10650692880153656,
+      "learning_rate": 4.568403908794788e-05,
+      "loss": 0.836,
+      "step": 1179
+    },
+    {
+      "epoch": 2.5919824272377814,
+      "grad_norm": 0.08229689300060272,
+      "learning_rate": 4.543973941368078e-05,
+      "loss": 0.6181,
+      "step": 1180
+    },
+    {
+      "epoch": 2.5941790225151014,
+      "grad_norm": 0.100438691675663,
+      "learning_rate": 4.519543973941368e-05,
+      "loss": 0.7967,
+      "step": 1181
+    },
+    {
+      "epoch": 2.5963756177924218,
+      "grad_norm": 0.09973705559968948,
+      "learning_rate": 4.495114006514658e-05,
+      "loss": 0.7559,
+      "step": 1182
+    },
+    {
+      "epoch": 2.598572213069742,
+      "grad_norm": 0.08667883276939392,
+      "learning_rate": 4.470684039087948e-05,
+      "loss": 0.7492,
+      "step": 1183
+    },
+    {
+      "epoch": 2.600768808347062,
+      "grad_norm": 0.07978794723749161,
+      "learning_rate": 4.446254071661238e-05,
+      "loss": 0.7361,
+      "step": 1184
+    },
+    {
+      "epoch": 2.602965403624382,
+      "grad_norm": 0.08310586214065552,
+      "learning_rate": 4.421824104234527e-05,
+      "loss": 0.7744,
+      "step": 1185
+    },
+    {
+      "epoch": 2.6051619989017025,
+      "grad_norm": 0.08208447694778442,
+      "learning_rate": 4.3973941368078175e-05,
+      "loss": 0.7859,
+      "step": 1186
+    },
+    {
+      "epoch": 2.6073585941790225,
+      "grad_norm": 0.08372897654771805,
+      "learning_rate": 4.372964169381107e-05,
+      "loss": 0.6409,
+      "step": 1187
+    },
+    {
+      "epoch": 2.6095551894563425,
+      "grad_norm": 0.07595013827085495,
+      "learning_rate": 4.3485342019543974e-05,
+      "loss": 0.6874,
+      "step": 1188
+    },
+    {
+      "epoch": 2.611751784733663,
+      "grad_norm": 0.11010180413722992,
+      "learning_rate": 4.324104234527687e-05,
+      "loss": 0.6602,
+      "step": 1189
+    },
+    {
+      "epoch": 2.613948380010983,
+      "grad_norm": 0.09158472716808319,
+      "learning_rate": 4.299674267100977e-05,
+      "loss": 0.7412,
+      "step": 1190
+    },
+    {
+      "epoch": 2.6161449752883033,
+      "grad_norm": 0.12209982424974442,
+      "learning_rate": 4.275244299674267e-05,
+      "loss": 0.6639,
+      "step": 1191
+    },
+    {
+      "epoch": 2.618341570565623,
+      "grad_norm": 0.06757701188325882,
+      "learning_rate": 4.250814332247557e-05,
+      "loss": 0.5704,
+      "step": 1192
+    },
+    {
+      "epoch": 2.6205381658429436,
+      "grad_norm": 0.1365954428911209,
+      "learning_rate": 4.226384364820847e-05,
+      "loss": 1.1802,
+      "step": 1193
+    },
+    {
+      "epoch": 2.6227347611202636,
+      "grad_norm": 0.26293429732322693,
+      "learning_rate": 4.2019543973941364e-05,
+      "loss": 0.8853,
+      "step": 1194
+    },
+    {
+      "epoch": 2.6249313563975836,
+      "grad_norm": 0.0646226704120636,
+      "learning_rate": 4.1775244299674266e-05,
+      "loss": 0.4727,
+      "step": 1195
+    },
+    {
+      "epoch": 2.627127951674904,
+      "grad_norm": 0.06954771280288696,
+      "learning_rate": 4.153094462540716e-05,
+      "loss": 0.7119,
+      "step": 1196
+    },
+    {
+      "epoch": 2.629324546952224,
+      "grad_norm": 0.1770591288805008,
+      "learning_rate": 4.1286644951140065e-05,
+      "loss": 0.7805,
+      "step": 1197
+    },
+    {
+      "epoch": 2.6315211422295444,
+      "grad_norm": 0.07363187521696091,
+      "learning_rate": 4.104234527687296e-05,
+      "loss": 0.7478,
+      "step": 1198
+    },
+    {
+      "epoch": 2.6337177375068643,
+      "grad_norm": 0.07933395355939865,
+      "learning_rate": 4.0798045602605864e-05,
+      "loss": 0.5766,
+      "step": 1199
+    },
+    {
+      "epoch": 2.6359143327841847,
+      "grad_norm": 0.09841504693031311,
+      "learning_rate": 4.055374592833876e-05,
+      "loss": 0.6924,
+      "step": 1200
+    },
+    {
+      "epoch": 2.6381109280615047,
+      "grad_norm": 0.0708954706788063,
+      "learning_rate": 4.0309446254071656e-05,
+      "loss": 0.5907,
+      "step": 1201
+    },
+    {
+      "epoch": 2.6403075233388247,
+      "grad_norm": 0.09200724214315414,
+      "learning_rate": 4.006514657980456e-05,
+      "loss": 0.5354,
+      "step": 1202
+    },
+    {
+      "epoch": 2.642504118616145,
+      "grad_norm": 0.09067467600107193,
+      "learning_rate": 3.9820846905537454e-05,
+      "loss": 0.937,
+      "step": 1203
+    },
+    {
+      "epoch": 2.644700713893465,
+      "grad_norm": 0.09036250412464142,
+      "learning_rate": 3.957654723127036e-05,
+      "loss": 0.6882,
+      "step": 1204
+    },
+    {
+      "epoch": 2.646897309170785,
+      "grad_norm": 0.07637454569339752,
+      "learning_rate": 3.933224755700325e-05,
+      "loss": 0.773,
+      "step": 1205
+    },
+    {
+      "epoch": 2.6490939044481054,
+      "grad_norm": 0.08656387031078339,
+      "learning_rate": 3.9087947882736156e-05,
+      "loss": 0.7962,
+      "step": 1206
+    },
+    {
+      "epoch": 2.651290499725426,
+      "grad_norm": 0.08390497416257858,
+      "learning_rate": 3.884364820846905e-05,
+      "loss": 0.7953,
+      "step": 1207
+    },
+    {
+      "epoch": 2.653487095002746,
+      "grad_norm": 0.09827643632888794,
+      "learning_rate": 3.8599348534201955e-05,
+      "loss": 0.7158,
+      "step": 1208
+    },
+    {
+      "epoch": 2.655683690280066,
+      "grad_norm": 0.08118298649787903,
+      "learning_rate": 3.835504885993485e-05,
+      "loss": 0.6013,
+      "step": 1209
+    },
+    {
+      "epoch": 2.657880285557386,
+      "grad_norm": 0.09417647123336792,
+      "learning_rate": 3.8110749185667747e-05,
+      "loss": 0.8925,
+      "step": 1210
+    },
+    {
+      "epoch": 2.660076880834706,
+      "grad_norm": 0.0860648974776268,
+      "learning_rate": 3.786644951140065e-05,
+      "loss": 0.6215,
+      "step": 1211
+    },
+    {
+      "epoch": 2.662273476112026,
+      "grad_norm": 0.1231558695435524,
+      "learning_rate": 3.7622149837133545e-05,
+      "loss": 0.7025,
+      "step": 1212
+    },
+    {
+      "epoch": 2.6644700713893466,
+      "grad_norm": 0.11858075112104416,
+      "learning_rate": 3.737785016286645e-05,
+      "loss": 0.8159,
+      "step": 1213
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 0.13391311466693878,
+      "learning_rate": 3.7133550488599344e-05,
+      "loss": 0.6076,
+      "step": 1214
+    },
+    {
+      "epoch": 2.668863261943987,
+      "grad_norm": 0.06996043771505356,
+      "learning_rate": 3.688925081433225e-05,
+      "loss": 0.7056,
+      "step": 1215
+    },
+    {
+      "epoch": 2.671059857221307,
+      "grad_norm": 0.16371381282806396,
+      "learning_rate": 3.664495114006514e-05,
+      "loss": 0.6911,
+      "step": 1216
+    },
+    {
+      "epoch": 2.6732564524986273,
+      "grad_norm": 0.06642128527164459,
+      "learning_rate": 3.640065146579804e-05,
+      "loss": 0.7789,
+      "step": 1217
+    },
+    {
+      "epoch": 2.6754530477759473,
+      "grad_norm": 0.08541740477085114,
+      "learning_rate": 3.615635179153094e-05,
+      "loss": 0.6973,
+      "step": 1218
+    },
+    {
+      "epoch": 2.6776496430532672,
+      "grad_norm": 0.07507028430700302,
+      "learning_rate": 3.591205211726384e-05,
+      "loss": 0.6663,
+      "step": 1219
+    },
+    {
+      "epoch": 2.6798462383305877,
+      "grad_norm": 0.08675278723239899,
+      "learning_rate": 3.566775244299674e-05,
+      "loss": 0.7055,
+      "step": 1220
+    },
+    {
+      "epoch": 2.6820428336079076,
+      "grad_norm": 0.07669603079557419,
+      "learning_rate": 3.5423452768729636e-05,
+      "loss": 0.8274,
+      "step": 1221
+    },
+    {
+      "epoch": 2.684239428885228,
+      "grad_norm": 0.12103569507598877,
+      "learning_rate": 3.517915309446254e-05,
+      "loss": 0.5332,
+      "step": 1222
+    },
+    {
+      "epoch": 2.686436024162548,
+      "grad_norm": 0.07096056640148163,
+      "learning_rate": 3.4934853420195435e-05,
+      "loss": 0.777,
+      "step": 1223
+    },
+    {
+      "epoch": 2.6886326194398684,
+      "grad_norm": 0.08103061467409134,
+      "learning_rate": 3.469055374592834e-05,
+      "loss": 0.7357,
+      "step": 1224
+    },
+    {
+      "epoch": 2.6908292147171884,
+      "grad_norm": 0.0876220241189003,
+      "learning_rate": 3.4446254071661234e-05,
+      "loss": 0.7664,
+      "step": 1225
+    },
+    {
+      "epoch": 2.6930258099945084,
+      "grad_norm": 0.11073236167430878,
+      "learning_rate": 3.420195439739413e-05,
+      "loss": 0.6229,
+      "step": 1226
+    },
+    {
+      "epoch": 2.6952224052718288,
+      "grad_norm": 0.11172617226839066,
+      "learning_rate": 3.395765472312703e-05,
+      "loss": 0.7816,
+      "step": 1227
+    },
+    {
+      "epoch": 2.6974190005491487,
+      "grad_norm": 0.06658000499010086,
+      "learning_rate": 3.371335504885993e-05,
+      "loss": 0.7035,
+      "step": 1228
+    },
+    {
+      "epoch": 2.6996155958264687,
+      "grad_norm": 0.09230737388134003,
+      "learning_rate": 3.346905537459283e-05,
+      "loss": 0.7769,
+      "step": 1229
+    },
+    {
+      "epoch": 2.701812191103789,
+      "grad_norm": 0.10880575329065323,
+      "learning_rate": 3.322475570032573e-05,
+      "loss": 0.6012,
+      "step": 1230
+    },
+    {
+      "epoch": 2.7040087863811095,
+      "grad_norm": 0.05940977856516838,
+      "learning_rate": 3.298045602605863e-05,
+      "loss": 0.5918,
+      "step": 1231
+    },
+    {
+      "epoch": 2.7062053816584295,
+      "grad_norm": 0.09063726663589478,
+      "learning_rate": 3.273615635179153e-05,
+      "loss": 0.813,
+      "step": 1232
+    },
+    {
+      "epoch": 2.7084019769357495,
+      "grad_norm": 0.08328960835933685,
+      "learning_rate": 3.249185667752443e-05,
+      "loss": 0.7105,
+      "step": 1233
+    },
+    {
+      "epoch": 2.71059857221307,
+      "grad_norm": 0.0767611414194107,
+      "learning_rate": 3.224755700325733e-05,
+      "loss": 0.7456,
+      "step": 1234
+    },
+    {
+      "epoch": 2.71279516749039,
+      "grad_norm": 0.10968983918428421,
+      "learning_rate": 3.200325732899023e-05,
+      "loss": 0.8173,
+      "step": 1235
+    },
+    {
+      "epoch": 2.71499176276771,
+      "grad_norm": 0.0895579606294632,
+      "learning_rate": 3.175895765472312e-05,
+      "loss": 0.8651,
+      "step": 1236
+    },
+    {
+      "epoch": 2.7171883580450302,
+      "grad_norm": 0.09464187175035477,
+      "learning_rate": 3.1514657980456026e-05,
+      "loss": 0.7562,
+      "step": 1237
+    },
+    {
+      "epoch": 2.71938495332235,
+      "grad_norm": 0.09358593076467514,
+      "learning_rate": 3.127035830618892e-05,
+      "loss": 0.736,
+      "step": 1238
+    },
+    {
+      "epoch": 2.7215815485996706,
+      "grad_norm": 0.09757359325885773,
+      "learning_rate": 3.1026058631921825e-05,
+      "loss": 0.7649,
+      "step": 1239
+    },
+    {
+      "epoch": 2.7237781438769906,
+      "grad_norm": 0.1129206046462059,
+      "learning_rate": 3.078175895765472e-05,
+      "loss": 0.6922,
+      "step": 1240
+    },
+    {
+      "epoch": 2.725974739154311,
+      "grad_norm": 0.09597048908472061,
+      "learning_rate": 3.0537459283387624e-05,
+      "loss": 0.5701,
+      "step": 1241
+    },
+    {
+      "epoch": 2.728171334431631,
+      "grad_norm": 0.1015196368098259,
+      "learning_rate": 3.029315960912052e-05,
+      "loss": 0.6635,
+      "step": 1242
+    },
+    {
+      "epoch": 2.730367929708951,
+      "grad_norm": 0.06939162313938141,
+      "learning_rate": 3.004885993485342e-05,
+      "loss": 0.7826,
+      "step": 1243
+    },
+    {
+      "epoch": 2.7325645249862713,
+      "grad_norm": 0.08454617857933044,
+      "learning_rate": 2.9804560260586318e-05,
+      "loss": 0.7339,
+      "step": 1244
+    },
+    {
+      "epoch": 2.7347611202635913,
+      "grad_norm": 0.09126226603984833,
+      "learning_rate": 2.9560260586319218e-05,
+      "loss": 0.7112,
+      "step": 1245
+    },
+    {
+      "epoch": 2.7369577155409117,
+      "grad_norm": 0.36390411853790283,
+      "learning_rate": 2.9315960912052114e-05,
+      "loss": 0.737,
+      "step": 1246
+    },
+    {
+      "epoch": 2.7391543108182317,
+      "grad_norm": 0.09418083727359772,
+      "learning_rate": 2.9071661237785013e-05,
+      "loss": 1.1566,
+      "step": 1247
+    },
+    {
+      "epoch": 2.741350906095552,
+      "grad_norm": 0.08405841886997223,
+      "learning_rate": 2.8827361563517912e-05,
+      "loss": 0.7215,
+      "step": 1248
+    },
+    {
+      "epoch": 2.743547501372872,
+      "grad_norm": 0.21146325767040253,
+      "learning_rate": 2.858306188925081e-05,
+      "loss": 0.8386,
+      "step": 1249
+    },
+    {
+      "epoch": 2.745744096650192,
+      "grad_norm": 0.10545873641967773,
+      "learning_rate": 2.833876221498371e-05,
+      "loss": 0.7685,
+      "step": 1250
+    },
+    {
+      "epoch": 2.7479406919275124,
+      "grad_norm": 0.06281019747257233,
+      "learning_rate": 2.809446254071661e-05,
+      "loss": 0.805,
+      "step": 1251
+    },
+    {
+      "epoch": 2.7501372872048324,
+      "grad_norm": 0.10891386866569519,
+      "learning_rate": 2.785016286644951e-05,
+      "loss": 0.7936,
+      "step": 1252
+    },
+    {
+      "epoch": 2.752333882482153,
+      "grad_norm": 0.09322663396596909,
+      "learning_rate": 2.760586319218241e-05,
+      "loss": 0.6482,
+      "step": 1253
+    },
+    {
+      "epoch": 2.754530477759473,
+      "grad_norm": 0.07085970044136047,
+      "learning_rate": 2.7361563517915305e-05,
+      "loss": 0.646,
+      "step": 1254
+    },
+    {
+      "epoch": 2.756727073036793,
+      "grad_norm": 0.08050304651260376,
+      "learning_rate": 2.7117263843648204e-05,
+      "loss": 0.8296,
+      "step": 1255
+    },
+    {
+      "epoch": 2.758923668314113,
+      "grad_norm": 0.090302474796772,
+      "learning_rate": 2.6872964169381104e-05,
+      "loss": 0.5525,
+      "step": 1256
+    },
+    {
+      "epoch": 2.761120263591433,
+      "grad_norm": 0.0808275043964386,
+      "learning_rate": 2.6628664495114003e-05,
+      "loss": 0.6991,
+      "step": 1257
+    },
+    {
+      "epoch": 2.7633168588687536,
+      "grad_norm": 0.07486400008201599,
+      "learning_rate": 2.6384364820846903e-05,
+      "loss": 0.583,
+      "step": 1258
+    },
+    {
+      "epoch": 2.7655134541460735,
+      "grad_norm": 0.09148278087377548,
+      "learning_rate": 2.6140065146579802e-05,
+      "loss": 0.8658,
+      "step": 1259
+    },
+    {
+      "epoch": 2.7677100494233935,
+      "grad_norm": 0.0961243137717247,
+      "learning_rate": 2.58957654723127e-05,
+      "loss": 0.7831,
+      "step": 1260
+    },
+    {
+      "epoch": 2.769906644700714,
+      "grad_norm": 0.09067551791667938,
+      "learning_rate": 2.56514657980456e-05,
+      "loss": 0.7322,
+      "step": 1261
+    },
+    {
+      "epoch": 2.7721032399780343,
+      "grad_norm": 0.08985911309719086,
+      "learning_rate": 2.5407166123778497e-05,
+      "loss": 0.7431,
+      "step": 1262
+    },
+    {
+      "epoch": 2.7742998352553543,
+      "grad_norm": 0.17587175965309143,
+      "learning_rate": 2.5162866449511396e-05,
+      "loss": 0.8184,
+      "step": 1263
+    },
+    {
+      "epoch": 2.7764964305326743,
+      "grad_norm": 0.09483988583087921,
+      "learning_rate": 2.4918566775244295e-05,
+      "loss": 0.8865,
+      "step": 1264
+    },
+    {
+      "epoch": 2.7786930258099947,
+      "grad_norm": 0.08679413050413132,
+      "learning_rate": 2.4674267100977195e-05,
+      "loss": 0.7801,
+      "step": 1265
+    },
+    {
+      "epoch": 2.7808896210873146,
+      "grad_norm": 0.06548160314559937,
+      "learning_rate": 2.4429967426710094e-05,
+      "loss": 0.6109,
+      "step": 1266
+    },
+    {
+      "epoch": 2.7830862163646346,
+      "grad_norm": 0.07188425213098526,
+      "learning_rate": 2.4185667752442993e-05,
+      "loss": 0.6419,
+      "step": 1267
+    },
+    {
+      "epoch": 2.785282811641955,
+      "grad_norm": 0.0663241371512413,
+      "learning_rate": 2.3941368078175893e-05,
+      "loss": 0.7207,
+      "step": 1268
+    },
+    {
+      "epoch": 2.787479406919275,
+      "grad_norm": 0.09383373707532883,
+      "learning_rate": 2.3697068403908792e-05,
+      "loss": 0.6554,
+      "step": 1269
+    },
+    {
+      "epoch": 2.7896760021965954,
+      "grad_norm": 0.06685321778059006,
+      "learning_rate": 2.3452768729641688e-05,
+      "loss": 0.6631,
+      "step": 1270
+    },
+    {
+      "epoch": 2.7918725974739154,
+      "grad_norm": 0.061118632555007935,
+      "learning_rate": 2.3208469055374594e-05,
+      "loss": 0.6207,
+      "step": 1271
+    },
+    {
+      "epoch": 2.7940691927512358,
+      "grad_norm": 0.09721321612596512,
+      "learning_rate": 2.296416938110749e-05,
+      "loss": 0.7763,
+      "step": 1272
+    },
+    {
+      "epoch": 2.7962657880285557,
+      "grad_norm": 0.06751512736082077,
+      "learning_rate": 2.271986970684039e-05,
+      "loss": 0.7306,
+      "step": 1273
+    },
+    {
+      "epoch": 2.7984623833058757,
+      "grad_norm": 0.08189290016889572,
+      "learning_rate": 2.247557003257329e-05,
+      "loss": 0.8933,
+      "step": 1274
+    },
+    {
+      "epoch": 2.800658978583196,
+      "grad_norm": 0.19471223652362823,
+      "learning_rate": 2.223127035830619e-05,
+      "loss": 0.6853,
+      "step": 1275
+    },
+    {
+      "epoch": 2.802855573860516,
+      "grad_norm": 0.0997140035033226,
+      "learning_rate": 2.1986970684039088e-05,
+      "loss": 0.8508,
+      "step": 1276
+    },
+    {
+      "epoch": 2.8050521691378365,
+      "grad_norm": 0.09638452529907227,
+      "learning_rate": 2.1742671009771987e-05,
+      "loss": 0.9514,
+      "step": 1277
+    },
+    {
+      "epoch": 2.8072487644151565,
+      "grad_norm": 0.08031744509935379,
+      "learning_rate": 2.1498371335504886e-05,
+      "loss": 0.7842,
+      "step": 1278
+    },
+    {
+      "epoch": 2.809445359692477,
+      "grad_norm": 0.09164081513881683,
+      "learning_rate": 2.1254071661237786e-05,
+      "loss": 0.7033,
+      "step": 1279
+    },
+    {
+      "epoch": 2.811641954969797,
+      "grad_norm": 0.057708434760570526,
+      "learning_rate": 2.1009771986970682e-05,
+      "loss": 0.6895,
+      "step": 1280
+    },
+    {
+      "epoch": 2.813838550247117,
+      "grad_norm": 0.10480155050754547,
+      "learning_rate": 2.076547231270358e-05,
+      "loss": 0.8366,
+      "step": 1281
+    },
+    {
+      "epoch": 2.8160351455244372,
+      "grad_norm": 0.13848979771137238,
+      "learning_rate": 2.052117263843648e-05,
+      "loss": 0.7983,
+      "step": 1282
+    },
+    {
+      "epoch": 2.818231740801757,
+      "grad_norm": 0.12757208943367004,
+      "learning_rate": 2.027687296416938e-05,
+      "loss": 0.8602,
+      "step": 1283
+    },
+    {
+      "epoch": 2.820428336079077,
+      "grad_norm": 0.08689507842063904,
+      "learning_rate": 2.003257328990228e-05,
+      "loss": 0.7943,
+      "step": 1284
+    },
+    {
+      "epoch": 2.8226249313563976,
+      "grad_norm": 0.08630877733230591,
+      "learning_rate": 1.978827361563518e-05,
+      "loss": 0.7805,
+      "step": 1285
+    },
+    {
+      "epoch": 2.824821526633718,
+      "grad_norm": 0.11185801774263382,
+      "learning_rate": 1.9543973941368078e-05,
+      "loss": 0.9272,
+      "step": 1286
+    },
+    {
+      "epoch": 2.827018121911038,
+      "grad_norm": 0.07941872626543045,
+      "learning_rate": 1.9299674267100977e-05,
+      "loss": 0.6406,
+      "step": 1287
+    },
+    {
+      "epoch": 2.829214717188358,
+      "grad_norm": 0.1542421579360962,
+      "learning_rate": 1.9055374592833873e-05,
+      "loss": 0.755,
+      "step": 1288
+    },
+    {
+      "epoch": 2.8314113124656783,
+      "grad_norm": 0.08039569109678268,
+      "learning_rate": 1.8811074918566773e-05,
+      "loss": 0.7034,
+      "step": 1289
+    },
+    {
+      "epoch": 2.8336079077429983,
+      "grad_norm": 0.07565472275018692,
+      "learning_rate": 1.8566775244299672e-05,
+      "loss": 0.7902,
+      "step": 1290
+    },
+    {
+      "epoch": 2.8358045030203183,
+      "grad_norm": 0.08182371407747269,
+      "learning_rate": 1.832247557003257e-05,
+      "loss": 0.6197,
+      "step": 1291
+    },
+    {
+      "epoch": 2.8380010982976387,
+      "grad_norm": 0.0733376294374466,
+      "learning_rate": 1.807817589576547e-05,
+      "loss": 0.7109,
+      "step": 1292
+    },
+    {
+      "epoch": 2.8401976935749587,
+      "grad_norm": 0.09757563471794128,
+      "learning_rate": 1.783387622149837e-05,
+      "loss": 0.5695,
+      "step": 1293
+    },
+    {
+      "epoch": 2.842394288852279,
+      "grad_norm": 0.07390463352203369,
+      "learning_rate": 1.758957654723127e-05,
+      "loss": 0.78,
+      "step": 1294
+    },
+    {
+      "epoch": 2.844590884129599,
+      "grad_norm": 0.21063832938671112,
+      "learning_rate": 1.734527687296417e-05,
+      "loss": 0.8195,
+      "step": 1295
+    },
+    {
+      "epoch": 2.8467874794069195,
+      "grad_norm": 0.08275119215250015,
+      "learning_rate": 1.7100977198697065e-05,
+      "loss": 0.8693,
+      "step": 1296
+    },
+    {
+      "epoch": 2.8489840746842394,
+      "grad_norm": 0.07229744642972946,
+      "learning_rate": 1.6856677524429964e-05,
+      "loss": 0.621,
+      "step": 1297
+    },
+    {
+      "epoch": 2.8511806699615594,
+      "grad_norm": 0.09600915759801865,
+      "learning_rate": 1.6612377850162864e-05,
+      "loss": 0.8802,
+      "step": 1298
+    },
+    {
+      "epoch": 2.85337726523888,
+      "grad_norm": 0.15338321030139923,
+      "learning_rate": 1.6368078175895766e-05,
+      "loss": 0.7454,
+      "step": 1299
+    },
+    {
+      "epoch": 2.8555738605161998,
+      "grad_norm": 0.0844864770770073,
+      "learning_rate": 1.6123778501628666e-05,
+      "loss": 0.8933,
+      "step": 1300
+    },
+    {
+      "epoch": 2.85777045579352,
+      "grad_norm": 0.07006143778562546,
+      "learning_rate": 1.587947882736156e-05,
+      "loss": 0.7092,
+      "step": 1301
+    },
+    {
+      "epoch": 2.85996705107084,
+      "grad_norm": 0.07498584687709808,
+      "learning_rate": 1.563517915309446e-05,
+      "loss": 0.6378,
+      "step": 1302
+    },
+    {
+      "epoch": 2.8621636463481606,
+      "grad_norm": 0.08135469257831573,
+      "learning_rate": 1.539087947882736e-05,
+      "loss": 0.9453,
+      "step": 1303
+    },
+    {
+      "epoch": 2.8643602416254805,
+      "grad_norm": 0.12629877030849457,
+      "learning_rate": 1.514657980456026e-05,
+      "loss": 0.6331,
+      "step": 1304
+    },
+    {
+      "epoch": 2.8665568369028005,
+      "grad_norm": 0.141535222530365,
+      "learning_rate": 1.4902280130293159e-05,
+      "loss": 0.7361,
+      "step": 1305
+    },
+    {
+      "epoch": 2.868753432180121,
+      "grad_norm": 0.0676809698343277,
+      "learning_rate": 1.4657980456026057e-05,
+      "loss": 0.7188,
+      "step": 1306
+    },
+    {
+      "epoch": 2.870950027457441,
+      "grad_norm": 0.07139524072408676,
+      "learning_rate": 1.4413680781758956e-05,
+      "loss": 0.7826,
+      "step": 1307
+    },
+    {
+      "epoch": 2.873146622734761,
+      "grad_norm": 0.07785090804100037,
+      "learning_rate": 1.4169381107491856e-05,
+      "loss": 0.5876,
+      "step": 1308
+    },
+    {
+      "epoch": 2.8753432180120813,
+      "grad_norm": 0.08373348414897919,
+      "learning_rate": 1.3925081433224755e-05,
+      "loss": 0.6802,
+      "step": 1309
+    },
+    {
+      "epoch": 2.8775398132894017,
+      "grad_norm": 0.14662790298461914,
+      "learning_rate": 1.3680781758957653e-05,
+      "loss": 0.8001,
+      "step": 1310
+    },
+    {
+      "epoch": 2.8797364085667216,
+      "grad_norm": 0.13497252762317657,
+      "learning_rate": 1.3436482084690552e-05,
+      "loss": 0.689,
+      "step": 1311
+    },
+    {
+      "epoch": 2.8819330038440416,
+      "grad_norm": 0.08835636079311371,
+      "learning_rate": 1.3192182410423451e-05,
+      "loss": 0.7591,
+      "step": 1312
+    },
+    {
+      "epoch": 2.884129599121362,
+      "grad_norm": 0.0955505296587944,
+      "learning_rate": 1.294788273615635e-05,
+      "loss": 0.7506,
+      "step": 1313
+    },
+    {
+      "epoch": 2.886326194398682,
+      "grad_norm": 0.08125445246696472,
+      "learning_rate": 1.2703583061889248e-05,
+      "loss": 0.7399,
+      "step": 1314
+    },
+    {
+      "epoch": 2.888522789676002,
+      "grad_norm": 0.0929594412446022,
+      "learning_rate": 1.2459283387622148e-05,
+      "loss": 0.649,
+      "step": 1315
+    },
+    {
+      "epoch": 2.8907193849533224,
+      "grad_norm": 0.08807020634412766,
+      "learning_rate": 1.2214983713355047e-05,
+      "loss": 0.7452,
+      "step": 1316
+    },
+    {
+      "epoch": 2.892915980230643,
+      "grad_norm": 0.13461244106292725,
+      "learning_rate": 1.1970684039087946e-05,
+      "loss": 0.6719,
+      "step": 1317
+    },
+    {
+      "epoch": 2.8951125755079627,
+      "grad_norm": 0.13088251650333405,
+      "learning_rate": 1.1726384364820844e-05,
+      "loss": 0.5588,
+      "step": 1318
+    },
+    {
+      "epoch": 2.8973091707852827,
+      "grad_norm": 0.1078072115778923,
+      "learning_rate": 1.1482084690553745e-05,
+      "loss": 0.5681,
+      "step": 1319
+    },
+    {
+      "epoch": 2.899505766062603,
+      "grad_norm": 0.1157631129026413,
+      "learning_rate": 1.1237785016286644e-05,
+      "loss": 0.7816,
+      "step": 1320
+    },
+    {
+      "epoch": 2.901702361339923,
+      "grad_norm": 0.10442047566175461,
+      "learning_rate": 1.0993485342019544e-05,
+      "loss": 0.5531,
+      "step": 1321
+    },
+    {
+      "epoch": 2.903898956617243,
+      "grad_norm": 0.10045961290597916,
+      "learning_rate": 1.0749185667752443e-05,
+      "loss": 0.744,
+      "step": 1322
+    },
+    {
+      "epoch": 2.9060955518945635,
+      "grad_norm": 0.08938507735729218,
+      "learning_rate": 1.0504885993485341e-05,
+      "loss": 0.6483,
+      "step": 1323
+    },
+    {
+      "epoch": 2.9082921471718834,
+      "grad_norm": 0.08280451595783234,
+      "learning_rate": 1.026058631921824e-05,
+      "loss": 0.7153,
+      "step": 1324
+    },
+    {
+      "epoch": 2.910488742449204,
+      "grad_norm": 0.06957967579364777,
+      "learning_rate": 1.001628664495114e-05,
+      "loss": 0.6086,
+      "step": 1325
+    },
+    {
+      "epoch": 2.912685337726524,
+      "grad_norm": 0.09963395446538925,
+      "learning_rate": 9.771986970684039e-06,
+      "loss": 0.7523,
+      "step": 1326
+    },
+    {
+      "epoch": 2.9148819330038442,
+      "grad_norm": 0.07673919945955276,
+      "learning_rate": 9.527687296416937e-06,
+      "loss": 0.7272,
+      "step": 1327
+    },
+    {
+      "epoch": 2.917078528281164,
+      "grad_norm": 0.2280912846326828,
+      "learning_rate": 9.283387622149836e-06,
+      "loss": 0.5491,
+      "step": 1328
+    },
+    {
+      "epoch": 2.919275123558484,
+      "grad_norm": 0.07022042572498322,
+      "learning_rate": 9.039087947882735e-06,
+      "loss": 0.5251,
+      "step": 1329
+    },
+    {
+      "epoch": 2.9214717188358046,
+      "grad_norm": 0.09710218012332916,
+      "learning_rate": 8.794788273615635e-06,
+      "loss": 0.6799,
+      "step": 1330
+    },
+    {
+      "epoch": 2.9236683141131246,
+      "grad_norm": 0.09736752510070801,
+      "learning_rate": 8.550488599348532e-06,
+      "loss": 0.8457,
+      "step": 1331
+    },
+    {
+      "epoch": 2.925864909390445,
+      "grad_norm": 0.0751236081123352,
+      "learning_rate": 8.306188925081432e-06,
+      "loss": 0.412,
+      "step": 1332
+    },
+    {
+      "epoch": 2.928061504667765,
+      "grad_norm": 0.09723926335573196,
+      "learning_rate": 8.061889250814333e-06,
+      "loss": 0.6629,
+      "step": 1333
+    },
+    {
+      "epoch": 2.9302580999450853,
+      "grad_norm": 0.06813426315784454,
+      "learning_rate": 7.81758957654723e-06,
+      "loss": 0.6108,
+      "step": 1334
+    },
+    {
+      "epoch": 2.9324546952224053,
+      "grad_norm": 0.08759930729866028,
+      "learning_rate": 7.57328990228013e-06,
+      "loss": 0.5458,
+      "step": 1335
+    },
+    {
+      "epoch": 2.9346512904997253,
+      "grad_norm": 0.1464262753725052,
+      "learning_rate": 7.328990228013028e-06,
+      "loss": 0.6136,
+      "step": 1336
+    },
+    {
+      "epoch": 2.9368478857770457,
+      "grad_norm": 0.08710359781980515,
+      "learning_rate": 7.084690553745928e-06,
+      "loss": 0.5978,
+      "step": 1337
+    },
+    {
+      "epoch": 2.9390444810543657,
+      "grad_norm": 0.07749280333518982,
+      "learning_rate": 6.840390879478826e-06,
+      "loss": 0.6742,
+      "step": 1338
+    },
+    {
+      "epoch": 2.9412410763316856,
+      "grad_norm": 0.0653919130563736,
+      "learning_rate": 6.596091205211726e-06,
+      "loss": 0.6895,
+      "step": 1339
+    },
+    {
+      "epoch": 2.943437671609006,
+      "grad_norm": 0.1575169563293457,
+      "learning_rate": 6.351791530944624e-06,
+      "loss": 0.5104,
+      "step": 1340
+    },
+    {
+      "epoch": 2.9456342668863265,
+      "grad_norm": 0.06528547406196594,
+      "learning_rate": 6.1074918566775235e-06,
+      "loss": 0.6594,
+      "step": 1341
+    },
+    {
+      "epoch": 2.9478308621636464,
+      "grad_norm": 0.08210776746273041,
+      "learning_rate": 5.863192182410422e-06,
+      "loss": 0.7457,
+      "step": 1342
+    },
+    {
+      "epoch": 2.9500274574409664,
+      "grad_norm": 0.09949994087219238,
+      "learning_rate": 5.618892508143322e-06,
+      "loss": 0.8239,
+      "step": 1343
+    },
+    {
+      "epoch": 2.952224052718287,
+      "grad_norm": 0.1490132361650467,
+      "learning_rate": 5.374592833876222e-06,
+      "loss": 0.8141,
+      "step": 1344
+    },
+    {
+      "epoch": 2.9544206479956068,
+      "grad_norm": 0.0738406553864479,
+      "learning_rate": 5.13029315960912e-06,
+      "loss": 0.7638,
+      "step": 1345
+    },
+    {
+      "epoch": 2.9566172432729267,
+      "grad_norm": 0.0978720560669899,
+      "learning_rate": 4.8859934853420195e-06,
+      "loss": 0.6641,
+      "step": 1346
+    },
+    {
+      "epoch": 2.958813838550247,
+      "grad_norm": 0.09276679158210754,
+      "learning_rate": 4.641693811074918e-06,
+      "loss": 0.645,
+      "step": 1347
+    },
+    {
+      "epoch": 2.961010433827567,
+      "grad_norm": 0.09173037856817245,
+      "learning_rate": 4.397394136807817e-06,
+      "loss": 0.6249,
+      "step": 1348
+    },
+    {
+      "epoch": 2.9632070291048875,
+      "grad_norm": 0.09970233589410782,
+      "learning_rate": 4.153094462540716e-06,
+      "loss": 0.9367,
+      "step": 1349
+    },
+    {
+      "epoch": 2.9654036243822075,
+      "grad_norm": 0.09109316766262054,
+      "learning_rate": 3.908794788273615e-06,
+      "loss": 0.6787,
+      "step": 1350
+    },
+    {
+      "epoch": 2.967600219659528,
+      "grad_norm": 0.07046660780906677,
+      "learning_rate": 3.664495114006514e-06,
+      "loss": 0.8386,
+      "step": 1351
+    },
+    {
+      "epoch": 2.969796814936848,
+      "grad_norm": 0.0911504402756691,
+      "learning_rate": 3.420195439739413e-06,
+      "loss": 0.7119,
+      "step": 1352
+    },
+    {
+      "epoch": 2.971993410214168,
+      "grad_norm": 0.07956632971763611,
+      "learning_rate": 3.175895765472312e-06,
+      "loss": 0.7665,
+      "step": 1353
+    },
+    {
+      "epoch": 2.9741900054914883,
+      "grad_norm": 0.09220395237207413,
+      "learning_rate": 2.931596091205211e-06,
+      "loss": 0.8344,
+      "step": 1354
+    },
+    {
+      "epoch": 2.9763866007688082,
+      "grad_norm": 0.09236311912536621,
+      "learning_rate": 2.687296416938111e-06,
+      "loss": 0.6423,
+      "step": 1355
+    },
+    {
+      "epoch": 2.9785831960461286,
+      "grad_norm": 0.11850301176309586,
+      "learning_rate": 2.4429967426710097e-06,
+      "loss": 0.8016,
+      "step": 1356
+    },
+    {
+      "epoch": 2.9807797913234486,
+      "grad_norm": 0.10890530794858932,
+      "learning_rate": 2.1986970684039087e-06,
+      "loss": 0.7553,
+      "step": 1357
+    },
+    {
+      "epoch": 2.982976386600769,
+      "grad_norm": 0.12959206104278564,
+      "learning_rate": 1.9543973941368076e-06,
+      "loss": 0.6865,
+      "step": 1358
+    },
+    {
+      "epoch": 2.985172981878089,
+      "grad_norm": 0.06862134486436844,
+      "learning_rate": 1.7100977198697066e-06,
+      "loss": 0.8421,
+      "step": 1359
+    },
+    {
+      "epoch": 2.987369577155409,
+      "grad_norm": 0.09025304019451141,
+      "learning_rate": 1.4657980456026055e-06,
+      "loss": 0.6394,
+      "step": 1360
+    },
+    {
+      "epoch": 2.9895661724327294,
+      "grad_norm": 0.0613892637193203,
+      "learning_rate": 1.2214983713355049e-06,
+      "loss": 0.7104,
+      "step": 1361
+    },
+    {
+      "epoch": 2.9917627677100493,
+      "grad_norm": 0.13201910257339478,
+      "learning_rate": 9.771986970684038e-07,
+      "loss": 0.8843,
+      "step": 1362
+    },
+    {
+      "epoch": 2.9939593629873693,
+      "grad_norm": 0.09035645425319672,
+      "learning_rate": 7.328990228013028e-07,
+      "loss": 0.6958,
+      "step": 1363
+    },
+    {
+      "epoch": 2.9961559582646897,
+      "grad_norm": 0.06832579523324966,
+      "learning_rate": 4.885993485342019e-07,
+      "loss": 0.7666,
+      "step": 1364
+    },
+    {
+      "epoch": 2.99835255354201,
+      "grad_norm": 0.06906723976135254,
+      "learning_rate": 2.4429967426710095e-07,
+      "loss": 0.7987,
+      "step": 1365
+    },
+    {
+      "epoch": 2.99835255354201,
+      "step": 1365,
+      "total_flos": 1.0363435831347118e+18,
+      "train_loss": 0.7699820332911425,
+      "train_runtime": 4969.1155,
+      "train_samples_per_second": 4.397,
+      "train_steps_per_second": 0.275
+    }
+  ],
+  "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": 1.0363435831347118e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}