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": 0.6867699027061462,
+      "learning_rate": 2.18978102189781e-06,
+      "loss": 3.8094,
+      "step": 1
+    },
+    {
+      "epoch": 0.004393190554640308,
+      "grad_norm": 0.4305570125579834,
+      "learning_rate": 4.37956204379562e-06,
+      "loss": 2.7958,
+      "step": 2
+    },
+    {
+      "epoch": 0.006589785831960461,
+      "grad_norm": 0.3998045325279236,
+      "learning_rate": 6.56934306569343e-06,
+      "loss": 2.1771,
+      "step": 3
+    },
+    {
+      "epoch": 0.008786381109280615,
+      "grad_norm": 0.5009977221488953,
+      "learning_rate": 8.75912408759124e-06,
+      "loss": 2.772,
+      "step": 4
+    },
+    {
+      "epoch": 0.010982976386600769,
+      "grad_norm": 0.5049038529396057,
+      "learning_rate": 1.0948905109489049e-05,
+      "loss": 2.7555,
+      "step": 5
+    },
+    {
+      "epoch": 0.013179571663920923,
+      "grad_norm": 0.4534657299518585,
+      "learning_rate": 1.313868613138686e-05,
+      "loss": 2.6643,
+      "step": 6
+    },
+    {
+      "epoch": 0.015376166941241077,
+      "grad_norm": 8.095786094665527,
+      "learning_rate": 1.532846715328467e-05,
+      "loss": 3.1614,
+      "step": 7
+    },
+    {
+      "epoch": 0.01757276221856123,
+      "grad_norm": 0.4472414255142212,
+      "learning_rate": 1.751824817518248e-05,
+      "loss": 2.441,
+      "step": 8
+    },
+    {
+      "epoch": 0.019769357495881382,
+      "grad_norm": 0.5243439078330994,
+      "learning_rate": 1.9708029197080292e-05,
+      "loss": 2.7972,
+      "step": 9
+    },
+    {
+      "epoch": 0.021965952773201538,
+      "grad_norm": 0.48234865069389343,
+      "learning_rate": 2.1897810218978098e-05,
+      "loss": 2.63,
+      "step": 10
+    },
+    {
+      "epoch": 0.02416254805052169,
+      "grad_norm": 0.45005694031715393,
+      "learning_rate": 2.408759124087591e-05,
+      "loss": 2.6311,
+      "step": 11
+    },
+    {
+      "epoch": 0.026359143327841845,
+      "grad_norm": 0.463713675737381,
+      "learning_rate": 2.627737226277372e-05,
+      "loss": 2.5901,
+      "step": 12
+    },
+    {
+      "epoch": 0.028555738605161998,
+      "grad_norm": 0.5905585289001465,
+      "learning_rate": 2.846715328467153e-05,
+      "loss": 3.014,
+      "step": 13
+    },
+    {
+      "epoch": 0.030752333882482153,
+      "grad_norm": 0.5893012881278992,
+      "learning_rate": 3.065693430656934e-05,
+      "loss": 2.4584,
+      "step": 14
+    },
+    {
+      "epoch": 0.032948929159802305,
+      "grad_norm": 0.5204203128814697,
+      "learning_rate": 3.284671532846715e-05,
+      "loss": 3.1182,
+      "step": 15
+    },
+    {
+      "epoch": 0.03514552443712246,
+      "grad_norm": 0.5976374745368958,
+      "learning_rate": 3.503649635036496e-05,
+      "loss": 2.8069,
+      "step": 16
+    },
+    {
+      "epoch": 0.037342119714442616,
+      "grad_norm": 0.6644695401191711,
+      "learning_rate": 3.722627737226277e-05,
+      "loss": 2.4451,
+      "step": 17
+    },
+    {
+      "epoch": 0.039538714991762765,
+      "grad_norm": 7.019594192504883,
+      "learning_rate": 3.9416058394160584e-05,
+      "loss": 2.4422,
+      "step": 18
+    },
+    {
+      "epoch": 0.04173531026908292,
+      "grad_norm": 0.649336576461792,
+      "learning_rate": 4.160583941605839e-05,
+      "loss": 3.2723,
+      "step": 19
+    },
+    {
+      "epoch": 0.043931905546403076,
+      "grad_norm": 0.41296225786209106,
+      "learning_rate": 4.3795620437956196e-05,
+      "loss": 2.7413,
+      "step": 20
+    },
+    {
+      "epoch": 0.04612850082372323,
+      "grad_norm": 0.9910441040992737,
+      "learning_rate": 4.5985401459854016e-05,
+      "loss": 2.8406,
+      "step": 21
+    },
+    {
+      "epoch": 0.04832509610104338,
+      "grad_norm": 0.8387014865875244,
+      "learning_rate": 4.817518248175182e-05,
+      "loss": 3.8058,
+      "step": 22
+    },
+    {
+      "epoch": 0.050521691378363535,
+      "grad_norm": 0.5711728930473328,
+      "learning_rate": 5.036496350364963e-05,
+      "loss": 2.8981,
+      "step": 23
+    },
+    {
+      "epoch": 0.05271828665568369,
+      "grad_norm": 0.4497218132019043,
+      "learning_rate": 5.255474452554744e-05,
+      "loss": 2.7871,
+      "step": 24
+    },
+    {
+      "epoch": 0.054914881933003847,
+      "grad_norm": 0.8127051591873169,
+      "learning_rate": 5.4744525547445253e-05,
+      "loss": 3.264,
+      "step": 25
+    },
+    {
+      "epoch": 0.057111477210323995,
+      "grad_norm": 0.5347107648849487,
+      "learning_rate": 5.693430656934306e-05,
+      "loss": 2.6335,
+      "step": 26
+    },
+    {
+      "epoch": 0.05930807248764415,
+      "grad_norm": 0.5810989737510681,
+      "learning_rate": 5.912408759124087e-05,
+      "loss": 2.727,
+      "step": 27
+    },
+    {
+      "epoch": 0.061504667764964306,
+      "grad_norm": 0.7369208931922913,
+      "learning_rate": 6.131386861313868e-05,
+      "loss": 2.8204,
+      "step": 28
+    },
+    {
+      "epoch": 0.06370126304228446,
+      "grad_norm": 0.6020238399505615,
+      "learning_rate": 6.350364963503648e-05,
+      "loss": 2.6757,
+      "step": 29
+    },
+    {
+      "epoch": 0.06589785831960461,
+      "grad_norm": 0.8717981576919556,
+      "learning_rate": 6.56934306569343e-05,
+      "loss": 3.3947,
+      "step": 30
+    },
+    {
+      "epoch": 0.06809445359692477,
+      "grad_norm": 0.5067639946937561,
+      "learning_rate": 6.788321167883211e-05,
+      "loss": 2.6693,
+      "step": 31
+    },
+    {
+      "epoch": 0.07029104887424492,
+      "grad_norm": 0.8791515231132507,
+      "learning_rate": 7.007299270072992e-05,
+      "loss": 2.6084,
+      "step": 32
+    },
+    {
+      "epoch": 0.07248764415156507,
+      "grad_norm": 5.650439262390137,
+      "learning_rate": 7.226277372262774e-05,
+      "loss": 3.2021,
+      "step": 33
+    },
+    {
+      "epoch": 0.07468423942888523,
+      "grad_norm": 0.4159945845603943,
+      "learning_rate": 7.445255474452554e-05,
+      "loss": 2.1879,
+      "step": 34
+    },
+    {
+      "epoch": 0.07688083470620538,
+      "grad_norm": 0.5776092410087585,
+      "learning_rate": 7.664233576642336e-05,
+      "loss": 2.9968,
+      "step": 35
+    },
+    {
+      "epoch": 0.07907742998352553,
+      "grad_norm": 0.7122467160224915,
+      "learning_rate": 7.883211678832117e-05,
+      "loss": 2.8563,
+      "step": 36
+    },
+    {
+      "epoch": 0.08127402526084569,
+      "grad_norm": 0.3632274568080902,
+      "learning_rate": 8.102189781021897e-05,
+      "loss": 2.2215,
+      "step": 37
+    },
+    {
+      "epoch": 0.08347062053816584,
+      "grad_norm": 0.4672868847846985,
+      "learning_rate": 8.321167883211678e-05,
+      "loss": 2.1369,
+      "step": 38
+    },
+    {
+      "epoch": 0.085667215815486,
+      "grad_norm": 0.5189034938812256,
+      "learning_rate": 8.540145985401459e-05,
+      "loss": 2.9934,
+      "step": 39
+    },
+    {
+      "epoch": 0.08786381109280615,
+      "grad_norm": 0.4526778757572174,
+      "learning_rate": 8.759124087591239e-05,
+      "loss": 2.1655,
+      "step": 40
+    },
+    {
+      "epoch": 0.0900604063701263,
+      "grad_norm": 0.7136516571044922,
+      "learning_rate": 8.978102189781021e-05,
+      "loss": 3.8504,
+      "step": 41
+    },
+    {
+      "epoch": 0.09225700164744646,
+      "grad_norm": 0.45867013931274414,
+      "learning_rate": 9.197080291970803e-05,
+      "loss": 2.6072,
+      "step": 42
+    },
+    {
+      "epoch": 0.09445359692476661,
+      "grad_norm": 0.636026918888092,
+      "learning_rate": 9.416058394160584e-05,
+      "loss": 2.8934,
+      "step": 43
+    },
+    {
+      "epoch": 0.09665019220208676,
+      "grad_norm": 3.3290586471557617,
+      "learning_rate": 9.635036496350364e-05,
+      "loss": 3.0198,
+      "step": 44
+    },
+    {
+      "epoch": 0.09884678747940692,
+      "grad_norm": 2.8718435764312744,
+      "learning_rate": 9.854014598540145e-05,
+      "loss": 3.0349,
+      "step": 45
+    },
+    {
+      "epoch": 0.10104338275672707,
+      "grad_norm": 0.6129201054573059,
+      "learning_rate": 0.00010072992700729926,
+      "loss": 2.8571,
+      "step": 46
+    },
+    {
+      "epoch": 0.10323997803404723,
+      "grad_norm": 0.5462786555290222,
+      "learning_rate": 0.00010291970802919708,
+      "loss": 2.8023,
+      "step": 47
+    },
+    {
+      "epoch": 0.10543657331136738,
+      "grad_norm": 24.708385467529297,
+      "learning_rate": 0.00010510948905109488,
+      "loss": 3.4579,
+      "step": 48
+    },
+    {
+      "epoch": 0.10763316858868753,
+      "grad_norm": 0.7014257907867432,
+      "learning_rate": 0.00010729927007299269,
+      "loss": 4.0115,
+      "step": 49
+    },
+    {
+      "epoch": 0.10982976386600769,
+      "grad_norm": 0.6989462375640869,
+      "learning_rate": 0.00010948905109489051,
+      "loss": 2.6788,
+      "step": 50
+    },
+    {
+      "epoch": 0.11202635914332784,
+      "grad_norm": 0.4397986829280853,
+      "learning_rate": 0.00011167883211678831,
+      "loss": 2.6642,
+      "step": 51
+    },
+    {
+      "epoch": 0.11422295442064799,
+      "grad_norm": 4.33657693862915,
+      "learning_rate": 0.00011386861313868612,
+      "loss": 2.9438,
+      "step": 52
+    },
+    {
+      "epoch": 0.11641954969796815,
+      "grad_norm": 0.5050129294395447,
+      "learning_rate": 0.00011605839416058394,
+      "loss": 2.7343,
+      "step": 53
+    },
+    {
+      "epoch": 0.1186161449752883,
+      "grad_norm": 5.0200347900390625,
+      "learning_rate": 0.00011824817518248174,
+      "loss": 2.7553,
+      "step": 54
+    },
+    {
+      "epoch": 0.12081274025260846,
+      "grad_norm": 0.47741836309432983,
+      "learning_rate": 0.00012043795620437955,
+      "loss": 2.8653,
+      "step": 55
+    },
+    {
+      "epoch": 0.12300933552992861,
+      "grad_norm": 0.998428225517273,
+      "learning_rate": 0.00012262773722627736,
+      "loss": 2.6805,
+      "step": 56
+    },
+    {
+      "epoch": 0.12520593080724876,
+      "grad_norm": 0.5642988085746765,
+      "learning_rate": 0.00012481751824817516,
+      "loss": 2.5517,
+      "step": 57
+    },
+    {
+      "epoch": 0.12740252608456892,
+      "grad_norm": 0.5111164450645447,
+      "learning_rate": 0.00012700729927007297,
+      "loss": 3.1598,
+      "step": 58
+    },
+    {
+      "epoch": 0.12959912136188906,
+      "grad_norm": 0.49121594429016113,
+      "learning_rate": 0.00012919708029197077,
+      "loss": 2.5436,
+      "step": 59
+    },
+    {
+      "epoch": 0.13179571663920922,
+      "grad_norm": 0.4762064814567566,
+      "learning_rate": 0.0001313868613138686,
+      "loss": 2.7901,
+      "step": 60
+    },
+    {
+      "epoch": 0.13399231191652938,
+      "grad_norm": 0.46534234285354614,
+      "learning_rate": 0.00013357664233576641,
+      "loss": 2.5091,
+      "step": 61
+    },
+    {
+      "epoch": 0.13618890719384955,
+      "grad_norm": 0.6299620866775513,
+      "learning_rate": 0.00013576642335766422,
+      "loss": 2.8292,
+      "step": 62
+    },
+    {
+      "epoch": 0.13838550247116968,
+      "grad_norm": 0.657383382320404,
+      "learning_rate": 0.00013795620437956203,
+      "loss": 2.2749,
+      "step": 63
+    },
+    {
+      "epoch": 0.14058209774848984,
+      "grad_norm": 0.4764951765537262,
+      "learning_rate": 0.00014014598540145983,
+      "loss": 2.3178,
+      "step": 64
+    },
+    {
+      "epoch": 0.14277869302581,
+      "grad_norm": 0.5892747640609741,
+      "learning_rate": 0.00014233576642335764,
+      "loss": 2.9363,
+      "step": 65
+    },
+    {
+      "epoch": 0.14497528830313014,
+      "grad_norm": 0.8434705138206482,
+      "learning_rate": 0.00014452554744525547,
+      "loss": 2.2338,
+      "step": 66
+    },
+    {
+      "epoch": 0.1471718835804503,
+      "grad_norm": 0.594817578792572,
+      "learning_rate": 0.00014671532846715328,
+      "loss": 2.6366,
+      "step": 67
+    },
+    {
+      "epoch": 0.14936847885777046,
+      "grad_norm": 0.985395073890686,
+      "learning_rate": 0.00014890510948905108,
+      "loss": 2.4862,
+      "step": 68
+    },
+    {
+      "epoch": 0.1515650741350906,
+      "grad_norm": 1.0945625305175781,
+      "learning_rate": 0.00015109489051094892,
+      "loss": 2.7649,
+      "step": 69
+    },
+    {
+      "epoch": 0.15376166941241076,
+      "grad_norm": 0.45858457684516907,
+      "learning_rate": 0.00015328467153284672,
+      "loss": 2.473,
+      "step": 70
+    },
+    {
+      "epoch": 0.15595826468973092,
+      "grad_norm": 0.5515007972717285,
+      "learning_rate": 0.00015547445255474453,
+      "loss": 2.8478,
+      "step": 71
+    },
+    {
+      "epoch": 0.15815485996705106,
+      "grad_norm": 1.4018892049789429,
+      "learning_rate": 0.00015766423357664234,
+      "loss": 2.5071,
+      "step": 72
+    },
+    {
+      "epoch": 0.16035145524437122,
+      "grad_norm": 1.6679786443710327,
+      "learning_rate": 0.00015985401459854014,
+      "loss": 2.5936,
+      "step": 73
+    },
+    {
+      "epoch": 0.16254805052169138,
+      "grad_norm": 0.3542213439941406,
+      "learning_rate": 0.00016204379562043795,
+      "loss": 2.3941,
+      "step": 74
+    },
+    {
+      "epoch": 0.16474464579901152,
+      "grad_norm": 2.5965020656585693,
+      "learning_rate": 0.00016423357664233575,
+      "loss": 2.6522,
+      "step": 75
+    },
+    {
+      "epoch": 0.16694124107633168,
+      "grad_norm": 0.4187857210636139,
+      "learning_rate": 0.00016642335766423356,
+      "loss": 2.603,
+      "step": 76
+    },
+    {
+      "epoch": 0.16913783635365184,
+      "grad_norm": 0.5375953912734985,
+      "learning_rate": 0.00016861313868613137,
+      "loss": 2.7434,
+      "step": 77
+    },
+    {
+      "epoch": 0.171334431630972,
+      "grad_norm": 0.5186806917190552,
+      "learning_rate": 0.00017080291970802917,
+      "loss": 1.9731,
+      "step": 78
+    },
+    {
+      "epoch": 0.17353102690829214,
+      "grad_norm": 0.36391544342041016,
+      "learning_rate": 0.00017299270072992698,
+      "loss": 2.2552,
+      "step": 79
+    },
+    {
+      "epoch": 0.1757276221856123,
+      "grad_norm": 0.4667130708694458,
+      "learning_rate": 0.00017518248175182478,
+      "loss": 3.0047,
+      "step": 80
+    },
+    {
+      "epoch": 0.17792421746293247,
+      "grad_norm": 1.1712918281555176,
+      "learning_rate": 0.00017737226277372262,
+      "loss": 2.5515,
+      "step": 81
+    },
+    {
+      "epoch": 0.1801208127402526,
+      "grad_norm": 0.34509119391441345,
+      "learning_rate": 0.00017956204379562042,
+      "loss": 2.158,
+      "step": 82
+    },
+    {
+      "epoch": 0.18231740801757276,
+      "grad_norm": 0.5572453737258911,
+      "learning_rate": 0.00018175182481751826,
+      "loss": 2.6059,
+      "step": 83
+    },
+    {
+      "epoch": 0.18451400329489293,
+      "grad_norm": 0.6319203972816467,
+      "learning_rate": 0.00018394160583941606,
+      "loss": 3.0351,
+      "step": 84
+    },
+    {
+      "epoch": 0.18671059857221306,
+      "grad_norm": 0.40414586663246155,
+      "learning_rate": 0.00018613138686131387,
+      "loss": 2.4334,
+      "step": 85
+    },
+    {
+      "epoch": 0.18890719384953322,
+      "grad_norm": 0.9658668637275696,
+      "learning_rate": 0.00018832116788321167,
+      "loss": 2.4,
+      "step": 86
+    },
+    {
+      "epoch": 0.19110378912685339,
+      "grad_norm": 0.3246234953403473,
+      "learning_rate": 0.00019051094890510948,
+      "loss": 2.186,
+      "step": 87
+    },
+    {
+      "epoch": 0.19330038440417352,
+      "grad_norm": 0.4000606834888458,
+      "learning_rate": 0.0001927007299270073,
+      "loss": 2.4948,
+      "step": 88
+    },
+    {
+      "epoch": 0.19549697968149368,
+      "grad_norm": 0.5539069771766663,
+      "learning_rate": 0.0001948905109489051,
+      "loss": 2.4681,
+      "step": 89
+    },
+    {
+      "epoch": 0.19769357495881384,
+      "grad_norm": 0.49171289801597595,
+      "learning_rate": 0.0001970802919708029,
+      "loss": 2.6894,
+      "step": 90
+    },
+    {
+      "epoch": 0.19989017023613398,
+      "grad_norm": 0.3739488124847412,
+      "learning_rate": 0.0001992700729927007,
+      "loss": 2.3717,
+      "step": 91
+    },
+    {
+      "epoch": 0.20208676551345414,
+      "grad_norm": 0.4669046401977539,
+      "learning_rate": 0.0002014598540145985,
+      "loss": 2.1638,
+      "step": 92
+    },
+    {
+      "epoch": 0.2042833607907743,
+      "grad_norm": 0.4423315227031708,
+      "learning_rate": 0.00020364963503649632,
+      "loss": 2.3953,
+      "step": 93
+    },
+    {
+      "epoch": 0.20647995606809447,
+      "grad_norm": 0.3435637950897217,
+      "learning_rate": 0.00020583941605839415,
+      "loss": 2.5799,
+      "step": 94
+    },
+    {
+      "epoch": 0.2086765513454146,
+      "grad_norm": 0.464108943939209,
+      "learning_rate": 0.00020802919708029196,
+      "loss": 2.7652,
+      "step": 95
+    },
+    {
+      "epoch": 0.21087314662273476,
+      "grad_norm": 0.38587018847465515,
+      "learning_rate": 0.00021021897810218976,
+      "loss": 2.554,
+      "step": 96
+    },
+    {
+      "epoch": 0.21306974190005493,
+      "grad_norm": 0.34112444519996643,
+      "learning_rate": 0.00021240875912408757,
+      "loss": 2.3431,
+      "step": 97
+    },
+    {
+      "epoch": 0.21526633717737506,
+      "grad_norm": 0.6684620380401611,
+      "learning_rate": 0.00021459854014598537,
+      "loss": 3.3252,
+      "step": 98
+    },
+    {
+      "epoch": 0.21746293245469522,
+      "grad_norm": 0.4697650671005249,
+      "learning_rate": 0.0002167883211678832,
+      "loss": 2.8338,
+      "step": 99
+    },
+    {
+      "epoch": 0.21965952773201539,
+      "grad_norm": 0.44773155450820923,
+      "learning_rate": 0.00021897810218978101,
+      "loss": 2.9343,
+      "step": 100
+    },
+    {
+      "epoch": 0.22185612300933552,
+      "grad_norm": 0.401836097240448,
+      "learning_rate": 0.00022116788321167882,
+      "loss": 2.6587,
+      "step": 101
+    },
+    {
+      "epoch": 0.22405271828665568,
+      "grad_norm": 0.4680890738964081,
+      "learning_rate": 0.00022335766423357663,
+      "loss": 2.6951,
+      "step": 102
+    },
+    {
+      "epoch": 0.22624931356397585,
+      "grad_norm": 0.6359214782714844,
+      "learning_rate": 0.00022554744525547443,
+      "loss": 3.0357,
+      "step": 103
+    },
+    {
+      "epoch": 0.22844590884129598,
+      "grad_norm": 0.34047117829322815,
+      "learning_rate": 0.00022773722627737224,
+      "loss": 2.1899,
+      "step": 104
+    },
+    {
+      "epoch": 0.23064250411861614,
+      "grad_norm": 0.42732179164886475,
+      "learning_rate": 0.00022992700729927004,
+      "loss": 2.7913,
+      "step": 105
+    },
+    {
+      "epoch": 0.2328390993959363,
+      "grad_norm": 0.416005939245224,
+      "learning_rate": 0.00023211678832116788,
+      "loss": 2.4601,
+      "step": 106
+    },
+    {
+      "epoch": 0.23503569467325644,
+      "grad_norm": 0.3134744465351105,
+      "learning_rate": 0.00023430656934306568,
+      "loss": 2.1341,
+      "step": 107
+    },
+    {
+      "epoch": 0.2372322899505766,
+      "grad_norm": 0.5194646120071411,
+      "learning_rate": 0.0002364963503649635,
+      "loss": 2.5627,
+      "step": 108
+    },
+    {
+      "epoch": 0.23942888522789676,
+      "grad_norm": 0.5569941401481628,
+      "learning_rate": 0.0002386861313868613,
+      "loss": 3.1268,
+      "step": 109
+    },
+    {
+      "epoch": 0.24162548050521693,
+      "grad_norm": 0.48139044642448425,
+      "learning_rate": 0.0002408759124087591,
+      "loss": 2.4277,
+      "step": 110
+    },
+    {
+      "epoch": 0.24382207578253706,
+      "grad_norm": 0.438280314207077,
+      "learning_rate": 0.0002430656934306569,
+      "loss": 2.3645,
+      "step": 111
+    },
+    {
+      "epoch": 0.24601867105985722,
+      "grad_norm": 0.4018980860710144,
+      "learning_rate": 0.0002452554744525547,
+      "loss": 2.5538,
+      "step": 112
+    },
+    {
+      "epoch": 0.2482152663371774,
+      "grad_norm": 0.3556693196296692,
+      "learning_rate": 0.0002474452554744525,
+      "loss": 2.1232,
+      "step": 113
+    },
+    {
+      "epoch": 0.2504118616144975,
+      "grad_norm": 0.9741861820220947,
+      "learning_rate": 0.0002496350364963503,
+      "loss": 2.6255,
+      "step": 114
+    },
+    {
+      "epoch": 0.25260845689181766,
+      "grad_norm": 0.7723566293716431,
+      "learning_rate": 0.00025182481751824813,
+      "loss": 1.9677,
+      "step": 115
+    },
+    {
+      "epoch": 0.25480505216913785,
+      "grad_norm": 0.3260561525821686,
+      "learning_rate": 0.00025401459854014594,
+      "loss": 2.2573,
+      "step": 116
+    },
+    {
+      "epoch": 0.257001647446458,
+      "grad_norm": 0.3814736008644104,
+      "learning_rate": 0.00025620437956204374,
+      "loss": 2.1968,
+      "step": 117
+    },
+    {
+      "epoch": 0.2591982427237781,
+      "grad_norm": 0.3188099265098572,
+      "learning_rate": 0.00025839416058394155,
+      "loss": 2.076,
+      "step": 118
+    },
+    {
+      "epoch": 0.2613948380010983,
+      "grad_norm": 0.7699891328811646,
+      "learning_rate": 0.0002605839416058394,
+      "loss": 2.9877,
+      "step": 119
+    },
+    {
+      "epoch": 0.26359143327841844,
+      "grad_norm": 0.33105942606925964,
+      "learning_rate": 0.0002627737226277372,
+      "loss": 1.8006,
+      "step": 120
+    },
+    {
+      "epoch": 0.26578802855573863,
+      "grad_norm": 0.7451686859130859,
+      "learning_rate": 0.000264963503649635,
+      "loss": 2.5211,
+      "step": 121
+    },
+    {
+      "epoch": 0.26798462383305877,
+      "grad_norm": 0.5171789526939392,
+      "learning_rate": 0.00026715328467153283,
+      "loss": 2.1813,
+      "step": 122
+    },
+    {
+      "epoch": 0.2701812191103789,
+      "grad_norm": 1.0024170875549316,
+      "learning_rate": 0.00026934306569343063,
+      "loss": 1.8798,
+      "step": 123
+    },
+    {
+      "epoch": 0.2723778143876991,
+      "grad_norm": 0.3116895258426666,
+      "learning_rate": 0.00027153284671532844,
+      "loss": 2.0405,
+      "step": 124
+    },
+    {
+      "epoch": 0.2745744096650192,
+      "grad_norm": 0.4436667263507843,
+      "learning_rate": 0.00027372262773722625,
+      "loss": 1.9623,
+      "step": 125
+    },
+    {
+      "epoch": 0.27677100494233936,
+      "grad_norm": 0.4368109107017517,
+      "learning_rate": 0.00027591240875912405,
+      "loss": 2.2034,
+      "step": 126
+    },
+    {
+      "epoch": 0.27896760021965955,
+      "grad_norm": 4.153882026672363,
+      "learning_rate": 0.00027810218978102186,
+      "loss": 2.5022,
+      "step": 127
+    },
+    {
+      "epoch": 0.2811641954969797,
+      "grad_norm": 6.639591693878174,
+      "learning_rate": 0.00028029197080291966,
+      "loss": 3.0398,
+      "step": 128
+    },
+    {
+      "epoch": 0.2833607907742998,
+      "grad_norm": 0.6586474180221558,
+      "learning_rate": 0.00028248175182481747,
+      "loss": 2.4114,
+      "step": 129
+    },
+    {
+      "epoch": 0.28555738605162,
+      "grad_norm": 0.6947298645973206,
+      "learning_rate": 0.0002846715328467153,
+      "loss": 2.7115,
+      "step": 130
+    },
+    {
+      "epoch": 0.28775398132894014,
+      "grad_norm": 1.2340610027313232,
+      "learning_rate": 0.00028686131386861314,
+      "loss": 2.1231,
+      "step": 131
+    },
+    {
+      "epoch": 0.2899505766062603,
+      "grad_norm": 0.37010684609413147,
+      "learning_rate": 0.00028905109489051094,
+      "loss": 2.4067,
+      "step": 132
+    },
+    {
+      "epoch": 0.29214717188358047,
+      "grad_norm": 0.36012178659439087,
+      "learning_rate": 0.00029124087591240875,
+      "loss": 2.0664,
+      "step": 133
+    },
+    {
+      "epoch": 0.2943437671609006,
+      "grad_norm": 0.7660682201385498,
+      "learning_rate": 0.00029343065693430656,
+      "loss": 1.9893,
+      "step": 134
+    },
+    {
+      "epoch": 0.29654036243822074,
+      "grad_norm": 0.296366423368454,
+      "learning_rate": 0.00029562043795620436,
+      "loss": 2.0618,
+      "step": 135
+    },
+    {
+      "epoch": 0.29873695771554093,
+      "grad_norm": 0.42295897006988525,
+      "learning_rate": 0.00029781021897810217,
+      "loss": 2.6448,
+      "step": 136
+    },
+    {
+      "epoch": 0.30093355299286106,
+      "grad_norm": 0.24549104273319244,
+      "learning_rate": 0.0003,
+      "loss": 1.8014,
+      "step": 137
+    },
+    {
+      "epoch": 0.3031301482701812,
+      "grad_norm": 0.2302277684211731,
+      "learning_rate": 0.00029975570032573286,
+      "loss": 1.8949,
+      "step": 138
+    },
+    {
+      "epoch": 0.3053267435475014,
+      "grad_norm": 0.29832273721694946,
+      "learning_rate": 0.00029951140065146575,
+      "loss": 1.9707,
+      "step": 139
+    },
+    {
+      "epoch": 0.3075233388248215,
+      "grad_norm": 0.3476719856262207,
+      "learning_rate": 0.0002992671009771987,
+      "loss": 2.3126,
+      "step": 140
+    },
+    {
+      "epoch": 0.30971993410214166,
+      "grad_norm": 0.32649660110473633,
+      "learning_rate": 0.0002990228013029316,
+      "loss": 2.3305,
+      "step": 141
+    },
+    {
+      "epoch": 0.31191652937946185,
+      "grad_norm": 0.2894906997680664,
+      "learning_rate": 0.0002987785016286645,
+      "loss": 2.0181,
+      "step": 142
+    },
+    {
+      "epoch": 0.314113124656782,
+      "grad_norm": 0.3129909932613373,
+      "learning_rate": 0.00029853420195439736,
+      "loss": 2.2065,
+      "step": 143
+    },
+    {
+      "epoch": 0.3163097199341021,
+      "grad_norm": 0.5649434328079224,
+      "learning_rate": 0.00029828990228013025,
+      "loss": 2.6106,
+      "step": 144
+    },
+    {
+      "epoch": 0.3185063152114223,
+      "grad_norm": 9.553032875061035,
+      "learning_rate": 0.00029804560260586314,
+      "loss": 2.686,
+      "step": 145
+    },
+    {
+      "epoch": 0.32070291048874244,
+      "grad_norm": 2.0201971530914307,
+      "learning_rate": 0.0002978013029315961,
+      "loss": 1.7872,
+      "step": 146
+    },
+    {
+      "epoch": 0.3228995057660626,
+      "grad_norm": 0.3196505010128021,
+      "learning_rate": 0.000297557003257329,
+      "loss": 1.7785,
+      "step": 147
+    },
+    {
+      "epoch": 0.32509610104338277,
+      "grad_norm": 0.35799074172973633,
+      "learning_rate": 0.00029731270358306186,
+      "loss": 2.244,
+      "step": 148
+    },
+    {
+      "epoch": 0.3272926963207029,
+      "grad_norm": 0.41830316185951233,
+      "learning_rate": 0.00029706840390879475,
+      "loss": 1.9262,
+      "step": 149
+    },
+    {
+      "epoch": 0.32948929159802304,
+      "grad_norm": 0.9637613892555237,
+      "learning_rate": 0.00029682410423452764,
+      "loss": 2.3908,
+      "step": 150
+    },
+    {
+      "epoch": 0.3316858868753432,
+      "grad_norm": 0.4740007817745209,
+      "learning_rate": 0.0002965798045602606,
+      "loss": 2.8223,
+      "step": 151
+    },
+    {
+      "epoch": 0.33388248215266336,
+      "grad_norm": 0.44443896412849426,
+      "learning_rate": 0.0002963355048859935,
+      "loss": 2.7445,
+      "step": 152
+    },
+    {
+      "epoch": 0.33607907742998355,
+      "grad_norm": 0.4331221580505371,
+      "learning_rate": 0.00029609120521172636,
+      "loss": 2.1885,
+      "step": 153
+    },
+    {
+      "epoch": 0.3382756727073037,
+      "grad_norm": 0.28629955649375916,
+      "learning_rate": 0.00029584690553745925,
+      "loss": 1.9067,
+      "step": 154
+    },
+    {
+      "epoch": 0.3404722679846238,
+      "grad_norm": 0.6570953130722046,
+      "learning_rate": 0.00029560260586319214,
+      "loss": 2.2199,
+      "step": 155
+    },
+    {
+      "epoch": 0.342668863261944,
+      "grad_norm": 0.3833053410053253,
+      "learning_rate": 0.00029535830618892503,
+      "loss": 2.4724,
+      "step": 156
+    },
+    {
+      "epoch": 0.34486545853926415,
+      "grad_norm": 0.6594868898391724,
+      "learning_rate": 0.000295114006514658,
+      "loss": 2.6274,
+      "step": 157
+    },
+    {
+      "epoch": 0.3470620538165843,
+      "grad_norm": 0.2988504469394684,
+      "learning_rate": 0.00029486970684039086,
+      "loss": 1.6105,
+      "step": 158
+    },
+    {
+      "epoch": 0.34925864909390447,
+      "grad_norm": 0.268501877784729,
+      "learning_rate": 0.00029462540716612375,
+      "loss": 1.9861,
+      "step": 159
+    },
+    {
+      "epoch": 0.3514552443712246,
+      "grad_norm": 0.33260461688041687,
+      "learning_rate": 0.00029438110749185664,
+      "loss": 2.157,
+      "step": 160
+    },
+    {
+      "epoch": 0.35365183964854474,
+      "grad_norm": 0.5266323089599609,
+      "learning_rate": 0.00029413680781758953,
+      "loss": 2.6445,
+      "step": 161
+    },
+    {
+      "epoch": 0.35584843492586493,
+      "grad_norm": 0.43333858251571655,
+      "learning_rate": 0.0002938925081433224,
+      "loss": 2.6215,
+      "step": 162
+    },
+    {
+      "epoch": 0.35804503020318507,
+      "grad_norm": 0.2660248577594757,
+      "learning_rate": 0.00029364820846905536,
+      "loss": 1.6085,
+      "step": 163
+    },
+    {
+      "epoch": 0.3602416254805052,
+      "grad_norm": 0.31763291358947754,
+      "learning_rate": 0.00029340390879478825,
+      "loss": 1.7981,
+      "step": 164
+    },
+    {
+      "epoch": 0.3624382207578254,
+      "grad_norm": 0.3440178632736206,
+      "learning_rate": 0.00029315960912052114,
+      "loss": 2.1721,
+      "step": 165
+    },
+    {
+      "epoch": 0.3646348160351455,
+      "grad_norm": 0.3001042604446411,
+      "learning_rate": 0.00029291530944625403,
+      "loss": 2.0286,
+      "step": 166
+    },
+    {
+      "epoch": 0.36683141131246566,
+      "grad_norm": 0.3227858543395996,
+      "learning_rate": 0.0002926710097719869,
+      "loss": 1.8412,
+      "step": 167
+    },
+    {
+      "epoch": 0.36902800658978585,
+      "grad_norm": 0.2738138735294342,
+      "learning_rate": 0.00029242671009771986,
+      "loss": 1.9062,
+      "step": 168
+    },
+    {
+      "epoch": 0.371224601867106,
+      "grad_norm": 0.37755200266838074,
+      "learning_rate": 0.00029218241042345275,
+      "loss": 2.3461,
+      "step": 169
+    },
+    {
+      "epoch": 0.3734211971444261,
+      "grad_norm": 0.5196016430854797,
+      "learning_rate": 0.00029193811074918564,
+      "loss": 2.1634,
+      "step": 170
+    },
+    {
+      "epoch": 0.3756177924217463,
+      "grad_norm": 0.6504445672035217,
+      "learning_rate": 0.00029169381107491853,
+      "loss": 2.0548,
+      "step": 171
+    },
+    {
+      "epoch": 0.37781438769906645,
+      "grad_norm": 1.4089365005493164,
+      "learning_rate": 0.0002914495114006514,
+      "loss": 2.2301,
+      "step": 172
+    },
+    {
+      "epoch": 0.3800109829763866,
+      "grad_norm": 0.3544122278690338,
+      "learning_rate": 0.0002912052117263843,
+      "loss": 2.1968,
+      "step": 173
+    },
+    {
+      "epoch": 0.38220757825370677,
+      "grad_norm": 0.2883487939834595,
+      "learning_rate": 0.00029096091205211725,
+      "loss": 2.2449,
+      "step": 174
+    },
+    {
+      "epoch": 0.3844041735310269,
+      "grad_norm": 0.3207280933856964,
+      "learning_rate": 0.00029071661237785014,
+      "loss": 1.8855,
+      "step": 175
+    },
+    {
+      "epoch": 0.38660076880834704,
+      "grad_norm": 0.6008182168006897,
+      "learning_rate": 0.00029047231270358303,
+      "loss": 2.283,
+      "step": 176
+    },
+    {
+      "epoch": 0.38879736408566723,
+      "grad_norm": 0.3999931216239929,
+      "learning_rate": 0.0002902280130293159,
+      "loss": 1.5404,
+      "step": 177
+    },
+    {
+      "epoch": 0.39099395936298736,
+      "grad_norm": 0.41997796297073364,
+      "learning_rate": 0.0002899837133550488,
+      "loss": 2.2081,
+      "step": 178
+    },
+    {
+      "epoch": 0.3931905546403075,
+      "grad_norm": 0.25927889347076416,
+      "learning_rate": 0.00028973941368078175,
+      "loss": 1.8162,
+      "step": 179
+    },
+    {
+      "epoch": 0.3953871499176277,
+      "grad_norm": 0.5086097717285156,
+      "learning_rate": 0.00028949511400651464,
+      "loss": 2.1914,
+      "step": 180
+    },
+    {
+      "epoch": 0.3975837451949478,
+      "grad_norm": 0.4042961299419403,
+      "learning_rate": 0.00028925081433224753,
+      "loss": 1.9328,
+      "step": 181
+    },
+    {
+      "epoch": 0.39978034047226796,
+      "grad_norm": 0.2584432363510132,
+      "learning_rate": 0.0002890065146579804,
+      "loss": 1.6583,
+      "step": 182
+    },
+    {
+      "epoch": 0.40197693574958815,
+      "grad_norm": 0.27398261427879333,
+      "learning_rate": 0.0002887622149837133,
+      "loss": 1.6225,
+      "step": 183
+    },
+    {
+      "epoch": 0.4041735310269083,
+      "grad_norm": 0.3208993673324585,
+      "learning_rate": 0.0002885179153094462,
+      "loss": 2.4086,
+      "step": 184
+    },
+    {
+      "epoch": 0.4063701263042284,
+      "grad_norm": 0.23842789232730865,
+      "learning_rate": 0.00028827361563517914,
+      "loss": 2.0104,
+      "step": 185
+    },
+    {
+      "epoch": 0.4085667215815486,
+      "grad_norm": 4.729008197784424,
+      "learning_rate": 0.00028802931596091203,
+      "loss": 1.8505,
+      "step": 186
+    },
+    {
+      "epoch": 0.41076331685886874,
+      "grad_norm": 0.9481481313705444,
+      "learning_rate": 0.0002877850162866449,
+      "loss": 2.1117,
+      "step": 187
+    },
+    {
+      "epoch": 0.41295991213618893,
+      "grad_norm": 1.2887364625930786,
+      "learning_rate": 0.0002875407166123778,
+      "loss": 2.1484,
+      "step": 188
+    },
+    {
+      "epoch": 0.41515650741350907,
+      "grad_norm": 0.2830604612827301,
+      "learning_rate": 0.0002872964169381107,
+      "loss": 1.8464,
+      "step": 189
+    },
+    {
+      "epoch": 0.4173531026908292,
+      "grad_norm": 0.3094532787799835,
+      "learning_rate": 0.00028705211726384364,
+      "loss": 1.5183,
+      "step": 190
+    },
+    {
+      "epoch": 0.4195496979681494,
+      "grad_norm": 0.655655026435852,
+      "learning_rate": 0.00028680781758957653,
+      "loss": 1.6178,
+      "step": 191
+    },
+    {
+      "epoch": 0.42174629324546953,
+      "grad_norm": 4.939327239990234,
+      "learning_rate": 0.0002865635179153094,
+      "loss": 2.245,
+      "step": 192
+    },
+    {
+      "epoch": 0.42394288852278966,
+      "grad_norm": 0.3241123557090759,
+      "learning_rate": 0.0002863192182410423,
+      "loss": 2.1361,
+      "step": 193
+    },
+    {
+      "epoch": 0.42613948380010985,
+      "grad_norm": 0.40930694341659546,
+      "learning_rate": 0.0002860749185667752,
+      "loss": 2.3224,
+      "step": 194
+    },
+    {
+      "epoch": 0.42833607907743,
+      "grad_norm": 0.43635666370391846,
+      "learning_rate": 0.0002858306188925081,
+      "loss": 2.593,
+      "step": 195
+    },
+    {
+      "epoch": 0.4305326743547501,
+      "grad_norm": 0.962032675743103,
+      "learning_rate": 0.00028558631921824103,
+      "loss": 2.3724,
+      "step": 196
+    },
+    {
+      "epoch": 0.4327292696320703,
+      "grad_norm": 0.32116198539733887,
+      "learning_rate": 0.0002853420195439739,
+      "loss": 2.0827,
+      "step": 197
+    },
+    {
+      "epoch": 0.43492586490939045,
+      "grad_norm": 0.30131450295448303,
+      "learning_rate": 0.0002850977198697068,
+      "loss": 1.6534,
+      "step": 198
+    },
+    {
+      "epoch": 0.4371224601867106,
+      "grad_norm": 0.35621631145477295,
+      "learning_rate": 0.0002848534201954397,
+      "loss": 2.1892,
+      "step": 199
+    },
+    {
+      "epoch": 0.43931905546403077,
+      "grad_norm": 0.23180830478668213,
+      "learning_rate": 0.0002846091205211726,
+      "loss": 2.0461,
+      "step": 200
+    },
+    {
+      "epoch": 0.4415156507413509,
+      "grad_norm": 0.48335835337638855,
+      "learning_rate": 0.0002843648208469055,
+      "loss": 2.1847,
+      "step": 201
+    },
+    {
+      "epoch": 0.44371224601867104,
+      "grad_norm": 3.1937615871429443,
+      "learning_rate": 0.0002841205211726384,
+      "loss": 2.5163,
+      "step": 202
+    },
+    {
+      "epoch": 0.44590884129599123,
+      "grad_norm": 0.9839800000190735,
+      "learning_rate": 0.0002838762214983713,
+      "loss": 2.1027,
+      "step": 203
+    },
+    {
+      "epoch": 0.44810543657331137,
+      "grad_norm": 1.3008617162704468,
+      "learning_rate": 0.0002836319218241042,
+      "loss": 2.3303,
+      "step": 204
+    },
+    {
+      "epoch": 0.4503020318506315,
+      "grad_norm": 0.30448833107948303,
+      "learning_rate": 0.0002833876221498371,
+      "loss": 2.0287,
+      "step": 205
+    },
+    {
+      "epoch": 0.4524986271279517,
+      "grad_norm": 0.575522780418396,
+      "learning_rate": 0.00028314332247557,
+      "loss": 1.9951,
+      "step": 206
+    },
+    {
+      "epoch": 0.4546952224052718,
+      "grad_norm": 3.374805212020874,
+      "learning_rate": 0.0002828990228013029,
+      "loss": 1.9257,
+      "step": 207
+    },
+    {
+      "epoch": 0.45689181768259196,
+      "grad_norm": 0.4268450140953064,
+      "learning_rate": 0.0002826547231270358,
+      "loss": 2.3055,
+      "step": 208
+    },
+    {
+      "epoch": 0.45908841295991215,
+      "grad_norm": 0.7651065587997437,
+      "learning_rate": 0.0002824104234527687,
+      "loss": 2.8775,
+      "step": 209
+    },
+    {
+      "epoch": 0.4612850082372323,
+      "grad_norm": 0.4056231677532196,
+      "learning_rate": 0.0002821661237785016,
+      "loss": 2.5683,
+      "step": 210
+    },
+    {
+      "epoch": 0.4634816035145524,
+      "grad_norm": 0.2436179667711258,
+      "learning_rate": 0.0002819218241042345,
+      "loss": 2.0719,
+      "step": 211
+    },
+    {
+      "epoch": 0.4656781987918726,
+      "grad_norm": 0.3772799074649811,
+      "learning_rate": 0.00028167752442996737,
+      "loss": 2.4026,
+      "step": 212
+    },
+    {
+      "epoch": 0.46787479406919275,
+      "grad_norm": 0.49247053265571594,
+      "learning_rate": 0.0002814332247557003,
+      "loss": 1.6893,
+      "step": 213
+    },
+    {
+      "epoch": 0.4700713893465129,
+      "grad_norm": 0.2046569585800171,
+      "learning_rate": 0.0002811889250814332,
+      "loss": 1.6147,
+      "step": 214
+    },
+    {
+      "epoch": 0.47226798462383307,
+      "grad_norm": 0.4080396890640259,
+      "learning_rate": 0.0002809446254071661,
+      "loss": 2.4344,
+      "step": 215
+    },
+    {
+      "epoch": 0.4744645799011532,
+      "grad_norm": 0.37259960174560547,
+      "learning_rate": 0.000280700325732899,
+      "loss": 2.3353,
+      "step": 216
+    },
+    {
+      "epoch": 0.47666117517847334,
+      "grad_norm": 0.25653886795043945,
+      "learning_rate": 0.00028045602605863187,
+      "loss": 1.7774,
+      "step": 217
+    },
+    {
+      "epoch": 0.47885777045579353,
+      "grad_norm": 0.48242172598838806,
+      "learning_rate": 0.0002802117263843648,
+      "loss": 2.3714,
+      "step": 218
+    },
+    {
+      "epoch": 0.48105436573311366,
+      "grad_norm": 0.8453269004821777,
+      "learning_rate": 0.0002799674267100977,
+      "loss": 2.4137,
+      "step": 219
+    },
+    {
+      "epoch": 0.48325096101043385,
+      "grad_norm": 0.3005157709121704,
+      "learning_rate": 0.0002797231270358306,
+      "loss": 2.0773,
+      "step": 220
+    },
+    {
+      "epoch": 0.485447556287754,
+      "grad_norm": 0.2650744616985321,
+      "learning_rate": 0.0002794788273615635,
+      "loss": 1.6764,
+      "step": 221
+    },
+    {
+      "epoch": 0.4876441515650741,
+      "grad_norm": 0.3024451434612274,
+      "learning_rate": 0.00027923452768729637,
+      "loss": 1.7028,
+      "step": 222
+    },
+    {
+      "epoch": 0.4898407468423943,
+      "grad_norm": 1.952035903930664,
+      "learning_rate": 0.00027899022801302926,
+      "loss": 1.7142,
+      "step": 223
+    },
+    {
+      "epoch": 0.49203734211971445,
+      "grad_norm": 0.45348402857780457,
+      "learning_rate": 0.0002787459283387622,
+      "loss": 2.0974,
+      "step": 224
+    },
+    {
+      "epoch": 0.4942339373970346,
+      "grad_norm": 0.24059057235717773,
+      "learning_rate": 0.0002785016286644951,
+      "loss": 1.3141,
+      "step": 225
+    },
+    {
+      "epoch": 0.4964305326743548,
+      "grad_norm": 0.21207170188426971,
+      "learning_rate": 0.000278257328990228,
+      "loss": 1.9957,
+      "step": 226
+    },
+    {
+      "epoch": 0.4986271279516749,
+      "grad_norm": 0.506279706954956,
+      "learning_rate": 0.00027801302931596087,
+      "loss": 1.6735,
+      "step": 227
+    },
+    {
+      "epoch": 0.500823723228995,
+      "grad_norm": 1.6554611921310425,
+      "learning_rate": 0.00027776872964169376,
+      "loss": 1.754,
+      "step": 228
+    },
+    {
+      "epoch": 0.5030203185063152,
+      "grad_norm": 0.26447927951812744,
+      "learning_rate": 0.0002775244299674267,
+      "loss": 1.9114,
+      "step": 229
+    },
+    {
+      "epoch": 0.5052169137836353,
+      "grad_norm": 0.317136287689209,
+      "learning_rate": 0.0002772801302931596,
+      "loss": 2.0925,
+      "step": 230
+    },
+    {
+      "epoch": 0.5074135090609555,
+      "grad_norm": 0.3000323474407196,
+      "learning_rate": 0.0002770358306188925,
+      "loss": 1.9522,
+      "step": 231
+    },
+    {
+      "epoch": 0.5096101043382757,
+      "grad_norm": 0.32324615120887756,
+      "learning_rate": 0.00027679153094462537,
+      "loss": 2.0915,
+      "step": 232
+    },
+    {
+      "epoch": 0.5118066996155958,
+      "grad_norm": 0.23328028619289398,
+      "learning_rate": 0.00027654723127035826,
+      "loss": 1.7136,
+      "step": 233
+    },
+    {
+      "epoch": 0.514003294892916,
+      "grad_norm": 1.4883815050125122,
+      "learning_rate": 0.00027630293159609115,
+      "loss": 1.8622,
+      "step": 234
+    },
+    {
+      "epoch": 0.5161998901702362,
+      "grad_norm": 0.25176090002059937,
+      "learning_rate": 0.0002760586319218241,
+      "loss": 1.8101,
+      "step": 235
+    },
+    {
+      "epoch": 0.5183964854475562,
+      "grad_norm": 0.22337031364440918,
+      "learning_rate": 0.000275814332247557,
+      "loss": 1.6828,
+      "step": 236
+    },
+    {
+      "epoch": 0.5205930807248764,
+      "grad_norm": 0.28729498386383057,
+      "learning_rate": 0.00027557003257328987,
+      "loss": 1.9663,
+      "step": 237
+    },
+    {
+      "epoch": 0.5227896760021966,
+      "grad_norm": 0.2963482439517975,
+      "learning_rate": 0.00027532573289902276,
+      "loss": 2.1517,
+      "step": 238
+    },
+    {
+      "epoch": 0.5249862712795168,
+      "grad_norm": 0.29167085886001587,
+      "learning_rate": 0.00027508143322475565,
+      "loss": 1.977,
+      "step": 239
+    },
+    {
+      "epoch": 0.5271828665568369,
+      "grad_norm": 0.9711644649505615,
+      "learning_rate": 0.00027483713355048854,
+      "loss": 2.1192,
+      "step": 240
+    },
+    {
+      "epoch": 0.5293794618341571,
+      "grad_norm": 0.4867478311061859,
+      "learning_rate": 0.0002745928338762215,
+      "loss": 1.8328,
+      "step": 241
+    },
+    {
+      "epoch": 0.5315760571114773,
+      "grad_norm": 0.29490819573402405,
+      "learning_rate": 0.00027434853420195437,
+      "loss": 1.555,
+      "step": 242
+    },
+    {
+      "epoch": 0.5337726523887973,
+      "grad_norm": 0.2698777914047241,
+      "learning_rate": 0.00027410423452768726,
+      "loss": 1.6627,
+      "step": 243
+    },
+    {
+      "epoch": 0.5359692476661175,
+      "grad_norm": 0.2947591543197632,
+      "learning_rate": 0.00027385993485342015,
+      "loss": 1.9527,
+      "step": 244
+    },
+    {
+      "epoch": 0.5381658429434377,
+      "grad_norm": 0.41817960143089294,
+      "learning_rate": 0.00027361563517915304,
+      "loss": 2.633,
+      "step": 245
+    },
+    {
+      "epoch": 0.5403624382207578,
+      "grad_norm": 0.377718448638916,
+      "learning_rate": 0.000273371335504886,
+      "loss": 2.4644,
+      "step": 246
+    },
+    {
+      "epoch": 0.542559033498078,
+      "grad_norm": 0.30354776978492737,
+      "learning_rate": 0.00027312703583061887,
+      "loss": 1.894,
+      "step": 247
+    },
+    {
+      "epoch": 0.5447556287753982,
+      "grad_norm": 0.2778910994529724,
+      "learning_rate": 0.00027288273615635176,
+      "loss": 1.7081,
+      "step": 248
+    },
+    {
+      "epoch": 0.5469522240527183,
+      "grad_norm": 0.26600056886672974,
+      "learning_rate": 0.00027263843648208465,
+      "loss": 1.7158,
+      "step": 249
+    },
+    {
+      "epoch": 0.5491488193300385,
+      "grad_norm": 0.3457019329071045,
+      "learning_rate": 0.00027239413680781754,
+      "loss": 1.9857,
+      "step": 250
+    },
+    {
+      "epoch": 0.5513454146073586,
+      "grad_norm": 0.3761509954929352,
+      "learning_rate": 0.0002721498371335504,
+      "loss": 2.2323,
+      "step": 251
+    },
+    {
+      "epoch": 0.5535420098846787,
+      "grad_norm": 0.26857590675354004,
+      "learning_rate": 0.00027190553745928337,
+      "loss": 2.0252,
+      "step": 252
+    },
+    {
+      "epoch": 0.5557386051619989,
+      "grad_norm": 0.3179982006549835,
+      "learning_rate": 0.00027166123778501626,
+      "loss": 2.212,
+      "step": 253
+    },
+    {
+      "epoch": 0.5579352004393191,
+      "grad_norm": 0.29780375957489014,
+      "learning_rate": 0.00027141693811074915,
+      "loss": 1.6237,
+      "step": 254
+    },
+    {
+      "epoch": 0.5601317957166392,
+      "grad_norm": 0.3720630705356598,
+      "learning_rate": 0.00027117263843648204,
+      "loss": 2.1751,
+      "step": 255
+    },
+    {
+      "epoch": 0.5623283909939594,
+      "grad_norm": 0.4460804760456085,
+      "learning_rate": 0.0002709283387622149,
+      "loss": 2.073,
+      "step": 256
+    },
+    {
+      "epoch": 0.5645249862712796,
+      "grad_norm": 0.23953329026699066,
+      "learning_rate": 0.00027068403908794787,
+      "loss": 1.5364,
+      "step": 257
+    },
+    {
+      "epoch": 0.5667215815485996,
+      "grad_norm": 0.43796446919441223,
+      "learning_rate": 0.00027043973941368076,
+      "loss": 1.8468,
+      "step": 258
+    },
+    {
+      "epoch": 0.5689181768259198,
+      "grad_norm": 0.4220910370349884,
+      "learning_rate": 0.00027019543973941365,
+      "loss": 2.4857,
+      "step": 259
+    },
+    {
+      "epoch": 0.57111477210324,
+      "grad_norm": 0.48905786871910095,
+      "learning_rate": 0.00026995114006514654,
+      "loss": 2.0072,
+      "step": 260
+    },
+    {
+      "epoch": 0.5733113673805601,
+      "grad_norm": 0.3668261170387268,
+      "learning_rate": 0.00026970684039087943,
+      "loss": 2.0333,
+      "step": 261
+    },
+    {
+      "epoch": 0.5755079626578803,
+      "grad_norm": 0.33565598726272583,
+      "learning_rate": 0.0002694625407166123,
+      "loss": 1.8137,
+      "step": 262
+    },
+    {
+      "epoch": 0.5777045579352005,
+      "grad_norm": 0.2737157940864563,
+      "learning_rate": 0.00026921824104234526,
+      "loss": 1.5304,
+      "step": 263
+    },
+    {
+      "epoch": 0.5799011532125206,
+      "grad_norm": 0.4503357410430908,
+      "learning_rate": 0.00026897394136807815,
+      "loss": 2.0004,
+      "step": 264
+    },
+    {
+      "epoch": 0.5820977484898407,
+      "grad_norm": 0.2231324166059494,
+      "learning_rate": 0.00026872964169381104,
+      "loss": 1.9711,
+      "step": 265
+    },
+    {
+      "epoch": 0.5842943437671609,
+      "grad_norm": 0.22645911574363708,
+      "learning_rate": 0.00026848534201954393,
+      "loss": 1.9809,
+      "step": 266
+    },
+    {
+      "epoch": 0.586490939044481,
+      "grad_norm": 0.3728168308734894,
+      "learning_rate": 0.0002682410423452768,
+      "loss": 2.0632,
+      "step": 267
+    },
+    {
+      "epoch": 0.5886875343218012,
+      "grad_norm": 0.22658048570156097,
+      "learning_rate": 0.00026799674267100976,
+      "loss": 1.804,
+      "step": 268
+    },
+    {
+      "epoch": 0.5908841295991214,
+      "grad_norm": 1.1121339797973633,
+      "learning_rate": 0.00026775244299674265,
+      "loss": 2.4445,
+      "step": 269
+    },
+    {
+      "epoch": 0.5930807248764415,
+      "grad_norm": 0.27379271388053894,
+      "learning_rate": 0.00026750814332247554,
+      "loss": 1.976,
+      "step": 270
+    },
+    {
+      "epoch": 0.5952773201537617,
+      "grad_norm": 0.23978272080421448,
+      "learning_rate": 0.00026726384364820843,
+      "loss": 1.5992,
+      "step": 271
+    },
+    {
+      "epoch": 0.5974739154310819,
+      "grad_norm": 0.2422962784767151,
+      "learning_rate": 0.0002670195439739413,
+      "loss": 1.4632,
+      "step": 272
+    },
+    {
+      "epoch": 0.5996705107084019,
+      "grad_norm": 0.42593052983283997,
+      "learning_rate": 0.0002667752442996742,
+      "loss": 1.8835,
+      "step": 273
+    },
+    {
+      "epoch": 0.6018671059857221,
+      "grad_norm": 0.37308213114738464,
+      "learning_rate": 0.00026653094462540715,
+      "loss": 1.9604,
+      "step": 274
+    },
+    {
+      "epoch": 0.6040637012630423,
+      "grad_norm": 0.5239186882972717,
+      "learning_rate": 0.00026628664495114004,
+      "loss": 2.0283,
+      "step": 275
+    },
+    {
+      "epoch": 0.6062602965403624,
+      "grad_norm": 0.27126190066337585,
+      "learning_rate": 0.00026604234527687293,
+      "loss": 1.6114,
+      "step": 276
+    },
+    {
+      "epoch": 0.6084568918176826,
+      "grad_norm": 0.2514207363128662,
+      "learning_rate": 0.0002657980456026058,
+      "loss": 1.8998,
+      "step": 277
+    },
+    {
+      "epoch": 0.6106534870950028,
+      "grad_norm": 0.6194104552268982,
+      "learning_rate": 0.0002655537459283387,
+      "loss": 2.0063,
+      "step": 278
+    },
+    {
+      "epoch": 0.6128500823723229,
+      "grad_norm": 0.20630212128162384,
+      "learning_rate": 0.00026530944625407165,
+      "loss": 1.6659,
+      "step": 279
+    },
+    {
+      "epoch": 0.615046677649643,
+      "grad_norm": 0.31481897830963135,
+      "learning_rate": 0.00026506514657980454,
+      "loss": 2.0444,
+      "step": 280
+    },
+    {
+      "epoch": 0.6172432729269632,
+      "grad_norm": 0.19248200953006744,
+      "learning_rate": 0.00026482084690553743,
+      "loss": 1.3379,
+      "step": 281
+    },
+    {
+      "epoch": 0.6194398682042833,
+      "grad_norm": 0.2798798978328705,
+      "learning_rate": 0.0002645765472312703,
+      "loss": 1.5631,
+      "step": 282
+    },
+    {
+      "epoch": 0.6216364634816035,
+      "grad_norm": 1.7760767936706543,
+      "learning_rate": 0.0002643322475570032,
+      "loss": 1.9501,
+      "step": 283
+    },
+    {
+      "epoch": 0.6238330587589237,
+      "grad_norm": 0.23137585818767548,
+      "learning_rate": 0.0002640879478827361,
+      "loss": 1.5628,
+      "step": 284
+    },
+    {
+      "epoch": 0.6260296540362438,
+      "grad_norm": 0.444173127412796,
+      "learning_rate": 0.00026384364820846904,
+      "loss": 2.2046,
+      "step": 285
+    },
+    {
+      "epoch": 0.628226249313564,
+      "grad_norm": 0.24467839300632477,
+      "learning_rate": 0.00026359934853420193,
+      "loss": 1.5266,
+      "step": 286
+    },
+    {
+      "epoch": 0.6304228445908842,
+      "grad_norm": 1.6656851768493652,
+      "learning_rate": 0.0002633550488599348,
+      "loss": 1.8701,
+      "step": 287
+    },
+    {
+      "epoch": 0.6326194398682042,
+      "grad_norm": 0.3342525064945221,
+      "learning_rate": 0.0002631107491856677,
+      "loss": 1.7194,
+      "step": 288
+    },
+    {
+      "epoch": 0.6348160351455244,
+      "grad_norm": 0.289370596408844,
+      "learning_rate": 0.0002628664495114006,
+      "loss": 1.5249,
+      "step": 289
+    },
+    {
+      "epoch": 0.6370126304228446,
+      "grad_norm": 0.416355699300766,
+      "learning_rate": 0.0002626221498371335,
+      "loss": 2.083,
+      "step": 290
+    },
+    {
+      "epoch": 0.6392092257001647,
+      "grad_norm": 2.3184218406677246,
+      "learning_rate": 0.00026237785016286643,
+      "loss": 2.0558,
+      "step": 291
+    },
+    {
+      "epoch": 0.6414058209774849,
+      "grad_norm": 0.3028144836425781,
+      "learning_rate": 0.0002621335504885993,
+      "loss": 1.6118,
+      "step": 292
+    },
+    {
+      "epoch": 0.6436024162548051,
+      "grad_norm": 0.2321065068244934,
+      "learning_rate": 0.00026188925081433226,
+      "loss": 1.6032,
+      "step": 293
+    },
+    {
+      "epoch": 0.6457990115321252,
+      "grad_norm": 0.36307665705680847,
+      "learning_rate": 0.00026164495114006515,
+      "loss": 1.7608,
+      "step": 294
+    },
+    {
+      "epoch": 0.6479956068094453,
+      "grad_norm": 1.0552006959915161,
+      "learning_rate": 0.00026140065146579804,
+      "loss": 1.7436,
+      "step": 295
+    },
+    {
+      "epoch": 0.6501922020867655,
+      "grad_norm": 0.25893399119377136,
+      "learning_rate": 0.00026115635179153093,
+      "loss": 1.8111,
+      "step": 296
+    },
+    {
+      "epoch": 0.6523887973640856,
+      "grad_norm": 0.3225669264793396,
+      "learning_rate": 0.0002609120521172638,
+      "loss": 1.6261,
+      "step": 297
+    },
+    {
+      "epoch": 0.6545853926414058,
+      "grad_norm": 0.3199833035469055,
+      "learning_rate": 0.00026066775244299676,
+      "loss": 1.7355,
+      "step": 298
+    },
+    {
+      "epoch": 0.656781987918726,
+      "grad_norm": 0.16786855459213257,
+      "learning_rate": 0.00026042345276872965,
+      "loss": 1.424,
+      "step": 299
+    },
+    {
+      "epoch": 0.6589785831960461,
+      "grad_norm": 0.30132219195365906,
+      "learning_rate": 0.00026017915309446254,
+      "loss": 1.572,
+      "step": 300
+    },
+    {
+      "epoch": 0.6611751784733663,
+      "grad_norm": 0.29592394828796387,
+      "learning_rate": 0.00025993485342019543,
+      "loss": 1.9169,
+      "step": 301
+    },
+    {
+      "epoch": 0.6633717737506865,
+      "grad_norm": 0.2513386309146881,
+      "learning_rate": 0.0002596905537459283,
+      "loss": 1.8102,
+      "step": 302
+    },
+    {
+      "epoch": 0.6655683690280065,
+      "grad_norm": 0.2882305681705475,
+      "learning_rate": 0.0002594462540716612,
+      "loss": 1.649,
+      "step": 303
+    },
+    {
+      "epoch": 0.6677649643053267,
+      "grad_norm": 0.20918822288513184,
+      "learning_rate": 0.00025920195439739415,
+      "loss": 1.7569,
+      "step": 304
+    },
+    {
+      "epoch": 0.6699615595826469,
+      "grad_norm": 0.36946746706962585,
+      "learning_rate": 0.00025895765472312704,
+      "loss": 1.8692,
+      "step": 305
+    },
+    {
+      "epoch": 0.6721581548599671,
+      "grad_norm": 0.3604501187801361,
+      "learning_rate": 0.00025871335504885993,
+      "loss": 1.7578,
+      "step": 306
+    },
+    {
+      "epoch": 0.6743547501372872,
+      "grad_norm": 0.39942219853401184,
+      "learning_rate": 0.0002584690553745928,
+      "loss": 1.6905,
+      "step": 307
+    },
+    {
+      "epoch": 0.6765513454146074,
+      "grad_norm": 0.43109774589538574,
+      "learning_rate": 0.0002582247557003257,
+      "loss": 1.785,
+      "step": 308
+    },
+    {
+      "epoch": 0.6787479406919276,
+      "grad_norm": 0.29537302255630493,
+      "learning_rate": 0.00025798045602605865,
+      "loss": 1.9658,
+      "step": 309
+    },
+    {
+      "epoch": 0.6809445359692476,
+      "grad_norm": 0.23576213419437408,
+      "learning_rate": 0.00025773615635179154,
+      "loss": 1.8133,
+      "step": 310
+    },
+    {
+      "epoch": 0.6831411312465678,
+      "grad_norm": 0.2506450116634369,
+      "learning_rate": 0.00025749185667752443,
+      "loss": 1.3471,
+      "step": 311
+    },
+    {
+      "epoch": 0.685337726523888,
+      "grad_norm": 0.22517234086990356,
+      "learning_rate": 0.0002572475570032573,
+      "loss": 1.7564,
+      "step": 312
+    },
+    {
+      "epoch": 0.6875343218012081,
+      "grad_norm": 0.49102678894996643,
+      "learning_rate": 0.0002570032573289902,
+      "loss": 2.0584,
+      "step": 313
+    },
+    {
+      "epoch": 0.6897309170785283,
+      "grad_norm": 0.38469281792640686,
+      "learning_rate": 0.0002567589576547231,
+      "loss": 1.6551,
+      "step": 314
+    },
+    {
+      "epoch": 0.6919275123558485,
+      "grad_norm": 0.2883450388908386,
+      "learning_rate": 0.00025651465798045604,
+      "loss": 1.22,
+      "step": 315
+    },
+    {
+      "epoch": 0.6941241076331686,
+      "grad_norm": 1.661759614944458,
+      "learning_rate": 0.00025627035830618893,
+      "loss": 2.1856,
+      "step": 316
+    },
+    {
+      "epoch": 0.6963207029104888,
+      "grad_norm": 0.26995956897735596,
+      "learning_rate": 0.0002560260586319218,
+      "loss": 1.8976,
+      "step": 317
+    },
+    {
+      "epoch": 0.6985172981878089,
+      "grad_norm": 0.26696524024009705,
+      "learning_rate": 0.0002557817589576547,
+      "loss": 1.8568,
+      "step": 318
+    },
+    {
+      "epoch": 0.700713893465129,
+      "grad_norm": 0.2914663851261139,
+      "learning_rate": 0.0002555374592833876,
+      "loss": 1.7797,
+      "step": 319
+    },
+    {
+      "epoch": 0.7029104887424492,
+      "grad_norm": 0.2516700029373169,
+      "learning_rate": 0.0002552931596091205,
+      "loss": 1.8716,
+      "step": 320
+    },
+    {
+      "epoch": 0.7051070840197694,
+      "grad_norm": 0.31859683990478516,
+      "learning_rate": 0.00025504885993485343,
+      "loss": 1.6661,
+      "step": 321
+    },
+    {
+      "epoch": 0.7073036792970895,
+      "grad_norm": 0.3403477966785431,
+      "learning_rate": 0.0002548045602605863,
+      "loss": 1.927,
+      "step": 322
+    },
+    {
+      "epoch": 0.7095002745744097,
+      "grad_norm": 0.28619951009750366,
+      "learning_rate": 0.0002545602605863192,
+      "loss": 1.4496,
+      "step": 323
+    },
+    {
+      "epoch": 0.7116968698517299,
+      "grad_norm": 1.439237356185913,
+      "learning_rate": 0.0002543159609120521,
+      "loss": 1.6355,
+      "step": 324
+    },
+    {
+      "epoch": 0.7138934651290499,
+      "grad_norm": 0.30764803290367126,
+      "learning_rate": 0.000254071661237785,
+      "loss": 1.48,
+      "step": 325
+    },
+    {
+      "epoch": 0.7160900604063701,
+      "grad_norm": 0.2986823618412018,
+      "learning_rate": 0.00025382736156351793,
+      "loss": 1.7087,
+      "step": 326
+    },
+    {
+      "epoch": 0.7182866556836903,
+      "grad_norm": 0.35139936208724976,
+      "learning_rate": 0.0002535830618892508,
+      "loss": 2.1898,
+      "step": 327
+    },
+    {
+      "epoch": 0.7204832509610104,
+      "grad_norm": 0.363787978887558,
+      "learning_rate": 0.0002533387622149837,
+      "loss": 2.3478,
+      "step": 328
+    },
+    {
+      "epoch": 0.7226798462383306,
+      "grad_norm": 0.26309481263160706,
+      "learning_rate": 0.0002530944625407166,
+      "loss": 1.8111,
+      "step": 329
+    },
+    {
+      "epoch": 0.7248764415156508,
+      "grad_norm": 0.26883387565612793,
+      "learning_rate": 0.0002528501628664495,
+      "loss": 1.9019,
+      "step": 330
+    },
+    {
+      "epoch": 0.7270730367929709,
+      "grad_norm": 0.3773690164089203,
+      "learning_rate": 0.0002526058631921824,
+      "loss": 1.9346,
+      "step": 331
+    },
+    {
+      "epoch": 0.729269632070291,
+      "grad_norm": 0.26342058181762695,
+      "learning_rate": 0.0002523615635179153,
+      "loss": 1.9755,
+      "step": 332
+    },
+    {
+      "epoch": 0.7314662273476112,
+      "grad_norm": 0.22879236936569214,
+      "learning_rate": 0.0002521172638436482,
+      "loss": 1.6739,
+      "step": 333
+    },
+    {
+      "epoch": 0.7336628226249313,
+      "grad_norm": 0.2745082378387451,
+      "learning_rate": 0.0002518729641693811,
+      "loss": 1.8863,
+      "step": 334
+    },
+    {
+      "epoch": 0.7358594179022515,
+      "grad_norm": 0.247549906373024,
+      "learning_rate": 0.000251628664495114,
+      "loss": 1.766,
+      "step": 335
+    },
+    {
+      "epoch": 0.7380560131795717,
+      "grad_norm": 0.359992653131485,
+      "learning_rate": 0.0002513843648208469,
+      "loss": 1.9052,
+      "step": 336
+    },
+    {
+      "epoch": 0.7402526084568918,
+      "grad_norm": 0.2775900363922119,
+      "learning_rate": 0.0002511400651465798,
+      "loss": 1.913,
+      "step": 337
+    },
+    {
+      "epoch": 0.742449203734212,
+      "grad_norm": 0.4396286904811859,
+      "learning_rate": 0.0002508957654723127,
+      "loss": 1.6597,
+      "step": 338
+    },
+    {
+      "epoch": 0.7446457990115322,
+      "grad_norm": 1.2676197290420532,
+      "learning_rate": 0.0002506514657980456,
+      "loss": 1.4836,
+      "step": 339
+    },
+    {
+      "epoch": 0.7468423942888522,
+      "grad_norm": 0.21433043479919434,
+      "learning_rate": 0.0002504071661237785,
+      "loss": 1.6793,
+      "step": 340
+    },
+    {
+      "epoch": 0.7490389895661724,
+      "grad_norm": 0.3791450560092926,
+      "learning_rate": 0.0002501628664495114,
+      "loss": 1.7938,
+      "step": 341
+    },
+    {
+      "epoch": 0.7512355848434926,
+      "grad_norm": 0.7319339513778687,
+      "learning_rate": 0.00024991856677524427,
+      "loss": 1.6075,
+      "step": 342
+    },
+    {
+      "epoch": 0.7534321801208127,
+      "grad_norm": 0.27512916922569275,
+      "learning_rate": 0.0002496742671009772,
+      "loss": 1.809,
+      "step": 343
+    },
+    {
+      "epoch": 0.7556287753981329,
+      "grad_norm": 0.3244725167751312,
+      "learning_rate": 0.0002494299674267101,
+      "loss": 1.7171,
+      "step": 344
+    },
+    {
+      "epoch": 0.7578253706754531,
+      "grad_norm": 0.2981887459754944,
+      "learning_rate": 0.000249185667752443,
+      "loss": 1.9597,
+      "step": 345
+    },
+    {
+      "epoch": 0.7600219659527732,
+      "grad_norm": 0.41594502329826355,
+      "learning_rate": 0.0002489413680781759,
+      "loss": 1.6825,
+      "step": 346
+    },
+    {
+      "epoch": 0.7622185612300933,
+      "grad_norm": 0.23714327812194824,
+      "learning_rate": 0.00024869706840390877,
+      "loss": 1.657,
+      "step": 347
+    },
+    {
+      "epoch": 0.7644151565074135,
+      "grad_norm": 0.33127591013908386,
+      "learning_rate": 0.0002484527687296417,
+      "loss": 1.8828,
+      "step": 348
+    },
+    {
+      "epoch": 0.7666117517847336,
+      "grad_norm": 0.27372831106185913,
+      "learning_rate": 0.0002482084690553746,
+      "loss": 1.2587,
+      "step": 349
+    },
+    {
+      "epoch": 0.7688083470620538,
+      "grad_norm": 0.214016392827034,
+      "learning_rate": 0.0002479641693811075,
+      "loss": 1.7531,
+      "step": 350
+    },
+    {
+      "epoch": 0.771004942339374,
+      "grad_norm": 0.26557856798171997,
+      "learning_rate": 0.0002477198697068404,
+      "loss": 2.0798,
+      "step": 351
+    },
+    {
+      "epoch": 0.7732015376166941,
+      "grad_norm": 0.31248384714126587,
+      "learning_rate": 0.00024747557003257327,
+      "loss": 1.7868,
+      "step": 352
+    },
+    {
+      "epoch": 0.7753981328940143,
+      "grad_norm": 0.2701241672039032,
+      "learning_rate": 0.00024723127035830616,
+      "loss": 1.5304,
+      "step": 353
+    },
+    {
+      "epoch": 0.7775947281713345,
+      "grad_norm": 0.2863544523715973,
+      "learning_rate": 0.0002469869706840391,
+      "loss": 1.8762,
+      "step": 354
+    },
+    {
+      "epoch": 0.7797913234486545,
+      "grad_norm": 0.317732036113739,
+      "learning_rate": 0.000246742671009772,
+      "loss": 1.2673,
+      "step": 355
+    },
+    {
+      "epoch": 0.7819879187259747,
+      "grad_norm": 0.3140071630477905,
+      "learning_rate": 0.0002464983713355049,
+      "loss": 1.7787,
+      "step": 356
+    },
+    {
+      "epoch": 0.7841845140032949,
+      "grad_norm": 0.8840736150741577,
+      "learning_rate": 0.00024625407166123777,
+      "loss": 1.805,
+      "step": 357
+    },
+    {
+      "epoch": 0.786381109280615,
+      "grad_norm": 0.21219655871391296,
+      "learning_rate": 0.00024600977198697066,
+      "loss": 1.5735,
+      "step": 358
+    },
+    {
+      "epoch": 0.7885777045579352,
+      "grad_norm": 0.2193101942539215,
+      "learning_rate": 0.00024576547231270354,
+      "loss": 1.3837,
+      "step": 359
+    },
+    {
+      "epoch": 0.7907742998352554,
+      "grad_norm": 0.238081693649292,
+      "learning_rate": 0.0002455211726384365,
+      "loss": 1.4646,
+      "step": 360
+    },
+    {
+      "epoch": 0.7929708951125755,
+      "grad_norm": 0.24821996688842773,
+      "learning_rate": 0.0002452768729641694,
+      "loss": 1.8037,
+      "step": 361
+    },
+    {
+      "epoch": 0.7951674903898956,
+      "grad_norm": 0.4130566418170929,
+      "learning_rate": 0.00024503257328990227,
+      "loss": 1.7878,
+      "step": 362
+    },
+    {
+      "epoch": 0.7973640856672158,
+      "grad_norm": 0.28218355774879456,
+      "learning_rate": 0.00024478827361563516,
+      "loss": 1.3709,
+      "step": 363
+    },
+    {
+      "epoch": 0.7995606809445359,
+      "grad_norm": 1.4110448360443115,
+      "learning_rate": 0.00024454397394136804,
+      "loss": 1.6775,
+      "step": 364
+    },
+    {
+      "epoch": 0.8017572762218561,
+      "grad_norm": 0.26909035444259644,
+      "learning_rate": 0.000244299674267101,
+      "loss": 1.2423,
+      "step": 365
+    },
+    {
+      "epoch": 0.8039538714991763,
+      "grad_norm": 0.27946969866752625,
+      "learning_rate": 0.00024405537459283385,
+      "loss": 1.8514,
+      "step": 366
+    },
+    {
+      "epoch": 0.8061504667764964,
+      "grad_norm": 0.28444793820381165,
+      "learning_rate": 0.00024381107491856677,
+      "loss": 1.5611,
+      "step": 367
+    },
+    {
+      "epoch": 0.8083470620538166,
+      "grad_norm": 0.2978900372982025,
+      "learning_rate": 0.00024356677524429966,
+      "loss": 1.4821,
+      "step": 368
+    },
+    {
+      "epoch": 0.8105436573311368,
+      "grad_norm": 0.7690660357475281,
+      "learning_rate": 0.00024332247557003257,
+      "loss": 1.8521,
+      "step": 369
+    },
+    {
+      "epoch": 0.8127402526084568,
+      "grad_norm": 0.20778357982635498,
+      "learning_rate": 0.00024307817589576546,
+      "loss": 1.2663,
+      "step": 370
+    },
+    {
+      "epoch": 0.814936847885777,
+      "grad_norm": 0.3188571035861969,
+      "learning_rate": 0.00024283387622149835,
+      "loss": 1.9455,
+      "step": 371
+    },
+    {
+      "epoch": 0.8171334431630972,
+      "grad_norm": 0.3040783405303955,
+      "learning_rate": 0.00024258957654723127,
+      "loss": 1.4808,
+      "step": 372
+    },
+    {
+      "epoch": 0.8193300384404174,
+      "grad_norm": 0.2951478660106659,
+      "learning_rate": 0.00024234527687296416,
+      "loss": 1.8996,
+      "step": 373
+    },
+    {
+      "epoch": 0.8215266337177375,
+      "grad_norm": 0.23818452656269073,
+      "learning_rate": 0.00024210097719869705,
+      "loss": 1.5897,
+      "step": 374
+    },
+    {
+      "epoch": 0.8237232289950577,
+      "grad_norm": 0.26143670082092285,
+      "learning_rate": 0.00024185667752442996,
+      "loss": 1.6846,
+      "step": 375
+    },
+    {
+      "epoch": 0.8259198242723779,
+      "grad_norm": 2.6398165225982666,
+      "learning_rate": 0.00024161237785016285,
+      "loss": 1.8354,
+      "step": 376
+    },
+    {
+      "epoch": 0.828116419549698,
+      "grad_norm": 0.4070051610469818,
+      "learning_rate": 0.00024136807817589574,
+      "loss": 1.7305,
+      "step": 377
+    },
+    {
+      "epoch": 0.8303130148270181,
+      "grad_norm": 0.28136730194091797,
+      "learning_rate": 0.00024112377850162866,
+      "loss": 1.7051,
+      "step": 378
+    },
+    {
+      "epoch": 0.8325096101043383,
+      "grad_norm": 0.21303951740264893,
+      "learning_rate": 0.00024087947882736155,
+      "loss": 1.4031,
+      "step": 379
+    },
+    {
+      "epoch": 0.8347062053816584,
+      "grad_norm": 0.2946576774120331,
+      "learning_rate": 0.00024063517915309446,
+      "loss": 1.7991,
+      "step": 380
+    },
+    {
+      "epoch": 0.8369028006589786,
+      "grad_norm": 0.2530032694339752,
+      "learning_rate": 0.00024039087947882735,
+      "loss": 1.6203,
+      "step": 381
+    },
+    {
+      "epoch": 0.8390993959362988,
+      "grad_norm": 0.2503615915775299,
+      "learning_rate": 0.00024014657980456024,
+      "loss": 1.6112,
+      "step": 382
+    },
+    {
+      "epoch": 0.8412959912136189,
+      "grad_norm": 0.33974409103393555,
+      "learning_rate": 0.00023990228013029316,
+      "loss": 1.7358,
+      "step": 383
+    },
+    {
+      "epoch": 0.8434925864909391,
+      "grad_norm": 0.362939715385437,
+      "learning_rate": 0.00023965798045602605,
+      "loss": 1.8923,
+      "step": 384
+    },
+    {
+      "epoch": 0.8456891817682592,
+      "grad_norm": 0.2244739532470703,
+      "learning_rate": 0.00023941368078175893,
+      "loss": 1.4727,
+      "step": 385
+    },
+    {
+      "epoch": 0.8478857770455793,
+      "grad_norm": 2.200756311416626,
+      "learning_rate": 0.00023916938110749185,
+      "loss": 2.1356,
+      "step": 386
+    },
+    {
+      "epoch": 0.8500823723228995,
+      "grad_norm": 0.2540011405944824,
+      "learning_rate": 0.00023892508143322474,
+      "loss": 1.743,
+      "step": 387
+    },
+    {
+      "epoch": 0.8522789676002197,
+      "grad_norm": 0.3036474287509918,
+      "learning_rate": 0.00023868078175895763,
+      "loss": 1.8361,
+      "step": 388
+    },
+    {
+      "epoch": 0.8544755628775398,
+      "grad_norm": 0.19831764698028564,
+      "learning_rate": 0.00023843648208469055,
+      "loss": 1.1846,
+      "step": 389
+    },
+    {
+      "epoch": 0.85667215815486,
+      "grad_norm": 0.2293327897787094,
+      "learning_rate": 0.00023819218241042343,
+      "loss": 1.6649,
+      "step": 390
+    },
+    {
+      "epoch": 0.8588687534321802,
+      "grad_norm": 0.2035629004240036,
+      "learning_rate": 0.00023794788273615632,
+      "loss": 1.5341,
+      "step": 391
+    },
+    {
+      "epoch": 0.8610653487095002,
+      "grad_norm": 0.21095000207424164,
+      "learning_rate": 0.00023770358306188924,
+      "loss": 1.904,
+      "step": 392
+    },
+    {
+      "epoch": 0.8632619439868204,
+      "grad_norm": 0.24570408463478088,
+      "learning_rate": 0.00023745928338762213,
+      "loss": 1.9608,
+      "step": 393
+    },
+    {
+      "epoch": 0.8654585392641406,
+      "grad_norm": 0.22200053930282593,
+      "learning_rate": 0.00023721498371335505,
+      "loss": 1.435,
+      "step": 394
+    },
+    {
+      "epoch": 0.8676551345414607,
+      "grad_norm": 0.3365512490272522,
+      "learning_rate": 0.00023697068403908794,
+      "loss": 2.0072,
+      "step": 395
+    },
+    {
+      "epoch": 0.8698517298187809,
+      "grad_norm": 0.2573312520980835,
+      "learning_rate": 0.00023672638436482082,
+      "loss": 1.4991,
+      "step": 396
+    },
+    {
+      "epoch": 0.8720483250961011,
+      "grad_norm": 0.3076959550380707,
+      "learning_rate": 0.00023648208469055374,
+      "loss": 2.0883,
+      "step": 397
+    },
+    {
+      "epoch": 0.8742449203734212,
+      "grad_norm": 1.0960148572921753,
+      "learning_rate": 0.00023623778501628663,
+      "loss": 1.4585,
+      "step": 398
+    },
+    {
+      "epoch": 0.8764415156507414,
+      "grad_norm": 0.2556944489479065,
+      "learning_rate": 0.00023599348534201952,
+      "loss": 1.7596,
+      "step": 399
+    },
+    {
+      "epoch": 0.8786381109280615,
+      "grad_norm": 0.21172761917114258,
+      "learning_rate": 0.00023574918566775244,
+      "loss": 1.5049,
+      "step": 400
+    },
+    {
+      "epoch": 0.8808347062053816,
+      "grad_norm": 0.2662028968334198,
+      "learning_rate": 0.00023550488599348532,
+      "loss": 1.5224,
+      "step": 401
+    },
+    {
+      "epoch": 0.8830313014827018,
+      "grad_norm": 0.3200950622558594,
+      "learning_rate": 0.00023526058631921821,
+      "loss": 2.1462,
+      "step": 402
+    },
+    {
+      "epoch": 0.885227896760022,
+      "grad_norm": 1.2355588674545288,
+      "learning_rate": 0.00023501628664495113,
+      "loss": 1.6096,
+      "step": 403
+    },
+    {
+      "epoch": 0.8874244920373421,
+      "grad_norm": 1.6774827241897583,
+      "learning_rate": 0.00023477198697068402,
+      "loss": 1.9455,
+      "step": 404
+    },
+    {
+      "epoch": 0.8896210873146623,
+      "grad_norm": 0.2593725323677063,
+      "learning_rate": 0.0002345276872964169,
+      "loss": 1.5961,
+      "step": 405
+    },
+    {
+      "epoch": 0.8918176825919825,
+      "grad_norm": 0.37221559882164,
+      "learning_rate": 0.00023428338762214982,
+      "loss": 1.6437,
+      "step": 406
+    },
+    {
+      "epoch": 0.8940142778693025,
+      "grad_norm": 0.45654961466789246,
+      "learning_rate": 0.00023403908794788271,
+      "loss": 1.8314,
+      "step": 407
+    },
+    {
+      "epoch": 0.8962108731466227,
+      "grad_norm": 0.24504287540912628,
+      "learning_rate": 0.00023379478827361563,
+      "loss": 1.6204,
+      "step": 408
+    },
+    {
+      "epoch": 0.8984074684239429,
+      "grad_norm": 0.693515956401825,
+      "learning_rate": 0.00023355048859934852,
+      "loss": 1.7317,
+      "step": 409
+    },
+    {
+      "epoch": 0.900604063701263,
+      "grad_norm": 0.2210521250963211,
+      "learning_rate": 0.0002333061889250814,
+      "loss": 1.6716,
+      "step": 410
+    },
+    {
+      "epoch": 0.9028006589785832,
+      "grad_norm": 0.29567384719848633,
+      "learning_rate": 0.00023306188925081432,
+      "loss": 1.8122,
+      "step": 411
+    },
+    {
+      "epoch": 0.9049972542559034,
+      "grad_norm": 0.2915588915348053,
+      "learning_rate": 0.00023281758957654721,
+      "loss": 1.4785,
+      "step": 412
+    },
+    {
+      "epoch": 0.9071938495332235,
+      "grad_norm": 0.27825218439102173,
+      "learning_rate": 0.0002325732899022801,
+      "loss": 1.8091,
+      "step": 413
+    },
+    {
+      "epoch": 0.9093904448105437,
+      "grad_norm": 0.9649204015731812,
+      "learning_rate": 0.00023232899022801302,
+      "loss": 1.7157,
+      "step": 414
+    },
+    {
+      "epoch": 0.9115870400878638,
+      "grad_norm": 0.26420706510543823,
+      "learning_rate": 0.0002320846905537459,
+      "loss": 1.3612,
+      "step": 415
+    },
+    {
+      "epoch": 0.9137836353651839,
+      "grad_norm": 0.2416657954454422,
+      "learning_rate": 0.0002318403908794788,
+      "loss": 1.7559,
+      "step": 416
+    },
+    {
+      "epoch": 0.9159802306425041,
+      "grad_norm": 0.25096720457077026,
+      "learning_rate": 0.00023159609120521171,
+      "loss": 1.4869,
+      "step": 417
+    },
+    {
+      "epoch": 0.9181768259198243,
+      "grad_norm": 0.942412793636322,
+      "learning_rate": 0.0002313517915309446,
+      "loss": 1.6754,
+      "step": 418
+    },
+    {
+      "epoch": 0.9203734211971444,
+      "grad_norm": 1.4666345119476318,
+      "learning_rate": 0.00023110749185667752,
+      "loss": 1.3626,
+      "step": 419
+    },
+    {
+      "epoch": 0.9225700164744646,
+      "grad_norm": 0.326686829328537,
+      "learning_rate": 0.0002308631921824104,
+      "loss": 1.9363,
+      "step": 420
+    },
+    {
+      "epoch": 0.9247666117517848,
+      "grad_norm": 0.5892528295516968,
+      "learning_rate": 0.0002306188925081433,
+      "loss": 1.4507,
+      "step": 421
+    },
+    {
+      "epoch": 0.9269632070291048,
+      "grad_norm": 0.24574226140975952,
+      "learning_rate": 0.00023037459283387621,
+      "loss": 1.6671,
+      "step": 422
+    },
+    {
+      "epoch": 0.929159802306425,
+      "grad_norm": 1.7222166061401367,
+      "learning_rate": 0.0002301302931596091,
+      "loss": 1.7336,
+      "step": 423
+    },
+    {
+      "epoch": 0.9313563975837452,
+      "grad_norm": 0.3384079933166504,
+      "learning_rate": 0.000229885993485342,
+      "loss": 1.7026,
+      "step": 424
+    },
+    {
+      "epoch": 0.9335529928610653,
+      "grad_norm": 0.22765927016735077,
+      "learning_rate": 0.0002296416938110749,
+      "loss": 1.3878,
+      "step": 425
+    },
+    {
+      "epoch": 0.9357495881383855,
+      "grad_norm": 0.35202109813690186,
+      "learning_rate": 0.0002293973941368078,
+      "loss": 1.9057,
+      "step": 426
+    },
+    {
+      "epoch": 0.9379461834157057,
+      "grad_norm": 0.19908814132213593,
+      "learning_rate": 0.0002291530944625407,
+      "loss": 1.4407,
+      "step": 427
+    },
+    {
+      "epoch": 0.9401427786930258,
+      "grad_norm": 0.40085896849632263,
+      "learning_rate": 0.0002289087947882736,
+      "loss": 1.1987,
+      "step": 428
+    },
+    {
+      "epoch": 0.942339373970346,
+      "grad_norm": 0.31224676966667175,
+      "learning_rate": 0.0002286644951140065,
+      "loss": 1.2761,
+      "step": 429
+    },
+    {
+      "epoch": 0.9445359692476661,
+      "grad_norm": 0.28462380170822144,
+      "learning_rate": 0.00022842019543973938,
+      "loss": 1.7321,
+      "step": 430
+    },
+    {
+      "epoch": 0.9467325645249862,
+      "grad_norm": 0.2645535171031952,
+      "learning_rate": 0.0002281758957654723,
+      "loss": 1.2466,
+      "step": 431
+    },
+    {
+      "epoch": 0.9489291598023064,
+      "grad_norm": 0.31242501735687256,
+      "learning_rate": 0.0002279315960912052,
+      "loss": 1.743,
+      "step": 432
+    },
+    {
+      "epoch": 0.9511257550796266,
+      "grad_norm": 2.431875228881836,
+      "learning_rate": 0.0002276872964169381,
+      "loss": 1.5049,
+      "step": 433
+    },
+    {
+      "epoch": 0.9533223503569467,
+      "grad_norm": 0.6512520909309387,
+      "learning_rate": 0.000227442996742671,
+      "loss": 1.1985,
+      "step": 434
+    },
+    {
+      "epoch": 0.9555189456342669,
+      "grad_norm": 0.2167208194732666,
+      "learning_rate": 0.00022719869706840388,
+      "loss": 1.4791,
+      "step": 435
+    },
+    {
+      "epoch": 0.9577155409115871,
+      "grad_norm": 0.21275602281093597,
+      "learning_rate": 0.0002269543973941368,
+      "loss": 1.1003,
+      "step": 436
+    },
+    {
+      "epoch": 0.9599121361889071,
+      "grad_norm": 0.19621990621089935,
+      "learning_rate": 0.0002267100977198697,
+      "loss": 1.4668,
+      "step": 437
+    },
+    {
+      "epoch": 0.9621087314662273,
+      "grad_norm": 0.3136390149593353,
+      "learning_rate": 0.00022646579804560258,
+      "loss": 1.7504,
+      "step": 438
+    },
+    {
+      "epoch": 0.9643053267435475,
+      "grad_norm": 0.23785126209259033,
+      "learning_rate": 0.0002262214983713355,
+      "loss": 1.742,
+      "step": 439
+    },
+    {
+      "epoch": 0.9665019220208677,
+      "grad_norm": 2.2050771713256836,
+      "learning_rate": 0.00022597719869706838,
+      "loss": 1.5767,
+      "step": 440
+    },
+    {
+      "epoch": 0.9686985172981878,
+      "grad_norm": 0.361826628446579,
+      "learning_rate": 0.00022573289902280127,
+      "loss": 1.7476,
+      "step": 441
+    },
+    {
+      "epoch": 0.970895112575508,
+      "grad_norm": 0.2659573554992676,
+      "learning_rate": 0.0002254885993485342,
+      "loss": 1.3809,
+      "step": 442
+    },
+    {
+      "epoch": 0.9730917078528282,
+      "grad_norm": 0.22771331667900085,
+      "learning_rate": 0.00022524429967426708,
+      "loss": 1.5044,
+      "step": 443
+    },
+    {
+      "epoch": 0.9752883031301482,
+      "grad_norm": 0.5552976131439209,
+      "learning_rate": 0.000225,
+      "loss": 2.0698,
+      "step": 444
+    },
+    {
+      "epoch": 0.9774848984074684,
+      "grad_norm": 0.28156644105911255,
+      "learning_rate": 0.00022475570032573288,
+      "loss": 1.4705,
+      "step": 445
+    },
+    {
+      "epoch": 0.9796814936847886,
+      "grad_norm": 0.34467193484306335,
+      "learning_rate": 0.00022451140065146577,
+      "loss": 1.6953,
+      "step": 446
+    },
+    {
+      "epoch": 0.9818780889621087,
+      "grad_norm": 0.24532712996006012,
+      "learning_rate": 0.0002242671009771987,
+      "loss": 1.9262,
+      "step": 447
+    },
+    {
+      "epoch": 0.9840746842394289,
+      "grad_norm": 0.1961548924446106,
+      "learning_rate": 0.00022402280130293158,
+      "loss": 1.4653,
+      "step": 448
+    },
+    {
+      "epoch": 0.9862712795167491,
+      "grad_norm": 0.20457522571086884,
+      "learning_rate": 0.00022377850162866447,
+      "loss": 1.6034,
+      "step": 449
+    },
+    {
+      "epoch": 0.9884678747940692,
+      "grad_norm": 0.18449504673480988,
+      "learning_rate": 0.00022353420195439738,
+      "loss": 1.366,
+      "step": 450
+    },
+    {
+      "epoch": 0.9906644700713894,
+      "grad_norm": 0.32383492588996887,
+      "learning_rate": 0.00022328990228013027,
+      "loss": 1.6299,
+      "step": 451
+    },
+    {
+      "epoch": 0.9928610653487095,
+      "grad_norm": 0.5974504947662354,
+      "learning_rate": 0.00022304560260586316,
+      "loss": 1.5247,
+      "step": 452
+    },
+    {
+      "epoch": 0.9950576606260296,
+      "grad_norm": 0.31966277956962585,
+      "learning_rate": 0.00022280130293159608,
+      "loss": 1.8135,
+      "step": 453
+    },
+    {
+      "epoch": 0.9972542559033498,
+      "grad_norm": 0.468244731426239,
+      "learning_rate": 0.00022255700325732897,
+      "loss": 1.8887,
+      "step": 454
+    },
+    {
+      "epoch": 0.99945085118067,
+      "grad_norm": 0.3096922039985657,
+      "learning_rate": 0.00022231270358306186,
+      "loss": 1.7351,
+      "step": 455
+    },
+    {
+      "epoch": 1.00164744645799,
+      "grad_norm": 0.2978070080280304,
+      "learning_rate": 0.00022206840390879477,
+      "loss": 1.8199,
+      "step": 456
+    },
+    {
+      "epoch": 1.0038440417353103,
+      "grad_norm": 0.25190699100494385,
+      "learning_rate": 0.00022182410423452766,
+      "loss": 1.6322,
+      "step": 457
+    },
+    {
+      "epoch": 1.0060406370126305,
+      "grad_norm": 0.21248610317707062,
+      "learning_rate": 0.00022157980456026058,
+      "loss": 1.3978,
+      "step": 458
+    },
+    {
+      "epoch": 1.0082372322899507,
+      "grad_norm": 0.3107396960258484,
+      "learning_rate": 0.00022133550488599347,
+      "loss": 1.4464,
+      "step": 459
+    },
+    {
+      "epoch": 1.0104338275672706,
+      "grad_norm": 0.23506712913513184,
+      "learning_rate": 0.00022109120521172636,
+      "loss": 1.5015,
+      "step": 460
+    },
+    {
+      "epoch": 1.0126304228445908,
+      "grad_norm": 0.2281513810157776,
+      "learning_rate": 0.00022084690553745927,
+      "loss": 1.2438,
+      "step": 461
+    },
+    {
+      "epoch": 1.014827018121911,
+      "grad_norm": 0.22616425156593323,
+      "learning_rate": 0.00022060260586319216,
+      "loss": 1.7366,
+      "step": 462
+    },
+    {
+      "epoch": 1.0170236133992312,
+      "grad_norm": 0.20376840233802795,
+      "learning_rate": 0.00022035830618892505,
+      "loss": 1.7806,
+      "step": 463
+    },
+    {
+      "epoch": 1.0192202086765514,
+      "grad_norm": 0.2397449016571045,
+      "learning_rate": 0.00022011400651465797,
+      "loss": 1.7793,
+      "step": 464
+    },
+    {
+      "epoch": 1.0214168039538716,
+      "grad_norm": 0.24638597667217255,
+      "learning_rate": 0.00021986970684039086,
+      "loss": 1.8908,
+      "step": 465
+    },
+    {
+      "epoch": 1.0236133992311915,
+      "grad_norm": 0.2102181762456894,
+      "learning_rate": 0.00021962540716612375,
+      "loss": 1.5963,
+      "step": 466
+    },
+    {
+      "epoch": 1.0258099945085117,
+      "grad_norm": 0.24757710099220276,
+      "learning_rate": 0.00021938110749185666,
+      "loss": 1.3865,
+      "step": 467
+    },
+    {
+      "epoch": 1.028006589785832,
+      "grad_norm": 0.22550068795681,
+      "learning_rate": 0.00021913680781758955,
+      "loss": 1.3863,
+      "step": 468
+    },
+    {
+      "epoch": 1.0302031850631521,
+      "grad_norm": 0.22517824172973633,
+      "learning_rate": 0.00021889250814332244,
+      "loss": 1.6623,
+      "step": 469
+    },
+    {
+      "epoch": 1.0323997803404723,
+      "grad_norm": 0.3622633218765259,
+      "learning_rate": 0.00021864820846905536,
+      "loss": 1.6141,
+      "step": 470
+    },
+    {
+      "epoch": 1.0345963756177925,
+      "grad_norm": 0.3289901912212372,
+      "learning_rate": 0.00021840390879478825,
+      "loss": 1.3277,
+      "step": 471
+    },
+    {
+      "epoch": 1.0367929708951125,
+      "grad_norm": 0.22940000891685486,
+      "learning_rate": 0.00021815960912052116,
+      "loss": 1.5327,
+      "step": 472
+    },
+    {
+      "epoch": 1.0389895661724327,
+      "grad_norm": 0.28177008032798767,
+      "learning_rate": 0.00021791530944625405,
+      "loss": 1.7185,
+      "step": 473
+    },
+    {
+      "epoch": 1.0411861614497528,
+      "grad_norm": 0.24486376345157623,
+      "learning_rate": 0.00021767100977198694,
+      "loss": 1.6732,
+      "step": 474
+    },
+    {
+      "epoch": 1.043382756727073,
+      "grad_norm": 0.238804891705513,
+      "learning_rate": 0.00021742671009771986,
+      "loss": 1.7674,
+      "step": 475
+    },
+    {
+      "epoch": 1.0455793520043932,
+      "grad_norm": 0.2188219428062439,
+      "learning_rate": 0.00021718241042345275,
+      "loss": 1.8207,
+      "step": 476
+    },
+    {
+      "epoch": 1.0477759472817134,
+      "grad_norm": 0.2144714593887329,
+      "learning_rate": 0.00021693811074918564,
+      "loss": 1.4587,
+      "step": 477
+    },
+    {
+      "epoch": 1.0499725425590336,
+      "grad_norm": 0.29802870750427246,
+      "learning_rate": 0.00021669381107491855,
+      "loss": 1.9801,
+      "step": 478
+    },
+    {
+      "epoch": 1.0521691378363536,
+      "grad_norm": 0.7072665691375732,
+      "learning_rate": 0.00021644951140065144,
+      "loss": 1.6092,
+      "step": 479
+    },
+    {
+      "epoch": 1.0543657331136738,
+      "grad_norm": 0.1766052395105362,
+      "learning_rate": 0.00021620521172638433,
+      "loss": 1.1077,
+      "step": 480
+    },
+    {
+      "epoch": 1.056562328390994,
+      "grad_norm": 0.5119725465774536,
+      "learning_rate": 0.00021596091205211725,
+      "loss": 1.4281,
+      "step": 481
+    },
+    {
+      "epoch": 1.0587589236683141,
+      "grad_norm": 1.2915948629379272,
+      "learning_rate": 0.00021571661237785014,
+      "loss": 1.3119,
+      "step": 482
+    },
+    {
+      "epoch": 1.0609555189456343,
+      "grad_norm": 0.2567638158798218,
+      "learning_rate": 0.00021547231270358305,
+      "loss": 1.6526,
+      "step": 483
+    },
+    {
+      "epoch": 1.0631521142229543,
+      "grad_norm": 0.2499484270811081,
+      "learning_rate": 0.00021522801302931594,
+      "loss": 1.7786,
+      "step": 484
+    },
+    {
+      "epoch": 1.0653487095002745,
+      "grad_norm": 0.17313426733016968,
+      "learning_rate": 0.00021498371335504883,
+      "loss": 1.313,
+      "step": 485
+    },
+    {
+      "epoch": 1.0675453047775947,
+      "grad_norm": 0.3492632508277893,
+      "learning_rate": 0.00021473941368078175,
+      "loss": 1.5828,
+      "step": 486
+    },
+    {
+      "epoch": 1.0697419000549149,
+      "grad_norm": 0.1732093095779419,
+      "learning_rate": 0.00021449511400651464,
+      "loss": 1.4776,
+      "step": 487
+    },
+    {
+      "epoch": 1.071938495332235,
+      "grad_norm": 0.23682935535907745,
+      "learning_rate": 0.00021425081433224753,
+      "loss": 1.7186,
+      "step": 488
+    },
+    {
+      "epoch": 1.0741350906095553,
+      "grad_norm": 0.6888928413391113,
+      "learning_rate": 0.00021400651465798044,
+      "loss": 1.5435,
+      "step": 489
+    },
+    {
+      "epoch": 1.0763316858868754,
+      "grad_norm": 0.20798350870609283,
+      "learning_rate": 0.00021376221498371333,
+      "loss": 1.3046,
+      "step": 490
+    },
+    {
+      "epoch": 1.0785282811641954,
+      "grad_norm": 0.4220634996891022,
+      "learning_rate": 0.00021351791530944622,
+      "loss": 1.8858,
+      "step": 491
+    },
+    {
+      "epoch": 1.0807248764415156,
+      "grad_norm": 0.21109972894191742,
+      "learning_rate": 0.00021327361563517914,
+      "loss": 1.4425,
+      "step": 492
+    },
+    {
+      "epoch": 1.0829214717188358,
+      "grad_norm": 0.3222923278808594,
+      "learning_rate": 0.00021302931596091203,
+      "loss": 1.6777,
+      "step": 493
+    },
+    {
+      "epoch": 1.085118066996156,
+      "grad_norm": 0.43307945132255554,
+      "learning_rate": 0.00021278501628664491,
+      "loss": 1.2985,
+      "step": 494
+    },
+    {
+      "epoch": 1.0873146622734762,
+      "grad_norm": 0.44205138087272644,
+      "learning_rate": 0.00021254071661237783,
+      "loss": 1.6253,
+      "step": 495
+    },
+    {
+      "epoch": 1.0895112575507964,
+      "grad_norm": 0.3212667405605316,
+      "learning_rate": 0.00021229641693811072,
+      "loss": 1.6246,
+      "step": 496
+    },
+    {
+      "epoch": 1.0917078528281163,
+      "grad_norm": 0.4754985570907593,
+      "learning_rate": 0.00021205211726384364,
+      "loss": 1.4124,
+      "step": 497
+    },
+    {
+      "epoch": 1.0939044481054365,
+      "grad_norm": 0.2065618932247162,
+      "learning_rate": 0.00021180781758957653,
+      "loss": 1.447,
+      "step": 498
+    },
+    {
+      "epoch": 1.0961010433827567,
+      "grad_norm": 2.5149717330932617,
+      "learning_rate": 0.00021156351791530941,
+      "loss": 1.7989,
+      "step": 499
+    },
+    {
+      "epoch": 1.098297638660077,
+      "grad_norm": 0.21608681976795197,
+      "learning_rate": 0.00021131921824104233,
+      "loss": 1.6197,
+      "step": 500
+    },
+    {
+      "epoch": 1.100494233937397,
+      "grad_norm": 0.20178812742233276,
+      "learning_rate": 0.00021107491856677522,
+      "loss": 1.7591,
+      "step": 501
+    },
+    {
+      "epoch": 1.1026908292147173,
+      "grad_norm": 0.21548449993133545,
+      "learning_rate": 0.0002108306188925081,
+      "loss": 1.4433,
+      "step": 502
+    },
+    {
+      "epoch": 1.1048874244920373,
+      "grad_norm": 0.29982197284698486,
+      "learning_rate": 0.00021058631921824103,
+      "loss": 1.736,
+      "step": 503
+    },
+    {
+      "epoch": 1.1070840197693574,
+      "grad_norm": 0.23096679151058197,
+      "learning_rate": 0.00021034201954397392,
+      "loss": 1.5644,
+      "step": 504
+    },
+    {
+      "epoch": 1.1092806150466776,
+      "grad_norm": 0.3034733533859253,
+      "learning_rate": 0.0002100977198697068,
+      "loss": 1.8251,
+      "step": 505
+    },
+    {
+      "epoch": 1.1114772103239978,
+      "grad_norm": 0.7842176556587219,
+      "learning_rate": 0.00020985342019543972,
+      "loss": 1.5628,
+      "step": 506
+    },
+    {
+      "epoch": 1.113673805601318,
+      "grad_norm": 0.2867143154144287,
+      "learning_rate": 0.0002096091205211726,
+      "loss": 1.8301,
+      "step": 507
+    },
+    {
+      "epoch": 1.1158704008786382,
+      "grad_norm": 0.18750140070915222,
+      "learning_rate": 0.0002093648208469055,
+      "loss": 1.5675,
+      "step": 508
+    },
+    {
+      "epoch": 1.1180669961559582,
+      "grad_norm": 0.33146658539772034,
+      "learning_rate": 0.00020912052117263842,
+      "loss": 1.4345,
+      "step": 509
+    },
+    {
+      "epoch": 1.1202635914332784,
+      "grad_norm": 0.332499235868454,
+      "learning_rate": 0.0002088762214983713,
+      "loss": 1.4928,
+      "step": 510
+    },
+    {
+      "epoch": 1.1224601867105986,
+      "grad_norm": 0.18482932448387146,
+      "learning_rate": 0.00020863192182410422,
+      "loss": 1.3006,
+      "step": 511
+    },
+    {
+      "epoch": 1.1246567819879187,
+      "grad_norm": 0.2448980063199997,
+      "learning_rate": 0.0002083876221498371,
+      "loss": 1.5272,
+      "step": 512
+    },
+    {
+      "epoch": 1.126853377265239,
+      "grad_norm": 0.28912651538848877,
+      "learning_rate": 0.00020814332247557,
+      "loss": 1.7425,
+      "step": 513
+    },
+    {
+      "epoch": 1.1290499725425591,
+      "grad_norm": 0.2543645203113556,
+      "learning_rate": 0.00020789902280130292,
+      "loss": 1.3965,
+      "step": 514
+    },
+    {
+      "epoch": 1.131246567819879,
+      "grad_norm": 0.2461393028497696,
+      "learning_rate": 0.0002076547231270358,
+      "loss": 1.6693,
+      "step": 515
+    },
+    {
+      "epoch": 1.1334431630971993,
+      "grad_norm": 0.2228182852268219,
+      "learning_rate": 0.0002074104234527687,
+      "loss": 1.5114,
+      "step": 516
+    },
+    {
+      "epoch": 1.1356397583745195,
+      "grad_norm": 0.3853662610054016,
+      "learning_rate": 0.0002071661237785016,
+      "loss": 1.3884,
+      "step": 517
+    },
+    {
+      "epoch": 1.1378363536518397,
+      "grad_norm": 0.5464024543762207,
+      "learning_rate": 0.0002069218241042345,
+      "loss": 1.6447,
+      "step": 518
+    },
+    {
+      "epoch": 1.1400329489291599,
+      "grad_norm": 0.27296561002731323,
+      "learning_rate": 0.0002066775244299674,
+      "loss": 1.3506,
+      "step": 519
+    },
+    {
+      "epoch": 1.14222954420648,
+      "grad_norm": 0.3924662470817566,
+      "learning_rate": 0.0002064332247557003,
+      "loss": 1.6451,
+      "step": 520
+    },
+    {
+      "epoch": 1.1444261394838002,
+      "grad_norm": 0.24612033367156982,
+      "learning_rate": 0.0002061889250814332,
+      "loss": 1.4206,
+      "step": 521
+    },
+    {
+      "epoch": 1.1466227347611202,
+      "grad_norm": 1.2090104818344116,
+      "learning_rate": 0.0002059446254071661,
+      "loss": 1.8171,
+      "step": 522
+    },
+    {
+      "epoch": 1.1488193300384404,
+      "grad_norm": 0.19782127439975739,
+      "learning_rate": 0.000205700325732899,
+      "loss": 1.2027,
+      "step": 523
+    },
+    {
+      "epoch": 1.1510159253157606,
+      "grad_norm": 0.23932522535324097,
+      "learning_rate": 0.0002054560260586319,
+      "loss": 1.6618,
+      "step": 524
+    },
+    {
+      "epoch": 1.1532125205930808,
+      "grad_norm": 0.7041940093040466,
+      "learning_rate": 0.0002052117263843648,
+      "loss": 1.7567,
+      "step": 525
+    },
+    {
+      "epoch": 1.155409115870401,
+      "grad_norm": 1.6351287364959717,
+      "learning_rate": 0.0002049674267100977,
+      "loss": 1.856,
+      "step": 526
+    },
+    {
+      "epoch": 1.157605711147721,
+      "grad_norm": 0.20006190240383148,
+      "learning_rate": 0.00020472312703583058,
+      "loss": 1.3139,
+      "step": 527
+    },
+    {
+      "epoch": 1.1598023064250411,
+      "grad_norm": 0.7060168981552124,
+      "learning_rate": 0.0002044788273615635,
+      "loss": 1.4103,
+      "step": 528
+    },
+    {
+      "epoch": 1.1619989017023613,
+      "grad_norm": 0.22440698742866516,
+      "learning_rate": 0.0002042345276872964,
+      "loss": 1.2358,
+      "step": 529
+    },
+    {
+      "epoch": 1.1641954969796815,
+      "grad_norm": 0.2695049047470093,
+      "learning_rate": 0.00020399022801302928,
+      "loss": 1.4645,
+      "step": 530
+    },
+    {
+      "epoch": 1.1663920922570017,
+      "grad_norm": 0.2839784026145935,
+      "learning_rate": 0.0002037459283387622,
+      "loss": 1.4995,
+      "step": 531
+    },
+    {
+      "epoch": 1.1685886875343219,
+      "grad_norm": 0.21016795933246613,
+      "learning_rate": 0.00020350162866449508,
+      "loss": 1.4888,
+      "step": 532
+    },
+    {
+      "epoch": 1.170785282811642,
+      "grad_norm": 0.33923205733299255,
+      "learning_rate": 0.00020325732899022797,
+      "loss": 1.5043,
+      "step": 533
+    },
+    {
+      "epoch": 1.172981878088962,
+      "grad_norm": 0.2953750789165497,
+      "learning_rate": 0.0002030130293159609,
+      "loss": 1.7876,
+      "step": 534
+    },
+    {
+      "epoch": 1.1751784733662822,
+      "grad_norm": 0.3070078194141388,
+      "learning_rate": 0.00020276872964169378,
+      "loss": 1.655,
+      "step": 535
+    },
+    {
+      "epoch": 1.1773750686436024,
+      "grad_norm": 0.4144851863384247,
+      "learning_rate": 0.0002025244299674267,
+      "loss": 1.5159,
+      "step": 536
+    },
+    {
+      "epoch": 1.1795716639209226,
+      "grad_norm": 0.9455963969230652,
+      "learning_rate": 0.00020228013029315958,
+      "loss": 1.6087,
+      "step": 537
+    },
+    {
+      "epoch": 1.1817682591982428,
+      "grad_norm": 0.19832828640937805,
+      "learning_rate": 0.00020203583061889247,
+      "loss": 1.1509,
+      "step": 538
+    },
+    {
+      "epoch": 1.1839648544755628,
+      "grad_norm": 0.9812628626823425,
+      "learning_rate": 0.0002017915309446254,
+      "loss": 2.236,
+      "step": 539
+    },
+    {
+      "epoch": 1.186161449752883,
+      "grad_norm": 0.17755061388015747,
+      "learning_rate": 0.00020154723127035828,
+      "loss": 1.1965,
+      "step": 540
+    },
+    {
+      "epoch": 1.1883580450302031,
+      "grad_norm": 0.21581527590751648,
+      "learning_rate": 0.00020130293159609117,
+      "loss": 1.5668,
+      "step": 541
+    },
+    {
+      "epoch": 1.1905546403075233,
+      "grad_norm": 0.3403506875038147,
+      "learning_rate": 0.00020105863192182408,
+      "loss": 1.6225,
+      "step": 542
+    },
+    {
+      "epoch": 1.1927512355848435,
+      "grad_norm": 1.1003831624984741,
+      "learning_rate": 0.00020081433224755697,
+      "loss": 1.6881,
+      "step": 543
+    },
+    {
+      "epoch": 1.1949478308621637,
+      "grad_norm": 0.23421354591846466,
+      "learning_rate": 0.00020057003257328986,
+      "loss": 1.6662,
+      "step": 544
+    },
+    {
+      "epoch": 1.197144426139484,
+      "grad_norm": 0.3825953006744385,
+      "learning_rate": 0.00020032573289902278,
+      "loss": 1.6887,
+      "step": 545
+    },
+    {
+      "epoch": 1.1993410214168039,
+      "grad_norm": 0.4678840637207031,
+      "learning_rate": 0.00020008143322475567,
+      "loss": 1.5962,
+      "step": 546
+    },
+    {
+      "epoch": 1.201537616694124,
+      "grad_norm": 0.26681530475616455,
+      "learning_rate": 0.00019983713355048856,
+      "loss": 1.8904,
+      "step": 547
+    },
+    {
+      "epoch": 1.2037342119714443,
+      "grad_norm": 0.2134128361940384,
+      "learning_rate": 0.00019959283387622147,
+      "loss": 1.5437,
+      "step": 548
+    },
+    {
+      "epoch": 1.2059308072487644,
+      "grad_norm": 0.4676864743232727,
+      "learning_rate": 0.00019934853420195436,
+      "loss": 1.9497,
+      "step": 549
+    },
+    {
+      "epoch": 1.2081274025260846,
+      "grad_norm": 0.27725517749786377,
+      "learning_rate": 0.00019910423452768728,
+      "loss": 1.6434,
+      "step": 550
+    },
+    {
+      "epoch": 1.2103239978034046,
+      "grad_norm": 0.9268063306808472,
+      "learning_rate": 0.00019885993485342017,
+      "loss": 1.6241,
+      "step": 551
+    },
+    {
+      "epoch": 1.2125205930807248,
+      "grad_norm": 0.8154603242874146,
+      "learning_rate": 0.00019861563517915306,
+      "loss": 1.7555,
+      "step": 552
+    },
+    {
+      "epoch": 1.214717188358045,
+      "grad_norm": 0.23990018665790558,
+      "learning_rate": 0.00019837133550488597,
+      "loss": 1.5566,
+      "step": 553
+    },
+    {
+      "epoch": 1.2169137836353652,
+      "grad_norm": 0.19225336611270905,
+      "learning_rate": 0.00019812703583061886,
+      "loss": 1.5519,
+      "step": 554
+    },
+    {
+      "epoch": 1.2191103789126854,
+      "grad_norm": 0.334241658449173,
+      "learning_rate": 0.00019788273615635175,
+      "loss": 1.8464,
+      "step": 555
+    },
+    {
+      "epoch": 1.2213069741900056,
+      "grad_norm": 0.3313543498516083,
+      "learning_rate": 0.00019763843648208467,
+      "loss": 1.6972,
+      "step": 556
+    },
+    {
+      "epoch": 1.2235035694673257,
+      "grad_norm": 0.20367823541164398,
+      "learning_rate": 0.00019739413680781756,
+      "loss": 1.4308,
+      "step": 557
+    },
+    {
+      "epoch": 1.2257001647446457,
+      "grad_norm": 0.3710614740848541,
+      "learning_rate": 0.00019714983713355045,
+      "loss": 1.5921,
+      "step": 558
+    },
+    {
+      "epoch": 1.227896760021966,
+      "grad_norm": 0.2657913565635681,
+      "learning_rate": 0.00019690553745928336,
+      "loss": 1.7072,
+      "step": 559
+    },
+    {
+      "epoch": 1.230093355299286,
+      "grad_norm": 0.18050478398799896,
+      "learning_rate": 0.00019666123778501625,
+      "loss": 1.3667,
+      "step": 560
+    },
+    {
+      "epoch": 1.2322899505766063,
+      "grad_norm": 0.2166796773672104,
+      "learning_rate": 0.00019641693811074917,
+      "loss": 1.4512,
+      "step": 561
+    },
+    {
+      "epoch": 1.2344865458539265,
+      "grad_norm": 2.5783748626708984,
+      "learning_rate": 0.00019617263843648206,
+      "loss": 1.6422,
+      "step": 562
+    },
+    {
+      "epoch": 1.2366831411312464,
+      "grad_norm": 0.2058299332857132,
+      "learning_rate": 0.00019592833876221495,
+      "loss": 1.5315,
+      "step": 563
+    },
+    {
+      "epoch": 1.2388797364085666,
+      "grad_norm": 2.7429044246673584,
+      "learning_rate": 0.00019568403908794786,
+      "loss": 1.4675,
+      "step": 564
+    },
+    {
+      "epoch": 1.2410763316858868,
+      "grad_norm": 0.33138278126716614,
+      "learning_rate": 0.00019543973941368075,
+      "loss": 1.6564,
+      "step": 565
+    },
+    {
+      "epoch": 1.243272926963207,
+      "grad_norm": 0.3432352840900421,
+      "learning_rate": 0.00019519543973941364,
+      "loss": 1.5867,
+      "step": 566
+    },
+    {
+      "epoch": 1.2454695222405272,
+      "grad_norm": 0.22188657522201538,
+      "learning_rate": 0.00019495114006514656,
+      "loss": 1.3163,
+      "step": 567
+    },
+    {
+      "epoch": 1.2476661175178474,
+      "grad_norm": 2.7905001640319824,
+      "learning_rate": 0.00019470684039087945,
+      "loss": 1.778,
+      "step": 568
+    },
+    {
+      "epoch": 1.2498627127951676,
+      "grad_norm": 1.197300672531128,
+      "learning_rate": 0.00019446254071661234,
+      "loss": 1.1947,
+      "step": 569
+    },
+    {
+      "epoch": 1.2520593080724876,
+      "grad_norm": 0.2688259482383728,
+      "learning_rate": 0.00019421824104234525,
+      "loss": 1.2032,
+      "step": 570
+    },
+    {
+      "epoch": 1.2542559033498077,
+      "grad_norm": 0.7616679072380066,
+      "learning_rate": 0.00019397394136807814,
+      "loss": 1.8771,
+      "step": 571
+    },
+    {
+      "epoch": 1.256452498627128,
+      "grad_norm": 0.31175944209098816,
+      "learning_rate": 0.00019372964169381103,
+      "loss": 1.8169,
+      "step": 572
+    },
+    {
+      "epoch": 1.2586490939044481,
+      "grad_norm": 0.30868232250213623,
+      "learning_rate": 0.00019348534201954395,
+      "loss": 1.7069,
+      "step": 573
+    },
+    {
+      "epoch": 1.2608456891817683,
+      "grad_norm": 0.7784240245819092,
+      "learning_rate": 0.00019324104234527684,
+      "loss": 1.8016,
+      "step": 574
+    },
+    {
+      "epoch": 1.2630422844590883,
+      "grad_norm": 0.2039678692817688,
+      "learning_rate": 0.00019299674267100975,
+      "loss": 1.7056,
+      "step": 575
+    },
+    {
+      "epoch": 1.2652388797364087,
+      "grad_norm": 0.2906222343444824,
+      "learning_rate": 0.00019275244299674264,
+      "loss": 1.6719,
+      "step": 576
+    },
+    {
+      "epoch": 1.2674354750137287,
+      "grad_norm": 0.6174877285957336,
+      "learning_rate": 0.00019250814332247553,
+      "loss": 1.8801,
+      "step": 577
+    },
+    {
+      "epoch": 1.2696320702910489,
+      "grad_norm": 0.22921861708164215,
+      "learning_rate": 0.00019226384364820845,
+      "loss": 1.074,
+      "step": 578
+    },
+    {
+      "epoch": 1.271828665568369,
+      "grad_norm": 0.28714317083358765,
+      "learning_rate": 0.00019201954397394134,
+      "loss": 1.3903,
+      "step": 579
+    },
+    {
+      "epoch": 1.2740252608456892,
+      "grad_norm": 0.20298270881175995,
+      "learning_rate": 0.00019177524429967423,
+      "loss": 1.4002,
+      "step": 580
+    },
+    {
+      "epoch": 1.2762218561230094,
+      "grad_norm": 0.2090635448694229,
+      "learning_rate": 0.00019153094462540714,
+      "loss": 1.7083,
+      "step": 581
+    },
+    {
+      "epoch": 1.2784184514003294,
+      "grad_norm": 0.26641401648521423,
+      "learning_rate": 0.00019128664495114003,
+      "loss": 1.6632,
+      "step": 582
+    },
+    {
+      "epoch": 1.2806150466776496,
+      "grad_norm": 0.3250808119773865,
+      "learning_rate": 0.00019104234527687292,
+      "loss": 1.5002,
+      "step": 583
+    },
+    {
+      "epoch": 1.2828116419549698,
+      "grad_norm": 0.2042284607887268,
+      "learning_rate": 0.00019079804560260584,
+      "loss": 1.2531,
+      "step": 584
+    },
+    {
+      "epoch": 1.28500823723229,
+      "grad_norm": 0.24577483534812927,
+      "learning_rate": 0.00019055374592833873,
+      "loss": 1.53,
+      "step": 585
+    },
+    {
+      "epoch": 1.2872048325096102,
+      "grad_norm": 0.43627724051475525,
+      "learning_rate": 0.00019030944625407164,
+      "loss": 1.4549,
+      "step": 586
+    },
+    {
+      "epoch": 1.2894014277869301,
+      "grad_norm": 0.3216465413570404,
+      "learning_rate": 0.00019006514657980453,
+      "loss": 1.4677,
+      "step": 587
+    },
+    {
+      "epoch": 1.2915980230642505,
+      "grad_norm": 0.2548219859600067,
+      "learning_rate": 0.00018982084690553742,
+      "loss": 1.8795,
+      "step": 588
+    },
+    {
+      "epoch": 1.2937946183415705,
+      "grad_norm": 0.3525017499923706,
+      "learning_rate": 0.00018957654723127034,
+      "loss": 1.6052,
+      "step": 589
+    },
+    {
+      "epoch": 1.2959912136188907,
+      "grad_norm": 0.21578241884708405,
+      "learning_rate": 0.00018933224755700323,
+      "loss": 1.4171,
+      "step": 590
+    },
+    {
+      "epoch": 1.2981878088962109,
+      "grad_norm": 0.31156131625175476,
+      "learning_rate": 0.00018908794788273612,
+      "loss": 1.6536,
+      "step": 591
+    },
+    {
+      "epoch": 1.300384404173531,
+      "grad_norm": 1.6835910081863403,
+      "learning_rate": 0.00018884364820846903,
+      "loss": 1.5284,
+      "step": 592
+    },
+    {
+      "epoch": 1.3025809994508513,
+      "grad_norm": 0.22443470358848572,
+      "learning_rate": 0.00018859934853420192,
+      "loss": 1.3553,
+      "step": 593
+    },
+    {
+      "epoch": 1.3047775947281712,
+      "grad_norm": 0.24995210766792297,
+      "learning_rate": 0.0001883550488599348,
+      "loss": 1.4075,
+      "step": 594
+    },
+    {
+      "epoch": 1.3069741900054914,
+      "grad_norm": 0.24818892776966095,
+      "learning_rate": 0.00018811074918566773,
+      "loss": 1.6966,
+      "step": 595
+    },
+    {
+      "epoch": 1.3091707852828116,
+      "grad_norm": 0.1782926768064499,
+      "learning_rate": 0.00018786644951140062,
+      "loss": 1.3093,
+      "step": 596
+    },
+    {
+      "epoch": 1.3113673805601318,
+      "grad_norm": 0.3287822902202606,
+      "learning_rate": 0.0001876221498371335,
+      "loss": 1.3642,
+      "step": 597
+    },
+    {
+      "epoch": 1.313563975837452,
+      "grad_norm": 0.2409510314464569,
+      "learning_rate": 0.00018737785016286645,
+      "loss": 1.3149,
+      "step": 598
+    },
+    {
+      "epoch": 1.3157605711147722,
+      "grad_norm": 0.16456027328968048,
+      "learning_rate": 0.00018713355048859934,
+      "loss": 1.2678,
+      "step": 599
+    },
+    {
+      "epoch": 1.3179571663920924,
+      "grad_norm": 0.1810493916273117,
+      "learning_rate": 0.00018688925081433225,
+      "loss": 1.6251,
+      "step": 600
+    },
+    {
+      "epoch": 1.3201537616694123,
+      "grad_norm": 0.18858911097049713,
+      "learning_rate": 0.00018664495114006514,
+      "loss": 1.4204,
+      "step": 601
+    },
+    {
+      "epoch": 1.3223503569467325,
+      "grad_norm": 0.19352254271507263,
+      "learning_rate": 0.00018640065146579803,
+      "loss": 1.3325,
+      "step": 602
+    },
+    {
+      "epoch": 1.3245469522240527,
+      "grad_norm": 0.1964646577835083,
+      "learning_rate": 0.00018615635179153095,
+      "loss": 1.5427,
+      "step": 603
+    },
+    {
+      "epoch": 1.326743547501373,
+      "grad_norm": 0.18267549574375153,
+      "learning_rate": 0.00018591205211726384,
+      "loss": 1.3247,
+      "step": 604
+    },
+    {
+      "epoch": 1.328940142778693,
+      "grad_norm": 0.2856385409832001,
+      "learning_rate": 0.00018566775244299675,
+      "loss": 1.9193,
+      "step": 605
+    },
+    {
+      "epoch": 1.331136738056013,
+      "grad_norm": 0.2759475111961365,
+      "learning_rate": 0.00018542345276872964,
+      "loss": 1.4201,
+      "step": 606
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 0.28060653805732727,
+      "learning_rate": 0.00018517915309446253,
+      "loss": 1.5524,
+      "step": 607
+    },
+    {
+      "epoch": 1.3355299286106534,
+      "grad_norm": 0.29499927163124084,
+      "learning_rate": 0.00018493485342019545,
+      "loss": 1.8082,
+      "step": 608
+    },
+    {
+      "epoch": 1.3377265238879736,
+      "grad_norm": 0.2193842977285385,
+      "learning_rate": 0.00018469055374592834,
+      "loss": 1.4362,
+      "step": 609
+    },
+    {
+      "epoch": 1.3399231191652938,
+      "grad_norm": 0.23178169131278992,
+      "learning_rate": 0.00018444625407166123,
+      "loss": 1.6218,
+      "step": 610
+    },
+    {
+      "epoch": 1.342119714442614,
+      "grad_norm": 0.3664883077144623,
+      "learning_rate": 0.00018420195439739414,
+      "loss": 1.7507,
+      "step": 611
+    },
+    {
+      "epoch": 1.3443163097199342,
+      "grad_norm": 0.27566832304000854,
+      "learning_rate": 0.00018395765472312703,
+      "loss": 1.3476,
+      "step": 612
+    },
+    {
+      "epoch": 1.3465129049972542,
+      "grad_norm": 1.0763379335403442,
+      "learning_rate": 0.00018371335504885992,
+      "loss": 1.5265,
+      "step": 613
+    },
+    {
+      "epoch": 1.3487095002745744,
+      "grad_norm": 0.24601931869983673,
+      "learning_rate": 0.00018346905537459284,
+      "loss": 1.396,
+      "step": 614
+    },
+    {
+      "epoch": 1.3509060955518946,
+      "grad_norm": 0.3085857331752777,
+      "learning_rate": 0.00018322475570032573,
+      "loss": 1.6203,
+      "step": 615
+    },
+    {
+      "epoch": 1.3531026908292147,
+      "grad_norm": 0.25747713446617126,
+      "learning_rate": 0.00018298045602605864,
+      "loss": 1.5711,
+      "step": 616
+    },
+    {
+      "epoch": 1.355299286106535,
+      "grad_norm": 0.19672615826129913,
+      "learning_rate": 0.00018273615635179153,
+      "loss": 1.2685,
+      "step": 617
+    },
+    {
+      "epoch": 1.357495881383855,
+      "grad_norm": 0.2396446019411087,
+      "learning_rate": 0.00018249185667752442,
+      "loss": 1.8329,
+      "step": 618
+    },
+    {
+      "epoch": 1.359692476661175,
+      "grad_norm": 0.37154537439346313,
+      "learning_rate": 0.00018224755700325734,
+      "loss": 1.3417,
+      "step": 619
+    },
+    {
+      "epoch": 1.3618890719384953,
+      "grad_norm": 0.1998070329427719,
+      "learning_rate": 0.00018200325732899023,
+      "loss": 1.315,
+      "step": 620
+    },
+    {
+      "epoch": 1.3640856672158155,
+      "grad_norm": 0.23325687646865845,
+      "learning_rate": 0.00018175895765472312,
+      "loss": 1.738,
+      "step": 621
+    },
+    {
+      "epoch": 1.3662822624931357,
+      "grad_norm": 0.17906086146831512,
+      "learning_rate": 0.00018151465798045603,
+      "loss": 1.1393,
+      "step": 622
+    },
+    {
+      "epoch": 1.3684788577704559,
+      "grad_norm": 0.208530992269516,
+      "learning_rate": 0.00018127035830618892,
+      "loss": 1.2864,
+      "step": 623
+    },
+    {
+      "epoch": 1.370675453047776,
+      "grad_norm": 0.18550193309783936,
+      "learning_rate": 0.0001810260586319218,
+      "loss": 1.3042,
+      "step": 624
+    },
+    {
+      "epoch": 1.372872048325096,
+      "grad_norm": 0.1899668276309967,
+      "learning_rate": 0.00018078175895765473,
+      "loss": 1.3217,
+      "step": 625
+    },
+    {
+      "epoch": 1.3750686436024162,
+      "grad_norm": 0.1339438259601593,
+      "learning_rate": 0.00018053745928338762,
+      "loss": 1.0858,
+      "step": 626
+    },
+    {
+      "epoch": 1.3772652388797364,
+      "grad_norm": 0.49044135212898254,
+      "learning_rate": 0.0001802931596091205,
+      "loss": 1.2837,
+      "step": 627
+    },
+    {
+      "epoch": 1.3794618341570566,
+      "grad_norm": 0.2338285893201828,
+      "learning_rate": 0.00018004885993485342,
+      "loss": 1.1011,
+      "step": 628
+    },
+    {
+      "epoch": 1.3816584294343768,
+      "grad_norm": 0.2337966412305832,
+      "learning_rate": 0.0001798045602605863,
+      "loss": 1.5607,
+      "step": 629
+    },
+    {
+      "epoch": 1.3838550247116967,
+      "grad_norm": 0.1881571263074875,
+      "learning_rate": 0.00017956026058631923,
+      "loss": 1.3885,
+      "step": 630
+    },
+    {
+      "epoch": 1.3860516199890172,
+      "grad_norm": 0.2737588584423065,
+      "learning_rate": 0.00017931596091205212,
+      "loss": 1.5026,
+      "step": 631
+    },
+    {
+      "epoch": 1.3882482152663371,
+      "grad_norm": 0.2206345945596695,
+      "learning_rate": 0.000179071661237785,
+      "loss": 1.637,
+      "step": 632
+    },
+    {
+      "epoch": 1.3904448105436573,
+      "grad_norm": 0.19553588330745697,
+      "learning_rate": 0.00017882736156351792,
+      "loss": 1.4522,
+      "step": 633
+    },
+    {
+      "epoch": 1.3926414058209775,
+      "grad_norm": 0.1784679889678955,
+      "learning_rate": 0.0001785830618892508,
+      "loss": 1.4278,
+      "step": 634
+    },
+    {
+      "epoch": 1.3948380010982977,
+      "grad_norm": 0.2899497449398041,
+      "learning_rate": 0.0001783387622149837,
+      "loss": 1.4458,
+      "step": 635
+    },
+    {
+      "epoch": 1.3970345963756179,
+      "grad_norm": 0.9294266700744629,
+      "learning_rate": 0.00017809446254071662,
+      "loss": 1.4051,
+      "step": 636
+    },
+    {
+      "epoch": 1.3992311916529379,
+      "grad_norm": 0.44814422726631165,
+      "learning_rate": 0.0001778501628664495,
+      "loss": 1.352,
+      "step": 637
+    },
+    {
+      "epoch": 1.401427786930258,
+      "grad_norm": 0.17504063248634338,
+      "learning_rate": 0.0001776058631921824,
+      "loss": 1.2403,
+      "step": 638
+    },
+    {
+      "epoch": 1.4036243822075782,
+      "grad_norm": 0.2671871781349182,
+      "learning_rate": 0.0001773615635179153,
+      "loss": 1.5047,
+      "step": 639
+    },
+    {
+      "epoch": 1.4058209774848984,
+      "grad_norm": 0.2416955679655075,
+      "learning_rate": 0.0001771172638436482,
+      "loss": 1.7912,
+      "step": 640
+    },
+    {
+      "epoch": 1.4080175727622186,
+      "grad_norm": 0.31299084424972534,
+      "learning_rate": 0.0001768729641693811,
+      "loss": 1.5149,
+      "step": 641
+    },
+    {
+      "epoch": 1.4102141680395386,
+      "grad_norm": 0.1942817121744156,
+      "learning_rate": 0.000176628664495114,
+      "loss": 1.5914,
+      "step": 642
+    },
+    {
+      "epoch": 1.412410763316859,
+      "grad_norm": 1.0172524452209473,
+      "learning_rate": 0.0001763843648208469,
+      "loss": 1.5636,
+      "step": 643
+    },
+    {
+      "epoch": 1.414607358594179,
+      "grad_norm": 0.3055048882961273,
+      "learning_rate": 0.0001761400651465798,
+      "loss": 1.4037,
+      "step": 644
+    },
+    {
+      "epoch": 1.4168039538714992,
+      "grad_norm": 0.942116916179657,
+      "learning_rate": 0.0001758957654723127,
+      "loss": 1.4638,
+      "step": 645
+    },
+    {
+      "epoch": 1.4190005491488193,
+      "grad_norm": 0.20707891881465912,
+      "learning_rate": 0.0001756514657980456,
+      "loss": 1.4504,
+      "step": 646
+    },
+    {
+      "epoch": 1.4211971444261395,
+      "grad_norm": 0.2763131856918335,
+      "learning_rate": 0.0001754071661237785,
+      "loss": 1.7616,
+      "step": 647
+    },
+    {
+      "epoch": 1.4233937397034597,
+      "grad_norm": 0.3040972054004669,
+      "learning_rate": 0.0001751628664495114,
+      "loss": 1.7308,
+      "step": 648
+    },
+    {
+      "epoch": 1.4255903349807797,
+      "grad_norm": 0.23299454152584076,
+      "learning_rate": 0.00017491856677524429,
+      "loss": 1.296,
+      "step": 649
+    },
+    {
+      "epoch": 1.4277869302580999,
+      "grad_norm": 0.17174780368804932,
+      "learning_rate": 0.0001746742671009772,
+      "loss": 1.0778,
+      "step": 650
+    },
+    {
+      "epoch": 1.42998352553542,
+      "grad_norm": 0.24531452357769012,
+      "learning_rate": 0.0001744299674267101,
+      "loss": 1.4582,
+      "step": 651
+    },
+    {
+      "epoch": 1.4321801208127403,
+      "grad_norm": 0.22440105676651,
+      "learning_rate": 0.00017418566775244298,
+      "loss": 1.5271,
+      "step": 652
+    },
+    {
+      "epoch": 1.4343767160900605,
+      "grad_norm": 0.18573051691055298,
+      "learning_rate": 0.0001739413680781759,
+      "loss": 1.4786,
+      "step": 653
+    },
+    {
+      "epoch": 1.4365733113673804,
+      "grad_norm": 0.3134148418903351,
+      "learning_rate": 0.00017369706840390879,
+      "loss": 1.7616,
+      "step": 654
+    },
+    {
+      "epoch": 1.4387699066447008,
+      "grad_norm": 1.014664649963379,
+      "learning_rate": 0.0001734527687296417,
+      "loss": 1.1514,
+      "step": 655
+    },
+    {
+      "epoch": 1.4409665019220208,
+      "grad_norm": 0.2553189992904663,
+      "learning_rate": 0.0001732084690553746,
+      "loss": 1.5287,
+      "step": 656
+    },
+    {
+      "epoch": 1.443163097199341,
+      "grad_norm": 0.2545950710773468,
+      "learning_rate": 0.00017296416938110748,
+      "loss": 1.4425,
+      "step": 657
+    },
+    {
+      "epoch": 1.4453596924766612,
+      "grad_norm": 1.5353608131408691,
+      "learning_rate": 0.0001727198697068404,
+      "loss": 1.9844,
+      "step": 658
+    },
+    {
+      "epoch": 1.4475562877539814,
+      "grad_norm": 1.3327744007110596,
+      "learning_rate": 0.00017247557003257329,
+      "loss": 1.8319,
+      "step": 659
+    },
+    {
+      "epoch": 1.4497528830313016,
+      "grad_norm": 0.17347146570682526,
+      "learning_rate": 0.00017223127035830618,
+      "loss": 1.3216,
+      "step": 660
+    },
+    {
+      "epoch": 1.4519494783086215,
+      "grad_norm": 0.22073277831077576,
+      "learning_rate": 0.0001719869706840391,
+      "loss": 1.3429,
+      "step": 661
+    },
+    {
+      "epoch": 1.4541460735859417,
+      "grad_norm": 0.36095699667930603,
+      "learning_rate": 0.00017174267100977198,
+      "loss": 1.6225,
+      "step": 662
+    },
+    {
+      "epoch": 1.456342668863262,
+      "grad_norm": 0.24609845876693726,
+      "learning_rate": 0.00017149837133550487,
+      "loss": 1.7886,
+      "step": 663
+    },
+    {
+      "epoch": 1.458539264140582,
+      "grad_norm": 0.5629845857620239,
+      "learning_rate": 0.00017125407166123779,
+      "loss": 1.4474,
+      "step": 664
+    },
+    {
+      "epoch": 1.4607358594179023,
+      "grad_norm": 0.28364306688308716,
+      "learning_rate": 0.00017100977198697068,
+      "loss": 1.7474,
+      "step": 665
+    },
+    {
+      "epoch": 1.4629324546952225,
+      "grad_norm": 0.22777429223060608,
+      "learning_rate": 0.00017076547231270356,
+      "loss": 1.6925,
+      "step": 666
+    },
+    {
+      "epoch": 1.4651290499725427,
+      "grad_norm": 0.19162385165691376,
+      "learning_rate": 0.00017052117263843648,
+      "loss": 1.2993,
+      "step": 667
+    },
+    {
+      "epoch": 1.4673256452498626,
+      "grad_norm": 0.402773380279541,
+      "learning_rate": 0.00017027687296416937,
+      "loss": 1.7588,
+      "step": 668
+    },
+    {
+      "epoch": 1.4695222405271828,
+      "grad_norm": 0.3195705711841583,
+      "learning_rate": 0.00017003257328990229,
+      "loss": 1.3975,
+      "step": 669
+    },
+    {
+      "epoch": 1.471718835804503,
+      "grad_norm": 0.21924123167991638,
+      "learning_rate": 0.00016978827361563518,
+      "loss": 1.4354,
+      "step": 670
+    },
+    {
+      "epoch": 1.4739154310818232,
+      "grad_norm": 0.4594910442829132,
+      "learning_rate": 0.00016954397394136806,
+      "loss": 1.6491,
+      "step": 671
+    },
+    {
+      "epoch": 1.4761120263591434,
+      "grad_norm": 0.6204549670219421,
+      "learning_rate": 0.00016929967426710098,
+      "loss": 1.6497,
+      "step": 672
+    },
+    {
+      "epoch": 1.4783086216364634,
+      "grad_norm": 0.21799476444721222,
+      "learning_rate": 0.00016905537459283387,
+      "loss": 1.3056,
+      "step": 673
+    },
+    {
+      "epoch": 1.4805052169137836,
+      "grad_norm": 0.23561517894268036,
+      "learning_rate": 0.00016881107491856676,
+      "loss": 1.6601,
+      "step": 674
+    },
+    {
+      "epoch": 1.4827018121911038,
+      "grad_norm": 0.20881032943725586,
+      "learning_rate": 0.00016856677524429968,
+      "loss": 1.3232,
+      "step": 675
+    },
+    {
+      "epoch": 1.484898407468424,
+      "grad_norm": 0.22069154679775238,
+      "learning_rate": 0.00016832247557003257,
+      "loss": 1.5653,
+      "step": 676
+    },
+    {
+      "epoch": 1.4870950027457441,
+      "grad_norm": 0.20432431995868683,
+      "learning_rate": 0.00016807817589576545,
+      "loss": 1.2535,
+      "step": 677
+    },
+    {
+      "epoch": 1.4892915980230643,
+      "grad_norm": 0.23453666269779205,
+      "learning_rate": 0.00016783387622149837,
+      "loss": 1.7483,
+      "step": 678
+    },
+    {
+      "epoch": 1.4914881933003845,
+      "grad_norm": 0.3612651824951172,
+      "learning_rate": 0.00016758957654723126,
+      "loss": 1.5832,
+      "step": 679
+    },
+    {
+      "epoch": 1.4936847885777045,
+      "grad_norm": 0.2592422366142273,
+      "learning_rate": 0.00016734527687296415,
+      "loss": 1.5452,
+      "step": 680
+    },
+    {
+      "epoch": 1.4958813838550247,
+      "grad_norm": 0.28201788663864136,
+      "learning_rate": 0.00016710097719869707,
+      "loss": 1.657,
+      "step": 681
+    },
+    {
+      "epoch": 1.4980779791323449,
+      "grad_norm": 0.24135351181030273,
+      "learning_rate": 0.00016685667752442995,
+      "loss": 1.4105,
+      "step": 682
+    },
+    {
+      "epoch": 1.500274574409665,
+      "grad_norm": 0.3412753641605377,
+      "learning_rate": 0.00016661237785016287,
+      "loss": 1.4407,
+      "step": 683
+    },
+    {
+      "epoch": 1.5024711696869852,
+      "grad_norm": 0.23166997730731964,
+      "learning_rate": 0.00016636807817589576,
+      "loss": 1.1341,
+      "step": 684
+    },
+    {
+      "epoch": 1.5046677649643052,
+      "grad_norm": 0.22654154896736145,
+      "learning_rate": 0.00016612377850162865,
+      "loss": 1.6125,
+      "step": 685
+    },
+    {
+      "epoch": 1.5068643602416256,
+      "grad_norm": 0.8772279024124146,
+      "learning_rate": 0.00016587947882736157,
+      "loss": 2.0432,
+      "step": 686
+    },
+    {
+      "epoch": 1.5090609555189456,
+      "grad_norm": 0.27497851848602295,
+      "learning_rate": 0.00016563517915309445,
+      "loss": 1.634,
+      "step": 687
+    },
+    {
+      "epoch": 1.5112575507962658,
+      "grad_norm": 0.14097030460834503,
+      "learning_rate": 0.00016539087947882734,
+      "loss": 1.3311,
+      "step": 688
+    },
+    {
+      "epoch": 1.513454146073586,
+      "grad_norm": 0.18364806473255157,
+      "learning_rate": 0.00016514657980456026,
+      "loss": 1.2961,
+      "step": 689
+    },
+    {
+      "epoch": 1.515650741350906,
+      "grad_norm": 0.1956978291273117,
+      "learning_rate": 0.00016490228013029315,
+      "loss": 1.1849,
+      "step": 690
+    },
+    {
+      "epoch": 1.5178473366282264,
+      "grad_norm": 1.7431960105895996,
+      "learning_rate": 0.00016465798045602604,
+      "loss": 1.7878,
+      "step": 691
+    },
+    {
+      "epoch": 1.5200439319055463,
+      "grad_norm": 0.3108193874359131,
+      "learning_rate": 0.00016441368078175895,
+      "loss": 1.6192,
+      "step": 692
+    },
+    {
+      "epoch": 1.5222405271828665,
+      "grad_norm": 0.4347732961177826,
+      "learning_rate": 0.00016416938110749184,
+      "loss": 1.4793,
+      "step": 693
+    },
+    {
+      "epoch": 1.5244371224601867,
+      "grad_norm": 0.28937646746635437,
+      "learning_rate": 0.00016392508143322476,
+      "loss": 1.6244,
+      "step": 694
+    },
+    {
+      "epoch": 1.526633717737507,
+      "grad_norm": 0.2161215990781784,
+      "learning_rate": 0.00016368078175895765,
+      "loss": 1.479,
+      "step": 695
+    },
+    {
+      "epoch": 1.528830313014827,
+      "grad_norm": 0.19767840206623077,
+      "learning_rate": 0.00016343648208469054,
+      "loss": 1.418,
+      "step": 696
+    },
+    {
+      "epoch": 1.531026908292147,
+      "grad_norm": 0.24696283042430878,
+      "learning_rate": 0.00016319218241042346,
+      "loss": 1.6182,
+      "step": 697
+    },
+    {
+      "epoch": 1.5332235035694675,
+      "grad_norm": 0.22354716062545776,
+      "learning_rate": 0.00016294788273615634,
+      "loss": 1.4676,
+      "step": 698
+    },
+    {
+      "epoch": 1.5354200988467874,
+      "grad_norm": 0.24923834204673767,
+      "learning_rate": 0.00016270358306188923,
+      "loss": 1.4434,
+      "step": 699
+    },
+    {
+      "epoch": 1.5376166941241076,
+      "grad_norm": 0.2457156628370285,
+      "learning_rate": 0.00016245928338762215,
+      "loss": 1.4381,
+      "step": 700
+    },
+    {
+      "epoch": 1.5398132894014278,
+      "grad_norm": 0.19205059111118317,
+      "learning_rate": 0.00016221498371335504,
+      "loss": 1.4332,
+      "step": 701
+    },
+    {
+      "epoch": 1.5420098846787478,
+      "grad_norm": 0.28883326053619385,
+      "learning_rate": 0.00016197068403908793,
+      "loss": 1.6975,
+      "step": 702
+    },
+    {
+      "epoch": 1.5442064799560682,
+      "grad_norm": 0.2536971867084503,
+      "learning_rate": 0.00016172638436482084,
+      "loss": 1.4172,
+      "step": 703
+    },
+    {
+      "epoch": 1.5464030752333882,
+      "grad_norm": 0.20173028111457825,
+      "learning_rate": 0.00016148208469055373,
+      "loss": 1.2212,
+      "step": 704
+    },
+    {
+      "epoch": 1.5485996705107083,
+      "grad_norm": 0.28440260887145996,
+      "learning_rate": 0.00016123778501628662,
+      "loss": 1.4696,
+      "step": 705
+    },
+    {
+      "epoch": 1.5507962657880285,
+      "grad_norm": 0.30404359102249146,
+      "learning_rate": 0.00016099348534201954,
+      "loss": 1.5108,
+      "step": 706
+    },
+    {
+      "epoch": 1.5529928610653487,
+      "grad_norm": 0.31012529134750366,
+      "learning_rate": 0.00016074918566775243,
+      "loss": 1.4227,
+      "step": 707
+    },
+    {
+      "epoch": 1.555189456342669,
+      "grad_norm": 0.22491522133350372,
+      "learning_rate": 0.00016050488599348534,
+      "loss": 1.406,
+      "step": 708
+    },
+    {
+      "epoch": 1.5573860516199889,
+      "grad_norm": 0.3146182894706726,
+      "learning_rate": 0.00016026058631921823,
+      "loss": 1.7731,
+      "step": 709
+    },
+    {
+      "epoch": 1.5595826468973093,
+      "grad_norm": 0.2549310326576233,
+      "learning_rate": 0.00016001628664495112,
+      "loss": 1.3538,
+      "step": 710
+    },
+    {
+      "epoch": 1.5617792421746293,
+      "grad_norm": 0.33023861050605774,
+      "learning_rate": 0.00015977198697068404,
+      "loss": 1.7219,
+      "step": 711
+    },
+    {
+      "epoch": 1.5639758374519495,
+      "grad_norm": 0.6675399541854858,
+      "learning_rate": 0.00015952768729641693,
+      "loss": 1.3225,
+      "step": 712
+    },
+    {
+      "epoch": 1.5661724327292696,
+      "grad_norm": 0.6012052893638611,
+      "learning_rate": 0.00015928338762214982,
+      "loss": 1.1211,
+      "step": 713
+    },
+    {
+      "epoch": 1.5683690280065898,
+      "grad_norm": 0.3987830579280853,
+      "learning_rate": 0.00015903908794788273,
+      "loss": 1.326,
+      "step": 714
+    },
+    {
+      "epoch": 1.57056562328391,
+      "grad_norm": 0.25261083245277405,
+      "learning_rate": 0.00015879478827361562,
+      "loss": 1.3131,
+      "step": 715
+    },
+    {
+      "epoch": 1.57276221856123,
+      "grad_norm": 0.2668202221393585,
+      "learning_rate": 0.0001585504885993485,
+      "loss": 1.7305,
+      "step": 716
+    },
+    {
+      "epoch": 1.5749588138385504,
+      "grad_norm": 0.25820016860961914,
+      "learning_rate": 0.00015830618892508143,
+      "loss": 1.7721,
+      "step": 717
+    },
+    {
+      "epoch": 1.5771554091158704,
+      "grad_norm": 0.2607474625110626,
+      "learning_rate": 0.00015806188925081432,
+      "loss": 1.4969,
+      "step": 718
+    },
+    {
+      "epoch": 1.5793520043931906,
+      "grad_norm": 0.22891727089881897,
+      "learning_rate": 0.00015781758957654723,
+      "loss": 1.5416,
+      "step": 719
+    },
+    {
+      "epoch": 1.5815485996705108,
+      "grad_norm": 0.22733044624328613,
+      "learning_rate": 0.00015757328990228012,
+      "loss": 1.4785,
+      "step": 720
+    },
+    {
+      "epoch": 1.5837451949478307,
+      "grad_norm": 0.28062447905540466,
+      "learning_rate": 0.000157328990228013,
+      "loss": 1.4515,
+      "step": 721
+    },
+    {
+      "epoch": 1.5859417902251511,
+      "grad_norm": 0.3594748377799988,
+      "learning_rate": 0.00015708469055374593,
+      "loss": 1.4193,
+      "step": 722
+    },
+    {
+      "epoch": 1.588138385502471,
+      "grad_norm": 0.3683021664619446,
+      "learning_rate": 0.00015684039087947882,
+      "loss": 1.3909,
+      "step": 723
+    },
+    {
+      "epoch": 1.5903349807797913,
+      "grad_norm": 0.18952369689941406,
+      "learning_rate": 0.0001565960912052117,
+      "loss": 1.2441,
+      "step": 724
+    },
+    {
+      "epoch": 1.5925315760571115,
+      "grad_norm": 0.23505713045597076,
+      "learning_rate": 0.00015635179153094462,
+      "loss": 1.5244,
+      "step": 725
+    },
+    {
+      "epoch": 1.5947281713344317,
+      "grad_norm": 0.7234029173851013,
+      "learning_rate": 0.0001561074918566775,
+      "loss": 1.4296,
+      "step": 726
+    },
+    {
+      "epoch": 1.5969247666117519,
+      "grad_norm": 0.282888263463974,
+      "learning_rate": 0.0001558631921824104,
+      "loss": 1.3513,
+      "step": 727
+    },
+    {
+      "epoch": 1.5991213618890718,
+      "grad_norm": 0.2394958883523941,
+      "learning_rate": 0.00015561889250814332,
+      "loss": 1.6444,
+      "step": 728
+    },
+    {
+      "epoch": 1.6013179571663922,
+      "grad_norm": 0.2930229902267456,
+      "learning_rate": 0.0001553745928338762,
+      "loss": 1.668,
+      "step": 729
+    },
+    {
+      "epoch": 1.6035145524437122,
+      "grad_norm": 0.6005991697311401,
+      "learning_rate": 0.0001551302931596091,
+      "loss": 1.4115,
+      "step": 730
+    },
+    {
+      "epoch": 1.6057111477210324,
+      "grad_norm": 0.18858914077281952,
+      "learning_rate": 0.000154885993485342,
+      "loss": 1.2548,
+      "step": 731
+    },
+    {
+      "epoch": 1.6079077429983526,
+      "grad_norm": 0.503218412399292,
+      "learning_rate": 0.0001546416938110749,
+      "loss": 1.3913,
+      "step": 732
+    },
+    {
+      "epoch": 1.6101043382756726,
+      "grad_norm": 0.19885681569576263,
+      "learning_rate": 0.00015439739413680782,
+      "loss": 1.5165,
+      "step": 733
+    },
+    {
+      "epoch": 1.612300933552993,
+      "grad_norm": 0.3013475835323334,
+      "learning_rate": 0.0001541530944625407,
+      "loss": 1.9134,
+      "step": 734
+    },
+    {
+      "epoch": 1.614497528830313,
+      "grad_norm": 0.21525722742080688,
+      "learning_rate": 0.0001539087947882736,
+      "loss": 1.2914,
+      "step": 735
+    },
+    {
+      "epoch": 1.6166941241076331,
+      "grad_norm": 0.19655828177928925,
+      "learning_rate": 0.0001536644951140065,
+      "loss": 1.5374,
+      "step": 736
+    },
+    {
+      "epoch": 1.6188907193849533,
+      "grad_norm": 0.292644739151001,
+      "learning_rate": 0.0001534201954397394,
+      "loss": 1.6444,
+      "step": 737
+    },
+    {
+      "epoch": 1.6210873146622735,
+      "grad_norm": 0.3042570650577545,
+      "learning_rate": 0.0001531758957654723,
+      "loss": 1.3042,
+      "step": 738
+    },
+    {
+      "epoch": 1.6232839099395937,
+      "grad_norm": 0.23533335328102112,
+      "learning_rate": 0.0001529315960912052,
+      "loss": 1.6056,
+      "step": 739
+    },
+    {
+      "epoch": 1.6254805052169137,
+      "grad_norm": 1.8164302110671997,
+      "learning_rate": 0.0001526872964169381,
+      "loss": 1.4015,
+      "step": 740
+    },
+    {
+      "epoch": 1.627677100494234,
+      "grad_norm": 7.149118900299072,
+      "learning_rate": 0.00015244299674267099,
+      "loss": 1.6312,
+      "step": 741
+    },
+    {
+      "epoch": 1.629873695771554,
+      "grad_norm": 0.1733080893754959,
+      "learning_rate": 0.0001521986970684039,
+      "loss": 1.1419,
+      "step": 742
+    },
+    {
+      "epoch": 1.6320702910488742,
+      "grad_norm": 0.22565101087093353,
+      "learning_rate": 0.0001519543973941368,
+      "loss": 1.2696,
+      "step": 743
+    },
+    {
+      "epoch": 1.6342668863261944,
+      "grad_norm": 1.1321609020233154,
+      "learning_rate": 0.00015171009771986968,
+      "loss": 1.1181,
+      "step": 744
+    },
+    {
+      "epoch": 1.6364634816035144,
+      "grad_norm": 0.6189210414886475,
+      "learning_rate": 0.0001514657980456026,
+      "loss": 1.3999,
+      "step": 745
+    },
+    {
+      "epoch": 1.6386600768808348,
+      "grad_norm": 1.009063482284546,
+      "learning_rate": 0.0001512214983713355,
+      "loss": 1.3277,
+      "step": 746
+    },
+    {
+      "epoch": 1.6408566721581548,
+      "grad_norm": 0.6081856489181519,
+      "learning_rate": 0.0001509771986970684,
+      "loss": 1.2188,
+      "step": 747
+    },
+    {
+      "epoch": 1.643053267435475,
+      "grad_norm": 0.2678440809249878,
+      "learning_rate": 0.0001507328990228013,
+      "loss": 1.5623,
+      "step": 748
+    },
+    {
+      "epoch": 1.6452498627127952,
+      "grad_norm": 0.2937515079975128,
+      "learning_rate": 0.00015048859934853418,
+      "loss": 1.4267,
+      "step": 749
+    },
+    {
+      "epoch": 1.6474464579901154,
+      "grad_norm": 0.19985835254192352,
+      "learning_rate": 0.0001502442996742671,
+      "loss": 1.392,
+      "step": 750
+    },
+    {
+      "epoch": 1.6496430532674355,
+      "grad_norm": 0.27106982469558716,
+      "learning_rate": 0.00015,
+      "loss": 1.5973,
+      "step": 751
+    },
+    {
+      "epoch": 1.6518396485447555,
+      "grad_norm": 0.27762484550476074,
+      "learning_rate": 0.00014975570032573288,
+      "loss": 1.4842,
+      "step": 752
+    },
+    {
+      "epoch": 1.654036243822076,
+      "grad_norm": 0.20539994537830353,
+      "learning_rate": 0.0001495114006514658,
+      "loss": 1.5102,
+      "step": 753
+    },
+    {
+      "epoch": 1.656232839099396,
+      "grad_norm": 0.2649374008178711,
+      "learning_rate": 0.00014926710097719868,
+      "loss": 1.1589,
+      "step": 754
+    },
+    {
+      "epoch": 1.658429434376716,
+      "grad_norm": 0.24394552409648895,
+      "learning_rate": 0.00014902280130293157,
+      "loss": 1.3184,
+      "step": 755
+    },
+    {
+      "epoch": 1.6606260296540363,
+      "grad_norm": 0.2517681121826172,
+      "learning_rate": 0.0001487785016286645,
+      "loss": 1.6584,
+      "step": 756
+    },
+    {
+      "epoch": 1.6628226249313562,
+      "grad_norm": 0.23089078068733215,
+      "learning_rate": 0.00014853420195439738,
+      "loss": 1.5137,
+      "step": 757
+    },
+    {
+      "epoch": 1.6650192202086767,
+      "grad_norm": 0.3136192560195923,
+      "learning_rate": 0.0001482899022801303,
+      "loss": 1.6885,
+      "step": 758
+    },
+    {
+      "epoch": 1.6672158154859966,
+      "grad_norm": 0.24763253331184387,
+      "learning_rate": 0.00014804560260586318,
+      "loss": 1.5538,
+      "step": 759
+    },
+    {
+      "epoch": 1.6694124107633168,
+      "grad_norm": 0.2483770102262497,
+      "learning_rate": 0.00014780130293159607,
+      "loss": 1.5205,
+      "step": 760
+    },
+    {
+      "epoch": 1.671609006040637,
+      "grad_norm": 0.2168157696723938,
+      "learning_rate": 0.000147557003257329,
+      "loss": 1.5632,
+      "step": 761
+    },
+    {
+      "epoch": 1.6738056013179572,
+      "grad_norm": 0.2264491617679596,
+      "learning_rate": 0.00014731270358306188,
+      "loss": 1.4893,
+      "step": 762
+    },
+    {
+      "epoch": 1.6760021965952774,
+      "grad_norm": 3.2434957027435303,
+      "learning_rate": 0.00014706840390879477,
+      "loss": 1.4821,
+      "step": 763
+    },
+    {
+      "epoch": 1.6781987918725974,
+      "grad_norm": 0.16092146933078766,
+      "learning_rate": 0.00014682410423452768,
+      "loss": 1.4758,
+      "step": 764
+    },
+    {
+      "epoch": 1.6803953871499178,
+      "grad_norm": 1.5099763870239258,
+      "learning_rate": 0.00014657980456026057,
+      "loss": 1.7472,
+      "step": 765
+    },
+    {
+      "epoch": 1.6825919824272377,
+      "grad_norm": 0.26016733050346375,
+      "learning_rate": 0.00014633550488599346,
+      "loss": 1.4812,
+      "step": 766
+    },
+    {
+      "epoch": 1.684788577704558,
+      "grad_norm": 0.7335305213928223,
+      "learning_rate": 0.00014609120521172638,
+      "loss": 1.6143,
+      "step": 767
+    },
+    {
+      "epoch": 1.6869851729818781,
+      "grad_norm": 0.23457981646060944,
+      "learning_rate": 0.00014584690553745927,
+      "loss": 1.5153,
+      "step": 768
+    },
+    {
+      "epoch": 1.689181768259198,
+      "grad_norm": 0.2525008022785187,
+      "learning_rate": 0.00014560260586319216,
+      "loss": 1.2015,
+      "step": 769
+    },
+    {
+      "epoch": 1.6913783635365185,
+      "grad_norm": 0.20113202929496765,
+      "learning_rate": 0.00014535830618892507,
+      "loss": 1.4268,
+      "step": 770
+    },
+    {
+      "epoch": 1.6935749588138385,
+      "grad_norm": 0.1633341759443283,
+      "learning_rate": 0.00014511400651465796,
+      "loss": 1.3866,
+      "step": 771
+    },
+    {
+      "epoch": 1.6957715540911587,
+      "grad_norm": 0.2505511939525604,
+      "learning_rate": 0.00014486970684039088,
+      "loss": 1.4497,
+      "step": 772
+    },
+    {
+      "epoch": 1.6979681493684788,
+      "grad_norm": 0.16012145578861237,
+      "learning_rate": 0.00014462540716612377,
+      "loss": 1.1285,
+      "step": 773
+    },
+    {
+      "epoch": 1.700164744645799,
+      "grad_norm": 0.207519993185997,
+      "learning_rate": 0.00014438110749185666,
+      "loss": 1.2798,
+      "step": 774
+    },
+    {
+      "epoch": 1.7023613399231192,
+      "grad_norm": 0.2635665237903595,
+      "learning_rate": 0.00014413680781758957,
+      "loss": 1.4949,
+      "step": 775
+    },
+    {
+      "epoch": 1.7045579352004392,
+      "grad_norm": 0.26190510392189026,
+      "learning_rate": 0.00014389250814332246,
+      "loss": 1.0798,
+      "step": 776
+    },
+    {
+      "epoch": 1.7067545304777596,
+      "grad_norm": 0.14722655713558197,
+      "learning_rate": 0.00014364820846905535,
+      "loss": 1.2883,
+      "step": 777
+    },
+    {
+      "epoch": 1.7089511257550796,
+      "grad_norm": 0.4619833528995514,
+      "learning_rate": 0.00014340390879478827,
+      "loss": 1.7755,
+      "step": 778
+    },
+    {
+      "epoch": 1.7111477210323998,
+      "grad_norm": 0.26691508293151855,
+      "learning_rate": 0.00014315960912052116,
+      "loss": 1.6575,
+      "step": 779
+    },
+    {
+      "epoch": 1.71334431630972,
+      "grad_norm": 0.22218242287635803,
+      "learning_rate": 0.00014291530944625404,
+      "loss": 1.4758,
+      "step": 780
+    },
+    {
+      "epoch": 1.7155409115870401,
+      "grad_norm": 1.4821361303329468,
+      "learning_rate": 0.00014267100977198696,
+      "loss": 1.4376,
+      "step": 781
+    },
+    {
+      "epoch": 1.7177375068643603,
+      "grad_norm": 0.23621997237205505,
+      "learning_rate": 0.00014242671009771985,
+      "loss": 1.5235,
+      "step": 782
+    },
+    {
+      "epoch": 1.7199341021416803,
+      "grad_norm": 0.23057015240192413,
+      "learning_rate": 0.00014218241042345274,
+      "loss": 1.6025,
+      "step": 783
+    },
+    {
+      "epoch": 1.7221306974190007,
+      "grad_norm": 0.42340490221977234,
+      "learning_rate": 0.00014193811074918566,
+      "loss": 1.347,
+      "step": 784
+    },
+    {
+      "epoch": 1.7243272926963207,
+      "grad_norm": 0.16306892037391663,
+      "learning_rate": 0.00014169381107491854,
+      "loss": 1.6087,
+      "step": 785
+    },
+    {
+      "epoch": 1.7265238879736409,
+      "grad_norm": 0.48266735672950745,
+      "learning_rate": 0.00014144951140065146,
+      "loss": 1.3599,
+      "step": 786
+    },
+    {
+      "epoch": 1.728720483250961,
+      "grad_norm": 0.24362435936927795,
+      "learning_rate": 0.00014120521172638435,
+      "loss": 1.384,
+      "step": 787
+    },
+    {
+      "epoch": 1.730917078528281,
+      "grad_norm": 0.21303275227546692,
+      "learning_rate": 0.00014096091205211724,
+      "loss": 1.312,
+      "step": 788
+    },
+    {
+      "epoch": 1.7331136738056014,
+      "grad_norm": 0.224172443151474,
+      "learning_rate": 0.00014071661237785016,
+      "loss": 1.5765,
+      "step": 789
+    },
+    {
+      "epoch": 1.7353102690829214,
+      "grad_norm": 0.25218847393989563,
+      "learning_rate": 0.00014047231270358305,
+      "loss": 1.5931,
+      "step": 790
+    },
+    {
+      "epoch": 1.7375068643602416,
+      "grad_norm": 0.3345753848552704,
+      "learning_rate": 0.00014022801302931593,
+      "loss": 1.7352,
+      "step": 791
+    },
+    {
+      "epoch": 1.7397034596375618,
+      "grad_norm": 0.15458346903324127,
+      "learning_rate": 0.00013998371335504885,
+      "loss": 1.0935,
+      "step": 792
+    },
+    {
+      "epoch": 1.741900054914882,
+      "grad_norm": 0.24607662856578827,
+      "learning_rate": 0.00013973941368078174,
+      "loss": 1.3351,
+      "step": 793
+    },
+    {
+      "epoch": 1.7440966501922022,
+      "grad_norm": 0.19723249971866608,
+      "learning_rate": 0.00013949511400651463,
+      "loss": 1.5731,
+      "step": 794
+    },
+    {
+      "epoch": 1.7462932454695221,
+      "grad_norm": 0.3881778120994568,
+      "learning_rate": 0.00013925081433224755,
+      "loss": 1.3905,
+      "step": 795
+    },
+    {
+      "epoch": 1.7484898407468425,
+      "grad_norm": 0.28490278124809265,
+      "learning_rate": 0.00013900651465798043,
+      "loss": 1.4801,
+      "step": 796
+    },
+    {
+      "epoch": 1.7506864360241625,
+      "grad_norm": 0.4950120747089386,
+      "learning_rate": 0.00013876221498371335,
+      "loss": 1.5029,
+      "step": 797
+    },
+    {
+      "epoch": 1.7528830313014827,
+      "grad_norm": 0.22113639116287231,
+      "learning_rate": 0.00013851791530944624,
+      "loss": 1.2736,
+      "step": 798
+    },
+    {
+      "epoch": 1.755079626578803,
+      "grad_norm": 0.2574387788772583,
+      "learning_rate": 0.00013827361563517913,
+      "loss": 1.5685,
+      "step": 799
+    },
+    {
+      "epoch": 1.7572762218561229,
+      "grad_norm": 0.19240427017211914,
+      "learning_rate": 0.00013802931596091205,
+      "loss": 1.6484,
+      "step": 800
+    },
+    {
+      "epoch": 1.7594728171334433,
+      "grad_norm": 0.4578224718570709,
+      "learning_rate": 0.00013778501628664493,
+      "loss": 1.4598,
+      "step": 801
+    },
+    {
+      "epoch": 1.7616694124107632,
+      "grad_norm": 0.24641212821006775,
+      "learning_rate": 0.00013754071661237782,
+      "loss": 1.8339,
+      "step": 802
+    },
+    {
+      "epoch": 1.7638660076880834,
+      "grad_norm": 0.27870142459869385,
+      "learning_rate": 0.00013729641693811074,
+      "loss": 1.758,
+      "step": 803
+    },
+    {
+      "epoch": 1.7660626029654036,
+      "grad_norm": 0.23762951791286469,
+      "learning_rate": 0.00013705211726384363,
+      "loss": 1.7083,
+      "step": 804
+    },
+    {
+      "epoch": 1.7682591982427238,
+      "grad_norm": 0.40864840149879456,
+      "learning_rate": 0.00013680781758957652,
+      "loss": 1.4738,
+      "step": 805
+    },
+    {
+      "epoch": 1.770455793520044,
+      "grad_norm": 0.22173675894737244,
+      "learning_rate": 0.00013656351791530943,
+      "loss": 1.4292,
+      "step": 806
+    },
+    {
+      "epoch": 1.772652388797364,
+      "grad_norm": 0.2720279097557068,
+      "learning_rate": 0.00013631921824104232,
+      "loss": 1.4522,
+      "step": 807
+    },
+    {
+      "epoch": 1.7748489840746844,
+      "grad_norm": 0.3518648147583008,
+      "learning_rate": 0.0001360749185667752,
+      "loss": 1.728,
+      "step": 808
+    },
+    {
+      "epoch": 1.7770455793520044,
+      "grad_norm": 0.18037134408950806,
+      "learning_rate": 0.00013583061889250813,
+      "loss": 1.3854,
+      "step": 809
+    },
+    {
+      "epoch": 1.7792421746293245,
+      "grad_norm": 0.27808713912963867,
+      "learning_rate": 0.00013558631921824102,
+      "loss": 1.079,
+      "step": 810
+    },
+    {
+      "epoch": 1.7814387699066447,
+      "grad_norm": 0.2400078922510147,
+      "learning_rate": 0.00013534201954397394,
+      "loss": 1.0406,
+      "step": 811
+    },
+    {
+      "epoch": 1.7836353651839647,
+      "grad_norm": 0.27571937441825867,
+      "learning_rate": 0.00013509771986970682,
+      "loss": 1.5092,
+      "step": 812
+    },
+    {
+      "epoch": 1.7858319604612851,
+      "grad_norm": 0.7129433155059814,
+      "learning_rate": 0.00013485342019543971,
+      "loss": 1.6137,
+      "step": 813
+    },
+    {
+      "epoch": 1.788028555738605,
+      "grad_norm": 0.3234291970729828,
+      "learning_rate": 0.00013460912052117263,
+      "loss": 1.4143,
+      "step": 814
+    },
+    {
+      "epoch": 1.7902251510159253,
+      "grad_norm": 0.2988220155239105,
+      "learning_rate": 0.00013436482084690552,
+      "loss": 1.7778,
+      "step": 815
+    },
+    {
+      "epoch": 1.7924217462932455,
+      "grad_norm": 0.2366095632314682,
+      "learning_rate": 0.0001341205211726384,
+      "loss": 1.5418,
+      "step": 816
+    },
+    {
+      "epoch": 1.7946183415705657,
+      "grad_norm": 0.23005251586437225,
+      "learning_rate": 0.00013387622149837132,
+      "loss": 1.0746,
+      "step": 817
+    },
+    {
+      "epoch": 1.7968149368478858,
+      "grad_norm": 3.893672227859497,
+      "learning_rate": 0.00013363192182410421,
+      "loss": 1.6367,
+      "step": 818
+    },
+    {
+      "epoch": 1.7990115321252058,
+      "grad_norm": 0.2869424521923065,
+      "learning_rate": 0.0001333876221498371,
+      "loss": 1.4128,
+      "step": 819
+    },
+    {
+      "epoch": 1.8012081274025262,
+      "grad_norm": 0.23572219908237457,
+      "learning_rate": 0.00013314332247557002,
+      "loss": 1.6882,
+      "step": 820
+    },
+    {
+      "epoch": 1.8034047226798462,
+      "grad_norm": 0.17067575454711914,
+      "learning_rate": 0.0001328990228013029,
+      "loss": 1.3391,
+      "step": 821
+    },
+    {
+      "epoch": 1.8056013179571664,
+      "grad_norm": 0.20899467170238495,
+      "learning_rate": 0.00013265472312703582,
+      "loss": 1.714,
+      "step": 822
+    },
+    {
+      "epoch": 1.8077979132344866,
+      "grad_norm": 0.3168517053127289,
+      "learning_rate": 0.00013241042345276871,
+      "loss": 1.8192,
+      "step": 823
+    },
+    {
+      "epoch": 1.8099945085118065,
+      "grad_norm": 0.23785385489463806,
+      "learning_rate": 0.0001321661237785016,
+      "loss": 1.4456,
+      "step": 824
+    },
+    {
+      "epoch": 1.812191103789127,
+      "grad_norm": 0.8433808088302612,
+      "learning_rate": 0.00013192182410423452,
+      "loss": 1.0583,
+      "step": 825
+    },
+    {
+      "epoch": 1.814387699066447,
+      "grad_norm": 3.2449138164520264,
+      "learning_rate": 0.0001316775244299674,
+      "loss": 1.1829,
+      "step": 826
+    },
+    {
+      "epoch": 1.8165842943437671,
+      "grad_norm": 0.18850171566009521,
+      "learning_rate": 0.0001314332247557003,
+      "loss": 1.4288,
+      "step": 827
+    },
+    {
+      "epoch": 1.8187808896210873,
+      "grad_norm": 0.3165040612220764,
+      "learning_rate": 0.00013118892508143321,
+      "loss": 1.2868,
+      "step": 828
+    },
+    {
+      "epoch": 1.8209774848984075,
+      "grad_norm": 0.3435731530189514,
+      "learning_rate": 0.00013094462540716613,
+      "loss": 1.4644,
+      "step": 829
+    },
+    {
+      "epoch": 1.8231740801757277,
+      "grad_norm": 0.29969531297683716,
+      "learning_rate": 0.00013070032573289902,
+      "loss": 1.24,
+      "step": 830
+    },
+    {
+      "epoch": 1.8253706754530477,
+      "grad_norm": 0.1890360563993454,
+      "learning_rate": 0.0001304560260586319,
+      "loss": 1.0531,
+      "step": 831
+    },
+    {
+      "epoch": 1.827567270730368,
+      "grad_norm": 0.29781246185302734,
+      "learning_rate": 0.00013021172638436483,
+      "loss": 1.5197,
+      "step": 832
+    },
+    {
+      "epoch": 1.829763866007688,
+      "grad_norm": 0.18762893974781036,
+      "learning_rate": 0.00012996742671009771,
+      "loss": 1.3841,
+      "step": 833
+    },
+    {
+      "epoch": 1.8319604612850082,
+      "grad_norm": 0.2002384066581726,
+      "learning_rate": 0.0001297231270358306,
+      "loss": 1.6597,
+      "step": 834
+    },
+    {
+      "epoch": 1.8341570565623284,
+      "grad_norm": 0.2557729184627533,
+      "learning_rate": 0.00012947882736156352,
+      "loss": 1.7044,
+      "step": 835
+    },
+    {
+      "epoch": 1.8363536518396484,
+      "grad_norm": 0.2332870066165924,
+      "learning_rate": 0.0001292345276872964,
+      "loss": 1.2411,
+      "step": 836
+    },
+    {
+      "epoch": 1.8385502471169688,
+      "grad_norm": 0.2559221386909485,
+      "learning_rate": 0.00012899022801302933,
+      "loss": 1.4441,
+      "step": 837
+    },
+    {
+      "epoch": 1.8407468423942888,
+      "grad_norm": 0.2570485472679138,
+      "learning_rate": 0.00012874592833876221,
+      "loss": 1.4514,
+      "step": 838
+    },
+    {
+      "epoch": 1.842943437671609,
+      "grad_norm": 0.4140826463699341,
+      "learning_rate": 0.0001285016286644951,
+      "loss": 1.2105,
+      "step": 839
+    },
+    {
+      "epoch": 1.8451400329489291,
+      "grad_norm": 0.20738688111305237,
+      "learning_rate": 0.00012825732899022802,
+      "loss": 1.274,
+      "step": 840
+    },
+    {
+      "epoch": 1.8473366282262493,
+      "grad_norm": 0.18406347930431366,
+      "learning_rate": 0.0001280130293159609,
+      "loss": 1.1901,
+      "step": 841
+    },
+    {
+      "epoch": 1.8495332235035695,
+      "grad_norm": 0.2968829274177551,
+      "learning_rate": 0.0001277687296416938,
+      "loss": 1.4104,
+      "step": 842
+    },
+    {
+      "epoch": 1.8517298187808895,
+      "grad_norm": 2.3050920963287354,
+      "learning_rate": 0.00012752442996742671,
+      "loss": 1.7315,
+      "step": 843
+    },
+    {
+      "epoch": 1.85392641405821,
+      "grad_norm": 0.20815357565879822,
+      "learning_rate": 0.0001272801302931596,
+      "loss": 1.2512,
+      "step": 844
+    },
+    {
+      "epoch": 1.8561230093355299,
+      "grad_norm": 0.447809100151062,
+      "learning_rate": 0.0001270358306188925,
+      "loss": 1.2763,
+      "step": 845
+    },
+    {
+      "epoch": 1.85831960461285,
+      "grad_norm": 0.2517823278903961,
+      "learning_rate": 0.0001267915309446254,
+      "loss": 1.4296,
+      "step": 846
+    },
+    {
+      "epoch": 1.8605161998901703,
+      "grad_norm": 0.1758531779050827,
+      "learning_rate": 0.0001265472312703583,
+      "loss": 1.4973,
+      "step": 847
+    },
+    {
+      "epoch": 1.8627127951674904,
+      "grad_norm": 1.0840387344360352,
+      "learning_rate": 0.0001263029315960912,
+      "loss": 1.3466,
+      "step": 848
+    },
+    {
+      "epoch": 1.8649093904448106,
+      "grad_norm": 0.17908209562301636,
+      "learning_rate": 0.0001260586319218241,
+      "loss": 1.1632,
+      "step": 849
+    },
+    {
+      "epoch": 1.8671059857221306,
+      "grad_norm": 0.3082078993320465,
+      "learning_rate": 0.000125814332247557,
+      "loss": 1.2108,
+      "step": 850
+    },
+    {
+      "epoch": 1.869302580999451,
+      "grad_norm": 0.21854417026042938,
+      "learning_rate": 0.0001255700325732899,
+      "loss": 1.1118,
+      "step": 851
+    },
+    {
+      "epoch": 1.871499176276771,
+      "grad_norm": 0.1575184464454651,
+      "learning_rate": 0.0001253257328990228,
+      "loss": 1.343,
+      "step": 852
+    },
+    {
+      "epoch": 1.8736957715540912,
+      "grad_norm": 0.14331933856010437,
+      "learning_rate": 0.0001250814332247557,
+      "loss": 1.4911,
+      "step": 853
+    },
+    {
+      "epoch": 1.8758923668314114,
+      "grad_norm": 0.1910061091184616,
+      "learning_rate": 0.0001248371335504886,
+      "loss": 1.5692,
+      "step": 854
+    },
+    {
+      "epoch": 1.8780889621087313,
+      "grad_norm": 0.20147591829299927,
+      "learning_rate": 0.0001245928338762215,
+      "loss": 1.3971,
+      "step": 855
+    },
+    {
+      "epoch": 1.8802855573860517,
+      "grad_norm": 0.16253185272216797,
+      "learning_rate": 0.00012434853420195438,
+      "loss": 1.4942,
+      "step": 856
+    },
+    {
+      "epoch": 1.8824821526633717,
+      "grad_norm": 0.22821293771266937,
+      "learning_rate": 0.0001241042345276873,
+      "loss": 1.4693,
+      "step": 857
+    },
+    {
+      "epoch": 1.884678747940692,
+      "grad_norm": 1.7836053371429443,
+      "learning_rate": 0.0001238599348534202,
+      "loss": 1.3615,
+      "step": 858
+    },
+    {
+      "epoch": 1.886875343218012,
+      "grad_norm": 0.16494309902191162,
+      "learning_rate": 0.00012361563517915308,
+      "loss": 1.1581,
+      "step": 859
+    },
+    {
+      "epoch": 1.8890719384953323,
+      "grad_norm": 0.15919899940490723,
+      "learning_rate": 0.000123371335504886,
+      "loss": 1.2949,
+      "step": 860
+    },
+    {
+      "epoch": 1.8912685337726525,
+      "grad_norm": 0.21696951985359192,
+      "learning_rate": 0.00012312703583061888,
+      "loss": 1.4701,
+      "step": 861
+    },
+    {
+      "epoch": 1.8934651290499724,
+      "grad_norm": 0.2164732664823532,
+      "learning_rate": 0.00012288273615635177,
+      "loss": 1.4951,
+      "step": 862
+    },
+    {
+      "epoch": 1.8956617243272929,
+      "grad_norm": 3.6681337356567383,
+      "learning_rate": 0.0001226384364820847,
+      "loss": 1.9395,
+      "step": 863
+    },
+    {
+      "epoch": 1.8978583196046128,
+      "grad_norm": 0.22124844789505005,
+      "learning_rate": 0.00012239413680781758,
+      "loss": 1.2718,
+      "step": 864
+    },
+    {
+      "epoch": 1.900054914881933,
+      "grad_norm": 0.22357521951198578,
+      "learning_rate": 0.0001221498371335505,
+      "loss": 1.5066,
+      "step": 865
+    },
+    {
+      "epoch": 1.9022515101592532,
+      "grad_norm": 0.2882751524448395,
+      "learning_rate": 0.00012190553745928338,
+      "loss": 1.5778,
+      "step": 866
+    },
+    {
+      "epoch": 1.9044481054365732,
+      "grad_norm": 0.16324131190776825,
+      "learning_rate": 0.00012166123778501629,
+      "loss": 1.3318,
+      "step": 867
+    },
+    {
+      "epoch": 1.9066447007138936,
+      "grad_norm": 0.2573060095310211,
+      "learning_rate": 0.00012141693811074918,
+      "loss": 1.7058,
+      "step": 868
+    },
+    {
+      "epoch": 1.9088412959912135,
+      "grad_norm": 0.15752385556697845,
+      "learning_rate": 0.00012117263843648208,
+      "loss": 1.6774,
+      "step": 869
+    },
+    {
+      "epoch": 1.9110378912685337,
+      "grad_norm": 0.29399770498275757,
+      "learning_rate": 0.00012092833876221498,
+      "loss": 1.287,
+      "step": 870
+    },
+    {
+      "epoch": 1.913234486545854,
+      "grad_norm": 2.259242057800293,
+      "learning_rate": 0.00012068403908794787,
+      "loss": 1.291,
+      "step": 871
+    },
+    {
+      "epoch": 1.9154310818231741,
+      "grad_norm": 0.22426477074623108,
+      "learning_rate": 0.00012043973941368077,
+      "loss": 1.3665,
+      "step": 872
+    },
+    {
+      "epoch": 1.9176276771004943,
+      "grad_norm": 0.30555784702301025,
+      "learning_rate": 0.00012019543973941368,
+      "loss": 1.4944,
+      "step": 873
+    },
+    {
+      "epoch": 1.9198242723778143,
+      "grad_norm": 0.151852086186409,
+      "learning_rate": 0.00011995114006514658,
+      "loss": 1.4912,
+      "step": 874
+    },
+    {
+      "epoch": 1.9220208676551347,
+      "grad_norm": 0.2249622792005539,
+      "learning_rate": 0.00011970684039087947,
+      "loss": 1.6221,
+      "step": 875
+    },
+    {
+      "epoch": 1.9242174629324547,
+      "grad_norm": 0.266010046005249,
+      "learning_rate": 0.00011946254071661237,
+      "loss": 1.6207,
+      "step": 876
+    },
+    {
+      "epoch": 1.9264140582097748,
+      "grad_norm": 0.17431454360485077,
+      "learning_rate": 0.00011921824104234527,
+      "loss": 1.257,
+      "step": 877
+    },
+    {
+      "epoch": 1.928610653487095,
+      "grad_norm": 0.24264472723007202,
+      "learning_rate": 0.00011897394136807816,
+      "loss": 1.6994,
+      "step": 878
+    },
+    {
+      "epoch": 1.930807248764415,
+      "grad_norm": 0.1962588131427765,
+      "learning_rate": 0.00011872964169381106,
+      "loss": 1.3671,
+      "step": 879
+    },
+    {
+      "epoch": 1.9330038440417354,
+      "grad_norm": 0.29242613911628723,
+      "learning_rate": 0.00011848534201954397,
+      "loss": 1.3577,
+      "step": 880
+    },
+    {
+      "epoch": 1.9352004393190554,
+      "grad_norm": 0.2002526968717575,
+      "learning_rate": 0.00011824104234527687,
+      "loss": 1.3039,
+      "step": 881
+    },
+    {
+      "epoch": 1.9373970345963756,
+      "grad_norm": 0.23892834782600403,
+      "learning_rate": 0.00011799674267100976,
+      "loss": 1.2026,
+      "step": 882
+    },
+    {
+      "epoch": 1.9395936298736958,
+      "grad_norm": 0.1750803291797638,
+      "learning_rate": 0.00011775244299674266,
+      "loss": 1.3776,
+      "step": 883
+    },
+    {
+      "epoch": 1.941790225151016,
+      "grad_norm": 0.28349369764328003,
+      "learning_rate": 0.00011750814332247557,
+      "loss": 1.4631,
+      "step": 884
+    },
+    {
+      "epoch": 1.9439868204283361,
+      "grad_norm": 0.4839487373828888,
+      "learning_rate": 0.00011726384364820845,
+      "loss": 1.9593,
+      "step": 885
+    },
+    {
+      "epoch": 1.9461834157056561,
+      "grad_norm": 0.2566339373588562,
+      "learning_rate": 0.00011701954397394136,
+      "loss": 1.4871,
+      "step": 886
+    },
+    {
+      "epoch": 1.9483800109829765,
+      "grad_norm": 0.19810737669467926,
+      "learning_rate": 0.00011677524429967426,
+      "loss": 1.3087,
+      "step": 887
+    },
+    {
+      "epoch": 1.9505766062602965,
+      "grad_norm": 0.15808707475662231,
+      "learning_rate": 0.00011653094462540716,
+      "loss": 1.0817,
+      "step": 888
+    },
+    {
+      "epoch": 1.9527732015376167,
+      "grad_norm": 0.2505858838558197,
+      "learning_rate": 0.00011628664495114005,
+      "loss": 1.4655,
+      "step": 889
+    },
+    {
+      "epoch": 1.9549697968149369,
+      "grad_norm": 0.2099035084247589,
+      "learning_rate": 0.00011604234527687295,
+      "loss": 1.2671,
+      "step": 890
+    },
+    {
+      "epoch": 1.9571663920922568,
+      "grad_norm": 0.15688221156597137,
+      "learning_rate": 0.00011579804560260586,
+      "loss": 1.3294,
+      "step": 891
+    },
+    {
+      "epoch": 1.9593629873695773,
+      "grad_norm": 0.2456396520137787,
+      "learning_rate": 0.00011555374592833876,
+      "loss": 1.4284,
+      "step": 892
+    },
+    {
+      "epoch": 1.9615595826468972,
+      "grad_norm": 0.26844170689582825,
+      "learning_rate": 0.00011530944625407165,
+      "loss": 1.5211,
+      "step": 893
+    },
+    {
+      "epoch": 1.9637561779242174,
+      "grad_norm": 0.20392146706581116,
+      "learning_rate": 0.00011506514657980455,
+      "loss": 1.438,
+      "step": 894
+    },
+    {
+      "epoch": 1.9659527732015376,
+      "grad_norm": 0.21144361793994904,
+      "learning_rate": 0.00011482084690553745,
+      "loss": 0.9624,
+      "step": 895
+    },
+    {
+      "epoch": 1.9681493684788578,
+      "grad_norm": 0.15580876171588898,
+      "learning_rate": 0.00011457654723127034,
+      "loss": 1.4617,
+      "step": 896
+    },
+    {
+      "epoch": 1.970345963756178,
+      "grad_norm": 0.16627080738544464,
+      "learning_rate": 0.00011433224755700325,
+      "loss": 1.1791,
+      "step": 897
+    },
+    {
+      "epoch": 1.972542559033498,
+      "grad_norm": 0.21740594506263733,
+      "learning_rate": 0.00011408794788273615,
+      "loss": 1.6171,
+      "step": 898
+    },
+    {
+      "epoch": 1.9747391543108184,
+      "grad_norm": 0.2720195949077606,
+      "learning_rate": 0.00011384364820846905,
+      "loss": 1.5237,
+      "step": 899
+    },
+    {
+      "epoch": 1.9769357495881383,
+      "grad_norm": 0.19744203984737396,
+      "learning_rate": 0.00011359934853420194,
+      "loss": 1.3974,
+      "step": 900
+    },
+    {
+      "epoch": 1.9791323448654585,
+      "grad_norm": 0.16138453781604767,
+      "learning_rate": 0.00011335504885993484,
+      "loss": 1.521,
+      "step": 901
+    },
+    {
+      "epoch": 1.9813289401427787,
+      "grad_norm": 0.20296061038970947,
+      "learning_rate": 0.00011311074918566775,
+      "loss": 1.3735,
+      "step": 902
+    },
+    {
+      "epoch": 1.9835255354200987,
+      "grad_norm": 0.1615646481513977,
+      "learning_rate": 0.00011286644951140064,
+      "loss": 1.3359,
+      "step": 903
+    },
+    {
+      "epoch": 1.985722130697419,
+      "grad_norm": 0.34788933396339417,
+      "learning_rate": 0.00011262214983713354,
+      "loss": 1.3649,
+      "step": 904
+    },
+    {
+      "epoch": 1.987918725974739,
+      "grad_norm": 0.2735278308391571,
+      "learning_rate": 0.00011237785016286644,
+      "loss": 1.3666,
+      "step": 905
+    },
+    {
+      "epoch": 1.9901153212520593,
+      "grad_norm": 0.21563218533992767,
+      "learning_rate": 0.00011213355048859934,
+      "loss": 1.4732,
+      "step": 906
+    },
+    {
+      "epoch": 1.9923119165293794,
+      "grad_norm": 0.17538166046142578,
+      "learning_rate": 0.00011188925081433223,
+      "loss": 1.0808,
+      "step": 907
+    },
+    {
+      "epoch": 1.9945085118066996,
+      "grad_norm": 0.5094255208969116,
+      "learning_rate": 0.00011164495114006514,
+      "loss": 1.1716,
+      "step": 908
+    },
+    {
+      "epoch": 1.9967051070840198,
+      "grad_norm": 0.21936166286468506,
+      "learning_rate": 0.00011140065146579804,
+      "loss": 1.1246,
+      "step": 909
+    },
+    {
+      "epoch": 1.9989017023613398,
+      "grad_norm": 0.2512020766735077,
+      "learning_rate": 0.00011115635179153093,
+      "loss": 1.5084,
+      "step": 910
+    },
+    {
+      "epoch": 2.00109829763866,
+      "grad_norm": 0.1653304547071457,
+      "learning_rate": 0.00011091205211726383,
+      "loss": 1.4345,
+      "step": 911
+    },
+    {
+      "epoch": 2.00329489291598,
+      "grad_norm": 0.25789299607276917,
+      "learning_rate": 0.00011066775244299673,
+      "loss": 1.4916,
+      "step": 912
+    },
+    {
+      "epoch": 2.0054914881933006,
+      "grad_norm": 0.17674526572227478,
+      "learning_rate": 0.00011042345276872964,
+      "loss": 1.1553,
+      "step": 913
+    },
+    {
+      "epoch": 2.0076880834706206,
+      "grad_norm": 0.2938138544559479,
+      "learning_rate": 0.00011017915309446253,
+      "loss": 1.4165,
+      "step": 914
+    },
+    {
+      "epoch": 2.0098846787479405,
+      "grad_norm": 0.6717258095741272,
+      "learning_rate": 0.00010993485342019543,
+      "loss": 1.3014,
+      "step": 915
+    },
+    {
+      "epoch": 2.012081274025261,
+      "grad_norm": 2.5162272453308105,
+      "learning_rate": 0.00010969055374592833,
+      "loss": 1.5399,
+      "step": 916
+    },
+    {
+      "epoch": 2.014277869302581,
+      "grad_norm": 0.20406237244606018,
+      "learning_rate": 0.00010944625407166122,
+      "loss": 1.5797,
+      "step": 917
+    },
+    {
+      "epoch": 2.0164744645799013,
+      "grad_norm": 0.1998022049665451,
+      "learning_rate": 0.00010920195439739412,
+      "loss": 1.6439,
+      "step": 918
+    },
+    {
+      "epoch": 2.0186710598572213,
+      "grad_norm": 0.2037169486284256,
+      "learning_rate": 0.00010895765472312703,
+      "loss": 1.3838,
+      "step": 919
+    },
+    {
+      "epoch": 2.0208676551345413,
+      "grad_norm": 0.19740107655525208,
+      "learning_rate": 0.00010871335504885993,
+      "loss": 1.2732,
+      "step": 920
+    },
+    {
+      "epoch": 2.0230642504118617,
+      "grad_norm": 0.20872588455677032,
+      "learning_rate": 0.00010846905537459282,
+      "loss": 1.2397,
+      "step": 921
+    },
+    {
+      "epoch": 2.0252608456891816,
+      "grad_norm": 0.22758692502975464,
+      "learning_rate": 0.00010822475570032572,
+      "loss": 1.4371,
+      "step": 922
+    },
+    {
+      "epoch": 2.027457440966502,
+      "grad_norm": 0.19487228989601135,
+      "learning_rate": 0.00010798045602605862,
+      "loss": 1.5518,
+      "step": 923
+    },
+    {
+      "epoch": 2.029654036243822,
+      "grad_norm": 0.17455589771270752,
+      "learning_rate": 0.00010773615635179153,
+      "loss": 1.3996,
+      "step": 924
+    },
+    {
+      "epoch": 2.0318506315211424,
+      "grad_norm": 0.27311432361602783,
+      "learning_rate": 0.00010749185667752442,
+      "loss": 1.1952,
+      "step": 925
+    },
+    {
+      "epoch": 2.0340472267984624,
+      "grad_norm": 0.8140767216682434,
+      "learning_rate": 0.00010724755700325732,
+      "loss": 1.452,
+      "step": 926
+    },
+    {
+      "epoch": 2.0362438220757824,
+      "grad_norm": 0.5986822247505188,
+      "learning_rate": 0.00010700325732899022,
+      "loss": 1.4775,
+      "step": 927
+    },
+    {
+      "epoch": 2.0384404173531028,
+      "grad_norm": 0.22461582720279694,
+      "learning_rate": 0.00010675895765472311,
+      "loss": 1.4012,
+      "step": 928
+    },
+    {
+      "epoch": 2.0406370126304227,
+      "grad_norm": 0.17101752758026123,
+      "learning_rate": 0.00010651465798045601,
+      "loss": 1.1783,
+      "step": 929
+    },
+    {
+      "epoch": 2.042833607907743,
+      "grad_norm": 0.47473663091659546,
+      "learning_rate": 0.00010627035830618892,
+      "loss": 1.2088,
+      "step": 930
+    },
+    {
+      "epoch": 2.045030203185063,
+      "grad_norm": 0.14159242808818817,
+      "learning_rate": 0.00010602605863192182,
+      "loss": 1.1302,
+      "step": 931
+    },
+    {
+      "epoch": 2.047226798462383,
+      "grad_norm": 0.17783978581428528,
+      "learning_rate": 0.00010578175895765471,
+      "loss": 1.3729,
+      "step": 932
+    },
+    {
+      "epoch": 2.0494233937397035,
+      "grad_norm": 0.2594006657600403,
+      "learning_rate": 0.00010553745928338761,
+      "loss": 1.2977,
+      "step": 933
+    },
+    {
+      "epoch": 2.0516199890170235,
+      "grad_norm": 0.7573102116584778,
+      "learning_rate": 0.00010529315960912051,
+      "loss": 1.4986,
+      "step": 934
+    },
+    {
+      "epoch": 2.053816584294344,
+      "grad_norm": 0.2645663917064667,
+      "learning_rate": 0.0001050488599348534,
+      "loss": 1.6041,
+      "step": 935
+    },
+    {
+      "epoch": 2.056013179571664,
+      "grad_norm": 0.17133016884326935,
+      "learning_rate": 0.0001048045602605863,
+      "loss": 1.4836,
+      "step": 936
+    },
+    {
+      "epoch": 2.0582097748489843,
+      "grad_norm": 0.19508536159992218,
+      "learning_rate": 0.00010456026058631921,
+      "loss": 1.4438,
+      "step": 937
+    },
+    {
+      "epoch": 2.0604063701263042,
+      "grad_norm": 0.21856260299682617,
+      "learning_rate": 0.00010431596091205211,
+      "loss": 1.148,
+      "step": 938
+    },
+    {
+      "epoch": 2.062602965403624,
+      "grad_norm": 0.2829470932483673,
+      "learning_rate": 0.000104071661237785,
+      "loss": 1.8325,
+      "step": 939
+    },
+    {
+      "epoch": 2.0647995606809446,
+      "grad_norm": 0.2196110039949417,
+      "learning_rate": 0.0001038273615635179,
+      "loss": 1.405,
+      "step": 940
+    },
+    {
+      "epoch": 2.0669961559582646,
+      "grad_norm": 0.2287091761827469,
+      "learning_rate": 0.0001035830618892508,
+      "loss": 1.3428,
+      "step": 941
+    },
+    {
+      "epoch": 2.069192751235585,
+      "grad_norm": 0.20146504044532776,
+      "learning_rate": 0.0001033387622149837,
+      "loss": 1.0069,
+      "step": 942
+    },
+    {
+      "epoch": 2.071389346512905,
+      "grad_norm": 0.19762928783893585,
+      "learning_rate": 0.0001030944625407166,
+      "loss": 1.3905,
+      "step": 943
+    },
+    {
+      "epoch": 2.073585941790225,
+      "grad_norm": 0.2882109582424164,
+      "learning_rate": 0.0001028501628664495,
+      "loss": 1.3744,
+      "step": 944
+    },
+    {
+      "epoch": 2.0757825370675453,
+      "grad_norm": 0.9888032674789429,
+      "learning_rate": 0.0001026058631921824,
+      "loss": 1.4272,
+      "step": 945
+    },
+    {
+      "epoch": 2.0779791323448653,
+      "grad_norm": 0.2057337909936905,
+      "learning_rate": 0.00010236156351791529,
+      "loss": 1.3739,
+      "step": 946
+    },
+    {
+      "epoch": 2.0801757276221857,
+      "grad_norm": 0.14969958364963531,
+      "learning_rate": 0.0001021172638436482,
+      "loss": 1.0219,
+      "step": 947
+    },
+    {
+      "epoch": 2.0823723228995057,
+      "grad_norm": 0.23668955266475677,
+      "learning_rate": 0.0001018729641693811,
+      "loss": 1.3787,
+      "step": 948
+    },
+    {
+      "epoch": 2.084568918176826,
+      "grad_norm": 0.22754451632499695,
+      "learning_rate": 0.00010162866449511399,
+      "loss": 1.7383,
+      "step": 949
+    },
+    {
+      "epoch": 2.086765513454146,
+      "grad_norm": 0.27770835161209106,
+      "learning_rate": 0.00010138436482084689,
+      "loss": 1.6724,
+      "step": 950
+    },
+    {
+      "epoch": 2.088962108731466,
+      "grad_norm": 0.19818635284900665,
+      "learning_rate": 0.00010114006514657979,
+      "loss": 1.332,
+      "step": 951
+    },
+    {
+      "epoch": 2.0911587040087865,
+      "grad_norm": 0.21416912972927094,
+      "learning_rate": 0.0001008957654723127,
+      "loss": 1.5037,
+      "step": 952
+    },
+    {
+      "epoch": 2.0933552992861064,
+      "grad_norm": 0.24095068871974945,
+      "learning_rate": 0.00010065146579804558,
+      "loss": 1.575,
+      "step": 953
+    },
+    {
+      "epoch": 2.095551894563427,
+      "grad_norm": 0.274183988571167,
+      "learning_rate": 0.00010040716612377849,
+      "loss": 1.5184,
+      "step": 954
+    },
+    {
+      "epoch": 2.097748489840747,
+      "grad_norm": 0.23575998842716217,
+      "learning_rate": 0.00010016286644951139,
+      "loss": 1.4606,
+      "step": 955
+    },
+    {
+      "epoch": 2.099945085118067,
+      "grad_norm": 0.15230488777160645,
+      "learning_rate": 9.991856677524428e-05,
+      "loss": 1.2231,
+      "step": 956
+    },
+    {
+      "epoch": 2.102141680395387,
+      "grad_norm": 0.2702975273132324,
+      "learning_rate": 9.967426710097718e-05,
+      "loss": 1.4522,
+      "step": 957
+    },
+    {
+      "epoch": 2.104338275672707,
+      "grad_norm": 0.20900417864322662,
+      "learning_rate": 9.942996742671008e-05,
+      "loss": 1.4258,
+      "step": 958
+    },
+    {
+      "epoch": 2.1065348709500276,
+      "grad_norm": 0.18970902264118195,
+      "learning_rate": 9.918566775244299e-05,
+      "loss": 1.445,
+      "step": 959
+    },
+    {
+      "epoch": 2.1087314662273475,
+      "grad_norm": 0.1459827721118927,
+      "learning_rate": 9.894136807817588e-05,
+      "loss": 1.218,
+      "step": 960
+    },
+    {
+      "epoch": 2.110928061504668,
+      "grad_norm": 0.38068506121635437,
+      "learning_rate": 9.869706840390878e-05,
+      "loss": 1.7456,
+      "step": 961
+    },
+    {
+      "epoch": 2.113124656781988,
+      "grad_norm": 0.18417230248451233,
+      "learning_rate": 9.845276872964168e-05,
+      "loss": 1.3638,
+      "step": 962
+    },
+    {
+      "epoch": 2.115321252059308,
+      "grad_norm": 0.19640502333641052,
+      "learning_rate": 9.820846905537458e-05,
+      "loss": 1.2976,
+      "step": 963
+    },
+    {
+      "epoch": 2.1175178473366283,
+      "grad_norm": 0.16095790266990662,
+      "learning_rate": 9.796416938110747e-05,
+      "loss": 1.2386,
+      "step": 964
+    },
+    {
+      "epoch": 2.1197144426139483,
+      "grad_norm": 0.2977880835533142,
+      "learning_rate": 9.771986970684038e-05,
+      "loss": 1.3604,
+      "step": 965
+    },
+    {
+      "epoch": 2.1219110378912687,
+      "grad_norm": 0.2996789216995239,
+      "learning_rate": 9.747557003257328e-05,
+      "loss": 1.1619,
+      "step": 966
+    },
+    {
+      "epoch": 2.1241076331685886,
+      "grad_norm": 0.19290515780448914,
+      "learning_rate": 9.723127035830617e-05,
+      "loss": 1.2773,
+      "step": 967
+    },
+    {
+      "epoch": 2.1263042284459086,
+      "grad_norm": 0.25365331768989563,
+      "learning_rate": 9.698697068403907e-05,
+      "loss": 1.271,
+      "step": 968
+    },
+    {
+      "epoch": 2.128500823723229,
+      "grad_norm": 0.2721065282821655,
+      "learning_rate": 9.674267100977197e-05,
+      "loss": 1.4255,
+      "step": 969
+    },
+    {
+      "epoch": 2.130697419000549,
+      "grad_norm": 0.27061453461647034,
+      "learning_rate": 9.649837133550488e-05,
+      "loss": 1.7893,
+      "step": 970
+    },
+    {
+      "epoch": 2.1328940142778694,
+      "grad_norm": 0.25385087728500366,
+      "learning_rate": 9.625407166123777e-05,
+      "loss": 1.4354,
+      "step": 971
+    },
+    {
+      "epoch": 2.1350906095551894,
+      "grad_norm": 0.24618680775165558,
+      "learning_rate": 9.600977198697067e-05,
+      "loss": 1.7048,
+      "step": 972
+    },
+    {
+      "epoch": 2.13728720483251,
+      "grad_norm": 0.20283547043800354,
+      "learning_rate": 9.576547231270357e-05,
+      "loss": 1.2615,
+      "step": 973
+    },
+    {
+      "epoch": 2.1394838001098297,
+      "grad_norm": 0.19419950246810913,
+      "learning_rate": 9.552117263843646e-05,
+      "loss": 1.3919,
+      "step": 974
+    },
+    {
+      "epoch": 2.1416803953871497,
+      "grad_norm": 0.22630715370178223,
+      "learning_rate": 9.527687296416936e-05,
+      "loss": 1.416,
+      "step": 975
+    },
+    {
+      "epoch": 2.14387699066447,
+      "grad_norm": 0.21820075809955597,
+      "learning_rate": 9.503257328990227e-05,
+      "loss": 1.4043,
+      "step": 976
+    },
+    {
+      "epoch": 2.14607358594179,
+      "grad_norm": 0.20700396597385406,
+      "learning_rate": 9.478827361563517e-05,
+      "loss": 1.3865,
+      "step": 977
+    },
+    {
+      "epoch": 2.1482701812191105,
+      "grad_norm": 0.17282232642173767,
+      "learning_rate": 9.454397394136806e-05,
+      "loss": 1.1912,
+      "step": 978
+    },
+    {
+      "epoch": 2.1504667764964305,
+      "grad_norm": 0.16802680492401123,
+      "learning_rate": 9.429967426710096e-05,
+      "loss": 1.2735,
+      "step": 979
+    },
+    {
+      "epoch": 2.152663371773751,
+      "grad_norm": 0.7697277069091797,
+      "learning_rate": 9.405537459283386e-05,
+      "loss": 1.7095,
+      "step": 980
+    },
+    {
+      "epoch": 2.154859967051071,
+      "grad_norm": 0.2105928510427475,
+      "learning_rate": 9.381107491856675e-05,
+      "loss": 1.2516,
+      "step": 981
+    },
+    {
+      "epoch": 2.157056562328391,
+      "grad_norm": 0.2143172025680542,
+      "learning_rate": 9.356677524429967e-05,
+      "loss": 1.2514,
+      "step": 982
+    },
+    {
+      "epoch": 2.1592531576057112,
+      "grad_norm": 0.2899986207485199,
+      "learning_rate": 9.332247557003257e-05,
+      "loss": 1.6095,
+      "step": 983
+    },
+    {
+      "epoch": 2.161449752883031,
+      "grad_norm": 0.23091678321361542,
+      "learning_rate": 9.307817589576547e-05,
+      "loss": 1.2887,
+      "step": 984
+    },
+    {
+      "epoch": 2.1636463481603516,
+      "grad_norm": 0.26528075337409973,
+      "learning_rate": 9.283387622149838e-05,
+      "loss": 1.2489,
+      "step": 985
+    },
+    {
+      "epoch": 2.1658429434376716,
+      "grad_norm": 0.2981375753879547,
+      "learning_rate": 9.258957654723127e-05,
+      "loss": 1.2603,
+      "step": 986
+    },
+    {
+      "epoch": 2.168039538714992,
+      "grad_norm": 0.16978327929973602,
+      "learning_rate": 9.234527687296417e-05,
+      "loss": 1.4059,
+      "step": 987
+    },
+    {
+      "epoch": 2.170236133992312,
+      "grad_norm": 0.2975841760635376,
+      "learning_rate": 9.210097719869707e-05,
+      "loss": 1.6477,
+      "step": 988
+    },
+    {
+      "epoch": 2.172432729269632,
+      "grad_norm": 0.14492321014404297,
+      "learning_rate": 9.185667752442996e-05,
+      "loss": 1.1661,
+      "step": 989
+    },
+    {
+      "epoch": 2.1746293245469523,
+      "grad_norm": 0.34362590312957764,
+      "learning_rate": 9.161237785016286e-05,
+      "loss": 1.9488,
+      "step": 990
+    },
+    {
+      "epoch": 2.1768259198242723,
+      "grad_norm": 0.16532684862613678,
+      "learning_rate": 9.136807817589577e-05,
+      "loss": 1.219,
+      "step": 991
+    },
+    {
+      "epoch": 2.1790225151015927,
+      "grad_norm": 0.15328644216060638,
+      "learning_rate": 9.112377850162867e-05,
+      "loss": 1.0672,
+      "step": 992
+    },
+    {
+      "epoch": 2.1812191103789127,
+      "grad_norm": 1.0922644138336182,
+      "learning_rate": 9.087947882736156e-05,
+      "loss": 1.5114,
+      "step": 993
+    },
+    {
+      "epoch": 2.1834157056562327,
+      "grad_norm": 0.4965292513370514,
+      "learning_rate": 9.063517915309446e-05,
+      "loss": 1.2653,
+      "step": 994
+    },
+    {
+      "epoch": 2.185612300933553,
+      "grad_norm": 0.1855444461107254,
+      "learning_rate": 9.039087947882736e-05,
+      "loss": 1.2504,
+      "step": 995
+    },
+    {
+      "epoch": 2.187808896210873,
+      "grad_norm": 0.22396041452884674,
+      "learning_rate": 9.014657980456025e-05,
+      "loss": 1.6787,
+      "step": 996
+    },
+    {
+      "epoch": 2.1900054914881935,
+      "grad_norm": 0.2425040900707245,
+      "learning_rate": 8.990228013029316e-05,
+      "loss": 1.4971,
+      "step": 997
+    },
+    {
+      "epoch": 2.1922020867655134,
+      "grad_norm": 0.1367364525794983,
+      "learning_rate": 8.965798045602606e-05,
+      "loss": 1.2388,
+      "step": 998
+    },
+    {
+      "epoch": 2.1943986820428334,
+      "grad_norm": 0.2183048129081726,
+      "learning_rate": 8.941368078175896e-05,
+      "loss": 1.497,
+      "step": 999
+    },
+    {
+      "epoch": 2.196595277320154,
+      "grad_norm": 0.2193196564912796,
+      "learning_rate": 8.916938110749185e-05,
+      "loss": 1.1793,
+      "step": 1000
+    },
+    {
+      "epoch": 2.1987918725974738,
+      "grad_norm": 0.8677994012832642,
+      "learning_rate": 8.892508143322475e-05,
+      "loss": 1.273,
+      "step": 1001
+    },
+    {
+      "epoch": 2.200988467874794,
+      "grad_norm": 0.24032939970493317,
+      "learning_rate": 8.868078175895766e-05,
+      "loss": 1.5052,
+      "step": 1002
+    },
+    {
+      "epoch": 2.203185063152114,
+      "grad_norm": 0.25967833399772644,
+      "learning_rate": 8.843648208469055e-05,
+      "loss": 1.4347,
+      "step": 1003
+    },
+    {
+      "epoch": 2.2053816584294346,
+      "grad_norm": 0.24665625393390656,
+      "learning_rate": 8.819218241042345e-05,
+      "loss": 1.339,
+      "step": 1004
+    },
+    {
+      "epoch": 2.2075782537067545,
+      "grad_norm": 0.14317521452903748,
+      "learning_rate": 8.794788273615635e-05,
+      "loss": 1.2216,
+      "step": 1005
+    },
+    {
+      "epoch": 2.2097748489840745,
+      "grad_norm": 0.19034574925899506,
+      "learning_rate": 8.770358306188925e-05,
+      "loss": 1.3159,
+      "step": 1006
+    },
+    {
+      "epoch": 2.211971444261395,
+      "grad_norm": 0.27595868706703186,
+      "learning_rate": 8.745928338762214e-05,
+      "loss": 1.1297,
+      "step": 1007
+    },
+    {
+      "epoch": 2.214168039538715,
+      "grad_norm": 1.479536771774292,
+      "learning_rate": 8.721498371335505e-05,
+      "loss": 1.9438,
+      "step": 1008
+    },
+    {
+      "epoch": 2.2163646348160353,
+      "grad_norm": 0.2818644642829895,
+      "learning_rate": 8.697068403908795e-05,
+      "loss": 1.7025,
+      "step": 1009
+    },
+    {
+      "epoch": 2.2185612300933553,
+      "grad_norm": 0.25675711035728455,
+      "learning_rate": 8.672638436482085e-05,
+      "loss": 1.365,
+      "step": 1010
+    },
+    {
+      "epoch": 2.2207578253706757,
+      "grad_norm": 0.19122309982776642,
+      "learning_rate": 8.648208469055374e-05,
+      "loss": 1.3485,
+      "step": 1011
+    },
+    {
+      "epoch": 2.2229544206479956,
+      "grad_norm": 0.1536233276128769,
+      "learning_rate": 8.623778501628664e-05,
+      "loss": 1.4106,
+      "step": 1012
+    },
+    {
+      "epoch": 2.2251510159253156,
+      "grad_norm": 0.36031123995780945,
+      "learning_rate": 8.599348534201955e-05,
+      "loss": 1.3296,
+      "step": 1013
+    },
+    {
+      "epoch": 2.227347611202636,
+      "grad_norm": 0.20472979545593262,
+      "learning_rate": 8.574918566775243e-05,
+      "loss": 1.6309,
+      "step": 1014
+    },
+    {
+      "epoch": 2.229544206479956,
+      "grad_norm": 0.2141575813293457,
+      "learning_rate": 8.550488599348534e-05,
+      "loss": 1.7298,
+      "step": 1015
+    },
+    {
+      "epoch": 2.2317408017572764,
+      "grad_norm": 0.131341814994812,
+      "learning_rate": 8.526058631921824e-05,
+      "loss": 1.243,
+      "step": 1016
+    },
+    {
+      "epoch": 2.2339373970345964,
+      "grad_norm": 0.6830413937568665,
+      "learning_rate": 8.501628664495114e-05,
+      "loss": 1.6791,
+      "step": 1017
+    },
+    {
+      "epoch": 2.2361339923119163,
+      "grad_norm": 0.16122624278068542,
+      "learning_rate": 8.477198697068403e-05,
+      "loss": 1.4547,
+      "step": 1018
+    },
+    {
+      "epoch": 2.2383305875892368,
+      "grad_norm": 0.19288095831871033,
+      "learning_rate": 8.452768729641694e-05,
+      "loss": 1.5143,
+      "step": 1019
+    },
+    {
+      "epoch": 2.2405271828665567,
+      "grad_norm": 0.2903972566127777,
+      "learning_rate": 8.428338762214984e-05,
+      "loss": 1.3747,
+      "step": 1020
+    },
+    {
+      "epoch": 2.242723778143877,
+      "grad_norm": 0.14779113233089447,
+      "learning_rate": 8.403908794788273e-05,
+      "loss": 1.3333,
+      "step": 1021
+    },
+    {
+      "epoch": 2.244920373421197,
+      "grad_norm": 0.18495352566242218,
+      "learning_rate": 8.379478827361563e-05,
+      "loss": 1.0218,
+      "step": 1022
+    },
+    {
+      "epoch": 2.247116968698517,
+      "grad_norm": 0.16002845764160156,
+      "learning_rate": 8.355048859934853e-05,
+      "loss": 1.322,
+      "step": 1023
+    },
+    {
+      "epoch": 2.2493135639758375,
+      "grad_norm": 0.21485884487628937,
+      "learning_rate": 8.330618892508144e-05,
+      "loss": 1.5236,
+      "step": 1024
+    },
+    {
+      "epoch": 2.2515101592531575,
+      "grad_norm": 0.20161975920200348,
+      "learning_rate": 8.306188925081432e-05,
+      "loss": 1.3696,
+      "step": 1025
+    },
+    {
+      "epoch": 2.253706754530478,
+      "grad_norm": 1.1814533472061157,
+      "learning_rate": 8.281758957654723e-05,
+      "loss": 1.211,
+      "step": 1026
+    },
+    {
+      "epoch": 2.255903349807798,
+      "grad_norm": 0.2981582581996918,
+      "learning_rate": 8.257328990228013e-05,
+      "loss": 1.3675,
+      "step": 1027
+    },
+    {
+      "epoch": 2.2580999450851182,
+      "grad_norm": 0.28699079155921936,
+      "learning_rate": 8.232899022801302e-05,
+      "loss": 1.7596,
+      "step": 1028
+    },
+    {
+      "epoch": 2.260296540362438,
+      "grad_norm": 0.9461281895637512,
+      "learning_rate": 8.208469055374592e-05,
+      "loss": 1.4862,
+      "step": 1029
+    },
+    {
+      "epoch": 2.262493135639758,
+      "grad_norm": 0.22394511103630066,
+      "learning_rate": 8.184039087947882e-05,
+      "loss": 1.2893,
+      "step": 1030
+    },
+    {
+      "epoch": 2.2646897309170786,
+      "grad_norm": 0.2492426633834839,
+      "learning_rate": 8.159609120521173e-05,
+      "loss": 1.2289,
+      "step": 1031
+    },
+    {
+      "epoch": 2.2668863261943986,
+      "grad_norm": 0.21610939502716064,
+      "learning_rate": 8.135179153094462e-05,
+      "loss": 1.2899,
+      "step": 1032
+    },
+    {
+      "epoch": 2.269082921471719,
+      "grad_norm": 0.2431950569152832,
+      "learning_rate": 8.110749185667752e-05,
+      "loss": 1.5459,
+      "step": 1033
+    },
+    {
+      "epoch": 2.271279516749039,
+      "grad_norm": 0.16477471590042114,
+      "learning_rate": 8.086319218241042e-05,
+      "loss": 1.1524,
+      "step": 1034
+    },
+    {
+      "epoch": 2.2734761120263594,
+      "grad_norm": 0.17133820056915283,
+      "learning_rate": 8.061889250814331e-05,
+      "loss": 1.2868,
+      "step": 1035
+    },
+    {
+      "epoch": 2.2756727073036793,
+      "grad_norm": 0.24375487864017487,
+      "learning_rate": 8.037459283387621e-05,
+      "loss": 1.6207,
+      "step": 1036
+    },
+    {
+      "epoch": 2.2778693025809993,
+      "grad_norm": 0.350642591714859,
+      "learning_rate": 8.013029315960912e-05,
+      "loss": 1.3041,
+      "step": 1037
+    },
+    {
+      "epoch": 2.2800658978583197,
+      "grad_norm": 0.3557656407356262,
+      "learning_rate": 7.988599348534202e-05,
+      "loss": 1.5255,
+      "step": 1038
+    },
+    {
+      "epoch": 2.2822624931356397,
+      "grad_norm": 0.8514796495437622,
+      "learning_rate": 7.964169381107491e-05,
+      "loss": 1.0529,
+      "step": 1039
+    },
+    {
+      "epoch": 2.28445908841296,
+      "grad_norm": 0.2063380479812622,
+      "learning_rate": 7.939739413680781e-05,
+      "loss": 1.4253,
+      "step": 1040
+    },
+    {
+      "epoch": 2.28665568369028,
+      "grad_norm": 0.23913481831550598,
+      "learning_rate": 7.915309446254071e-05,
+      "loss": 1.6009,
+      "step": 1041
+    },
+    {
+      "epoch": 2.2888522789676005,
+      "grad_norm": 0.28252214193344116,
+      "learning_rate": 7.890879478827362e-05,
+      "loss": 1.5007,
+      "step": 1042
+    },
+    {
+      "epoch": 2.2910488742449204,
+      "grad_norm": 0.20608732104301453,
+      "learning_rate": 7.86644951140065e-05,
+      "loss": 1.1633,
+      "step": 1043
+    },
+    {
+      "epoch": 2.2932454695222404,
+      "grad_norm": 0.18775761127471924,
+      "learning_rate": 7.842019543973941e-05,
+      "loss": 1.3277,
+      "step": 1044
+    },
+    {
+      "epoch": 2.295442064799561,
+      "grad_norm": 0.21078310906887054,
+      "learning_rate": 7.817589576547231e-05,
+      "loss": 1.5311,
+      "step": 1045
+    },
+    {
+      "epoch": 2.297638660076881,
+      "grad_norm": 0.47881367802619934,
+      "learning_rate": 7.79315960912052e-05,
+      "loss": 1.4311,
+      "step": 1046
+    },
+    {
+      "epoch": 2.2998352553542007,
+      "grad_norm": 0.305379182100296,
+      "learning_rate": 7.76872964169381e-05,
+      "loss": 1.3023,
+      "step": 1047
+    },
+    {
+      "epoch": 2.302031850631521,
+      "grad_norm": 0.329461932182312,
+      "learning_rate": 7.7442996742671e-05,
+      "loss": 1.4581,
+      "step": 1048
+    },
+    {
+      "epoch": 2.304228445908841,
+      "grad_norm": 0.2726741433143616,
+      "learning_rate": 7.719869706840391e-05,
+      "loss": 1.5956,
+      "step": 1049
+    },
+    {
+      "epoch": 2.3064250411861615,
+      "grad_norm": 0.12619316577911377,
+      "learning_rate": 7.69543973941368e-05,
+      "loss": 1.256,
+      "step": 1050
+    },
+    {
+      "epoch": 2.3086216364634815,
+      "grad_norm": 3.8506016731262207,
+      "learning_rate": 7.67100977198697e-05,
+      "loss": 1.2065,
+      "step": 1051
+    },
+    {
+      "epoch": 2.310818231740802,
+      "grad_norm": 0.2045181691646576,
+      "learning_rate": 7.64657980456026e-05,
+      "loss": 1.3298,
+      "step": 1052
+    },
+    {
+      "epoch": 2.313014827018122,
+      "grad_norm": 0.5041709542274475,
+      "learning_rate": 7.622149837133549e-05,
+      "loss": 1.5257,
+      "step": 1053
+    },
+    {
+      "epoch": 2.315211422295442,
+      "grad_norm": 0.36025121808052063,
+      "learning_rate": 7.59771986970684e-05,
+      "loss": 1.227,
+      "step": 1054
+    },
+    {
+      "epoch": 2.3174080175727623,
+      "grad_norm": 0.1555452197790146,
+      "learning_rate": 7.57328990228013e-05,
+      "loss": 1.5067,
+      "step": 1055
+    },
+    {
+      "epoch": 2.3196046128500822,
+      "grad_norm": 0.14404194056987762,
+      "learning_rate": 7.54885993485342e-05,
+      "loss": 1.143,
+      "step": 1056
+    },
+    {
+      "epoch": 2.3218012081274026,
+      "grad_norm": 0.23977619409561157,
+      "learning_rate": 7.524429967426709e-05,
+      "loss": 1.4733,
+      "step": 1057
+    },
+    {
+      "epoch": 2.3239978034047226,
+      "grad_norm": 0.17418934404850006,
+      "learning_rate": 7.5e-05,
+      "loss": 1.1721,
+      "step": 1058
+    },
+    {
+      "epoch": 2.326194398682043,
+      "grad_norm": 0.3196136951446533,
+      "learning_rate": 7.47557003257329e-05,
+      "loss": 1.5294,
+      "step": 1059
+    },
+    {
+      "epoch": 2.328390993959363,
+      "grad_norm": 0.2646423280239105,
+      "learning_rate": 7.451140065146579e-05,
+      "loss": 1.3242,
+      "step": 1060
+    },
+    {
+      "epoch": 2.330587589236683,
+      "grad_norm": 0.2307281196117401,
+      "learning_rate": 7.426710097719869e-05,
+      "loss": 1.5279,
+      "step": 1061
+    },
+    {
+      "epoch": 2.3327841845140034,
+      "grad_norm": 0.2387910783290863,
+      "learning_rate": 7.402280130293159e-05,
+      "loss": 1.4852,
+      "step": 1062
+    },
+    {
+      "epoch": 2.3349807797913233,
+      "grad_norm": 0.20486928522586823,
+      "learning_rate": 7.37785016286645e-05,
+      "loss": 1.3534,
+      "step": 1063
+    },
+    {
+      "epoch": 2.3371773750686438,
+      "grad_norm": 0.18170353770256042,
+      "learning_rate": 7.353420195439738e-05,
+      "loss": 1.4916,
+      "step": 1064
+    },
+    {
+      "epoch": 2.3393739703459637,
+      "grad_norm": 0.17407332360744476,
+      "learning_rate": 7.328990228013029e-05,
+      "loss": 1.3328,
+      "step": 1065
+    },
+    {
+      "epoch": 2.341570565623284,
+      "grad_norm": 0.15711259841918945,
+      "learning_rate": 7.304560260586319e-05,
+      "loss": 1.3829,
+      "step": 1066
+    },
+    {
+      "epoch": 2.343767160900604,
+      "grad_norm": 0.15377655625343323,
+      "learning_rate": 7.280130293159608e-05,
+      "loss": 1.1151,
+      "step": 1067
+    },
+    {
+      "epoch": 2.345963756177924,
+      "grad_norm": 0.169809490442276,
+      "learning_rate": 7.255700325732898e-05,
+      "loss": 1.245,
+      "step": 1068
+    },
+    {
+      "epoch": 2.3481603514552445,
+      "grad_norm": 0.22954536974430084,
+      "learning_rate": 7.231270358306188e-05,
+      "loss": 1.5231,
+      "step": 1069
+    },
+    {
+      "epoch": 2.3503569467325645,
+      "grad_norm": 0.21440736949443817,
+      "learning_rate": 7.206840390879479e-05,
+      "loss": 1.3658,
+      "step": 1070
+    },
+    {
+      "epoch": 2.352553542009885,
+      "grad_norm": 0.25058168172836304,
+      "learning_rate": 7.182410423452768e-05,
+      "loss": 1.3695,
+      "step": 1071
+    },
+    {
+      "epoch": 2.354750137287205,
+      "grad_norm": 0.21611596643924713,
+      "learning_rate": 7.157980456026058e-05,
+      "loss": 1.375,
+      "step": 1072
+    },
+    {
+      "epoch": 2.356946732564525,
+      "grad_norm": 0.2139333039522171,
+      "learning_rate": 7.133550488599348e-05,
+      "loss": 1.1989,
+      "step": 1073
+    },
+    {
+      "epoch": 2.359143327841845,
+      "grad_norm": 0.22309629619121552,
+      "learning_rate": 7.109120521172637e-05,
+      "loss": 1.3279,
+      "step": 1074
+    },
+    {
+      "epoch": 2.361339923119165,
+      "grad_norm": 0.1871998906135559,
+      "learning_rate": 7.084690553745927e-05,
+      "loss": 1.4558,
+      "step": 1075
+    },
+    {
+      "epoch": 2.3635365183964856,
+      "grad_norm": 0.29660937190055847,
+      "learning_rate": 7.060260586319218e-05,
+      "loss": 1.6098,
+      "step": 1076
+    },
+    {
+      "epoch": 2.3657331136738056,
+      "grad_norm": 0.44715559482574463,
+      "learning_rate": 7.035830618892508e-05,
+      "loss": 1.476,
+      "step": 1077
+    },
+    {
+      "epoch": 2.3679297089511255,
+      "grad_norm": 0.19169501960277557,
+      "learning_rate": 7.011400651465797e-05,
+      "loss": 1.3898,
+      "step": 1078
+    },
+    {
+      "epoch": 2.370126304228446,
+      "grad_norm": 0.23507751524448395,
+      "learning_rate": 6.986970684039087e-05,
+      "loss": 1.0683,
+      "step": 1079
+    },
+    {
+      "epoch": 2.372322899505766,
+      "grad_norm": 0.21713146567344666,
+      "learning_rate": 6.962540716612377e-05,
+      "loss": 1.2207,
+      "step": 1080
+    },
+    {
+      "epoch": 2.3745194947830863,
+      "grad_norm": 6.272311687469482,
+      "learning_rate": 6.938110749185668e-05,
+      "loss": 1.874,
+      "step": 1081
+    },
+    {
+      "epoch": 2.3767160900604063,
+      "grad_norm": 0.25933757424354553,
+      "learning_rate": 6.913680781758956e-05,
+      "loss": 1.3152,
+      "step": 1082
+    },
+    {
+      "epoch": 2.3789126853377267,
+      "grad_norm": 0.33141347765922546,
+      "learning_rate": 6.889250814332247e-05,
+      "loss": 1.6363,
+      "step": 1083
+    },
+    {
+      "epoch": 2.3811092806150467,
+      "grad_norm": 0.7714502215385437,
+      "learning_rate": 6.864820846905537e-05,
+      "loss": 1.6472,
+      "step": 1084
+    },
+    {
+      "epoch": 2.3833058758923666,
+      "grad_norm": 0.34010133147239685,
+      "learning_rate": 6.840390879478826e-05,
+      "loss": 1.7797,
+      "step": 1085
+    },
+    {
+      "epoch": 2.385502471169687,
+      "grad_norm": 0.18810725212097168,
+      "learning_rate": 6.815960912052116e-05,
+      "loss": 1.2744,
+      "step": 1086
+    },
+    {
+      "epoch": 2.387699066447007,
+      "grad_norm": 1.4064626693725586,
+      "learning_rate": 6.791530944625406e-05,
+      "loss": 1.7187,
+      "step": 1087
+    },
+    {
+      "epoch": 2.3898956617243274,
+      "grad_norm": 0.24149908125400543,
+      "learning_rate": 6.767100977198697e-05,
+      "loss": 1.3521,
+      "step": 1088
+    },
+    {
+      "epoch": 2.3920922570016474,
+      "grad_norm": 0.21091431379318237,
+      "learning_rate": 6.742671009771986e-05,
+      "loss": 1.3952,
+      "step": 1089
+    },
+    {
+      "epoch": 2.394288852278968,
+      "grad_norm": 0.1861375868320465,
+      "learning_rate": 6.718241042345276e-05,
+      "loss": 1.1767,
+      "step": 1090
+    },
+    {
+      "epoch": 2.396485447556288,
+      "grad_norm": 0.22415637969970703,
+      "learning_rate": 6.693811074918566e-05,
+      "loss": 1.4149,
+      "step": 1091
+    },
+    {
+      "epoch": 2.3986820428336078,
+      "grad_norm": 0.3325531780719757,
+      "learning_rate": 6.669381107491855e-05,
+      "loss": 1.4957,
+      "step": 1092
+    },
+    {
+      "epoch": 2.400878638110928,
+      "grad_norm": 0.14337149262428284,
+      "learning_rate": 6.644951140065145e-05,
+      "loss": 1.2559,
+      "step": 1093
+    },
+    {
+      "epoch": 2.403075233388248,
+      "grad_norm": 0.17911843955516815,
+      "learning_rate": 6.620521172638436e-05,
+      "loss": 1.483,
+      "step": 1094
+    },
+    {
+      "epoch": 2.4052718286655685,
+      "grad_norm": 0.234444722533226,
+      "learning_rate": 6.596091205211726e-05,
+      "loss": 1.2228,
+      "step": 1095
+    },
+    {
+      "epoch": 2.4074684239428885,
+      "grad_norm": 0.17206354439258575,
+      "learning_rate": 6.571661237785015e-05,
+      "loss": 1.224,
+      "step": 1096
+    },
+    {
+      "epoch": 2.409665019220209,
+      "grad_norm": 2.2190234661102295,
+      "learning_rate": 6.547231270358307e-05,
+      "loss": 1.5791,
+      "step": 1097
+    },
+    {
+      "epoch": 2.411861614497529,
+      "grad_norm": 0.7150613069534302,
+      "learning_rate": 6.522801302931595e-05,
+      "loss": 1.3123,
+      "step": 1098
+    },
+    {
+      "epoch": 2.414058209774849,
+      "grad_norm": 0.3869498074054718,
+      "learning_rate": 6.498371335504886e-05,
+      "loss": 1.7737,
+      "step": 1099
+    },
+    {
+      "epoch": 2.4162548050521693,
+      "grad_norm": 0.2531861960887909,
+      "learning_rate": 6.473941368078176e-05,
+      "loss": 1.5269,
+      "step": 1100
+    },
+    {
+      "epoch": 2.4184514003294892,
+      "grad_norm": 0.2593514621257782,
+      "learning_rate": 6.449511400651466e-05,
+      "loss": 1.4892,
+      "step": 1101
+    },
+    {
+      "epoch": 2.420647995606809,
+      "grad_norm": 0.1788308322429657,
+      "learning_rate": 6.425081433224755e-05,
+      "loss": 1.3875,
+      "step": 1102
+    },
+    {
+      "epoch": 2.4228445908841296,
+      "grad_norm": 0.2724817395210266,
+      "learning_rate": 6.400651465798045e-05,
+      "loss": 1.3431,
+      "step": 1103
+    },
+    {
+      "epoch": 2.4250411861614496,
+      "grad_norm": 0.20094607770442963,
+      "learning_rate": 6.376221498371336e-05,
+      "loss": 1.3406,
+      "step": 1104
+    },
+    {
+      "epoch": 2.42723778143877,
+      "grad_norm": 0.2948630452156067,
+      "learning_rate": 6.351791530944625e-05,
+      "loss": 1.4169,
+      "step": 1105
+    },
+    {
+      "epoch": 2.42943437671609,
+      "grad_norm": 0.18700957298278809,
+      "learning_rate": 6.327361563517915e-05,
+      "loss": 1.1564,
+      "step": 1106
+    },
+    {
+      "epoch": 2.4316309719934104,
+      "grad_norm": 0.4981689453125,
+      "learning_rate": 6.302931596091205e-05,
+      "loss": 1.5646,
+      "step": 1107
+    },
+    {
+      "epoch": 2.4338275672707304,
+      "grad_norm": 0.22117353975772858,
+      "learning_rate": 6.278501628664495e-05,
+      "loss": 1.6978,
+      "step": 1108
+    },
+    {
+      "epoch": 2.4360241625480503,
+      "grad_norm": 0.16674844920635223,
+      "learning_rate": 6.254071661237784e-05,
+      "loss": 1.2655,
+      "step": 1109
+    },
+    {
+      "epoch": 2.4382207578253707,
+      "grad_norm": 0.15081451833248138,
+      "learning_rate": 6.229641693811075e-05,
+      "loss": 1.2796,
+      "step": 1110
+    },
+    {
+      "epoch": 2.4404173531026907,
+      "grad_norm": 0.1461818963289261,
+      "learning_rate": 6.205211726384365e-05,
+      "loss": 1.2428,
+      "step": 1111
+    },
+    {
+      "epoch": 2.442613948380011,
+      "grad_norm": 0.28263601660728455,
+      "learning_rate": 6.180781758957654e-05,
+      "loss": 1.1012,
+      "step": 1112
+    },
+    {
+      "epoch": 2.444810543657331,
+      "grad_norm": 2.4464070796966553,
+      "learning_rate": 6.156351791530944e-05,
+      "loss": 1.9667,
+      "step": 1113
+    },
+    {
+      "epoch": 2.4470071389346515,
+      "grad_norm": 0.18488647043704987,
+      "learning_rate": 6.131921824104234e-05,
+      "loss": 1.47,
+      "step": 1114
+    },
+    {
+      "epoch": 2.4492037342119715,
+      "grad_norm": 0.16672752797603607,
+      "learning_rate": 6.107491856677525e-05,
+      "loss": 1.392,
+      "step": 1115
+    },
+    {
+      "epoch": 2.4514003294892914,
+      "grad_norm": 0.3268401622772217,
+      "learning_rate": 6.083061889250814e-05,
+      "loss": 1.1438,
+      "step": 1116
+    },
+    {
+      "epoch": 2.453596924766612,
+      "grad_norm": 0.5543538928031921,
+      "learning_rate": 6.058631921824104e-05,
+      "loss": 1.4569,
+      "step": 1117
+    },
+    {
+      "epoch": 2.455793520043932,
+      "grad_norm": 1.67349374294281,
+      "learning_rate": 6.0342019543973935e-05,
+      "loss": 1.3915,
+      "step": 1118
+    },
+    {
+      "epoch": 2.4579901153212522,
+      "grad_norm": 0.19056463241577148,
+      "learning_rate": 6.009771986970684e-05,
+      "loss": 1.3632,
+      "step": 1119
+    },
+    {
+      "epoch": 2.460186710598572,
+      "grad_norm": 2.632965564727783,
+      "learning_rate": 5.9853420195439734e-05,
+      "loss": 1.2753,
+      "step": 1120
+    },
+    {
+      "epoch": 2.4623833058758926,
+      "grad_norm": 0.873747706413269,
+      "learning_rate": 5.9609120521172636e-05,
+      "loss": 1.3508,
+      "step": 1121
+    },
+    {
+      "epoch": 2.4645799011532126,
+      "grad_norm": 0.26368191838264465,
+      "learning_rate": 5.936482084690553e-05,
+      "loss": 1.8327,
+      "step": 1122
+    },
+    {
+      "epoch": 2.4667764964305325,
+      "grad_norm": 0.16192562878131866,
+      "learning_rate": 5.9120521172638435e-05,
+      "loss": 1.2338,
+      "step": 1123
+    },
+    {
+      "epoch": 2.468973091707853,
+      "grad_norm": 0.2656403183937073,
+      "learning_rate": 5.887622149837133e-05,
+      "loss": 1.5502,
+      "step": 1124
+    },
+    {
+      "epoch": 2.471169686985173,
+      "grad_norm": 0.16509686410427094,
+      "learning_rate": 5.863192182410423e-05,
+      "loss": 1.186,
+      "step": 1125
+    },
+    {
+      "epoch": 2.473366282262493,
+      "grad_norm": 0.22797688841819763,
+      "learning_rate": 5.838762214983713e-05,
+      "loss": 1.5835,
+      "step": 1126
+    },
+    {
+      "epoch": 2.4755628775398133,
+      "grad_norm": 0.1898186057806015,
+      "learning_rate": 5.8143322475570026e-05,
+      "loss": 1.1098,
+      "step": 1127
+    },
+    {
+      "epoch": 2.4777594728171333,
+      "grad_norm": 0.14500586688518524,
+      "learning_rate": 5.789902280130293e-05,
+      "loss": 1.2934,
+      "step": 1128
+    },
+    {
+      "epoch": 2.4799560680944537,
+      "grad_norm": 0.27298539876937866,
+      "learning_rate": 5.7654723127035825e-05,
+      "loss": 1.92,
+      "step": 1129
+    },
+    {
+      "epoch": 2.4821526633717736,
+      "grad_norm": 0.22553841769695282,
+      "learning_rate": 5.741042345276873e-05,
+      "loss": 1.5544,
+      "step": 1130
+    },
+    {
+      "epoch": 2.484349258649094,
+      "grad_norm": 0.2400488555431366,
+      "learning_rate": 5.716612377850162e-05,
+      "loss": 1.5767,
+      "step": 1131
+    },
+    {
+      "epoch": 2.486545853926414,
+      "grad_norm": 0.16928952932357788,
+      "learning_rate": 5.6921824104234526e-05,
+      "loss": 1.3058,
+      "step": 1132
+    },
+    {
+      "epoch": 2.488742449203734,
+      "grad_norm": 0.259355366230011,
+      "learning_rate": 5.667752442996742e-05,
+      "loss": 1.0904,
+      "step": 1133
+    },
+    {
+      "epoch": 2.4909390444810544,
+      "grad_norm": 1.004323124885559,
+      "learning_rate": 5.643322475570032e-05,
+      "loss": 1.4417,
+      "step": 1134
+    },
+    {
+      "epoch": 2.4931356397583744,
+      "grad_norm": 0.2255350649356842,
+      "learning_rate": 5.618892508143322e-05,
+      "loss": 1.4313,
+      "step": 1135
+    },
+    {
+      "epoch": 2.495332235035695,
+      "grad_norm": 4.243278980255127,
+      "learning_rate": 5.594462540716612e-05,
+      "loss": 1.2421,
+      "step": 1136
+    },
+    {
+      "epoch": 2.4975288303130148,
+      "grad_norm": 0.1794518083333969,
+      "learning_rate": 5.570032573289902e-05,
+      "loss": 1.0562,
+      "step": 1137
+    },
+    {
+      "epoch": 2.499725425590335,
+      "grad_norm": 0.19731611013412476,
+      "learning_rate": 5.5456026058631915e-05,
+      "loss": 1.3982,
+      "step": 1138
+    },
+    {
+      "epoch": 2.501922020867655,
+      "grad_norm": 0.14244674146175385,
+      "learning_rate": 5.521172638436482e-05,
+      "loss": 1.3165,
+      "step": 1139
+    },
+    {
+      "epoch": 2.504118616144975,
+      "grad_norm": 0.5229745507240295,
+      "learning_rate": 5.4967426710097714e-05,
+      "loss": 1.2491,
+      "step": 1140
+    },
+    {
+      "epoch": 2.5063152114222955,
+      "grad_norm": 0.19223928451538086,
+      "learning_rate": 5.472312703583061e-05,
+      "loss": 1.2914,
+      "step": 1141
+    },
+    {
+      "epoch": 2.5085118066996155,
+      "grad_norm": 0.21131464838981628,
+      "learning_rate": 5.447882736156351e-05,
+      "loss": 1.3862,
+      "step": 1142
+    },
+    {
+      "epoch": 2.510708401976936,
+      "grad_norm": 0.16256894171237946,
+      "learning_rate": 5.423452768729641e-05,
+      "loss": 1.4503,
+      "step": 1143
+    },
+    {
+      "epoch": 2.512904997254256,
+      "grad_norm": 0.18625320494174957,
+      "learning_rate": 5.399022801302931e-05,
+      "loss": 1.4381,
+      "step": 1144
+    },
+    {
+      "epoch": 2.5151015925315763,
+      "grad_norm": 0.23135897517204285,
+      "learning_rate": 5.374592833876221e-05,
+      "loss": 1.3015,
+      "step": 1145
+    },
+    {
+      "epoch": 2.5172981878088962,
+      "grad_norm": 0.2197522521018982,
+      "learning_rate": 5.350162866449511e-05,
+      "loss": 1.2598,
+      "step": 1146
+    },
+    {
+      "epoch": 2.519494783086216,
+      "grad_norm": 0.22324113547801971,
+      "learning_rate": 5.3257328990228006e-05,
+      "loss": 1.3558,
+      "step": 1147
+    },
+    {
+      "epoch": 2.5216913783635366,
+      "grad_norm": 0.19673176109790802,
+      "learning_rate": 5.301302931596091e-05,
+      "loss": 1.2952,
+      "step": 1148
+    },
+    {
+      "epoch": 2.5238879736408566,
+      "grad_norm": 0.2057698518037796,
+      "learning_rate": 5.2768729641693805e-05,
+      "loss": 1.5265,
+      "step": 1149
+    },
+    {
+      "epoch": 2.5260845689181766,
+      "grad_norm": 0.28150543570518494,
+      "learning_rate": 5.25244299674267e-05,
+      "loss": 1.7445,
+      "step": 1150
+    },
+    {
+      "epoch": 2.528281164195497,
+      "grad_norm": 0.2568766474723816,
+      "learning_rate": 5.2280130293159604e-05,
+      "loss": 1.5031,
+      "step": 1151
+    },
+    {
+      "epoch": 2.5304777594728174,
+      "grad_norm": 0.21097829937934875,
+      "learning_rate": 5.20358306188925e-05,
+      "loss": 1.3976,
+      "step": 1152
+    },
+    {
+      "epoch": 2.5326743547501374,
+      "grad_norm": 0.19125327467918396,
+      "learning_rate": 5.17915309446254e-05,
+      "loss": 1.3197,
+      "step": 1153
+    },
+    {
+      "epoch": 2.5348709500274573,
+      "grad_norm": 0.17177650332450867,
+      "learning_rate": 5.15472312703583e-05,
+      "loss": 1.2558,
+      "step": 1154
+    },
+    {
+      "epoch": 2.5370675453047777,
+      "grad_norm": 0.19873207807540894,
+      "learning_rate": 5.13029315960912e-05,
+      "loss": 1.3715,
+      "step": 1155
+    },
+    {
+      "epoch": 2.5392641405820977,
+      "grad_norm": 0.5406712293624878,
+      "learning_rate": 5.10586319218241e-05,
+      "loss": 1.7213,
+      "step": 1156
+    },
+    {
+      "epoch": 2.5414607358594177,
+      "grad_norm": 0.3400755822658539,
+      "learning_rate": 5.081433224755699e-05,
+      "loss": 1.591,
+      "step": 1157
+    },
+    {
+      "epoch": 2.543657331136738,
+      "grad_norm": 0.17038463056087494,
+      "learning_rate": 5.0570032573289896e-05,
+      "loss": 1.241,
+      "step": 1158
+    },
+    {
+      "epoch": 2.545853926414058,
+      "grad_norm": 0.2030758261680603,
+      "learning_rate": 5.032573289902279e-05,
+      "loss": 1.438,
+      "step": 1159
+    },
+    {
+      "epoch": 2.5480505216913785,
+      "grad_norm": 0.15319815278053284,
+      "learning_rate": 5.0081433224755695e-05,
+      "loss": 1.2487,
+      "step": 1160
+    },
+    {
+      "epoch": 2.5502471169686984,
+      "grad_norm": 0.22111478447914124,
+      "learning_rate": 4.983713355048859e-05,
+      "loss": 1.1426,
+      "step": 1161
+    },
+    {
+      "epoch": 2.552443712246019,
+      "grad_norm": 0.1913571059703827,
+      "learning_rate": 4.9592833876221493e-05,
+      "loss": 1.3721,
+      "step": 1162
+    },
+    {
+      "epoch": 2.554640307523339,
+      "grad_norm": 0.23786605894565582,
+      "learning_rate": 4.934853420195439e-05,
+      "loss": 1.1621,
+      "step": 1163
+    },
+    {
+      "epoch": 2.556836902800659,
+      "grad_norm": 0.558089017868042,
+      "learning_rate": 4.910423452768729e-05,
+      "loss": 1.3955,
+      "step": 1164
+    },
+    {
+      "epoch": 2.559033498077979,
+      "grad_norm": 0.1649811565876007,
+      "learning_rate": 4.885993485342019e-05,
+      "loss": 1.4955,
+      "step": 1165
+    },
+    {
+      "epoch": 2.561230093355299,
+      "grad_norm": 0.21886514127254486,
+      "learning_rate": 4.8615635179153084e-05,
+      "loss": 1.2865,
+      "step": 1166
+    },
+    {
+      "epoch": 2.5634266886326196,
+      "grad_norm": 0.18356750905513763,
+      "learning_rate": 4.837133550488599e-05,
+      "loss": 1.0189,
+      "step": 1167
+    },
+    {
+      "epoch": 2.5656232839099395,
+      "grad_norm": 0.4854435324668884,
+      "learning_rate": 4.812703583061888e-05,
+      "loss": 1.4723,
+      "step": 1168
+    },
+    {
+      "epoch": 2.56781987918726,
+      "grad_norm": 0.22648486495018005,
+      "learning_rate": 4.7882736156351786e-05,
+      "loss": 1.3402,
+      "step": 1169
+    },
+    {
+      "epoch": 2.57001647446458,
+      "grad_norm": 0.2627280652523041,
+      "learning_rate": 4.763843648208468e-05,
+      "loss": 1.325,
+      "step": 1170
+    },
+    {
+      "epoch": 2.5722130697419,
+      "grad_norm": 0.2345982789993286,
+      "learning_rate": 4.7394136807817584e-05,
+      "loss": 1.3292,
+      "step": 1171
+    },
+    {
+      "epoch": 2.5744096650192203,
+      "grad_norm": 0.21541334688663483,
+      "learning_rate": 4.714983713355048e-05,
+      "loss": 1.273,
+      "step": 1172
+    },
+    {
+      "epoch": 2.5766062602965403,
+      "grad_norm": 0.2103474736213684,
+      "learning_rate": 4.6905537459283376e-05,
+      "loss": 1.3426,
+      "step": 1173
+    },
+    {
+      "epoch": 2.5788028555738602,
+      "grad_norm": 0.16749581694602966,
+      "learning_rate": 4.6661237785016286e-05,
+      "loss": 1.2242,
+      "step": 1174
+    },
+    {
+      "epoch": 2.5809994508511807,
+      "grad_norm": 0.17730233073234558,
+      "learning_rate": 4.641693811074919e-05,
+      "loss": 1.2853,
+      "step": 1175
+    },
+    {
+      "epoch": 2.583196046128501,
+      "grad_norm": 0.22141699492931366,
+      "learning_rate": 4.6172638436482085e-05,
+      "loss": 1.3527,
+      "step": 1176
+    },
+    {
+      "epoch": 2.585392641405821,
+      "grad_norm": 0.19063033163547516,
+      "learning_rate": 4.592833876221498e-05,
+      "loss": 1.6095,
+      "step": 1177
+    },
+    {
+      "epoch": 2.587589236683141,
+      "grad_norm": 0.28201329708099365,
+      "learning_rate": 4.568403908794788e-05,
+      "loss": 1.2788,
+      "step": 1178
+    },
+    {
+      "epoch": 2.5897858319604614,
+      "grad_norm": 0.2236359417438507,
+      "learning_rate": 4.543973941368078e-05,
+      "loss": 1.5177,
+      "step": 1179
+    },
+    {
+      "epoch": 2.5919824272377814,
+      "grad_norm": 0.32610806822776794,
+      "learning_rate": 4.519543973941368e-05,
+      "loss": 1.1558,
+      "step": 1180
+    },
+    {
+      "epoch": 2.5941790225151014,
+      "grad_norm": 2.7527480125427246,
+      "learning_rate": 4.495114006514658e-05,
+      "loss": 1.6848,
+      "step": 1181
+    },
+    {
+      "epoch": 2.5963756177924218,
+      "grad_norm": 0.30730244517326355,
+      "learning_rate": 4.470684039087948e-05,
+      "loss": 1.7142,
+      "step": 1182
+    },
+    {
+      "epoch": 2.598572213069742,
+      "grad_norm": 0.14672213792800903,
+      "learning_rate": 4.446254071661238e-05,
+      "loss": 1.3046,
+      "step": 1183
+    },
+    {
+      "epoch": 2.600768808347062,
+      "grad_norm": 0.19130246341228485,
+      "learning_rate": 4.421824104234527e-05,
+      "loss": 1.3585,
+      "step": 1184
+    },
+    {
+      "epoch": 2.602965403624382,
+      "grad_norm": 0.19376350939273834,
+      "learning_rate": 4.3973941368078175e-05,
+      "loss": 1.3475,
+      "step": 1185
+    },
+    {
+      "epoch": 2.6051619989017025,
+      "grad_norm": 0.2983350455760956,
+      "learning_rate": 4.372964169381107e-05,
+      "loss": 1.1782,
+      "step": 1186
+    },
+    {
+      "epoch": 2.6073585941790225,
+      "grad_norm": 0.757777214050293,
+      "learning_rate": 4.3485342019543974e-05,
+      "loss": 1.1976,
+      "step": 1187
+    },
+    {
+      "epoch": 2.6095551894563425,
+      "grad_norm": 0.2225475311279297,
+      "learning_rate": 4.324104234527687e-05,
+      "loss": 1.5311,
+      "step": 1188
+    },
+    {
+      "epoch": 2.611751784733663,
+      "grad_norm": 0.17348453402519226,
+      "learning_rate": 4.299674267100977e-05,
+      "loss": 1.1365,
+      "step": 1189
+    },
+    {
+      "epoch": 2.613948380010983,
+      "grad_norm": 0.2843315899372101,
+      "learning_rate": 4.275244299674267e-05,
+      "loss": 1.6258,
+      "step": 1190
+    },
+    {
+      "epoch": 2.6161449752883033,
+      "grad_norm": 0.27499476075172424,
+      "learning_rate": 4.250814332247557e-05,
+      "loss": 1.2293,
+      "step": 1191
+    },
+    {
+      "epoch": 2.618341570565623,
+      "grad_norm": 0.34709349274635315,
+      "learning_rate": 4.226384364820847e-05,
+      "loss": 1.1359,
+      "step": 1192
+    },
+    {
+      "epoch": 2.6205381658429436,
+      "grad_norm": 0.26161327958106995,
+      "learning_rate": 4.2019543973941364e-05,
+      "loss": 2.2024,
+      "step": 1193
+    },
+    {
+      "epoch": 2.6227347611202636,
+      "grad_norm": 0.2144954353570938,
+      "learning_rate": 4.1775244299674266e-05,
+      "loss": 1.4153,
+      "step": 1194
+    },
+    {
+      "epoch": 2.6249313563975836,
+      "grad_norm": 0.19181737303733826,
+      "learning_rate": 4.153094462540716e-05,
+      "loss": 0.9028,
+      "step": 1195
+    },
+    {
+      "epoch": 2.627127951674904,
+      "grad_norm": 0.28709062933921814,
+      "learning_rate": 4.1286644951140065e-05,
+      "loss": 1.4736,
+      "step": 1196
+    },
+    {
+      "epoch": 2.629324546952224,
+      "grad_norm": 0.3244180977344513,
+      "learning_rate": 4.104234527687296e-05,
+      "loss": 1.1751,
+      "step": 1197
+    },
+    {
+      "epoch": 2.6315211422295444,
+      "grad_norm": 0.16790807247161865,
+      "learning_rate": 4.0798045602605864e-05,
+      "loss": 1.3708,
+      "step": 1198
+    },
+    {
+      "epoch": 2.6337177375068643,
+      "grad_norm": 0.18200695514678955,
+      "learning_rate": 4.055374592833876e-05,
+      "loss": 1.1898,
+      "step": 1199
+    },
+    {
+      "epoch": 2.6359143327841847,
+      "grad_norm": 0.29705825448036194,
+      "learning_rate": 4.0309446254071656e-05,
+      "loss": 1.3749,
+      "step": 1200
+    },
+    {
+      "epoch": 2.6381109280615047,
+      "grad_norm": 1.334060549736023,
+      "learning_rate": 4.006514657980456e-05,
+      "loss": 1.2672,
+      "step": 1201
+    },
+    {
+      "epoch": 2.6403075233388247,
+      "grad_norm": 0.2507898807525635,
+      "learning_rate": 3.9820846905537454e-05,
+      "loss": 1.1254,
+      "step": 1202
+    },
+    {
+      "epoch": 2.642504118616145,
+      "grad_norm": 0.1816466897726059,
+      "learning_rate": 3.957654723127036e-05,
+      "loss": 1.6106,
+      "step": 1203
+    },
+    {
+      "epoch": 2.644700713893465,
+      "grad_norm": 0.18348409235477448,
+      "learning_rate": 3.933224755700325e-05,
+      "loss": 1.3564,
+      "step": 1204
+    },
+    {
+      "epoch": 2.646897309170785,
+      "grad_norm": 0.6410112977027893,
+      "learning_rate": 3.9087947882736156e-05,
+      "loss": 1.558,
+      "step": 1205
+    },
+    {
+      "epoch": 2.6490939044481054,
+      "grad_norm": 0.19577385485172272,
+      "learning_rate": 3.884364820846905e-05,
+      "loss": 1.5423,
+      "step": 1206
+    },
+    {
+      "epoch": 2.651290499725426,
+      "grad_norm": 0.20101633667945862,
+      "learning_rate": 3.8599348534201955e-05,
+      "loss": 1.3804,
+      "step": 1207
+    },
+    {
+      "epoch": 2.653487095002746,
+      "grad_norm": 0.24588535726070404,
+      "learning_rate": 3.835504885993485e-05,
+      "loss": 1.4992,
+      "step": 1208
+    },
+    {
+      "epoch": 2.655683690280066,
+      "grad_norm": 0.6213909983634949,
+      "learning_rate": 3.8110749185667747e-05,
+      "loss": 1.3359,
+      "step": 1209
+    },
+    {
+      "epoch": 2.657880285557386,
+      "grad_norm": 0.18846510350704193,
+      "learning_rate": 3.786644951140065e-05,
+      "loss": 1.5062,
+      "step": 1210
+    },
+    {
+      "epoch": 2.660076880834706,
+      "grad_norm": 0.18832895159721375,
+      "learning_rate": 3.7622149837133545e-05,
+      "loss": 1.2799,
+      "step": 1211
+    },
+    {
+      "epoch": 2.662273476112026,
+      "grad_norm": 0.28692084550857544,
+      "learning_rate": 3.737785016286645e-05,
+      "loss": 1.3425,
+      "step": 1212
+    },
+    {
+      "epoch": 2.6644700713893466,
+      "grad_norm": 0.22624464333057404,
+      "learning_rate": 3.7133550488599344e-05,
+      "loss": 1.3175,
+      "step": 1213
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 0.2292485386133194,
+      "learning_rate": 3.688925081433225e-05,
+      "loss": 1.3352,
+      "step": 1214
+    },
+    {
+      "epoch": 2.668863261943987,
+      "grad_norm": 0.2791121304035187,
+      "learning_rate": 3.664495114006514e-05,
+      "loss": 1.2544,
+      "step": 1215
+    },
+    {
+      "epoch": 2.671059857221307,
+      "grad_norm": 0.6891103386878967,
+      "learning_rate": 3.640065146579804e-05,
+      "loss": 1.4097,
+      "step": 1216
+    },
+    {
+      "epoch": 2.6732564524986273,
+      "grad_norm": 0.1824502944946289,
+      "learning_rate": 3.615635179153094e-05,
+      "loss": 1.2962,
+      "step": 1217
+    },
+    {
+      "epoch": 2.6754530477759473,
+      "grad_norm": 0.18341611325740814,
+      "learning_rate": 3.591205211726384e-05,
+      "loss": 1.3036,
+      "step": 1218
+    },
+    {
+      "epoch": 2.6776496430532672,
+      "grad_norm": 0.1843859702348709,
+      "learning_rate": 3.566775244299674e-05,
+      "loss": 1.1627,
+      "step": 1219
+    },
+    {
+      "epoch": 2.6798462383305877,
+      "grad_norm": 0.17254804074764252,
+      "learning_rate": 3.5423452768729636e-05,
+      "loss": 1.1228,
+      "step": 1220
+    },
+    {
+      "epoch": 2.6820428336079076,
+      "grad_norm": 1.8023040294647217,
+      "learning_rate": 3.517915309446254e-05,
+      "loss": 1.7643,
+      "step": 1221
+    },
+    {
+      "epoch": 2.684239428885228,
+      "grad_norm": 0.41080448031425476,
+      "learning_rate": 3.4934853420195435e-05,
+      "loss": 1.1207,
+      "step": 1222
+    },
+    {
+      "epoch": 2.686436024162548,
+      "grad_norm": 0.14066874980926514,
+      "learning_rate": 3.469055374592834e-05,
+      "loss": 1.3565,
+      "step": 1223
+    },
+    {
+      "epoch": 2.6886326194398684,
+      "grad_norm": 0.1875414401292801,
+      "learning_rate": 3.4446254071661234e-05,
+      "loss": 1.3574,
+      "step": 1224
+    },
+    {
+      "epoch": 2.6908292147171884,
+      "grad_norm": 0.6480764746665955,
+      "learning_rate": 3.420195439739413e-05,
+      "loss": 1.4946,
+      "step": 1225
+    },
+    {
+      "epoch": 2.6930258099945084,
+      "grad_norm": 0.21886175870895386,
+      "learning_rate": 3.395765472312703e-05,
+      "loss": 1.0586,
+      "step": 1226
+    },
+    {
+      "epoch": 2.6952224052718288,
+      "grad_norm": 0.30352267622947693,
+      "learning_rate": 3.371335504885993e-05,
+      "loss": 1.6504,
+      "step": 1227
+    },
+    {
+      "epoch": 2.6974190005491487,
+      "grad_norm": 0.14534388482570648,
+      "learning_rate": 3.346905537459283e-05,
+      "loss": 1.1565,
+      "step": 1228
+    },
+    {
+      "epoch": 2.6996155958264687,
+      "grad_norm": 0.20326562225818634,
+      "learning_rate": 3.322475570032573e-05,
+      "loss": 1.7595,
+      "step": 1229
+    },
+    {
+      "epoch": 2.701812191103789,
+      "grad_norm": 0.21642982959747314,
+      "learning_rate": 3.298045602605863e-05,
+      "loss": 1.2617,
+      "step": 1230
+    },
+    {
+      "epoch": 2.7040087863811095,
+      "grad_norm": 0.15716630220413208,
+      "learning_rate": 3.273615635179153e-05,
+      "loss": 1.0701,
+      "step": 1231
+    },
+    {
+      "epoch": 2.7062053816584295,
+      "grad_norm": 0.19693447649478912,
+      "learning_rate": 3.249185667752443e-05,
+      "loss": 1.4012,
+      "step": 1232
+    },
+    {
+      "epoch": 2.7084019769357495,
+      "grad_norm": 0.8351868391036987,
+      "learning_rate": 3.224755700325733e-05,
+      "loss": 1.3601,
+      "step": 1233
+    },
+    {
+      "epoch": 2.71059857221307,
+      "grad_norm": 0.3937622308731079,
+      "learning_rate": 3.200325732899023e-05,
+      "loss": 1.2861,
+      "step": 1234
+    },
+    {
+      "epoch": 2.71279516749039,
+      "grad_norm": 0.18516786396503448,
+      "learning_rate": 3.175895765472312e-05,
+      "loss": 1.4057,
+      "step": 1235
+    },
+    {
+      "epoch": 2.71499176276771,
+      "grad_norm": 0.2850317358970642,
+      "learning_rate": 3.1514657980456026e-05,
+      "loss": 1.7239,
+      "step": 1236
+    },
+    {
+      "epoch": 2.7171883580450302,
+      "grad_norm": 0.1785927265882492,
+      "learning_rate": 3.127035830618892e-05,
+      "loss": 1.3401,
+      "step": 1237
+    },
+    {
+      "epoch": 2.71938495332235,
+      "grad_norm": 0.18877708911895752,
+      "learning_rate": 3.1026058631921825e-05,
+      "loss": 1.3602,
+      "step": 1238
+    },
+    {
+      "epoch": 2.7215815485996706,
+      "grad_norm": 0.5423057675361633,
+      "learning_rate": 3.078175895765472e-05,
+      "loss": 1.3301,
+      "step": 1239
+    },
+    {
+      "epoch": 2.7237781438769906,
+      "grad_norm": 0.1610414832830429,
+      "learning_rate": 3.0537459283387624e-05,
+      "loss": 1.3729,
+      "step": 1240
+    },
+    {
+      "epoch": 2.725974739154311,
+      "grad_norm": 0.34224772453308105,
+      "learning_rate": 3.029315960912052e-05,
+      "loss": 1.2707,
+      "step": 1241
+    },
+    {
+      "epoch": 2.728171334431631,
+      "grad_norm": 0.20006833970546722,
+      "learning_rate": 3.004885993485342e-05,
+      "loss": 1.2727,
+      "step": 1242
+    },
+    {
+      "epoch": 2.730367929708951,
+      "grad_norm": 0.1688380241394043,
+      "learning_rate": 2.9804560260586318e-05,
+      "loss": 1.4203,
+      "step": 1243
+    },
+    {
+      "epoch": 2.7325645249862713,
+      "grad_norm": 0.22224003076553345,
+      "learning_rate": 2.9560260586319218e-05,
+      "loss": 1.4129,
+      "step": 1244
+    },
+    {
+      "epoch": 2.7347611202635913,
+      "grad_norm": 0.3277818262577057,
+      "learning_rate": 2.9315960912052114e-05,
+      "loss": 1.5047,
+      "step": 1245
+    },
+    {
+      "epoch": 2.7369577155409117,
+      "grad_norm": 0.20604190230369568,
+      "learning_rate": 2.9071661237785013e-05,
+      "loss": 1.3863,
+      "step": 1246
+    },
+    {
+      "epoch": 2.7391543108182317,
+      "grad_norm": 0.19053234159946442,
+      "learning_rate": 2.8827361563517912e-05,
+      "loss": 1.8521,
+      "step": 1247
+    },
+    {
+      "epoch": 2.741350906095552,
+      "grad_norm": 0.2088519185781479,
+      "learning_rate": 2.858306188925081e-05,
+      "loss": 1.3624,
+      "step": 1248
+    },
+    {
+      "epoch": 2.743547501372872,
+      "grad_norm": 0.3598547875881195,
+      "learning_rate": 2.833876221498371e-05,
+      "loss": 1.4652,
+      "step": 1249
+    },
+    {
+      "epoch": 2.745744096650192,
+      "grad_norm": 0.2859017848968506,
+      "learning_rate": 2.809446254071661e-05,
+      "loss": 1.3566,
+      "step": 1250
+    },
+    {
+      "epoch": 2.7479406919275124,
+      "grad_norm": 0.20389781892299652,
+      "learning_rate": 2.785016286644951e-05,
+      "loss": 1.587,
+      "step": 1251
+    },
+    {
+      "epoch": 2.7501372872048324,
+      "grad_norm": 0.30276355147361755,
+      "learning_rate": 2.760586319218241e-05,
+      "loss": 1.6867,
+      "step": 1252
+    },
+    {
+      "epoch": 2.752333882482153,
+      "grad_norm": 0.23027262091636658,
+      "learning_rate": 2.7361563517915305e-05,
+      "loss": 1.3588,
+      "step": 1253
+    },
+    {
+      "epoch": 2.754530477759473,
+      "grad_norm": 0.4418986439704895,
+      "learning_rate": 2.7117263843648204e-05,
+      "loss": 1.3687,
+      "step": 1254
+    },
+    {
+      "epoch": 2.756727073036793,
+      "grad_norm": 0.3999635875225067,
+      "learning_rate": 2.6872964169381104e-05,
+      "loss": 1.497,
+      "step": 1255
+    },
+    {
+      "epoch": 2.758923668314113,
+      "grad_norm": 0.22622758150100708,
+      "learning_rate": 2.6628664495114003e-05,
+      "loss": 1.1507,
+      "step": 1256
+    },
+    {
+      "epoch": 2.761120263591433,
+      "grad_norm": 0.1939053237438202,
+      "learning_rate": 2.6384364820846903e-05,
+      "loss": 1.3661,
+      "step": 1257
+    },
+    {
+      "epoch": 2.7633168588687536,
+      "grad_norm": 0.19766634702682495,
+      "learning_rate": 2.6140065146579802e-05,
+      "loss": 1.1472,
+      "step": 1258
+    },
+    {
+      "epoch": 2.7655134541460735,
+      "grad_norm": 0.34489259123802185,
+      "learning_rate": 2.58957654723127e-05,
+      "loss": 1.4098,
+      "step": 1259
+    },
+    {
+      "epoch": 2.7677100494233935,
+      "grad_norm": 0.23982088267803192,
+      "learning_rate": 2.56514657980456e-05,
+      "loss": 1.6733,
+      "step": 1260
+    },
+    {
+      "epoch": 2.769906644700714,
+      "grad_norm": 0.17743347585201263,
+      "learning_rate": 2.5407166123778497e-05,
+      "loss": 1.3386,
+      "step": 1261
+    },
+    {
+      "epoch": 2.7721032399780343,
+      "grad_norm": 0.4670315384864807,
+      "learning_rate": 2.5162866449511396e-05,
+      "loss": 1.2987,
+      "step": 1262
+    },
+    {
+      "epoch": 2.7742998352553543,
+      "grad_norm": 0.23021093010902405,
+      "learning_rate": 2.4918566775244295e-05,
+      "loss": 1.3906,
+      "step": 1263
+    },
+    {
+      "epoch": 2.7764964305326743,
+      "grad_norm": 0.21674062311649323,
+      "learning_rate": 2.4674267100977195e-05,
+      "loss": 1.6916,
+      "step": 1264
+    },
+    {
+      "epoch": 2.7786930258099947,
+      "grad_norm": 1.0417847633361816,
+      "learning_rate": 2.4429967426710094e-05,
+      "loss": 1.2382,
+      "step": 1265
+    },
+    {
+      "epoch": 2.7808896210873146,
+      "grad_norm": 0.17795318365097046,
+      "learning_rate": 2.4185667752442993e-05,
+      "loss": 1.2259,
+      "step": 1266
+    },
+    {
+      "epoch": 2.7830862163646346,
+      "grad_norm": 0.24960289895534515,
+      "learning_rate": 2.3941368078175893e-05,
+      "loss": 1.3525,
+      "step": 1267
+    },
+    {
+      "epoch": 2.785282811641955,
+      "grad_norm": 0.17415030300617218,
+      "learning_rate": 2.3697068403908792e-05,
+      "loss": 1.2235,
+      "step": 1268
+    },
+    {
+      "epoch": 2.787479406919275,
+      "grad_norm": 0.2175319641828537,
+      "learning_rate": 2.3452768729641688e-05,
+      "loss": 1.0961,
+      "step": 1269
+    },
+    {
+      "epoch": 2.7896760021965954,
+      "grad_norm": 0.20007994771003723,
+      "learning_rate": 2.3208469055374594e-05,
+      "loss": 1.3646,
+      "step": 1270
+    },
+    {
+      "epoch": 2.7918725974739154,
+      "grad_norm": 0.15729592740535736,
+      "learning_rate": 2.296416938110749e-05,
+      "loss": 1.1315,
+      "step": 1271
+    },
+    {
+      "epoch": 2.7940691927512358,
+      "grad_norm": 0.15994858741760254,
+      "learning_rate": 2.271986970684039e-05,
+      "loss": 1.2836,
+      "step": 1272
+    },
+    {
+      "epoch": 2.7962657880285557,
+      "grad_norm": 0.2275433987379074,
+      "learning_rate": 2.247557003257329e-05,
+      "loss": 1.4107,
+      "step": 1273
+    },
+    {
+      "epoch": 2.7984623833058757,
+      "grad_norm": 0.15423938632011414,
+      "learning_rate": 2.223127035830619e-05,
+      "loss": 1.5123,
+      "step": 1274
+    },
+    {
+      "epoch": 2.800658978583196,
+      "grad_norm": 0.3233983516693115,
+      "learning_rate": 2.1986970684039088e-05,
+      "loss": 1.2529,
+      "step": 1275
+    },
+    {
+      "epoch": 2.802855573860516,
+      "grad_norm": 0.15965090692043304,
+      "learning_rate": 2.1742671009771987e-05,
+      "loss": 1.4492,
+      "step": 1276
+    },
+    {
+      "epoch": 2.8050521691378365,
+      "grad_norm": 0.17559552192687988,
+      "learning_rate": 2.1498371335504886e-05,
+      "loss": 1.4506,
+      "step": 1277
+    },
+    {
+      "epoch": 2.8072487644151565,
+      "grad_norm": 0.19457119703292847,
+      "learning_rate": 2.1254071661237786e-05,
+      "loss": 1.4964,
+      "step": 1278
+    },
+    {
+      "epoch": 2.809445359692477,
+      "grad_norm": 0.19493229687213898,
+      "learning_rate": 2.1009771986970682e-05,
+      "loss": 1.2766,
+      "step": 1279
+    },
+    {
+      "epoch": 2.811641954969797,
+      "grad_norm": 0.27747842669487,
+      "learning_rate": 2.076547231270358e-05,
+      "loss": 1.3159,
+      "step": 1280
+    },
+    {
+      "epoch": 2.813838550247117,
+      "grad_norm": 0.17320483922958374,
+      "learning_rate": 2.052117263843648e-05,
+      "loss": 1.3457,
+      "step": 1281
+    },
+    {
+      "epoch": 2.8160351455244372,
+      "grad_norm": 1.3414702415466309,
+      "learning_rate": 2.027687296416938e-05,
+      "loss": 1.201,
+      "step": 1282
+    },
+    {
+      "epoch": 2.818231740801757,
+      "grad_norm": 0.29095005989074707,
+      "learning_rate": 2.003257328990228e-05,
+      "loss": 1.6405,
+      "step": 1283
+    },
+    {
+      "epoch": 2.820428336079077,
+      "grad_norm": 0.20117023587226868,
+      "learning_rate": 1.978827361563518e-05,
+      "loss": 1.6701,
+      "step": 1284
+    },
+    {
+      "epoch": 2.8226249313563976,
+      "grad_norm": 1.7894169092178345,
+      "learning_rate": 1.9543973941368078e-05,
+      "loss": 1.516,
+      "step": 1285
+    },
+    {
+      "epoch": 2.824821526633718,
+      "grad_norm": 0.22227610647678375,
+      "learning_rate": 1.9299674267100977e-05,
+      "loss": 1.5905,
+      "step": 1286
+    },
+    {
+      "epoch": 2.827018121911038,
+      "grad_norm": 0.2208711802959442,
+      "learning_rate": 1.9055374592833873e-05,
+      "loss": 1.3667,
+      "step": 1287
+    },
+    {
+      "epoch": 2.829214717188358,
+      "grad_norm": 0.2377820611000061,
+      "learning_rate": 1.8811074918566773e-05,
+      "loss": 1.3712,
+      "step": 1288
+    },
+    {
+      "epoch": 2.8314113124656783,
+      "grad_norm": 0.278088241815567,
+      "learning_rate": 1.8566775244299672e-05,
+      "loss": 1.508,
+      "step": 1289
+    },
+    {
+      "epoch": 2.8336079077429983,
+      "grad_norm": 0.1981515884399414,
+      "learning_rate": 1.832247557003257e-05,
+      "loss": 1.3008,
+      "step": 1290
+    },
+    {
+      "epoch": 2.8358045030203183,
+      "grad_norm": 0.15273354947566986,
+      "learning_rate": 1.807817589576547e-05,
+      "loss": 1.0475,
+      "step": 1291
+    },
+    {
+      "epoch": 2.8380010982976387,
+      "grad_norm": 0.19819973409175873,
+      "learning_rate": 1.783387622149837e-05,
+      "loss": 1.2457,
+      "step": 1292
+    },
+    {
+      "epoch": 2.8401976935749587,
+      "grad_norm": 0.20077751576900482,
+      "learning_rate": 1.758957654723127e-05,
+      "loss": 1.1039,
+      "step": 1293
+    },
+    {
+      "epoch": 2.842394288852279,
+      "grad_norm": 0.2422526478767395,
+      "learning_rate": 1.734527687296417e-05,
+      "loss": 1.5941,
+      "step": 1294
+    },
+    {
+      "epoch": 2.844590884129599,
+      "grad_norm": 0.3003119230270386,
+      "learning_rate": 1.7100977198697065e-05,
+      "loss": 1.3363,
+      "step": 1295
+    },
+    {
+      "epoch": 2.8467874794069195,
+      "grad_norm": 0.20843830704689026,
+      "learning_rate": 1.6856677524429964e-05,
+      "loss": 1.6147,
+      "step": 1296
+    },
+    {
+      "epoch": 2.8489840746842394,
+      "grad_norm": 0.2142113894224167,
+      "learning_rate": 1.6612377850162864e-05,
+      "loss": 1.2372,
+      "step": 1297
+    },
+    {
+      "epoch": 2.8511806699615594,
+      "grad_norm": 0.2853759825229645,
+      "learning_rate": 1.6368078175895766e-05,
+      "loss": 1.648,
+      "step": 1298
+    },
+    {
+      "epoch": 2.85337726523888,
+      "grad_norm": 0.22733373939990997,
+      "learning_rate": 1.6123778501628666e-05,
+      "loss": 1.388,
+      "step": 1299
+    },
+    {
+      "epoch": 2.8555738605161998,
+      "grad_norm": 0.3281534016132355,
+      "learning_rate": 1.587947882736156e-05,
+      "loss": 1.4308,
+      "step": 1300
+    },
+    {
+      "epoch": 2.85777045579352,
+      "grad_norm": 0.16984234750270844,
+      "learning_rate": 1.563517915309446e-05,
+      "loss": 1.4661,
+      "step": 1301
+    },
+    {
+      "epoch": 2.85996705107084,
+      "grad_norm": 0.2253715544939041,
+      "learning_rate": 1.539087947882736e-05,
+      "loss": 1.3447,
+      "step": 1302
+    },
+    {
+      "epoch": 2.8621636463481606,
+      "grad_norm": 0.1885456144809723,
+      "learning_rate": 1.514657980456026e-05,
+      "loss": 1.6502,
+      "step": 1303
+    },
+    {
+      "epoch": 2.8643602416254805,
+      "grad_norm": 0.293314129114151,
+      "learning_rate": 1.4902280130293159e-05,
+      "loss": 1.4008,
+      "step": 1304
+    },
+    {
+      "epoch": 2.8665568369028005,
+      "grad_norm": 0.8386337161064148,
+      "learning_rate": 1.4657980456026057e-05,
+      "loss": 1.2499,
+      "step": 1305
+    },
+    {
+      "epoch": 2.868753432180121,
+      "grad_norm": 0.1734490990638733,
+      "learning_rate": 1.4413680781758956e-05,
+      "loss": 1.3636,
+      "step": 1306
+    },
+    {
+      "epoch": 2.870950027457441,
+      "grad_norm": 0.28471821546554565,
+      "learning_rate": 1.4169381107491856e-05,
+      "loss": 1.426,
+      "step": 1307
+    },
+    {
+      "epoch": 2.873146622734761,
+      "grad_norm": 0.15671572089195251,
+      "learning_rate": 1.3925081433224755e-05,
+      "loss": 1.0689,
+      "step": 1308
+    },
+    {
+      "epoch": 2.8753432180120813,
+      "grad_norm": 0.184584841132164,
+      "learning_rate": 1.3680781758957653e-05,
+      "loss": 1.3194,
+      "step": 1309
+    },
+    {
+      "epoch": 2.8775398132894017,
+      "grad_norm": 0.34168484807014465,
+      "learning_rate": 1.3436482084690552e-05,
+      "loss": 1.5105,
+      "step": 1310
+    },
+    {
+      "epoch": 2.8797364085667216,
+      "grad_norm": 0.16754253208637238,
+      "learning_rate": 1.3192182410423451e-05,
+      "loss": 1.2317,
+      "step": 1311
+    },
+    {
+      "epoch": 2.8819330038440416,
+      "grad_norm": 1.41954505443573,
+      "learning_rate": 1.294788273615635e-05,
+      "loss": 1.3588,
+      "step": 1312
+    },
+    {
+      "epoch": 2.884129599121362,
+      "grad_norm": 0.15078102052211761,
+      "learning_rate": 1.2703583061889248e-05,
+      "loss": 1.2515,
+      "step": 1313
+    },
+    {
+      "epoch": 2.886326194398682,
+      "grad_norm": 0.18679364025592804,
+      "learning_rate": 1.2459283387622148e-05,
+      "loss": 1.2782,
+      "step": 1314
+    },
+    {
+      "epoch": 2.888522789676002,
+      "grad_norm": 0.25878170132637024,
+      "learning_rate": 1.2214983713355047e-05,
+      "loss": 1.4061,
+      "step": 1315
+    },
+    {
+      "epoch": 2.8907193849533224,
+      "grad_norm": 0.28154006600379944,
+      "learning_rate": 1.1970684039087946e-05,
+      "loss": 1.2712,
+      "step": 1316
+    },
+    {
+      "epoch": 2.892915980230643,
+      "grad_norm": 1.0967358350753784,
+      "learning_rate": 1.1726384364820844e-05,
+      "loss": 1.4573,
+      "step": 1317
+    },
+    {
+      "epoch": 2.8951125755079627,
+      "grad_norm": 0.3252203166484833,
+      "learning_rate": 1.1482084690553745e-05,
+      "loss": 1.2722,
+      "step": 1318
+    },
+    {
+      "epoch": 2.8973091707852827,
+      "grad_norm": 0.22345015406608582,
+      "learning_rate": 1.1237785016286644e-05,
+      "loss": 1.1501,
+      "step": 1319
+    },
+    {
+      "epoch": 2.899505766062603,
+      "grad_norm": 0.28263670206069946,
+      "learning_rate": 1.0993485342019544e-05,
+      "loss": 1.6963,
+      "step": 1320
+    },
+    {
+      "epoch": 2.901702361339923,
+      "grad_norm": 0.20358389616012573,
+      "learning_rate": 1.0749185667752443e-05,
+      "loss": 1.1436,
+      "step": 1321
+    },
+    {
+      "epoch": 2.903898956617243,
+      "grad_norm": 0.21308307349681854,
+      "learning_rate": 1.0504885993485341e-05,
+      "loss": 1.2807,
+      "step": 1322
+    },
+    {
+      "epoch": 2.9060955518945635,
+      "grad_norm": 0.3190558850765228,
+      "learning_rate": 1.026058631921824e-05,
+      "loss": 1.1563,
+      "step": 1323
+    },
+    {
+      "epoch": 2.9082921471718834,
+      "grad_norm": 0.32753413915634155,
+      "learning_rate": 1.001628664495114e-05,
+      "loss": 1.3948,
+      "step": 1324
+    },
+    {
+      "epoch": 2.910488742449204,
+      "grad_norm": 0.2112674117088318,
+      "learning_rate": 9.771986970684039e-06,
+      "loss": 1.4209,
+      "step": 1325
+    },
+    {
+      "epoch": 2.912685337726524,
+      "grad_norm": 0.322493314743042,
+      "learning_rate": 9.527687296416937e-06,
+      "loss": 1.1309,
+      "step": 1326
+    },
+    {
+      "epoch": 2.9148819330038442,
+      "grad_norm": 0.2189081758260727,
+      "learning_rate": 9.283387622149836e-06,
+      "loss": 1.4787,
+      "step": 1327
+    },
+    {
+      "epoch": 2.917078528281164,
+      "grad_norm": 0.22521480917930603,
+      "learning_rate": 9.039087947882735e-06,
+      "loss": 1.1971,
+      "step": 1328
+    },
+    {
+      "epoch": 2.919275123558484,
+      "grad_norm": 1.6270073652267456,
+      "learning_rate": 8.794788273615635e-06,
+      "loss": 1.1047,
+      "step": 1329
+    },
+    {
+      "epoch": 2.9214717188358046,
+      "grad_norm": 0.21770431101322174,
+      "learning_rate": 8.550488599348532e-06,
+      "loss": 1.2067,
+      "step": 1330
+    },
+    {
+      "epoch": 2.9236683141131246,
+      "grad_norm": 1.7811280488967896,
+      "learning_rate": 8.306188925081432e-06,
+      "loss": 1.7333,
+      "step": 1331
+    },
+    {
+      "epoch": 2.925864909390445,
+      "grad_norm": 0.44827425479888916,
+      "learning_rate": 8.061889250814333e-06,
+      "loss": 0.7931,
+      "step": 1332
+    },
+    {
+      "epoch": 2.928061504667765,
+      "grad_norm": 0.25730210542678833,
+      "learning_rate": 7.81758957654723e-06,
+      "loss": 1.548,
+      "step": 1333
+    },
+    {
+      "epoch": 2.9302580999450853,
+      "grad_norm": 0.22819942235946655,
+      "learning_rate": 7.57328990228013e-06,
+      "loss": 1.2598,
+      "step": 1334
+    },
+    {
+      "epoch": 2.9324546952224053,
+      "grad_norm": 0.28666070103645325,
+      "learning_rate": 7.328990228013028e-06,
+      "loss": 1.2361,
+      "step": 1335
+    },
+    {
+      "epoch": 2.9346512904997253,
+      "grad_norm": 0.23839588463306427,
+      "learning_rate": 7.084690553745928e-06,
+      "loss": 1.2594,
+      "step": 1336
+    },
+    {
+      "epoch": 2.9368478857770457,
+      "grad_norm": 0.1970425397157669,
+      "learning_rate": 6.840390879478826e-06,
+      "loss": 1.3193,
+      "step": 1337
+    },
+    {
+      "epoch": 2.9390444810543657,
+      "grad_norm": 0.6420565247535706,
+      "learning_rate": 6.596091205211726e-06,
+      "loss": 1.2279,
+      "step": 1338
+    },
+    {
+      "epoch": 2.9412410763316856,
+      "grad_norm": 0.40793150663375854,
+      "learning_rate": 6.351791530944624e-06,
+      "loss": 1.1993,
+      "step": 1339
+    },
+    {
+      "epoch": 2.943437671609006,
+      "grad_norm": 0.2739318907260895,
+      "learning_rate": 6.1074918566775235e-06,
+      "loss": 0.9406,
+      "step": 1340
+    },
+    {
+      "epoch": 2.9456342668863265,
+      "grad_norm": 0.16102901101112366,
+      "learning_rate": 5.863192182410422e-06,
+      "loss": 1.1616,
+      "step": 1341
+    },
+    {
+      "epoch": 2.9478308621636464,
+      "grad_norm": 0.15330617129802704,
+      "learning_rate": 5.618892508143322e-06,
+      "loss": 1.3064,
+      "step": 1342
+    },
+    {
+      "epoch": 2.9500274574409664,
+      "grad_norm": 0.5324276685714722,
+      "learning_rate": 5.374592833876222e-06,
+      "loss": 1.3901,
+      "step": 1343
+    },
+    {
+      "epoch": 2.952224052718287,
+      "grad_norm": 0.15149946510791779,
+      "learning_rate": 5.13029315960912e-06,
+      "loss": 1.346,
+      "step": 1344
+    },
+    {
+      "epoch": 2.9544206479956068,
+      "grad_norm": 0.3765592575073242,
+      "learning_rate": 4.8859934853420195e-06,
+      "loss": 1.7075,
+      "step": 1345
+    },
+    {
+      "epoch": 2.9566172432729267,
+      "grad_norm": 0.2435656040906906,
+      "learning_rate": 4.641693811074918e-06,
+      "loss": 1.4176,
+      "step": 1346
+    },
+    {
+      "epoch": 2.958813838550247,
+      "grad_norm": 0.4647420048713684,
+      "learning_rate": 4.397394136807817e-06,
+      "loss": 1.4917,
+      "step": 1347
+    },
+    {
+      "epoch": 2.961010433827567,
+      "grad_norm": 0.23206600546836853,
+      "learning_rate": 4.153094462540716e-06,
+      "loss": 1.2828,
+      "step": 1348
+    },
+    {
+      "epoch": 2.9632070291048875,
+      "grad_norm": 0.1564568281173706,
+      "learning_rate": 3.908794788273615e-06,
+      "loss": 1.5231,
+      "step": 1349
+    },
+    {
+      "epoch": 2.9654036243822075,
+      "grad_norm": 0.31009599566459656,
+      "learning_rate": 3.664495114006514e-06,
+      "loss": 1.3838,
+      "step": 1350
+    },
+    {
+      "epoch": 2.967600219659528,
+      "grad_norm": 0.1965751200914383,
+      "learning_rate": 3.420195439739413e-06,
+      "loss": 1.5125,
+      "step": 1351
+    },
+    {
+      "epoch": 2.969796814936848,
+      "grad_norm": 0.19770434498786926,
+      "learning_rate": 3.175895765472312e-06,
+      "loss": 1.4252,
+      "step": 1352
+    },
+    {
+      "epoch": 2.971993410214168,
+      "grad_norm": 0.2041786015033722,
+      "learning_rate": 2.931596091205211e-06,
+      "loss": 1.2683,
+      "step": 1353
+    },
+    {
+      "epoch": 2.9741900054914883,
+      "grad_norm": 0.20583733916282654,
+      "learning_rate": 2.687296416938111e-06,
+      "loss": 1.5769,
+      "step": 1354
+    },
+    {
+      "epoch": 2.9763866007688082,
+      "grad_norm": 0.2343320995569229,
+      "learning_rate": 2.4429967426710097e-06,
+      "loss": 1.2326,
+      "step": 1355
+    },
+    {
+      "epoch": 2.9785831960461286,
+      "grad_norm": 0.22887475788593292,
+      "learning_rate": 2.1986970684039087e-06,
+      "loss": 1.5565,
+      "step": 1356
+    },
+    {
+      "epoch": 2.9807797913234486,
+      "grad_norm": 0.301126629114151,
+      "learning_rate": 1.9543973941368076e-06,
+      "loss": 1.5802,
+      "step": 1357
+    },
+    {
+      "epoch": 2.982976386600769,
+      "grad_norm": 0.24683403968811035,
+      "learning_rate": 1.7100977198697066e-06,
+      "loss": 1.2981,
+      "step": 1358
+    },
+    {
+      "epoch": 2.985172981878089,
+      "grad_norm": 0.18973901867866516,
+      "learning_rate": 1.4657980456026055e-06,
+      "loss": 1.4462,
+      "step": 1359
+    },
+    {
+      "epoch": 2.987369577155409,
+      "grad_norm": 0.16842716932296753,
+      "learning_rate": 1.2214983713355049e-06,
+      "loss": 1.1468,
+      "step": 1360
+    },
+    {
+      "epoch": 2.9895661724327294,
+      "grad_norm": 0.1672341376543045,
+      "learning_rate": 9.771986970684038e-07,
+      "loss": 1.457,
+      "step": 1361
+    },
+    {
+      "epoch": 2.9917627677100493,
+      "grad_norm": 0.27890270948410034,
+      "learning_rate": 7.328990228013028e-07,
+      "loss": 1.7561,
+      "step": 1362
+    },
+    {
+      "epoch": 2.9939593629873693,
+      "grad_norm": 2.8956868648529053,
+      "learning_rate": 4.885993485342019e-07,
+      "loss": 1.2543,
+      "step": 1363
+    },
+    {
+      "epoch": 2.9961559582646897,
+      "grad_norm": 0.1428501009941101,
+      "learning_rate": 2.4429967426710095e-07,
+      "loss": 1.215,
+      "step": 1364
+    },
+    {
+      "epoch": 2.99835255354201,
+      "grad_norm": 0.13521011173725128,
+      "learning_rate": 0.0,
+      "loss": 1.2596,
+      "step": 1365
+    },
+    {
+      "epoch": 2.99835255354201,
+      "step": 1365,
+      "total_flos": 1.90616367529984e+17,
+      "train_loss": 1.6517934967746666,
+      "train_runtime": 2824.2972,
+      "train_samples_per_second": 7.736,
+      "train_steps_per_second": 0.483
+    }
+  ],
+  "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": 1.90616367529984e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}