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": 0.15748536586761475,
+      "learning_rate": 0.0,
+      "loss": 0.8401,
+      "step": 1
+    },
+    {
+      "epoch": 0.004393190554640308,
+      "grad_norm": 0.1566241979598999,
+      "learning_rate": 2.18978102189781e-06,
+      "loss": 1.0464,
+      "step": 2
+    },
+    {
+      "epoch": 0.006589785831960461,
+      "grad_norm": 0.12935101985931396,
+      "learning_rate": 4.37956204379562e-06,
+      "loss": 0.8651,
+      "step": 3
+    },
+    {
+      "epoch": 0.008786381109280615,
+      "grad_norm": 0.11286711692810059,
+      "learning_rate": 6.56934306569343e-06,
+      "loss": 0.744,
+      "step": 4
+    },
+    {
+      "epoch": 0.010982976386600769,
+      "grad_norm": 0.09725913405418396,
+      "learning_rate": 8.75912408759124e-06,
+      "loss": 0.8383,
+      "step": 5
+    },
+    {
+      "epoch": 0.013179571663920923,
+      "grad_norm": 0.0977504774928093,
+      "learning_rate": 1.0948905109489049e-05,
+      "loss": 0.6631,
+      "step": 6
+    },
+    {
+      "epoch": 0.015376166941241077,
+      "grad_norm": 0.10078343749046326,
+      "learning_rate": 1.313868613138686e-05,
+      "loss": 0.6671,
+      "step": 7
+    },
+    {
+      "epoch": 0.01757276221856123,
+      "grad_norm": 0.1003885269165039,
+      "learning_rate": 1.532846715328467e-05,
+      "loss": 0.7274,
+      "step": 8
+    },
+    {
+      "epoch": 0.019769357495881382,
+      "grad_norm": 0.1751730740070343,
+      "learning_rate": 1.751824817518248e-05,
+      "loss": 1.0782,
+      "step": 9
+    },
+    {
+      "epoch": 0.021965952773201538,
+      "grad_norm": 0.11849772185087204,
+      "learning_rate": 1.9708029197080292e-05,
+      "loss": 0.7394,
+      "step": 10
+    },
+    {
+      "epoch": 0.02416254805052169,
+      "grad_norm": 0.1626003533601761,
+      "learning_rate": 2.1897810218978098e-05,
+      "loss": 0.9005,
+      "step": 11
+    },
+    {
+      "epoch": 0.026359143327841845,
+      "grad_norm": 0.09926193952560425,
+      "learning_rate": 2.408759124087591e-05,
+      "loss": 0.8164,
+      "step": 12
+    },
+    {
+      "epoch": 0.028555738605161998,
+      "grad_norm": 0.12030395120382309,
+      "learning_rate": 2.627737226277372e-05,
+      "loss": 0.7881,
+      "step": 13
+    },
+    {
+      "epoch": 0.030752333882482153,
+      "grad_norm": 0.10779309272766113,
+      "learning_rate": 2.846715328467153e-05,
+      "loss": 0.7969,
+      "step": 14
+    },
+    {
+      "epoch": 0.032948929159802305,
+      "grad_norm": 0.13166797161102295,
+      "learning_rate": 3.065693430656934e-05,
+      "loss": 0.8575,
+      "step": 15
+    },
+    {
+      "epoch": 0.03514552443712246,
+      "grad_norm": 0.11057678610086441,
+      "learning_rate": 3.284671532846715e-05,
+      "loss": 0.6935,
+      "step": 16
+    },
+    {
+      "epoch": 0.037342119714442616,
+      "grad_norm": 0.1497359573841095,
+      "learning_rate": 3.503649635036496e-05,
+      "loss": 0.811,
+      "step": 17
+    },
+    {
+      "epoch": 0.039538714991762765,
+      "grad_norm": 0.16224654018878937,
+      "learning_rate": 3.722627737226277e-05,
+      "loss": 0.8342,
+      "step": 18
+    },
+    {
+      "epoch": 0.04173531026908292,
+      "grad_norm": 0.11993195116519928,
+      "learning_rate": 3.9416058394160584e-05,
+      "loss": 0.8995,
+      "step": 19
+    },
+    {
+      "epoch": 0.043931905546403076,
+      "grad_norm": 0.1456042230129242,
+      "learning_rate": 4.160583941605839e-05,
+      "loss": 0.8057,
+      "step": 20
+    },
+    {
+      "epoch": 0.04612850082372323,
+      "grad_norm": 0.16391023993492126,
+      "learning_rate": 4.3795620437956196e-05,
+      "loss": 0.8953,
+      "step": 21
+    },
+    {
+      "epoch": 0.04832509610104338,
+      "grad_norm": 0.12463803589344025,
+      "learning_rate": 4.5985401459854016e-05,
+      "loss": 0.7705,
+      "step": 22
+    },
+    {
+      "epoch": 0.050521691378363535,
+      "grad_norm": 0.22539383172988892,
+      "learning_rate": 4.817518248175182e-05,
+      "loss": 0.9801,
+      "step": 23
+    },
+    {
+      "epoch": 0.05271828665568369,
+      "grad_norm": 0.17554739117622375,
+      "learning_rate": 5.036496350364963e-05,
+      "loss": 0.8502,
+      "step": 24
+    },
+    {
+      "epoch": 0.054914881933003847,
+      "grad_norm": 0.16498945653438568,
+      "learning_rate": 5.255474452554744e-05,
+      "loss": 0.7728,
+      "step": 25
+    },
+    {
+      "epoch": 0.057111477210323995,
+      "grad_norm": 0.24642176926136017,
+      "learning_rate": 5.4744525547445253e-05,
+      "loss": 0.7592,
+      "step": 26
+    },
+    {
+      "epoch": 0.05930807248764415,
+      "grad_norm": 0.3069516718387604,
+      "learning_rate": 5.693430656934306e-05,
+      "loss": 0.7982,
+      "step": 27
+    },
+    {
+      "epoch": 0.061504667764964306,
+      "grad_norm": 0.1885368525981903,
+      "learning_rate": 5.912408759124087e-05,
+      "loss": 0.7617,
+      "step": 28
+    },
+    {
+      "epoch": 0.06370126304228446,
+      "grad_norm": 0.1696944534778595,
+      "learning_rate": 6.131386861313868e-05,
+      "loss": 0.7347,
+      "step": 29
+    },
+    {
+      "epoch": 0.06589785831960461,
+      "grad_norm": 0.22931723296642303,
+      "learning_rate": 6.350364963503648e-05,
+      "loss": 0.7461,
+      "step": 30
+    },
+    {
+      "epoch": 0.06809445359692477,
+      "grad_norm": 0.23805610835552216,
+      "learning_rate": 6.56934306569343e-05,
+      "loss": 0.7315,
+      "step": 31
+    },
+    {
+      "epoch": 0.07029104887424492,
+      "grad_norm": 0.38632869720458984,
+      "learning_rate": 6.788321167883211e-05,
+      "loss": 0.9196,
+      "step": 32
+    },
+    {
+      "epoch": 0.07248764415156507,
+      "grad_norm": 0.26859116554260254,
+      "learning_rate": 7.007299270072992e-05,
+      "loss": 1.022,
+      "step": 33
+    },
+    {
+      "epoch": 0.07468423942888523,
+      "grad_norm": 0.30093973875045776,
+      "learning_rate": 7.226277372262774e-05,
+      "loss": 0.8109,
+      "step": 34
+    },
+    {
+      "epoch": 0.07688083470620538,
+      "grad_norm": 0.2643658518791199,
+      "learning_rate": 7.445255474452554e-05,
+      "loss": 0.757,
+      "step": 35
+    },
+    {
+      "epoch": 0.07907742998352553,
+      "grad_norm": 0.3769732117652893,
+      "learning_rate": 7.664233576642336e-05,
+      "loss": 0.8267,
+      "step": 36
+    },
+    {
+      "epoch": 0.08127402526084569,
+      "grad_norm": 0.26617488265037537,
+      "learning_rate": 7.883211678832117e-05,
+      "loss": 0.7577,
+      "step": 37
+    },
+    {
+      "epoch": 0.08347062053816584,
+      "grad_norm": 0.3899919390678406,
+      "learning_rate": 8.102189781021897e-05,
+      "loss": 0.8496,
+      "step": 38
+    },
+    {
+      "epoch": 0.085667215815486,
+      "grad_norm": 0.36231234669685364,
+      "learning_rate": 8.321167883211678e-05,
+      "loss": 0.8756,
+      "step": 39
+    },
+    {
+      "epoch": 0.08786381109280615,
+      "grad_norm": 0.3995247483253479,
+      "learning_rate": 8.540145985401459e-05,
+      "loss": 0.5903,
+      "step": 40
+    },
+    {
+      "epoch": 0.0900604063701263,
+      "grad_norm": 0.35032156109809875,
+      "learning_rate": 8.759124087591239e-05,
+      "loss": 0.7711,
+      "step": 41
+    },
+    {
+      "epoch": 0.09225700164744646,
+      "grad_norm": 0.44692444801330566,
+      "learning_rate": 8.978102189781021e-05,
+      "loss": 0.7563,
+      "step": 42
+    },
+    {
+      "epoch": 0.09445359692476661,
+      "grad_norm": 0.34704864025115967,
+      "learning_rate": 9.197080291970803e-05,
+      "loss": 0.6482,
+      "step": 43
+    },
+    {
+      "epoch": 0.09665019220208676,
+      "grad_norm": 0.34864211082458496,
+      "learning_rate": 9.416058394160584e-05,
+      "loss": 0.7607,
+      "step": 44
+    },
+    {
+      "epoch": 0.09884678747940692,
+      "grad_norm": 0.3925572335720062,
+      "learning_rate": 9.635036496350364e-05,
+      "loss": 0.7979,
+      "step": 45
+    },
+    {
+      "epoch": 0.10104338275672707,
+      "grad_norm": 0.31112727522850037,
+      "learning_rate": 9.854014598540145e-05,
+      "loss": 0.6628,
+      "step": 46
+    },
+    {
+      "epoch": 0.10323997803404723,
+      "grad_norm": 0.2704273760318756,
+      "learning_rate": 0.00010072992700729926,
+      "loss": 0.6792,
+      "step": 47
+    },
+    {
+      "epoch": 0.10543657331136738,
+      "grad_norm": 0.3343430161476135,
+      "learning_rate": 0.00010291970802919708,
+      "loss": 0.7592,
+      "step": 48
+    },
+    {
+      "epoch": 0.10763316858868753,
+      "grad_norm": 0.40160536766052246,
+      "learning_rate": 0.00010510948905109488,
+      "loss": 0.7992,
+      "step": 49
+    },
+    {
+      "epoch": 0.10982976386600769,
+      "grad_norm": 0.39016401767730713,
+      "learning_rate": 0.00010729927007299269,
+      "loss": 0.711,
+      "step": 50
+    },
+    {
+      "epoch": 0.11202635914332784,
+      "grad_norm": 0.56025230884552,
+      "learning_rate": 0.00010948905109489051,
+      "loss": 0.7123,
+      "step": 51
+    },
+    {
+      "epoch": 0.11422295442064799,
+      "grad_norm": 0.4479077458381653,
+      "learning_rate": 0.00011167883211678831,
+      "loss": 0.6478,
+      "step": 52
+    },
+    {
+      "epoch": 0.11641954969796815,
+      "grad_norm": 0.5601754784584045,
+      "learning_rate": 0.00011386861313868612,
+      "loss": 0.7053,
+      "step": 53
+    },
+    {
+      "epoch": 0.1186161449752883,
+      "grad_norm": 0.5117803812026978,
+      "learning_rate": 0.00011605839416058394,
+      "loss": 0.5364,
+      "step": 54
+    },
+    {
+      "epoch": 0.12081274025260846,
+      "grad_norm": 0.34400907158851624,
+      "learning_rate": 0.00011824817518248174,
+      "loss": 0.6605,
+      "step": 55
+    },
+    {
+      "epoch": 0.12300933552992861,
+      "grad_norm": 0.6189077496528625,
+      "learning_rate": 0.00012043795620437955,
+      "loss": 0.7678,
+      "step": 56
+    },
+    {
+      "epoch": 0.12520593080724876,
+      "grad_norm": 0.2851059138774872,
+      "learning_rate": 0.00012262773722627736,
+      "loss": 0.5739,
+      "step": 57
+    },
+    {
+      "epoch": 0.12740252608456892,
+      "grad_norm": 0.49091941118240356,
+      "learning_rate": 0.00012481751824817516,
+      "loss": 0.7891,
+      "step": 58
+    },
+    {
+      "epoch": 0.12959912136188906,
+      "grad_norm": 0.33804330229759216,
+      "learning_rate": 0.00012700729927007297,
+      "loss": 0.5649,
+      "step": 59
+    },
+    {
+      "epoch": 0.13179571663920922,
+      "grad_norm": 0.3179687559604645,
+      "learning_rate": 0.00012919708029197077,
+      "loss": 0.6431,
+      "step": 60
+    },
+    {
+      "epoch": 0.13399231191652938,
+      "grad_norm": 0.4577971398830414,
+      "learning_rate": 0.0001313868613138686,
+      "loss": 0.7097,
+      "step": 61
+    },
+    {
+      "epoch": 0.13618890719384955,
+      "grad_norm": 0.29803401231765747,
+      "learning_rate": 0.00013357664233576641,
+      "loss": 0.5639,
+      "step": 62
+    },
+    {
+      "epoch": 0.13838550247116968,
+      "grad_norm": 0.37666216492652893,
+      "learning_rate": 0.00013576642335766422,
+      "loss": 0.5534,
+      "step": 63
+    },
+    {
+      "epoch": 0.14058209774848984,
+      "grad_norm": 0.43315786123275757,
+      "learning_rate": 0.00013795620437956203,
+      "loss": 0.655,
+      "step": 64
+    },
+    {
+      "epoch": 0.14277869302581,
+      "grad_norm": 0.277567595243454,
+      "learning_rate": 0.00014014598540145983,
+      "loss": 0.6415,
+      "step": 65
+    },
+    {
+      "epoch": 0.14497528830313014,
+      "grad_norm": 0.2811828553676605,
+      "learning_rate": 0.00014233576642335764,
+      "loss": 0.5278,
+      "step": 66
+    },
+    {
+      "epoch": 0.1471718835804503,
+      "grad_norm": 0.4014721214771271,
+      "learning_rate": 0.00014452554744525547,
+      "loss": 0.6033,
+      "step": 67
+    },
+    {
+      "epoch": 0.14936847885777046,
+      "grad_norm": 0.49722668528556824,
+      "learning_rate": 0.00014671532846715328,
+      "loss": 0.6104,
+      "step": 68
+    },
+    {
+      "epoch": 0.1515650741350906,
+      "grad_norm": 0.5049996972084045,
+      "learning_rate": 0.00014890510948905108,
+      "loss": 0.7662,
+      "step": 69
+    },
+    {
+      "epoch": 0.15376166941241076,
+      "grad_norm": 0.42294803261756897,
+      "learning_rate": 0.00015109489051094892,
+      "loss": 0.5599,
+      "step": 70
+    },
+    {
+      "epoch": 0.15595826468973092,
+      "grad_norm": 0.33769503235816956,
+      "learning_rate": 0.00015328467153284672,
+      "loss": 0.7431,
+      "step": 71
+    },
+    {
+      "epoch": 0.15815485996705106,
+      "grad_norm": 0.4182431399822235,
+      "learning_rate": 0.00015547445255474453,
+      "loss": 0.6548,
+      "step": 72
+    },
+    {
+      "epoch": 0.16035145524437122,
+      "grad_norm": 0.5675399899482727,
+      "learning_rate": 0.00015766423357664234,
+      "loss": 0.7614,
+      "step": 73
+    },
+    {
+      "epoch": 0.16254805052169138,
+      "grad_norm": 0.3355293869972229,
+      "learning_rate": 0.00015985401459854014,
+      "loss": 0.7131,
+      "step": 74
+    },
+    {
+      "epoch": 0.16474464579901152,
+      "grad_norm": 0.3530334532260895,
+      "learning_rate": 0.00016204379562043795,
+      "loss": 0.5189,
+      "step": 75
+    },
+    {
+      "epoch": 0.16694124107633168,
+      "grad_norm": 0.4291476905345917,
+      "learning_rate": 0.00016423357664233575,
+      "loss": 0.7305,
+      "step": 76
+    },
+    {
+      "epoch": 0.16913783635365184,
+      "grad_norm": 0.4080694615840912,
+      "learning_rate": 0.00016642335766423356,
+      "loss": 0.6555,
+      "step": 77
+    },
+    {
+      "epoch": 0.171334431630972,
+      "grad_norm": 0.3551289141178131,
+      "learning_rate": 0.00016861313868613137,
+      "loss": 0.4435,
+      "step": 78
+    },
+    {
+      "epoch": 0.17353102690829214,
+      "grad_norm": 0.3615769147872925,
+      "learning_rate": 0.00017080291970802917,
+      "loss": 0.6833,
+      "step": 79
+    },
+    {
+      "epoch": 0.1757276221856123,
+      "grad_norm": 0.41205713152885437,
+      "learning_rate": 0.00017299270072992698,
+      "loss": 0.694,
+      "step": 80
+    },
+    {
+      "epoch": 0.17792421746293247,
+      "grad_norm": 0.3597385287284851,
+      "learning_rate": 0.00017518248175182478,
+      "loss": 0.6104,
+      "step": 81
+    },
+    {
+      "epoch": 0.1801208127402526,
+      "grad_norm": 0.37500330805778503,
+      "learning_rate": 0.00017737226277372262,
+      "loss": 0.5848,
+      "step": 82
+    },
+    {
+      "epoch": 0.18231740801757276,
+      "grad_norm": 0.2954053282737732,
+      "learning_rate": 0.00017956204379562042,
+      "loss": 0.5777,
+      "step": 83
+    },
+    {
+      "epoch": 0.18451400329489293,
+      "grad_norm": 0.6109694838523865,
+      "learning_rate": 0.00018175182481751826,
+      "loss": 0.6718,
+      "step": 84
+    },
+    {
+      "epoch": 0.18671059857221306,
+      "grad_norm": 0.3358462154865265,
+      "learning_rate": 0.00018394160583941606,
+      "loss": 0.7317,
+      "step": 85
+    },
+    {
+      "epoch": 0.18890719384953322,
+      "grad_norm": 0.4268472492694855,
+      "learning_rate": 0.00018613138686131387,
+      "loss": 0.8915,
+      "step": 86
+    },
+    {
+      "epoch": 0.19110378912685339,
+      "grad_norm": 0.2842489778995514,
+      "learning_rate": 0.00018832116788321167,
+      "loss": 0.7149,
+      "step": 87
+    },
+    {
+      "epoch": 0.19330038440417352,
+      "grad_norm": 0.5656724572181702,
+      "learning_rate": 0.00019051094890510948,
+      "loss": 0.7391,
+      "step": 88
+    },
+    {
+      "epoch": 0.19549697968149368,
+      "grad_norm": 0.34295400977134705,
+      "learning_rate": 0.0001927007299270073,
+      "loss": 0.5211,
+      "step": 89
+    },
+    {
+      "epoch": 0.19769357495881384,
+      "grad_norm": 0.5301206707954407,
+      "learning_rate": 0.0001948905109489051,
+      "loss": 0.8407,
+      "step": 90
+    },
+    {
+      "epoch": 0.19989017023613398,
+      "grad_norm": 0.2414446771144867,
+      "learning_rate": 0.0001970802919708029,
+      "loss": 0.6247,
+      "step": 91
+    },
+    {
+      "epoch": 0.20208676551345414,
+      "grad_norm": 0.27806249260902405,
+      "learning_rate": 0.0001992700729927007,
+      "loss": 0.665,
+      "step": 92
+    },
+    {
+      "epoch": 0.2042833607907743,
+      "grad_norm": 0.3400893211364746,
+      "learning_rate": 0.0002014598540145985,
+      "loss": 0.6042,
+      "step": 93
+    },
+    {
+      "epoch": 0.20647995606809447,
+      "grad_norm": 0.4327530562877655,
+      "learning_rate": 0.00020364963503649632,
+      "loss": 0.8196,
+      "step": 94
+    },
+    {
+      "epoch": 0.2086765513454146,
+      "grad_norm": 0.28770408034324646,
+      "learning_rate": 0.00020583941605839415,
+      "loss": 0.6251,
+      "step": 95
+    },
+    {
+      "epoch": 0.21087314662273476,
+      "grad_norm": 0.4040006399154663,
+      "learning_rate": 0.00020802919708029196,
+      "loss": 0.6818,
+      "step": 96
+    },
+    {
+      "epoch": 0.21306974190005493,
+      "grad_norm": 0.35075584053993225,
+      "learning_rate": 0.00021021897810218976,
+      "loss": 0.7605,
+      "step": 97
+    },
+    {
+      "epoch": 0.21526633717737506,
+      "grad_norm": 0.4216291904449463,
+      "learning_rate": 0.00021240875912408757,
+      "loss": 0.6548,
+      "step": 98
+    },
+    {
+      "epoch": 0.21746293245469522,
+      "grad_norm": 0.3403765559196472,
+      "learning_rate": 0.00021459854014598537,
+      "loss": 0.6848,
+      "step": 99
+    },
+    {
+      "epoch": 0.21965952773201539,
+      "grad_norm": 0.4702026844024658,
+      "learning_rate": 0.0002167883211678832,
+      "loss": 0.7066,
+      "step": 100
+    },
+    {
+      "epoch": 0.22185612300933552,
+      "grad_norm": 0.5880163908004761,
+      "learning_rate": 0.00021897810218978101,
+      "loss": 0.7677,
+      "step": 101
+    },
+    {
+      "epoch": 0.22405271828665568,
+      "grad_norm": 0.4032009243965149,
+      "learning_rate": 0.00022116788321167882,
+      "loss": 0.634,
+      "step": 102
+    },
+    {
+      "epoch": 0.22624931356397585,
+      "grad_norm": 0.25544553995132446,
+      "learning_rate": 0.00022335766423357663,
+      "loss": 0.7035,
+      "step": 103
+    },
+    {
+      "epoch": 0.22844590884129598,
+      "grad_norm": 0.31988269090652466,
+      "learning_rate": 0.00022554744525547443,
+      "loss": 0.5396,
+      "step": 104
+    },
+    {
+      "epoch": 0.23064250411861614,
+      "grad_norm": 0.27701693773269653,
+      "learning_rate": 0.00022773722627737224,
+      "loss": 0.7162,
+      "step": 105
+    },
+    {
+      "epoch": 0.2328390993959363,
+      "grad_norm": 0.30583328008651733,
+      "learning_rate": 0.00022992700729927004,
+      "loss": 0.7728,
+      "step": 106
+    },
+    {
+      "epoch": 0.23503569467325644,
+      "grad_norm": 0.4964165687561035,
+      "learning_rate": 0.00023211678832116788,
+      "loss": 0.6303,
+      "step": 107
+    },
+    {
+      "epoch": 0.2372322899505766,
+      "grad_norm": 0.2638928294181824,
+      "learning_rate": 0.00023430656934306568,
+      "loss": 0.5687,
+      "step": 108
+    },
+    {
+      "epoch": 0.23942888522789676,
+      "grad_norm": 0.27152299880981445,
+      "learning_rate": 0.0002364963503649635,
+      "loss": 0.5819,
+      "step": 109
+    },
+    {
+      "epoch": 0.24162548050521693,
+      "grad_norm": 0.256440132856369,
+      "learning_rate": 0.0002386861313868613,
+      "loss": 0.6375,
+      "step": 110
+    },
+    {
+      "epoch": 0.24382207578253706,
+      "grad_norm": 0.3033052384853363,
+      "learning_rate": 0.0002408759124087591,
+      "loss": 0.6269,
+      "step": 111
+    },
+    {
+      "epoch": 0.24601867105985722,
+      "grad_norm": 0.29481041431427,
+      "learning_rate": 0.0002430656934306569,
+      "loss": 0.6471,
+      "step": 112
+    },
+    {
+      "epoch": 0.2482152663371774,
+      "grad_norm": 0.2520420253276825,
+      "learning_rate": 0.0002452554744525547,
+      "loss": 0.6023,
+      "step": 113
+    },
+    {
+      "epoch": 0.2504118616144975,
+      "grad_norm": 0.4020513892173767,
+      "learning_rate": 0.0002474452554744525,
+      "loss": 0.5811,
+      "step": 114
+    },
+    {
+      "epoch": 0.25260845689181766,
+      "grad_norm": 0.4276483952999115,
+      "learning_rate": 0.0002496350364963503,
+      "loss": 0.6746,
+      "step": 115
+    },
+    {
+      "epoch": 0.25480505216913785,
+      "grad_norm": 0.28798580169677734,
+      "learning_rate": 0.00025182481751824813,
+      "loss": 0.775,
+      "step": 116
+    },
+    {
+      "epoch": 0.257001647446458,
+      "grad_norm": 0.4346650242805481,
+      "learning_rate": 0.00025401459854014594,
+      "loss": 0.7344,
+      "step": 117
+    },
+    {
+      "epoch": 0.2591982427237781,
+      "grad_norm": 0.2660764157772064,
+      "learning_rate": 0.00025620437956204374,
+      "loss": 0.668,
+      "step": 118
+    },
+    {
+      "epoch": 0.2613948380010983,
+      "grad_norm": 0.428676962852478,
+      "learning_rate": 0.00025839416058394155,
+      "loss": 0.6283,
+      "step": 119
+    },
+    {
+      "epoch": 0.26359143327841844,
+      "grad_norm": 0.273325115442276,
+      "learning_rate": 0.0002605839416058394,
+      "loss": 0.5178,
+      "step": 120
+    },
+    {
+      "epoch": 0.26578802855573863,
+      "grad_norm": 0.2068067342042923,
+      "learning_rate": 0.0002627737226277372,
+      "loss": 0.5933,
+      "step": 121
+    },
+    {
+      "epoch": 0.26798462383305877,
+      "grad_norm": 0.23313117027282715,
+      "learning_rate": 0.000264963503649635,
+      "loss": 0.5773,
+      "step": 122
+    },
+    {
+      "epoch": 0.2701812191103789,
+      "grad_norm": 0.8602606058120728,
+      "learning_rate": 0.00026715328467153283,
+      "loss": 0.6048,
+      "step": 123
+    },
+    {
+      "epoch": 0.2723778143876991,
+      "grad_norm": 0.32625848054885864,
+      "learning_rate": 0.00026934306569343063,
+      "loss": 0.5598,
+      "step": 124
+    },
+    {
+      "epoch": 0.2745744096650192,
+      "grad_norm": 0.48582178354263306,
+      "learning_rate": 0.00027153284671532844,
+      "loss": 0.7393,
+      "step": 125
+    },
+    {
+      "epoch": 0.27677100494233936,
+      "grad_norm": 0.2123667150735855,
+      "learning_rate": 0.00027372262773722625,
+      "loss": 0.466,
+      "step": 126
+    },
+    {
+      "epoch": 0.27896760021965955,
+      "grad_norm": 0.337178498506546,
+      "learning_rate": 0.00027591240875912405,
+      "loss": 0.6515,
+      "step": 127
+    },
+    {
+      "epoch": 0.2811641954969797,
+      "grad_norm": 0.34165921807289124,
+      "learning_rate": 0.00027810218978102186,
+      "loss": 0.8295,
+      "step": 128
+    },
+    {
+      "epoch": 0.2833607907742998,
+      "grad_norm": 0.29111403226852417,
+      "learning_rate": 0.00028029197080291966,
+      "loss": 0.7319,
+      "step": 129
+    },
+    {
+      "epoch": 0.28555738605162,
+      "grad_norm": 0.3523664176464081,
+      "learning_rate": 0.00028248175182481747,
+      "loss": 0.5731,
+      "step": 130
+    },
+    {
+      "epoch": 0.28775398132894014,
+      "grad_norm": 0.2988019585609436,
+      "learning_rate": 0.0002846715328467153,
+      "loss": 0.6765,
+      "step": 131
+    },
+    {
+      "epoch": 0.2899505766062603,
+      "grad_norm": 0.23230232298374176,
+      "learning_rate": 0.00028686131386861314,
+      "loss": 0.5574,
+      "step": 132
+    },
+    {
+      "epoch": 0.29214717188358047,
+      "grad_norm": 0.3545559346675873,
+      "learning_rate": 0.00028905109489051094,
+      "loss": 0.581,
+      "step": 133
+    },
+    {
+      "epoch": 0.2943437671609006,
+      "grad_norm": 0.32541900873184204,
+      "learning_rate": 0.00029124087591240875,
+      "loss": 0.604,
+      "step": 134
+    },
+    {
+      "epoch": 0.29654036243822074,
+      "grad_norm": 0.20449653267860413,
+      "learning_rate": 0.00029343065693430656,
+      "loss": 0.6732,
+      "step": 135
+    },
+    {
+      "epoch": 0.29873695771554093,
+      "grad_norm": 0.44041532278060913,
+      "learning_rate": 0.00029562043795620436,
+      "loss": 0.7346,
+      "step": 136
+    },
+    {
+      "epoch": 0.30093355299286106,
+      "grad_norm": 0.30077028274536133,
+      "learning_rate": 0.00029781021897810217,
+      "loss": 0.5806,
+      "step": 137
+    },
+    {
+      "epoch": 0.3031301482701812,
+      "grad_norm": 0.22556346654891968,
+      "learning_rate": 0.0003,
+      "loss": 0.6678,
+      "step": 138
+    },
+    {
+      "epoch": 0.3053267435475014,
+      "grad_norm": 0.2511537969112396,
+      "learning_rate": 0.00029975570032573286,
+      "loss": 0.5792,
+      "step": 139
+    },
+    {
+      "epoch": 0.3075233388248215,
+      "grad_norm": 0.2401726394891739,
+      "learning_rate": 0.00029951140065146575,
+      "loss": 0.7172,
+      "step": 140
+    },
+    {
+      "epoch": 0.30971993410214166,
+      "grad_norm": 0.2528117597103119,
+      "learning_rate": 0.0002992671009771987,
+      "loss": 0.6255,
+      "step": 141
+    },
+    {
+      "epoch": 0.31191652937946185,
+      "grad_norm": 0.2821625769138336,
+      "learning_rate": 0.0002990228013029316,
+      "loss": 0.7031,
+      "step": 142
+    },
+    {
+      "epoch": 0.314113124656782,
+      "grad_norm": 0.34014376997947693,
+      "learning_rate": 0.0002987785016286645,
+      "loss": 0.6987,
+      "step": 143
+    },
+    {
+      "epoch": 0.3163097199341021,
+      "grad_norm": 0.37961485981941223,
+      "learning_rate": 0.00029853420195439736,
+      "loss": 0.7961,
+      "step": 144
+    },
+    {
+      "epoch": 0.3185063152114223,
+      "grad_norm": 0.28274771571159363,
+      "learning_rate": 0.00029828990228013025,
+      "loss": 0.6544,
+      "step": 145
+    },
+    {
+      "epoch": 0.32070291048874244,
+      "grad_norm": 0.2571098208427429,
+      "learning_rate": 0.00029804560260586314,
+      "loss": 0.7047,
+      "step": 146
+    },
+    {
+      "epoch": 0.3228995057660626,
+      "grad_norm": 0.30160510540008545,
+      "learning_rate": 0.0002978013029315961,
+      "loss": 0.5117,
+      "step": 147
+    },
+    {
+      "epoch": 0.32509610104338277,
+      "grad_norm": 0.19304673373699188,
+      "learning_rate": 0.000297557003257329,
+      "loss": 0.5729,
+      "step": 148
+    },
+    {
+      "epoch": 0.3272926963207029,
+      "grad_norm": 0.2948278486728668,
+      "learning_rate": 0.00029731270358306186,
+      "loss": 0.6269,
+      "step": 149
+    },
+    {
+      "epoch": 0.32948929159802304,
+      "grad_norm": 0.2918180823326111,
+      "learning_rate": 0.00029706840390879475,
+      "loss": 0.8671,
+      "step": 150
+    },
+    {
+      "epoch": 0.3316858868753432,
+      "grad_norm": 0.22447623312473297,
+      "learning_rate": 0.00029682410423452764,
+      "loss": 0.7231,
+      "step": 151
+    },
+    {
+      "epoch": 0.33388248215266336,
+      "grad_norm": 0.2767757773399353,
+      "learning_rate": 0.0002965798045602606,
+      "loss": 0.7579,
+      "step": 152
+    },
+    {
+      "epoch": 0.33607907742998355,
+      "grad_norm": 0.2532203197479248,
+      "learning_rate": 0.0002963355048859935,
+      "loss": 0.557,
+      "step": 153
+    },
+    {
+      "epoch": 0.3382756727073037,
+      "grad_norm": 0.3206186294555664,
+      "learning_rate": 0.00029609120521172636,
+      "loss": 0.629,
+      "step": 154
+    },
+    {
+      "epoch": 0.3404722679846238,
+      "grad_norm": 0.26047345995903015,
+      "learning_rate": 0.00029584690553745925,
+      "loss": 0.5892,
+      "step": 155
+    },
+    {
+      "epoch": 0.342668863261944,
+      "grad_norm": 0.20632953941822052,
+      "learning_rate": 0.00029560260586319214,
+      "loss": 0.6273,
+      "step": 156
+    },
+    {
+      "epoch": 0.34486545853926415,
+      "grad_norm": 0.6605954766273499,
+      "learning_rate": 0.00029535830618892503,
+      "loss": 0.5583,
+      "step": 157
+    },
+    {
+      "epoch": 0.3470620538165843,
+      "grad_norm": 0.24968554079532623,
+      "learning_rate": 0.000295114006514658,
+      "loss": 0.6315,
+      "step": 158
+    },
+    {
+      "epoch": 0.34925864909390447,
+      "grad_norm": 0.2880593240261078,
+      "learning_rate": 0.00029486970684039086,
+      "loss": 0.7649,
+      "step": 159
+    },
+    {
+      "epoch": 0.3514552443712246,
+      "grad_norm": 0.33606216311454773,
+      "learning_rate": 0.00029462540716612375,
+      "loss": 0.6606,
+      "step": 160
+    },
+    {
+      "epoch": 0.35365183964854474,
+      "grad_norm": 0.3336750864982605,
+      "learning_rate": 0.00029438110749185664,
+      "loss": 0.6203,
+      "step": 161
+    },
+    {
+      "epoch": 0.35584843492586493,
+      "grad_norm": 0.22258640825748444,
+      "learning_rate": 0.00029413680781758953,
+      "loss": 0.5365,
+      "step": 162
+    },
+    {
+      "epoch": 0.35804503020318507,
+      "grad_norm": 0.22713644802570343,
+      "learning_rate": 0.0002938925081433224,
+      "loss": 0.425,
+      "step": 163
+    },
+    {
+      "epoch": 0.3602416254805052,
+      "grad_norm": 0.29401063919067383,
+      "learning_rate": 0.00029364820846905536,
+      "loss": 0.5546,
+      "step": 164
+    },
+    {
+      "epoch": 0.3624382207578254,
+      "grad_norm": 0.24572621285915375,
+      "learning_rate": 0.00029340390879478825,
+      "loss": 0.7232,
+      "step": 165
+    },
+    {
+      "epoch": 0.3646348160351455,
+      "grad_norm": 0.29458099603652954,
+      "learning_rate": 0.00029315960912052114,
+      "loss": 0.6651,
+      "step": 166
+    },
+    {
+      "epoch": 0.36683141131246566,
+      "grad_norm": 0.21029439568519592,
+      "learning_rate": 0.00029291530944625403,
+      "loss": 0.6038,
+      "step": 167
+    },
+    {
+      "epoch": 0.36902800658978585,
+      "grad_norm": 0.3289354145526886,
+      "learning_rate": 0.0002926710097719869,
+      "loss": 0.6823,
+      "step": 168
+    },
+    {
+      "epoch": 0.371224601867106,
+      "grad_norm": 0.23331189155578613,
+      "learning_rate": 0.00029242671009771986,
+      "loss": 0.7014,
+      "step": 169
+    },
+    {
+      "epoch": 0.3734211971444261,
+      "grad_norm": 0.3129960596561432,
+      "learning_rate": 0.00029218241042345275,
+      "loss": 0.7512,
+      "step": 170
+    },
+    {
+      "epoch": 0.3756177924217463,
+      "grad_norm": 0.484561026096344,
+      "learning_rate": 0.00029193811074918564,
+      "loss": 0.5811,
+      "step": 171
+    },
+    {
+      "epoch": 0.37781438769906645,
+      "grad_norm": 0.23281951248645782,
+      "learning_rate": 0.00029169381107491853,
+      "loss": 0.5217,
+      "step": 172
+    },
+    {
+      "epoch": 0.3800109829763866,
+      "grad_norm": 0.23013974726200104,
+      "learning_rate": 0.0002914495114006514,
+      "loss": 0.6613,
+      "step": 173
+    },
+    {
+      "epoch": 0.38220757825370677,
+      "grad_norm": 0.27919653058052063,
+      "learning_rate": 0.0002912052117263843,
+      "loss": 0.8525,
+      "step": 174
+    },
+    {
+      "epoch": 0.3844041735310269,
+      "grad_norm": 0.24366801977157593,
+      "learning_rate": 0.00029096091205211725,
+      "loss": 0.5344,
+      "step": 175
+    },
+    {
+      "epoch": 0.38660076880834704,
+      "grad_norm": 0.23048312962055206,
+      "learning_rate": 0.00029071661237785014,
+      "loss": 0.769,
+      "step": 176
+    },
+    {
+      "epoch": 0.38879736408566723,
+      "grad_norm": 0.21273106336593628,
+      "learning_rate": 0.00029047231270358303,
+      "loss": 0.4036,
+      "step": 177
+    },
+    {
+      "epoch": 0.39099395936298736,
+      "grad_norm": 0.30006176233291626,
+      "learning_rate": 0.0002902280130293159,
+      "loss": 0.6675,
+      "step": 178
+    },
+    {
+      "epoch": 0.3931905546403075,
+      "grad_norm": 0.2932669222354889,
+      "learning_rate": 0.0002899837133550488,
+      "loss": 0.6451,
+      "step": 179
+    },
+    {
+      "epoch": 0.3953871499176277,
+      "grad_norm": 0.20935624837875366,
+      "learning_rate": 0.00028973941368078175,
+      "loss": 0.544,
+      "step": 180
+    },
+    {
+      "epoch": 0.3975837451949478,
+      "grad_norm": 0.25854918360710144,
+      "learning_rate": 0.00028949511400651464,
+      "loss": 0.4777,
+      "step": 181
+    },
+    {
+      "epoch": 0.39978034047226796,
+      "grad_norm": 0.20923751592636108,
+      "learning_rate": 0.00028925081433224753,
+      "loss": 0.5232,
+      "step": 182
+    },
+    {
+      "epoch": 0.40197693574958815,
+      "grad_norm": 0.3148953318595886,
+      "learning_rate": 0.0002890065146579804,
+      "loss": 0.7018,
+      "step": 183
+    },
+    {
+      "epoch": 0.4041735310269083,
+      "grad_norm": 0.24374817311763763,
+      "learning_rate": 0.0002887622149837133,
+      "loss": 0.9343,
+      "step": 184
+    },
+    {
+      "epoch": 0.4063701263042284,
+      "grad_norm": 0.21448668837547302,
+      "learning_rate": 0.0002885179153094462,
+      "loss": 0.672,
+      "step": 185
+    },
+    {
+      "epoch": 0.4085667215815486,
+      "grad_norm": 0.20949478447437286,
+      "learning_rate": 0.00028827361563517914,
+      "loss": 0.6461,
+      "step": 186
+    },
+    {
+      "epoch": 0.41076331685886874,
+      "grad_norm": 0.29779893159866333,
+      "learning_rate": 0.00028802931596091203,
+      "loss": 0.4763,
+      "step": 187
+    },
+    {
+      "epoch": 0.41295991213618893,
+      "grad_norm": 0.29939642548561096,
+      "learning_rate": 0.0002877850162866449,
+      "loss": 0.6827,
+      "step": 188
+    },
+    {
+      "epoch": 0.41515650741350907,
+      "grad_norm": 0.41549426317214966,
+      "learning_rate": 0.0002875407166123778,
+      "loss": 0.7007,
+      "step": 189
+    },
+    {
+      "epoch": 0.4173531026908292,
+      "grad_norm": 0.24645191431045532,
+      "learning_rate": 0.0002872964169381107,
+      "loss": 0.6252,
+      "step": 190
+    },
+    {
+      "epoch": 0.4195496979681494,
+      "grad_norm": 0.42897850275039673,
+      "learning_rate": 0.00028705211726384364,
+      "loss": 0.5542,
+      "step": 191
+    },
+    {
+      "epoch": 0.42174629324546953,
+      "grad_norm": 0.23049034178256989,
+      "learning_rate": 0.00028680781758957653,
+      "loss": 0.7258,
+      "step": 192
+    },
+    {
+      "epoch": 0.42394288852278966,
+      "grad_norm": 0.31065866351127625,
+      "learning_rate": 0.0002865635179153094,
+      "loss": 0.7053,
+      "step": 193
+    },
+    {
+      "epoch": 0.42613948380010985,
+      "grad_norm": 0.23515336215496063,
+      "learning_rate": 0.0002863192182410423,
+      "loss": 0.6941,
+      "step": 194
+    },
+    {
+      "epoch": 0.42833607907743,
+      "grad_norm": 0.26374128460884094,
+      "learning_rate": 0.0002860749185667752,
+      "loss": 0.8289,
+      "step": 195
+    },
+    {
+      "epoch": 0.4305326743547501,
+      "grad_norm": 0.24964147806167603,
+      "learning_rate": 0.0002858306188925081,
+      "loss": 0.9764,
+      "step": 196
+    },
+    {
+      "epoch": 0.4327292696320703,
+      "grad_norm": 0.22735856473445892,
+      "learning_rate": 0.00028558631921824103,
+      "loss": 0.5817,
+      "step": 197
+    },
+    {
+      "epoch": 0.43492586490939045,
+      "grad_norm": 0.2793346047401428,
+      "learning_rate": 0.0002853420195439739,
+      "loss": 0.5422,
+      "step": 198
+    },
+    {
+      "epoch": 0.4371224601867106,
+      "grad_norm": 0.2854657769203186,
+      "learning_rate": 0.0002850977198697068,
+      "loss": 0.6223,
+      "step": 199
+    },
+    {
+      "epoch": 0.43931905546403077,
+      "grad_norm": 0.21625939011573792,
+      "learning_rate": 0.0002848534201954397,
+      "loss": 0.6949,
+      "step": 200
+    },
+    {
+      "epoch": 0.4415156507413509,
+      "grad_norm": 0.25202566385269165,
+      "learning_rate": 0.0002846091205211726,
+      "loss": 0.5635,
+      "step": 201
+    },
+    {
+      "epoch": 0.44371224601867104,
+      "grad_norm": 0.18184080719947815,
+      "learning_rate": 0.0002843648208469055,
+      "loss": 0.5972,
+      "step": 202
+    },
+    {
+      "epoch": 0.44590884129599123,
+      "grad_norm": 0.258818656206131,
+      "learning_rate": 0.0002841205211726384,
+      "loss": 0.6396,
+      "step": 203
+    },
+    {
+      "epoch": 0.44810543657331137,
+      "grad_norm": 0.36491551995277405,
+      "learning_rate": 0.0002838762214983713,
+      "loss": 0.6435,
+      "step": 204
+    },
+    {
+      "epoch": 0.4503020318506315,
+      "grad_norm": 0.395247220993042,
+      "learning_rate": 0.0002836319218241042,
+      "loss": 0.7867,
+      "step": 205
+    },
+    {
+      "epoch": 0.4524986271279517,
+      "grad_norm": 0.3144085705280304,
+      "learning_rate": 0.0002833876221498371,
+      "loss": 0.788,
+      "step": 206
+    },
+    {
+      "epoch": 0.4546952224052718,
+      "grad_norm": 0.23340974748134613,
+      "learning_rate": 0.00028314332247557,
+      "loss": 0.6658,
+      "step": 207
+    },
+    {
+      "epoch": 0.45689181768259196,
+      "grad_norm": 0.18852849304676056,
+      "learning_rate": 0.0002828990228013029,
+      "loss": 0.5672,
+      "step": 208
+    },
+    {
+      "epoch": 0.45908841295991215,
+      "grad_norm": 0.9333329796791077,
+      "learning_rate": 0.0002826547231270358,
+      "loss": 0.6435,
+      "step": 209
+    },
+    {
+      "epoch": 0.4612850082372323,
+      "grad_norm": 0.3410152196884155,
+      "learning_rate": 0.0002824104234527687,
+      "loss": 0.7752,
+      "step": 210
+    },
+    {
+      "epoch": 0.4634816035145524,
+      "grad_norm": 0.21588265895843506,
+      "learning_rate": 0.0002821661237785016,
+      "loss": 0.7607,
+      "step": 211
+    },
+    {
+      "epoch": 0.4656781987918726,
+      "grad_norm": 0.4102458953857422,
+      "learning_rate": 0.0002819218241042345,
+      "loss": 0.701,
+      "step": 212
+    },
+    {
+      "epoch": 0.46787479406919275,
+      "grad_norm": 0.26939937472343445,
+      "learning_rate": 0.00028167752442996737,
+      "loss": 0.4709,
+      "step": 213
+    },
+    {
+      "epoch": 0.4700713893465129,
+      "grad_norm": 0.17174175381660461,
+      "learning_rate": 0.0002814332247557003,
+      "loss": 0.6141,
+      "step": 214
+    },
+    {
+      "epoch": 0.47226798462383307,
+      "grad_norm": 0.23832230269908905,
+      "learning_rate": 0.0002811889250814332,
+      "loss": 0.6997,
+      "step": 215
+    },
+    {
+      "epoch": 0.4744645799011532,
+      "grad_norm": 0.2893271744251251,
+      "learning_rate": 0.0002809446254071661,
+      "loss": 0.6828,
+      "step": 216
+    },
+    {
+      "epoch": 0.47666117517847334,
+      "grad_norm": 0.23587121069431305,
+      "learning_rate": 0.000280700325732899,
+      "loss": 0.5511,
+      "step": 217
+    },
+    {
+      "epoch": 0.47885777045579353,
+      "grad_norm": 0.21264143288135529,
+      "learning_rate": 0.00028045602605863187,
+      "loss": 0.7433,
+      "step": 218
+    },
+    {
+      "epoch": 0.48105436573311366,
+      "grad_norm": 0.269033282995224,
+      "learning_rate": 0.0002802117263843648,
+      "loss": 0.6118,
+      "step": 219
+    },
+    {
+      "epoch": 0.48325096101043385,
+      "grad_norm": 0.221666619181633,
+      "learning_rate": 0.0002799674267100977,
+      "loss": 0.717,
+      "step": 220
+    },
+    {
+      "epoch": 0.485447556287754,
+      "grad_norm": 0.3358190655708313,
+      "learning_rate": 0.0002797231270358306,
+      "loss": 0.6192,
+      "step": 221
+    },
+    {
+      "epoch": 0.4876441515650741,
+      "grad_norm": 0.28939563035964966,
+      "learning_rate": 0.0002794788273615635,
+      "loss": 0.5246,
+      "step": 222
+    },
+    {
+      "epoch": 0.4898407468423943,
+      "grad_norm": 0.26786231994628906,
+      "learning_rate": 0.00027923452768729637,
+      "loss": 0.7101,
+      "step": 223
+    },
+    {
+      "epoch": 0.49203734211971445,
+      "grad_norm": 0.2974752187728882,
+      "learning_rate": 0.00027899022801302926,
+      "loss": 0.7153,
+      "step": 224
+    },
+    {
+      "epoch": 0.4942339373970346,
+      "grad_norm": 0.21427251398563385,
+      "learning_rate": 0.0002787459283387622,
+      "loss": 0.4682,
+      "step": 225
+    },
+    {
+      "epoch": 0.4964305326743548,
+      "grad_norm": 0.22402474284172058,
+      "learning_rate": 0.0002785016286644951,
+      "loss": 0.9029,
+      "step": 226
+    },
+    {
+      "epoch": 0.4986271279516749,
+      "grad_norm": 0.25051653385162354,
+      "learning_rate": 0.000278257328990228,
+      "loss": 0.6428,
+      "step": 227
+    },
+    {
+      "epoch": 0.500823723228995,
+      "grad_norm": 0.23938477039337158,
+      "learning_rate": 0.00027801302931596087,
+      "loss": 0.6227,
+      "step": 228
+    },
+    {
+      "epoch": 0.5030203185063152,
+      "grad_norm": 0.2304198145866394,
+      "learning_rate": 0.00027776872964169376,
+      "loss": 0.6632,
+      "step": 229
+    },
+    {
+      "epoch": 0.5052169137836353,
+      "grad_norm": 0.3043380677700043,
+      "learning_rate": 0.0002775244299674267,
+      "loss": 0.7494,
+      "step": 230
+    },
+    {
+      "epoch": 0.5074135090609555,
+      "grad_norm": 0.2606238126754761,
+      "learning_rate": 0.0002772801302931596,
+      "loss": 0.708,
+      "step": 231
+    },
+    {
+      "epoch": 0.5096101043382757,
+      "grad_norm": 0.2550143301486969,
+      "learning_rate": 0.0002770358306188925,
+      "loss": 0.5945,
+      "step": 232
+    },
+    {
+      "epoch": 0.5118066996155958,
+      "grad_norm": 0.2605602741241455,
+      "learning_rate": 0.00027679153094462537,
+      "loss": 0.7173,
+      "step": 233
+    },
+    {
+      "epoch": 0.514003294892916,
+      "grad_norm": 0.2413637787103653,
+      "learning_rate": 0.00027654723127035826,
+      "loss": 0.6554,
+      "step": 234
+    },
+    {
+      "epoch": 0.5161998901702362,
+      "grad_norm": 0.27595633268356323,
+      "learning_rate": 0.00027630293159609115,
+      "loss": 0.665,
+      "step": 235
+    },
+    {
+      "epoch": 0.5183964854475562,
+      "grad_norm": 0.16411711275577545,
+      "learning_rate": 0.0002760586319218241,
+      "loss": 0.581,
+      "step": 236
+    },
+    {
+      "epoch": 0.5205930807248764,
+      "grad_norm": 0.40575161576271057,
+      "learning_rate": 0.000275814332247557,
+      "loss": 0.8982,
+      "step": 237
+    },
+    {
+      "epoch": 0.5227896760021966,
+      "grad_norm": 0.2569533586502075,
+      "learning_rate": 0.00027557003257328987,
+      "loss": 0.7441,
+      "step": 238
+    },
+    {
+      "epoch": 0.5249862712795168,
+      "grad_norm": 0.2130047231912613,
+      "learning_rate": 0.00027532573289902276,
+      "loss": 0.6245,
+      "step": 239
+    },
+    {
+      "epoch": 0.5271828665568369,
+      "grad_norm": 0.25668343901634216,
+      "learning_rate": 0.00027508143322475565,
+      "loss": 0.6829,
+      "step": 240
+    },
+    {
+      "epoch": 0.5293794618341571,
+      "grad_norm": 0.2509118914604187,
+      "learning_rate": 0.00027483713355048854,
+      "loss": 0.6852,
+      "step": 241
+    },
+    {
+      "epoch": 0.5315760571114773,
+      "grad_norm": 0.23890118300914764,
+      "learning_rate": 0.0002745928338762215,
+      "loss": 0.6074,
+      "step": 242
+    },
+    {
+      "epoch": 0.5337726523887973,
+      "grad_norm": 0.2057567983865738,
+      "learning_rate": 0.00027434853420195437,
+      "loss": 0.5886,
+      "step": 243
+    },
+    {
+      "epoch": 0.5359692476661175,
+      "grad_norm": 0.2694491744041443,
+      "learning_rate": 0.00027410423452768726,
+      "loss": 0.7477,
+      "step": 244
+    },
+    {
+      "epoch": 0.5381658429434377,
+      "grad_norm": 0.25715163350105286,
+      "learning_rate": 0.00027385993485342015,
+      "loss": 0.6526,
+      "step": 245
+    },
+    {
+      "epoch": 0.5403624382207578,
+      "grad_norm": 0.3362589478492737,
+      "learning_rate": 0.00027361563517915304,
+      "loss": 0.6669,
+      "step": 246
+    },
+    {
+      "epoch": 0.542559033498078,
+      "grad_norm": 0.2794806659221649,
+      "learning_rate": 0.000273371335504886,
+      "loss": 0.5734,
+      "step": 247
+    },
+    {
+      "epoch": 0.5447556287753982,
+      "grad_norm": 0.4383026957511902,
+      "learning_rate": 0.00027312703583061887,
+      "loss": 0.7065,
+      "step": 248
+    },
+    {
+      "epoch": 0.5469522240527183,
+      "grad_norm": 0.22355012595653534,
+      "learning_rate": 0.00027288273615635176,
+      "loss": 0.5965,
+      "step": 249
+    },
+    {
+      "epoch": 0.5491488193300385,
+      "grad_norm": 0.2873373329639435,
+      "learning_rate": 0.00027263843648208465,
+      "loss": 0.6678,
+      "step": 250
+    },
+    {
+      "epoch": 0.5513454146073586,
+      "grad_norm": 0.3472447395324707,
+      "learning_rate": 0.00027239413680781754,
+      "loss": 0.6765,
+      "step": 251
+    },
+    {
+      "epoch": 0.5535420098846787,
+      "grad_norm": 0.2879409193992615,
+      "learning_rate": 0.0002721498371335504,
+      "loss": 0.658,
+      "step": 252
+    },
+    {
+      "epoch": 0.5557386051619989,
+      "grad_norm": 0.21323907375335693,
+      "learning_rate": 0.00027190553745928337,
+      "loss": 0.7701,
+      "step": 253
+    },
+    {
+      "epoch": 0.5579352004393191,
+      "grad_norm": 0.2700759470462799,
+      "learning_rate": 0.00027166123778501626,
+      "loss": 0.6522,
+      "step": 254
+    },
+    {
+      "epoch": 0.5601317957166392,
+      "grad_norm": 0.28444936871528625,
+      "learning_rate": 0.00027141693811074915,
+      "loss": 0.6803,
+      "step": 255
+    },
+    {
+      "epoch": 0.5623283909939594,
+      "grad_norm": 0.4065149426460266,
+      "learning_rate": 0.00027117263843648204,
+      "loss": 0.644,
+      "step": 256
+    },
+    {
+      "epoch": 0.5645249862712796,
+      "grad_norm": 0.28332868218421936,
+      "learning_rate": 0.0002709283387622149,
+      "loss": 0.5671,
+      "step": 257
+    },
+    {
+      "epoch": 0.5667215815485996,
+      "grad_norm": 0.2807782292366028,
+      "learning_rate": 0.00027068403908794787,
+      "loss": 0.7866,
+      "step": 258
+    },
+    {
+      "epoch": 0.5689181768259198,
+      "grad_norm": 0.26401036977767944,
+      "learning_rate": 0.00027043973941368076,
+      "loss": 0.5871,
+      "step": 259
+    },
+    {
+      "epoch": 0.57111477210324,
+      "grad_norm": 0.24250145256519318,
+      "learning_rate": 0.00027019543973941365,
+      "loss": 0.6628,
+      "step": 260
+    },
+    {
+      "epoch": 0.5733113673805601,
+      "grad_norm": 0.23073849081993103,
+      "learning_rate": 0.00026995114006514654,
+      "loss": 0.7092,
+      "step": 261
+    },
+    {
+      "epoch": 0.5755079626578803,
+      "grad_norm": 0.1891871988773346,
+      "learning_rate": 0.00026970684039087943,
+      "loss": 0.6749,
+      "step": 262
+    },
+    {
+      "epoch": 0.5777045579352005,
+      "grad_norm": 0.3806025981903076,
+      "learning_rate": 0.0002694625407166123,
+      "loss": 0.5696,
+      "step": 263
+    },
+    {
+      "epoch": 0.5799011532125206,
+      "grad_norm": 0.3086026608943939,
+      "learning_rate": 0.00026921824104234526,
+      "loss": 0.6546,
+      "step": 264
+    },
+    {
+      "epoch": 0.5820977484898407,
+      "grad_norm": 0.24817922711372375,
+      "learning_rate": 0.00026897394136807815,
+      "loss": 0.9384,
+      "step": 265
+    },
+    {
+      "epoch": 0.5842943437671609,
+      "grad_norm": 0.2545439302921295,
+      "learning_rate": 0.00026872964169381104,
+      "loss": 0.7617,
+      "step": 266
+    },
+    {
+      "epoch": 0.586490939044481,
+      "grad_norm": 0.19701135158538818,
+      "learning_rate": 0.00026848534201954393,
+      "loss": 0.6791,
+      "step": 267
+    },
+    {
+      "epoch": 0.5886875343218012,
+      "grad_norm": 0.25485432147979736,
+      "learning_rate": 0.0002682410423452768,
+      "loss": 0.7095,
+      "step": 268
+    },
+    {
+      "epoch": 0.5908841295991214,
+      "grad_norm": 0.40982499718666077,
+      "learning_rate": 0.00026799674267100976,
+      "loss": 0.7928,
+      "step": 269
+    },
+    {
+      "epoch": 0.5930807248764415,
+      "grad_norm": 0.299757182598114,
+      "learning_rate": 0.00026775244299674265,
+      "loss": 0.7094,
+      "step": 270
+    },
+    {
+      "epoch": 0.5952773201537617,
+      "grad_norm": 0.2621508240699768,
+      "learning_rate": 0.00026750814332247554,
+      "loss": 0.6057,
+      "step": 271
+    },
+    {
+      "epoch": 0.5974739154310819,
+      "grad_norm": 0.2649948000907898,
+      "learning_rate": 0.00026726384364820843,
+      "loss": 0.663,
+      "step": 272
+    },
+    {
+      "epoch": 0.5996705107084019,
+      "grad_norm": 0.32136768102645874,
+      "learning_rate": 0.0002670195439739413,
+      "loss": 0.5737,
+      "step": 273
+    },
+    {
+      "epoch": 0.6018671059857221,
+      "grad_norm": 0.3689887821674347,
+      "learning_rate": 0.0002667752442996742,
+      "loss": 0.6521,
+      "step": 274
+    },
+    {
+      "epoch": 0.6040637012630423,
+      "grad_norm": 0.5081011056900024,
+      "learning_rate": 0.00026653094462540715,
+      "loss": 0.6792,
+      "step": 275
+    },
+    {
+      "epoch": 0.6062602965403624,
+      "grad_norm": 0.23460817337036133,
+      "learning_rate": 0.00026628664495114004,
+      "loss": 0.4988,
+      "step": 276
+    },
+    {
+      "epoch": 0.6084568918176826,
+      "grad_norm": 0.20828036963939667,
+      "learning_rate": 0.00026604234527687293,
+      "loss": 0.6365,
+      "step": 277
+    },
+    {
+      "epoch": 0.6106534870950028,
+      "grad_norm": 0.33490219712257385,
+      "learning_rate": 0.0002657980456026058,
+      "loss": 0.7975,
+      "step": 278
+    },
+    {
+      "epoch": 0.6128500823723229,
+      "grad_norm": 0.34623801708221436,
+      "learning_rate": 0.0002655537459283387,
+      "loss": 0.6433,
+      "step": 279
+    },
+    {
+      "epoch": 0.615046677649643,
+      "grad_norm": 0.25095421075820923,
+      "learning_rate": 0.00026530944625407165,
+      "loss": 0.614,
+      "step": 280
+    },
+    {
+      "epoch": 0.6172432729269632,
+      "grad_norm": 0.3376641273498535,
+      "learning_rate": 0.00026506514657980454,
+      "loss": 0.5448,
+      "step": 281
+    },
+    {
+      "epoch": 0.6194398682042833,
+      "grad_norm": 0.21586130559444427,
+      "learning_rate": 0.00026482084690553743,
+      "loss": 0.5908,
+      "step": 282
+    },
+    {
+      "epoch": 0.6216364634816035,
+      "grad_norm": 0.29025331139564514,
+      "learning_rate": 0.0002645765472312703,
+      "loss": 0.7326,
+      "step": 283
+    },
+    {
+      "epoch": 0.6238330587589237,
+      "grad_norm": 0.23376524448394775,
+      "learning_rate": 0.0002643322475570032,
+      "loss": 0.5887,
+      "step": 284
+    },
+    {
+      "epoch": 0.6260296540362438,
+      "grad_norm": 0.22396379709243774,
+      "learning_rate": 0.0002640879478827361,
+      "loss": 0.6082,
+      "step": 285
+    },
+    {
+      "epoch": 0.628226249313564,
+      "grad_norm": 0.24109943211078644,
+      "learning_rate": 0.00026384364820846904,
+      "loss": 0.6059,
+      "step": 286
+    },
+    {
+      "epoch": 0.6304228445908842,
+      "grad_norm": 0.21724393963813782,
+      "learning_rate": 0.00026359934853420193,
+      "loss": 0.6981,
+      "step": 287
+    },
+    {
+      "epoch": 0.6326194398682042,
+      "grad_norm": 0.22276242077350616,
+      "learning_rate": 0.0002633550488599348,
+      "loss": 0.6156,
+      "step": 288
+    },
+    {
+      "epoch": 0.6348160351455244,
+      "grad_norm": 0.2747544050216675,
+      "learning_rate": 0.0002631107491856677,
+      "loss": 0.5327,
+      "step": 289
+    },
+    {
+      "epoch": 0.6370126304228446,
+      "grad_norm": 0.29766714572906494,
+      "learning_rate": 0.0002628664495114006,
+      "loss": 0.5632,
+      "step": 290
+    },
+    {
+      "epoch": 0.6392092257001647,
+      "grad_norm": 0.5123509764671326,
+      "learning_rate": 0.0002626221498371335,
+      "loss": 0.7338,
+      "step": 291
+    },
+    {
+      "epoch": 0.6414058209774849,
+      "grad_norm": 0.4441981911659241,
+      "learning_rate": 0.00026237785016286643,
+      "loss": 0.7253,
+      "step": 292
+    },
+    {
+      "epoch": 0.6436024162548051,
+      "grad_norm": 0.33704692125320435,
+      "learning_rate": 0.0002621335504885993,
+      "loss": 0.7562,
+      "step": 293
+    },
+    {
+      "epoch": 0.6457990115321252,
+      "grad_norm": 0.3116348087787628,
+      "learning_rate": 0.00026188925081433226,
+      "loss": 0.7313,
+      "step": 294
+    },
+    {
+      "epoch": 0.6479956068094453,
+      "grad_norm": 0.3933788537979126,
+      "learning_rate": 0.00026164495114006515,
+      "loss": 0.6774,
+      "step": 295
+    },
+    {
+      "epoch": 0.6501922020867655,
+      "grad_norm": 0.25686022639274597,
+      "learning_rate": 0.00026140065146579804,
+      "loss": 0.6521,
+      "step": 296
+    },
+    {
+      "epoch": 0.6523887973640856,
+      "grad_norm": 0.2514008581638336,
+      "learning_rate": 0.00026115635179153093,
+      "loss": 0.5821,
+      "step": 297
+    },
+    {
+      "epoch": 0.6545853926414058,
+      "grad_norm": 0.4609866142272949,
+      "learning_rate": 0.0002609120521172638,
+      "loss": 0.5969,
+      "step": 298
+    },
+    {
+      "epoch": 0.656781987918726,
+      "grad_norm": 0.23484556376934052,
+      "learning_rate": 0.00026066775244299676,
+      "loss": 0.7001,
+      "step": 299
+    },
+    {
+      "epoch": 0.6589785831960461,
+      "grad_norm": 0.30440106987953186,
+      "learning_rate": 0.00026042345276872965,
+      "loss": 0.6245,
+      "step": 300
+    },
+    {
+      "epoch": 0.6611751784733663,
+      "grad_norm": 0.3201594948768616,
+      "learning_rate": 0.00026017915309446254,
+      "loss": 0.7195,
+      "step": 301
+    },
+    {
+      "epoch": 0.6633717737506865,
+      "grad_norm": 0.303623229265213,
+      "learning_rate": 0.00025993485342019543,
+      "loss": 0.6169,
+      "step": 302
+    },
+    {
+      "epoch": 0.6655683690280065,
+      "grad_norm": 0.20629887282848358,
+      "learning_rate": 0.0002596905537459283,
+      "loss": 0.5615,
+      "step": 303
+    },
+    {
+      "epoch": 0.6677649643053267,
+      "grad_norm": 0.36723387241363525,
+      "learning_rate": 0.0002594462540716612,
+      "loss": 0.7426,
+      "step": 304
+    },
+    {
+      "epoch": 0.6699615595826469,
+      "grad_norm": 0.2982316017150879,
+      "learning_rate": 0.00025920195439739415,
+      "loss": 0.6249,
+      "step": 305
+    },
+    {
+      "epoch": 0.6721581548599671,
+      "grad_norm": 0.3072456121444702,
+      "learning_rate": 0.00025895765472312704,
+      "loss": 0.6714,
+      "step": 306
+    },
+    {
+      "epoch": 0.6743547501372872,
+      "grad_norm": 0.2769891321659088,
+      "learning_rate": 0.00025871335504885993,
+      "loss": 0.6913,
+      "step": 307
+    },
+    {
+      "epoch": 0.6765513454146074,
+      "grad_norm": 0.31844112277030945,
+      "learning_rate": 0.0002584690553745928,
+      "loss": 0.7036,
+      "step": 308
+    },
+    {
+      "epoch": 0.6787479406919276,
+      "grad_norm": 0.27636781334877014,
+      "learning_rate": 0.0002582247557003257,
+      "loss": 0.6599,
+      "step": 309
+    },
+    {
+      "epoch": 0.6809445359692476,
+      "grad_norm": 0.2366364449262619,
+      "learning_rate": 0.00025798045602605865,
+      "loss": 0.7003,
+      "step": 310
+    },
+    {
+      "epoch": 0.6831411312465678,
+      "grad_norm": 0.39780211448669434,
+      "learning_rate": 0.00025773615635179154,
+      "loss": 0.5436,
+      "step": 311
+    },
+    {
+      "epoch": 0.685337726523888,
+      "grad_norm": 0.37236976623535156,
+      "learning_rate": 0.00025749185667752443,
+      "loss": 0.7434,
+      "step": 312
+    },
+    {
+      "epoch": 0.6875343218012081,
+      "grad_norm": 0.2848716378211975,
+      "learning_rate": 0.0002572475570032573,
+      "loss": 0.6194,
+      "step": 313
+    },
+    {
+      "epoch": 0.6897309170785283,
+      "grad_norm": 0.22966192662715912,
+      "learning_rate": 0.0002570032573289902,
+      "loss": 0.4599,
+      "step": 314
+    },
+    {
+      "epoch": 0.6919275123558485,
+      "grad_norm": 0.2632063627243042,
+      "learning_rate": 0.0002567589576547231,
+      "loss": 0.4752,
+      "step": 315
+    },
+    {
+      "epoch": 0.6941241076331686,
+      "grad_norm": 0.32491621375083923,
+      "learning_rate": 0.00025651465798045604,
+      "loss": 0.7206,
+      "step": 316
+    },
+    {
+      "epoch": 0.6963207029104888,
+      "grad_norm": 0.26535525918006897,
+      "learning_rate": 0.00025627035830618893,
+      "loss": 0.8002,
+      "step": 317
+    },
+    {
+      "epoch": 0.6985172981878089,
+      "grad_norm": 0.2586904764175415,
+      "learning_rate": 0.0002560260586319218,
+      "loss": 0.7599,
+      "step": 318
+    },
+    {
+      "epoch": 0.700713893465129,
+      "grad_norm": 0.2392316460609436,
+      "learning_rate": 0.0002557817589576547,
+      "loss": 0.6144,
+      "step": 319
+    },
+    {
+      "epoch": 0.7029104887424492,
+      "grad_norm": 0.20571915805339813,
+      "learning_rate": 0.0002555374592833876,
+      "loss": 0.6279,
+      "step": 320
+    },
+    {
+      "epoch": 0.7051070840197694,
+      "grad_norm": 0.20141111314296722,
+      "learning_rate": 0.0002552931596091205,
+      "loss": 0.689,
+      "step": 321
+    },
+    {
+      "epoch": 0.7073036792970895,
+      "grad_norm": 0.21566997468471527,
+      "learning_rate": 0.00025504885993485343,
+      "loss": 0.6986,
+      "step": 322
+    },
+    {
+      "epoch": 0.7095002745744097,
+      "grad_norm": 0.2472105771303177,
+      "learning_rate": 0.0002548045602605863,
+      "loss": 0.555,
+      "step": 323
+    },
+    {
+      "epoch": 0.7116968698517299,
+      "grad_norm": 0.2734464704990387,
+      "learning_rate": 0.0002545602605863192,
+      "loss": 0.7447,
+      "step": 324
+    },
+    {
+      "epoch": 0.7138934651290499,
+      "grad_norm": 0.33236053586006165,
+      "learning_rate": 0.0002543159609120521,
+      "loss": 0.6297,
+      "step": 325
+    },
+    {
+      "epoch": 0.7160900604063701,
+      "grad_norm": 0.3248502016067505,
+      "learning_rate": 0.000254071661237785,
+      "loss": 0.6572,
+      "step": 326
+    },
+    {
+      "epoch": 0.7182866556836903,
+      "grad_norm": 0.2228022664785385,
+      "learning_rate": 0.00025382736156351793,
+      "loss": 0.6646,
+      "step": 327
+    },
+    {
+      "epoch": 0.7204832509610104,
+      "grad_norm": 0.31670036911964417,
+      "learning_rate": 0.0002535830618892508,
+      "loss": 0.7559,
+      "step": 328
+    },
+    {
+      "epoch": 0.7226798462383306,
+      "grad_norm": 0.20176252722740173,
+      "learning_rate": 0.0002533387622149837,
+      "loss": 0.6417,
+      "step": 329
+    },
+    {
+      "epoch": 0.7248764415156508,
+      "grad_norm": 0.26634252071380615,
+      "learning_rate": 0.0002530944625407166,
+      "loss": 0.6054,
+      "step": 330
+    },
+    {
+      "epoch": 0.7270730367929709,
+      "grad_norm": 0.3679504692554474,
+      "learning_rate": 0.0002528501628664495,
+      "loss": 0.7664,
+      "step": 331
+    },
+    {
+      "epoch": 0.729269632070291,
+      "grad_norm": 0.23203566670417786,
+      "learning_rate": 0.0002526058631921824,
+      "loss": 0.6089,
+      "step": 332
+    },
+    {
+      "epoch": 0.7314662273476112,
+      "grad_norm": 0.2532273530960083,
+      "learning_rate": 0.0002523615635179153,
+      "loss": 0.7031,
+      "step": 333
+    },
+    {
+      "epoch": 0.7336628226249313,
+      "grad_norm": 0.30534666776657104,
+      "learning_rate": 0.0002521172638436482,
+      "loss": 0.6983,
+      "step": 334
+    },
+    {
+      "epoch": 0.7358594179022515,
+      "grad_norm": 0.22172865271568298,
+      "learning_rate": 0.0002518729641693811,
+      "loss": 0.6501,
+      "step": 335
+    },
+    {
+      "epoch": 0.7380560131795717,
+      "grad_norm": 0.2655487656593323,
+      "learning_rate": 0.000251628664495114,
+      "loss": 0.6469,
+      "step": 336
+    },
+    {
+      "epoch": 0.7402526084568918,
+      "grad_norm": 0.21879547834396362,
+      "learning_rate": 0.0002513843648208469,
+      "loss": 0.794,
+      "step": 337
+    },
+    {
+      "epoch": 0.742449203734212,
+      "grad_norm": 0.264906108379364,
+      "learning_rate": 0.0002511400651465798,
+      "loss": 0.6263,
+      "step": 338
+    },
+    {
+      "epoch": 0.7446457990115322,
+      "grad_norm": 0.3371405005455017,
+      "learning_rate": 0.0002508957654723127,
+      "loss": 0.7331,
+      "step": 339
+    },
+    {
+      "epoch": 0.7468423942888522,
+      "grad_norm": 0.2580586373806,
+      "learning_rate": 0.0002506514657980456,
+      "loss": 0.6835,
+      "step": 340
+    },
+    {
+      "epoch": 0.7490389895661724,
+      "grad_norm": 0.252922385931015,
+      "learning_rate": 0.0002504071661237785,
+      "loss": 0.6124,
+      "step": 341
+    },
+    {
+      "epoch": 0.7512355848434926,
+      "grad_norm": 0.22605293989181519,
+      "learning_rate": 0.0002501628664495114,
+      "loss": 0.5497,
+      "step": 342
+    },
+    {
+      "epoch": 0.7534321801208127,
+      "grad_norm": 0.21860457956790924,
+      "learning_rate": 0.00024991856677524427,
+      "loss": 0.6299,
+      "step": 343
+    },
+    {
+      "epoch": 0.7556287753981329,
+      "grad_norm": 0.32161346077919006,
+      "learning_rate": 0.0002496742671009772,
+      "loss": 0.6569,
+      "step": 344
+    },
+    {
+      "epoch": 0.7578253706754531,
+      "grad_norm": 0.24863888323307037,
+      "learning_rate": 0.0002494299674267101,
+      "loss": 0.5729,
+      "step": 345
+    },
+    {
+      "epoch": 0.7600219659527732,
+      "grad_norm": 0.32666417956352234,
+      "learning_rate": 0.000249185667752443,
+      "loss": 0.6684,
+      "step": 346
+    },
+    {
+      "epoch": 0.7622185612300933,
+      "grad_norm": 0.21403762698173523,
+      "learning_rate": 0.0002489413680781759,
+      "loss": 0.6633,
+      "step": 347
+    },
+    {
+      "epoch": 0.7644151565074135,
+      "grad_norm": 0.2976447641849518,
+      "learning_rate": 0.00024869706840390877,
+      "loss": 0.6632,
+      "step": 348
+    },
+    {
+      "epoch": 0.7666117517847336,
+      "grad_norm": 0.19317321479320526,
+      "learning_rate": 0.0002484527687296417,
+      "loss": 0.4828,
+      "step": 349
+    },
+    {
+      "epoch": 0.7688083470620538,
+      "grad_norm": 0.27904486656188965,
+      "learning_rate": 0.0002482084690553746,
+      "loss": 0.7672,
+      "step": 350
+    },
+    {
+      "epoch": 0.771004942339374,
+      "grad_norm": 0.27140355110168457,
+      "learning_rate": 0.0002479641693811075,
+      "loss": 0.8469,
+      "step": 351
+    },
+    {
+      "epoch": 0.7732015376166941,
+      "grad_norm": 0.47070470452308655,
+      "learning_rate": 0.0002477198697068404,
+      "loss": 0.6845,
+      "step": 352
+    },
+    {
+      "epoch": 0.7753981328940143,
+      "grad_norm": 0.21045289933681488,
+      "learning_rate": 0.00024747557003257327,
+      "loss": 0.5284,
+      "step": 353
+    },
+    {
+      "epoch": 0.7775947281713345,
+      "grad_norm": 0.27455511689186096,
+      "learning_rate": 0.00024723127035830616,
+      "loss": 0.6647,
+      "step": 354
+    },
+    {
+      "epoch": 0.7797913234486545,
+      "grad_norm": 0.30961373448371887,
+      "learning_rate": 0.0002469869706840391,
+      "loss": 0.4736,
+      "step": 355
+    },
+    {
+      "epoch": 0.7819879187259747,
+      "grad_norm": 0.24248731136322021,
+      "learning_rate": 0.000246742671009772,
+      "loss": 0.571,
+      "step": 356
+    },
+    {
+      "epoch": 0.7841845140032949,
+      "grad_norm": 0.1999034434556961,
+      "learning_rate": 0.0002464983713355049,
+      "loss": 0.7033,
+      "step": 357
+    },
+    {
+      "epoch": 0.786381109280615,
+      "grad_norm": 0.2284795194864273,
+      "learning_rate": 0.00024625407166123777,
+      "loss": 0.6227,
+      "step": 358
+    },
+    {
+      "epoch": 0.7885777045579352,
+      "grad_norm": 0.41865697503089905,
+      "learning_rate": 0.00024600977198697066,
+      "loss": 0.5828,
+      "step": 359
+    },
+    {
+      "epoch": 0.7907742998352554,
+      "grad_norm": 0.18465656042099,
+      "learning_rate": 0.00024576547231270354,
+      "loss": 0.574,
+      "step": 360
+    },
+    {
+      "epoch": 0.7929708951125755,
+      "grad_norm": 0.23776018619537354,
+      "learning_rate": 0.0002455211726384365,
+      "loss": 0.7043,
+      "step": 361
+    },
+    {
+      "epoch": 0.7951674903898956,
+      "grad_norm": 0.2736996114253998,
+      "learning_rate": 0.0002452768729641694,
+      "loss": 0.6874,
+      "step": 362
+    },
+    {
+      "epoch": 0.7973640856672158,
+      "grad_norm": 0.24799343943595886,
+      "learning_rate": 0.00024503257328990227,
+      "loss": 0.5501,
+      "step": 363
+    },
+    {
+      "epoch": 0.7995606809445359,
+      "grad_norm": 0.31103914976119995,
+      "learning_rate": 0.00024478827361563516,
+      "loss": 0.5875,
+      "step": 364
+    },
+    {
+      "epoch": 0.8017572762218561,
+      "grad_norm": 0.25656548142433167,
+      "learning_rate": 0.00024454397394136804,
+      "loss": 0.4904,
+      "step": 365
+    },
+    {
+      "epoch": 0.8039538714991763,
+      "grad_norm": 0.24041813611984253,
+      "learning_rate": 0.000244299674267101,
+      "loss": 0.7227,
+      "step": 366
+    },
+    {
+      "epoch": 0.8061504667764964,
+      "grad_norm": 0.5465381145477295,
+      "learning_rate": 0.00024405537459283385,
+      "loss": 0.6738,
+      "step": 367
+    },
+    {
+      "epoch": 0.8083470620538166,
+      "grad_norm": 0.2519221603870392,
+      "learning_rate": 0.00024381107491856677,
+      "loss": 0.5178,
+      "step": 368
+    },
+    {
+      "epoch": 0.8105436573311368,
+      "grad_norm": 0.19826839864253998,
+      "learning_rate": 0.00024356677524429966,
+      "loss": 0.5782,
+      "step": 369
+    },
+    {
+      "epoch": 0.8127402526084568,
+      "grad_norm": 0.1782648265361786,
+      "learning_rate": 0.00024332247557003257,
+      "loss": 0.4674,
+      "step": 370
+    },
+    {
+      "epoch": 0.814936847885777,
+      "grad_norm": 0.18223823606967926,
+      "learning_rate": 0.00024307817589576546,
+      "loss": 0.7485,
+      "step": 371
+    },
+    {
+      "epoch": 0.8171334431630972,
+      "grad_norm": 0.27911701798439026,
+      "learning_rate": 0.00024283387622149835,
+      "loss": 0.6583,
+      "step": 372
+    },
+    {
+      "epoch": 0.8193300384404174,
+      "grad_norm": 0.19583971798419952,
+      "learning_rate": 0.00024258957654723127,
+      "loss": 0.6343,
+      "step": 373
+    },
+    {
+      "epoch": 0.8215266337177375,
+      "grad_norm": 0.3335391879081726,
+      "learning_rate": 0.00024234527687296416,
+      "loss": 0.6222,
+      "step": 374
+    },
+    {
+      "epoch": 0.8237232289950577,
+      "grad_norm": 0.2782208323478699,
+      "learning_rate": 0.00024210097719869705,
+      "loss": 0.748,
+      "step": 375
+    },
+    {
+      "epoch": 0.8259198242723779,
+      "grad_norm": 0.26846539974212646,
+      "learning_rate": 0.00024185667752442996,
+      "loss": 0.7366,
+      "step": 376
+    },
+    {
+      "epoch": 0.828116419549698,
+      "grad_norm": 0.29050639271736145,
+      "learning_rate": 0.00024161237785016285,
+      "loss": 0.6431,
+      "step": 377
+    },
+    {
+      "epoch": 0.8303130148270181,
+      "grad_norm": 0.20523902773857117,
+      "learning_rate": 0.00024136807817589574,
+      "loss": 0.622,
+      "step": 378
+    },
+    {
+      "epoch": 0.8325096101043383,
+      "grad_norm": 0.2321215718984604,
+      "learning_rate": 0.00024112377850162866,
+      "loss": 0.4937,
+      "step": 379
+    },
+    {
+      "epoch": 0.8347062053816584,
+      "grad_norm": 0.2826741635799408,
+      "learning_rate": 0.00024087947882736155,
+      "loss": 0.7109,
+      "step": 380
+    },
+    {
+      "epoch": 0.8369028006589786,
+      "grad_norm": 0.28134724497795105,
+      "learning_rate": 0.00024063517915309446,
+      "loss": 0.6189,
+      "step": 381
+    },
+    {
+      "epoch": 0.8390993959362988,
+      "grad_norm": 0.23492133617401123,
+      "learning_rate": 0.00024039087947882735,
+      "loss": 0.7149,
+      "step": 382
+    },
+    {
+      "epoch": 0.8412959912136189,
+      "grad_norm": 0.21543096005916595,
+      "learning_rate": 0.00024014657980456024,
+      "loss": 0.721,
+      "step": 383
+    },
+    {
+      "epoch": 0.8434925864909391,
+      "grad_norm": 0.40665197372436523,
+      "learning_rate": 0.00023990228013029316,
+      "loss": 0.7509,
+      "step": 384
+    },
+    {
+      "epoch": 0.8456891817682592,
+      "grad_norm": 0.36472800374031067,
+      "learning_rate": 0.00023965798045602605,
+      "loss": 0.7802,
+      "step": 385
+    },
+    {
+      "epoch": 0.8478857770455793,
+      "grad_norm": 0.3685151934623718,
+      "learning_rate": 0.00023941368078175893,
+      "loss": 0.7019,
+      "step": 386
+    },
+    {
+      "epoch": 0.8500823723228995,
+      "grad_norm": 0.2089424580335617,
+      "learning_rate": 0.00023916938110749185,
+      "loss": 0.5904,
+      "step": 387
+    },
+    {
+      "epoch": 0.8522789676002197,
+      "grad_norm": 0.26246070861816406,
+      "learning_rate": 0.00023892508143322474,
+      "loss": 0.709,
+      "step": 388
+    },
+    {
+      "epoch": 0.8544755628775398,
+      "grad_norm": 0.33440127968788147,
+      "learning_rate": 0.00023868078175895763,
+      "loss": 0.5349,
+      "step": 389
+    },
+    {
+      "epoch": 0.85667215815486,
+      "grad_norm": 0.20990434288978577,
+      "learning_rate": 0.00023843648208469055,
+      "loss": 0.5728,
+      "step": 390
+    },
+    {
+      "epoch": 0.8588687534321802,
+      "grad_norm": 0.219412162899971,
+      "learning_rate": 0.00023819218241042343,
+      "loss": 0.666,
+      "step": 391
+    },
+    {
+      "epoch": 0.8610653487095002,
+      "grad_norm": 0.23335079848766327,
+      "learning_rate": 0.00023794788273615632,
+      "loss": 0.8763,
+      "step": 392
+    },
+    {
+      "epoch": 0.8632619439868204,
+      "grad_norm": 0.3229334056377411,
+      "learning_rate": 0.00023770358306188924,
+      "loss": 0.9441,
+      "step": 393
+    },
+    {
+      "epoch": 0.8654585392641406,
+      "grad_norm": 0.3056965172290802,
+      "learning_rate": 0.00023745928338762213,
+      "loss": 0.6201,
+      "step": 394
+    },
+    {
+      "epoch": 0.8676551345414607,
+      "grad_norm": 0.19122830033302307,
+      "learning_rate": 0.00023721498371335505,
+      "loss": 0.5433,
+      "step": 395
+    },
+    {
+      "epoch": 0.8698517298187809,
+      "grad_norm": 0.21614734828472137,
+      "learning_rate": 0.00023697068403908794,
+      "loss": 0.6362,
+      "step": 396
+    },
+    {
+      "epoch": 0.8720483250961011,
+      "grad_norm": 0.253456175327301,
+      "learning_rate": 0.00023672638436482082,
+      "loss": 0.626,
+      "step": 397
+    },
+    {
+      "epoch": 0.8742449203734212,
+      "grad_norm": 0.1684822142124176,
+      "learning_rate": 0.00023648208469055374,
+      "loss": 0.5681,
+      "step": 398
+    },
+    {
+      "epoch": 0.8764415156507414,
+      "grad_norm": 0.27529001235961914,
+      "learning_rate": 0.00023623778501628663,
+      "loss": 0.6869,
+      "step": 399
+    },
+    {
+      "epoch": 0.8786381109280615,
+      "grad_norm": 0.2589411735534668,
+      "learning_rate": 0.00023599348534201952,
+      "loss": 0.6928,
+      "step": 400
+    },
+    {
+      "epoch": 0.8808347062053816,
+      "grad_norm": 0.31063005328178406,
+      "learning_rate": 0.00023574918566775244,
+      "loss": 0.5345,
+      "step": 401
+    },
+    {
+      "epoch": 0.8830313014827018,
+      "grad_norm": 0.18234965205192566,
+      "learning_rate": 0.00023550488599348532,
+      "loss": 0.6332,
+      "step": 402
+    },
+    {
+      "epoch": 0.885227896760022,
+      "grad_norm": 0.28171342611312866,
+      "learning_rate": 0.00023526058631921821,
+      "loss": 0.7823,
+      "step": 403
+    },
+    {
+      "epoch": 0.8874244920373421,
+      "grad_norm": 0.1757378727197647,
+      "learning_rate": 0.00023501628664495113,
+      "loss": 0.7809,
+      "step": 404
+    },
+    {
+      "epoch": 0.8896210873146623,
+      "grad_norm": 0.2226286232471466,
+      "learning_rate": 0.00023477198697068402,
+      "loss": 0.541,
+      "step": 405
+    },
+    {
+      "epoch": 0.8918176825919825,
+      "grad_norm": 0.224228173494339,
+      "learning_rate": 0.0002345276872964169,
+      "loss": 0.5744,
+      "step": 406
+    },
+    {
+      "epoch": 0.8940142778693025,
+      "grad_norm": 0.26919159293174744,
+      "learning_rate": 0.00023428338762214982,
+      "loss": 0.5189,
+      "step": 407
+    },
+    {
+      "epoch": 0.8962108731466227,
+      "grad_norm": 0.22223412990570068,
+      "learning_rate": 0.00023403908794788271,
+      "loss": 0.5565,
+      "step": 408
+    },
+    {
+      "epoch": 0.8984074684239429,
+      "grad_norm": 0.15144801139831543,
+      "learning_rate": 0.00023379478827361563,
+      "loss": 0.5894,
+      "step": 409
+    },
+    {
+      "epoch": 0.900604063701263,
+      "grad_norm": 0.3811337649822235,
+      "learning_rate": 0.00023355048859934852,
+      "loss": 0.732,
+      "step": 410
+    },
+    {
+      "epoch": 0.9028006589785832,
+      "grad_norm": 0.20232048630714417,
+      "learning_rate": 0.0002333061889250814,
+      "loss": 0.5375,
+      "step": 411
+    },
+    {
+      "epoch": 0.9049972542559034,
+      "grad_norm": 0.3878488838672638,
+      "learning_rate": 0.00023306188925081432,
+      "loss": 0.5928,
+      "step": 412
+    },
+    {
+      "epoch": 0.9071938495332235,
+      "grad_norm": 0.19909335672855377,
+      "learning_rate": 0.00023281758957654721,
+      "loss": 0.6441,
+      "step": 413
+    },
+    {
+      "epoch": 0.9093904448105437,
+      "grad_norm": 0.2696763873100281,
+      "learning_rate": 0.0002325732899022801,
+      "loss": 0.7046,
+      "step": 414
+    },
+    {
+      "epoch": 0.9115870400878638,
+      "grad_norm": 0.2511988580226898,
+      "learning_rate": 0.00023232899022801302,
+      "loss": 0.6144,
+      "step": 415
+    },
+    {
+      "epoch": 0.9137836353651839,
+      "grad_norm": 0.2677559554576874,
+      "learning_rate": 0.0002320846905537459,
+      "loss": 0.7622,
+      "step": 416
+    },
+    {
+      "epoch": 0.9159802306425041,
+      "grad_norm": 0.26745161414146423,
+      "learning_rate": 0.0002318403908794788,
+      "loss": 0.6546,
+      "step": 417
+    },
+    {
+      "epoch": 0.9181768259198243,
+      "grad_norm": 0.6473090648651123,
+      "learning_rate": 0.00023159609120521171,
+      "loss": 0.5058,
+      "step": 418
+    },
+    {
+      "epoch": 0.9203734211971444,
+      "grad_norm": 0.26495614647865295,
+      "learning_rate": 0.0002313517915309446,
+      "loss": 0.5467,
+      "step": 419
+    },
+    {
+      "epoch": 0.9225700164744646,
+      "grad_norm": 0.23127774894237518,
+      "learning_rate": 0.00023110749185667752,
+      "loss": 0.6621,
+      "step": 420
+    },
+    {
+      "epoch": 0.9247666117517848,
+      "grad_norm": 0.3387419581413269,
+      "learning_rate": 0.0002308631921824104,
+      "loss": 0.5939,
+      "step": 421
+    },
+    {
+      "epoch": 0.9269632070291048,
+      "grad_norm": 0.25289273262023926,
+      "learning_rate": 0.0002306188925081433,
+      "loss": 0.6301,
+      "step": 422
+    },
+    {
+      "epoch": 0.929159802306425,
+      "grad_norm": 0.2828865945339203,
+      "learning_rate": 0.00023037459283387621,
+      "loss": 0.801,
+      "step": 423
+    },
+    {
+      "epoch": 0.9313563975837452,
+      "grad_norm": 0.36047419905662537,
+      "learning_rate": 0.0002301302931596091,
+      "loss": 0.6538,
+      "step": 424
+    },
+    {
+      "epoch": 0.9335529928610653,
+      "grad_norm": 0.20596139132976532,
+      "learning_rate": 0.000229885993485342,
+      "loss": 0.5612,
+      "step": 425
+    },
+    {
+      "epoch": 0.9357495881383855,
+      "grad_norm": 0.28171420097351074,
+      "learning_rate": 0.0002296416938110749,
+      "loss": 0.5958,
+      "step": 426
+    },
+    {
+      "epoch": 0.9379461834157057,
+      "grad_norm": 0.2218017876148224,
+      "learning_rate": 0.0002293973941368078,
+      "loss": 0.6339,
+      "step": 427
+    },
+    {
+      "epoch": 0.9401427786930258,
+      "grad_norm": 0.2413358837366104,
+      "learning_rate": 0.0002291530944625407,
+      "loss": 0.524,
+      "step": 428
+    },
+    {
+      "epoch": 0.942339373970346,
+      "grad_norm": 0.2129582017660141,
+      "learning_rate": 0.0002289087947882736,
+      "loss": 0.4704,
+      "step": 429
+    },
+    {
+      "epoch": 0.9445359692476661,
+      "grad_norm": 0.26698023080825806,
+      "learning_rate": 0.0002286644951140065,
+      "loss": 0.6956,
+      "step": 430
+    },
+    {
+      "epoch": 0.9467325645249862,
+      "grad_norm": 0.2373342663049698,
+      "learning_rate": 0.00022842019543973938,
+      "loss": 0.5087,
+      "step": 431
+    },
+    {
+      "epoch": 0.9489291598023064,
+      "grad_norm": 0.28995776176452637,
+      "learning_rate": 0.0002281758957654723,
+      "loss": 0.7717,
+      "step": 432
+    },
+    {
+      "epoch": 0.9511257550796266,
+      "grad_norm": 0.23951876163482666,
+      "learning_rate": 0.0002279315960912052,
+      "loss": 0.649,
+      "step": 433
+    },
+    {
+      "epoch": 0.9533223503569467,
+      "grad_norm": 0.2564714550971985,
+      "learning_rate": 0.0002276872964169381,
+      "loss": 0.541,
+      "step": 434
+    },
+    {
+      "epoch": 0.9555189456342669,
+      "grad_norm": 0.1765207052230835,
+      "learning_rate": 0.000227442996742671,
+      "loss": 0.5613,
+      "step": 435
+    },
+    {
+      "epoch": 0.9577155409115871,
+      "grad_norm": 0.26193463802337646,
+      "learning_rate": 0.00022719869706840388,
+      "loss": 0.4931,
+      "step": 436
+    },
+    {
+      "epoch": 0.9599121361889071,
+      "grad_norm": 0.23861880600452423,
+      "learning_rate": 0.0002269543973941368,
+      "loss": 0.6243,
+      "step": 437
+    },
+    {
+      "epoch": 0.9621087314662273,
+      "grad_norm": 0.22465889155864716,
+      "learning_rate": 0.0002267100977198697,
+      "loss": 0.5287,
+      "step": 438
+    },
+    {
+      "epoch": 0.9643053267435475,
+      "grad_norm": 0.24283123016357422,
+      "learning_rate": 0.00022646579804560258,
+      "loss": 0.6793,
+      "step": 439
+    },
+    {
+      "epoch": 0.9665019220208677,
+      "grad_norm": 0.32185986638069153,
+      "learning_rate": 0.0002262214983713355,
+      "loss": 0.7162,
+      "step": 440
+    },
+    {
+      "epoch": 0.9686985172981878,
+      "grad_norm": 0.3535987138748169,
+      "learning_rate": 0.00022597719869706838,
+      "loss": 0.5804,
+      "step": 441
+    },
+    {
+      "epoch": 0.970895112575508,
+      "grad_norm": 0.27203071117401123,
+      "learning_rate": 0.00022573289902280127,
+      "loss": 0.5682,
+      "step": 442
+    },
+    {
+      "epoch": 0.9730917078528282,
+      "grad_norm": 0.1956549733877182,
+      "learning_rate": 0.0002254885993485342,
+      "loss": 0.6613,
+      "step": 443
+    },
+    {
+      "epoch": 0.9752883031301482,
+      "grad_norm": 0.2391206920146942,
+      "learning_rate": 0.00022524429967426708,
+      "loss": 0.7182,
+      "step": 444
+    },
+    {
+      "epoch": 0.9774848984074684,
+      "grad_norm": 0.38639745116233826,
+      "learning_rate": 0.000225,
+      "loss": 0.6728,
+      "step": 445
+    },
+    {
+      "epoch": 0.9796814936847886,
+      "grad_norm": 0.7342166900634766,
+      "learning_rate": 0.00022475570032573288,
+      "loss": 0.741,
+      "step": 446
+    },
+    {
+      "epoch": 0.9818780889621087,
+      "grad_norm": 0.4025481641292572,
+      "learning_rate": 0.00022451140065146577,
+      "loss": 0.7566,
+      "step": 447
+    },
+    {
+      "epoch": 0.9840746842394289,
+      "grad_norm": 0.25724342465400696,
+      "learning_rate": 0.0002242671009771987,
+      "loss": 0.6401,
+      "step": 448
+    },
+    {
+      "epoch": 0.9862712795167491,
+      "grad_norm": 0.2377576380968094,
+      "learning_rate": 0.00022402280130293158,
+      "loss": 0.7034,
+      "step": 449
+    },
+    {
+      "epoch": 0.9884678747940692,
+      "grad_norm": 0.23696520924568176,
+      "learning_rate": 0.00022377850162866447,
+      "loss": 0.6146,
+      "step": 450
+    },
+    {
+      "epoch": 0.9906644700713894,
+      "grad_norm": 0.27091917395591736,
+      "learning_rate": 0.00022353420195439738,
+      "loss": 0.5849,
+      "step": 451
+    },
+    {
+      "epoch": 0.9928610653487095,
+      "grad_norm": 0.23507006466388702,
+      "learning_rate": 0.00022328990228013027,
+      "loss": 0.6694,
+      "step": 452
+    },
+    {
+      "epoch": 0.9950576606260296,
+      "grad_norm": 0.21350957453250885,
+      "learning_rate": 0.00022304560260586316,
+      "loss": 0.6957,
+      "step": 453
+    },
+    {
+      "epoch": 0.9972542559033498,
+      "grad_norm": 0.2614002227783203,
+      "learning_rate": 0.00022280130293159608,
+      "loss": 0.5294,
+      "step": 454
+    },
+    {
+      "epoch": 0.99945085118067,
+      "grad_norm": 0.23171333968639374,
+      "learning_rate": 0.00022255700325732897,
+      "loss": 0.6098,
+      "step": 455
+    },
+    {
+      "epoch": 1.00164744645799,
+      "grad_norm": 0.28078219294548035,
+      "learning_rate": 0.00022231270358306186,
+      "loss": 0.6972,
+      "step": 456
+    },
+    {
+      "epoch": 1.0038440417353103,
+      "grad_norm": 0.2107093781232834,
+      "learning_rate": 0.00022206840390879477,
+      "loss": 0.7031,
+      "step": 457
+    },
+    {
+      "epoch": 1.0060406370126305,
+      "grad_norm": 0.2080126702785492,
+      "learning_rate": 0.00022182410423452766,
+      "loss": 0.6249,
+      "step": 458
+    },
+    {
+      "epoch": 1.0082372322899507,
+      "grad_norm": 0.23544473946094513,
+      "learning_rate": 0.00022157980456026058,
+      "loss": 0.5519,
+      "step": 459
+    },
+    {
+      "epoch": 1.0104338275672706,
+      "grad_norm": 0.20219573378562927,
+      "learning_rate": 0.00022133550488599347,
+      "loss": 0.5681,
+      "step": 460
+    },
+    {
+      "epoch": 1.0126304228445908,
+      "grad_norm": 0.19428123533725739,
+      "learning_rate": 0.00022109120521172636,
+      "loss": 0.5015,
+      "step": 461
+    },
+    {
+      "epoch": 1.014827018121911,
+      "grad_norm": 0.26935046911239624,
+      "learning_rate": 0.00022084690553745927,
+      "loss": 0.7994,
+      "step": 462
+    },
+    {
+      "epoch": 1.0170236133992312,
+      "grad_norm": 0.24201099574565887,
+      "learning_rate": 0.00022060260586319216,
+      "loss": 0.7245,
+      "step": 463
+    },
+    {
+      "epoch": 1.0192202086765514,
+      "grad_norm": 0.273305207490921,
+      "learning_rate": 0.00022035830618892505,
+      "loss": 0.5755,
+      "step": 464
+    },
+    {
+      "epoch": 1.0214168039538716,
+      "grad_norm": 0.2769998610019684,
+      "learning_rate": 0.00022011400651465797,
+      "loss": 0.7564,
+      "step": 465
+    },
+    {
+      "epoch": 1.0236133992311915,
+      "grad_norm": 0.23386965692043304,
+      "learning_rate": 0.00021986970684039086,
+      "loss": 0.7521,
+      "step": 466
+    },
+    {
+      "epoch": 1.0258099945085117,
+      "grad_norm": 0.26884371042251587,
+      "learning_rate": 0.00021962540716612375,
+      "loss": 0.6332,
+      "step": 467
+    },
+    {
+      "epoch": 1.028006589785832,
+      "grad_norm": 0.2686975598335266,
+      "learning_rate": 0.00021938110749185666,
+      "loss": 0.5853,
+      "step": 468
+    },
+    {
+      "epoch": 1.0302031850631521,
+      "grad_norm": 0.24127250909805298,
+      "learning_rate": 0.00021913680781758955,
+      "loss": 0.6844,
+      "step": 469
+    },
+    {
+      "epoch": 1.0323997803404723,
+      "grad_norm": 0.24710983037948608,
+      "learning_rate": 0.00021889250814332244,
+      "loss": 0.6794,
+      "step": 470
+    },
+    {
+      "epoch": 1.0345963756177925,
+      "grad_norm": 0.29141977429389954,
+      "learning_rate": 0.00021864820846905536,
+      "loss": 0.5223,
+      "step": 471
+    },
+    {
+      "epoch": 1.0367929708951125,
+      "grad_norm": 0.3217121958732605,
+      "learning_rate": 0.00021840390879478825,
+      "loss": 0.6072,
+      "step": 472
+    },
+    {
+      "epoch": 1.0389895661724327,
+      "grad_norm": 0.2607187032699585,
+      "learning_rate": 0.00021815960912052116,
+      "loss": 0.7811,
+      "step": 473
+    },
+    {
+      "epoch": 1.0411861614497528,
+      "grad_norm": 0.21983908116817474,
+      "learning_rate": 0.00021791530944625405,
+      "loss": 0.5989,
+      "step": 474
+    },
+    {
+      "epoch": 1.043382756727073,
+      "grad_norm": 0.203995943069458,
+      "learning_rate": 0.00021767100977198694,
+      "loss": 0.6149,
+      "step": 475
+    },
+    {
+      "epoch": 1.0455793520043932,
+      "grad_norm": 0.22711966931819916,
+      "learning_rate": 0.00021742671009771986,
+      "loss": 0.7011,
+      "step": 476
+    },
+    {
+      "epoch": 1.0477759472817134,
+      "grad_norm": 0.2459600418806076,
+      "learning_rate": 0.00021718241042345275,
+      "loss": 0.5322,
+      "step": 477
+    },
+    {
+      "epoch": 1.0499725425590336,
+      "grad_norm": 0.2453494817018509,
+      "learning_rate": 0.00021693811074918564,
+      "loss": 0.6645,
+      "step": 478
+    },
+    {
+      "epoch": 1.0521691378363536,
+      "grad_norm": 0.2935507893562317,
+      "learning_rate": 0.00021669381107491855,
+      "loss": 0.6353,
+      "step": 479
+    },
+    {
+      "epoch": 1.0543657331136738,
+      "grad_norm": 0.19175612926483154,
+      "learning_rate": 0.00021644951140065144,
+      "loss": 0.5643,
+      "step": 480
+    },
+    {
+      "epoch": 1.056562328390994,
+      "grad_norm": 0.15534724295139313,
+      "learning_rate": 0.00021620521172638433,
+      "loss": 0.566,
+      "step": 481
+    },
+    {
+      "epoch": 1.0587589236683141,
+      "grad_norm": 0.2306113988161087,
+      "learning_rate": 0.00021596091205211725,
+      "loss": 0.567,
+      "step": 482
+    },
+    {
+      "epoch": 1.0609555189456343,
+      "grad_norm": 0.2719849646091461,
+      "learning_rate": 0.00021571661237785014,
+      "loss": 0.5887,
+      "step": 483
+    },
+    {
+      "epoch": 1.0631521142229543,
+      "grad_norm": 0.20400957763195038,
+      "learning_rate": 0.00021547231270358305,
+      "loss": 0.6445,
+      "step": 484
+    },
+    {
+      "epoch": 1.0653487095002745,
+      "grad_norm": 0.2154066413640976,
+      "learning_rate": 0.00021522801302931594,
+      "loss": 0.6192,
+      "step": 485
+    },
+    {
+      "epoch": 1.0675453047775947,
+      "grad_norm": 0.23527441918849945,
+      "learning_rate": 0.00021498371335504883,
+      "loss": 0.4818,
+      "step": 486
+    },
+    {
+      "epoch": 1.0697419000549149,
+      "grad_norm": 0.2430892139673233,
+      "learning_rate": 0.00021473941368078175,
+      "loss": 0.6013,
+      "step": 487
+    },
+    {
+      "epoch": 1.071938495332235,
+      "grad_norm": 0.25507187843322754,
+      "learning_rate": 0.00021449511400651464,
+      "loss": 0.6764,
+      "step": 488
+    },
+    {
+      "epoch": 1.0741350906095553,
+      "grad_norm": 0.21335454285144806,
+      "learning_rate": 0.00021425081433224753,
+      "loss": 0.5671,
+      "step": 489
+    },
+    {
+      "epoch": 1.0763316858868754,
+      "grad_norm": 0.17528115212917328,
+      "learning_rate": 0.00021400651465798044,
+      "loss": 0.5526,
+      "step": 490
+    },
+    {
+      "epoch": 1.0785282811641954,
+      "grad_norm": 0.3730625808238983,
+      "learning_rate": 0.00021376221498371333,
+      "loss": 0.7861,
+      "step": 491
+    },
+    {
+      "epoch": 1.0807248764415156,
+      "grad_norm": 0.24294064939022064,
+      "learning_rate": 0.00021351791530944622,
+      "loss": 0.5481,
+      "step": 492
+    },
+    {
+      "epoch": 1.0829214717188358,
+      "grad_norm": 0.33950790762901306,
+      "learning_rate": 0.00021327361563517914,
+      "loss": 0.658,
+      "step": 493
+    },
+    {
+      "epoch": 1.085118066996156,
+      "grad_norm": 0.20844106376171112,
+      "learning_rate": 0.00021302931596091203,
+      "loss": 0.4991,
+      "step": 494
+    },
+    {
+      "epoch": 1.0873146622734762,
+      "grad_norm": 0.326637327671051,
+      "learning_rate": 0.00021278501628664491,
+      "loss": 0.6395,
+      "step": 495
+    },
+    {
+      "epoch": 1.0895112575507964,
+      "grad_norm": 0.228445366024971,
+      "learning_rate": 0.00021254071661237783,
+      "loss": 0.5756,
+      "step": 496
+    },
+    {
+      "epoch": 1.0917078528281163,
+      "grad_norm": 0.18142180144786835,
+      "learning_rate": 0.00021229641693811072,
+      "loss": 0.5695,
+      "step": 497
+    },
+    {
+      "epoch": 1.0939044481054365,
+      "grad_norm": 0.28032442927360535,
+      "learning_rate": 0.00021205211726384364,
+      "loss": 0.5606,
+      "step": 498
+    },
+    {
+      "epoch": 1.0961010433827567,
+      "grad_norm": 0.27549880743026733,
+      "learning_rate": 0.00021180781758957653,
+      "loss": 0.6348,
+      "step": 499
+    },
+    {
+      "epoch": 1.098297638660077,
+      "grad_norm": 0.2927318215370178,
+      "learning_rate": 0.00021156351791530941,
+      "loss": 0.6735,
+      "step": 500
+    },
+    {
+      "epoch": 1.100494233937397,
+      "grad_norm": 0.2442750632762909,
+      "learning_rate": 0.00021131921824104233,
+      "loss": 0.7149,
+      "step": 501
+    },
+    {
+      "epoch": 1.1026908292147173,
+      "grad_norm": 0.22121095657348633,
+      "learning_rate": 0.00021107491856677522,
+      "loss": 0.5388,
+      "step": 502
+    },
+    {
+      "epoch": 1.1048874244920373,
+      "grad_norm": 0.39968645572662354,
+      "learning_rate": 0.0002108306188925081,
+      "loss": 0.6293,
+      "step": 503
+    },
+    {
+      "epoch": 1.1070840197693574,
+      "grad_norm": 0.239824116230011,
+      "learning_rate": 0.00021058631921824103,
+      "loss": 0.5627,
+      "step": 504
+    },
+    {
+      "epoch": 1.1092806150466776,
+      "grad_norm": 0.2966967225074768,
+      "learning_rate": 0.00021034201954397392,
+      "loss": 0.7827,
+      "step": 505
+    },
+    {
+      "epoch": 1.1114772103239978,
+      "grad_norm": 0.24256278574466705,
+      "learning_rate": 0.0002100977198697068,
+      "loss": 0.6161,
+      "step": 506
+    },
+    {
+      "epoch": 1.113673805601318,
+      "grad_norm": 0.24170801043510437,
+      "learning_rate": 0.00020985342019543972,
+      "loss": 0.6266,
+      "step": 507
+    },
+    {
+      "epoch": 1.1158704008786382,
+      "grad_norm": 0.3593856692314148,
+      "learning_rate": 0.0002096091205211726,
+      "loss": 0.6836,
+      "step": 508
+    },
+    {
+      "epoch": 1.1180669961559582,
+      "grad_norm": 0.20119130611419678,
+      "learning_rate": 0.0002093648208469055,
+      "loss": 0.6714,
+      "step": 509
+    },
+    {
+      "epoch": 1.1202635914332784,
+      "grad_norm": 0.32899942994117737,
+      "learning_rate": 0.00020912052117263842,
+      "loss": 0.7524,
+      "step": 510
+    },
+    {
+      "epoch": 1.1224601867105986,
+      "grad_norm": 0.3130359649658203,
+      "learning_rate": 0.0002088762214983713,
+      "loss": 0.6052,
+      "step": 511
+    },
+    {
+      "epoch": 1.1246567819879187,
+      "grad_norm": 0.3245997428894043,
+      "learning_rate": 0.00020863192182410422,
+      "loss": 0.6389,
+      "step": 512
+    },
+    {
+      "epoch": 1.126853377265239,
+      "grad_norm": 0.28697583079338074,
+      "learning_rate": 0.0002083876221498371,
+      "loss": 0.7449,
+      "step": 513
+    },
+    {
+      "epoch": 1.1290499725425591,
+      "grad_norm": 0.3154304027557373,
+      "learning_rate": 0.00020814332247557,
+      "loss": 0.6277,
+      "step": 514
+    },
+    {
+      "epoch": 1.131246567819879,
+      "grad_norm": 0.3123483657836914,
+      "learning_rate": 0.00020789902280130292,
+      "loss": 0.6372,
+      "step": 515
+    },
+    {
+      "epoch": 1.1334431630971993,
+      "grad_norm": 0.44807755947113037,
+      "learning_rate": 0.0002076547231270358,
+      "loss": 0.6219,
+      "step": 516
+    },
+    {
+      "epoch": 1.1356397583745195,
+      "grad_norm": 0.31425386667251587,
+      "learning_rate": 0.0002074104234527687,
+      "loss": 0.6152,
+      "step": 517
+    },
+    {
+      "epoch": 1.1378363536518397,
+      "grad_norm": 0.316241055727005,
+      "learning_rate": 0.0002071661237785016,
+      "loss": 0.4323,
+      "step": 518
+    },
+    {
+      "epoch": 1.1400329489291599,
+      "grad_norm": 0.3034268915653229,
+      "learning_rate": 0.0002069218241042345,
+      "loss": 0.4482,
+      "step": 519
+    },
+    {
+      "epoch": 1.14222954420648,
+      "grad_norm": 0.3242630362510681,
+      "learning_rate": 0.0002066775244299674,
+      "loss": 0.6411,
+      "step": 520
+    },
+    {
+      "epoch": 1.1444261394838002,
+      "grad_norm": 0.24422337114810944,
+      "learning_rate": 0.0002064332247557003,
+      "loss": 0.5386,
+      "step": 521
+    },
+    {
+      "epoch": 1.1466227347611202,
+      "grad_norm": 0.31941747665405273,
+      "learning_rate": 0.0002061889250814332,
+      "loss": 0.5543,
+      "step": 522
+    },
+    {
+      "epoch": 1.1488193300384404,
+      "grad_norm": 0.24687407910823822,
+      "learning_rate": 0.0002059446254071661,
+      "loss": 0.4754,
+      "step": 523
+    },
+    {
+      "epoch": 1.1510159253157606,
+      "grad_norm": 0.2494855374097824,
+      "learning_rate": 0.000205700325732899,
+      "loss": 0.6946,
+      "step": 524
+    },
+    {
+      "epoch": 1.1532125205930808,
+      "grad_norm": 0.2201654613018036,
+      "learning_rate": 0.0002054560260586319,
+      "loss": 0.6018,
+      "step": 525
+    },
+    {
+      "epoch": 1.155409115870401,
+      "grad_norm": 0.3383132517337799,
+      "learning_rate": 0.0002052117263843648,
+      "loss": 0.8209,
+      "step": 526
+    },
+    {
+      "epoch": 1.157605711147721,
+      "grad_norm": 0.24267181754112244,
+      "learning_rate": 0.0002049674267100977,
+      "loss": 0.5275,
+      "step": 527
+    },
+    {
+      "epoch": 1.1598023064250411,
+      "grad_norm": 0.29713040590286255,
+      "learning_rate": 0.00020472312703583058,
+      "loss": 0.6879,
+      "step": 528
+    },
+    {
+      "epoch": 1.1619989017023613,
+      "grad_norm": 0.26054662466049194,
+      "learning_rate": 0.0002044788273615635,
+      "loss": 0.5191,
+      "step": 529
+    },
+    {
+      "epoch": 1.1641954969796815,
+      "grad_norm": 0.24083611369132996,
+      "learning_rate": 0.0002042345276872964,
+      "loss": 0.4728,
+      "step": 530
+    },
+    {
+      "epoch": 1.1663920922570017,
+      "grad_norm": 0.26093512773513794,
+      "learning_rate": 0.00020399022801302928,
+      "loss": 0.5469,
+      "step": 531
+    },
+    {
+      "epoch": 1.1685886875343219,
+      "grad_norm": 0.2655316889286041,
+      "learning_rate": 0.0002037459283387622,
+      "loss": 0.7343,
+      "step": 532
+    },
+    {
+      "epoch": 1.170785282811642,
+      "grad_norm": 0.2698550224304199,
+      "learning_rate": 0.00020350162866449508,
+      "loss": 0.6522,
+      "step": 533
+    },
+    {
+      "epoch": 1.172981878088962,
+      "grad_norm": 0.33289939165115356,
+      "learning_rate": 0.00020325732899022797,
+      "loss": 0.6999,
+      "step": 534
+    },
+    {
+      "epoch": 1.1751784733662822,
+      "grad_norm": 0.30757299065589905,
+      "learning_rate": 0.0002030130293159609,
+      "loss": 0.596,
+      "step": 535
+    },
+    {
+      "epoch": 1.1773750686436024,
+      "grad_norm": 0.3720068335533142,
+      "learning_rate": 0.00020276872964169378,
+      "loss": 0.5237,
+      "step": 536
+    },
+    {
+      "epoch": 1.1795716639209226,
+      "grad_norm": 0.2637775242328644,
+      "learning_rate": 0.0002025244299674267,
+      "loss": 0.5863,
+      "step": 537
+    },
+    {
+      "epoch": 1.1817682591982428,
+      "grad_norm": 0.3159791827201843,
+      "learning_rate": 0.00020228013029315958,
+      "loss": 0.5236,
+      "step": 538
+    },
+    {
+      "epoch": 1.1839648544755628,
+      "grad_norm": 0.2702966034412384,
+      "learning_rate": 0.00020203583061889247,
+      "loss": 0.8043,
+      "step": 539
+    },
+    {
+      "epoch": 1.186161449752883,
+      "grad_norm": 0.3391665816307068,
+      "learning_rate": 0.0002017915309446254,
+      "loss": 0.5504,
+      "step": 540
+    },
+    {
+      "epoch": 1.1883580450302031,
+      "grad_norm": 0.2930944561958313,
+      "learning_rate": 0.00020154723127035828,
+      "loss": 0.6143,
+      "step": 541
+    },
+    {
+      "epoch": 1.1905546403075233,
+      "grad_norm": 0.2322131246328354,
+      "learning_rate": 0.00020130293159609117,
+      "loss": 0.5242,
+      "step": 542
+    },
+    {
+      "epoch": 1.1927512355848435,
+      "grad_norm": 0.29261189699172974,
+      "learning_rate": 0.00020105863192182408,
+      "loss": 0.6781,
+      "step": 543
+    },
+    {
+      "epoch": 1.1949478308621637,
+      "grad_norm": 0.24949099123477936,
+      "learning_rate": 0.00020081433224755697,
+      "loss": 0.629,
+      "step": 544
+    },
+    {
+      "epoch": 1.197144426139484,
+      "grad_norm": 0.32845792174339294,
+      "learning_rate": 0.00020057003257328986,
+      "loss": 0.6543,
+      "step": 545
+    },
+    {
+      "epoch": 1.1993410214168039,
+      "grad_norm": 0.22020256519317627,
+      "learning_rate": 0.00020032573289902278,
+      "loss": 0.6444,
+      "step": 546
+    },
+    {
+      "epoch": 1.201537616694124,
+      "grad_norm": 0.36626219749450684,
+      "learning_rate": 0.00020008143322475567,
+      "loss": 0.8552,
+      "step": 547
+    },
+    {
+      "epoch": 1.2037342119714443,
+      "grad_norm": 0.2300000935792923,
+      "learning_rate": 0.00019983713355048856,
+      "loss": 0.6076,
+      "step": 548
+    },
+    {
+      "epoch": 1.2059308072487644,
+      "grad_norm": 0.29556599259376526,
+      "learning_rate": 0.00019959283387622147,
+      "loss": 0.8374,
+      "step": 549
+    },
+    {
+      "epoch": 1.2081274025260846,
+      "grad_norm": 0.5969923138618469,
+      "learning_rate": 0.00019934853420195436,
+      "loss": 0.691,
+      "step": 550
+    },
+    {
+      "epoch": 1.2103239978034046,
+      "grad_norm": 0.33564502000808716,
+      "learning_rate": 0.00019910423452768728,
+      "loss": 0.5999,
+      "step": 551
+    },
+    {
+      "epoch": 1.2125205930807248,
+      "grad_norm": 0.30871856212615967,
+      "learning_rate": 0.00019885993485342017,
+      "loss": 0.6732,
+      "step": 552
+    },
+    {
+      "epoch": 1.214717188358045,
+      "grad_norm": 0.2994288206100464,
+      "learning_rate": 0.00019861563517915306,
+      "loss": 0.6876,
+      "step": 553
+    },
+    {
+      "epoch": 1.2169137836353652,
+      "grad_norm": 0.2464524507522583,
+      "learning_rate": 0.00019837133550488597,
+      "loss": 0.6489,
+      "step": 554
+    },
+    {
+      "epoch": 1.2191103789126854,
+      "grad_norm": 0.4448976218700409,
+      "learning_rate": 0.00019812703583061886,
+      "loss": 0.713,
+      "step": 555
+    },
+    {
+      "epoch": 1.2213069741900056,
+      "grad_norm": 0.309722900390625,
+      "learning_rate": 0.00019788273615635175,
+      "loss": 0.5181,
+      "step": 556
+    },
+    {
+      "epoch": 1.2235035694673257,
+      "grad_norm": 0.33176374435424805,
+      "learning_rate": 0.00019763843648208467,
+      "loss": 0.6338,
+      "step": 557
+    },
+    {
+      "epoch": 1.2257001647446457,
+      "grad_norm": 0.32494696974754333,
+      "learning_rate": 0.00019739413680781756,
+      "loss": 0.5799,
+      "step": 558
+    },
+    {
+      "epoch": 1.227896760021966,
+      "grad_norm": 0.6247962117195129,
+      "learning_rate": 0.00019714983713355045,
+      "loss": 0.7473,
+      "step": 559
+    },
+    {
+      "epoch": 1.230093355299286,
+      "grad_norm": 0.2604236602783203,
+      "learning_rate": 0.00019690553745928336,
+      "loss": 0.6953,
+      "step": 560
+    },
+    {
+      "epoch": 1.2322899505766063,
+      "grad_norm": 0.3901156187057495,
+      "learning_rate": 0.00019666123778501625,
+      "loss": 0.5913,
+      "step": 561
+    },
+    {
+      "epoch": 1.2344865458539265,
+      "grad_norm": 0.3233148753643036,
+      "learning_rate": 0.00019641693811074917,
+      "loss": 0.5988,
+      "step": 562
+    },
+    {
+      "epoch": 1.2366831411312464,
+      "grad_norm": 0.39019376039505005,
+      "learning_rate": 0.00019617263843648206,
+      "loss": 0.7419,
+      "step": 563
+    },
+    {
+      "epoch": 1.2388797364085666,
+      "grad_norm": 0.39409011602401733,
+      "learning_rate": 0.00019592833876221495,
+      "loss": 0.5695,
+      "step": 564
+    },
+    {
+      "epoch": 1.2410763316858868,
+      "grad_norm": 0.37627455592155457,
+      "learning_rate": 0.00019568403908794786,
+      "loss": 0.6623,
+      "step": 565
+    },
+    {
+      "epoch": 1.243272926963207,
+      "grad_norm": 0.4057425856590271,
+      "learning_rate": 0.00019543973941368075,
+      "loss": 0.5859,
+      "step": 566
+    },
+    {
+      "epoch": 1.2454695222405272,
+      "grad_norm": 0.2544190585613251,
+      "learning_rate": 0.00019519543973941364,
+      "loss": 0.6259,
+      "step": 567
+    },
+    {
+      "epoch": 1.2476661175178474,
+      "grad_norm": 0.26313352584838867,
+      "learning_rate": 0.00019495114006514656,
+      "loss": 0.6458,
+      "step": 568
+    },
+    {
+      "epoch": 1.2498627127951676,
+      "grad_norm": 0.25184911489486694,
+      "learning_rate": 0.00019470684039087945,
+      "loss": 0.5668,
+      "step": 569
+    },
+    {
+      "epoch": 1.2520593080724876,
+      "grad_norm": 0.28255292773246765,
+      "learning_rate": 0.00019446254071661234,
+      "loss": 0.4272,
+      "step": 570
+    },
+    {
+      "epoch": 1.2542559033498077,
+      "grad_norm": 0.28654971718788147,
+      "learning_rate": 0.00019421824104234525,
+      "loss": 0.753,
+      "step": 571
+    },
+    {
+      "epoch": 1.256452498627128,
+      "grad_norm": 0.3245311975479126,
+      "learning_rate": 0.00019397394136807814,
+      "loss": 0.632,
+      "step": 572
+    },
+    {
+      "epoch": 1.2586490939044481,
+      "grad_norm": 0.393248587846756,
+      "learning_rate": 0.00019372964169381103,
+      "loss": 0.5824,
+      "step": 573
+    },
+    {
+      "epoch": 1.2608456891817683,
+      "grad_norm": 0.6012445092201233,
+      "learning_rate": 0.00019348534201954395,
+      "loss": 0.6949,
+      "step": 574
+    },
+    {
+      "epoch": 1.2630422844590883,
+      "grad_norm": 0.2230839878320694,
+      "learning_rate": 0.00019324104234527684,
+      "loss": 0.7238,
+      "step": 575
+    },
+    {
+      "epoch": 1.2652388797364087,
+      "grad_norm": 0.2652684152126312,
+      "learning_rate": 0.00019299674267100975,
+      "loss": 0.5898,
+      "step": 576
+    },
+    {
+      "epoch": 1.2674354750137287,
+      "grad_norm": 0.34143152832984924,
+      "learning_rate": 0.00019275244299674264,
+      "loss": 0.8115,
+      "step": 577
+    },
+    {
+      "epoch": 1.2696320702910489,
+      "grad_norm": 0.3003377914428711,
+      "learning_rate": 0.00019250814332247553,
+      "loss": 0.4729,
+      "step": 578
+    },
+    {
+      "epoch": 1.271828665568369,
+      "grad_norm": 0.28741148114204407,
+      "learning_rate": 0.00019226384364820845,
+      "loss": 0.4625,
+      "step": 579
+    },
+    {
+      "epoch": 1.2740252608456892,
+      "grad_norm": 0.22336037456989288,
+      "learning_rate": 0.00019201954397394134,
+      "loss": 0.4885,
+      "step": 580
+    },
+    {
+      "epoch": 1.2762218561230094,
+      "grad_norm": 0.25551778078079224,
+      "learning_rate": 0.00019177524429967423,
+      "loss": 0.641,
+      "step": 581
+    },
+    {
+      "epoch": 1.2784184514003294,
+      "grad_norm": 0.40498101711273193,
+      "learning_rate": 0.00019153094462540714,
+      "loss": 0.5463,
+      "step": 582
+    },
+    {
+      "epoch": 1.2806150466776496,
+      "grad_norm": 0.21880993247032166,
+      "learning_rate": 0.00019128664495114003,
+      "loss": 0.5707,
+      "step": 583
+    },
+    {
+      "epoch": 1.2828116419549698,
+      "grad_norm": 0.38854214549064636,
+      "learning_rate": 0.00019104234527687292,
+      "loss": 0.4861,
+      "step": 584
+    },
+    {
+      "epoch": 1.28500823723229,
+      "grad_norm": 0.45537322759628296,
+      "learning_rate": 0.00019079804560260584,
+      "loss": 0.64,
+      "step": 585
+    },
+    {
+      "epoch": 1.2872048325096102,
+      "grad_norm": 0.24819771945476532,
+      "learning_rate": 0.00019055374592833873,
+      "loss": 0.6348,
+      "step": 586
+    },
+    {
+      "epoch": 1.2894014277869301,
+      "grad_norm": 0.2755352556705475,
+      "learning_rate": 0.00019030944625407164,
+      "loss": 0.5839,
+      "step": 587
+    },
+    {
+      "epoch": 1.2915980230642505,
+      "grad_norm": 0.3702441155910492,
+      "learning_rate": 0.00019006514657980453,
+      "loss": 0.6219,
+      "step": 588
+    },
+    {
+      "epoch": 1.2937946183415705,
+      "grad_norm": 0.33641189336776733,
+      "learning_rate": 0.00018982084690553742,
+      "loss": 0.5754,
+      "step": 589
+    },
+    {
+      "epoch": 1.2959912136188907,
+      "grad_norm": 0.25161850452423096,
+      "learning_rate": 0.00018957654723127034,
+      "loss": 0.6415,
+      "step": 590
+    },
+    {
+      "epoch": 1.2981878088962109,
+      "grad_norm": 0.21588459610939026,
+      "learning_rate": 0.00018933224755700323,
+      "loss": 0.4814,
+      "step": 591
+    },
+    {
+      "epoch": 1.300384404173531,
+      "grad_norm": 0.3121974766254425,
+      "learning_rate": 0.00018908794788273612,
+      "loss": 0.6383,
+      "step": 592
+    },
+    {
+      "epoch": 1.3025809994508513,
+      "grad_norm": 0.22696368396282196,
+      "learning_rate": 0.00018884364820846903,
+      "loss": 0.5242,
+      "step": 593
+    },
+    {
+      "epoch": 1.3047775947281712,
+      "grad_norm": 0.21158656477928162,
+      "learning_rate": 0.00018859934853420192,
+      "loss": 0.6592,
+      "step": 594
+    },
+    {
+      "epoch": 1.3069741900054914,
+      "grad_norm": 0.27026697993278503,
+      "learning_rate": 0.0001883550488599348,
+      "loss": 0.7014,
+      "step": 595
+    },
+    {
+      "epoch": 1.3091707852828116,
+      "grad_norm": 0.2758921980857849,
+      "learning_rate": 0.00018811074918566773,
+      "loss": 0.5577,
+      "step": 596
+    },
+    {
+      "epoch": 1.3113673805601318,
+      "grad_norm": 0.3476666510105133,
+      "learning_rate": 0.00018786644951140062,
+      "loss": 0.5819,
+      "step": 597
+    },
+    {
+      "epoch": 1.313563975837452,
+      "grad_norm": 0.2886625826358795,
+      "learning_rate": 0.0001876221498371335,
+      "loss": 0.5309,
+      "step": 598
+    },
+    {
+      "epoch": 1.3157605711147722,
+      "grad_norm": 0.30730798840522766,
+      "learning_rate": 0.00018737785016286645,
+      "loss": 0.5981,
+      "step": 599
+    },
+    {
+      "epoch": 1.3179571663920924,
+      "grad_norm": 0.2137092649936676,
+      "learning_rate": 0.00018713355048859934,
+      "loss": 0.7442,
+      "step": 600
+    },
+    {
+      "epoch": 1.3201537616694123,
+      "grad_norm": 0.29109349846839905,
+      "learning_rate": 0.00018688925081433225,
+      "loss": 0.5948,
+      "step": 601
+    },
+    {
+      "epoch": 1.3223503569467325,
+      "grad_norm": 0.22900830209255219,
+      "learning_rate": 0.00018664495114006514,
+      "loss": 0.6371,
+      "step": 602
+    },
+    {
+      "epoch": 1.3245469522240527,
+      "grad_norm": 0.24927644431591034,
+      "learning_rate": 0.00018640065146579803,
+      "loss": 0.649,
+      "step": 603
+    },
+    {
+      "epoch": 1.326743547501373,
+      "grad_norm": 0.31492671370506287,
+      "learning_rate": 0.00018615635179153095,
+      "loss": 0.6534,
+      "step": 604
+    },
+    {
+      "epoch": 1.328940142778693,
+      "grad_norm": 0.27633586525917053,
+      "learning_rate": 0.00018591205211726384,
+      "loss": 0.7102,
+      "step": 605
+    },
+    {
+      "epoch": 1.331136738056013,
+      "grad_norm": 0.3094451129436493,
+      "learning_rate": 0.00018566775244299675,
+      "loss": 0.5398,
+      "step": 606
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 0.28441157937049866,
+      "learning_rate": 0.00018542345276872964,
+      "loss": 0.4747,
+      "step": 607
+    },
+    {
+      "epoch": 1.3355299286106534,
+      "grad_norm": 0.26143428683280945,
+      "learning_rate": 0.00018517915309446253,
+      "loss": 0.5823,
+      "step": 608
+    },
+    {
+      "epoch": 1.3377265238879736,
+      "grad_norm": 0.34576717019081116,
+      "learning_rate": 0.00018493485342019545,
+      "loss": 0.6141,
+      "step": 609
+    },
+    {
+      "epoch": 1.3399231191652938,
+      "grad_norm": 0.25988277792930603,
+      "learning_rate": 0.00018469055374592834,
+      "loss": 0.646,
+      "step": 610
+    },
+    {
+      "epoch": 1.342119714442614,
+      "grad_norm": 0.3158949315547943,
+      "learning_rate": 0.00018444625407166123,
+      "loss": 0.6636,
+      "step": 611
+    },
+    {
+      "epoch": 1.3443163097199342,
+      "grad_norm": 0.2994331419467926,
+      "learning_rate": 0.00018420195439739414,
+      "loss": 0.5607,
+      "step": 612
+    },
+    {
+      "epoch": 1.3465129049972542,
+      "grad_norm": 0.24911248683929443,
+      "learning_rate": 0.00018395765472312703,
+      "loss": 0.6823,
+      "step": 613
+    },
+    {
+      "epoch": 1.3487095002745744,
+      "grad_norm": 0.22547048330307007,
+      "learning_rate": 0.00018371335504885992,
+      "loss": 0.5487,
+      "step": 614
+    },
+    {
+      "epoch": 1.3509060955518946,
+      "grad_norm": 0.3022003769874573,
+      "learning_rate": 0.00018346905537459284,
+      "loss": 0.6847,
+      "step": 615
+    },
+    {
+      "epoch": 1.3531026908292147,
+      "grad_norm": 0.4155174493789673,
+      "learning_rate": 0.00018322475570032573,
+      "loss": 0.7165,
+      "step": 616
+    },
+    {
+      "epoch": 1.355299286106535,
+      "grad_norm": 0.32970142364501953,
+      "learning_rate": 0.00018298045602605864,
+      "loss": 0.6365,
+      "step": 617
+    },
+    {
+      "epoch": 1.357495881383855,
+      "grad_norm": 0.3691022992134094,
+      "learning_rate": 0.00018273615635179153,
+      "loss": 0.8507,
+      "step": 618
+    },
+    {
+      "epoch": 1.359692476661175,
+      "grad_norm": 0.2730911672115326,
+      "learning_rate": 0.00018249185667752442,
+      "loss": 0.4883,
+      "step": 619
+    },
+    {
+      "epoch": 1.3618890719384953,
+      "grad_norm": 0.2999876141548157,
+      "learning_rate": 0.00018224755700325734,
+      "loss": 0.5825,
+      "step": 620
+    },
+    {
+      "epoch": 1.3640856672158155,
+      "grad_norm": 0.33087801933288574,
+      "learning_rate": 0.00018200325732899023,
+      "loss": 0.8109,
+      "step": 621
+    },
+    {
+      "epoch": 1.3662822624931357,
+      "grad_norm": 0.29036375880241394,
+      "learning_rate": 0.00018175895765472312,
+      "loss": 0.5852,
+      "step": 622
+    },
+    {
+      "epoch": 1.3684788577704559,
+      "grad_norm": 0.27295586466789246,
+      "learning_rate": 0.00018151465798045603,
+      "loss": 0.5135,
+      "step": 623
+    },
+    {
+      "epoch": 1.370675453047776,
+      "grad_norm": 0.36017584800720215,
+      "learning_rate": 0.00018127035830618892,
+      "loss": 0.5901,
+      "step": 624
+    },
+    {
+      "epoch": 1.372872048325096,
+      "grad_norm": 0.39633694291114807,
+      "learning_rate": 0.0001810260586319218,
+      "loss": 0.7028,
+      "step": 625
+    },
+    {
+      "epoch": 1.3750686436024162,
+      "grad_norm": 0.2307877540588379,
+      "learning_rate": 0.00018078175895765473,
+      "loss": 0.5295,
+      "step": 626
+    },
+    {
+      "epoch": 1.3772652388797364,
+      "grad_norm": 0.24561668932437897,
+      "learning_rate": 0.00018053745928338762,
+      "loss": 0.6041,
+      "step": 627
+    },
+    {
+      "epoch": 1.3794618341570566,
+      "grad_norm": 0.3117298483848572,
+      "learning_rate": 0.0001802931596091205,
+      "loss": 0.4198,
+      "step": 628
+    },
+    {
+      "epoch": 1.3816584294343768,
+      "grad_norm": 0.2746444046497345,
+      "learning_rate": 0.00018004885993485342,
+      "loss": 0.6042,
+      "step": 629
+    },
+    {
+      "epoch": 1.3838550247116967,
+      "grad_norm": 0.3963146209716797,
+      "learning_rate": 0.0001798045602605863,
+      "loss": 0.5576,
+      "step": 630
+    },
+    {
+      "epoch": 1.3860516199890172,
+      "grad_norm": 0.24994483590126038,
+      "learning_rate": 0.00017956026058631923,
+      "loss": 0.601,
+      "step": 631
+    },
+    {
+      "epoch": 1.3882482152663371,
+      "grad_norm": 0.27220186591148376,
+      "learning_rate": 0.00017931596091205212,
+      "loss": 0.7411,
+      "step": 632
+    },
+    {
+      "epoch": 1.3904448105436573,
+      "grad_norm": 0.2554558217525482,
+      "learning_rate": 0.000179071661237785,
+      "loss": 0.645,
+      "step": 633
+    },
+    {
+      "epoch": 1.3926414058209775,
+      "grad_norm": 0.28812485933303833,
+      "learning_rate": 0.00017882736156351792,
+      "loss": 0.549,
+      "step": 634
+    },
+    {
+      "epoch": 1.3948380010982977,
+      "grad_norm": 0.5013327598571777,
+      "learning_rate": 0.0001785830618892508,
+      "loss": 0.6658,
+      "step": 635
+    },
+    {
+      "epoch": 1.3970345963756179,
+      "grad_norm": 0.24490439891815186,
+      "learning_rate": 0.0001783387622149837,
+      "loss": 0.7053,
+      "step": 636
+    },
+    {
+      "epoch": 1.3992311916529379,
+      "grad_norm": 0.3620946705341339,
+      "learning_rate": 0.00017809446254071662,
+      "loss": 0.5418,
+      "step": 637
+    },
+    {
+      "epoch": 1.401427786930258,
+      "grad_norm": 0.27582746744155884,
+      "learning_rate": 0.0001778501628664495,
+      "loss": 0.5234,
+      "step": 638
+    },
+    {
+      "epoch": 1.4036243822075782,
+      "grad_norm": 0.3775025010108948,
+      "learning_rate": 0.0001776058631921824,
+      "loss": 0.5629,
+      "step": 639
+    },
+    {
+      "epoch": 1.4058209774848984,
+      "grad_norm": 0.3011874854564667,
+      "learning_rate": 0.0001773615635179153,
+      "loss": 0.8416,
+      "step": 640
+    },
+    {
+      "epoch": 1.4080175727622186,
+      "grad_norm": 0.29876771569252014,
+      "learning_rate": 0.0001771172638436482,
+      "loss": 0.6106,
+      "step": 641
+    },
+    {
+      "epoch": 1.4102141680395386,
+      "grad_norm": 0.2435743510723114,
+      "learning_rate": 0.0001768729641693811,
+      "loss": 0.7133,
+      "step": 642
+    },
+    {
+      "epoch": 1.412410763316859,
+      "grad_norm": 0.4375503659248352,
+      "learning_rate": 0.000176628664495114,
+      "loss": 0.5089,
+      "step": 643
+    },
+    {
+      "epoch": 1.414607358594179,
+      "grad_norm": 0.2969537675380707,
+      "learning_rate": 0.0001763843648208469,
+      "loss": 0.5059,
+      "step": 644
+    },
+    {
+      "epoch": 1.4168039538714992,
+      "grad_norm": 0.3530534505844116,
+      "learning_rate": 0.0001761400651465798,
+      "loss": 0.5312,
+      "step": 645
+    },
+    {
+      "epoch": 1.4190005491488193,
+      "grad_norm": 0.30920514464378357,
+      "learning_rate": 0.0001758957654723127,
+      "loss": 0.5282,
+      "step": 646
+    },
+    {
+      "epoch": 1.4211971444261395,
+      "grad_norm": 0.4011867940425873,
+      "learning_rate": 0.0001756514657980456,
+      "loss": 0.5913,
+      "step": 647
+    },
+    {
+      "epoch": 1.4233937397034597,
+      "grad_norm": 0.3105852007865906,
+      "learning_rate": 0.0001754071661237785,
+      "loss": 0.558,
+      "step": 648
+    },
+    {
+      "epoch": 1.4255903349807797,
+      "grad_norm": 0.3915203809738159,
+      "learning_rate": 0.0001751628664495114,
+      "loss": 0.6793,
+      "step": 649
+    },
+    {
+      "epoch": 1.4277869302580999,
+      "grad_norm": 0.27737900614738464,
+      "learning_rate": 0.00017491856677524429,
+      "loss": 0.4995,
+      "step": 650
+    },
+    {
+      "epoch": 1.42998352553542,
+      "grad_norm": 0.2909097373485565,
+      "learning_rate": 0.0001746742671009772,
+      "loss": 0.5888,
+      "step": 651
+    },
+    {
+      "epoch": 1.4321801208127403,
+      "grad_norm": 0.25826534628868103,
+      "learning_rate": 0.0001744299674267101,
+      "loss": 0.571,
+      "step": 652
+    },
+    {
+      "epoch": 1.4343767160900605,
+      "grad_norm": 0.2521781921386719,
+      "learning_rate": 0.00017418566775244298,
+      "loss": 0.5878,
+      "step": 653
+    },
+    {
+      "epoch": 1.4365733113673804,
+      "grad_norm": 0.2738262414932251,
+      "learning_rate": 0.0001739413680781759,
+      "loss": 0.5563,
+      "step": 654
+    },
+    {
+      "epoch": 1.4387699066447008,
+      "grad_norm": 0.33885934948921204,
+      "learning_rate": 0.00017369706840390879,
+      "loss": 0.6024,
+      "step": 655
+    },
+    {
+      "epoch": 1.4409665019220208,
+      "grad_norm": 0.3954622745513916,
+      "learning_rate": 0.0001734527687296417,
+      "loss": 0.6705,
+      "step": 656
+    },
+    {
+      "epoch": 1.443163097199341,
+      "grad_norm": 0.2776380479335785,
+      "learning_rate": 0.0001732084690553746,
+      "loss": 0.7622,
+      "step": 657
+    },
+    {
+      "epoch": 1.4453596924766612,
+      "grad_norm": 0.25171712040901184,
+      "learning_rate": 0.00017296416938110748,
+      "loss": 0.7241,
+      "step": 658
+    },
+    {
+      "epoch": 1.4475562877539814,
+      "grad_norm": 0.2903113067150116,
+      "learning_rate": 0.0001727198697068404,
+      "loss": 0.7672,
+      "step": 659
+    },
+    {
+      "epoch": 1.4497528830313016,
+      "grad_norm": 0.22792574763298035,
+      "learning_rate": 0.00017247557003257329,
+      "loss": 0.5305,
+      "step": 660
+    },
+    {
+      "epoch": 1.4519494783086215,
+      "grad_norm": 0.2526301145553589,
+      "learning_rate": 0.00017223127035830618,
+      "loss": 0.5573,
+      "step": 661
+    },
+    {
+      "epoch": 1.4541460735859417,
+      "grad_norm": 0.33352550864219666,
+      "learning_rate": 0.0001719869706840391,
+      "loss": 0.6404,
+      "step": 662
+    },
+    {
+      "epoch": 1.456342668863262,
+      "grad_norm": 0.2948947250843048,
+      "learning_rate": 0.00017174267100977198,
+      "loss": 0.6157,
+      "step": 663
+    },
+    {
+      "epoch": 1.458539264140582,
+      "grad_norm": 0.24244947731494904,
+      "learning_rate": 0.00017149837133550487,
+      "loss": 0.5437,
+      "step": 664
+    },
+    {
+      "epoch": 1.4607358594179023,
+      "grad_norm": 0.3552567660808563,
+      "learning_rate": 0.00017125407166123779,
+      "loss": 0.7909,
+      "step": 665
+    },
+    {
+      "epoch": 1.4629324546952225,
+      "grad_norm": 0.37764763832092285,
+      "learning_rate": 0.00017100977198697068,
+      "loss": 0.7283,
+      "step": 666
+    },
+    {
+      "epoch": 1.4651290499725427,
+      "grad_norm": 0.36119577288627625,
+      "learning_rate": 0.00017076547231270356,
+      "loss": 0.5586,
+      "step": 667
+    },
+    {
+      "epoch": 1.4673256452498626,
+      "grad_norm": 0.27259424328804016,
+      "learning_rate": 0.00017052117263843648,
+      "loss": 0.757,
+      "step": 668
+    },
+    {
+      "epoch": 1.4695222405271828,
+      "grad_norm": 0.30097636580467224,
+      "learning_rate": 0.00017027687296416937,
+      "loss": 0.5771,
+      "step": 669
+    },
+    {
+      "epoch": 1.471718835804503,
+      "grad_norm": 0.28512677550315857,
+      "learning_rate": 0.00017003257328990229,
+      "loss": 0.6112,
+      "step": 670
+    },
+    {
+      "epoch": 1.4739154310818232,
+      "grad_norm": 0.3389562964439392,
+      "learning_rate": 0.00016978827361563518,
+      "loss": 0.7254,
+      "step": 671
+    },
+    {
+      "epoch": 1.4761120263591434,
+      "grad_norm": 0.2517207860946655,
+      "learning_rate": 0.00016954397394136806,
+      "loss": 0.6422,
+      "step": 672
+    },
+    {
+      "epoch": 1.4783086216364634,
+      "grad_norm": 0.3356015384197235,
+      "learning_rate": 0.00016929967426710098,
+      "loss": 0.6283,
+      "step": 673
+    },
+    {
+      "epoch": 1.4805052169137836,
+      "grad_norm": 0.21315288543701172,
+      "learning_rate": 0.00016905537459283387,
+      "loss": 0.6604,
+      "step": 674
+    },
+    {
+      "epoch": 1.4827018121911038,
+      "grad_norm": 0.25601252913475037,
+      "learning_rate": 0.00016881107491856676,
+      "loss": 0.5474,
+      "step": 675
+    },
+    {
+      "epoch": 1.484898407468424,
+      "grad_norm": 0.26255157589912415,
+      "learning_rate": 0.00016856677524429968,
+      "loss": 0.6266,
+      "step": 676
+    },
+    {
+      "epoch": 1.4870950027457441,
+      "grad_norm": 0.26035693287849426,
+      "learning_rate": 0.00016832247557003257,
+      "loss": 0.6205,
+      "step": 677
+    },
+    {
+      "epoch": 1.4892915980230643,
+      "grad_norm": 0.3519388735294342,
+      "learning_rate": 0.00016807817589576545,
+      "loss": 0.6616,
+      "step": 678
+    },
+    {
+      "epoch": 1.4914881933003845,
+      "grad_norm": 0.2628263831138611,
+      "learning_rate": 0.00016783387622149837,
+      "loss": 0.6743,
+      "step": 679
+    },
+    {
+      "epoch": 1.4936847885777045,
+      "grad_norm": 0.6105637550354004,
+      "learning_rate": 0.00016758957654723126,
+      "loss": 0.743,
+      "step": 680
+    },
+    {
+      "epoch": 1.4958813838550247,
+      "grad_norm": 0.29936155676841736,
+      "learning_rate": 0.00016734527687296415,
+      "loss": 0.5813,
+      "step": 681
+    },
+    {
+      "epoch": 1.4980779791323449,
+      "grad_norm": 0.3049934208393097,
+      "learning_rate": 0.00016710097719869707,
+      "loss": 0.8088,
+      "step": 682
+    },
+    {
+      "epoch": 1.500274574409665,
+      "grad_norm": 0.30023303627967834,
+      "learning_rate": 0.00016685667752442995,
+      "loss": 0.4126,
+      "step": 683
+    },
+    {
+      "epoch": 1.5024711696869852,
+      "grad_norm": 0.2981858551502228,
+      "learning_rate": 0.00016661237785016287,
+      "loss": 0.5395,
+      "step": 684
+    },
+    {
+      "epoch": 1.5046677649643052,
+      "grad_norm": 0.3480345904827118,
+      "learning_rate": 0.00016636807817589576,
+      "loss": 0.7007,
+      "step": 685
+    },
+    {
+      "epoch": 1.5068643602416256,
+      "grad_norm": 0.40120676159858704,
+      "learning_rate": 0.00016612377850162865,
+      "loss": 0.7738,
+      "step": 686
+    },
+    {
+      "epoch": 1.5090609555189456,
+      "grad_norm": 0.3745495080947876,
+      "learning_rate": 0.00016587947882736157,
+      "loss": 0.6137,
+      "step": 687
+    },
+    {
+      "epoch": 1.5112575507962658,
+      "grad_norm": 0.27560093998908997,
+      "learning_rate": 0.00016563517915309445,
+      "loss": 0.7115,
+      "step": 688
+    },
+    {
+      "epoch": 1.513454146073586,
+      "grad_norm": 0.2649567127227783,
+      "learning_rate": 0.00016539087947882734,
+      "loss": 0.544,
+      "step": 689
+    },
+    {
+      "epoch": 1.515650741350906,
+      "grad_norm": 0.23219019174575806,
+      "learning_rate": 0.00016514657980456026,
+      "loss": 0.5964,
+      "step": 690
+    },
+    {
+      "epoch": 1.5178473366282264,
+      "grad_norm": 0.2690134048461914,
+      "learning_rate": 0.00016490228013029315,
+      "loss": 0.6205,
+      "step": 691
+    },
+    {
+      "epoch": 1.5200439319055463,
+      "grad_norm": 0.31884023547172546,
+      "learning_rate": 0.00016465798045602604,
+      "loss": 0.6624,
+      "step": 692
+    },
+    {
+      "epoch": 1.5222405271828665,
+      "grad_norm": 0.30244123935699463,
+      "learning_rate": 0.00016441368078175895,
+      "loss": 0.6475,
+      "step": 693
+    },
+    {
+      "epoch": 1.5244371224601867,
+      "grad_norm": 0.2731572687625885,
+      "learning_rate": 0.00016416938110749184,
+      "loss": 0.5411,
+      "step": 694
+    },
+    {
+      "epoch": 1.526633717737507,
+      "grad_norm": 0.2560082972049713,
+      "learning_rate": 0.00016392508143322476,
+      "loss": 0.6536,
+      "step": 695
+    },
+    {
+      "epoch": 1.528830313014827,
+      "grad_norm": 0.24629110097885132,
+      "learning_rate": 0.00016368078175895765,
+      "loss": 0.5667,
+      "step": 696
+    },
+    {
+      "epoch": 1.531026908292147,
+      "grad_norm": 0.3002482056617737,
+      "learning_rate": 0.00016343648208469054,
+      "loss": 0.6016,
+      "step": 697
+    },
+    {
+      "epoch": 1.5332235035694675,
+      "grad_norm": 0.2304675430059433,
+      "learning_rate": 0.00016319218241042346,
+      "loss": 0.5838,
+      "step": 698
+    },
+    {
+      "epoch": 1.5354200988467874,
+      "grad_norm": 0.3193989396095276,
+      "learning_rate": 0.00016294788273615634,
+      "loss": 0.6536,
+      "step": 699
+    },
+    {
+      "epoch": 1.5376166941241076,
+      "grad_norm": 0.2352873533964157,
+      "learning_rate": 0.00016270358306188923,
+      "loss": 0.7291,
+      "step": 700
+    },
+    {
+      "epoch": 1.5398132894014278,
+      "grad_norm": 0.3113371431827545,
+      "learning_rate": 0.00016245928338762215,
+      "loss": 0.567,
+      "step": 701
+    },
+    {
+      "epoch": 1.5420098846787478,
+      "grad_norm": 0.2885269522666931,
+      "learning_rate": 0.00016221498371335504,
+      "loss": 0.595,
+      "step": 702
+    },
+    {
+      "epoch": 1.5442064799560682,
+      "grad_norm": 0.3357740044593811,
+      "learning_rate": 0.00016197068403908793,
+      "loss": 0.5443,
+      "step": 703
+    },
+    {
+      "epoch": 1.5464030752333882,
+      "grad_norm": 0.3329252004623413,
+      "learning_rate": 0.00016172638436482084,
+      "loss": 0.5036,
+      "step": 704
+    },
+    {
+      "epoch": 1.5485996705107083,
+      "grad_norm": 0.45284631848335266,
+      "learning_rate": 0.00016148208469055373,
+      "loss": 0.6286,
+      "step": 705
+    },
+    {
+      "epoch": 1.5507962657880285,
+      "grad_norm": 0.2770507037639618,
+      "learning_rate": 0.00016123778501628662,
+      "loss": 0.5876,
+      "step": 706
+    },
+    {
+      "epoch": 1.5529928610653487,
+      "grad_norm": 0.3554888665676117,
+      "learning_rate": 0.00016099348534201954,
+      "loss": 0.6283,
+      "step": 707
+    },
+    {
+      "epoch": 1.555189456342669,
+      "grad_norm": 0.5835204720497131,
+      "learning_rate": 0.00016074918566775243,
+      "loss": 0.621,
+      "step": 708
+    },
+    {
+      "epoch": 1.5573860516199889,
+      "grad_norm": 0.31040385365486145,
+      "learning_rate": 0.00016050488599348534,
+      "loss": 0.7368,
+      "step": 709
+    },
+    {
+      "epoch": 1.5595826468973093,
+      "grad_norm": 0.33235064148902893,
+      "learning_rate": 0.00016026058631921823,
+      "loss": 0.4383,
+      "step": 710
+    },
+    {
+      "epoch": 1.5617792421746293,
+      "grad_norm": 0.28510627150535583,
+      "learning_rate": 0.00016001628664495112,
+      "loss": 0.7321,
+      "step": 711
+    },
+    {
+      "epoch": 1.5639758374519495,
+      "grad_norm": 0.3064011037349701,
+      "learning_rate": 0.00015977198697068404,
+      "loss": 0.4582,
+      "step": 712
+    },
+    {
+      "epoch": 1.5661724327292696,
+      "grad_norm": 0.3007277846336365,
+      "learning_rate": 0.00015952768729641693,
+      "loss": 0.5331,
+      "step": 713
+    },
+    {
+      "epoch": 1.5683690280065898,
+      "grad_norm": 0.2235771119594574,
+      "learning_rate": 0.00015928338762214982,
+      "loss": 0.4968,
+      "step": 714
+    },
+    {
+      "epoch": 1.57056562328391,
+      "grad_norm": 0.33795225620269775,
+      "learning_rate": 0.00015903908794788273,
+      "loss": 0.6536,
+      "step": 715
+    },
+    {
+      "epoch": 1.57276221856123,
+      "grad_norm": 0.4655829071998596,
+      "learning_rate": 0.00015879478827361562,
+      "loss": 0.7938,
+      "step": 716
+    },
+    {
+      "epoch": 1.5749588138385504,
+      "grad_norm": 0.2759786546230316,
+      "learning_rate": 0.0001585504885993485,
+      "loss": 0.622,
+      "step": 717
+    },
+    {
+      "epoch": 1.5771554091158704,
+      "grad_norm": 0.41292333602905273,
+      "learning_rate": 0.00015830618892508143,
+      "loss": 0.5318,
+      "step": 718
+    },
+    {
+      "epoch": 1.5793520043931906,
+      "grad_norm": 0.3668091893196106,
+      "learning_rate": 0.00015806188925081432,
+      "loss": 0.6211,
+      "step": 719
+    },
+    {
+      "epoch": 1.5815485996705108,
+      "grad_norm": 0.2840741276741028,
+      "learning_rate": 0.00015781758957654723,
+      "loss": 0.5465,
+      "step": 720
+    },
+    {
+      "epoch": 1.5837451949478307,
+      "grad_norm": 0.29478296637535095,
+      "learning_rate": 0.00015757328990228012,
+      "loss": 0.7277,
+      "step": 721
+    },
+    {
+      "epoch": 1.5859417902251511,
+      "grad_norm": 0.3577365279197693,
+      "learning_rate": 0.000157328990228013,
+      "loss": 0.7465,
+      "step": 722
+    },
+    {
+      "epoch": 1.588138385502471,
+      "grad_norm": 0.24068699777126312,
+      "learning_rate": 0.00015708469055374593,
+      "loss": 0.6192,
+      "step": 723
+    },
+    {
+      "epoch": 1.5903349807797913,
+      "grad_norm": 0.27783656120300293,
+      "learning_rate": 0.00015684039087947882,
+      "loss": 0.4855,
+      "step": 724
+    },
+    {
+      "epoch": 1.5925315760571115,
+      "grad_norm": 0.31329163908958435,
+      "learning_rate": 0.0001565960912052117,
+      "loss": 0.6595,
+      "step": 725
+    },
+    {
+      "epoch": 1.5947281713344317,
+      "grad_norm": 0.2875850796699524,
+      "learning_rate": 0.00015635179153094462,
+      "loss": 0.6217,
+      "step": 726
+    },
+    {
+      "epoch": 1.5969247666117519,
+      "grad_norm": 0.28909721970558167,
+      "learning_rate": 0.0001561074918566775,
+      "loss": 0.6968,
+      "step": 727
+    },
+    {
+      "epoch": 1.5991213618890718,
+      "grad_norm": 0.2586701214313507,
+      "learning_rate": 0.0001558631921824104,
+      "loss": 0.6945,
+      "step": 728
+    },
+    {
+      "epoch": 1.6013179571663922,
+      "grad_norm": 0.44186872243881226,
+      "learning_rate": 0.00015561889250814332,
+      "loss": 0.6171,
+      "step": 729
+    },
+    {
+      "epoch": 1.6035145524437122,
+      "grad_norm": 0.33587637543678284,
+      "learning_rate": 0.0001553745928338762,
+      "loss": 0.6425,
+      "step": 730
+    },
+    {
+      "epoch": 1.6057111477210324,
+      "grad_norm": 0.2617303729057312,
+      "learning_rate": 0.0001551302931596091,
+      "loss": 0.4408,
+      "step": 731
+    },
+    {
+      "epoch": 1.6079077429983526,
+      "grad_norm": 0.30631333589553833,
+      "learning_rate": 0.000154885993485342,
+      "loss": 0.603,
+      "step": 732
+    },
+    {
+      "epoch": 1.6101043382756726,
+      "grad_norm": 0.3789410889148712,
+      "learning_rate": 0.0001546416938110749,
+      "loss": 0.7554,
+      "step": 733
+    },
+    {
+      "epoch": 1.612300933552993,
+      "grad_norm": 0.25523197650909424,
+      "learning_rate": 0.00015439739413680782,
+      "loss": 0.6058,
+      "step": 734
+    },
+    {
+      "epoch": 1.614497528830313,
+      "grad_norm": 0.44026675820350647,
+      "learning_rate": 0.0001541530944625407,
+      "loss": 0.6971,
+      "step": 735
+    },
+    {
+      "epoch": 1.6166941241076331,
+      "grad_norm": 0.2667173445224762,
+      "learning_rate": 0.0001539087947882736,
+      "loss": 0.6116,
+      "step": 736
+    },
+    {
+      "epoch": 1.6188907193849533,
+      "grad_norm": 0.32267916202545166,
+      "learning_rate": 0.0001536644951140065,
+      "loss": 0.6901,
+      "step": 737
+    },
+    {
+      "epoch": 1.6210873146622735,
+      "grad_norm": 0.4425192177295685,
+      "learning_rate": 0.0001534201954397394,
+      "loss": 0.5923,
+      "step": 738
+    },
+    {
+      "epoch": 1.6232839099395937,
+      "grad_norm": 0.29255667328834534,
+      "learning_rate": 0.0001531758957654723,
+      "loss": 0.5927,
+      "step": 739
+    },
+    {
+      "epoch": 1.6254805052169137,
+      "grad_norm": 0.34101757407188416,
+      "learning_rate": 0.0001529315960912052,
+      "loss": 0.6434,
+      "step": 740
+    },
+    {
+      "epoch": 1.627677100494234,
+      "grad_norm": 0.2917497754096985,
+      "learning_rate": 0.0001526872964169381,
+      "loss": 0.69,
+      "step": 741
+    },
+    {
+      "epoch": 1.629873695771554,
+      "grad_norm": 0.37623515725135803,
+      "learning_rate": 0.00015244299674267099,
+      "loss": 0.603,
+      "step": 742
+    },
+    {
+      "epoch": 1.6320702910488742,
+      "grad_norm": 0.36102744936943054,
+      "learning_rate": 0.0001521986970684039,
+      "loss": 0.5286,
+      "step": 743
+    },
+    {
+      "epoch": 1.6342668863261944,
+      "grad_norm": 0.32013267278671265,
+      "learning_rate": 0.0001519543973941368,
+      "loss": 0.4552,
+      "step": 744
+    },
+    {
+      "epoch": 1.6364634816035144,
+      "grad_norm": 0.2944977283477783,
+      "learning_rate": 0.00015171009771986968,
+      "loss": 0.5184,
+      "step": 745
+    },
+    {
+      "epoch": 1.6386600768808348,
+      "grad_norm": 0.22503727674484253,
+      "learning_rate": 0.0001514657980456026,
+      "loss": 0.3991,
+      "step": 746
+    },
+    {
+      "epoch": 1.6408566721581548,
+      "grad_norm": 0.3357175588607788,
+      "learning_rate": 0.0001512214983713355,
+      "loss": 0.6048,
+      "step": 747
+    },
+    {
+      "epoch": 1.643053267435475,
+      "grad_norm": 0.33931994438171387,
+      "learning_rate": 0.0001509771986970684,
+      "loss": 0.4785,
+      "step": 748
+    },
+    {
+      "epoch": 1.6452498627127952,
+      "grad_norm": 0.3109085261821747,
+      "learning_rate": 0.0001507328990228013,
+      "loss": 0.4825,
+      "step": 749
+    },
+    {
+      "epoch": 1.6474464579901154,
+      "grad_norm": 0.2905789613723755,
+      "learning_rate": 0.00015048859934853418,
+      "loss": 0.5364,
+      "step": 750
+    },
+    {
+      "epoch": 1.6496430532674355,
+      "grad_norm": 0.25822383165359497,
+      "learning_rate": 0.0001502442996742671,
+      "loss": 0.5862,
+      "step": 751
+    },
+    {
+      "epoch": 1.6518396485447555,
+      "grad_norm": 0.4649774432182312,
+      "learning_rate": 0.00015,
+      "loss": 0.6869,
+      "step": 752
+    },
+    {
+      "epoch": 1.654036243822076,
+      "grad_norm": 0.527242124080658,
+      "learning_rate": 0.00014975570032573288,
+      "loss": 0.8521,
+      "step": 753
+    },
+    {
+      "epoch": 1.656232839099396,
+      "grad_norm": 0.2887413799762726,
+      "learning_rate": 0.0001495114006514658,
+      "loss": 0.4893,
+      "step": 754
+    },
+    {
+      "epoch": 1.658429434376716,
+      "grad_norm": 0.2745928168296814,
+      "learning_rate": 0.00014926710097719868,
+      "loss": 0.5866,
+      "step": 755
+    },
+    {
+      "epoch": 1.6606260296540363,
+      "grad_norm": 0.20508626103401184,
+      "learning_rate": 0.00014902280130293157,
+      "loss": 0.6061,
+      "step": 756
+    },
+    {
+      "epoch": 1.6628226249313562,
+      "grad_norm": 0.3084701597690582,
+      "learning_rate": 0.0001487785016286645,
+      "loss": 0.7018,
+      "step": 757
+    },
+    {
+      "epoch": 1.6650192202086767,
+      "grad_norm": 0.3058534562587738,
+      "learning_rate": 0.00014853420195439738,
+      "loss": 0.6823,
+      "step": 758
+    },
+    {
+      "epoch": 1.6672158154859966,
+      "grad_norm": 0.4641686975955963,
+      "learning_rate": 0.0001482899022801303,
+      "loss": 0.7298,
+      "step": 759
+    },
+    {
+      "epoch": 1.6694124107633168,
+      "grad_norm": 0.5710025429725647,
+      "learning_rate": 0.00014804560260586318,
+      "loss": 0.5641,
+      "step": 760
+    },
+    {
+      "epoch": 1.671609006040637,
+      "grad_norm": 0.2564079761505127,
+      "learning_rate": 0.00014780130293159607,
+      "loss": 0.6281,
+      "step": 761
+    },
+    {
+      "epoch": 1.6738056013179572,
+      "grad_norm": 0.2962080240249634,
+      "learning_rate": 0.000147557003257329,
+      "loss": 0.6396,
+      "step": 762
+    },
+    {
+      "epoch": 1.6760021965952774,
+      "grad_norm": 0.2927751839160919,
+      "learning_rate": 0.00014731270358306188,
+      "loss": 0.6383,
+      "step": 763
+    },
+    {
+      "epoch": 1.6781987918725974,
+      "grad_norm": 0.23479032516479492,
+      "learning_rate": 0.00014706840390879477,
+      "loss": 0.7043,
+      "step": 764
+    },
+    {
+      "epoch": 1.6803953871499178,
+      "grad_norm": 0.3385325074195862,
+      "learning_rate": 0.00014682410423452768,
+      "loss": 0.6306,
+      "step": 765
+    },
+    {
+      "epoch": 1.6825919824272377,
+      "grad_norm": 0.30713969469070435,
+      "learning_rate": 0.00014657980456026057,
+      "loss": 0.5617,
+      "step": 766
+    },
+    {
+      "epoch": 1.684788577704558,
+      "grad_norm": 0.3150871992111206,
+      "learning_rate": 0.00014633550488599346,
+      "loss": 0.6326,
+      "step": 767
+    },
+    {
+      "epoch": 1.6869851729818781,
+      "grad_norm": 0.2927306592464447,
+      "learning_rate": 0.00014609120521172638,
+      "loss": 0.5461,
+      "step": 768
+    },
+    {
+      "epoch": 1.689181768259198,
+      "grad_norm": 0.2745507061481476,
+      "learning_rate": 0.00014584690553745927,
+      "loss": 0.6249,
+      "step": 769
+    },
+    {
+      "epoch": 1.6913783635365185,
+      "grad_norm": 0.264901340007782,
+      "learning_rate": 0.00014560260586319216,
+      "loss": 0.6665,
+      "step": 770
+    },
+    {
+      "epoch": 1.6935749588138385,
+      "grad_norm": 0.3595081865787506,
+      "learning_rate": 0.00014535830618892507,
+      "loss": 0.7573,
+      "step": 771
+    },
+    {
+      "epoch": 1.6957715540911587,
+      "grad_norm": 0.28307056427001953,
+      "learning_rate": 0.00014511400651465796,
+      "loss": 0.6853,
+      "step": 772
+    },
+    {
+      "epoch": 1.6979681493684788,
+      "grad_norm": 0.26563799381256104,
+      "learning_rate": 0.00014486970684039088,
+      "loss": 0.5657,
+      "step": 773
+    },
+    {
+      "epoch": 1.700164744645799,
+      "grad_norm": 0.24004624783992767,
+      "learning_rate": 0.00014462540716612377,
+      "loss": 0.5179,
+      "step": 774
+    },
+    {
+      "epoch": 1.7023613399231192,
+      "grad_norm": 0.3450486660003662,
+      "learning_rate": 0.00014438110749185666,
+      "loss": 0.64,
+      "step": 775
+    },
+    {
+      "epoch": 1.7045579352004392,
+      "grad_norm": 0.2740166187286377,
+      "learning_rate": 0.00014413680781758957,
+      "loss": 0.4319,
+      "step": 776
+    },
+    {
+      "epoch": 1.7067545304777596,
+      "grad_norm": 0.34098851680755615,
+      "learning_rate": 0.00014389250814332246,
+      "loss": 0.7096,
+      "step": 777
+    },
+    {
+      "epoch": 1.7089511257550796,
+      "grad_norm": 0.32044702768325806,
+      "learning_rate": 0.00014364820846905535,
+      "loss": 0.6824,
+      "step": 778
+    },
+    {
+      "epoch": 1.7111477210323998,
+      "grad_norm": 0.30336469411849976,
+      "learning_rate": 0.00014340390879478827,
+      "loss": 0.5951,
+      "step": 779
+    },
+    {
+      "epoch": 1.71334431630972,
+      "grad_norm": 0.2571478486061096,
+      "learning_rate": 0.00014315960912052116,
+      "loss": 0.64,
+      "step": 780
+    },
+    {
+      "epoch": 1.7155409115870401,
+      "grad_norm": 0.3304470479488373,
+      "learning_rate": 0.00014291530944625404,
+      "loss": 0.6097,
+      "step": 781
+    },
+    {
+      "epoch": 1.7177375068643603,
+      "grad_norm": 0.3318883180618286,
+      "learning_rate": 0.00014267100977198696,
+      "loss": 0.5403,
+      "step": 782
+    },
+    {
+      "epoch": 1.7199341021416803,
+      "grad_norm": 0.27189263701438904,
+      "learning_rate": 0.00014242671009771985,
+      "loss": 0.6271,
+      "step": 783
+    },
+    {
+      "epoch": 1.7221306974190007,
+      "grad_norm": 0.2758796513080597,
+      "learning_rate": 0.00014218241042345274,
+      "loss": 0.5621,
+      "step": 784
+    },
+    {
+      "epoch": 1.7243272926963207,
+      "grad_norm": 0.3286420404911041,
+      "learning_rate": 0.00014193811074918566,
+      "loss": 0.8403,
+      "step": 785
+    },
+    {
+      "epoch": 1.7265238879736409,
+      "grad_norm": 0.2251243144273758,
+      "learning_rate": 0.00014169381107491854,
+      "loss": 0.532,
+      "step": 786
+    },
+    {
+      "epoch": 1.728720483250961,
+      "grad_norm": 0.3269249200820923,
+      "learning_rate": 0.00014144951140065146,
+      "loss": 0.5197,
+      "step": 787
+    },
+    {
+      "epoch": 1.730917078528281,
+      "grad_norm": 0.24294660985469818,
+      "learning_rate": 0.00014120521172638435,
+      "loss": 0.6162,
+      "step": 788
+    },
+    {
+      "epoch": 1.7331136738056014,
+      "grad_norm": 0.24577254056930542,
+      "learning_rate": 0.00014096091205211724,
+      "loss": 0.6488,
+      "step": 789
+    },
+    {
+      "epoch": 1.7353102690829214,
+      "grad_norm": 0.27060091495513916,
+      "learning_rate": 0.00014071661237785016,
+      "loss": 0.7942,
+      "step": 790
+    },
+    {
+      "epoch": 1.7375068643602416,
+      "grad_norm": 0.39916133880615234,
+      "learning_rate": 0.00014047231270358305,
+      "loss": 0.5826,
+      "step": 791
+    },
+    {
+      "epoch": 1.7397034596375618,
+      "grad_norm": 0.2174745351076126,
+      "learning_rate": 0.00014022801302931593,
+      "loss": 0.582,
+      "step": 792
+    },
+    {
+      "epoch": 1.741900054914882,
+      "grad_norm": 0.3330744504928589,
+      "learning_rate": 0.00013998371335504885,
+      "loss": 0.5524,
+      "step": 793
+    },
+    {
+      "epoch": 1.7440966501922022,
+      "grad_norm": 0.33084964752197266,
+      "learning_rate": 0.00013973941368078174,
+      "loss": 0.7174,
+      "step": 794
+    },
+    {
+      "epoch": 1.7462932454695221,
+      "grad_norm": 0.22110159695148468,
+      "learning_rate": 0.00013949511400651463,
+      "loss": 0.5762,
+      "step": 795
+    },
+    {
+      "epoch": 1.7484898407468425,
+      "grad_norm": 1.185044765472412,
+      "learning_rate": 0.00013925081433224755,
+      "loss": 0.5447,
+      "step": 796
+    },
+    {
+      "epoch": 1.7506864360241625,
+      "grad_norm": 0.22130067646503448,
+      "learning_rate": 0.00013900651465798043,
+      "loss": 0.6019,
+      "step": 797
+    },
+    {
+      "epoch": 1.7528830313014827,
+      "grad_norm": 0.27342772483825684,
+      "learning_rate": 0.00013876221498371335,
+      "loss": 0.634,
+      "step": 798
+    },
+    {
+      "epoch": 1.755079626578803,
+      "grad_norm": 0.46273374557495117,
+      "learning_rate": 0.00013851791530944624,
+      "loss": 0.692,
+      "step": 799
+    },
+    {
+      "epoch": 1.7572762218561229,
+      "grad_norm": 0.31564822793006897,
+      "learning_rate": 0.00013827361563517913,
+      "loss": 0.8289,
+      "step": 800
+    },
+    {
+      "epoch": 1.7594728171334433,
+      "grad_norm": 0.33761343359947205,
+      "learning_rate": 0.00013802931596091205,
+      "loss": 0.7512,
+      "step": 801
+    },
+    {
+      "epoch": 1.7616694124107632,
+      "grad_norm": 0.34530550241470337,
+      "learning_rate": 0.00013778501628664493,
+      "loss": 0.7999,
+      "step": 802
+    },
+    {
+      "epoch": 1.7638660076880834,
+      "grad_norm": 0.5441341996192932,
+      "learning_rate": 0.00013754071661237782,
+      "loss": 0.8995,
+      "step": 803
+    },
+    {
+      "epoch": 1.7660626029654036,
+      "grad_norm": 0.37606433033943176,
+      "learning_rate": 0.00013729641693811074,
+      "loss": 0.8042,
+      "step": 804
+    },
+    {
+      "epoch": 1.7682591982427238,
+      "grad_norm": 0.31094858050346375,
+      "learning_rate": 0.00013705211726384363,
+      "loss": 0.639,
+      "step": 805
+    },
+    {
+      "epoch": 1.770455793520044,
+      "grad_norm": 0.31574147939682007,
+      "learning_rate": 0.00013680781758957652,
+      "loss": 0.5852,
+      "step": 806
+    },
+    {
+      "epoch": 1.772652388797364,
+      "grad_norm": 0.2851109802722931,
+      "learning_rate": 0.00013656351791530943,
+      "loss": 0.6395,
+      "step": 807
+    },
+    {
+      "epoch": 1.7748489840746844,
+      "grad_norm": 0.41837596893310547,
+      "learning_rate": 0.00013631921824104232,
+      "loss": 0.6685,
+      "step": 808
+    },
+    {
+      "epoch": 1.7770455793520044,
+      "grad_norm": 0.3092561960220337,
+      "learning_rate": 0.0001360749185667752,
+      "loss": 0.7168,
+      "step": 809
+    },
+    {
+      "epoch": 1.7792421746293245,
+      "grad_norm": 0.2859450876712799,
+      "learning_rate": 0.00013583061889250813,
+      "loss": 0.4274,
+      "step": 810
+    },
+    {
+      "epoch": 1.7814387699066447,
+      "grad_norm": 0.31098270416259766,
+      "learning_rate": 0.00013558631921824102,
+      "loss": 0.4916,
+      "step": 811
+    },
+    {
+      "epoch": 1.7836353651839647,
+      "grad_norm": 0.31832170486450195,
+      "learning_rate": 0.00013534201954397394,
+      "loss": 0.5875,
+      "step": 812
+    },
+    {
+      "epoch": 1.7858319604612851,
+      "grad_norm": 0.39608514308929443,
+      "learning_rate": 0.00013509771986970682,
+      "loss": 0.7692,
+      "step": 813
+    },
+    {
+      "epoch": 1.788028555738605,
+      "grad_norm": 0.30884218215942383,
+      "learning_rate": 0.00013485342019543971,
+      "loss": 0.649,
+      "step": 814
+    },
+    {
+      "epoch": 1.7902251510159253,
+      "grad_norm": 0.3142588138580322,
+      "learning_rate": 0.00013460912052117263,
+      "loss": 0.6549,
+      "step": 815
+    },
+    {
+      "epoch": 1.7924217462932455,
+      "grad_norm": 0.2957165539264679,
+      "learning_rate": 0.00013436482084690552,
+      "loss": 0.6646,
+      "step": 816
+    },
+    {
+      "epoch": 1.7946183415705657,
+      "grad_norm": 0.4314785599708557,
+      "learning_rate": 0.0001341205211726384,
+      "loss": 0.5389,
+      "step": 817
+    },
+    {
+      "epoch": 1.7968149368478858,
+      "grad_norm": 0.32042449712753296,
+      "learning_rate": 0.00013387622149837132,
+      "loss": 0.6247,
+      "step": 818
+    },
+    {
+      "epoch": 1.7990115321252058,
+      "grad_norm": 0.29641464352607727,
+      "learning_rate": 0.00013363192182410421,
+      "loss": 0.5955,
+      "step": 819
+    },
+    {
+      "epoch": 1.8012081274025262,
+      "grad_norm": 0.29636019468307495,
+      "learning_rate": 0.0001333876221498371,
+      "loss": 0.652,
+      "step": 820
+    },
+    {
+      "epoch": 1.8034047226798462,
+      "grad_norm": 0.2582562267780304,
+      "learning_rate": 0.00013314332247557002,
+      "loss": 0.6947,
+      "step": 821
+    },
+    {
+      "epoch": 1.8056013179571664,
+      "grad_norm": 0.35147929191589355,
+      "learning_rate": 0.0001328990228013029,
+      "loss": 0.7378,
+      "step": 822
+    },
+    {
+      "epoch": 1.8077979132344866,
+      "grad_norm": 0.2972518801689148,
+      "learning_rate": 0.00013265472312703582,
+      "loss": 0.6612,
+      "step": 823
+    },
+    {
+      "epoch": 1.8099945085118065,
+      "grad_norm": 0.29565125703811646,
+      "learning_rate": 0.00013241042345276871,
+      "loss": 0.5869,
+      "step": 824
+    },
+    {
+      "epoch": 1.812191103789127,
+      "grad_norm": 0.3595902621746063,
+      "learning_rate": 0.0001321661237785016,
+      "loss": 0.5728,
+      "step": 825
+    },
+    {
+      "epoch": 1.814387699066447,
+      "grad_norm": 0.3137751519680023,
+      "learning_rate": 0.00013192182410423452,
+      "loss": 0.4421,
+      "step": 826
+    },
+    {
+      "epoch": 1.8165842943437671,
+      "grad_norm": 0.33113762736320496,
+      "learning_rate": 0.0001316775244299674,
+      "loss": 0.6949,
+      "step": 827
+    },
+    {
+      "epoch": 1.8187808896210873,
+      "grad_norm": 0.24384188652038574,
+      "learning_rate": 0.0001314332247557003,
+      "loss": 0.5737,
+      "step": 828
+    },
+    {
+      "epoch": 1.8209774848984075,
+      "grad_norm": 0.3388102352619171,
+      "learning_rate": 0.00013118892508143321,
+      "loss": 0.5326,
+      "step": 829
+    },
+    {
+      "epoch": 1.8231740801757277,
+      "grad_norm": 0.38650086522102356,
+      "learning_rate": 0.00013094462540716613,
+      "loss": 0.471,
+      "step": 830
+    },
+    {
+      "epoch": 1.8253706754530477,
+      "grad_norm": 0.2583390474319458,
+      "learning_rate": 0.00013070032573289902,
+      "loss": 0.4215,
+      "step": 831
+    },
+    {
+      "epoch": 1.827567270730368,
+      "grad_norm": 0.2933657467365265,
+      "learning_rate": 0.0001304560260586319,
+      "loss": 0.6442,
+      "step": 832
+    },
+    {
+      "epoch": 1.829763866007688,
+      "grad_norm": 0.28765028715133667,
+      "learning_rate": 0.00013021172638436483,
+      "loss": 0.6617,
+      "step": 833
+    },
+    {
+      "epoch": 1.8319604612850082,
+      "grad_norm": 0.27898189425468445,
+      "learning_rate": 0.00012996742671009771,
+      "loss": 0.7256,
+      "step": 834
+    },
+    {
+      "epoch": 1.8341570565623284,
+      "grad_norm": 0.2909148335456848,
+      "learning_rate": 0.0001297231270358306,
+      "loss": 0.7148,
+      "step": 835
+    },
+    {
+      "epoch": 1.8363536518396484,
+      "grad_norm": 0.3123301863670349,
+      "learning_rate": 0.00012947882736156352,
+      "loss": 0.4878,
+      "step": 836
+    },
+    {
+      "epoch": 1.8385502471169688,
+      "grad_norm": 0.31307247281074524,
+      "learning_rate": 0.0001292345276872964,
+      "loss": 0.5883,
+      "step": 837
+    },
+    {
+      "epoch": 1.8407468423942888,
+      "grad_norm": 0.3377425968647003,
+      "learning_rate": 0.00012899022801302933,
+      "loss": 0.574,
+      "step": 838
+    },
+    {
+      "epoch": 1.842943437671609,
+      "grad_norm": 0.2906651496887207,
+      "learning_rate": 0.00012874592833876221,
+      "loss": 0.5223,
+      "step": 839
+    },
+    {
+      "epoch": 1.8451400329489291,
+      "grad_norm": 0.4251576364040375,
+      "learning_rate": 0.0001285016286644951,
+      "loss": 0.6401,
+      "step": 840
+    },
+    {
+      "epoch": 1.8473366282262493,
+      "grad_norm": 0.26640015840530396,
+      "learning_rate": 0.00012825732899022802,
+      "loss": 0.6023,
+      "step": 841
+    },
+    {
+      "epoch": 1.8495332235035695,
+      "grad_norm": 0.4165840148925781,
+      "learning_rate": 0.0001280130293159609,
+      "loss": 0.5831,
+      "step": 842
+    },
+    {
+      "epoch": 1.8517298187808895,
+      "grad_norm": 0.3378630578517914,
+      "learning_rate": 0.0001277687296416938,
+      "loss": 0.802,
+      "step": 843
+    },
+    {
+      "epoch": 1.85392641405821,
+      "grad_norm": 0.32427728176116943,
+      "learning_rate": 0.00012752442996742671,
+      "loss": 0.6531,
+      "step": 844
+    },
+    {
+      "epoch": 1.8561230093355299,
+      "grad_norm": 0.3648292124271393,
+      "learning_rate": 0.0001272801302931596,
+      "loss": 0.6439,
+      "step": 845
+    },
+    {
+      "epoch": 1.85831960461285,
+      "grad_norm": 0.2850014269351959,
+      "learning_rate": 0.0001270358306188925,
+      "loss": 0.538,
+      "step": 846
+    },
+    {
+      "epoch": 1.8605161998901703,
+      "grad_norm": 0.3368457555770874,
+      "learning_rate": 0.0001267915309446254,
+      "loss": 0.7405,
+      "step": 847
+    },
+    {
+      "epoch": 1.8627127951674904,
+      "grad_norm": 0.28684937953948975,
+      "learning_rate": 0.0001265472312703583,
+      "loss": 0.5623,
+      "step": 848
+    },
+    {
+      "epoch": 1.8649093904448106,
+      "grad_norm": 0.2165999710559845,
+      "learning_rate": 0.0001263029315960912,
+      "loss": 0.5335,
+      "step": 849
+    },
+    {
+      "epoch": 1.8671059857221306,
+      "grad_norm": 0.22919365763664246,
+      "learning_rate": 0.0001260586319218241,
+      "loss": 0.5844,
+      "step": 850
+    },
+    {
+      "epoch": 1.869302580999451,
+      "grad_norm": 0.3021653890609741,
+      "learning_rate": 0.000125814332247557,
+      "loss": 0.4969,
+      "step": 851
+    },
+    {
+      "epoch": 1.871499176276771,
+      "grad_norm": 0.2282877117395401,
+      "learning_rate": 0.0001255700325732899,
+      "loss": 0.6438,
+      "step": 852
+    },
+    {
+      "epoch": 1.8736957715540912,
+      "grad_norm": 0.3082697093486786,
+      "learning_rate": 0.0001253257328990228,
+      "loss": 0.6594,
+      "step": 853
+    },
+    {
+      "epoch": 1.8758923668314114,
+      "grad_norm": 0.3801448345184326,
+      "learning_rate": 0.0001250814332247557,
+      "loss": 0.7952,
+      "step": 854
+    },
+    {
+      "epoch": 1.8780889621087313,
+      "grad_norm": 0.23287056386470795,
+      "learning_rate": 0.0001248371335504886,
+      "loss": 0.5176,
+      "step": 855
+    },
+    {
+      "epoch": 1.8802855573860517,
+      "grad_norm": 0.3359370231628418,
+      "learning_rate": 0.0001245928338762215,
+      "loss": 0.7629,
+      "step": 856
+    },
+    {
+      "epoch": 1.8824821526633717,
+      "grad_norm": 0.3249419331550598,
+      "learning_rate": 0.00012434853420195438,
+      "loss": 0.6017,
+      "step": 857
+    },
+    {
+      "epoch": 1.884678747940692,
+      "grad_norm": 0.31045034527778625,
+      "learning_rate": 0.0001241042345276873,
+      "loss": 0.6812,
+      "step": 858
+    },
+    {
+      "epoch": 1.886875343218012,
+      "grad_norm": 0.2801130414009094,
+      "learning_rate": 0.0001238599348534202,
+      "loss": 0.5598,
+      "step": 859
+    },
+    {
+      "epoch": 1.8890719384953323,
+      "grad_norm": 0.22003044188022614,
+      "learning_rate": 0.00012361563517915308,
+      "loss": 0.5745,
+      "step": 860
+    },
+    {
+      "epoch": 1.8912685337726525,
+      "grad_norm": 0.34684598445892334,
+      "learning_rate": 0.000123371335504886,
+      "loss": 0.7818,
+      "step": 861
+    },
+    {
+      "epoch": 1.8934651290499724,
+      "grad_norm": 0.28532010316848755,
+      "learning_rate": 0.00012312703583061888,
+      "loss": 0.6137,
+      "step": 862
+    },
+    {
+      "epoch": 1.8956617243272929,
+      "grad_norm": 0.31024038791656494,
+      "learning_rate": 0.00012288273615635177,
+      "loss": 0.817,
+      "step": 863
+    },
+    {
+      "epoch": 1.8978583196046128,
+      "grad_norm": 0.27985459566116333,
+      "learning_rate": 0.0001226384364820847,
+      "loss": 0.531,
+      "step": 864
+    },
+    {
+      "epoch": 1.900054914881933,
+      "grad_norm": 0.3009069859981537,
+      "learning_rate": 0.00012239413680781758,
+      "loss": 0.6864,
+      "step": 865
+    },
+    {
+      "epoch": 1.9022515101592532,
+      "grad_norm": 0.32191112637519836,
+      "learning_rate": 0.0001221498371335505,
+      "loss": 0.5852,
+      "step": 866
+    },
+    {
+      "epoch": 1.9044481054365732,
+      "grad_norm": 0.33307084441185,
+      "learning_rate": 0.00012190553745928338,
+      "loss": 0.5695,
+      "step": 867
+    },
+    {
+      "epoch": 1.9066447007138936,
+      "grad_norm": 0.2647794783115387,
+      "learning_rate": 0.00012166123778501629,
+      "loss": 0.7818,
+      "step": 868
+    },
+    {
+      "epoch": 1.9088412959912135,
+      "grad_norm": 0.29898229241371155,
+      "learning_rate": 0.00012141693811074918,
+      "loss": 0.7345,
+      "step": 869
+    },
+    {
+      "epoch": 1.9110378912685337,
+      "grad_norm": 0.36632969975471497,
+      "learning_rate": 0.00012117263843648208,
+      "loss": 0.5927,
+      "step": 870
+    },
+    {
+      "epoch": 1.913234486545854,
+      "grad_norm": 0.3301393687725067,
+      "learning_rate": 0.00012092833876221498,
+      "loss": 0.5956,
+      "step": 871
+    },
+    {
+      "epoch": 1.9154310818231741,
+      "grad_norm": 0.24661529064178467,
+      "learning_rate": 0.00012068403908794787,
+      "loss": 0.6978,
+      "step": 872
+    },
+    {
+      "epoch": 1.9176276771004943,
+      "grad_norm": 0.33883655071258545,
+      "learning_rate": 0.00012043973941368077,
+      "loss": 0.6868,
+      "step": 873
+    },
+    {
+      "epoch": 1.9198242723778143,
+      "grad_norm": 0.24409429728984833,
+      "learning_rate": 0.00012019543973941368,
+      "loss": 0.8246,
+      "step": 874
+    },
+    {
+      "epoch": 1.9220208676551347,
+      "grad_norm": 0.36363038420677185,
+      "learning_rate": 0.00011995114006514658,
+      "loss": 0.7233,
+      "step": 875
+    },
+    {
+      "epoch": 1.9242174629324547,
+      "grad_norm": 0.3607602119445801,
+      "learning_rate": 0.00011970684039087947,
+      "loss": 0.7028,
+      "step": 876
+    },
+    {
+      "epoch": 1.9264140582097748,
+      "grad_norm": 0.31431251764297485,
+      "learning_rate": 0.00011946254071661237,
+      "loss": 0.5625,
+      "step": 877
+    },
+    {
+      "epoch": 1.928610653487095,
+      "grad_norm": 0.4127863645553589,
+      "learning_rate": 0.00011921824104234527,
+      "loss": 0.6875,
+      "step": 878
+    },
+    {
+      "epoch": 1.930807248764415,
+      "grad_norm": 0.2782525420188904,
+      "learning_rate": 0.00011897394136807816,
+      "loss": 0.6824,
+      "step": 879
+    },
+    {
+      "epoch": 1.9330038440417354,
+      "grad_norm": 0.33698660135269165,
+      "learning_rate": 0.00011872964169381106,
+      "loss": 0.5582,
+      "step": 880
+    },
+    {
+      "epoch": 1.9352004393190554,
+      "grad_norm": 0.33458802103996277,
+      "learning_rate": 0.00011848534201954397,
+      "loss": 0.6698,
+      "step": 881
+    },
+    {
+      "epoch": 1.9373970345963756,
+      "grad_norm": 0.23767663538455963,
+      "learning_rate": 0.00011824104234527687,
+      "loss": 0.4841,
+      "step": 882
+    },
+    {
+      "epoch": 1.9395936298736958,
+      "grad_norm": 0.40156811475753784,
+      "learning_rate": 0.00011799674267100976,
+      "loss": 0.7339,
+      "step": 883
+    },
+    {
+      "epoch": 1.941790225151016,
+      "grad_norm": 0.36377251148223877,
+      "learning_rate": 0.00011775244299674266,
+      "loss": 0.6647,
+      "step": 884
+    },
+    {
+      "epoch": 1.9439868204283361,
+      "grad_norm": 0.2833596169948578,
+      "learning_rate": 0.00011750814332247557,
+      "loss": 0.9841,
+      "step": 885
+    },
+    {
+      "epoch": 1.9461834157056561,
+      "grad_norm": 0.3663980960845947,
+      "learning_rate": 0.00011726384364820845,
+      "loss": 0.6655,
+      "step": 886
+    },
+    {
+      "epoch": 1.9483800109829765,
+      "grad_norm": 0.20633137226104736,
+      "learning_rate": 0.00011701954397394136,
+      "loss": 0.5684,
+      "step": 887
+    },
+    {
+      "epoch": 1.9505766062602965,
+      "grad_norm": 0.297671377658844,
+      "learning_rate": 0.00011677524429967426,
+      "loss": 0.5663,
+      "step": 888
+    },
+    {
+      "epoch": 1.9527732015376167,
+      "grad_norm": 0.3512963354587555,
+      "learning_rate": 0.00011653094462540716,
+      "loss": 0.6862,
+      "step": 889
+    },
+    {
+      "epoch": 1.9549697968149369,
+      "grad_norm": 0.30056527256965637,
+      "learning_rate": 0.00011628664495114005,
+      "loss": 0.6111,
+      "step": 890
+    },
+    {
+      "epoch": 1.9571663920922568,
+      "grad_norm": 0.2516704797744751,
+      "learning_rate": 0.00011604234527687295,
+      "loss": 0.7549,
+      "step": 891
+    },
+    {
+      "epoch": 1.9593629873695773,
+      "grad_norm": 0.2964627146720886,
+      "learning_rate": 0.00011579804560260586,
+      "loss": 0.5529,
+      "step": 892
+    },
+    {
+      "epoch": 1.9615595826468972,
+      "grad_norm": 0.31984639167785645,
+      "learning_rate": 0.00011555374592833876,
+      "loss": 0.6185,
+      "step": 893
+    },
+    {
+      "epoch": 1.9637561779242174,
+      "grad_norm": 0.3613375127315521,
+      "learning_rate": 0.00011530944625407165,
+      "loss": 0.6803,
+      "step": 894
+    },
+    {
+      "epoch": 1.9659527732015376,
+      "grad_norm": 0.29901495575904846,
+      "learning_rate": 0.00011506514657980455,
+      "loss": 0.4784,
+      "step": 895
+    },
+    {
+      "epoch": 1.9681493684788578,
+      "grad_norm": 0.2912847101688385,
+      "learning_rate": 0.00011482084690553745,
+      "loss": 0.7561,
+      "step": 896
+    },
+    {
+      "epoch": 1.970345963756178,
+      "grad_norm": 0.20477470755577087,
+      "learning_rate": 0.00011457654723127034,
+      "loss": 0.5536,
+      "step": 897
+    },
+    {
+      "epoch": 1.972542559033498,
+      "grad_norm": 0.2915339469909668,
+      "learning_rate": 0.00011433224755700325,
+      "loss": 0.7609,
+      "step": 898
+    },
+    {
+      "epoch": 1.9747391543108184,
+      "grad_norm": 0.37481367588043213,
+      "learning_rate": 0.00011408794788273615,
+      "loss": 0.7014,
+      "step": 899
+    },
+    {
+      "epoch": 1.9769357495881383,
+      "grad_norm": 0.27598080039024353,
+      "learning_rate": 0.00011384364820846905,
+      "loss": 0.5314,
+      "step": 900
+    },
+    {
+      "epoch": 1.9791323448654585,
+      "grad_norm": 0.23892323672771454,
+      "learning_rate": 0.00011359934853420194,
+      "loss": 0.6736,
+      "step": 901
+    },
+    {
+      "epoch": 1.9813289401427787,
+      "grad_norm": 0.2880638539791107,
+      "learning_rate": 0.00011335504885993484,
+      "loss": 0.7233,
+      "step": 902
+    },
+    {
+      "epoch": 1.9835255354200987,
+      "grad_norm": 0.31268733739852905,
+      "learning_rate": 0.00011311074918566775,
+      "loss": 0.6619,
+      "step": 903
+    },
+    {
+      "epoch": 1.985722130697419,
+      "grad_norm": 0.2768239676952362,
+      "learning_rate": 0.00011286644951140064,
+      "loss": 0.5512,
+      "step": 904
+    },
+    {
+      "epoch": 1.987918725974739,
+      "grad_norm": 0.31970319151878357,
+      "learning_rate": 0.00011262214983713354,
+      "loss": 0.5264,
+      "step": 905
+    },
+    {
+      "epoch": 1.9901153212520593,
+      "grad_norm": 0.2745014727115631,
+      "learning_rate": 0.00011237785016286644,
+      "loss": 0.7418,
+      "step": 906
+    },
+    {
+      "epoch": 1.9923119165293794,
+      "grad_norm": 0.3232331871986389,
+      "learning_rate": 0.00011213355048859934,
+      "loss": 0.482,
+      "step": 907
+    },
+    {
+      "epoch": 1.9945085118066996,
+      "grad_norm": 0.22175395488739014,
+      "learning_rate": 0.00011188925081433223,
+      "loss": 0.5237,
+      "step": 908
+    },
+    {
+      "epoch": 1.9967051070840198,
+      "grad_norm": 0.2541004419326782,
+      "learning_rate": 0.00011164495114006514,
+      "loss": 0.4611,
+      "step": 909
+    },
+    {
+      "epoch": 1.9989017023613398,
+      "grad_norm": 0.683760941028595,
+      "learning_rate": 0.00011140065146579804,
+      "loss": 0.5701,
+      "step": 910
+    },
+    {
+      "epoch": 2.00109829763866,
+      "grad_norm": 0.29126739501953125,
+      "learning_rate": 0.00011115635179153093,
+      "loss": 0.7327,
+      "step": 911
+    },
+    {
+      "epoch": 2.00329489291598,
+      "grad_norm": 0.2815655767917633,
+      "learning_rate": 0.00011091205211726383,
+      "loss": 0.6164,
+      "step": 912
+    },
+    {
+      "epoch": 2.0054914881933006,
+      "grad_norm": 0.3038989305496216,
+      "learning_rate": 0.00011066775244299673,
+      "loss": 0.5305,
+      "step": 913
+    },
+    {
+      "epoch": 2.0076880834706206,
+      "grad_norm": 0.30940869450569153,
+      "learning_rate": 0.00011042345276872964,
+      "loss": 0.609,
+      "step": 914
+    },
+    {
+      "epoch": 2.0098846787479405,
+      "grad_norm": 0.4073452055454254,
+      "learning_rate": 0.00011017915309446253,
+      "loss": 0.5682,
+      "step": 915
+    },
+    {
+      "epoch": 2.012081274025261,
+      "grad_norm": 0.31348660588264465,
+      "learning_rate": 0.00010993485342019543,
+      "loss": 0.7675,
+      "step": 916
+    },
+    {
+      "epoch": 2.014277869302581,
+      "grad_norm": 0.306452214717865,
+      "learning_rate": 0.00010969055374592833,
+      "loss": 0.7985,
+      "step": 917
+    },
+    {
+      "epoch": 2.0164744645799013,
+      "grad_norm": 0.2873900234699249,
+      "learning_rate": 0.00010944625407166122,
+      "loss": 0.7201,
+      "step": 918
+    },
+    {
+      "epoch": 2.0186710598572213,
+      "grad_norm": 0.2569607198238373,
+      "learning_rate": 0.00010920195439739412,
+      "loss": 0.572,
+      "step": 919
+    },
+    {
+      "epoch": 2.0208676551345413,
+      "grad_norm": 0.5690404176712036,
+      "learning_rate": 0.00010895765472312703,
+      "loss": 0.5876,
+      "step": 920
+    },
+    {
+      "epoch": 2.0230642504118617,
+      "grad_norm": 0.2954144775867462,
+      "learning_rate": 0.00010871335504885993,
+      "loss": 0.5312,
+      "step": 921
+    },
+    {
+      "epoch": 2.0252608456891816,
+      "grad_norm": 0.2523924708366394,
+      "learning_rate": 0.00010846905537459282,
+      "loss": 0.6433,
+      "step": 922
+    },
+    {
+      "epoch": 2.027457440966502,
+      "grad_norm": 0.285146564245224,
+      "learning_rate": 0.00010822475570032572,
+      "loss": 0.5501,
+      "step": 923
+    },
+    {
+      "epoch": 2.029654036243822,
+      "grad_norm": 0.3737488090991974,
+      "learning_rate": 0.00010798045602605862,
+      "loss": 0.7172,
+      "step": 924
+    },
+    {
+      "epoch": 2.0318506315211424,
+      "grad_norm": 0.3101334273815155,
+      "learning_rate": 0.00010773615635179153,
+      "loss": 0.5891,
+      "step": 925
+    },
+    {
+      "epoch": 2.0340472267984624,
+      "grad_norm": 0.2619282603263855,
+      "learning_rate": 0.00010749185667752442,
+      "loss": 0.5975,
+      "step": 926
+    },
+    {
+      "epoch": 2.0362438220757824,
+      "grad_norm": 0.32030749320983887,
+      "learning_rate": 0.00010724755700325732,
+      "loss": 0.7491,
+      "step": 927
+    },
+    {
+      "epoch": 2.0384404173531028,
+      "grad_norm": 0.23549683392047882,
+      "learning_rate": 0.00010700325732899022,
+      "loss": 0.6251,
+      "step": 928
+    },
+    {
+      "epoch": 2.0406370126304227,
+      "grad_norm": 0.2145562618970871,
+      "learning_rate": 0.00010675895765472311,
+      "loss": 0.5365,
+      "step": 929
+    },
+    {
+      "epoch": 2.042833607907743,
+      "grad_norm": 0.3381960690021515,
+      "learning_rate": 0.00010651465798045601,
+      "loss": 0.5167,
+      "step": 930
+    },
+    {
+      "epoch": 2.045030203185063,
+      "grad_norm": 0.2788982093334198,
+      "learning_rate": 0.00010627035830618892,
+      "loss": 0.5692,
+      "step": 931
+    },
+    {
+      "epoch": 2.047226798462383,
+      "grad_norm": 0.35798248648643494,
+      "learning_rate": 0.00010602605863192182,
+      "loss": 0.6022,
+      "step": 932
+    },
+    {
+      "epoch": 2.0494233937397035,
+      "grad_norm": 0.30149203538894653,
+      "learning_rate": 0.00010578175895765471,
+      "loss": 0.644,
+      "step": 933
+    },
+    {
+      "epoch": 2.0516199890170235,
+      "grad_norm": 0.2656877040863037,
+      "learning_rate": 0.00010553745928338761,
+      "loss": 0.5175,
+      "step": 934
+    },
+    {
+      "epoch": 2.053816584294344,
+      "grad_norm": 0.22816424071788788,
+      "learning_rate": 0.00010529315960912051,
+      "loss": 0.5573,
+      "step": 935
+    },
+    {
+      "epoch": 2.056013179571664,
+      "grad_norm": 0.2823135554790497,
+      "learning_rate": 0.0001050488599348534,
+      "loss": 0.7369,
+      "step": 936
+    },
+    {
+      "epoch": 2.0582097748489843,
+      "grad_norm": 0.26900506019592285,
+      "learning_rate": 0.0001048045602605863,
+      "loss": 0.6803,
+      "step": 937
+    },
+    {
+      "epoch": 2.0604063701263042,
+      "grad_norm": 0.24996565282344818,
+      "learning_rate": 0.00010456026058631921,
+      "loss": 0.4292,
+      "step": 938
+    },
+    {
+      "epoch": 2.062602965403624,
+      "grad_norm": 0.366100013256073,
+      "learning_rate": 0.00010431596091205211,
+      "loss": 0.7071,
+      "step": 939
+    },
+    {
+      "epoch": 2.0647995606809446,
+      "grad_norm": 0.2762909233570099,
+      "learning_rate": 0.000104071661237785,
+      "loss": 0.5816,
+      "step": 940
+    },
+    {
+      "epoch": 2.0669961559582646,
+      "grad_norm": 0.2231253981590271,
+      "learning_rate": 0.0001038273615635179,
+      "loss": 0.6069,
+      "step": 941
+    },
+    {
+      "epoch": 2.069192751235585,
+      "grad_norm": 0.33811599016189575,
+      "learning_rate": 0.0001035830618892508,
+      "loss": 0.4772,
+      "step": 942
+    },
+    {
+      "epoch": 2.071389346512905,
+      "grad_norm": 0.2521122694015503,
+      "learning_rate": 0.0001033387622149837,
+      "loss": 0.644,
+      "step": 943
+    },
+    {
+      "epoch": 2.073585941790225,
+      "grad_norm": 0.42457273602485657,
+      "learning_rate": 0.0001030944625407166,
+      "loss": 0.5237,
+      "step": 944
+    },
+    {
+      "epoch": 2.0757825370675453,
+      "grad_norm": 0.24401000142097473,
+      "learning_rate": 0.0001028501628664495,
+      "loss": 0.6146,
+      "step": 945
+    },
+    {
+      "epoch": 2.0779791323448653,
+      "grad_norm": 0.2809639871120453,
+      "learning_rate": 0.0001026058631921824,
+      "loss": 0.599,
+      "step": 946
+    },
+    {
+      "epoch": 2.0801757276221857,
+      "grad_norm": 0.21833446621894836,
+      "learning_rate": 0.00010236156351791529,
+      "loss": 0.4907,
+      "step": 947
+    },
+    {
+      "epoch": 2.0823723228995057,
+      "grad_norm": 0.35902145504951477,
+      "learning_rate": 0.0001021172638436482,
+      "loss": 0.5084,
+      "step": 948
+    },
+    {
+      "epoch": 2.084568918176826,
+      "grad_norm": 0.3082290291786194,
+      "learning_rate": 0.0001018729641693811,
+      "loss": 0.7394,
+      "step": 949
+    },
+    {
+      "epoch": 2.086765513454146,
+      "grad_norm": 0.2334074079990387,
+      "learning_rate": 0.00010162866449511399,
+      "loss": 0.6297,
+      "step": 950
+    },
+    {
+      "epoch": 2.088962108731466,
+      "grad_norm": 0.2637062072753906,
+      "learning_rate": 0.00010138436482084689,
+      "loss": 0.6305,
+      "step": 951
+    },
+    {
+      "epoch": 2.0911587040087865,
+      "grad_norm": 0.43729156255722046,
+      "learning_rate": 0.00010114006514657979,
+      "loss": 0.6498,
+      "step": 952
+    },
+    {
+      "epoch": 2.0933552992861064,
+      "grad_norm": 0.3261928856372833,
+      "learning_rate": 0.0001008957654723127,
+      "loss": 0.6656,
+      "step": 953
+    },
+    {
+      "epoch": 2.095551894563427,
+      "grad_norm": 0.24635173380374908,
+      "learning_rate": 0.00010065146579804558,
+      "loss": 0.5432,
+      "step": 954
+    },
+    {
+      "epoch": 2.097748489840747,
+      "grad_norm": 0.32413849234580994,
+      "learning_rate": 0.00010040716612377849,
+      "loss": 0.6746,
+      "step": 955
+    },
+    {
+      "epoch": 2.099945085118067,
+      "grad_norm": 0.30958375334739685,
+      "learning_rate": 0.00010016286644951139,
+      "loss": 0.6089,
+      "step": 956
+    },
+    {
+      "epoch": 2.102141680395387,
+      "grad_norm": 0.31989455223083496,
+      "learning_rate": 9.991856677524428e-05,
+      "loss": 0.6244,
+      "step": 957
+    },
+    {
+      "epoch": 2.104338275672707,
+      "grad_norm": 0.27133166790008545,
+      "learning_rate": 9.967426710097718e-05,
+      "loss": 0.567,
+      "step": 958
+    },
+    {
+      "epoch": 2.1065348709500276,
+      "grad_norm": 0.27256739139556885,
+      "learning_rate": 9.942996742671008e-05,
+      "loss": 0.5735,
+      "step": 959
+    },
+    {
+      "epoch": 2.1087314662273475,
+      "grad_norm": 0.2944307029247284,
+      "learning_rate": 9.918566775244299e-05,
+      "loss": 0.5817,
+      "step": 960
+    },
+    {
+      "epoch": 2.110928061504668,
+      "grad_norm": 0.37318965792655945,
+      "learning_rate": 9.894136807817588e-05,
+      "loss": 0.7223,
+      "step": 961
+    },
+    {
+      "epoch": 2.113124656781988,
+      "grad_norm": 0.3693804144859314,
+      "learning_rate": 9.869706840390878e-05,
+      "loss": 0.5729,
+      "step": 962
+    },
+    {
+      "epoch": 2.115321252059308,
+      "grad_norm": 0.31479042768478394,
+      "learning_rate": 9.845276872964168e-05,
+      "loss": 0.4292,
+      "step": 963
+    },
+    {
+      "epoch": 2.1175178473366283,
+      "grad_norm": 0.26124605536460876,
+      "learning_rate": 9.820846905537458e-05,
+      "loss": 0.5815,
+      "step": 964
+    },
+    {
+      "epoch": 2.1197144426139483,
+      "grad_norm": 0.26337531208992004,
+      "learning_rate": 9.796416938110747e-05,
+      "loss": 0.552,
+      "step": 965
+    },
+    {
+      "epoch": 2.1219110378912687,
+      "grad_norm": 0.44966036081314087,
+      "learning_rate": 9.771986970684038e-05,
+      "loss": 0.5577,
+      "step": 966
+    },
+    {
+      "epoch": 2.1241076331685886,
+      "grad_norm": 0.31915444135665894,
+      "learning_rate": 9.747557003257328e-05,
+      "loss": 0.606,
+      "step": 967
+    },
+    {
+      "epoch": 2.1263042284459086,
+      "grad_norm": 0.24085840582847595,
+      "learning_rate": 9.723127035830617e-05,
+      "loss": 0.5569,
+      "step": 968
+    },
+    {
+      "epoch": 2.128500823723229,
+      "grad_norm": 0.2575514018535614,
+      "learning_rate": 9.698697068403907e-05,
+      "loss": 0.5126,
+      "step": 969
+    },
+    {
+      "epoch": 2.130697419000549,
+      "grad_norm": 0.35073891282081604,
+      "learning_rate": 9.674267100977197e-05,
+      "loss": 0.671,
+      "step": 970
+    },
+    {
+      "epoch": 2.1328940142778694,
+      "grad_norm": 0.5013448596000671,
+      "learning_rate": 9.649837133550488e-05,
+      "loss": 0.5278,
+      "step": 971
+    },
+    {
+      "epoch": 2.1350906095551894,
+      "grad_norm": 0.41064876317977905,
+      "learning_rate": 9.625407166123777e-05,
+      "loss": 0.5847,
+      "step": 972
+    },
+    {
+      "epoch": 2.13728720483251,
+      "grad_norm": 0.2834590673446655,
+      "learning_rate": 9.600977198697067e-05,
+      "loss": 0.5789,
+      "step": 973
+    },
+    {
+      "epoch": 2.1394838001098297,
+      "grad_norm": 0.35682836174964905,
+      "learning_rate": 9.576547231270357e-05,
+      "loss": 0.6498,
+      "step": 974
+    },
+    {
+      "epoch": 2.1416803953871497,
+      "grad_norm": 0.38761988282203674,
+      "learning_rate": 9.552117263843646e-05,
+      "loss": 0.6678,
+      "step": 975
+    },
+    {
+      "epoch": 2.14387699066447,
+      "grad_norm": 0.3560211658477783,
+      "learning_rate": 9.527687296416936e-05,
+      "loss": 0.5912,
+      "step": 976
+    },
+    {
+      "epoch": 2.14607358594179,
+      "grad_norm": 0.26209142804145813,
+      "learning_rate": 9.503257328990227e-05,
+      "loss": 0.6162,
+      "step": 977
+    },
+    {
+      "epoch": 2.1482701812191105,
+      "grad_norm": 0.3332812190055847,
+      "learning_rate": 9.478827361563517e-05,
+      "loss": 0.5516,
+      "step": 978
+    },
+    {
+      "epoch": 2.1504667764964305,
+      "grad_norm": 0.2781393527984619,
+      "learning_rate": 9.454397394136806e-05,
+      "loss": 0.6246,
+      "step": 979
+    },
+    {
+      "epoch": 2.152663371773751,
+      "grad_norm": 0.3723507225513458,
+      "learning_rate": 9.429967426710096e-05,
+      "loss": 0.6652,
+      "step": 980
+    },
+    {
+      "epoch": 2.154859967051071,
+      "grad_norm": 0.29219406843185425,
+      "learning_rate": 9.405537459283386e-05,
+      "loss": 0.4181,
+      "step": 981
+    },
+    {
+      "epoch": 2.157056562328391,
+      "grad_norm": 0.3464544713497162,
+      "learning_rate": 9.381107491856675e-05,
+      "loss": 0.5142,
+      "step": 982
+    },
+    {
+      "epoch": 2.1592531576057112,
+      "grad_norm": 0.39875712990760803,
+      "learning_rate": 9.356677524429967e-05,
+      "loss": 0.688,
+      "step": 983
+    },
+    {
+      "epoch": 2.161449752883031,
+      "grad_norm": 0.3801586329936981,
+      "learning_rate": 9.332247557003257e-05,
+      "loss": 0.5269,
+      "step": 984
+    },
+    {
+      "epoch": 2.1636463481603516,
+      "grad_norm": 0.31335434317588806,
+      "learning_rate": 9.307817589576547e-05,
+      "loss": 0.5037,
+      "step": 985
+    },
+    {
+      "epoch": 2.1658429434376716,
+      "grad_norm": 0.48261263966560364,
+      "learning_rate": 9.283387622149838e-05,
+      "loss": 0.515,
+      "step": 986
+    },
+    {
+      "epoch": 2.168039538714992,
+      "grad_norm": 0.3493369221687317,
+      "learning_rate": 9.258957654723127e-05,
+      "loss": 0.7235,
+      "step": 987
+    },
+    {
+      "epoch": 2.170236133992312,
+      "grad_norm": 0.3027682602405548,
+      "learning_rate": 9.234527687296417e-05,
+      "loss": 0.4981,
+      "step": 988
+    },
+    {
+      "epoch": 2.172432729269632,
+      "grad_norm": 0.28331682085990906,
+      "learning_rate": 9.210097719869707e-05,
+      "loss": 0.5831,
+      "step": 989
+    },
+    {
+      "epoch": 2.1746293245469523,
+      "grad_norm": 0.36998894810676575,
+      "learning_rate": 9.185667752442996e-05,
+      "loss": 0.741,
+      "step": 990
+    },
+    {
+      "epoch": 2.1768259198242723,
+      "grad_norm": 0.3569234609603882,
+      "learning_rate": 9.161237785016286e-05,
+      "loss": 0.537,
+      "step": 991
+    },
+    {
+      "epoch": 2.1790225151015927,
+      "grad_norm": 0.2555007040500641,
+      "learning_rate": 9.136807817589577e-05,
+      "loss": 0.5525,
+      "step": 992
+    },
+    {
+      "epoch": 2.1812191103789127,
+      "grad_norm": 0.24925877153873444,
+      "learning_rate": 9.112377850162867e-05,
+      "loss": 0.6942,
+      "step": 993
+    },
+    {
+      "epoch": 2.1834157056562327,
+      "grad_norm": 0.34174925088882446,
+      "learning_rate": 9.087947882736156e-05,
+      "loss": 0.5238,
+      "step": 994
+    },
+    {
+      "epoch": 2.185612300933553,
+      "grad_norm": 0.25993165373802185,
+      "learning_rate": 9.063517915309446e-05,
+      "loss": 0.5323,
+      "step": 995
+    },
+    {
+      "epoch": 2.187808896210873,
+      "grad_norm": 0.515430748462677,
+      "learning_rate": 9.039087947882736e-05,
+      "loss": 0.6959,
+      "step": 996
+    },
+    {
+      "epoch": 2.1900054914881935,
+      "grad_norm": 0.337978720664978,
+      "learning_rate": 9.014657980456025e-05,
+      "loss": 0.6296,
+      "step": 997
+    },
+    {
+      "epoch": 2.1922020867655134,
+      "grad_norm": 0.3303092420101166,
+      "learning_rate": 8.990228013029316e-05,
+      "loss": 0.6237,
+      "step": 998
+    },
+    {
+      "epoch": 2.1943986820428334,
+      "grad_norm": 0.33312517404556274,
+      "learning_rate": 8.965798045602606e-05,
+      "loss": 0.7386,
+      "step": 999
+    },
+    {
+      "epoch": 2.196595277320154,
+      "grad_norm": 0.32515770196914673,
+      "learning_rate": 8.941368078175896e-05,
+      "loss": 0.5635,
+      "step": 1000
+    },
+    {
+      "epoch": 2.1987918725974738,
+      "grad_norm": 0.3189857006072998,
+      "learning_rate": 8.916938110749185e-05,
+      "loss": 0.4292,
+      "step": 1001
+    },
+    {
+      "epoch": 2.200988467874794,
+      "grad_norm": 0.27129366993904114,
+      "learning_rate": 8.892508143322475e-05,
+      "loss": 0.5477,
+      "step": 1002
+    },
+    {
+      "epoch": 2.203185063152114,
+      "grad_norm": 0.3820481598377228,
+      "learning_rate": 8.868078175895766e-05,
+      "loss": 0.5429,
+      "step": 1003
+    },
+    {
+      "epoch": 2.2053816584294346,
+      "grad_norm": 0.34725454449653625,
+      "learning_rate": 8.843648208469055e-05,
+      "loss": 0.5057,
+      "step": 1004
+    },
+    {
+      "epoch": 2.2075782537067545,
+      "grad_norm": 0.29410260915756226,
+      "learning_rate": 8.819218241042345e-05,
+      "loss": 0.5784,
+      "step": 1005
+    },
+    {
+      "epoch": 2.2097748489840745,
+      "grad_norm": 0.4241052567958832,
+      "learning_rate": 8.794788273615635e-05,
+      "loss": 0.5956,
+      "step": 1006
+    },
+    {
+      "epoch": 2.211971444261395,
+      "grad_norm": 0.31482958793640137,
+      "learning_rate": 8.770358306188925e-05,
+      "loss": 0.4974,
+      "step": 1007
+    },
+    {
+      "epoch": 2.214168039538715,
+      "grad_norm": 0.33919915556907654,
+      "learning_rate": 8.745928338762214e-05,
+      "loss": 0.7327,
+      "step": 1008
+    },
+    {
+      "epoch": 2.2163646348160353,
+      "grad_norm": 0.2845192849636078,
+      "learning_rate": 8.721498371335505e-05,
+      "loss": 0.6391,
+      "step": 1009
+    },
+    {
+      "epoch": 2.2185612300933553,
+      "grad_norm": 0.3469097912311554,
+      "learning_rate": 8.697068403908795e-05,
+      "loss": 0.4916,
+      "step": 1010
+    },
+    {
+      "epoch": 2.2207578253706757,
+      "grad_norm": 0.30652695894241333,
+      "learning_rate": 8.672638436482085e-05,
+      "loss": 0.6293,
+      "step": 1011
+    },
+    {
+      "epoch": 2.2229544206479956,
+      "grad_norm": 0.3529328405857086,
+      "learning_rate": 8.648208469055374e-05,
+      "loss": 0.7946,
+      "step": 1012
+    },
+    {
+      "epoch": 2.2251510159253156,
+      "grad_norm": 0.355424165725708,
+      "learning_rate": 8.623778501628664e-05,
+      "loss": 0.5263,
+      "step": 1013
+    },
+    {
+      "epoch": 2.227347611202636,
+      "grad_norm": 0.45320528745651245,
+      "learning_rate": 8.599348534201955e-05,
+      "loss": 0.6727,
+      "step": 1014
+    },
+    {
+      "epoch": 2.229544206479956,
+      "grad_norm": 0.4962396025657654,
+      "learning_rate": 8.574918566775243e-05,
+      "loss": 0.7042,
+      "step": 1015
+    },
+    {
+      "epoch": 2.2317408017572764,
+      "grad_norm": 0.2847757935523987,
+      "learning_rate": 8.550488599348534e-05,
+      "loss": 0.6367,
+      "step": 1016
+    },
+    {
+      "epoch": 2.2339373970345964,
+      "grad_norm": 0.44635120034217834,
+      "learning_rate": 8.526058631921824e-05,
+      "loss": 0.6484,
+      "step": 1017
+    },
+    {
+      "epoch": 2.2361339923119163,
+      "grad_norm": 0.41171061992645264,
+      "learning_rate": 8.501628664495114e-05,
+      "loss": 0.7118,
+      "step": 1018
+    },
+    {
+      "epoch": 2.2383305875892368,
+      "grad_norm": 0.3229183852672577,
+      "learning_rate": 8.477198697068403e-05,
+      "loss": 0.6792,
+      "step": 1019
+    },
+    {
+      "epoch": 2.2405271828665567,
+      "grad_norm": 1.3522107601165771,
+      "learning_rate": 8.452768729641694e-05,
+      "loss": 0.5759,
+      "step": 1020
+    },
+    {
+      "epoch": 2.242723778143877,
+      "grad_norm": 0.2898223102092743,
+      "learning_rate": 8.428338762214984e-05,
+      "loss": 0.5468,
+      "step": 1021
+    },
+    {
+      "epoch": 2.244920373421197,
+      "grad_norm": 0.3508555293083191,
+      "learning_rate": 8.403908794788273e-05,
+      "loss": 0.4892,
+      "step": 1022
+    },
+    {
+      "epoch": 2.247116968698517,
+      "grad_norm": 0.5690311193466187,
+      "learning_rate": 8.379478827361563e-05,
+      "loss": 0.6875,
+      "step": 1023
+    },
+    {
+      "epoch": 2.2493135639758375,
+      "grad_norm": 0.5027151703834534,
+      "learning_rate": 8.355048859934853e-05,
+      "loss": 0.5623,
+      "step": 1024
+    },
+    {
+      "epoch": 2.2515101592531575,
+      "grad_norm": 0.3075565695762634,
+      "learning_rate": 8.330618892508144e-05,
+      "loss": 0.5805,
+      "step": 1025
+    },
+    {
+      "epoch": 2.253706754530478,
+      "grad_norm": 0.28475990891456604,
+      "learning_rate": 8.306188925081432e-05,
+      "loss": 0.5261,
+      "step": 1026
+    },
+    {
+      "epoch": 2.255903349807798,
+      "grad_norm": 0.36897411942481995,
+      "learning_rate": 8.281758957654723e-05,
+      "loss": 0.613,
+      "step": 1027
+    },
+    {
+      "epoch": 2.2580999450851182,
+      "grad_norm": 0.24904020130634308,
+      "learning_rate": 8.257328990228013e-05,
+      "loss": 0.6908,
+      "step": 1028
+    },
+    {
+      "epoch": 2.260296540362438,
+      "grad_norm": 0.377989798784256,
+      "learning_rate": 8.232899022801302e-05,
+      "loss": 0.6399,
+      "step": 1029
+    },
+    {
+      "epoch": 2.262493135639758,
+      "grad_norm": 0.3288625180721283,
+      "learning_rate": 8.208469055374592e-05,
+      "loss": 0.453,
+      "step": 1030
+    },
+    {
+      "epoch": 2.2646897309170786,
+      "grad_norm": 0.31279629468917847,
+      "learning_rate": 8.184039087947882e-05,
+      "loss": 0.5149,
+      "step": 1031
+    },
+    {
+      "epoch": 2.2668863261943986,
+      "grad_norm": 0.348426878452301,
+      "learning_rate": 8.159609120521173e-05,
+      "loss": 0.4982,
+      "step": 1032
+    },
+    {
+      "epoch": 2.269082921471719,
+      "grad_norm": 0.333123117685318,
+      "learning_rate": 8.135179153094462e-05,
+      "loss": 0.7462,
+      "step": 1033
+    },
+    {
+      "epoch": 2.271279516749039,
+      "grad_norm": 0.4721328914165497,
+      "learning_rate": 8.110749185667752e-05,
+      "loss": 0.5673,
+      "step": 1034
+    },
+    {
+      "epoch": 2.2734761120263594,
+      "grad_norm": 0.30867999792099,
+      "learning_rate": 8.086319218241042e-05,
+      "loss": 0.526,
+      "step": 1035
+    },
+    {
+      "epoch": 2.2756727073036793,
+      "grad_norm": 0.2876860797405243,
+      "learning_rate": 8.061889250814331e-05,
+      "loss": 0.574,
+      "step": 1036
+    },
+    {
+      "epoch": 2.2778693025809993,
+      "grad_norm": 0.274228572845459,
+      "learning_rate": 8.037459283387621e-05,
+      "loss": 0.5598,
+      "step": 1037
+    },
+    {
+      "epoch": 2.2800658978583197,
+      "grad_norm": 0.3044149577617645,
+      "learning_rate": 8.013029315960912e-05,
+      "loss": 0.5563,
+      "step": 1038
+    },
+    {
+      "epoch": 2.2822624931356397,
+      "grad_norm": 0.3240131139755249,
+      "learning_rate": 7.988599348534202e-05,
+      "loss": 0.5107,
+      "step": 1039
+    },
+    {
+      "epoch": 2.28445908841296,
+      "grad_norm": 0.363832950592041,
+      "learning_rate": 7.964169381107491e-05,
+      "loss": 0.6284,
+      "step": 1040
+    },
+    {
+      "epoch": 2.28665568369028,
+      "grad_norm": 0.48597678542137146,
+      "learning_rate": 7.939739413680781e-05,
+      "loss": 0.6068,
+      "step": 1041
+    },
+    {
+      "epoch": 2.2888522789676005,
+      "grad_norm": 0.36999255418777466,
+      "learning_rate": 7.915309446254071e-05,
+      "loss": 0.6745,
+      "step": 1042
+    },
+    {
+      "epoch": 2.2910488742449204,
+      "grad_norm": 0.27040785551071167,
+      "learning_rate": 7.890879478827362e-05,
+      "loss": 0.485,
+      "step": 1043
+    },
+    {
+      "epoch": 2.2932454695222404,
+      "grad_norm": 0.38252371549606323,
+      "learning_rate": 7.86644951140065e-05,
+      "loss": 0.5942,
+      "step": 1044
+    },
+    {
+      "epoch": 2.295442064799561,
+      "grad_norm": 0.29662197828292847,
+      "learning_rate": 7.842019543973941e-05,
+      "loss": 0.6824,
+      "step": 1045
+    },
+    {
+      "epoch": 2.297638660076881,
+      "grad_norm": 0.3458291292190552,
+      "learning_rate": 7.817589576547231e-05,
+      "loss": 0.5822,
+      "step": 1046
+    },
+    {
+      "epoch": 2.2998352553542007,
+      "grad_norm": 0.5660237669944763,
+      "learning_rate": 7.79315960912052e-05,
+      "loss": 0.456,
+      "step": 1047
+    },
+    {
+      "epoch": 2.302031850631521,
+      "grad_norm": 0.3648456633090973,
+      "learning_rate": 7.76872964169381e-05,
+      "loss": 0.5235,
+      "step": 1048
+    },
+    {
+      "epoch": 2.304228445908841,
+      "grad_norm": 0.3552340567111969,
+      "learning_rate": 7.7442996742671e-05,
+      "loss": 0.5664,
+      "step": 1049
+    },
+    {
+      "epoch": 2.3064250411861615,
+      "grad_norm": 0.2951726019382477,
+      "learning_rate": 7.719869706840391e-05,
+      "loss": 0.6395,
+      "step": 1050
+    },
+    {
+      "epoch": 2.3086216364634815,
+      "grad_norm": 0.35674580931663513,
+      "learning_rate": 7.69543973941368e-05,
+      "loss": 0.4544,
+      "step": 1051
+    },
+    {
+      "epoch": 2.310818231740802,
+      "grad_norm": 0.3143481910228729,
+      "learning_rate": 7.67100977198697e-05,
+      "loss": 0.5466,
+      "step": 1052
+    },
+    {
+      "epoch": 2.313014827018122,
+      "grad_norm": 0.3722371459007263,
+      "learning_rate": 7.64657980456026e-05,
+      "loss": 0.6301,
+      "step": 1053
+    },
+    {
+      "epoch": 2.315211422295442,
+      "grad_norm": 0.3105016350746155,
+      "learning_rate": 7.622149837133549e-05,
+      "loss": 0.5802,
+      "step": 1054
+    },
+    {
+      "epoch": 2.3174080175727623,
+      "grad_norm": 0.309731662273407,
+      "learning_rate": 7.59771986970684e-05,
+      "loss": 0.7584,
+      "step": 1055
+    },
+    {
+      "epoch": 2.3196046128500822,
+      "grad_norm": 0.368158221244812,
+      "learning_rate": 7.57328990228013e-05,
+      "loss": 0.5738,
+      "step": 1056
+    },
+    {
+      "epoch": 2.3218012081274026,
+      "grad_norm": 0.3946864604949951,
+      "learning_rate": 7.54885993485342e-05,
+      "loss": 0.7717,
+      "step": 1057
+    },
+    {
+      "epoch": 2.3239978034047226,
+      "grad_norm": 0.2715712785720825,
+      "learning_rate": 7.524429967426709e-05,
+      "loss": 0.4115,
+      "step": 1058
+    },
+    {
+      "epoch": 2.326194398682043,
+      "grad_norm": 0.33181726932525635,
+      "learning_rate": 7.5e-05,
+      "loss": 0.5202,
+      "step": 1059
+    },
+    {
+      "epoch": 2.328390993959363,
+      "grad_norm": 0.2959939241409302,
+      "learning_rate": 7.47557003257329e-05,
+      "loss": 0.5391,
+      "step": 1060
+    },
+    {
+      "epoch": 2.330587589236683,
+      "grad_norm": 0.5121694207191467,
+      "learning_rate": 7.451140065146579e-05,
+      "loss": 0.5683,
+      "step": 1061
+    },
+    {
+      "epoch": 2.3327841845140034,
+      "grad_norm": 0.35229071974754333,
+      "learning_rate": 7.426710097719869e-05,
+      "loss": 0.6441,
+      "step": 1062
+    },
+    {
+      "epoch": 2.3349807797913233,
+      "grad_norm": 0.32757681608200073,
+      "learning_rate": 7.402280130293159e-05,
+      "loss": 0.6965,
+      "step": 1063
+    },
+    {
+      "epoch": 2.3371773750686438,
+      "grad_norm": 0.3974643349647522,
+      "learning_rate": 7.37785016286645e-05,
+      "loss": 0.6261,
+      "step": 1064
+    },
+    {
+      "epoch": 2.3393739703459637,
+      "grad_norm": 0.27362626791000366,
+      "learning_rate": 7.353420195439738e-05,
+      "loss": 0.5959,
+      "step": 1065
+    },
+    {
+      "epoch": 2.341570565623284,
+      "grad_norm": 0.3045765161514282,
+      "learning_rate": 7.328990228013029e-05,
+      "loss": 0.6903,
+      "step": 1066
+    },
+    {
+      "epoch": 2.343767160900604,
+      "grad_norm": 0.24681130051612854,
+      "learning_rate": 7.304560260586319e-05,
+      "loss": 0.4895,
+      "step": 1067
+    },
+    {
+      "epoch": 2.345963756177924,
+      "grad_norm": 0.29551970958709717,
+      "learning_rate": 7.280130293159608e-05,
+      "loss": 0.6477,
+      "step": 1068
+    },
+    {
+      "epoch": 2.3481603514552445,
+      "grad_norm": 0.39739498496055603,
+      "learning_rate": 7.255700325732898e-05,
+      "loss": 0.6562,
+      "step": 1069
+    },
+    {
+      "epoch": 2.3503569467325645,
+      "grad_norm": 0.3182360827922821,
+      "learning_rate": 7.231270358306188e-05,
+      "loss": 0.5479,
+      "step": 1070
+    },
+    {
+      "epoch": 2.352553542009885,
+      "grad_norm": 0.3348866105079651,
+      "learning_rate": 7.206840390879479e-05,
+      "loss": 0.6135,
+      "step": 1071
+    },
+    {
+      "epoch": 2.354750137287205,
+      "grad_norm": 0.3930214047431946,
+      "learning_rate": 7.182410423452768e-05,
+      "loss": 0.4899,
+      "step": 1072
+    },
+    {
+      "epoch": 2.356946732564525,
+      "grad_norm": 0.33119136095046997,
+      "learning_rate": 7.157980456026058e-05,
+      "loss": 0.5271,
+      "step": 1073
+    },
+    {
+      "epoch": 2.359143327841845,
+      "grad_norm": 0.32269689440727234,
+      "learning_rate": 7.133550488599348e-05,
+      "loss": 0.5729,
+      "step": 1074
+    },
+    {
+      "epoch": 2.361339923119165,
+      "grad_norm": 0.29347434639930725,
+      "learning_rate": 7.109120521172637e-05,
+      "loss": 0.6957,
+      "step": 1075
+    },
+    {
+      "epoch": 2.3635365183964856,
+      "grad_norm": 0.39845579862594604,
+      "learning_rate": 7.084690553745927e-05,
+      "loss": 0.6395,
+      "step": 1076
+    },
+    {
+      "epoch": 2.3657331136738056,
+      "grad_norm": 0.3080476224422455,
+      "learning_rate": 7.060260586319218e-05,
+      "loss": 0.5766,
+      "step": 1077
+    },
+    {
+      "epoch": 2.3679297089511255,
+      "grad_norm": 0.3723848760128021,
+      "learning_rate": 7.035830618892508e-05,
+      "loss": 0.6447,
+      "step": 1078
+    },
+    {
+      "epoch": 2.370126304228446,
+      "grad_norm": 0.31864669919013977,
+      "learning_rate": 7.011400651465797e-05,
+      "loss": 0.4612,
+      "step": 1079
+    },
+    {
+      "epoch": 2.372322899505766,
+      "grad_norm": 0.33365413546562195,
+      "learning_rate": 6.986970684039087e-05,
+      "loss": 0.6147,
+      "step": 1080
+    },
+    {
+      "epoch": 2.3745194947830863,
+      "grad_norm": 0.4464552700519562,
+      "learning_rate": 6.962540716612377e-05,
+      "loss": 0.7844,
+      "step": 1081
+    },
+    {
+      "epoch": 2.3767160900604063,
+      "grad_norm": 0.41667115688323975,
+      "learning_rate": 6.938110749185668e-05,
+      "loss": 0.4942,
+      "step": 1082
+    },
+    {
+      "epoch": 2.3789126853377267,
+      "grad_norm": 0.31530165672302246,
+      "learning_rate": 6.913680781758956e-05,
+      "loss": 0.5468,
+      "step": 1083
+    },
+    {
+      "epoch": 2.3811092806150467,
+      "grad_norm": 0.31491440534591675,
+      "learning_rate": 6.889250814332247e-05,
+      "loss": 0.7305,
+      "step": 1084
+    },
+    {
+      "epoch": 2.3833058758923666,
+      "grad_norm": 0.40166211128234863,
+      "learning_rate": 6.864820846905537e-05,
+      "loss": 0.7864,
+      "step": 1085
+    },
+    {
+      "epoch": 2.385502471169687,
+      "grad_norm": 0.31348833441734314,
+      "learning_rate": 6.840390879478826e-05,
+      "loss": 0.494,
+      "step": 1086
+    },
+    {
+      "epoch": 2.387699066447007,
+      "grad_norm": 0.40460991859436035,
+      "learning_rate": 6.815960912052116e-05,
+      "loss": 0.7467,
+      "step": 1087
+    },
+    {
+      "epoch": 2.3898956617243274,
+      "grad_norm": 0.4652503430843353,
+      "learning_rate": 6.791530944625406e-05,
+      "loss": 0.4911,
+      "step": 1088
+    },
+    {
+      "epoch": 2.3920922570016474,
+      "grad_norm": 0.40283307433128357,
+      "learning_rate": 6.767100977198697e-05,
+      "loss": 0.6702,
+      "step": 1089
+    },
+    {
+      "epoch": 2.394288852278968,
+      "grad_norm": 0.3422105014324188,
+      "learning_rate": 6.742671009771986e-05,
+      "loss": 0.5801,
+      "step": 1090
+    },
+    {
+      "epoch": 2.396485447556288,
+      "grad_norm": 0.38644954562187195,
+      "learning_rate": 6.718241042345276e-05,
+      "loss": 0.5537,
+      "step": 1091
+    },
+    {
+      "epoch": 2.3986820428336078,
+      "grad_norm": 0.36957839131355286,
+      "learning_rate": 6.693811074918566e-05,
+      "loss": 0.606,
+      "step": 1092
+    },
+    {
+      "epoch": 2.400878638110928,
+      "grad_norm": 0.3754534423351288,
+      "learning_rate": 6.669381107491855e-05,
+      "loss": 0.6351,
+      "step": 1093
+    },
+    {
+      "epoch": 2.403075233388248,
+      "grad_norm": 0.31378570199012756,
+      "learning_rate": 6.644951140065145e-05,
+      "loss": 0.6636,
+      "step": 1094
+    },
+    {
+      "epoch": 2.4052718286655685,
+      "grad_norm": 0.5148921608924866,
+      "learning_rate": 6.620521172638436e-05,
+      "loss": 0.5421,
+      "step": 1095
+    },
+    {
+      "epoch": 2.4074684239428885,
+      "grad_norm": 0.32462626695632935,
+      "learning_rate": 6.596091205211726e-05,
+      "loss": 0.46,
+      "step": 1096
+    },
+    {
+      "epoch": 2.409665019220209,
+      "grad_norm": 0.4241209328174591,
+      "learning_rate": 6.571661237785015e-05,
+      "loss": 0.6578,
+      "step": 1097
+    },
+    {
+      "epoch": 2.411861614497529,
+      "grad_norm": 0.3013802468776703,
+      "learning_rate": 6.547231270358307e-05,
+      "loss": 0.6542,
+      "step": 1098
+    },
+    {
+      "epoch": 2.414058209774849,
+      "grad_norm": 0.395439475774765,
+      "learning_rate": 6.522801302931595e-05,
+      "loss": 0.6774,
+      "step": 1099
+    },
+    {
+      "epoch": 2.4162548050521693,
+      "grad_norm": 0.35957175493240356,
+      "learning_rate": 6.498371335504886e-05,
+      "loss": 0.6436,
+      "step": 1100
+    },
+    {
+      "epoch": 2.4184514003294892,
+      "grad_norm": 0.3322821259498596,
+      "learning_rate": 6.473941368078176e-05,
+      "loss": 0.5119,
+      "step": 1101
+    },
+    {
+      "epoch": 2.420647995606809,
+      "grad_norm": 0.37965264916419983,
+      "learning_rate": 6.449511400651466e-05,
+      "loss": 0.6335,
+      "step": 1102
+    },
+    {
+      "epoch": 2.4228445908841296,
+      "grad_norm": 0.39819997549057007,
+      "learning_rate": 6.425081433224755e-05,
+      "loss": 0.6017,
+      "step": 1103
+    },
+    {
+      "epoch": 2.4250411861614496,
+      "grad_norm": 0.40206262469291687,
+      "learning_rate": 6.400651465798045e-05,
+      "loss": 0.6422,
+      "step": 1104
+    },
+    {
+      "epoch": 2.42723778143877,
+      "grad_norm": 0.4805324673652649,
+      "learning_rate": 6.376221498371336e-05,
+      "loss": 0.6004,
+      "step": 1105
+    },
+    {
+      "epoch": 2.42943437671609,
+      "grad_norm": 0.30040624737739563,
+      "learning_rate": 6.351791530944625e-05,
+      "loss": 0.529,
+      "step": 1106
+    },
+    {
+      "epoch": 2.4316309719934104,
+      "grad_norm": 0.3511275053024292,
+      "learning_rate": 6.327361563517915e-05,
+      "loss": 0.6492,
+      "step": 1107
+    },
+    {
+      "epoch": 2.4338275672707304,
+      "grad_norm": 0.42680490016937256,
+      "learning_rate": 6.302931596091205e-05,
+      "loss": 0.7039,
+      "step": 1108
+    },
+    {
+      "epoch": 2.4360241625480503,
+      "grad_norm": 0.3264611065387726,
+      "learning_rate": 6.278501628664495e-05,
+      "loss": 0.6312,
+      "step": 1109
+    },
+    {
+      "epoch": 2.4382207578253707,
+      "grad_norm": 0.3092852532863617,
+      "learning_rate": 6.254071661237784e-05,
+      "loss": 0.6293,
+      "step": 1110
+    },
+    {
+      "epoch": 2.4404173531026907,
+      "grad_norm": 0.2942371964454651,
+      "learning_rate": 6.229641693811075e-05,
+      "loss": 0.6003,
+      "step": 1111
+    },
+    {
+      "epoch": 2.442613948380011,
+      "grad_norm": 0.35826870799064636,
+      "learning_rate": 6.205211726384365e-05,
+      "loss": 0.4129,
+      "step": 1112
+    },
+    {
+      "epoch": 2.444810543657331,
+      "grad_norm": 0.3668456971645355,
+      "learning_rate": 6.180781758957654e-05,
+      "loss": 0.7337,
+      "step": 1113
+    },
+    {
+      "epoch": 2.4470071389346515,
+      "grad_norm": 0.44200536608695984,
+      "learning_rate": 6.156351791530944e-05,
+      "loss": 0.7607,
+      "step": 1114
+    },
+    {
+      "epoch": 2.4492037342119715,
+      "grad_norm": 0.3001335561275482,
+      "learning_rate": 6.131921824104234e-05,
+      "loss": 0.5349,
+      "step": 1115
+    },
+    {
+      "epoch": 2.4514003294892914,
+      "grad_norm": 0.29789355397224426,
+      "learning_rate": 6.107491856677525e-05,
+      "loss": 0.5434,
+      "step": 1116
+    },
+    {
+      "epoch": 2.453596924766612,
+      "grad_norm": 0.3011775314807892,
+      "learning_rate": 6.083061889250814e-05,
+      "loss": 0.5388,
+      "step": 1117
+    },
+    {
+      "epoch": 2.455793520043932,
+      "grad_norm": 0.3254930377006531,
+      "learning_rate": 6.058631921824104e-05,
+      "loss": 0.5427,
+      "step": 1118
+    },
+    {
+      "epoch": 2.4579901153212522,
+      "grad_norm": 0.3494664430618286,
+      "learning_rate": 6.0342019543973935e-05,
+      "loss": 0.6992,
+      "step": 1119
+    },
+    {
+      "epoch": 2.460186710598572,
+      "grad_norm": 0.3502826988697052,
+      "learning_rate": 6.009771986970684e-05,
+      "loss": 0.5768,
+      "step": 1120
+    },
+    {
+      "epoch": 2.4623833058758926,
+      "grad_norm": 0.34961405396461487,
+      "learning_rate": 5.9853420195439734e-05,
+      "loss": 0.5584,
+      "step": 1121
+    },
+    {
+      "epoch": 2.4645799011532126,
+      "grad_norm": 0.4981488883495331,
+      "learning_rate": 5.9609120521172636e-05,
+      "loss": 0.9293,
+      "step": 1122
+    },
+    {
+      "epoch": 2.4667764964305325,
+      "grad_norm": 0.41492149233818054,
+      "learning_rate": 5.936482084690553e-05,
+      "loss": 0.5664,
+      "step": 1123
+    },
+    {
+      "epoch": 2.468973091707853,
+      "grad_norm": 0.4649076461791992,
+      "learning_rate": 5.9120521172638435e-05,
+      "loss": 0.5977,
+      "step": 1124
+    },
+    {
+      "epoch": 2.471169686985173,
+      "grad_norm": 0.33978936076164246,
+      "learning_rate": 5.887622149837133e-05,
+      "loss": 0.5334,
+      "step": 1125
+    },
+    {
+      "epoch": 2.473366282262493,
+      "grad_norm": 0.4572916030883789,
+      "learning_rate": 5.863192182410423e-05,
+      "loss": 0.6832,
+      "step": 1126
+    },
+    {
+      "epoch": 2.4755628775398133,
+      "grad_norm": 0.3624480366706848,
+      "learning_rate": 5.838762214983713e-05,
+      "loss": 0.5113,
+      "step": 1127
+    },
+    {
+      "epoch": 2.4777594728171333,
+      "grad_norm": 0.24833893775939941,
+      "learning_rate": 5.8143322475570026e-05,
+      "loss": 0.599,
+      "step": 1128
+    },
+    {
+      "epoch": 2.4799560680944537,
+      "grad_norm": 0.35467055439949036,
+      "learning_rate": 5.789902280130293e-05,
+      "loss": 0.7665,
+      "step": 1129
+    },
+    {
+      "epoch": 2.4821526633717736,
+      "grad_norm": 0.6054608225822449,
+      "learning_rate": 5.7654723127035825e-05,
+      "loss": 0.5703,
+      "step": 1130
+    },
+    {
+      "epoch": 2.484349258649094,
+      "grad_norm": 0.3156985342502594,
+      "learning_rate": 5.741042345276873e-05,
+      "loss": 0.5546,
+      "step": 1131
+    },
+    {
+      "epoch": 2.486545853926414,
+      "grad_norm": 0.31863078474998474,
+      "learning_rate": 5.716612377850162e-05,
+      "loss": 0.6165,
+      "step": 1132
+    },
+    {
+      "epoch": 2.488742449203734,
+      "grad_norm": 0.44531193375587463,
+      "learning_rate": 5.6921824104234526e-05,
+      "loss": 0.484,
+      "step": 1133
+    },
+    {
+      "epoch": 2.4909390444810544,
+      "grad_norm": 0.38687118887901306,
+      "learning_rate": 5.667752442996742e-05,
+      "loss": 0.5863,
+      "step": 1134
+    },
+    {
+      "epoch": 2.4931356397583744,
+      "grad_norm": 0.33428195118904114,
+      "learning_rate": 5.643322475570032e-05,
+      "loss": 0.6131,
+      "step": 1135
+    },
+    {
+      "epoch": 2.495332235035695,
+      "grad_norm": 0.33822137117385864,
+      "learning_rate": 5.618892508143322e-05,
+      "loss": 0.4938,
+      "step": 1136
+    },
+    {
+      "epoch": 2.4975288303130148,
+      "grad_norm": 0.3004172146320343,
+      "learning_rate": 5.594462540716612e-05,
+      "loss": 0.3972,
+      "step": 1137
+    },
+    {
+      "epoch": 2.499725425590335,
+      "grad_norm": 0.4346812963485718,
+      "learning_rate": 5.570032573289902e-05,
+      "loss": 0.5745,
+      "step": 1138
+    },
+    {
+      "epoch": 2.501922020867655,
+      "grad_norm": 0.3716161549091339,
+      "learning_rate": 5.5456026058631915e-05,
+      "loss": 0.7312,
+      "step": 1139
+    },
+    {
+      "epoch": 2.504118616144975,
+      "grad_norm": 0.7298053503036499,
+      "learning_rate": 5.521172638436482e-05,
+      "loss": 0.4984,
+      "step": 1140
+    },
+    {
+      "epoch": 2.5063152114222955,
+      "grad_norm": 0.24788261950016022,
+      "learning_rate": 5.4967426710097714e-05,
+      "loss": 0.5482,
+      "step": 1141
+    },
+    {
+      "epoch": 2.5085118066996155,
+      "grad_norm": 0.29723817110061646,
+      "learning_rate": 5.472312703583061e-05,
+      "loss": 0.6537,
+      "step": 1142
+    },
+    {
+      "epoch": 2.510708401976936,
+      "grad_norm": 0.37248557806015015,
+      "learning_rate": 5.447882736156351e-05,
+      "loss": 0.7063,
+      "step": 1143
+    },
+    {
+      "epoch": 2.512904997254256,
+      "grad_norm": 0.31541529297828674,
+      "learning_rate": 5.423452768729641e-05,
+      "loss": 0.5846,
+      "step": 1144
+    },
+    {
+      "epoch": 2.5151015925315763,
+      "grad_norm": 0.3811168372631073,
+      "learning_rate": 5.399022801302931e-05,
+      "loss": 0.5779,
+      "step": 1145
+    },
+    {
+      "epoch": 2.5172981878088962,
+      "grad_norm": 0.4978001117706299,
+      "learning_rate": 5.374592833876221e-05,
+      "loss": 0.5899,
+      "step": 1146
+    },
+    {
+      "epoch": 2.519494783086216,
+      "grad_norm": 0.34890079498291016,
+      "learning_rate": 5.350162866449511e-05,
+      "loss": 0.5704,
+      "step": 1147
+    },
+    {
+      "epoch": 2.5216913783635366,
+      "grad_norm": 0.3254714012145996,
+      "learning_rate": 5.3257328990228006e-05,
+      "loss": 0.4747,
+      "step": 1148
+    },
+    {
+      "epoch": 2.5238879736408566,
+      "grad_norm": 0.3114301562309265,
+      "learning_rate": 5.301302931596091e-05,
+      "loss": 0.6483,
+      "step": 1149
+    },
+    {
+      "epoch": 2.5260845689181766,
+      "grad_norm": 0.32955828309059143,
+      "learning_rate": 5.2768729641693805e-05,
+      "loss": 0.8356,
+      "step": 1150
+    },
+    {
+      "epoch": 2.528281164195497,
+      "grad_norm": 0.4147564172744751,
+      "learning_rate": 5.25244299674267e-05,
+      "loss": 0.5992,
+      "step": 1151
+    },
+    {
+      "epoch": 2.5304777594728174,
+      "grad_norm": 0.3547995090484619,
+      "learning_rate": 5.2280130293159604e-05,
+      "loss": 0.5916,
+      "step": 1152
+    },
+    {
+      "epoch": 2.5326743547501374,
+      "grad_norm": 0.3431404232978821,
+      "learning_rate": 5.20358306188925e-05,
+      "loss": 0.6262,
+      "step": 1153
+    },
+    {
+      "epoch": 2.5348709500274573,
+      "grad_norm": 0.31657135486602783,
+      "learning_rate": 5.17915309446254e-05,
+      "loss": 0.5716,
+      "step": 1154
+    },
+    {
+      "epoch": 2.5370675453047777,
+      "grad_norm": 0.4484020471572876,
+      "learning_rate": 5.15472312703583e-05,
+      "loss": 0.5563,
+      "step": 1155
+    },
+    {
+      "epoch": 2.5392641405820977,
+      "grad_norm": 0.42113980650901794,
+      "learning_rate": 5.13029315960912e-05,
+      "loss": 0.68,
+      "step": 1156
+    },
+    {
+      "epoch": 2.5414607358594177,
+      "grad_norm": 0.6648101806640625,
+      "learning_rate": 5.10586319218241e-05,
+      "loss": 0.6797,
+      "step": 1157
+    },
+    {
+      "epoch": 2.543657331136738,
+      "grad_norm": 0.38691243529319763,
+      "learning_rate": 5.081433224755699e-05,
+      "loss": 0.561,
+      "step": 1158
+    },
+    {
+      "epoch": 2.545853926414058,
+      "grad_norm": 0.3410184979438782,
+      "learning_rate": 5.0570032573289896e-05,
+      "loss": 0.5228,
+      "step": 1159
+    },
+    {
+      "epoch": 2.5480505216913785,
+      "grad_norm": 0.3449324071407318,
+      "learning_rate": 5.032573289902279e-05,
+      "loss": 0.5696,
+      "step": 1160
+    },
+    {
+      "epoch": 2.5502471169686984,
+      "grad_norm": 0.37889084219932556,
+      "learning_rate": 5.0081433224755695e-05,
+      "loss": 0.4846,
+      "step": 1161
+    },
+    {
+      "epoch": 2.552443712246019,
+      "grad_norm": 0.34902289509773254,
+      "learning_rate": 4.983713355048859e-05,
+      "loss": 0.6407,
+      "step": 1162
+    },
+    {
+      "epoch": 2.554640307523339,
+      "grad_norm": 0.3935790956020355,
+      "learning_rate": 4.9592833876221493e-05,
+      "loss": 0.5032,
+      "step": 1163
+    },
+    {
+      "epoch": 2.556836902800659,
+      "grad_norm": 0.3602747619152069,
+      "learning_rate": 4.934853420195439e-05,
+      "loss": 0.6473,
+      "step": 1164
+    },
+    {
+      "epoch": 2.559033498077979,
+      "grad_norm": 0.3557460606098175,
+      "learning_rate": 4.910423452768729e-05,
+      "loss": 0.6515,
+      "step": 1165
+    },
+    {
+      "epoch": 2.561230093355299,
+      "grad_norm": 0.28969329595565796,
+      "learning_rate": 4.885993485342019e-05,
+      "loss": 0.5918,
+      "step": 1166
+    },
+    {
+      "epoch": 2.5634266886326196,
+      "grad_norm": 0.3520037829875946,
+      "learning_rate": 4.8615635179153084e-05,
+      "loss": 0.4916,
+      "step": 1167
+    },
+    {
+      "epoch": 2.5656232839099395,
+      "grad_norm": 0.3820478916168213,
+      "learning_rate": 4.837133550488599e-05,
+      "loss": 0.5909,
+      "step": 1168
+    },
+    {
+      "epoch": 2.56781987918726,
+      "grad_norm": 0.3452962040901184,
+      "learning_rate": 4.812703583061888e-05,
+      "loss": 0.5783,
+      "step": 1169
+    },
+    {
+      "epoch": 2.57001647446458,
+      "grad_norm": 0.2983270287513733,
+      "learning_rate": 4.7882736156351786e-05,
+      "loss": 0.5515,
+      "step": 1170
+    },
+    {
+      "epoch": 2.5722130697419,
+      "grad_norm": 0.3056553304195404,
+      "learning_rate": 4.763843648208468e-05,
+      "loss": 0.53,
+      "step": 1171
+    },
+    {
+      "epoch": 2.5744096650192203,
+      "grad_norm": 0.4891917109489441,
+      "learning_rate": 4.7394136807817584e-05,
+      "loss": 0.4992,
+      "step": 1172
+    },
+    {
+      "epoch": 2.5766062602965403,
+      "grad_norm": 0.3521086275577545,
+      "learning_rate": 4.714983713355048e-05,
+      "loss": 0.5525,
+      "step": 1173
+    },
+    {
+      "epoch": 2.5788028555738602,
+      "grad_norm": 0.38371115922927856,
+      "learning_rate": 4.6905537459283376e-05,
+      "loss": 0.6337,
+      "step": 1174
+    },
+    {
+      "epoch": 2.5809994508511807,
+      "grad_norm": 0.37448614835739136,
+      "learning_rate": 4.6661237785016286e-05,
+      "loss": 0.6265,
+      "step": 1175
+    },
+    {
+      "epoch": 2.583196046128501,
+      "grad_norm": 0.372931569814682,
+      "learning_rate": 4.641693811074919e-05,
+      "loss": 0.5297,
+      "step": 1176
+    },
+    {
+      "epoch": 2.585392641405821,
+      "grad_norm": 0.4340185225009918,
+      "learning_rate": 4.6172638436482085e-05,
+      "loss": 0.7585,
+      "step": 1177
+    },
+    {
+      "epoch": 2.587589236683141,
+      "grad_norm": 0.43668967485427856,
+      "learning_rate": 4.592833876221498e-05,
+      "loss": 0.5284,
+      "step": 1178
+    },
+    {
+      "epoch": 2.5897858319604614,
+      "grad_norm": 0.44207245111465454,
+      "learning_rate": 4.568403908794788e-05,
+      "loss": 0.6241,
+      "step": 1179
+    },
+    {
+      "epoch": 2.5919824272377814,
+      "grad_norm": 0.33144840598106384,
+      "learning_rate": 4.543973941368078e-05,
+      "loss": 0.5446,
+      "step": 1180
+    },
+    {
+      "epoch": 2.5941790225151014,
+      "grad_norm": 0.43423134088516235,
+      "learning_rate": 4.519543973941368e-05,
+      "loss": 0.6233,
+      "step": 1181
+    },
+    {
+      "epoch": 2.5963756177924218,
+      "grad_norm": 0.5007559061050415,
+      "learning_rate": 4.495114006514658e-05,
+      "loss": 0.6136,
+      "step": 1182
+    },
+    {
+      "epoch": 2.598572213069742,
+      "grad_norm": 0.32267042994499207,
+      "learning_rate": 4.470684039087948e-05,
+      "loss": 0.6038,
+      "step": 1183
+    },
+    {
+      "epoch": 2.600768808347062,
+      "grad_norm": 0.38702747225761414,
+      "learning_rate": 4.446254071661238e-05,
+      "loss": 0.6132,
+      "step": 1184
+    },
+    {
+      "epoch": 2.602965403624382,
+      "grad_norm": 0.36762750148773193,
+      "learning_rate": 4.421824104234527e-05,
+      "loss": 0.647,
+      "step": 1185
+    },
+    {
+      "epoch": 2.6051619989017025,
+      "grad_norm": 0.29947736859321594,
+      "learning_rate": 4.3973941368078175e-05,
+      "loss": 0.6579,
+      "step": 1186
+    },
+    {
+      "epoch": 2.6073585941790225,
+      "grad_norm": 0.284198522567749,
+      "learning_rate": 4.372964169381107e-05,
+      "loss": 0.5215,
+      "step": 1187
+    },
+    {
+      "epoch": 2.6095551894563425,
+      "grad_norm": 0.37136346101760864,
+      "learning_rate": 4.3485342019543974e-05,
+      "loss": 0.5761,
+      "step": 1188
+    },
+    {
+      "epoch": 2.611751784733663,
+      "grad_norm": 0.3389601409435272,
+      "learning_rate": 4.324104234527687e-05,
+      "loss": 0.5251,
+      "step": 1189
+    },
+    {
+      "epoch": 2.613948380010983,
+      "grad_norm": 0.3875013291835785,
+      "learning_rate": 4.299674267100977e-05,
+      "loss": 0.5843,
+      "step": 1190
+    },
+    {
+      "epoch": 2.6161449752883033,
+      "grad_norm": 0.3551155924797058,
+      "learning_rate": 4.275244299674267e-05,
+      "loss": 0.5923,
+      "step": 1191
+    },
+    {
+      "epoch": 2.618341570565623,
+      "grad_norm": 0.36455950140953064,
+      "learning_rate": 4.250814332247557e-05,
+      "loss": 0.4685,
+      "step": 1192
+    },
+    {
+      "epoch": 2.6205381658429436,
+      "grad_norm": 0.5868074893951416,
+      "learning_rate": 4.226384364820847e-05,
+      "loss": 0.8865,
+      "step": 1193
+    },
+    {
+      "epoch": 2.6227347611202636,
+      "grad_norm": 0.5389795303344727,
+      "learning_rate": 4.2019543973941364e-05,
+      "loss": 0.6944,
+      "step": 1194
+    },
+    {
+      "epoch": 2.6249313563975836,
+      "grad_norm": 0.2596234083175659,
+      "learning_rate": 4.1775244299674266e-05,
+      "loss": 0.4079,
+      "step": 1195
+    },
+    {
+      "epoch": 2.627127951674904,
+      "grad_norm": 0.44579586386680603,
+      "learning_rate": 4.153094462540716e-05,
+      "loss": 0.5659,
+      "step": 1196
+    },
+    {
+      "epoch": 2.629324546952224,
+      "grad_norm": 0.4281311333179474,
+      "learning_rate": 4.1286644951140065e-05,
+      "loss": 0.5264,
+      "step": 1197
+    },
+    {
+      "epoch": 2.6315211422295444,
+      "grad_norm": 0.28948143124580383,
+      "learning_rate": 4.104234527687296e-05,
+      "loss": 0.5969,
+      "step": 1198
+    },
+    {
+      "epoch": 2.6337177375068643,
+      "grad_norm": 0.3329457938671112,
+      "learning_rate": 4.0798045602605864e-05,
+      "loss": 0.5106,
+      "step": 1199
+    },
+    {
+      "epoch": 2.6359143327841847,
+      "grad_norm": 0.3458653688430786,
+      "learning_rate": 4.055374592833876e-05,
+      "loss": 0.5403,
+      "step": 1200
+    },
+    {
+      "epoch": 2.6381109280615047,
+      "grad_norm": 0.36975568532943726,
+      "learning_rate": 4.0309446254071656e-05,
+      "loss": 0.4706,
+      "step": 1201
+    },
+    {
+      "epoch": 2.6403075233388247,
+      "grad_norm": 0.4200529158115387,
+      "learning_rate": 4.006514657980456e-05,
+      "loss": 0.4622,
+      "step": 1202
+    },
+    {
+      "epoch": 2.642504118616145,
+      "grad_norm": 0.4237093925476074,
+      "learning_rate": 3.9820846905537454e-05,
+      "loss": 0.7444,
+      "step": 1203
+    },
+    {
+      "epoch": 2.644700713893465,
+      "grad_norm": 0.4348616898059845,
+      "learning_rate": 3.957654723127036e-05,
+      "loss": 0.556,
+      "step": 1204
+    },
+    {
+      "epoch": 2.646897309170785,
+      "grad_norm": 0.357363224029541,
+      "learning_rate": 3.933224755700325e-05,
+      "loss": 0.6235,
+      "step": 1205
+    },
+    {
+      "epoch": 2.6490939044481054,
+      "grad_norm": 0.37834882736206055,
+      "learning_rate": 3.9087947882736156e-05,
+      "loss": 0.7009,
+      "step": 1206
+    },
+    {
+      "epoch": 2.651290499725426,
+      "grad_norm": 0.370535284280777,
+      "learning_rate": 3.884364820846905e-05,
+      "loss": 0.6342,
+      "step": 1207
+    },
+    {
+      "epoch": 2.653487095002746,
+      "grad_norm": 0.42786186933517456,
+      "learning_rate": 3.8599348534201955e-05,
+      "loss": 0.5878,
+      "step": 1208
+    },
+    {
+      "epoch": 2.655683690280066,
+      "grad_norm": 0.428110271692276,
+      "learning_rate": 3.835504885993485e-05,
+      "loss": 0.4761,
+      "step": 1209
+    },
+    {
+      "epoch": 2.657880285557386,
+      "grad_norm": 0.39182230830192566,
+      "learning_rate": 3.8110749185667747e-05,
+      "loss": 0.7707,
+      "step": 1210
+    },
+    {
+      "epoch": 2.660076880834706,
+      "grad_norm": 0.3728565573692322,
+      "learning_rate": 3.786644951140065e-05,
+      "loss": 0.4932,
+      "step": 1211
+    },
+    {
+      "epoch": 2.662273476112026,
+      "grad_norm": 0.4355725049972534,
+      "learning_rate": 3.7622149837133545e-05,
+      "loss": 0.5237,
+      "step": 1212
+    },
+    {
+      "epoch": 2.6644700713893466,
+      "grad_norm": 0.37621086835861206,
+      "learning_rate": 3.737785016286645e-05,
+      "loss": 0.6512,
+      "step": 1213
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 0.4229162335395813,
+      "learning_rate": 3.7133550488599344e-05,
+      "loss": 0.4736,
+      "step": 1214
+    },
+    {
+      "epoch": 2.668863261943987,
+      "grad_norm": 0.36932286620140076,
+      "learning_rate": 3.688925081433225e-05,
+      "loss": 0.584,
+      "step": 1215
+    },
+    {
+      "epoch": 2.671059857221307,
+      "grad_norm": 0.37565213441848755,
+      "learning_rate": 3.664495114006514e-05,
+      "loss": 0.5812,
+      "step": 1216
+    },
+    {
+      "epoch": 2.6732564524986273,
+      "grad_norm": 0.308076411485672,
+      "learning_rate": 3.640065146579804e-05,
+      "loss": 0.6724,
+      "step": 1217
+    },
+    {
+      "epoch": 2.6754530477759473,
+      "grad_norm": 0.3448914587497711,
+      "learning_rate": 3.615635179153094e-05,
+      "loss": 0.6031,
+      "step": 1218
+    },
+    {
+      "epoch": 2.6776496430532672,
+      "grad_norm": 0.5128368735313416,
+      "learning_rate": 3.591205211726384e-05,
+      "loss": 0.5583,
+      "step": 1219
+    },
+    {
+      "epoch": 2.6798462383305877,
+      "grad_norm": 0.34365490078926086,
+      "learning_rate": 3.566775244299674e-05,
+      "loss": 0.571,
+      "step": 1220
+    },
+    {
+      "epoch": 2.6820428336079076,
+      "grad_norm": 0.4177529215812683,
+      "learning_rate": 3.5423452768729636e-05,
+      "loss": 0.713,
+      "step": 1221
+    },
+    {
+      "epoch": 2.684239428885228,
+      "grad_norm": 0.35535329580307007,
+      "learning_rate": 3.517915309446254e-05,
+      "loss": 0.4053,
+      "step": 1222
+    },
+    {
+      "epoch": 2.686436024162548,
+      "grad_norm": 0.3490915596485138,
+      "learning_rate": 3.4934853420195435e-05,
+      "loss": 0.6483,
+      "step": 1223
+    },
+    {
+      "epoch": 2.6886326194398684,
+      "grad_norm": 0.4012199640274048,
+      "learning_rate": 3.469055374592834e-05,
+      "loss": 0.6399,
+      "step": 1224
+    },
+    {
+      "epoch": 2.6908292147171884,
+      "grad_norm": 0.32689276337623596,
+      "learning_rate": 3.4446254071661234e-05,
+      "loss": 0.6704,
+      "step": 1225
+    },
+    {
+      "epoch": 2.6930258099945084,
+      "grad_norm": 0.3250572979450226,
+      "learning_rate": 3.420195439739413e-05,
+      "loss": 0.4988,
+      "step": 1226
+    },
+    {
+      "epoch": 2.6952224052718288,
+      "grad_norm": 0.4365735352039337,
+      "learning_rate": 3.395765472312703e-05,
+      "loss": 0.6192,
+      "step": 1227
+    },
+    {
+      "epoch": 2.6974190005491487,
+      "grad_norm": 0.28177183866500854,
+      "learning_rate": 3.371335504885993e-05,
+      "loss": 0.565,
+      "step": 1228
+    },
+    {
+      "epoch": 2.6996155958264687,
+      "grad_norm": 0.43977609276771545,
+      "learning_rate": 3.346905537459283e-05,
+      "loss": 0.6414,
+      "step": 1229
+    },
+    {
+      "epoch": 2.701812191103789,
+      "grad_norm": 0.41717398166656494,
+      "learning_rate": 3.322475570032573e-05,
+      "loss": 0.466,
+      "step": 1230
+    },
+    {
+      "epoch": 2.7040087863811095,
+      "grad_norm": 0.24823954701423645,
+      "learning_rate": 3.298045602605863e-05,
+      "loss": 0.5045,
+      "step": 1231
+    },
+    {
+      "epoch": 2.7062053816584295,
+      "grad_norm": 0.38355401158332825,
+      "learning_rate": 3.273615635179153e-05,
+      "loss": 0.6763,
+      "step": 1232
+    },
+    {
+      "epoch": 2.7084019769357495,
+      "grad_norm": 0.32537418603897095,
+      "learning_rate": 3.249185667752443e-05,
+      "loss": 0.5307,
+      "step": 1233
+    },
+    {
+      "epoch": 2.71059857221307,
+      "grad_norm": 0.4678703248500824,
+      "learning_rate": 3.224755700325733e-05,
+      "loss": 0.6079,
+      "step": 1234
+    },
+    {
+      "epoch": 2.71279516749039,
+      "grad_norm": 0.3226070702075958,
+      "learning_rate": 3.200325732899023e-05,
+      "loss": 0.7202,
+      "step": 1235
+    },
+    {
+      "epoch": 2.71499176276771,
+      "grad_norm": 0.3895830512046814,
+      "learning_rate": 3.175895765472312e-05,
+      "loss": 0.6611,
+      "step": 1236
+    },
+    {
+      "epoch": 2.7171883580450302,
+      "grad_norm": 0.31069839000701904,
+      "learning_rate": 3.1514657980456026e-05,
+      "loss": 0.5791,
+      "step": 1237
+    },
+    {
+      "epoch": 2.71938495332235,
+      "grad_norm": 0.4009678065776825,
+      "learning_rate": 3.127035830618892e-05,
+      "loss": 0.609,
+      "step": 1238
+    },
+    {
+      "epoch": 2.7215815485996706,
+      "grad_norm": 0.448982834815979,
+      "learning_rate": 3.1026058631921825e-05,
+      "loss": 0.6248,
+      "step": 1239
+    },
+    {
+      "epoch": 2.7237781438769906,
+      "grad_norm": 0.31311914324760437,
+      "learning_rate": 3.078175895765472e-05,
+      "loss": 0.5605,
+      "step": 1240
+    },
+    {
+      "epoch": 2.725974739154311,
+      "grad_norm": 0.39911043643951416,
+      "learning_rate": 3.0537459283387624e-05,
+      "loss": 0.4456,
+      "step": 1241
+    },
+    {
+      "epoch": 2.728171334431631,
+      "grad_norm": 0.3986220359802246,
+      "learning_rate": 3.029315960912052e-05,
+      "loss": 0.4989,
+      "step": 1242
+    },
+    {
+      "epoch": 2.730367929708951,
+      "grad_norm": 0.3716065585613251,
+      "learning_rate": 3.004885993485342e-05,
+      "loss": 0.6724,
+      "step": 1243
+    },
+    {
+      "epoch": 2.7325645249862713,
+      "grad_norm": 0.4374796450138092,
+      "learning_rate": 2.9804560260586318e-05,
+      "loss": 0.6144,
+      "step": 1244
+    },
+    {
+      "epoch": 2.7347611202635913,
+      "grad_norm": 0.39002498984336853,
+      "learning_rate": 2.9560260586319218e-05,
+      "loss": 0.5604,
+      "step": 1245
+    },
+    {
+      "epoch": 2.7369577155409117,
+      "grad_norm": 0.4173770248889923,
+      "learning_rate": 2.9315960912052114e-05,
+      "loss": 0.6033,
+      "step": 1246
+    },
+    {
+      "epoch": 2.7391543108182317,
+      "grad_norm": 0.44223037362098694,
+      "learning_rate": 2.9071661237785013e-05,
+      "loss": 0.7887,
+      "step": 1247
+    },
+    {
+      "epoch": 2.741350906095552,
+      "grad_norm": 0.3783401548862457,
+      "learning_rate": 2.8827361563517912e-05,
+      "loss": 0.5745,
+      "step": 1248
+    },
+    {
+      "epoch": 2.743547501372872,
+      "grad_norm": 0.35366612672805786,
+      "learning_rate": 2.858306188925081e-05,
+      "loss": 0.7071,
+      "step": 1249
+    },
+    {
+      "epoch": 2.745744096650192,
+      "grad_norm": 0.31281355023384094,
+      "learning_rate": 2.833876221498371e-05,
+      "loss": 0.6347,
+      "step": 1250
+    },
+    {
+      "epoch": 2.7479406919275124,
+      "grad_norm": 0.31258246302604675,
+      "learning_rate": 2.809446254071661e-05,
+      "loss": 0.717,
+      "step": 1251
+    },
+    {
+      "epoch": 2.7501372872048324,
+      "grad_norm": 0.4531727135181427,
+      "learning_rate": 2.785016286644951e-05,
+      "loss": 0.5994,
+      "step": 1252
+    },
+    {
+      "epoch": 2.752333882482153,
+      "grad_norm": 0.36134952306747437,
+      "learning_rate": 2.760586319218241e-05,
+      "loss": 0.4816,
+      "step": 1253
+    },
+    {
+      "epoch": 2.754530477759473,
+      "grad_norm": 0.37271133065223694,
+      "learning_rate": 2.7361563517915305e-05,
+      "loss": 0.5398,
+      "step": 1254
+    },
+    {
+      "epoch": 2.756727073036793,
+      "grad_norm": 0.33823123574256897,
+      "learning_rate": 2.7117263843648204e-05,
+      "loss": 0.726,
+      "step": 1255
+    },
+    {
+      "epoch": 2.758923668314113,
+      "grad_norm": 0.32507631182670593,
+      "learning_rate": 2.6872964169381104e-05,
+      "loss": 0.4293,
+      "step": 1256
+    },
+    {
+      "epoch": 2.761120263591433,
+      "grad_norm": 0.3639959394931793,
+      "learning_rate": 2.6628664495114003e-05,
+      "loss": 0.5658,
+      "step": 1257
+    },
+    {
+      "epoch": 2.7633168588687536,
+      "grad_norm": 0.3325282335281372,
+      "learning_rate": 2.6384364820846903e-05,
+      "loss": 0.4962,
+      "step": 1258
+    },
+    {
+      "epoch": 2.7655134541460735,
+      "grad_norm": 0.4300753176212311,
+      "learning_rate": 2.6140065146579802e-05,
+      "loss": 0.7064,
+      "step": 1259
+    },
+    {
+      "epoch": 2.7677100494233935,
+      "grad_norm": 0.38355934619903564,
+      "learning_rate": 2.58957654723127e-05,
+      "loss": 0.6407,
+      "step": 1260
+    },
+    {
+      "epoch": 2.769906644700714,
+      "grad_norm": 0.3721751570701599,
+      "learning_rate": 2.56514657980456e-05,
+      "loss": 0.5351,
+      "step": 1261
+    },
+    {
+      "epoch": 2.7721032399780343,
+      "grad_norm": 0.4482388496398926,
+      "learning_rate": 2.5407166123778497e-05,
+      "loss": 0.5606,
+      "step": 1262
+    },
+    {
+      "epoch": 2.7742998352553543,
+      "grad_norm": 0.5166032314300537,
+      "learning_rate": 2.5162866449511396e-05,
+      "loss": 0.6381,
+      "step": 1263
+    },
+    {
+      "epoch": 2.7764964305326743,
+      "grad_norm": 0.31258314847946167,
+      "learning_rate": 2.4918566775244295e-05,
+      "loss": 0.7673,
+      "step": 1264
+    },
+    {
+      "epoch": 2.7786930258099947,
+      "grad_norm": 0.284380167722702,
+      "learning_rate": 2.4674267100977195e-05,
+      "loss": 0.7015,
+      "step": 1265
+    },
+    {
+      "epoch": 2.7808896210873146,
+      "grad_norm": 0.2771058678627014,
+      "learning_rate": 2.4429967426710094e-05,
+      "loss": 0.5066,
+      "step": 1266
+    },
+    {
+      "epoch": 2.7830862163646346,
+      "grad_norm": 0.41136935353279114,
+      "learning_rate": 2.4185667752442993e-05,
+      "loss": 0.5091,
+      "step": 1267
+    },
+    {
+      "epoch": 2.785282811641955,
+      "grad_norm": 0.3271212875843048,
+      "learning_rate": 2.3941368078175893e-05,
+      "loss": 0.6221,
+      "step": 1268
+    },
+    {
+      "epoch": 2.787479406919275,
+      "grad_norm": 0.40301135182380676,
+      "learning_rate": 2.3697068403908792e-05,
+      "loss": 0.5041,
+      "step": 1269
+    },
+    {
+      "epoch": 2.7896760021965954,
+      "grad_norm": 0.4785580635070801,
+      "learning_rate": 2.3452768729641688e-05,
+      "loss": 0.4977,
+      "step": 1270
+    },
+    {
+      "epoch": 2.7918725974739154,
+      "grad_norm": 0.3381398022174835,
+      "learning_rate": 2.3208469055374594e-05,
+      "loss": 0.5375,
+      "step": 1271
+    },
+    {
+      "epoch": 2.7940691927512358,
+      "grad_norm": 0.40477830171585083,
+      "learning_rate": 2.296416938110749e-05,
+      "loss": 0.6236,
+      "step": 1272
+    },
+    {
+      "epoch": 2.7962657880285557,
+      "grad_norm": 0.3422625958919525,
+      "learning_rate": 2.271986970684039e-05,
+      "loss": 0.6258,
+      "step": 1273
+    },
+    {
+      "epoch": 2.7984623833058757,
+      "grad_norm": 0.339061975479126,
+      "learning_rate": 2.247557003257329e-05,
+      "loss": 0.748,
+      "step": 1274
+    },
+    {
+      "epoch": 2.800658978583196,
+      "grad_norm": 0.43954479694366455,
+      "learning_rate": 2.223127035830619e-05,
+      "loss": 0.5707,
+      "step": 1275
+    },
+    {
+      "epoch": 2.802855573860516,
+      "grad_norm": 0.44815370440483093,
+      "learning_rate": 2.1986970684039088e-05,
+      "loss": 0.7244,
+      "step": 1276
+    },
+    {
+      "epoch": 2.8050521691378365,
+      "grad_norm": 0.48473912477493286,
+      "learning_rate": 2.1742671009771987e-05,
+      "loss": 0.808,
+      "step": 1277
+    },
+    {
+      "epoch": 2.8072487644151565,
+      "grad_norm": 0.4030792713165283,
+      "learning_rate": 2.1498371335504886e-05,
+      "loss": 0.6473,
+      "step": 1278
+    },
+    {
+      "epoch": 2.809445359692477,
+      "grad_norm": 0.3813606798648834,
+      "learning_rate": 2.1254071661237786e-05,
+      "loss": 0.5758,
+      "step": 1279
+    },
+    {
+      "epoch": 2.811641954969797,
+      "grad_norm": 0.29771262407302856,
+      "learning_rate": 2.1009771986970682e-05,
+      "loss": 0.573,
+      "step": 1280
+    },
+    {
+      "epoch": 2.813838550247117,
+      "grad_norm": 0.39586323499679565,
+      "learning_rate": 2.076547231270358e-05,
+      "loss": 0.6021,
+      "step": 1281
+    },
+    {
+      "epoch": 2.8160351455244372,
+      "grad_norm": 0.4622383415699005,
+      "learning_rate": 2.052117263843648e-05,
+      "loss": 0.6241,
+      "step": 1282
+    },
+    {
+      "epoch": 2.818231740801757,
+      "grad_norm": 0.42414385080337524,
+      "learning_rate": 2.027687296416938e-05,
+      "loss": 0.7131,
+      "step": 1283
+    },
+    {
+      "epoch": 2.820428336079077,
+      "grad_norm": 0.3358052372932434,
+      "learning_rate": 2.003257328990228e-05,
+      "loss": 0.6692,
+      "step": 1284
+    },
+    {
+      "epoch": 2.8226249313563976,
+      "grad_norm": 0.3923681080341339,
+      "learning_rate": 1.978827361563518e-05,
+      "loss": 0.6248,
+      "step": 1285
+    },
+    {
+      "epoch": 2.824821526633718,
+      "grad_norm": 0.4191087782382965,
+      "learning_rate": 1.9543973941368078e-05,
+      "loss": 0.791,
+      "step": 1286
+    },
+    {
+      "epoch": 2.827018121911038,
+      "grad_norm": 0.3246961832046509,
+      "learning_rate": 1.9299674267100977e-05,
+      "loss": 0.5202,
+      "step": 1287
+    },
+    {
+      "epoch": 2.829214717188358,
+      "grad_norm": 0.49594706296920776,
+      "learning_rate": 1.9055374592833873e-05,
+      "loss": 0.5779,
+      "step": 1288
+    },
+    {
+      "epoch": 2.8314113124656783,
+      "grad_norm": 0.4633212089538574,
+      "learning_rate": 1.8811074918566773e-05,
+      "loss": 0.603,
+      "step": 1289
+    },
+    {
+      "epoch": 2.8336079077429983,
+      "grad_norm": 0.4270968735218048,
+      "learning_rate": 1.8566775244299672e-05,
+      "loss": 0.6617,
+      "step": 1290
+    },
+    {
+      "epoch": 2.8358045030203183,
+      "grad_norm": 0.4147087633609772,
+      "learning_rate": 1.832247557003257e-05,
+      "loss": 0.5361,
+      "step": 1291
+    },
+    {
+      "epoch": 2.8380010982976387,
+      "grad_norm": 0.3119422197341919,
+      "learning_rate": 1.807817589576547e-05,
+      "loss": 0.5883,
+      "step": 1292
+    },
+    {
+      "epoch": 2.8401976935749587,
+      "grad_norm": 0.3216208517551422,
+      "learning_rate": 1.783387622149837e-05,
+      "loss": 0.4242,
+      "step": 1293
+    },
+    {
+      "epoch": 2.842394288852279,
+      "grad_norm": 0.34990546107292175,
+      "learning_rate": 1.758957654723127e-05,
+      "loss": 0.647,
+      "step": 1294
+    },
+    {
+      "epoch": 2.844590884129599,
+      "grad_norm": 0.4455469846725464,
+      "learning_rate": 1.734527687296417e-05,
+      "loss": 0.6171,
+      "step": 1295
+    },
+    {
+      "epoch": 2.8467874794069195,
+      "grad_norm": 0.3917944133281708,
+      "learning_rate": 1.7100977198697065e-05,
+      "loss": 0.6752,
+      "step": 1296
+    },
+    {
+      "epoch": 2.8489840746842394,
+      "grad_norm": 0.28565680980682373,
+      "learning_rate": 1.6856677524429964e-05,
+      "loss": 0.5244,
+      "step": 1297
+    },
+    {
+      "epoch": 2.8511806699615594,
+      "grad_norm": 0.36126193404197693,
+      "learning_rate": 1.6612377850162864e-05,
+      "loss": 0.6637,
+      "step": 1298
+    },
+    {
+      "epoch": 2.85337726523888,
+      "grad_norm": 0.45588603615760803,
+      "learning_rate": 1.6368078175895766e-05,
+      "loss": 0.5985,
+      "step": 1299
+    },
+    {
+      "epoch": 2.8555738605161998,
+      "grad_norm": 0.4465961754322052,
+      "learning_rate": 1.6123778501628666e-05,
+      "loss": 0.7739,
+      "step": 1300
+    },
+    {
+      "epoch": 2.85777045579352,
+      "grad_norm": 0.319690078496933,
+      "learning_rate": 1.587947882736156e-05,
+      "loss": 0.6263,
+      "step": 1301
+    },
+    {
+      "epoch": 2.85996705107084,
+      "grad_norm": 0.30340343713760376,
+      "learning_rate": 1.563517915309446e-05,
+      "loss": 0.537,
+      "step": 1302
+    },
+    {
+      "epoch": 2.8621636463481606,
+      "grad_norm": 0.341239333152771,
+      "learning_rate": 1.539087947882736e-05,
+      "loss": 0.7815,
+      "step": 1303
+    },
+    {
+      "epoch": 2.8643602416254805,
+      "grad_norm": 0.3735162615776062,
+      "learning_rate": 1.514657980456026e-05,
+      "loss": 0.518,
+      "step": 1304
+    },
+    {
+      "epoch": 2.8665568369028005,
+      "grad_norm": 0.41930490732192993,
+      "learning_rate": 1.4902280130293159e-05,
+      "loss": 0.6013,
+      "step": 1305
+    },
+    {
+      "epoch": 2.868753432180121,
+      "grad_norm": 0.3290257453918457,
+      "learning_rate": 1.4657980456026057e-05,
+      "loss": 0.6015,
+      "step": 1306
+    },
+    {
+      "epoch": 2.870950027457441,
+      "grad_norm": 0.3936542272567749,
+      "learning_rate": 1.4413680781758956e-05,
+      "loss": 0.6591,
+      "step": 1307
+    },
+    {
+      "epoch": 2.873146622734761,
+      "grad_norm": 0.3079110085964203,
+      "learning_rate": 1.4169381107491856e-05,
+      "loss": 0.4738,
+      "step": 1308
+    },
+    {
+      "epoch": 2.8753432180120813,
+      "grad_norm": 0.3470717966556549,
+      "learning_rate": 1.3925081433224755e-05,
+      "loss": 0.5711,
+      "step": 1309
+    },
+    {
+      "epoch": 2.8775398132894017,
+      "grad_norm": 0.4416014552116394,
+      "learning_rate": 1.3680781758957653e-05,
+      "loss": 0.6302,
+      "step": 1310
+    },
+    {
+      "epoch": 2.8797364085667216,
+      "grad_norm": 0.3373110592365265,
+      "learning_rate": 1.3436482084690552e-05,
+      "loss": 0.5813,
+      "step": 1311
+    },
+    {
+      "epoch": 2.8819330038440416,
+      "grad_norm": 0.38170263171195984,
+      "learning_rate": 1.3192182410423451e-05,
+      "loss": 0.6429,
+      "step": 1312
+    },
+    {
+      "epoch": 2.884129599121362,
+      "grad_norm": 0.44573941826820374,
+      "learning_rate": 1.294788273615635e-05,
+      "loss": 0.6255,
+      "step": 1313
+    },
+    {
+      "epoch": 2.886326194398682,
+      "grad_norm": 0.3631455898284912,
+      "learning_rate": 1.2703583061889248e-05,
+      "loss": 0.617,
+      "step": 1314
+    },
+    {
+      "epoch": 2.888522789676002,
+      "grad_norm": 0.4141990542411804,
+      "learning_rate": 1.2459283387622148e-05,
+      "loss": 0.5328,
+      "step": 1315
+    },
+    {
+      "epoch": 2.8907193849533224,
+      "grad_norm": 0.46746328473091125,
+      "learning_rate": 1.2214983713355047e-05,
+      "loss": 0.6221,
+      "step": 1316
+    },
+    {
+      "epoch": 2.892915980230643,
+      "grad_norm": 0.49625125527381897,
+      "learning_rate": 1.1970684039087946e-05,
+      "loss": 0.4914,
+      "step": 1317
+    },
+    {
+      "epoch": 2.8951125755079627,
+      "grad_norm": 0.3066498339176178,
+      "learning_rate": 1.1726384364820844e-05,
+      "loss": 0.4353,
+      "step": 1318
+    },
+    {
+      "epoch": 2.8973091707852827,
+      "grad_norm": 0.3838805556297302,
+      "learning_rate": 1.1482084690553745e-05,
+      "loss": 0.465,
+      "step": 1319
+    },
+    {
+      "epoch": 2.899505766062603,
+      "grad_norm": 0.31246620416641235,
+      "learning_rate": 1.1237785016286644e-05,
+      "loss": 0.5914,
+      "step": 1320
+    },
+    {
+      "epoch": 2.901702361339923,
+      "grad_norm": 0.3723578453063965,
+      "learning_rate": 1.0993485342019544e-05,
+      "loss": 0.438,
+      "step": 1321
+    },
+    {
+      "epoch": 2.903898956617243,
+      "grad_norm": 0.5644330978393555,
+      "learning_rate": 1.0749185667752443e-05,
+      "loss": 0.5884,
+      "step": 1322
+    },
+    {
+      "epoch": 2.9060955518945635,
+      "grad_norm": 0.42743244767189026,
+      "learning_rate": 1.0504885993485341e-05,
+      "loss": 0.5378,
+      "step": 1323
+    },
+    {
+      "epoch": 2.9082921471718834,
+      "grad_norm": 0.5095661878585815,
+      "learning_rate": 1.026058631921824e-05,
+      "loss": 0.563,
+      "step": 1324
+    },
+    {
+      "epoch": 2.910488742449204,
+      "grad_norm": 0.2788356840610504,
+      "learning_rate": 1.001628664495114e-05,
+      "loss": 0.4719,
+      "step": 1325
+    },
+    {
+      "epoch": 2.912685337726524,
+      "grad_norm": 0.3998756408691406,
+      "learning_rate": 9.771986970684039e-06,
+      "loss": 0.5757,
+      "step": 1326
+    },
+    {
+      "epoch": 2.9148819330038442,
+      "grad_norm": 0.3741336464881897,
+      "learning_rate": 9.527687296416937e-06,
+      "loss": 0.6131,
+      "step": 1327
+    },
+    {
+      "epoch": 2.917078528281164,
+      "grad_norm": 0.3467644453048706,
+      "learning_rate": 9.283387622149836e-06,
+      "loss": 0.4352,
+      "step": 1328
+    },
+    {
+      "epoch": 2.919275123558484,
+      "grad_norm": 0.326631635427475,
+      "learning_rate": 9.039087947882735e-06,
+      "loss": 0.3992,
+      "step": 1329
+    },
+    {
+      "epoch": 2.9214717188358046,
+      "grad_norm": 0.35358214378356934,
+      "learning_rate": 8.794788273615635e-06,
+      "loss": 0.4986,
+      "step": 1330
+    },
+    {
+      "epoch": 2.9236683141131246,
+      "grad_norm": 0.4555149972438812,
+      "learning_rate": 8.550488599348532e-06,
+      "loss": 0.7148,
+      "step": 1331
+    },
+    {
+      "epoch": 2.925864909390445,
+      "grad_norm": 0.33451521396636963,
+      "learning_rate": 8.306188925081432e-06,
+      "loss": 0.3274,
+      "step": 1332
+    },
+    {
+      "epoch": 2.928061504667765,
+      "grad_norm": 0.5291953086853027,
+      "learning_rate": 8.061889250814333e-06,
+      "loss": 0.475,
+      "step": 1333
+    },
+    {
+      "epoch": 2.9302580999450853,
+      "grad_norm": 0.27760300040245056,
+      "learning_rate": 7.81758957654723e-06,
+      "loss": 0.5093,
+      "step": 1334
+    },
+    {
+      "epoch": 2.9324546952224053,
+      "grad_norm": 0.35058730840682983,
+      "learning_rate": 7.57328990228013e-06,
+      "loss": 0.4649,
+      "step": 1335
+    },
+    {
+      "epoch": 2.9346512904997253,
+      "grad_norm": 0.37483546137809753,
+      "learning_rate": 7.328990228013028e-06,
+      "loss": 0.4832,
+      "step": 1336
+    },
+    {
+      "epoch": 2.9368478857770457,
+      "grad_norm": 0.28210797905921936,
+      "learning_rate": 7.084690553745928e-06,
+      "loss": 0.488,
+      "step": 1337
+    },
+    {
+      "epoch": 2.9390444810543657,
+      "grad_norm": 0.30135488510131836,
+      "learning_rate": 6.840390879478826e-06,
+      "loss": 0.5468,
+      "step": 1338
+    },
+    {
+      "epoch": 2.9412410763316856,
+      "grad_norm": 0.3030971586704254,
+      "learning_rate": 6.596091205211726e-06,
+      "loss": 0.5726,
+      "step": 1339
+    },
+    {
+      "epoch": 2.943437671609006,
+      "grad_norm": 0.31649550795555115,
+      "learning_rate": 6.351791530944624e-06,
+      "loss": 0.4053,
+      "step": 1340
+    },
+    {
+      "epoch": 2.9456342668863265,
+      "grad_norm": 0.3474193215370178,
+      "learning_rate": 6.1074918566775235e-06,
+      "loss": 0.5773,
+      "step": 1341
+    },
+    {
+      "epoch": 2.9478308621636464,
+      "grad_norm": 0.2980194687843323,
+      "learning_rate": 5.863192182410422e-06,
+      "loss": 0.6185,
+      "step": 1342
+    },
+    {
+      "epoch": 2.9500274574409664,
+      "grad_norm": 0.40653666853904724,
+      "learning_rate": 5.618892508143322e-06,
+      "loss": 0.738,
+      "step": 1343
+    },
+    {
+      "epoch": 2.952224052718287,
+      "grad_norm": 0.37222039699554443,
+      "learning_rate": 5.374592833876222e-06,
+      "loss": 0.6942,
+      "step": 1344
+    },
+    {
+      "epoch": 2.9544206479956068,
+      "grad_norm": 0.39097172021865845,
+      "learning_rate": 5.13029315960912e-06,
+      "loss": 0.6028,
+      "step": 1345
+    },
+    {
+      "epoch": 2.9566172432729267,
+      "grad_norm": 0.49179306626319885,
+      "learning_rate": 4.8859934853420195e-06,
+      "loss": 0.4885,
+      "step": 1346
+    },
+    {
+      "epoch": 2.958813838550247,
+      "grad_norm": 0.342683881521225,
+      "learning_rate": 4.641693811074918e-06,
+      "loss": 0.4749,
+      "step": 1347
+    },
+    {
+      "epoch": 2.961010433827567,
+      "grad_norm": 0.44037333130836487,
+      "learning_rate": 4.397394136807817e-06,
+      "loss": 0.4794,
+      "step": 1348
+    },
+    {
+      "epoch": 2.9632070291048875,
+      "grad_norm": 0.3877731263637543,
+      "learning_rate": 4.153094462540716e-06,
+      "loss": 0.7047,
+      "step": 1349
+    },
+    {
+      "epoch": 2.9654036243822075,
+      "grad_norm": 0.4986109137535095,
+      "learning_rate": 3.908794788273615e-06,
+      "loss": 0.532,
+      "step": 1350
+    },
+    {
+      "epoch": 2.967600219659528,
+      "grad_norm": 0.32422760128974915,
+      "learning_rate": 3.664495114006514e-06,
+      "loss": 0.7126,
+      "step": 1351
+    },
+    {
+      "epoch": 2.969796814936848,
+      "grad_norm": 0.3965681493282318,
+      "learning_rate": 3.420195439739413e-06,
+      "loss": 0.5441,
+      "step": 1352
+    },
+    {
+      "epoch": 2.971993410214168,
+      "grad_norm": 0.3819817900657654,
+      "learning_rate": 3.175895765472312e-06,
+      "loss": 0.6599,
+      "step": 1353
+    },
+    {
+      "epoch": 2.9741900054914883,
+      "grad_norm": 0.3484026789665222,
+      "learning_rate": 2.931596091205211e-06,
+      "loss": 0.6935,
+      "step": 1354
+    },
+    {
+      "epoch": 2.9763866007688082,
+      "grad_norm": 0.39374682307243347,
+      "learning_rate": 2.687296416938111e-06,
+      "loss": 0.4945,
+      "step": 1355
+    },
+    {
+      "epoch": 2.9785831960461286,
+      "grad_norm": 0.5244062542915344,
+      "learning_rate": 2.4429967426710097e-06,
+      "loss": 0.5717,
+      "step": 1356
+    },
+    {
+      "epoch": 2.9807797913234486,
+      "grad_norm": 0.3763648271560669,
+      "learning_rate": 2.1986970684039087e-06,
+      "loss": 0.6156,
+      "step": 1357
+    },
+    {
+      "epoch": 2.982976386600769,
+      "grad_norm": 0.5366123914718628,
+      "learning_rate": 1.9543973941368076e-06,
+      "loss": 0.4156,
+      "step": 1358
+    },
+    {
+      "epoch": 2.985172981878089,
+      "grad_norm": 0.35754984617233276,
+      "learning_rate": 1.7100977198697066e-06,
+      "loss": 0.7282,
+      "step": 1359
+    },
+    {
+      "epoch": 2.987369577155409,
+      "grad_norm": 0.3933640718460083,
+      "learning_rate": 1.4657980456026055e-06,
+      "loss": 0.533,
+      "step": 1360
+    },
+    {
+      "epoch": 2.9895661724327294,
+      "grad_norm": 0.3363213539123535,
+      "learning_rate": 1.2214983713355049e-06,
+      "loss": 0.6006,
+      "step": 1361
+    },
+    {
+      "epoch": 2.9917627677100493,
+      "grad_norm": 0.49599722027778625,
+      "learning_rate": 9.771986970684038e-07,
+      "loss": 0.6247,
+      "step": 1362
+    },
+    {
+      "epoch": 2.9939593629873693,
+      "grad_norm": 0.443291574716568,
+      "learning_rate": 7.328990228013028e-07,
+      "loss": 0.5244,
+      "step": 1363
+    },
+    {
+      "epoch": 2.9961559582646897,
+      "grad_norm": 0.35655391216278076,
+      "learning_rate": 4.885993485342019e-07,
+      "loss": 0.6374,
+      "step": 1364
+    },
+    {
+      "epoch": 2.99835255354201,
+      "grad_norm": 0.3855736255645752,
+      "learning_rate": 2.4429967426710095e-07,
+      "loss": 0.6764,
+      "step": 1365
+    },
+    {
+      "epoch": 2.99835255354201,
+      "step": 1365,
+      "total_flos": 1.0043491055199846e+18,
+      "train_loss": 0.6286539919647105,
+      "train_runtime": 5254.6218,
+      "train_samples_per_second": 4.158,
+      "train_steps_per_second": 0.26
+    }
+  ],
+  "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": 1.0043491055199846e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}