diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,9492 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9991668980838657,
+  "eval_steps": 500,
+  "global_step": 1350,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0022216051096917524,
+      "grad_norm": 0.010440913029015064,
+      "learning_rate": 2.2222222222222223e-05,
+      "loss": 0.8403,
+      "step": 1
+    },
+    {
+      "epoch": 0.004443210219383505,
+      "grad_norm": 0.007241299841552973,
+      "learning_rate": 4.4444444444444447e-05,
+      "loss": 0.6888,
+      "step": 2
+    },
+    {
+      "epoch": 0.006664815329075257,
+      "grad_norm": 0.005549824796617031,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 0.966,
+      "step": 3
+    },
+    {
+      "epoch": 0.00888642043876701,
+      "grad_norm": 0.008592728525400162,
+      "learning_rate": 8.888888888888889e-05,
+      "loss": 0.8665,
+      "step": 4
+    },
+    {
+      "epoch": 0.011108025548458762,
+      "grad_norm": 0.007739076856523752,
+      "learning_rate": 0.0001111111111111111,
+      "loss": 0.7687,
+      "step": 5
+    },
+    {
+      "epoch": 0.013329630658150514,
+      "grad_norm": 0.008957313373684883,
+      "learning_rate": 0.00013333333333333334,
+      "loss": 0.8916,
+      "step": 6
+    },
+    {
+      "epoch": 0.015551235767842266,
+      "grad_norm": 0.009202866815030575,
+      "learning_rate": 0.00015555555555555556,
+      "loss": 0.8653,
+      "step": 7
+    },
+    {
+      "epoch": 0.01777284087753402,
+      "grad_norm": 0.007248671259731054,
+      "learning_rate": 0.00017777777777777779,
+      "loss": 0.861,
+      "step": 8
+    },
+    {
+      "epoch": 0.01999444598722577,
+      "grad_norm": 0.007237965241074562,
+      "learning_rate": 0.0002,
+      "loss": 0.946,
+      "step": 9
+    },
+    {
+      "epoch": 0.022216051096917523,
+      "grad_norm": 0.006773823872208595,
+      "learning_rate": 0.0002222222222222222,
+      "loss": 0.8272,
+      "step": 10
+    },
+    {
+      "epoch": 0.024437656206609277,
+      "grad_norm": 0.004907544236630201,
+      "learning_rate": 0.0002444444444444445,
+      "loss": 0.7749,
+      "step": 11
+    },
+    {
+      "epoch": 0.026659261316301027,
+      "grad_norm": 0.006352128926664591,
+      "learning_rate": 0.0002666666666666667,
+      "loss": 0.8051,
+      "step": 12
+    },
+    {
+      "epoch": 0.02888086642599278,
+      "grad_norm": 0.007070545107126236,
+      "learning_rate": 0.0002888888888888889,
+      "loss": 0.729,
+      "step": 13
+    },
+    {
+      "epoch": 0.03110247153568453,
+      "grad_norm": 0.014602522365748882,
+      "learning_rate": 0.0003111111111111111,
+      "loss": 0.872,
+      "step": 14
+    },
+    {
+      "epoch": 0.03332407664537628,
+      "grad_norm": 0.006523902527987957,
+      "learning_rate": 0.0003333333333333333,
+      "loss": 0.8664,
+      "step": 15
+    },
+    {
+      "epoch": 0.03554568175506804,
+      "grad_norm": 0.005539918318390846,
+      "learning_rate": 0.00035555555555555557,
+      "loss": 0.8333,
+      "step": 16
+    },
+    {
+      "epoch": 0.03776728686475979,
+      "grad_norm": 0.008334115147590637,
+      "learning_rate": 0.00037777777777777777,
+      "loss": 0.8471,
+      "step": 17
+    },
+    {
+      "epoch": 0.03998889197445154,
+      "grad_norm": 0.009435688145458698,
+      "learning_rate": 0.0004,
+      "loss": 0.9297,
+      "step": 18
+    },
+    {
+      "epoch": 0.042210497084143296,
+      "grad_norm": 0.010521095246076584,
+      "learning_rate": 0.00042222222222222227,
+      "loss": 0.9572,
+      "step": 19
+    },
+    {
+      "epoch": 0.044432102193835046,
+      "grad_norm": 0.008772826753556728,
+      "learning_rate": 0.0004444444444444444,
+      "loss": 1.0062,
+      "step": 20
+    },
+    {
+      "epoch": 0.0466537073035268,
+      "grad_norm": 0.006737489253282547,
+      "learning_rate": 0.00046666666666666666,
+      "loss": 0.699,
+      "step": 21
+    },
+    {
+      "epoch": 0.048875312413218554,
+      "grad_norm": 0.003660662332549691,
+      "learning_rate": 0.000488888888888889,
+      "loss": 0.7278,
+      "step": 22
+    },
+    {
+      "epoch": 0.051096917522910304,
+      "grad_norm": 0.007398749236017466,
+      "learning_rate": 0.0005111111111111112,
+      "loss": 0.8316,
+      "step": 23
+    },
+    {
+      "epoch": 0.053318522632602054,
+      "grad_norm": 0.006491729989647865,
+      "learning_rate": 0.0005333333333333334,
+      "loss": 0.8999,
+      "step": 24
+    },
+    {
+      "epoch": 0.055540127742293804,
+      "grad_norm": 0.007386940531432629,
+      "learning_rate": 0.0005555555555555556,
+      "loss": 0.8758,
+      "step": 25
+    },
+    {
+      "epoch": 0.05776173285198556,
+      "grad_norm": 0.006012436002492905,
+      "learning_rate": 0.0005777777777777778,
+      "loss": 0.9802,
+      "step": 26
+    },
+    {
+      "epoch": 0.05998333796167731,
+      "grad_norm": 0.007025507278740406,
+      "learning_rate": 0.0006000000000000001,
+      "loss": 0.7958,
+      "step": 27
+    },
+    {
+      "epoch": 0.06220494307136906,
+      "grad_norm": 0.008443865925073624,
+      "learning_rate": 0.0006222222222222223,
+      "loss": 1.0211,
+      "step": 28
+    },
+    {
+      "epoch": 0.06442654818106082,
+      "grad_norm": 0.007126843556761742,
+      "learning_rate": 0.0006444444444444444,
+      "loss": 0.9817,
+      "step": 29
+    },
+    {
+      "epoch": 0.06664815329075256,
+      "grad_norm": 0.007003214210271835,
+      "learning_rate": 0.0006666666666666666,
+      "loss": 0.8581,
+      "step": 30
+    },
+    {
+      "epoch": 0.06886975840044432,
+      "grad_norm": 0.00639823405072093,
+      "learning_rate": 0.000688888888888889,
+      "loss": 0.867,
+      "step": 31
+    },
+    {
+      "epoch": 0.07109136351013608,
+      "grad_norm": 0.006634861696511507,
+      "learning_rate": 0.0007111111111111111,
+      "loss": 0.9614,
+      "step": 32
+    },
+    {
+      "epoch": 0.07331296861982782,
+      "grad_norm": 0.006281793583184481,
+      "learning_rate": 0.0007333333333333333,
+      "loss": 0.8645,
+      "step": 33
+    },
+    {
+      "epoch": 0.07553457372951958,
+      "grad_norm": 0.009016958996653557,
+      "learning_rate": 0.0007555555555555555,
+      "loss": 0.9221,
+      "step": 34
+    },
+    {
+      "epoch": 0.07775617883921133,
+      "grad_norm": 0.007040904834866524,
+      "learning_rate": 0.0007777777777777777,
+      "loss": 0.9473,
+      "step": 35
+    },
+    {
+      "epoch": 0.07997778394890308,
+      "grad_norm": 0.008552225306630135,
+      "learning_rate": 0.0008,
+      "loss": 0.7651,
+      "step": 36
+    },
+    {
+      "epoch": 0.08219938905859484,
+      "grad_norm": 0.007091434672474861,
+      "learning_rate": 0.0008222222222222222,
+      "loss": 0.8234,
+      "step": 37
+    },
+    {
+      "epoch": 0.08442099416828659,
+      "grad_norm": 0.009340236894786358,
+      "learning_rate": 0.0008444444444444445,
+      "loss": 0.7472,
+      "step": 38
+    },
+    {
+      "epoch": 0.08664259927797834,
+      "grad_norm": 0.010021640919148922,
+      "learning_rate": 0.0008666666666666666,
+      "loss": 1.039,
+      "step": 39
+    },
+    {
+      "epoch": 0.08886420438767009,
+      "grad_norm": 0.011557168327271938,
+      "learning_rate": 0.0008888888888888888,
+      "loss": 0.8556,
+      "step": 40
+    },
+    {
+      "epoch": 0.09108580949736185,
+      "grad_norm": 0.006138915196061134,
+      "learning_rate": 0.0009111111111111111,
+      "loss": 0.8866,
+      "step": 41
+    },
+    {
+      "epoch": 0.0933074146070536,
+      "grad_norm": 0.005869209300726652,
+      "learning_rate": 0.0009333333333333333,
+      "loss": 0.9876,
+      "step": 42
+    },
+    {
+      "epoch": 0.09552901971674535,
+      "grad_norm": 0.006708982866257429,
+      "learning_rate": 0.0009555555555555556,
+      "loss": 1.0414,
+      "step": 43
+    },
+    {
+      "epoch": 0.09775062482643711,
+      "grad_norm": 0.006898691412061453,
+      "learning_rate": 0.000977777777777778,
+      "loss": 0.826,
+      "step": 44
+    },
+    {
+      "epoch": 0.09997222993612885,
+      "grad_norm": 0.007414904423058033,
+      "learning_rate": 0.001,
+      "loss": 0.7444,
+      "step": 45
+    },
+    {
+      "epoch": 0.10219383504582061,
+      "grad_norm": 0.005232702940702438,
+      "learning_rate": 0.0010222222222222223,
+      "loss": 0.8942,
+      "step": 46
+    },
+    {
+      "epoch": 0.10441544015551235,
+      "grad_norm": 0.006096681114286184,
+      "learning_rate": 0.0010444444444444444,
+      "loss": 1.0935,
+      "step": 47
+    },
+    {
+      "epoch": 0.10663704526520411,
+      "grad_norm": 0.008140468038618565,
+      "learning_rate": 0.0010666666666666667,
+      "loss": 0.8532,
+      "step": 48
+    },
+    {
+      "epoch": 0.10885865037489587,
+      "grad_norm": 0.007381773088127375,
+      "learning_rate": 0.001088888888888889,
+      "loss": 0.869,
+      "step": 49
+    },
+    {
+      "epoch": 0.11108025548458761,
+      "grad_norm": 0.011090748943388462,
+      "learning_rate": 0.0011111111111111111,
+      "loss": 0.782,
+      "step": 50
+    },
+    {
+      "epoch": 0.11330186059427937,
+      "grad_norm": 0.012807847931981087,
+      "learning_rate": 0.0011333333333333334,
+      "loss": 0.9066,
+      "step": 51
+    },
+    {
+      "epoch": 0.11552346570397112,
+      "grad_norm": 0.0077123180963099,
+      "learning_rate": 0.0011555555555555555,
+      "loss": 0.7865,
+      "step": 52
+    },
+    {
+      "epoch": 0.11774507081366287,
+      "grad_norm": 0.006746234837919474,
+      "learning_rate": 0.0011777777777777778,
+      "loss": 0.8238,
+      "step": 53
+    },
+    {
+      "epoch": 0.11996667592335462,
+      "grad_norm": 0.00544540723785758,
+      "learning_rate": 0.0012000000000000001,
+      "loss": 0.8159,
+      "step": 54
+    },
+    {
+      "epoch": 0.12218828103304638,
+      "grad_norm": 0.004280385095626116,
+      "learning_rate": 0.0012222222222222222,
+      "loss": 0.7477,
+      "step": 55
+    },
+    {
+      "epoch": 0.12440988614273812,
+      "grad_norm": 0.006144772749394178,
+      "learning_rate": 0.0012444444444444445,
+      "loss": 0.7885,
+      "step": 56
+    },
+    {
+      "epoch": 0.12663149125242987,
+      "grad_norm": 0.007963410578668118,
+      "learning_rate": 0.0012666666666666666,
+      "loss": 0.8448,
+      "step": 57
+    },
+    {
+      "epoch": 0.12885309636212164,
+      "grad_norm": 0.005704255774617195,
+      "learning_rate": 0.001288888888888889,
+      "loss": 0.7554,
+      "step": 58
+    },
+    {
+      "epoch": 0.13107470147181338,
+      "grad_norm": 0.0052896905690431595,
+      "learning_rate": 0.0013111111111111112,
+      "loss": 0.7285,
+      "step": 59
+    },
+    {
+      "epoch": 0.13329630658150513,
+      "grad_norm": 0.005001865327358246,
+      "learning_rate": 0.0013333333333333333,
+      "loss": 0.9311,
+      "step": 60
+    },
+    {
+      "epoch": 0.1355179116911969,
+      "grad_norm": 0.006869412958621979,
+      "learning_rate": 0.0013555555555555556,
+      "loss": 0.9407,
+      "step": 61
+    },
+    {
+      "epoch": 0.13773951680088864,
+      "grad_norm": 0.009640970267355442,
+      "learning_rate": 0.001377777777777778,
+      "loss": 0.8013,
+      "step": 62
+    },
+    {
+      "epoch": 0.13996112191058038,
+      "grad_norm": 0.009351574815809727,
+      "learning_rate": 0.0014,
+      "loss": 0.9175,
+      "step": 63
+    },
+    {
+      "epoch": 0.14218272702027215,
+      "grad_norm": 0.006775525864213705,
+      "learning_rate": 0.0014222222222222223,
+      "loss": 1.021,
+      "step": 64
+    },
+    {
+      "epoch": 0.1444043321299639,
+      "grad_norm": 0.005892643239349127,
+      "learning_rate": 0.0014444444444444444,
+      "loss": 0.7416,
+      "step": 65
+    },
+    {
+      "epoch": 0.14662593723965564,
+      "grad_norm": 0.008734880946576595,
+      "learning_rate": 0.0014666666666666667,
+      "loss": 0.9335,
+      "step": 66
+    },
+    {
+      "epoch": 0.1488475423493474,
+      "grad_norm": 0.005164718721061945,
+      "learning_rate": 0.001488888888888889,
+      "loss": 1.1011,
+      "step": 67
+    },
+    {
+      "epoch": 0.15106914745903915,
+      "grad_norm": 0.004635809920728207,
+      "learning_rate": 0.001511111111111111,
+      "loss": 0.8691,
+      "step": 68
+    },
+    {
+      "epoch": 0.1532907525687309,
+      "grad_norm": 0.006990849506109953,
+      "learning_rate": 0.0015333333333333332,
+      "loss": 0.9086,
+      "step": 69
+    },
+    {
+      "epoch": 0.15551235767842267,
+      "grad_norm": 0.0074639637023210526,
+      "learning_rate": 0.0015555555555555555,
+      "loss": 0.6275,
+      "step": 70
+    },
+    {
+      "epoch": 0.1577339627881144,
+      "grad_norm": 0.00890981312841177,
+      "learning_rate": 0.0015777777777777778,
+      "loss": 0.7847,
+      "step": 71
+    },
+    {
+      "epoch": 0.15995556789780616,
+      "grad_norm": 0.006609317846596241,
+      "learning_rate": 0.0016,
+      "loss": 0.7674,
+      "step": 72
+    },
+    {
+      "epoch": 0.16217717300749793,
+      "grad_norm": 0.012736441567540169,
+      "learning_rate": 0.0016222222222222222,
+      "loss": 0.8132,
+      "step": 73
+    },
+    {
+      "epoch": 0.16439877811718967,
+      "grad_norm": 0.008025340735912323,
+      "learning_rate": 0.0016444444444444445,
+      "loss": 0.842,
+      "step": 74
+    },
+    {
+      "epoch": 0.1666203832268814,
+      "grad_norm": 0.010593456216156483,
+      "learning_rate": 0.0016666666666666668,
+      "loss": 0.8742,
+      "step": 75
+    },
+    {
+      "epoch": 0.16884198833657318,
+      "grad_norm": 0.007079805713146925,
+      "learning_rate": 0.001688888888888889,
+      "loss": 0.7933,
+      "step": 76
+    },
+    {
+      "epoch": 0.17106359344626493,
+      "grad_norm": 0.006508518476039171,
+      "learning_rate": 0.0017111111111111114,
+      "loss": 0.7358,
+      "step": 77
+    },
+    {
+      "epoch": 0.17328519855595667,
+      "grad_norm": 0.007266649045050144,
+      "learning_rate": 0.0017333333333333333,
+      "loss": 0.7264,
+      "step": 78
+    },
+    {
+      "epoch": 0.17550680366564844,
+      "grad_norm": 0.014574556611478329,
+      "learning_rate": 0.0017555555555555556,
+      "loss": 0.9533,
+      "step": 79
+    },
+    {
+      "epoch": 0.17772840877534019,
+      "grad_norm": 0.007840900681912899,
+      "learning_rate": 0.0017777777777777776,
+      "loss": 0.8228,
+      "step": 80
+    },
+    {
+      "epoch": 0.17995001388503193,
+      "grad_norm": 0.005117777734994888,
+      "learning_rate": 0.0018,
+      "loss": 0.8011,
+      "step": 81
+    },
+    {
+      "epoch": 0.1821716189947237,
+      "grad_norm": 0.012980911880731583,
+      "learning_rate": 0.0018222222222222223,
+      "loss": 0.7205,
+      "step": 82
+    },
+    {
+      "epoch": 0.18439322410441544,
+      "grad_norm": 0.008068774826824665,
+      "learning_rate": 0.0018444444444444446,
+      "loss": 0.8121,
+      "step": 83
+    },
+    {
+      "epoch": 0.1866148292141072,
+      "grad_norm": 0.007991770282387733,
+      "learning_rate": 0.0018666666666666666,
+      "loss": 0.8732,
+      "step": 84
+    },
+    {
+      "epoch": 0.18883643432379896,
+      "grad_norm": 0.00741413002833724,
+      "learning_rate": 0.001888888888888889,
+      "loss": 0.7656,
+      "step": 85
+    },
+    {
+      "epoch": 0.1910580394334907,
+      "grad_norm": 0.011837368831038475,
+      "learning_rate": 0.0019111111111111113,
+      "loss": 0.8856,
+      "step": 86
+    },
+    {
+      "epoch": 0.19327964454318244,
+      "grad_norm": 0.006025332026183605,
+      "learning_rate": 0.0019333333333333336,
+      "loss": 0.7813,
+      "step": 87
+    },
+    {
+      "epoch": 0.19550124965287421,
+      "grad_norm": 0.006810932885855436,
+      "learning_rate": 0.001955555555555556,
+      "loss": 0.7852,
+      "step": 88
+    },
+    {
+      "epoch": 0.19772285476256596,
+      "grad_norm": 0.01033777091652155,
+      "learning_rate": 0.0019777777777777775,
+      "loss": 0.9799,
+      "step": 89
+    },
+    {
+      "epoch": 0.1999444598722577,
+      "grad_norm": 0.007290220353752375,
+      "learning_rate": 0.002,
+      "loss": 0.8157,
+      "step": 90
+    },
+    {
+      "epoch": 0.20216606498194944,
+      "grad_norm": 0.0066700279712677,
+      "learning_rate": 0.002022222222222222,
+      "loss": 0.8832,
+      "step": 91
+    },
+    {
+      "epoch": 0.20438767009164122,
+      "grad_norm": 0.006774111185222864,
+      "learning_rate": 0.0020444444444444447,
+      "loss": 0.8649,
+      "step": 92
+    },
+    {
+      "epoch": 0.20660927520133296,
+      "grad_norm": 0.014007099904119968,
+      "learning_rate": 0.0020666666666666667,
+      "loss": 0.7803,
+      "step": 93
+    },
+    {
+      "epoch": 0.2088308803110247,
+      "grad_norm": 0.013977098278701305,
+      "learning_rate": 0.002088888888888889,
+      "loss": 0.8531,
+      "step": 94
+    },
+    {
+      "epoch": 0.21105248542071647,
+      "grad_norm": 0.010178284719586372,
+      "learning_rate": 0.0021111111111111113,
+      "loss": 0.7527,
+      "step": 95
+    },
+    {
+      "epoch": 0.21327409053040822,
+      "grad_norm": 0.014048648998141289,
+      "learning_rate": 0.0021333333333333334,
+      "loss": 0.8357,
+      "step": 96
+    },
+    {
+      "epoch": 0.21549569564009996,
+      "grad_norm": 0.010518389753997326,
+      "learning_rate": 0.0021555555555555555,
+      "loss": 0.892,
+      "step": 97
+    },
+    {
+      "epoch": 0.21771730074979173,
+      "grad_norm": 0.006465825717896223,
+      "learning_rate": 0.002177777777777778,
+      "loss": 0.7175,
+      "step": 98
+    },
+    {
+      "epoch": 0.21993890585948347,
+      "grad_norm": 0.008858314715325832,
+      "learning_rate": 0.0021999999999999997,
+      "loss": 0.8786,
+      "step": 99
+    },
+    {
+      "epoch": 0.22216051096917522,
+      "grad_norm": 0.022068988531827927,
+      "learning_rate": 0.0022222222222222222,
+      "loss": 0.9224,
+      "step": 100
+    },
+    {
+      "epoch": 0.224382116078867,
+      "grad_norm": 0.00683838315308094,
+      "learning_rate": 0.0022444444444444443,
+      "loss": 0.7949,
+      "step": 101
+    },
+    {
+      "epoch": 0.22660372118855873,
+      "grad_norm": 0.008592939004302025,
+      "learning_rate": 0.002266666666666667,
+      "loss": 0.6778,
+      "step": 102
+    },
+    {
+      "epoch": 0.22882532629825048,
+      "grad_norm": 0.005525896325707436,
+      "learning_rate": 0.002288888888888889,
+      "loss": 0.6661,
+      "step": 103
+    },
+    {
+      "epoch": 0.23104693140794225,
+      "grad_norm": 0.006523781456053257,
+      "learning_rate": 0.002311111111111111,
+      "loss": 0.852,
+      "step": 104
+    },
+    {
+      "epoch": 0.233268536517634,
+      "grad_norm": 0.007678350433707237,
+      "learning_rate": 0.0023333333333333335,
+      "loss": 0.7352,
+      "step": 105
+    },
+    {
+      "epoch": 0.23549014162732573,
+      "grad_norm": 0.005192214157432318,
+      "learning_rate": 0.0023555555555555556,
+      "loss": 0.8651,
+      "step": 106
+    },
+    {
+      "epoch": 0.2377117467370175,
+      "grad_norm": 0.007413934450596571,
+      "learning_rate": 0.002377777777777778,
+      "loss": 0.8005,
+      "step": 107
+    },
+    {
+      "epoch": 0.23993335184670925,
+      "grad_norm": 0.004861793946474791,
+      "learning_rate": 0.0024000000000000002,
+      "loss": 0.8363,
+      "step": 108
+    },
+    {
+      "epoch": 0.242154956956401,
+      "grad_norm": 0.005079795140773058,
+      "learning_rate": 0.0024222222222222223,
+      "loss": 0.8443,
+      "step": 109
+    },
+    {
+      "epoch": 0.24437656206609276,
+      "grad_norm": 0.0050173490308225155,
+      "learning_rate": 0.0024444444444444444,
+      "loss": 0.8945,
+      "step": 110
+    },
+    {
+      "epoch": 0.2465981671757845,
+      "grad_norm": 0.005878186784684658,
+      "learning_rate": 0.0024666666666666665,
+      "loss": 0.7684,
+      "step": 111
+    },
+    {
+      "epoch": 0.24881977228547625,
+      "grad_norm": 0.005863640923053026,
+      "learning_rate": 0.002488888888888889,
+      "loss": 0.8823,
+      "step": 112
+    },
+    {
+      "epoch": 0.251041377395168,
+      "grad_norm": 0.008415856398642063,
+      "learning_rate": 0.002511111111111111,
+      "loss": 0.7852,
+      "step": 113
+    },
+    {
+      "epoch": 0.25326298250485973,
+      "grad_norm": 0.006449985783547163,
+      "learning_rate": 0.002533333333333333,
+      "loss": 0.8724,
+      "step": 114
+    },
+    {
+      "epoch": 0.2554845876145515,
+      "grad_norm": 0.006899280473589897,
+      "learning_rate": 0.0025555555555555557,
+      "loss": 0.7742,
+      "step": 115
+    },
+    {
+      "epoch": 0.2577061927242433,
+      "grad_norm": 0.005553306546062231,
+      "learning_rate": 0.002577777777777778,
+      "loss": 0.9467,
+      "step": 116
+    },
+    {
+      "epoch": 0.259927797833935,
+      "grad_norm": 0.004053581040352583,
+      "learning_rate": 0.0026000000000000003,
+      "loss": 0.7767,
+      "step": 117
+    },
+    {
+      "epoch": 0.26214940294362676,
+      "grad_norm": 0.006906622089445591,
+      "learning_rate": 0.0026222222222222224,
+      "loss": 0.8731,
+      "step": 118
+    },
+    {
+      "epoch": 0.26437100805331853,
+      "grad_norm": 0.004571325611323118,
+      "learning_rate": 0.0026444444444444445,
+      "loss": 0.7887,
+      "step": 119
+    },
+    {
+      "epoch": 0.26659261316301025,
+      "grad_norm": 0.01691054366528988,
+      "learning_rate": 0.0026666666666666666,
+      "loss": 1.2046,
+      "step": 120
+    },
+    {
+      "epoch": 0.268814218272702,
+      "grad_norm": 0.006552200298756361,
+      "learning_rate": 0.0026888888888888887,
+      "loss": 1.078,
+      "step": 121
+    },
+    {
+      "epoch": 0.2710358233823938,
+      "grad_norm": 0.007155867293477058,
+      "learning_rate": 0.002711111111111111,
+      "loss": 0.8023,
+      "step": 122
+    },
+    {
+      "epoch": 0.2732574284920855,
+      "grad_norm": 0.006113141775131226,
+      "learning_rate": 0.0027333333333333333,
+      "loss": 0.8907,
+      "step": 123
+    },
+    {
+      "epoch": 0.2754790336017773,
+      "grad_norm": 0.004279575310647488,
+      "learning_rate": 0.002755555555555556,
+      "loss": 0.8393,
+      "step": 124
+    },
+    {
+      "epoch": 0.27770063871146905,
+      "grad_norm": 0.007299073971807957,
+      "learning_rate": 0.002777777777777778,
+      "loss": 0.7685,
+      "step": 125
+    },
+    {
+      "epoch": 0.27992224382116077,
+      "grad_norm": 0.005895877256989479,
+      "learning_rate": 0.0028,
+      "loss": 0.7787,
+      "step": 126
+    },
+    {
+      "epoch": 0.28214384893085254,
+      "grad_norm": 0.005973066668957472,
+      "learning_rate": 0.0028222222222222225,
+      "loss": 0.873,
+      "step": 127
+    },
+    {
+      "epoch": 0.2843654540405443,
+      "grad_norm": 0.006746681872755289,
+      "learning_rate": 0.0028444444444444446,
+      "loss": 0.8828,
+      "step": 128
+    },
+    {
+      "epoch": 0.286587059150236,
+      "grad_norm": 0.0047051082365214825,
+      "learning_rate": 0.0028666666666666667,
+      "loss": 0.8526,
+      "step": 129
+    },
+    {
+      "epoch": 0.2888086642599278,
+      "grad_norm": 0.0053818318992853165,
+      "learning_rate": 0.0028888888888888888,
+      "loss": 0.7842,
+      "step": 130
+    },
+    {
+      "epoch": 0.29103026936961957,
+      "grad_norm": 0.00510416692122817,
+      "learning_rate": 0.002911111111111111,
+      "loss": 0.8911,
+      "step": 131
+    },
+    {
+      "epoch": 0.2932518744793113,
+      "grad_norm": 0.00566230108961463,
+      "learning_rate": 0.0029333333333333334,
+      "loss": 0.8506,
+      "step": 132
+    },
+    {
+      "epoch": 0.29547347958900305,
+      "grad_norm": 0.007671915460377932,
+      "learning_rate": 0.0029555555555555555,
+      "loss": 0.8499,
+      "step": 133
+    },
+    {
+      "epoch": 0.2976950846986948,
+      "grad_norm": 0.005723770707845688,
+      "learning_rate": 0.002977777777777778,
+      "loss": 0.9035,
+      "step": 134
+    },
+    {
+      "epoch": 0.29991668980838654,
+      "grad_norm": 0.008367186412215233,
+      "learning_rate": 0.003,
+      "loss": 1.0066,
+      "step": 135
+    },
+    {
+      "epoch": 0.3021382949180783,
+      "grad_norm": 0.00792329665273428,
+      "learning_rate": 0.0029975308641975312,
+      "loss": 0.8272,
+      "step": 136
+    },
+    {
+      "epoch": 0.3043599000277701,
+      "grad_norm": 0.00764816626906395,
+      "learning_rate": 0.0029950617283950615,
+      "loss": 0.8652,
+      "step": 137
+    },
+    {
+      "epoch": 0.3065815051374618,
+      "grad_norm": 0.007859203033149242,
+      "learning_rate": 0.0029925925925925927,
+      "loss": 0.9529,
+      "step": 138
+    },
+    {
+      "epoch": 0.30880311024715357,
+      "grad_norm": 0.006429871078580618,
+      "learning_rate": 0.0029901234567901234,
+      "loss": 1.022,
+      "step": 139
+    },
+    {
+      "epoch": 0.31102471535684534,
+      "grad_norm": 0.007074374705553055,
+      "learning_rate": 0.0029876543209876546,
+      "loss": 0.8192,
+      "step": 140
+    },
+    {
+      "epoch": 0.31324632046653705,
+      "grad_norm": 0.006051207892596722,
+      "learning_rate": 0.002985185185185185,
+      "loss": 0.6763,
+      "step": 141
+    },
+    {
+      "epoch": 0.3154679255762288,
+      "grad_norm": 0.005065048113465309,
+      "learning_rate": 0.002982716049382716,
+      "loss": 0.8691,
+      "step": 142
+    },
+    {
+      "epoch": 0.3176895306859206,
+      "grad_norm": 0.006148957181721926,
+      "learning_rate": 0.0029802469135802472,
+      "loss": 0.6874,
+      "step": 143
+    },
+    {
+      "epoch": 0.3199111357956123,
+      "grad_norm": 0.007386965677142143,
+      "learning_rate": 0.002977777777777778,
+      "loss": 1.0042,
+      "step": 144
+    },
+    {
+      "epoch": 0.3221327409053041,
+      "grad_norm": 0.006896166130900383,
+      "learning_rate": 0.0029753086419753087,
+      "loss": 0.8665,
+      "step": 145
+    },
+    {
+      "epoch": 0.32435434601499585,
+      "grad_norm": 0.010677389800548553,
+      "learning_rate": 0.0029728395061728394,
+      "loss": 0.8079,
+      "step": 146
+    },
+    {
+      "epoch": 0.32657595112468757,
+      "grad_norm": 0.0062642525881528854,
+      "learning_rate": 0.0029703703703703706,
+      "loss": 0.7643,
+      "step": 147
+    },
+    {
+      "epoch": 0.32879755623437934,
+      "grad_norm": 0.013006279245018959,
+      "learning_rate": 0.0029679012345679013,
+      "loss": 0.839,
+      "step": 148
+    },
+    {
+      "epoch": 0.3310191613440711,
+      "grad_norm": 0.011782053858041763,
+      "learning_rate": 0.002965432098765432,
+      "loss": 0.7516,
+      "step": 149
+    },
+    {
+      "epoch": 0.3332407664537628,
+      "grad_norm": 0.0076681142672896385,
+      "learning_rate": 0.002962962962962963,
+      "loss": 0.9896,
+      "step": 150
+    },
+    {
+      "epoch": 0.3354623715634546,
+      "grad_norm": 0.006526409648358822,
+      "learning_rate": 0.002960493827160494,
+      "loss": 0.8881,
+      "step": 151
+    },
+    {
+      "epoch": 0.33768397667314637,
+      "grad_norm": 0.00649358332157135,
+      "learning_rate": 0.0029580246913580247,
+      "loss": 0.7183,
+      "step": 152
+    },
+    {
+      "epoch": 0.3399055817828381,
+      "grad_norm": 0.005959691014140844,
+      "learning_rate": 0.0029555555555555555,
+      "loss": 0.8335,
+      "step": 153
+    },
+    {
+      "epoch": 0.34212718689252986,
+      "grad_norm": 0.005547534208744764,
+      "learning_rate": 0.0029530864197530866,
+      "loss": 0.7954,
+      "step": 154
+    },
+    {
+      "epoch": 0.3443487920022216,
+      "grad_norm": 0.005612339824438095,
+      "learning_rate": 0.0029506172839506174,
+      "loss": 0.7647,
+      "step": 155
+    },
+    {
+      "epoch": 0.34657039711191334,
+      "grad_norm": 0.00677449582144618,
+      "learning_rate": 0.002948148148148148,
+      "loss": 0.9327,
+      "step": 156
+    },
+    {
+      "epoch": 0.3487920022216051,
+      "grad_norm": 0.007437689695507288,
+      "learning_rate": 0.002945679012345679,
+      "loss": 0.8094,
+      "step": 157
+    },
+    {
+      "epoch": 0.3510136073312969,
+      "grad_norm": 0.005581557285040617,
+      "learning_rate": 0.00294320987654321,
+      "loss": 0.9587,
+      "step": 158
+    },
+    {
+      "epoch": 0.3532352124409886,
+      "grad_norm": 0.007026318926364183,
+      "learning_rate": 0.0029407407407407407,
+      "loss": 0.9297,
+      "step": 159
+    },
+    {
+      "epoch": 0.35545681755068037,
+      "grad_norm": 0.007772298529744148,
+      "learning_rate": 0.0029382716049382715,
+      "loss": 0.6309,
+      "step": 160
+    },
+    {
+      "epoch": 0.35767842266037214,
+      "grad_norm": 0.005535977426916361,
+      "learning_rate": 0.0029358024691358026,
+      "loss": 0.8712,
+      "step": 161
+    },
+    {
+      "epoch": 0.35990002777006386,
+      "grad_norm": 0.0064979312010109425,
+      "learning_rate": 0.0029333333333333334,
+      "loss": 0.7258,
+      "step": 162
+    },
+    {
+      "epoch": 0.36212163287975563,
+      "grad_norm": 0.009692924097180367,
+      "learning_rate": 0.002930864197530864,
+      "loss": 0.8411,
+      "step": 163
+    },
+    {
+      "epoch": 0.3643432379894474,
+      "grad_norm": 0.0048782131634652615,
+      "learning_rate": 0.002928395061728395,
+      "loss": 0.8574,
+      "step": 164
+    },
+    {
+      "epoch": 0.3665648430991391,
+      "grad_norm": 0.005760766100138426,
+      "learning_rate": 0.002925925925925926,
+      "loss": 0.771,
+      "step": 165
+    },
+    {
+      "epoch": 0.3687864482088309,
+      "grad_norm": 0.0057001737877726555,
+      "learning_rate": 0.002923456790123457,
+      "loss": 0.8197,
+      "step": 166
+    },
+    {
+      "epoch": 0.37100805331852266,
+      "grad_norm": 0.0068498472683131695,
+      "learning_rate": 0.0029209876543209875,
+      "loss": 0.8001,
+      "step": 167
+    },
+    {
+      "epoch": 0.3732296584282144,
+      "grad_norm": 0.00659180385991931,
+      "learning_rate": 0.0029185185185185186,
+      "loss": 0.6573,
+      "step": 168
+    },
+    {
+      "epoch": 0.37545126353790614,
+      "grad_norm": 0.005685858894139528,
+      "learning_rate": 0.0029160493827160494,
+      "loss": 0.8331,
+      "step": 169
+    },
+    {
+      "epoch": 0.3776728686475979,
+      "grad_norm": 0.005140091758221388,
+      "learning_rate": 0.0029135802469135805,
+      "loss": 0.6738,
+      "step": 170
+    },
+    {
+      "epoch": 0.37989447375728963,
+      "grad_norm": 0.006014992017298937,
+      "learning_rate": 0.002911111111111111,
+      "loss": 0.692,
+      "step": 171
+    },
+    {
+      "epoch": 0.3821160788669814,
+      "grad_norm": 0.00642600329592824,
+      "learning_rate": 0.002908641975308642,
+      "loss": 0.6947,
+      "step": 172
+    },
+    {
+      "epoch": 0.3843376839766732,
+      "grad_norm": 0.0054779211059212685,
+      "learning_rate": 0.002906172839506173,
+      "loss": 0.8936,
+      "step": 173
+    },
+    {
+      "epoch": 0.3865592890863649,
+      "grad_norm": 0.006306678056716919,
+      "learning_rate": 0.002903703703703704,
+      "loss": 0.8784,
+      "step": 174
+    },
+    {
+      "epoch": 0.38878089419605666,
+      "grad_norm": 0.006951971910893917,
+      "learning_rate": 0.0029012345679012346,
+      "loss": 0.8032,
+      "step": 175
+    },
+    {
+      "epoch": 0.39100249930574843,
+      "grad_norm": 0.011171508580446243,
+      "learning_rate": 0.0028987654320987654,
+      "loss": 0.7842,
+      "step": 176
+    },
+    {
+      "epoch": 0.39322410441544015,
+      "grad_norm": 0.008988376706838608,
+      "learning_rate": 0.0028962962962962966,
+      "loss": 0.9934,
+      "step": 177
+    },
+    {
+      "epoch": 0.3954457095251319,
+      "grad_norm": 0.008020875044167042,
+      "learning_rate": 0.0028938271604938273,
+      "loss": 0.887,
+      "step": 178
+    },
+    {
+      "epoch": 0.3976673146348237,
+      "grad_norm": 0.00882386602461338,
+      "learning_rate": 0.002891358024691358,
+      "loss": 0.8859,
+      "step": 179
+    },
+    {
+      "epoch": 0.3998889197445154,
+      "grad_norm": 0.0066726855002343655,
+      "learning_rate": 0.0028888888888888888,
+      "loss": 0.7581,
+      "step": 180
+    },
+    {
+      "epoch": 0.4021105248542072,
+      "grad_norm": 0.011035891249775887,
+      "learning_rate": 0.00288641975308642,
+      "loss": 0.9224,
+      "step": 181
+    },
+    {
+      "epoch": 0.4043321299638989,
+      "grad_norm": 0.0045290132984519005,
+      "learning_rate": 0.0028839506172839507,
+      "loss": 0.6422,
+      "step": 182
+    },
+    {
+      "epoch": 0.40655373507359066,
+      "grad_norm": 0.006385261192917824,
+      "learning_rate": 0.0028814814814814814,
+      "loss": 0.7726,
+      "step": 183
+    },
+    {
+      "epoch": 0.40877534018328243,
+      "grad_norm": 0.006436359137296677,
+      "learning_rate": 0.0028790123456790126,
+      "loss": 0.7782,
+      "step": 184
+    },
+    {
+      "epoch": 0.41099694529297415,
+      "grad_norm": 0.006331396754831076,
+      "learning_rate": 0.0028765432098765433,
+      "loss": 0.8402,
+      "step": 185
+    },
+    {
+      "epoch": 0.4132185504026659,
+      "grad_norm": 0.009812875650823116,
+      "learning_rate": 0.002874074074074074,
+      "loss": 0.7303,
+      "step": 186
+    },
+    {
+      "epoch": 0.4154401555123577,
+      "grad_norm": 0.008723249658942223,
+      "learning_rate": 0.0028716049382716048,
+      "loss": 0.8702,
+      "step": 187
+    },
+    {
+      "epoch": 0.4176617606220494,
+      "grad_norm": 0.007981473580002785,
+      "learning_rate": 0.002869135802469136,
+      "loss": 0.7417,
+      "step": 188
+    },
+    {
+      "epoch": 0.4198833657317412,
+      "grad_norm": 0.00769078591838479,
+      "learning_rate": 0.0028666666666666667,
+      "loss": 0.9278,
+      "step": 189
+    },
+    {
+      "epoch": 0.42210497084143295,
+      "grad_norm": 0.006087169982492924,
+      "learning_rate": 0.0028641975308641974,
+      "loss": 0.7726,
+      "step": 190
+    },
+    {
+      "epoch": 0.42432657595112466,
+      "grad_norm": 0.012171044945716858,
+      "learning_rate": 0.0028617283950617286,
+      "loss": 0.9587,
+      "step": 191
+    },
+    {
+      "epoch": 0.42654818106081643,
+      "grad_norm": 0.0077499267645180225,
+      "learning_rate": 0.0028592592592592593,
+      "loss": 0.9549,
+      "step": 192
+    },
+    {
+      "epoch": 0.4287697861705082,
+      "grad_norm": 0.012135319411754608,
+      "learning_rate": 0.00285679012345679,
+      "loss": 0.868,
+      "step": 193
+    },
+    {
+      "epoch": 0.4309913912801999,
+      "grad_norm": 0.01298799179494381,
+      "learning_rate": 0.0028543209876543208,
+      "loss": 0.6943,
+      "step": 194
+    },
+    {
+      "epoch": 0.4332129963898917,
+      "grad_norm": 0.009397965855896473,
+      "learning_rate": 0.002851851851851852,
+      "loss": 0.5948,
+      "step": 195
+    },
+    {
+      "epoch": 0.43543460149958346,
+      "grad_norm": 0.008861496113240719,
+      "learning_rate": 0.002849382716049383,
+      "loss": 0.8092,
+      "step": 196
+    },
+    {
+      "epoch": 0.4376562066092752,
+      "grad_norm": 0.011813944205641747,
+      "learning_rate": 0.0028469135802469134,
+      "loss": 0.8379,
+      "step": 197
+    },
+    {
+      "epoch": 0.43987781171896695,
+      "grad_norm": 0.008746293373405933,
+      "learning_rate": 0.0028444444444444446,
+      "loss": 0.7222,
+      "step": 198
+    },
+    {
+      "epoch": 0.4420994168286587,
+      "grad_norm": 0.006701031234115362,
+      "learning_rate": 0.0028419753086419753,
+      "loss": 0.7471,
+      "step": 199
+    },
+    {
+      "epoch": 0.44432102193835044,
+      "grad_norm": 0.00861276499927044,
+      "learning_rate": 0.0028395061728395065,
+      "loss": 0.7922,
+      "step": 200
+    },
+    {
+      "epoch": 0.4465426270480422,
+      "grad_norm": 0.009883517399430275,
+      "learning_rate": 0.002837037037037037,
+      "loss": 0.7278,
+      "step": 201
+    },
+    {
+      "epoch": 0.448764232157734,
+      "grad_norm": 0.009246917441487312,
+      "learning_rate": 0.002834567901234568,
+      "loss": 0.7314,
+      "step": 202
+    },
+    {
+      "epoch": 0.4509858372674257,
+      "grad_norm": 0.008673238568007946,
+      "learning_rate": 0.002832098765432099,
+      "loss": 0.9159,
+      "step": 203
+    },
+    {
+      "epoch": 0.45320744237711746,
+      "grad_norm": 0.011417444795370102,
+      "learning_rate": 0.00282962962962963,
+      "loss": 0.7767,
+      "step": 204
+    },
+    {
+      "epoch": 0.45542904748680924,
+      "grad_norm": 0.008244509808719158,
+      "learning_rate": 0.0028271604938271606,
+      "loss": 0.5515,
+      "step": 205
+    },
+    {
+      "epoch": 0.45765065259650095,
+      "grad_norm": 0.007160137873142958,
+      "learning_rate": 0.0028246913580246913,
+      "loss": 0.8073,
+      "step": 206
+    },
+    {
+      "epoch": 0.4598722577061927,
+      "grad_norm": 0.00974318664520979,
+      "learning_rate": 0.0028222222222222225,
+      "loss": 0.8644,
+      "step": 207
+    },
+    {
+      "epoch": 0.4620938628158845,
+      "grad_norm": 0.012410882860422134,
+      "learning_rate": 0.0028197530864197532,
+      "loss": 0.8899,
+      "step": 208
+    },
+    {
+      "epoch": 0.4643154679255762,
+      "grad_norm": 0.008747541345655918,
+      "learning_rate": 0.002817283950617284,
+      "loss": 0.7164,
+      "step": 209
+    },
+    {
+      "epoch": 0.466537073035268,
+      "grad_norm": 0.006024535745382309,
+      "learning_rate": 0.0028148148148148147,
+      "loss": 0.8533,
+      "step": 210
+    },
+    {
+      "epoch": 0.46875867814495975,
+      "grad_norm": 0.015827184543013573,
+      "learning_rate": 0.002812345679012346,
+      "loss": 0.8671,
+      "step": 211
+    },
+    {
+      "epoch": 0.47098028325465147,
+      "grad_norm": 0.013850335963070393,
+      "learning_rate": 0.0028098765432098766,
+      "loss": 0.8823,
+      "step": 212
+    },
+    {
+      "epoch": 0.47320188836434324,
+      "grad_norm": 0.0074915774166584015,
+      "learning_rate": 0.0028074074074074073,
+      "loss": 0.7789,
+      "step": 213
+    },
+    {
+      "epoch": 0.475423493474035,
+      "grad_norm": 0.011638727970421314,
+      "learning_rate": 0.0028049382716049385,
+      "loss": 0.7609,
+      "step": 214
+    },
+    {
+      "epoch": 0.4776450985837267,
+      "grad_norm": 0.007853872142732143,
+      "learning_rate": 0.0028024691358024692,
+      "loss": 0.8372,
+      "step": 215
+    },
+    {
+      "epoch": 0.4798667036934185,
+      "grad_norm": 0.010911183431744576,
+      "learning_rate": 0.0028,
+      "loss": 0.8292,
+      "step": 216
+    },
+    {
+      "epoch": 0.48208830880311027,
+      "grad_norm": 0.012936594896018505,
+      "learning_rate": 0.0027975308641975307,
+      "loss": 0.7752,
+      "step": 217
+    },
+    {
+      "epoch": 0.484309913912802,
+      "grad_norm": 0.008519921451807022,
+      "learning_rate": 0.002795061728395062,
+      "loss": 0.8553,
+      "step": 218
+    },
+    {
+      "epoch": 0.48653151902249375,
+      "grad_norm": 0.009515728801488876,
+      "learning_rate": 0.0027925925925925926,
+      "loss": 0.87,
+      "step": 219
+    },
+    {
+      "epoch": 0.4887531241321855,
+      "grad_norm": 0.01014457643032074,
+      "learning_rate": 0.0027901234567901233,
+      "loss": 0.7561,
+      "step": 220
+    },
+    {
+      "epoch": 0.49097472924187724,
+      "grad_norm": 0.007946905680000782,
+      "learning_rate": 0.0027876543209876545,
+      "loss": 0.7532,
+      "step": 221
+    },
+    {
+      "epoch": 0.493196334351569,
+      "grad_norm": 0.0068723889999091625,
+      "learning_rate": 0.0027851851851851852,
+      "loss": 0.8536,
+      "step": 222
+    },
+    {
+      "epoch": 0.4954179394612608,
+      "grad_norm": 0.006598350126296282,
+      "learning_rate": 0.002782716049382716,
+      "loss": 0.7341,
+      "step": 223
+    },
+    {
+      "epoch": 0.4976395445709525,
+      "grad_norm": 0.00824499037116766,
+      "learning_rate": 0.0027802469135802467,
+      "loss": 0.7093,
+      "step": 224
+    },
+    {
+      "epoch": 0.49986114968064427,
+      "grad_norm": 0.007247340399771929,
+      "learning_rate": 0.002777777777777778,
+      "loss": 0.9103,
+      "step": 225
+    },
+    {
+      "epoch": 0.502082754790336,
+      "grad_norm": 0.008267593570053577,
+      "learning_rate": 0.002775308641975309,
+      "loss": 0.8218,
+      "step": 226
+    },
+    {
+      "epoch": 0.5043043599000278,
+      "grad_norm": 0.00702422671020031,
+      "learning_rate": 0.0027728395061728394,
+      "loss": 0.8115,
+      "step": 227
+    },
+    {
+      "epoch": 0.5065259650097195,
+      "grad_norm": 0.005334863904863596,
+      "learning_rate": 0.0027703703703703705,
+      "loss": 0.7843,
+      "step": 228
+    },
+    {
+      "epoch": 0.5087475701194113,
+      "grad_norm": 0.009140919893980026,
+      "learning_rate": 0.0027679012345679013,
+      "loss": 0.8132,
+      "step": 229
+    },
+    {
+      "epoch": 0.510969175229103,
+      "grad_norm": 0.007120530121028423,
+      "learning_rate": 0.0027654320987654324,
+      "loss": 0.7962,
+      "step": 230
+    },
+    {
+      "epoch": 0.5131907803387947,
+      "grad_norm": 0.012158839032053947,
+      "learning_rate": 0.0027629629629629627,
+      "loss": 0.9541,
+      "step": 231
+    },
+    {
+      "epoch": 0.5154123854484866,
+      "grad_norm": 0.015124938450753689,
+      "learning_rate": 0.002760493827160494,
+      "loss": 0.8044,
+      "step": 232
+    },
+    {
+      "epoch": 0.5176339905581783,
+      "grad_norm": 0.007999619469046593,
+      "learning_rate": 0.0027580246913580246,
+      "loss": 0.8035,
+      "step": 233
+    },
+    {
+      "epoch": 0.51985559566787,
+      "grad_norm": 0.02217845618724823,
+      "learning_rate": 0.002755555555555556,
+      "loss": 1.1012,
+      "step": 234
+    },
+    {
+      "epoch": 0.5220772007775618,
+      "grad_norm": 0.010226511396467686,
+      "learning_rate": 0.0027530864197530865,
+      "loss": 0.6606,
+      "step": 235
+    },
+    {
+      "epoch": 0.5242988058872535,
+      "grad_norm": 0.006293569225817919,
+      "learning_rate": 0.0027506172839506173,
+      "loss": 0.7629,
+      "step": 236
+    },
+    {
+      "epoch": 0.5265204109969452,
+      "grad_norm": 0.012506984174251556,
+      "learning_rate": 0.0027481481481481484,
+      "loss": 0.9344,
+      "step": 237
+    },
+    {
+      "epoch": 0.5287420161066371,
+      "grad_norm": 0.009668968617916107,
+      "learning_rate": 0.002745679012345679,
+      "loss": 0.9924,
+      "step": 238
+    },
+    {
+      "epoch": 0.5309636212163288,
+      "grad_norm": 0.01322484202682972,
+      "learning_rate": 0.00274320987654321,
+      "loss": 0.904,
+      "step": 239
+    },
+    {
+      "epoch": 0.5331852263260205,
+      "grad_norm": 0.006201568990945816,
+      "learning_rate": 0.0027407407407407406,
+      "loss": 0.8552,
+      "step": 240
+    },
+    {
+      "epoch": 0.5354068314357123,
+      "grad_norm": 0.013609252870082855,
+      "learning_rate": 0.002738271604938272,
+      "loss": 0.7899,
+      "step": 241
+    },
+    {
+      "epoch": 0.537628436545404,
+      "grad_norm": 0.007919220253825188,
+      "learning_rate": 0.0027358024691358025,
+      "loss": 0.7527,
+      "step": 242
+    },
+    {
+      "epoch": 0.5398500416550958,
+      "grad_norm": 0.00696425000205636,
+      "learning_rate": 0.0027333333333333333,
+      "loss": 0.719,
+      "step": 243
+    },
+    {
+      "epoch": 0.5420716467647876,
+      "grad_norm": 0.008950626477599144,
+      "learning_rate": 0.0027308641975308644,
+      "loss": 0.7014,
+      "step": 244
+    },
+    {
+      "epoch": 0.5442932518744793,
+      "grad_norm": 0.012288985773921013,
+      "learning_rate": 0.002728395061728395,
+      "loss": 0.7899,
+      "step": 245
+    },
+    {
+      "epoch": 0.546514856984171,
+      "grad_norm": 0.011310501024127007,
+      "learning_rate": 0.002725925925925926,
+      "loss": 0.9283,
+      "step": 246
+    },
+    {
+      "epoch": 0.5487364620938628,
+      "grad_norm": 0.015484470874071121,
+      "learning_rate": 0.0027234567901234566,
+      "loss": 0.9007,
+      "step": 247
+    },
+    {
+      "epoch": 0.5509580672035546,
+      "grad_norm": 0.008781633339822292,
+      "learning_rate": 0.002720987654320988,
+      "loss": 0.9231,
+      "step": 248
+    },
+    {
+      "epoch": 0.5531796723132463,
+      "grad_norm": 0.009064835496246815,
+      "learning_rate": 0.0027185185185185185,
+      "loss": 0.7347,
+      "step": 249
+    },
+    {
+      "epoch": 0.5554012774229381,
+      "grad_norm": 0.011060320772230625,
+      "learning_rate": 0.0027160493827160493,
+      "loss": 0.5795,
+      "step": 250
+    },
+    {
+      "epoch": 0.5576228825326298,
+      "grad_norm": 0.00757238594815135,
+      "learning_rate": 0.0027135802469135805,
+      "loss": 0.8312,
+      "step": 251
+    },
+    {
+      "epoch": 0.5598444876423215,
+      "grad_norm": 0.008306944742798805,
+      "learning_rate": 0.002711111111111111,
+      "loss": 0.9226,
+      "step": 252
+    },
+    {
+      "epoch": 0.5620660927520134,
+      "grad_norm": 0.010423055849969387,
+      "learning_rate": 0.002708641975308642,
+      "loss": 0.8448,
+      "step": 253
+    },
+    {
+      "epoch": 0.5642876978617051,
+      "grad_norm": 0.010584446601569653,
+      "learning_rate": 0.0027061728395061727,
+      "loss": 0.8116,
+      "step": 254
+    },
+    {
+      "epoch": 0.5665093029713968,
+      "grad_norm": 0.00733089679852128,
+      "learning_rate": 0.002703703703703704,
+      "loss": 0.6264,
+      "step": 255
+    },
+    {
+      "epoch": 0.5687309080810886,
+      "grad_norm": 0.007233286276459694,
+      "learning_rate": 0.002701234567901235,
+      "loss": 0.9293,
+      "step": 256
+    },
+    {
+      "epoch": 0.5709525131907803,
+      "grad_norm": 0.005947368685156107,
+      "learning_rate": 0.0026987654320987653,
+      "loss": 0.8702,
+      "step": 257
+    },
+    {
+      "epoch": 0.573174118300472,
+      "grad_norm": 0.007871137000620365,
+      "learning_rate": 0.0026962962962962965,
+      "loss": 0.7136,
+      "step": 258
+    },
+    {
+      "epoch": 0.5753957234101639,
+      "grad_norm": 0.012100719846785069,
+      "learning_rate": 0.002693827160493827,
+      "loss": 1.0855,
+      "step": 259
+    },
+    {
+      "epoch": 0.5776173285198556,
+      "grad_norm": 0.012036778964102268,
+      "learning_rate": 0.0026913580246913584,
+      "loss": 0.6768,
+      "step": 260
+    },
+    {
+      "epoch": 0.5798389336295473,
+      "grad_norm": 0.015129570849239826,
+      "learning_rate": 0.0026888888888888887,
+      "loss": 0.8401,
+      "step": 261
+    },
+    {
+      "epoch": 0.5820605387392391,
+      "grad_norm": 0.0090151596814394,
+      "learning_rate": 0.00268641975308642,
+      "loss": 0.729,
+      "step": 262
+    },
+    {
+      "epoch": 0.5842821438489308,
+      "grad_norm": 0.01668744534254074,
+      "learning_rate": 0.0026839506172839506,
+      "loss": 0.8578,
+      "step": 263
+    },
+    {
+      "epoch": 0.5865037489586226,
+      "grad_norm": 0.024113476276397705,
+      "learning_rate": 0.0026814814814814817,
+      "loss": 0.9802,
+      "step": 264
+    },
+    {
+      "epoch": 0.5887253540683144,
+      "grad_norm": 0.006910145748406649,
+      "learning_rate": 0.0026790123456790125,
+      "loss": 0.8169,
+      "step": 265
+    },
+    {
+      "epoch": 0.5909469591780061,
+      "grad_norm": 0.006911164615303278,
+      "learning_rate": 0.002676543209876543,
+      "loss": 0.8629,
+      "step": 266
+    },
+    {
+      "epoch": 0.5931685642876978,
+      "grad_norm": 0.010877076536417007,
+      "learning_rate": 0.0026740740740740744,
+      "loss": 0.8885,
+      "step": 267
+    },
+    {
+      "epoch": 0.5953901693973896,
+      "grad_norm": 0.01698371395468712,
+      "learning_rate": 0.0026716049382716047,
+      "loss": 0.8002,
+      "step": 268
+    },
+    {
+      "epoch": 0.5976117745070814,
+      "grad_norm": 0.008402695879340172,
+      "learning_rate": 0.002669135802469136,
+      "loss": 0.7057,
+      "step": 269
+    },
+    {
+      "epoch": 0.5998333796167731,
+      "grad_norm": 0.010379761457443237,
+      "learning_rate": 0.0026666666666666666,
+      "loss": 0.829,
+      "step": 270
+    },
+    {
+      "epoch": 0.6020549847264649,
+      "grad_norm": 0.009576470591127872,
+      "learning_rate": 0.0026641975308641977,
+      "loss": 0.9202,
+      "step": 271
+    },
+    {
+      "epoch": 0.6042765898361566,
+      "grad_norm": 0.008488965220749378,
+      "learning_rate": 0.0026617283950617285,
+      "loss": 0.7697,
+      "step": 272
+    },
+    {
+      "epoch": 0.6064981949458483,
+      "grad_norm": 0.011320218443870544,
+      "learning_rate": 0.002659259259259259,
+      "loss": 0.756,
+      "step": 273
+    },
+    {
+      "epoch": 0.6087198000555402,
+      "grad_norm": 0.02209623157978058,
+      "learning_rate": 0.0026567901234567904,
+      "loss": 1.0533,
+      "step": 274
+    },
+    {
+      "epoch": 0.6109414051652319,
+      "grad_norm": 0.01179931778460741,
+      "learning_rate": 0.002654320987654321,
+      "loss": 0.8604,
+      "step": 275
+    },
+    {
+      "epoch": 0.6131630102749236,
+      "grad_norm": 0.013143769465386868,
+      "learning_rate": 0.002651851851851852,
+      "loss": 0.8984,
+      "step": 276
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 0.010282885283231735,
+      "learning_rate": 0.0026493827160493826,
+      "loss": 0.8802,
+      "step": 277
+    },
+    {
+      "epoch": 0.6176062204943071,
+      "grad_norm": 0.007673004176467657,
+      "learning_rate": 0.0026469135802469138,
+      "loss": 0.8439,
+      "step": 278
+    },
+    {
+      "epoch": 0.6198278256039988,
+      "grad_norm": 0.008318427950143814,
+      "learning_rate": 0.0026444444444444445,
+      "loss": 0.7776,
+      "step": 279
+    },
+    {
+      "epoch": 0.6220494307136907,
+      "grad_norm": 0.012036367319524288,
+      "learning_rate": 0.0026419753086419752,
+      "loss": 0.7192,
+      "step": 280
+    },
+    {
+      "epoch": 0.6242710358233824,
+      "grad_norm": 0.0068825045600533485,
+      "learning_rate": 0.0026395061728395064,
+      "loss": 0.7847,
+      "step": 281
+    },
+    {
+      "epoch": 0.6264926409330741,
+      "grad_norm": 0.007046104408800602,
+      "learning_rate": 0.002637037037037037,
+      "loss": 0.7069,
+      "step": 282
+    },
+    {
+      "epoch": 0.6287142460427659,
+      "grad_norm": 0.00804830901324749,
+      "learning_rate": 0.002634567901234568,
+      "loss": 0.9622,
+      "step": 283
+    },
+    {
+      "epoch": 0.6309358511524576,
+      "grad_norm": 0.0072870319709181786,
+      "learning_rate": 0.0026320987654320986,
+      "loss": 0.6801,
+      "step": 284
+    },
+    {
+      "epoch": 0.6331574562621494,
+      "grad_norm": 0.00952050369232893,
+      "learning_rate": 0.0026296296296296298,
+      "loss": 0.8667,
+      "step": 285
+    },
+    {
+      "epoch": 0.6353790613718412,
+      "grad_norm": 0.006871790159493685,
+      "learning_rate": 0.0026271604938271605,
+      "loss": 0.7638,
+      "step": 286
+    },
+    {
+      "epoch": 0.6376006664815329,
+      "grad_norm": 0.006217732559889555,
+      "learning_rate": 0.0026246913580246912,
+      "loss": 0.7272,
+      "step": 287
+    },
+    {
+      "epoch": 0.6398222715912246,
+      "grad_norm": 0.006713525392115116,
+      "learning_rate": 0.0026222222222222224,
+      "loss": 0.7885,
+      "step": 288
+    },
+    {
+      "epoch": 0.6420438767009164,
+      "grad_norm": 0.004922814667224884,
+      "learning_rate": 0.002619753086419753,
+      "loss": 0.6974,
+      "step": 289
+    },
+    {
+      "epoch": 0.6442654818106082,
+      "grad_norm": 0.00727014010772109,
+      "learning_rate": 0.002617283950617284,
+      "loss": 0.9864,
+      "step": 290
+    },
+    {
+      "epoch": 0.6464870869202999,
+      "grad_norm": 0.010591446422040462,
+      "learning_rate": 0.0026148148148148146,
+      "loss": 0.8506,
+      "step": 291
+    },
+    {
+      "epoch": 0.6487086920299917,
+      "grad_norm": 0.009507875889539719,
+      "learning_rate": 0.0026123456790123458,
+      "loss": 0.754,
+      "step": 292
+    },
+    {
+      "epoch": 0.6509302971396834,
+      "grad_norm": 0.00874133687466383,
+      "learning_rate": 0.0026098765432098765,
+      "loss": 0.9275,
+      "step": 293
+    },
+    {
+      "epoch": 0.6531519022493751,
+      "grad_norm": 0.008199427276849747,
+      "learning_rate": 0.0026074074074074072,
+      "loss": 0.8286,
+      "step": 294
+    },
+    {
+      "epoch": 0.655373507359067,
+      "grad_norm": 0.010877394117414951,
+      "learning_rate": 0.0026049382716049384,
+      "loss": 0.6721,
+      "step": 295
+    },
+    {
+      "epoch": 0.6575951124687587,
+      "grad_norm": 0.007632778026163578,
+      "learning_rate": 0.002602469135802469,
+      "loss": 0.796,
+      "step": 296
+    },
+    {
+      "epoch": 0.6598167175784504,
+      "grad_norm": 0.007029394619166851,
+      "learning_rate": 0.0026000000000000003,
+      "loss": 0.7297,
+      "step": 297
+    },
+    {
+      "epoch": 0.6620383226881422,
+      "grad_norm": 0.008049899712204933,
+      "learning_rate": 0.0025975308641975306,
+      "loss": 0.7303,
+      "step": 298
+    },
+    {
+      "epoch": 0.6642599277978339,
+      "grad_norm": 0.009279787540435791,
+      "learning_rate": 0.002595061728395062,
+      "loss": 0.7327,
+      "step": 299
+    },
+    {
+      "epoch": 0.6664815329075257,
+      "grad_norm": 0.006137878634035587,
+      "learning_rate": 0.0025925925925925925,
+      "loss": 0.7083,
+      "step": 300
+    },
+    {
+      "epoch": 0.6687031380172175,
+      "grad_norm": 0.008568949066102505,
+      "learning_rate": 0.0025901234567901237,
+      "loss": 0.9256,
+      "step": 301
+    },
+    {
+      "epoch": 0.6709247431269092,
+      "grad_norm": 0.006636477075517178,
+      "learning_rate": 0.0025876543209876544,
+      "loss": 0.7854,
+      "step": 302
+    },
+    {
+      "epoch": 0.6731463482366009,
+      "grad_norm": 0.008452818728983402,
+      "learning_rate": 0.002585185185185185,
+      "loss": 0.8342,
+      "step": 303
+    },
+    {
+      "epoch": 0.6753679533462927,
+      "grad_norm": 0.0126169603317976,
+      "learning_rate": 0.0025827160493827163,
+      "loss": 0.7861,
+      "step": 304
+    },
+    {
+      "epoch": 0.6775895584559845,
+      "grad_norm": 0.007217437960207462,
+      "learning_rate": 0.002580246913580247,
+      "loss": 0.7743,
+      "step": 305
+    },
+    {
+      "epoch": 0.6798111635656762,
+      "grad_norm": 0.008002759888768196,
+      "learning_rate": 0.002577777777777778,
+      "loss": 0.6823,
+      "step": 306
+    },
+    {
+      "epoch": 0.682032768675368,
+      "grad_norm": 0.013185406103730202,
+      "learning_rate": 0.0025753086419753085,
+      "loss": 0.8276,
+      "step": 307
+    },
+    {
+      "epoch": 0.6842543737850597,
+      "grad_norm": 0.006368585396558046,
+      "learning_rate": 0.0025728395061728397,
+      "loss": 0.6713,
+      "step": 308
+    },
+    {
+      "epoch": 0.6864759788947514,
+      "grad_norm": 0.007478953339159489,
+      "learning_rate": 0.0025703703703703704,
+      "loss": 0.7786,
+      "step": 309
+    },
+    {
+      "epoch": 0.6886975840044433,
+      "grad_norm": 0.008412404917180538,
+      "learning_rate": 0.002567901234567901,
+      "loss": 0.6874,
+      "step": 310
+    },
+    {
+      "epoch": 0.690919189114135,
+      "grad_norm": 0.0067495848052203655,
+      "learning_rate": 0.0025654320987654323,
+      "loss": 0.7412,
+      "step": 311
+    },
+    {
+      "epoch": 0.6931407942238267,
+      "grad_norm": 0.007151488680392504,
+      "learning_rate": 0.002562962962962963,
+      "loss": 0.6853,
+      "step": 312
+    },
+    {
+      "epoch": 0.6953623993335185,
+      "grad_norm": 0.009860935620963573,
+      "learning_rate": 0.002560493827160494,
+      "loss": 0.7972,
+      "step": 313
+    },
+    {
+      "epoch": 0.6975840044432102,
+      "grad_norm": 0.007228904403746128,
+      "learning_rate": 0.0025580246913580245,
+      "loss": 0.816,
+      "step": 314
+    },
+    {
+      "epoch": 0.6998056095529019,
+      "grad_norm": 0.008685276843607426,
+      "learning_rate": 0.0025555555555555557,
+      "loss": 0.9024,
+      "step": 315
+    },
+    {
+      "epoch": 0.7020272146625938,
+      "grad_norm": 0.011443904601037502,
+      "learning_rate": 0.0025530864197530864,
+      "loss": 0.8276,
+      "step": 316
+    },
+    {
+      "epoch": 0.7042488197722855,
+      "grad_norm": 0.00936659425497055,
+      "learning_rate": 0.002550617283950617,
+      "loss": 0.8549,
+      "step": 317
+    },
+    {
+      "epoch": 0.7064704248819772,
+      "grad_norm": 0.007054576184600592,
+      "learning_rate": 0.0025481481481481483,
+      "loss": 0.7386,
+      "step": 318
+    },
+    {
+      "epoch": 0.708692029991669,
+      "grad_norm": 0.008555485866963863,
+      "learning_rate": 0.002545679012345679,
+      "loss": 0.6551,
+      "step": 319
+    },
+    {
+      "epoch": 0.7109136351013607,
+      "grad_norm": 0.02001245878636837,
+      "learning_rate": 0.00254320987654321,
+      "loss": 0.9046,
+      "step": 320
+    },
+    {
+      "epoch": 0.7131352402110525,
+      "grad_norm": 0.018720177933573723,
+      "learning_rate": 0.0025407407407407405,
+      "loss": 0.8019,
+      "step": 321
+    },
+    {
+      "epoch": 0.7153568453207443,
+      "grad_norm": 0.0072449627332389355,
+      "learning_rate": 0.0025382716049382717,
+      "loss": 0.8341,
+      "step": 322
+    },
+    {
+      "epoch": 0.717578450430436,
+      "grad_norm": 0.010015754029154778,
+      "learning_rate": 0.0025358024691358024,
+      "loss": 0.8745,
+      "step": 323
+    },
+    {
+      "epoch": 0.7198000555401277,
+      "grad_norm": 0.008088629692792892,
+      "learning_rate": 0.002533333333333333,
+      "loss": 0.8579,
+      "step": 324
+    },
+    {
+      "epoch": 0.7220216606498195,
+      "grad_norm": 0.014426672831177711,
+      "learning_rate": 0.0025308641975308644,
+      "loss": 0.8284,
+      "step": 325
+    },
+    {
+      "epoch": 0.7242432657595113,
+      "grad_norm": 0.006323180161416531,
+      "learning_rate": 0.002528395061728395,
+      "loss": 0.5411,
+      "step": 326
+    },
+    {
+      "epoch": 0.726464870869203,
+      "grad_norm": 0.0061256056651473045,
+      "learning_rate": 0.0025259259259259263,
+      "loss": 0.7823,
+      "step": 327
+    },
+    {
+      "epoch": 0.7286864759788948,
+      "grad_norm": 0.008604188449680805,
+      "learning_rate": 0.0025234567901234566,
+      "loss": 0.748,
+      "step": 328
+    },
+    {
+      "epoch": 0.7309080810885865,
+      "grad_norm": 0.006540137343108654,
+      "learning_rate": 0.0025209876543209877,
+      "loss": 0.6014,
+      "step": 329
+    },
+    {
+      "epoch": 0.7331296861982782,
+      "grad_norm": 0.007149749435484409,
+      "learning_rate": 0.0025185185185185185,
+      "loss": 0.7551,
+      "step": 330
+    },
+    {
+      "epoch": 0.7353512913079701,
+      "grad_norm": 0.005437437444925308,
+      "learning_rate": 0.0025160493827160496,
+      "loss": 0.7338,
+      "step": 331
+    },
+    {
+      "epoch": 0.7375728964176618,
+      "grad_norm": 0.010258332826197147,
+      "learning_rate": 0.0025135802469135804,
+      "loss": 0.7954,
+      "step": 332
+    },
+    {
+      "epoch": 0.7397945015273535,
+      "grad_norm": 0.007653366308659315,
+      "learning_rate": 0.002511111111111111,
+      "loss": 0.9539,
+      "step": 333
+    },
+    {
+      "epoch": 0.7420161066370453,
+      "grad_norm": 0.0065800040028989315,
+      "learning_rate": 0.0025086419753086423,
+      "loss": 0.7384,
+      "step": 334
+    },
+    {
+      "epoch": 0.744237711746737,
+      "grad_norm": 0.0080635417252779,
+      "learning_rate": 0.002506172839506173,
+      "loss": 0.976,
+      "step": 335
+    },
+    {
+      "epoch": 0.7464593168564287,
+      "grad_norm": 0.007618797477334738,
+      "learning_rate": 0.0025037037037037037,
+      "loss": 0.9015,
+      "step": 336
+    },
+    {
+      "epoch": 0.7486809219661206,
+      "grad_norm": 0.006701778620481491,
+      "learning_rate": 0.0025012345679012345,
+      "loss": 0.6757,
+      "step": 337
+    },
+    {
+      "epoch": 0.7509025270758123,
+      "grad_norm": 0.008706500753760338,
+      "learning_rate": 0.0024987654320987656,
+      "loss": 1.0567,
+      "step": 338
+    },
+    {
+      "epoch": 0.753124132185504,
+      "grad_norm": 0.006723032798618078,
+      "learning_rate": 0.0024962962962962964,
+      "loss": 0.7917,
+      "step": 339
+    },
+    {
+      "epoch": 0.7553457372951958,
+      "grad_norm": 0.004636780824512243,
+      "learning_rate": 0.002493827160493827,
+      "loss": 0.5242,
+      "step": 340
+    },
+    {
+      "epoch": 0.7575673424048875,
+      "grad_norm": 0.006264273542910814,
+      "learning_rate": 0.0024913580246913583,
+      "loss": 1.0525,
+      "step": 341
+    },
+    {
+      "epoch": 0.7597889475145793,
+      "grad_norm": 0.007951164618134499,
+      "learning_rate": 0.002488888888888889,
+      "loss": 0.7083,
+      "step": 342
+    },
+    {
+      "epoch": 0.7620105526242711,
+      "grad_norm": 0.004694493021816015,
+      "learning_rate": 0.0024864197530864197,
+      "loss": 0.776,
+      "step": 343
+    },
+    {
+      "epoch": 0.7642321577339628,
+      "grad_norm": 0.014506370760500431,
+      "learning_rate": 0.0024839506172839505,
+      "loss": 0.7931,
+      "step": 344
+    },
+    {
+      "epoch": 0.7664537628436545,
+      "grad_norm": 0.008699407801032066,
+      "learning_rate": 0.0024814814814814816,
+      "loss": 0.8931,
+      "step": 345
+    },
+    {
+      "epoch": 0.7686753679533463,
+      "grad_norm": 0.00661693187430501,
+      "learning_rate": 0.0024790123456790124,
+      "loss": 0.8803,
+      "step": 346
+    },
+    {
+      "epoch": 0.7708969730630381,
+      "grad_norm": 0.009916028007864952,
+      "learning_rate": 0.002476543209876543,
+      "loss": 0.7263,
+      "step": 347
+    },
+    {
+      "epoch": 0.7731185781727298,
+      "grad_norm": 0.009556926786899567,
+      "learning_rate": 0.0024740740740740743,
+      "loss": 0.8132,
+      "step": 348
+    },
+    {
+      "epoch": 0.7753401832824216,
+      "grad_norm": 0.011330212466418743,
+      "learning_rate": 0.002471604938271605,
+      "loss": 0.7426,
+      "step": 349
+    },
+    {
+      "epoch": 0.7775617883921133,
+      "grad_norm": 0.00721346540376544,
+      "learning_rate": 0.0024691358024691358,
+      "loss": 0.7328,
+      "step": 350
+    },
+    {
+      "epoch": 0.779783393501805,
+      "grad_norm": 0.007094467990100384,
+      "learning_rate": 0.0024666666666666665,
+      "loss": 0.7336,
+      "step": 351
+    },
+    {
+      "epoch": 0.7820049986114969,
+      "grad_norm": 0.0060173156671226025,
+      "learning_rate": 0.0024641975308641977,
+      "loss": 0.6998,
+      "step": 352
+    },
+    {
+      "epoch": 0.7842266037211886,
+      "grad_norm": 0.006421815603971481,
+      "learning_rate": 0.0024617283950617284,
+      "loss": 0.8042,
+      "step": 353
+    },
+    {
+      "epoch": 0.7864482088308803,
+      "grad_norm": 0.005537498742341995,
+      "learning_rate": 0.002459259259259259,
+      "loss": 0.6824,
+      "step": 354
+    },
+    {
+      "epoch": 0.7886698139405721,
+      "grad_norm": 0.007721144240349531,
+      "learning_rate": 0.0024567901234567903,
+      "loss": 0.7393,
+      "step": 355
+    },
+    {
+      "epoch": 0.7908914190502638,
+      "grad_norm": 0.006249431986361742,
+      "learning_rate": 0.002454320987654321,
+      "loss": 0.8343,
+      "step": 356
+    },
+    {
+      "epoch": 0.7931130241599555,
+      "grad_norm": 0.007465182337909937,
+      "learning_rate": 0.002451851851851852,
+      "loss": 0.6554,
+      "step": 357
+    },
+    {
+      "epoch": 0.7953346292696474,
+      "grad_norm": 0.007041784934699535,
+      "learning_rate": 0.0024493827160493825,
+      "loss": 0.7373,
+      "step": 358
+    },
+    {
+      "epoch": 0.7975562343793391,
+      "grad_norm": 0.006842833943665028,
+      "learning_rate": 0.0024469135802469137,
+      "loss": 0.8867,
+      "step": 359
+    },
+    {
+      "epoch": 0.7997778394890308,
+      "grad_norm": 0.009905919432640076,
+      "learning_rate": 0.0024444444444444444,
+      "loss": 0.7637,
+      "step": 360
+    },
+    {
+      "epoch": 0.8019994445987225,
+      "grad_norm": 0.010288221761584282,
+      "learning_rate": 0.0024419753086419756,
+      "loss": 0.8162,
+      "step": 361
+    },
+    {
+      "epoch": 0.8042210497084143,
+      "grad_norm": 0.010223974473774433,
+      "learning_rate": 0.0024395061728395063,
+      "loss": 0.8779,
+      "step": 362
+    },
+    {
+      "epoch": 0.8064426548181061,
+      "grad_norm": 0.012591023929417133,
+      "learning_rate": 0.002437037037037037,
+      "loss": 0.8765,
+      "step": 363
+    },
+    {
+      "epoch": 0.8086642599277978,
+      "grad_norm": 0.005684548057615757,
+      "learning_rate": 0.002434567901234568,
+      "loss": 0.7797,
+      "step": 364
+    },
+    {
+      "epoch": 0.8108858650374896,
+      "grad_norm": 0.0055415453389286995,
+      "learning_rate": 0.002432098765432099,
+      "loss": 0.6947,
+      "step": 365
+    },
+    {
+      "epoch": 0.8131074701471813,
+      "grad_norm": 0.006009736098349094,
+      "learning_rate": 0.0024296296296296297,
+      "loss": 0.9479,
+      "step": 366
+    },
+    {
+      "epoch": 0.815329075256873,
+      "grad_norm": 0.009372019208967686,
+      "learning_rate": 0.0024271604938271604,
+      "loss": 0.8523,
+      "step": 367
+    },
+    {
+      "epoch": 0.8175506803665649,
+      "grad_norm": 0.007912317290902138,
+      "learning_rate": 0.0024246913580246916,
+      "loss": 0.6513,
+      "step": 368
+    },
+    {
+      "epoch": 0.8197722854762566,
+      "grad_norm": 0.011661085300147533,
+      "learning_rate": 0.0024222222222222223,
+      "loss": 0.8989,
+      "step": 369
+    },
+    {
+      "epoch": 0.8219938905859483,
+      "grad_norm": 0.011349090375006199,
+      "learning_rate": 0.002419753086419753,
+      "loss": 0.9422,
+      "step": 370
+    },
+    {
+      "epoch": 0.8242154956956401,
+      "grad_norm": 0.006582579109817743,
+      "learning_rate": 0.002417283950617284,
+      "loss": 0.7312,
+      "step": 371
+    },
+    {
+      "epoch": 0.8264371008053318,
+      "grad_norm": 0.006679638754576445,
+      "learning_rate": 0.002414814814814815,
+      "loss": 0.889,
+      "step": 372
+    },
+    {
+      "epoch": 0.8286587059150236,
+      "grad_norm": 0.006373196840286255,
+      "learning_rate": 0.0024123456790123457,
+      "loss": 0.6953,
+      "step": 373
+    },
+    {
+      "epoch": 0.8308803110247154,
+      "grad_norm": 0.0081919701769948,
+      "learning_rate": 0.0024098765432098764,
+      "loss": 0.9879,
+      "step": 374
+    },
+    {
+      "epoch": 0.8331019161344071,
+      "grad_norm": 0.010226764716207981,
+      "learning_rate": 0.0024074074074074076,
+      "loss": 1.0487,
+      "step": 375
+    },
+    {
+      "epoch": 0.8353235212440988,
+      "grad_norm": 0.00779406912624836,
+      "learning_rate": 0.0024049382716049383,
+      "loss": 0.8523,
+      "step": 376
+    },
+    {
+      "epoch": 0.8375451263537906,
+      "grad_norm": 0.007026501465588808,
+      "learning_rate": 0.002402469135802469,
+      "loss": 0.6678,
+      "step": 377
+    },
+    {
+      "epoch": 0.8397667314634824,
+      "grad_norm": 0.014233388006687164,
+      "learning_rate": 0.0024000000000000002,
+      "loss": 1.0843,
+      "step": 378
+    },
+    {
+      "epoch": 0.8419883365731741,
+      "grad_norm": 0.009694437496364117,
+      "learning_rate": 0.002397530864197531,
+      "loss": 0.7636,
+      "step": 379
+    },
+    {
+      "epoch": 0.8442099416828659,
+      "grad_norm": 0.007724369410425425,
+      "learning_rate": 0.0023950617283950617,
+      "loss": 0.6192,
+      "step": 380
+    },
+    {
+      "epoch": 0.8464315467925576,
+      "grad_norm": 0.009011819027364254,
+      "learning_rate": 0.0023925925925925924,
+      "loss": 0.8457,
+      "step": 381
+    },
+    {
+      "epoch": 0.8486531519022493,
+      "grad_norm": 0.010344129055738449,
+      "learning_rate": 0.0023901234567901236,
+      "loss": 0.7277,
+      "step": 382
+    },
+    {
+      "epoch": 0.8508747570119412,
+      "grad_norm": 0.005570253822952509,
+      "learning_rate": 0.0023876543209876543,
+      "loss": 0.8237,
+      "step": 383
+    },
+    {
+      "epoch": 0.8530963621216329,
+      "grad_norm": 0.00868288055062294,
+      "learning_rate": 0.002385185185185185,
+      "loss": 0.7281,
+      "step": 384
+    },
+    {
+      "epoch": 0.8553179672313246,
+      "grad_norm": 0.009439078159630299,
+      "learning_rate": 0.0023827160493827162,
+      "loss": 0.9108,
+      "step": 385
+    },
+    {
+      "epoch": 0.8575395723410164,
+      "grad_norm": 0.01069068256765604,
+      "learning_rate": 0.002380246913580247,
+      "loss": 0.7893,
+      "step": 386
+    },
+    {
+      "epoch": 0.8597611774507081,
+      "grad_norm": 0.006505821831524372,
+      "learning_rate": 0.002377777777777778,
+      "loss": 0.7375,
+      "step": 387
+    },
+    {
+      "epoch": 0.8619827825603998,
+      "grad_norm": 0.006728131324052811,
+      "learning_rate": 0.0023753086419753084,
+      "loss": 0.883,
+      "step": 388
+    },
+    {
+      "epoch": 0.8642043876700917,
+      "grad_norm": 0.008597608655691147,
+      "learning_rate": 0.0023728395061728396,
+      "loss": 0.7026,
+      "step": 389
+    },
+    {
+      "epoch": 0.8664259927797834,
+      "grad_norm": 0.008809217251837254,
+      "learning_rate": 0.0023703703703703703,
+      "loss": 0.881,
+      "step": 390
+    },
+    {
+      "epoch": 0.8686475978894751,
+      "grad_norm": 0.008603409864008427,
+      "learning_rate": 0.0023679012345679015,
+      "loss": 0.6806,
+      "step": 391
+    },
+    {
+      "epoch": 0.8708692029991669,
+      "grad_norm": 0.011070380918681622,
+      "learning_rate": 0.0023654320987654322,
+      "loss": 0.8248,
+      "step": 392
+    },
+    {
+      "epoch": 0.8730908081088586,
+      "grad_norm": 0.011338909156620502,
+      "learning_rate": 0.002362962962962963,
+      "loss": 0.7086,
+      "step": 393
+    },
+    {
+      "epoch": 0.8753124132185504,
+      "grad_norm": 0.00947530660778284,
+      "learning_rate": 0.002360493827160494,
+      "loss": 1.0779,
+      "step": 394
+    },
+    {
+      "epoch": 0.8775340183282422,
+      "grad_norm": 0.004716199357062578,
+      "learning_rate": 0.002358024691358025,
+      "loss": 0.7149,
+      "step": 395
+    },
+    {
+      "epoch": 0.8797556234379339,
+      "grad_norm": 0.009315990842878819,
+      "learning_rate": 0.0023555555555555556,
+      "loss": 0.7395,
+      "step": 396
+    },
+    {
+      "epoch": 0.8819772285476256,
+      "grad_norm": 0.007294144481420517,
+      "learning_rate": 0.0023530864197530863,
+      "loss": 0.7305,
+      "step": 397
+    },
+    {
+      "epoch": 0.8841988336573174,
+      "grad_norm": 0.015177429653704166,
+      "learning_rate": 0.0023506172839506175,
+      "loss": 0.8675,
+      "step": 398
+    },
+    {
+      "epoch": 0.8864204387670092,
+      "grad_norm": 0.007899096235632896,
+      "learning_rate": 0.002348148148148148,
+      "loss": 0.7365,
+      "step": 399
+    },
+    {
+      "epoch": 0.8886420438767009,
+      "grad_norm": 0.007007989101111889,
+      "learning_rate": 0.002345679012345679,
+      "loss": 0.7847,
+      "step": 400
+    },
+    {
+      "epoch": 0.8908636489863927,
+      "grad_norm": 0.006175495218485594,
+      "learning_rate": 0.00234320987654321,
+      "loss": 0.8364,
+      "step": 401
+    },
+    {
+      "epoch": 0.8930852540960844,
+      "grad_norm": 0.00996660441160202,
+      "learning_rate": 0.002340740740740741,
+      "loss": 0.824,
+      "step": 402
+    },
+    {
+      "epoch": 0.8953068592057761,
+      "grad_norm": 0.007697720546275377,
+      "learning_rate": 0.0023382716049382716,
+      "loss": 0.8092,
+      "step": 403
+    },
+    {
+      "epoch": 0.897528464315468,
+      "grad_norm": 0.02166207879781723,
+      "learning_rate": 0.0023358024691358024,
+      "loss": 0.8932,
+      "step": 404
+    },
+    {
+      "epoch": 0.8997500694251597,
+      "grad_norm": 0.011461895890533924,
+      "learning_rate": 0.0023333333333333335,
+      "loss": 0.7087,
+      "step": 405
+    },
+    {
+      "epoch": 0.9019716745348514,
+      "grad_norm": 0.007513254880905151,
+      "learning_rate": 0.0023308641975308643,
+      "loss": 0.9315,
+      "step": 406
+    },
+    {
+      "epoch": 0.9041932796445432,
+      "grad_norm": 0.005459482315927744,
+      "learning_rate": 0.002328395061728395,
+      "loss": 0.7973,
+      "step": 407
+    },
+    {
+      "epoch": 0.9064148847542349,
+      "grad_norm": 0.01074253860861063,
+      "learning_rate": 0.002325925925925926,
+      "loss": 0.7004,
+      "step": 408
+    },
+    {
+      "epoch": 0.9086364898639266,
+      "grad_norm": 0.011036701500415802,
+      "learning_rate": 0.002323456790123457,
+      "loss": 1.0297,
+      "step": 409
+    },
+    {
+      "epoch": 0.9108580949736185,
+      "grad_norm": 0.0065698218531906605,
+      "learning_rate": 0.0023209876543209876,
+      "loss": 0.8523,
+      "step": 410
+    },
+    {
+      "epoch": 0.9130797000833102,
+      "grad_norm": 0.008030000142753124,
+      "learning_rate": 0.0023185185185185184,
+      "loss": 0.6409,
+      "step": 411
+    },
+    {
+      "epoch": 0.9153013051930019,
+      "grad_norm": 0.010795329697430134,
+      "learning_rate": 0.0023160493827160495,
+      "loss": 0.8743,
+      "step": 412
+    },
+    {
+      "epoch": 0.9175229103026937,
+      "grad_norm": 0.00822600070387125,
+      "learning_rate": 0.0023135802469135803,
+      "loss": 0.8858,
+      "step": 413
+    },
+    {
+      "epoch": 0.9197445154123854,
+      "grad_norm": 0.006053009070456028,
+      "learning_rate": 0.002311111111111111,
+      "loss": 0.8098,
+      "step": 414
+    },
+    {
+      "epoch": 0.9219661205220772,
+      "grad_norm": 0.012354751117527485,
+      "learning_rate": 0.002308641975308642,
+      "loss": 0.8258,
+      "step": 415
+    },
+    {
+      "epoch": 0.924187725631769,
+      "grad_norm": 0.007067359518259764,
+      "learning_rate": 0.002306172839506173,
+      "loss": 0.6538,
+      "step": 416
+    },
+    {
+      "epoch": 0.9264093307414607,
+      "grad_norm": 0.005133168306201696,
+      "learning_rate": 0.0023037037037037036,
+      "loss": 0.7096,
+      "step": 417
+    },
+    {
+      "epoch": 0.9286309358511524,
+      "grad_norm": 0.00972819048911333,
+      "learning_rate": 0.0023012345679012344,
+      "loss": 0.7409,
+      "step": 418
+    },
+    {
+      "epoch": 0.9308525409608442,
+      "grad_norm": 0.009975786320865154,
+      "learning_rate": 0.0022987654320987655,
+      "loss": 0.6634,
+      "step": 419
+    },
+    {
+      "epoch": 0.933074146070536,
+      "grad_norm": 0.007423476316034794,
+      "learning_rate": 0.0022962962962962963,
+      "loss": 0.893,
+      "step": 420
+    },
+    {
+      "epoch": 0.9352957511802277,
+      "grad_norm": 0.011143303476274014,
+      "learning_rate": 0.002293827160493827,
+      "loss": 0.9115,
+      "step": 421
+    },
+    {
+      "epoch": 0.9375173562899195,
+      "grad_norm": 0.011022884398698807,
+      "learning_rate": 0.002291358024691358,
+      "loss": 0.8602,
+      "step": 422
+    },
+    {
+      "epoch": 0.9397389613996112,
+      "grad_norm": 0.012729224748909473,
+      "learning_rate": 0.002288888888888889,
+      "loss": 0.8307,
+      "step": 423
+    },
+    {
+      "epoch": 0.9419605665093029,
+      "grad_norm": 0.011760844849050045,
+      "learning_rate": 0.00228641975308642,
+      "loss": 0.806,
+      "step": 424
+    },
+    {
+      "epoch": 0.9441821716189948,
+      "grad_norm": 0.007908307015895844,
+      "learning_rate": 0.0022839506172839504,
+      "loss": 0.8382,
+      "step": 425
+    },
+    {
+      "epoch": 0.9464037767286865,
+      "grad_norm": 0.008884398266673088,
+      "learning_rate": 0.0022814814814814816,
+      "loss": 0.9276,
+      "step": 426
+    },
+    {
+      "epoch": 0.9486253818383782,
+      "grad_norm": 0.008932596072554588,
+      "learning_rate": 0.0022790123456790123,
+      "loss": 0.838,
+      "step": 427
+    },
+    {
+      "epoch": 0.95084698694807,
+      "grad_norm": 0.008600243367254734,
+      "learning_rate": 0.0022765432098765435,
+      "loss": 0.726,
+      "step": 428
+    },
+    {
+      "epoch": 0.9530685920577617,
+      "grad_norm": 0.006019454915076494,
+      "learning_rate": 0.0022740740740740738,
+      "loss": 0.6559,
+      "step": 429
+    },
+    {
+      "epoch": 0.9552901971674534,
+      "grad_norm": 0.0076181404292583466,
+      "learning_rate": 0.002271604938271605,
+      "loss": 0.7481,
+      "step": 430
+    },
+    {
+      "epoch": 0.9575118022771453,
+      "grad_norm": 0.012168084271252155,
+      "learning_rate": 0.002269135802469136,
+      "loss": 0.8652,
+      "step": 431
+    },
+    {
+      "epoch": 0.959733407386837,
+      "grad_norm": 0.007883667014539242,
+      "learning_rate": 0.002266666666666667,
+      "loss": 0.8239,
+      "step": 432
+    },
+    {
+      "epoch": 0.9619550124965287,
+      "grad_norm": 0.009150558151304722,
+      "learning_rate": 0.0022641975308641976,
+      "loss": 0.6909,
+      "step": 433
+    },
+    {
+      "epoch": 0.9641766176062205,
+      "grad_norm": 0.010145551525056362,
+      "learning_rate": 0.0022617283950617283,
+      "loss": 0.6662,
+      "step": 434
+    },
+    {
+      "epoch": 0.9663982227159122,
+      "grad_norm": 0.00680849514901638,
+      "learning_rate": 0.0022592592592592595,
+      "loss": 0.6533,
+      "step": 435
+    },
+    {
+      "epoch": 0.968619827825604,
+      "grad_norm": 0.01036826428025961,
+      "learning_rate": 0.00225679012345679,
+      "loss": 0.8156,
+      "step": 436
+    },
+    {
+      "epoch": 0.9708414329352958,
+      "grad_norm": 0.008677355945110321,
+      "learning_rate": 0.002254320987654321,
+      "loss": 0.8441,
+      "step": 437
+    },
+    {
+      "epoch": 0.9730630380449875,
+      "grad_norm": 0.00549481064081192,
+      "learning_rate": 0.002251851851851852,
+      "loss": 0.7917,
+      "step": 438
+    },
+    {
+      "epoch": 0.9752846431546792,
+      "grad_norm": 0.006068227346986532,
+      "learning_rate": 0.002249382716049383,
+      "loss": 0.8223,
+      "step": 439
+    },
+    {
+      "epoch": 0.977506248264371,
+      "grad_norm": 0.011117229238152504,
+      "learning_rate": 0.0022469135802469136,
+      "loss": 0.8663,
+      "step": 440
+    },
+    {
+      "epoch": 0.9797278533740628,
+      "grad_norm": 0.004921234678477049,
+      "learning_rate": 0.0022444444444444443,
+      "loss": 0.9496,
+      "step": 441
+    },
+    {
+      "epoch": 0.9819494584837545,
+      "grad_norm": 0.007072269916534424,
+      "learning_rate": 0.0022419753086419755,
+      "loss": 0.8655,
+      "step": 442
+    },
+    {
+      "epoch": 0.9841710635934463,
+      "grad_norm": 0.008353911340236664,
+      "learning_rate": 0.002239506172839506,
+      "loss": 0.6893,
+      "step": 443
+    },
+    {
+      "epoch": 0.986392668703138,
+      "grad_norm": 0.021165434271097183,
+      "learning_rate": 0.002237037037037037,
+      "loss": 0.6942,
+      "step": 444
+    },
+    {
+      "epoch": 0.9886142738128297,
+      "grad_norm": 0.008496430702507496,
+      "learning_rate": 0.002234567901234568,
+      "loss": 0.5909,
+      "step": 445
+    },
+    {
+      "epoch": 0.9908358789225216,
+      "grad_norm": 0.008113761432468891,
+      "learning_rate": 0.002232098765432099,
+      "loss": 0.9439,
+      "step": 446
+    },
+    {
+      "epoch": 0.9930574840322133,
+      "grad_norm": 0.013667792081832886,
+      "learning_rate": 0.0022296296296296296,
+      "loss": 0.961,
+      "step": 447
+    },
+    {
+      "epoch": 0.995279089141905,
+      "grad_norm": 0.010663739405572414,
+      "learning_rate": 0.0022271604938271603,
+      "loss": 0.9028,
+      "step": 448
+    },
+    {
+      "epoch": 0.9975006942515968,
+      "grad_norm": 0.008401602506637573,
+      "learning_rate": 0.0022246913580246915,
+      "loss": 0.9042,
+      "step": 449
+    },
+    {
+      "epoch": 0.9997222993612885,
+      "grad_norm": 0.005140175111591816,
+      "learning_rate": 0.0022222222222222222,
+      "loss": 0.7971,
+      "step": 450
+    },
+    {
+      "epoch": 1.0019439044709804,
+      "grad_norm": 0.007235647179186344,
+      "learning_rate": 0.002219753086419753,
+      "loss": 0.9074,
+      "step": 451
+    },
+    {
+      "epoch": 1.004165509580672,
+      "grad_norm": 0.008099445141851902,
+      "learning_rate": 0.002217283950617284,
+      "loss": 0.7123,
+      "step": 452
+    },
+    {
+      "epoch": 1.0063871146903638,
+      "grad_norm": 0.008628906682133675,
+      "learning_rate": 0.002214814814814815,
+      "loss": 0.7415,
+      "step": 453
+    },
+    {
+      "epoch": 1.0086087198000555,
+      "grad_norm": 0.006053466349840164,
+      "learning_rate": 0.002212345679012346,
+      "loss": 0.7232,
+      "step": 454
+    },
+    {
+      "epoch": 1.0108303249097472,
+      "grad_norm": 0.008410786278545856,
+      "learning_rate": 0.0022098765432098763,
+      "loss": 0.6775,
+      "step": 455
+    },
+    {
+      "epoch": 1.013051930019439,
+      "grad_norm": 0.005019874777644873,
+      "learning_rate": 0.0022074074074074075,
+      "loss": 0.6729,
+      "step": 456
+    },
+    {
+      "epoch": 1.0152735351291309,
+      "grad_norm": 0.00676839891821146,
+      "learning_rate": 0.0022049382716049382,
+      "loss": 0.7891,
+      "step": 457
+    },
+    {
+      "epoch": 1.0174951402388226,
+      "grad_norm": 0.006478031165897846,
+      "learning_rate": 0.0022024691358024694,
+      "loss": 0.5783,
+      "step": 458
+    },
+    {
+      "epoch": 1.0197167453485143,
+      "grad_norm": 0.010523135773837566,
+      "learning_rate": 0.0021999999999999997,
+      "loss": 0.7024,
+      "step": 459
+    },
+    {
+      "epoch": 1.021938350458206,
+      "grad_norm": 0.012737292796373367,
+      "learning_rate": 0.002197530864197531,
+      "loss": 0.8262,
+      "step": 460
+    },
+    {
+      "epoch": 1.0241599555678977,
+      "grad_norm": 0.0061058285646140575,
+      "learning_rate": 0.002195061728395062,
+      "loss": 0.8292,
+      "step": 461
+    },
+    {
+      "epoch": 1.0263815606775895,
+      "grad_norm": 0.01105250883847475,
+      "learning_rate": 0.0021925925925925928,
+      "loss": 0.7996,
+      "step": 462
+    },
+    {
+      "epoch": 1.0286031657872814,
+      "grad_norm": 0.008844085969030857,
+      "learning_rate": 0.0021901234567901235,
+      "loss": 0.7622,
+      "step": 463
+    },
+    {
+      "epoch": 1.030824770896973,
+      "grad_norm": 0.005652655381709337,
+      "learning_rate": 0.0021876543209876542,
+      "loss": 0.8272,
+      "step": 464
+    },
+    {
+      "epoch": 1.0330463760066648,
+      "grad_norm": 0.009466007351875305,
+      "learning_rate": 0.0021851851851851854,
+      "loss": 0.9157,
+      "step": 465
+    },
+    {
+      "epoch": 1.0352679811163565,
+      "grad_norm": 0.007364574354141951,
+      "learning_rate": 0.002182716049382716,
+      "loss": 0.6422,
+      "step": 466
+    },
+    {
+      "epoch": 1.0374895862260483,
+      "grad_norm": 0.008415921591222286,
+      "learning_rate": 0.002180246913580247,
+      "loss": 0.7811,
+      "step": 467
+    },
+    {
+      "epoch": 1.03971119133574,
+      "grad_norm": 0.007536835968494415,
+      "learning_rate": 0.002177777777777778,
+      "loss": 0.6505,
+      "step": 468
+    },
+    {
+      "epoch": 1.041932796445432,
+      "grad_norm": 0.006799593102186918,
+      "learning_rate": 0.0021753086419753088,
+      "loss": 0.7923,
+      "step": 469
+    },
+    {
+      "epoch": 1.0441544015551236,
+      "grad_norm": 0.00823921337723732,
+      "learning_rate": 0.0021728395061728395,
+      "loss": 0.7098,
+      "step": 470
+    },
+    {
+      "epoch": 1.0463760066648153,
+      "grad_norm": 0.008536463603377342,
+      "learning_rate": 0.0021703703703703702,
+      "loss": 0.7643,
+      "step": 471
+    },
+    {
+      "epoch": 1.048597611774507,
+      "grad_norm": 0.008986718021333218,
+      "learning_rate": 0.0021679012345679014,
+      "loss": 0.8968,
+      "step": 472
+    },
+    {
+      "epoch": 1.0508192168841988,
+      "grad_norm": 0.010182259604334831,
+      "learning_rate": 0.002165432098765432,
+      "loss": 0.7392,
+      "step": 473
+    },
+    {
+      "epoch": 1.0530408219938905,
+      "grad_norm": 0.007251216098666191,
+      "learning_rate": 0.002162962962962963,
+      "loss": 0.7097,
+      "step": 474
+    },
+    {
+      "epoch": 1.0552624271035824,
+      "grad_norm": 0.007560486439615488,
+      "learning_rate": 0.002160493827160494,
+      "loss": 0.7698,
+      "step": 475
+    },
+    {
+      "epoch": 1.0574840322132741,
+      "grad_norm": 0.0077867708168923855,
+      "learning_rate": 0.002158024691358025,
+      "loss": 0.8747,
+      "step": 476
+    },
+    {
+      "epoch": 1.0597056373229659,
+      "grad_norm": 0.013194149360060692,
+      "learning_rate": 0.0021555555555555555,
+      "loss": 0.8222,
+      "step": 477
+    },
+    {
+      "epoch": 1.0619272424326576,
+      "grad_norm": 0.00997414905577898,
+      "learning_rate": 0.0021530864197530863,
+      "loss": 0.663,
+      "step": 478
+    },
+    {
+      "epoch": 1.0641488475423493,
+      "grad_norm": 0.011852915398776531,
+      "learning_rate": 0.0021506172839506174,
+      "loss": 0.6943,
+      "step": 479
+    },
+    {
+      "epoch": 1.066370452652041,
+      "grad_norm": 0.008246763609349728,
+      "learning_rate": 0.002148148148148148,
+      "loss": 0.6128,
+      "step": 480
+    },
+    {
+      "epoch": 1.068592057761733,
+      "grad_norm": 0.007343011442571878,
+      "learning_rate": 0.002145679012345679,
+      "loss": 0.8747,
+      "step": 481
+    },
+    {
+      "epoch": 1.0708136628714247,
+      "grad_norm": 0.007888510823249817,
+      "learning_rate": 0.0021432098765432096,
+      "loss": 0.8691,
+      "step": 482
+    },
+    {
+      "epoch": 1.0730352679811164,
+      "grad_norm": 0.008748864755034447,
+      "learning_rate": 0.002140740740740741,
+      "loss": 0.8701,
+      "step": 483
+    },
+    {
+      "epoch": 1.075256873090808,
+      "grad_norm": 0.006701997015625238,
+      "learning_rate": 0.002138271604938272,
+      "loss": 0.9312,
+      "step": 484
+    },
+    {
+      "epoch": 1.0774784782004998,
+      "grad_norm": 0.006520512979477644,
+      "learning_rate": 0.0021358024691358023,
+      "loss": 0.9224,
+      "step": 485
+    },
+    {
+      "epoch": 1.0797000833101915,
+      "grad_norm": 0.011394795030355453,
+      "learning_rate": 0.0021333333333333334,
+      "loss": 0.8161,
+      "step": 486
+    },
+    {
+      "epoch": 1.0819216884198835,
+      "grad_norm": 0.012948388233780861,
+      "learning_rate": 0.002130864197530864,
+      "loss": 0.8619,
+      "step": 487
+    },
+    {
+      "epoch": 1.0841432935295752,
+      "grad_norm": 0.005069487262517214,
+      "learning_rate": 0.0021283950617283953,
+      "loss": 0.8027,
+      "step": 488
+    },
+    {
+      "epoch": 1.0863648986392669,
+      "grad_norm": 0.008600343950092793,
+      "learning_rate": 0.0021259259259259256,
+      "loss": 0.8324,
+      "step": 489
+    },
+    {
+      "epoch": 1.0885865037489586,
+      "grad_norm": 0.00814684759825468,
+      "learning_rate": 0.002123456790123457,
+      "loss": 0.6381,
+      "step": 490
+    },
+    {
+      "epoch": 1.0908081088586503,
+      "grad_norm": 0.008729801513254642,
+      "learning_rate": 0.002120987654320988,
+      "loss": 0.6205,
+      "step": 491
+    },
+    {
+      "epoch": 1.093029713968342,
+      "grad_norm": 0.008048581890761852,
+      "learning_rate": 0.0021185185185185187,
+      "loss": 0.7394,
+      "step": 492
+    },
+    {
+      "epoch": 1.095251319078034,
+      "grad_norm": 0.008639739826321602,
+      "learning_rate": 0.0021160493827160494,
+      "loss": 0.9498,
+      "step": 493
+    },
+    {
+      "epoch": 1.0974729241877257,
+      "grad_norm": 0.006795950699597597,
+      "learning_rate": 0.00211358024691358,
+      "loss": 0.9062,
+      "step": 494
+    },
+    {
+      "epoch": 1.0996945292974174,
+      "grad_norm": 0.007198169827461243,
+      "learning_rate": 0.0021111111111111113,
+      "loss": 0.8697,
+      "step": 495
+    },
+    {
+      "epoch": 1.1019161344071091,
+      "grad_norm": 0.007898751646280289,
+      "learning_rate": 0.002108641975308642,
+      "loss": 0.7633,
+      "step": 496
+    },
+    {
+      "epoch": 1.1041377395168008,
+      "grad_norm": 0.008064448833465576,
+      "learning_rate": 0.002106172839506173,
+      "loss": 0.7843,
+      "step": 497
+    },
+    {
+      "epoch": 1.1063593446264925,
+      "grad_norm": 0.006913089193403721,
+      "learning_rate": 0.002103703703703704,
+      "loss": 0.7572,
+      "step": 498
+    },
+    {
+      "epoch": 1.1085809497361845,
+      "grad_norm": 0.007993088103830814,
+      "learning_rate": 0.0021012345679012347,
+      "loss": 0.7287,
+      "step": 499
+    },
+    {
+      "epoch": 1.1108025548458762,
+      "grad_norm": 0.007741628214716911,
+      "learning_rate": 0.0020987654320987655,
+      "loss": 0.7852,
+      "step": 500
+    },
+    {
+      "epoch": 1.113024159955568,
+      "grad_norm": 0.009709767997264862,
+      "learning_rate": 0.002096296296296296,
+      "loss": 0.8637,
+      "step": 501
+    },
+    {
+      "epoch": 1.1152457650652596,
+      "grad_norm": 0.010323646478354931,
+      "learning_rate": 0.0020938271604938274,
+      "loss": 0.8176,
+      "step": 502
+    },
+    {
+      "epoch": 1.1174673701749513,
+      "grad_norm": 0.013374350033700466,
+      "learning_rate": 0.002091358024691358,
+      "loss": 0.8289,
+      "step": 503
+    },
+    {
+      "epoch": 1.119688975284643,
+      "grad_norm": 0.010094034485518932,
+      "learning_rate": 0.002088888888888889,
+      "loss": 0.7587,
+      "step": 504
+    },
+    {
+      "epoch": 1.121910580394335,
+      "grad_norm": 0.006756503600627184,
+      "learning_rate": 0.00208641975308642,
+      "loss": 0.8676,
+      "step": 505
+    },
+    {
+      "epoch": 1.1241321855040267,
+      "grad_norm": 0.00773257901892066,
+      "learning_rate": 0.0020839506172839507,
+      "loss": 0.6533,
+      "step": 506
+    },
+    {
+      "epoch": 1.1263537906137184,
+      "grad_norm": 0.0079817408695817,
+      "learning_rate": 0.0020814814814814815,
+      "loss": 0.8949,
+      "step": 507
+    },
+    {
+      "epoch": 1.1285753957234101,
+      "grad_norm": 0.0096563920378685,
+      "learning_rate": 0.002079012345679012,
+      "loss": 0.8016,
+      "step": 508
+    },
+    {
+      "epoch": 1.1307970008331019,
+      "grad_norm": 0.006322795990854502,
+      "learning_rate": 0.0020765432098765434,
+      "loss": 0.7104,
+      "step": 509
+    },
+    {
+      "epoch": 1.1330186059427936,
+      "grad_norm": 0.008746332488954067,
+      "learning_rate": 0.002074074074074074,
+      "loss": 0.9368,
+      "step": 510
+    },
+    {
+      "epoch": 1.1352402110524855,
+      "grad_norm": 0.012176689691841602,
+      "learning_rate": 0.002071604938271605,
+      "loss": 0.8829,
+      "step": 511
+    },
+    {
+      "epoch": 1.1374618161621772,
+      "grad_norm": 0.00898228120058775,
+      "learning_rate": 0.0020691358024691356,
+      "loss": 0.7631,
+      "step": 512
+    },
+    {
+      "epoch": 1.139683421271869,
+      "grad_norm": 0.00945940800011158,
+      "learning_rate": 0.0020666666666666667,
+      "loss": 0.606,
+      "step": 513
+    },
+    {
+      "epoch": 1.1419050263815607,
+      "grad_norm": 0.00997233297675848,
+      "learning_rate": 0.002064197530864198,
+      "loss": 0.9461,
+      "step": 514
+    },
+    {
+      "epoch": 1.1441266314912524,
+      "grad_norm": 0.0064164213836193085,
+      "learning_rate": 0.002061728395061728,
+      "loss": 0.742,
+      "step": 515
+    },
+    {
+      "epoch": 1.146348236600944,
+      "grad_norm": 0.007206726353615522,
+      "learning_rate": 0.0020592592592592594,
+      "loss": 0.6104,
+      "step": 516
+    },
+    {
+      "epoch": 1.148569841710636,
+      "grad_norm": 0.008765126578509808,
+      "learning_rate": 0.00205679012345679,
+      "loss": 0.8508,
+      "step": 517
+    },
+    {
+      "epoch": 1.1507914468203277,
+      "grad_norm": 0.009035077877342701,
+      "learning_rate": 0.0020543209876543213,
+      "loss": 0.6652,
+      "step": 518
+    },
+    {
+      "epoch": 1.1530130519300195,
+      "grad_norm": 0.007268165238201618,
+      "learning_rate": 0.0020518518518518516,
+      "loss": 0.7535,
+      "step": 519
+    },
+    {
+      "epoch": 1.1552346570397112,
+      "grad_norm": 0.007550420239567757,
+      "learning_rate": 0.0020493827160493827,
+      "loss": 0.7689,
+      "step": 520
+    },
+    {
+      "epoch": 1.157456262149403,
+      "grad_norm": 0.008769848383963108,
+      "learning_rate": 0.002046913580246914,
+      "loss": 0.9409,
+      "step": 521
+    },
+    {
+      "epoch": 1.1596778672590946,
+      "grad_norm": 0.007353247608989477,
+      "learning_rate": 0.0020444444444444447,
+      "loss": 0.6995,
+      "step": 522
+    },
+    {
+      "epoch": 1.1618994723687863,
+      "grad_norm": 0.010628338903188705,
+      "learning_rate": 0.0020419753086419754,
+      "loss": 0.9438,
+      "step": 523
+    },
+    {
+      "epoch": 1.1641210774784783,
+      "grad_norm": 0.011336445808410645,
+      "learning_rate": 0.002039506172839506,
+      "loss": 0.6474,
+      "step": 524
+    },
+    {
+      "epoch": 1.16634268258817,
+      "grad_norm": 0.008600319735705853,
+      "learning_rate": 0.0020370370370370373,
+      "loss": 0.676,
+      "step": 525
+    },
+    {
+      "epoch": 1.1685642876978617,
+      "grad_norm": 0.007791724521666765,
+      "learning_rate": 0.002034567901234568,
+      "loss": 0.8491,
+      "step": 526
+    },
+    {
+      "epoch": 1.1707858928075534,
+      "grad_norm": 0.008302255533635616,
+      "learning_rate": 0.0020320987654320988,
+      "loss": 0.5874,
+      "step": 527
+    },
+    {
+      "epoch": 1.1730074979172451,
+      "grad_norm": 0.007213059347122908,
+      "learning_rate": 0.00202962962962963,
+      "loss": 0.6567,
+      "step": 528
+    },
+    {
+      "epoch": 1.175229103026937,
+      "grad_norm": 0.006716989912092686,
+      "learning_rate": 0.0020271604938271607,
+      "loss": 0.7785,
+      "step": 529
+    },
+    {
+      "epoch": 1.1774507081366288,
+      "grad_norm": 0.006225584540516138,
+      "learning_rate": 0.0020246913580246914,
+      "loss": 0.8877,
+      "step": 530
+    },
+    {
+      "epoch": 1.1796723132463205,
+      "grad_norm": 0.009407371282577515,
+      "learning_rate": 0.002022222222222222,
+      "loss": 0.6993,
+      "step": 531
+    },
+    {
+      "epoch": 1.1818939183560122,
+      "grad_norm": 0.004865545313805342,
+      "learning_rate": 0.0020197530864197533,
+      "loss": 0.7627,
+      "step": 532
+    },
+    {
+      "epoch": 1.184115523465704,
+      "grad_norm": 0.010101933032274246,
+      "learning_rate": 0.002017283950617284,
+      "loss": 0.8246,
+      "step": 533
+    },
+    {
+      "epoch": 1.1863371285753956,
+      "grad_norm": 0.011003025807440281,
+      "learning_rate": 0.0020148148148148148,
+      "loss": 0.6756,
+      "step": 534
+    },
+    {
+      "epoch": 1.1885587336850874,
+      "grad_norm": 0.01154054794460535,
+      "learning_rate": 0.002012345679012346,
+      "loss": 0.6787,
+      "step": 535
+    },
+    {
+      "epoch": 1.1907803387947793,
+      "grad_norm": 0.012816593050956726,
+      "learning_rate": 0.0020098765432098767,
+      "loss": 0.75,
+      "step": 536
+    },
+    {
+      "epoch": 1.193001943904471,
+      "grad_norm": 0.00780332088470459,
+      "learning_rate": 0.0020074074074074074,
+      "loss": 0.8453,
+      "step": 537
+    },
+    {
+      "epoch": 1.1952235490141627,
+      "grad_norm": 0.006366015411913395,
+      "learning_rate": 0.002004938271604938,
+      "loss": 0.6712,
+      "step": 538
+    },
+    {
+      "epoch": 1.1974451541238544,
+      "grad_norm": 0.004924730863422155,
+      "learning_rate": 0.0020024691358024693,
+      "loss": 0.8163,
+      "step": 539
+    },
+    {
+      "epoch": 1.1996667592335462,
+      "grad_norm": 0.005040746182203293,
+      "learning_rate": 0.002,
+      "loss": 0.7244,
+      "step": 540
+    },
+    {
+      "epoch": 1.201888364343238,
+      "grad_norm": 0.0060347081162035465,
+      "learning_rate": 0.0019975308641975308,
+      "loss": 0.7421,
+      "step": 541
+    },
+    {
+      "epoch": 1.2041099694529298,
+      "grad_norm": 0.00597890792414546,
+      "learning_rate": 0.0019950617283950615,
+      "loss": 0.7418,
+      "step": 542
+    },
+    {
+      "epoch": 1.2063315745626215,
+      "grad_norm": 0.0050376360304653645,
+      "learning_rate": 0.0019925925925925927,
+      "loss": 0.7606,
+      "step": 543
+    },
+    {
+      "epoch": 1.2085531796723132,
+      "grad_norm": 0.0056321881711483,
+      "learning_rate": 0.001990123456790124,
+      "loss": 0.8881,
+      "step": 544
+    },
+    {
+      "epoch": 1.210774784782005,
+      "grad_norm": 0.006516937632113695,
+      "learning_rate": 0.001987654320987654,
+      "loss": 0.7048,
+      "step": 545
+    },
+    {
+      "epoch": 1.2129963898916967,
+      "grad_norm": 0.004874023143202066,
+      "learning_rate": 0.0019851851851851853,
+      "loss": 0.8859,
+      "step": 546
+    },
+    {
+      "epoch": 1.2152179950013884,
+      "grad_norm": 0.007384075317531824,
+      "learning_rate": 0.001982716049382716,
+      "loss": 0.6933,
+      "step": 547
+    },
+    {
+      "epoch": 1.2174396001110803,
+      "grad_norm": 0.005381579045206308,
+      "learning_rate": 0.001980246913580247,
+      "loss": 0.7698,
+      "step": 548
+    },
+    {
+      "epoch": 1.219661205220772,
+      "grad_norm": 0.008545870892703533,
+      "learning_rate": 0.0019777777777777775,
+      "loss": 0.6051,
+      "step": 549
+    },
+    {
+      "epoch": 1.2218828103304638,
+      "grad_norm": 0.007204385939985514,
+      "learning_rate": 0.0019753086419753087,
+      "loss": 0.668,
+      "step": 550
+    },
+    {
+      "epoch": 1.2241044154401555,
+      "grad_norm": 0.021104220300912857,
+      "learning_rate": 0.00197283950617284,
+      "loss": 0.7877,
+      "step": 551
+    },
+    {
+      "epoch": 1.2263260205498472,
+      "grad_norm": 0.009946838952600956,
+      "learning_rate": 0.00197037037037037,
+      "loss": 0.6948,
+      "step": 552
+    },
+    {
+      "epoch": 1.2285476256595391,
+      "grad_norm": 0.007187278009951115,
+      "learning_rate": 0.0019679012345679013,
+      "loss": 0.6236,
+      "step": 553
+    },
+    {
+      "epoch": 1.2307692307692308,
+      "grad_norm": 0.009974513202905655,
+      "learning_rate": 0.001965432098765432,
+      "loss": 0.9304,
+      "step": 554
+    },
+    {
+      "epoch": 1.2329908358789226,
+      "grad_norm": 0.007199294865131378,
+      "learning_rate": 0.0019629629629629632,
+      "loss": 0.8051,
+      "step": 555
+    },
+    {
+      "epoch": 1.2352124409886143,
+      "grad_norm": 0.012498236261308193,
+      "learning_rate": 0.0019604938271604935,
+      "loss": 0.9383,
+      "step": 556
+    },
+    {
+      "epoch": 1.237434046098306,
+      "grad_norm": 0.009159721434116364,
+      "learning_rate": 0.0019580246913580247,
+      "loss": 0.8871,
+      "step": 557
+    },
+    {
+      "epoch": 1.2396556512079977,
+      "grad_norm": 0.006070270668715239,
+      "learning_rate": 0.001955555555555556,
+      "loss": 0.7775,
+      "step": 558
+    },
+    {
+      "epoch": 1.2418772563176894,
+      "grad_norm": 0.009294399060308933,
+      "learning_rate": 0.0019530864197530864,
+      "loss": 0.8176,
+      "step": 559
+    },
+    {
+      "epoch": 1.2440988614273814,
+      "grad_norm": 0.009688743390142918,
+      "learning_rate": 0.0019506172839506173,
+      "loss": 0.9653,
+      "step": 560
+    },
+    {
+      "epoch": 1.246320466537073,
+      "grad_norm": 0.011101442389190197,
+      "learning_rate": 0.001948148148148148,
+      "loss": 1.107,
+      "step": 561
+    },
+    {
+      "epoch": 1.2485420716467648,
+      "grad_norm": 0.006938232574611902,
+      "learning_rate": 0.001945679012345679,
+      "loss": 0.723,
+      "step": 562
+    },
+    {
+      "epoch": 1.2507636767564565,
+      "grad_norm": 0.010605933144688606,
+      "learning_rate": 0.0019432098765432098,
+      "loss": 0.7292,
+      "step": 563
+    },
+    {
+      "epoch": 1.2529852818661482,
+      "grad_norm": 0.008306719362735748,
+      "learning_rate": 0.0019407407407407407,
+      "loss": 0.7772,
+      "step": 564
+    },
+    {
+      "epoch": 1.2552068869758402,
+      "grad_norm": 0.006135668605566025,
+      "learning_rate": 0.0019382716049382714,
+      "loss": 0.6693,
+      "step": 565
+    },
+    {
+      "epoch": 1.2574284920855319,
+      "grad_norm": 0.01065669022500515,
+      "learning_rate": 0.0019358024691358024,
+      "loss": 0.7522,
+      "step": 566
+    },
+    {
+      "epoch": 1.2596500971952236,
+      "grad_norm": 0.009806640446186066,
+      "learning_rate": 0.0019333333333333336,
+      "loss": 0.9129,
+      "step": 567
+    },
+    {
+      "epoch": 1.2618717023049153,
+      "grad_norm": 0.014492296613752842,
+      "learning_rate": 0.001930864197530864,
+      "loss": 0.8762,
+      "step": 568
+    },
+    {
+      "epoch": 1.264093307414607,
+      "grad_norm": 0.008053292520344257,
+      "learning_rate": 0.0019283950617283952,
+      "loss": 0.7453,
+      "step": 569
+    },
+    {
+      "epoch": 1.2663149125242987,
+      "grad_norm": 0.011233272030949593,
+      "learning_rate": 0.0019259259259259258,
+      "loss": 0.7827,
+      "step": 570
+    },
+    {
+      "epoch": 1.2685365176339904,
+      "grad_norm": 0.008643749170005322,
+      "learning_rate": 0.001923456790123457,
+      "loss": 0.8582,
+      "step": 571
+    },
+    {
+      "epoch": 1.2707581227436824,
+      "grad_norm": 0.008974653668701649,
+      "learning_rate": 0.0019209876543209875,
+      "loss": 0.8184,
+      "step": 572
+    },
+    {
+      "epoch": 1.272979727853374,
+      "grad_norm": 0.007029388565570116,
+      "learning_rate": 0.0019185185185185186,
+      "loss": 0.8485,
+      "step": 573
+    },
+    {
+      "epoch": 1.2752013329630658,
+      "grad_norm": 0.008042087778449059,
+      "learning_rate": 0.0019160493827160496,
+      "loss": 0.8451,
+      "step": 574
+    },
+    {
+      "epoch": 1.2774229380727575,
+      "grad_norm": 0.009854896925389767,
+      "learning_rate": 0.0019135802469135803,
+      "loss": 0.9475,
+      "step": 575
+    },
+    {
+      "epoch": 1.2796445431824492,
+      "grad_norm": 0.011435446329414845,
+      "learning_rate": 0.0019111111111111113,
+      "loss": 0.8207,
+      "step": 576
+    },
+    {
+      "epoch": 1.2818661482921412,
+      "grad_norm": 0.013869104906916618,
+      "learning_rate": 0.001908641975308642,
+      "loss": 0.647,
+      "step": 577
+    },
+    {
+      "epoch": 1.284087753401833,
+      "grad_norm": 0.008760456927120686,
+      "learning_rate": 0.001906172839506173,
+      "loss": 0.8233,
+      "step": 578
+    },
+    {
+      "epoch": 1.2863093585115246,
+      "grad_norm": 0.011544842272996902,
+      "learning_rate": 0.0019037037037037037,
+      "loss": 0.6646,
+      "step": 579
+    },
+    {
+      "epoch": 1.2885309636212163,
+      "grad_norm": 0.010519598610699177,
+      "learning_rate": 0.0019012345679012346,
+      "loss": 0.6739,
+      "step": 580
+    },
+    {
+      "epoch": 1.290752568730908,
+      "grad_norm": 0.008361653424799442,
+      "learning_rate": 0.0018987654320987656,
+      "loss": 0.7724,
+      "step": 581
+    },
+    {
+      "epoch": 1.2929741738405998,
+      "grad_norm": 0.007549158297479153,
+      "learning_rate": 0.0018962962962962963,
+      "loss": 0.6881,
+      "step": 582
+    },
+    {
+      "epoch": 1.2951957789502915,
+      "grad_norm": 0.0081331180408597,
+      "learning_rate": 0.0018938271604938273,
+      "loss": 0.754,
+      "step": 583
+    },
+    {
+      "epoch": 1.2974173840599834,
+      "grad_norm": 0.008421644568443298,
+      "learning_rate": 0.001891358024691358,
+      "loss": 0.772,
+      "step": 584
+    },
+    {
+      "epoch": 1.2996389891696751,
+      "grad_norm": 0.00717672286555171,
+      "learning_rate": 0.001888888888888889,
+      "loss": 0.79,
+      "step": 585
+    },
+    {
+      "epoch": 1.3018605942793668,
+      "grad_norm": 0.00759734446182847,
+      "learning_rate": 0.0018864197530864197,
+      "loss": 0.9811,
+      "step": 586
+    },
+    {
+      "epoch": 1.3040821993890586,
+      "grad_norm": 0.009103677235543728,
+      "learning_rate": 0.0018839506172839506,
+      "loss": 0.7995,
+      "step": 587
+    },
+    {
+      "epoch": 1.3063038044987503,
+      "grad_norm": 0.009735452011227608,
+      "learning_rate": 0.0018814814814814816,
+      "loss": 0.8571,
+      "step": 588
+    },
+    {
+      "epoch": 1.3085254096084422,
+      "grad_norm": 0.010691888630390167,
+      "learning_rate": 0.0018790123456790123,
+      "loss": 0.7097,
+      "step": 589
+    },
+    {
+      "epoch": 1.310747014718134,
+      "grad_norm": 0.006676527205854654,
+      "learning_rate": 0.0018765432098765433,
+      "loss": 0.6608,
+      "step": 590
+    },
+    {
+      "epoch": 1.3129686198278256,
+      "grad_norm": 0.008960582315921783,
+      "learning_rate": 0.001874074074074074,
+      "loss": 0.6465,
+      "step": 591
+    },
+    {
+      "epoch": 1.3151902249375174,
+      "grad_norm": 0.0076140440069139,
+      "learning_rate": 0.001871604938271605,
+      "loss": 0.8781,
+      "step": 592
+    },
+    {
+      "epoch": 1.317411830047209,
+      "grad_norm": 0.006687042769044638,
+      "learning_rate": 0.0018691358024691357,
+      "loss": 0.8268,
+      "step": 593
+    },
+    {
+      "epoch": 1.3196334351569008,
+      "grad_norm": 0.011486892588436604,
+      "learning_rate": 0.0018666666666666666,
+      "loss": 0.8591,
+      "step": 594
+    },
+    {
+      "epoch": 1.3218550402665925,
+      "grad_norm": 0.007649209350347519,
+      "learning_rate": 0.0018641975308641974,
+      "loss": 0.8245,
+      "step": 595
+    },
+    {
+      "epoch": 1.3240766453762844,
+      "grad_norm": 0.010610325261950493,
+      "learning_rate": 0.0018617283950617283,
+      "loss": 0.8198,
+      "step": 596
+    },
+    {
+      "epoch": 1.3262982504859762,
+      "grad_norm": 0.013199282810091972,
+      "learning_rate": 0.0018592592592592595,
+      "loss": 0.7064,
+      "step": 597
+    },
+    {
+      "epoch": 1.3285198555956679,
+      "grad_norm": 0.006374049000442028,
+      "learning_rate": 0.00185679012345679,
+      "loss": 0.7449,
+      "step": 598
+    },
+    {
+      "epoch": 1.3307414607053596,
+      "grad_norm": 0.012544116005301476,
+      "learning_rate": 0.0018543209876543212,
+      "loss": 0.5865,
+      "step": 599
+    },
+    {
+      "epoch": 1.3329630658150513,
+      "grad_norm": 0.007943814620375633,
+      "learning_rate": 0.0018518518518518517,
+      "loss": 0.8927,
+      "step": 600
+    },
+    {
+      "epoch": 1.3351846709247432,
+      "grad_norm": 0.007226669695228338,
+      "learning_rate": 0.0018493827160493829,
+      "loss": 0.9163,
+      "step": 601
+    },
+    {
+      "epoch": 1.337406276034435,
+      "grad_norm": 0.006219184026122093,
+      "learning_rate": 0.0018469135802469134,
+      "loss": 0.8312,
+      "step": 602
+    },
+    {
+      "epoch": 1.3396278811441267,
+      "grad_norm": 0.009386156685650349,
+      "learning_rate": 0.0018444444444444446,
+      "loss": 0.62,
+      "step": 603
+    },
+    {
+      "epoch": 1.3418494862538184,
+      "grad_norm": 0.009648386389017105,
+      "learning_rate": 0.0018419753086419755,
+      "loss": 0.8502,
+      "step": 604
+    },
+    {
+      "epoch": 1.34407109136351,
+      "grad_norm": 0.007513147313147783,
+      "learning_rate": 0.0018395061728395062,
+      "loss": 0.7968,
+      "step": 605
+    },
+    {
+      "epoch": 1.3462926964732018,
+      "grad_norm": 0.007452718913555145,
+      "learning_rate": 0.0018370370370370372,
+      "loss": 0.8427,
+      "step": 606
+    },
+    {
+      "epoch": 1.3485143015828935,
+      "grad_norm": 0.0075603355653584,
+      "learning_rate": 0.001834567901234568,
+      "loss": 0.7846,
+      "step": 607
+    },
+    {
+      "epoch": 1.3507359066925855,
+      "grad_norm": 0.008041512221097946,
+      "learning_rate": 0.0018320987654320989,
+      "loss": 1.0172,
+      "step": 608
+    },
+    {
+      "epoch": 1.3529575118022772,
+      "grad_norm": 0.009016762487590313,
+      "learning_rate": 0.0018296296296296296,
+      "loss": 0.7735,
+      "step": 609
+    },
+    {
+      "epoch": 1.355179116911969,
+      "grad_norm": 0.005660157650709152,
+      "learning_rate": 0.0018271604938271606,
+      "loss": 0.704,
+      "step": 610
+    },
+    {
+      "epoch": 1.3574007220216606,
+      "grad_norm": 0.02231772243976593,
+      "learning_rate": 0.0018246913580246915,
+      "loss": 0.9194,
+      "step": 611
+    },
+    {
+      "epoch": 1.3596223271313523,
+      "grad_norm": 0.009612826630473137,
+      "learning_rate": 0.0018222222222222223,
+      "loss": 0.805,
+      "step": 612
+    },
+    {
+      "epoch": 1.3618439322410443,
+      "grad_norm": 0.008082101121544838,
+      "learning_rate": 0.0018197530864197532,
+      "loss": 0.8137,
+      "step": 613
+    },
+    {
+      "epoch": 1.364065537350736,
+      "grad_norm": 0.0074988240376114845,
+      "learning_rate": 0.001817283950617284,
+      "loss": 0.7579,
+      "step": 614
+    },
+    {
+      "epoch": 1.3662871424604277,
+      "grad_norm": 0.008431846275925636,
+      "learning_rate": 0.001814814814814815,
+      "loss": 0.6704,
+      "step": 615
+    },
+    {
+      "epoch": 1.3685087475701194,
+      "grad_norm": 0.00818427000194788,
+      "learning_rate": 0.0018123456790123456,
+      "loss": 0.7728,
+      "step": 616
+    },
+    {
+      "epoch": 1.3707303526798111,
+      "grad_norm": 0.009126820601522923,
+      "learning_rate": 0.0018098765432098766,
+      "loss": 0.5695,
+      "step": 617
+    },
+    {
+      "epoch": 1.3729519577895029,
+      "grad_norm": 0.010422170162200928,
+      "learning_rate": 0.0018074074074074075,
+      "loss": 0.7908,
+      "step": 618
+    },
+    {
+      "epoch": 1.3751735628991946,
+      "grad_norm": 0.00857476331293583,
+      "learning_rate": 0.0018049382716049383,
+      "loss": 0.7064,
+      "step": 619
+    },
+    {
+      "epoch": 1.3773951680088863,
+      "grad_norm": 0.014797719195485115,
+      "learning_rate": 0.0018024691358024692,
+      "loss": 0.9911,
+      "step": 620
+    },
+    {
+      "epoch": 1.3796167731185782,
+      "grad_norm": 0.009074145928025246,
+      "learning_rate": 0.0018,
+      "loss": 0.9399,
+      "step": 621
+    },
+    {
+      "epoch": 1.38183837822827,
+      "grad_norm": 0.00829361006617546,
+      "learning_rate": 0.001797530864197531,
+      "loss": 0.7555,
+      "step": 622
+    },
+    {
+      "epoch": 1.3840599833379617,
+      "grad_norm": 0.008414024487137794,
+      "learning_rate": 0.0017950617283950616,
+      "loss": 0.7842,
+      "step": 623
+    },
+    {
+      "epoch": 1.3862815884476534,
+      "grad_norm": 0.008322815410792828,
+      "learning_rate": 0.0017925925925925926,
+      "loss": 0.7748,
+      "step": 624
+    },
+    {
+      "epoch": 1.3885031935573453,
+      "grad_norm": 0.010673588141798973,
+      "learning_rate": 0.0017901234567901233,
+      "loss": 0.7959,
+      "step": 625
+    },
+    {
+      "epoch": 1.390724798667037,
+      "grad_norm": 0.016830846667289734,
+      "learning_rate": 0.0017876543209876543,
+      "loss": 0.6871,
+      "step": 626
+    },
+    {
+      "epoch": 1.3929464037767287,
+      "grad_norm": 0.031158100813627243,
+      "learning_rate": 0.0017851851851851854,
+      "loss": 1.0221,
+      "step": 627
+    },
+    {
+      "epoch": 1.3951680088864205,
+      "grad_norm": 0.0064673833549022675,
+      "learning_rate": 0.001782716049382716,
+      "loss": 0.8553,
+      "step": 628
+    },
+    {
+      "epoch": 1.3973896139961122,
+      "grad_norm": 0.008676216006278992,
+      "learning_rate": 0.0017802469135802471,
+      "loss": 0.7436,
+      "step": 629
+    },
+    {
+      "epoch": 1.3996112191058039,
+      "grad_norm": 0.006582185626029968,
+      "learning_rate": 0.0017777777777777776,
+      "loss": 0.7052,
+      "step": 630
+    },
+    {
+      "epoch": 1.4018328242154956,
+      "grad_norm": 0.009261371567845345,
+      "learning_rate": 0.0017753086419753088,
+      "loss": 0.8944,
+      "step": 631
+    },
+    {
+      "epoch": 1.4040544293251873,
+      "grad_norm": 0.005905612837523222,
+      "learning_rate": 0.0017728395061728393,
+      "loss": 0.8222,
+      "step": 632
+    },
+    {
+      "epoch": 1.4062760344348793,
+      "grad_norm": 0.0069193001836538315,
+      "learning_rate": 0.0017703703703703705,
+      "loss": 0.7179,
+      "step": 633
+    },
+    {
+      "epoch": 1.408497639544571,
+      "grad_norm": 0.01019218284636736,
+      "learning_rate": 0.0017679012345679015,
+      "loss": 0.8184,
+      "step": 634
+    },
+    {
+      "epoch": 1.4107192446542627,
+      "grad_norm": 0.008972816169261932,
+      "learning_rate": 0.0017654320987654322,
+      "loss": 0.7132,
+      "step": 635
+    },
+    {
+      "epoch": 1.4129408497639544,
+      "grad_norm": 0.011042306199669838,
+      "learning_rate": 0.0017629629629629631,
+      "loss": 0.7855,
+      "step": 636
+    },
+    {
+      "epoch": 1.4151624548736463,
+      "grad_norm": 0.010043452493846416,
+      "learning_rate": 0.0017604938271604939,
+      "loss": 0.9929,
+      "step": 637
+    },
+    {
+      "epoch": 1.417384059983338,
+      "grad_norm": 0.006690305657684803,
+      "learning_rate": 0.0017580246913580248,
+      "loss": 0.8618,
+      "step": 638
+    },
+    {
+      "epoch": 1.4196056650930298,
+      "grad_norm": 0.0102543905377388,
+      "learning_rate": 0.0017555555555555556,
+      "loss": 0.7471,
+      "step": 639
+    },
+    {
+      "epoch": 1.4218272702027215,
+      "grad_norm": 0.011543518863618374,
+      "learning_rate": 0.0017530864197530865,
+      "loss": 0.8827,
+      "step": 640
+    },
+    {
+      "epoch": 1.4240488753124132,
+      "grad_norm": 0.0073388186283409595,
+      "learning_rate": 0.0017506172839506175,
+      "loss": 0.8361,
+      "step": 641
+    },
+    {
+      "epoch": 1.426270480422105,
+      "grad_norm": 0.009625883772969246,
+      "learning_rate": 0.0017481481481481482,
+      "loss": 0.756,
+      "step": 642
+    },
+    {
+      "epoch": 1.4284920855317966,
+      "grad_norm": 0.008923063054680824,
+      "learning_rate": 0.0017456790123456791,
+      "loss": 0.7119,
+      "step": 643
+    },
+    {
+      "epoch": 1.4307136906414883,
+      "grad_norm": 0.008398482576012611,
+      "learning_rate": 0.0017432098765432099,
+      "loss": 0.5496,
+      "step": 644
+    },
+    {
+      "epoch": 1.4329352957511803,
+      "grad_norm": 0.013783407397568226,
+      "learning_rate": 0.0017407407407407408,
+      "loss": 0.7134,
+      "step": 645
+    },
+    {
+      "epoch": 1.435156900860872,
+      "grad_norm": 0.008702287450432777,
+      "learning_rate": 0.0017382716049382716,
+      "loss": 0.8801,
+      "step": 646
+    },
+    {
+      "epoch": 1.4373785059705637,
+      "grad_norm": 0.007398973684757948,
+      "learning_rate": 0.0017358024691358025,
+      "loss": 0.7561,
+      "step": 647
+    },
+    {
+      "epoch": 1.4396001110802554,
+      "grad_norm": 0.009800056926906109,
+      "learning_rate": 0.0017333333333333333,
+      "loss": 0.7184,
+      "step": 648
+    },
+    {
+      "epoch": 1.4418217161899474,
+      "grad_norm": 0.006798363756388426,
+      "learning_rate": 0.0017308641975308642,
+      "loss": 0.8016,
+      "step": 649
+    },
+    {
+      "epoch": 1.444043321299639,
+      "grad_norm": 0.005563552491366863,
+      "learning_rate": 0.0017283950617283952,
+      "loss": 0.727,
+      "step": 650
+    },
+    {
+      "epoch": 1.4462649264093308,
+      "grad_norm": 0.032814882695674896,
+      "learning_rate": 0.0017259259259259259,
+      "loss": 0.733,
+      "step": 651
+    },
+    {
+      "epoch": 1.4484865315190225,
+      "grad_norm": 0.007777430582791567,
+      "learning_rate": 0.0017234567901234568,
+      "loss": 0.8456,
+      "step": 652
+    },
+    {
+      "epoch": 1.4507081366287142,
+      "grad_norm": 0.01004884298890829,
+      "learning_rate": 0.0017209876543209876,
+      "loss": 0.8893,
+      "step": 653
+    },
+    {
+      "epoch": 1.452929741738406,
+      "grad_norm": 0.005339653231203556,
+      "learning_rate": 0.0017185185185185185,
+      "loss": 0.6644,
+      "step": 654
+    },
+    {
+      "epoch": 1.4551513468480977,
+      "grad_norm": 0.009580488316714764,
+      "learning_rate": 0.0017160493827160493,
+      "loss": 0.78,
+      "step": 655
+    },
+    {
+      "epoch": 1.4573729519577894,
+      "grad_norm": 0.008997698314487934,
+      "learning_rate": 0.0017135802469135802,
+      "loss": 0.7838,
+      "step": 656
+    },
+    {
+      "epoch": 1.4595945570674813,
+      "grad_norm": 0.008436165750026703,
+      "learning_rate": 0.0017111111111111114,
+      "loss": 0.7232,
+      "step": 657
+    },
+    {
+      "epoch": 1.461816162177173,
+      "grad_norm": 0.009556714445352554,
+      "learning_rate": 0.001708641975308642,
+      "loss": 0.815,
+      "step": 658
+    },
+    {
+      "epoch": 1.4640377672868647,
+      "grad_norm": 0.01159976702183485,
+      "learning_rate": 0.001706172839506173,
+      "loss": 0.8061,
+      "step": 659
+    },
+    {
+      "epoch": 1.4662593723965565,
+      "grad_norm": 0.0057432996109128,
+      "learning_rate": 0.0017037037037037036,
+      "loss": 0.8892,
+      "step": 660
+    },
+    {
+      "epoch": 1.4684809775062484,
+      "grad_norm": 0.007714425213634968,
+      "learning_rate": 0.0017012345679012348,
+      "loss": 0.8771,
+      "step": 661
+    },
+    {
+      "epoch": 1.4707025826159401,
+      "grad_norm": 0.00997792650014162,
+      "learning_rate": 0.0016987654320987653,
+      "loss": 0.7399,
+      "step": 662
+    },
+    {
+      "epoch": 1.4729241877256318,
+      "grad_norm": 0.008112712763249874,
+      "learning_rate": 0.0016962962962962964,
+      "loss": 0.7373,
+      "step": 663
+    },
+    {
+      "epoch": 1.4751457928353235,
+      "grad_norm": 0.009659236297011375,
+      "learning_rate": 0.0016938271604938274,
+      "loss": 0.823,
+      "step": 664
+    },
+    {
+      "epoch": 1.4773673979450153,
+      "grad_norm": 0.021559547632932663,
+      "learning_rate": 0.0016913580246913581,
+      "loss": 0.8344,
+      "step": 665
+    },
+    {
+      "epoch": 1.479589003054707,
+      "grad_norm": 0.007330529857426882,
+      "learning_rate": 0.001688888888888889,
+      "loss": 0.7308,
+      "step": 666
+    },
+    {
+      "epoch": 1.4818106081643987,
+      "grad_norm": 0.009981613606214523,
+      "learning_rate": 0.0016864197530864198,
+      "loss": 0.7725,
+      "step": 667
+    },
+    {
+      "epoch": 1.4840322132740904,
+      "grad_norm": 0.011183936148881912,
+      "learning_rate": 0.0016839506172839508,
+      "loss": 0.785,
+      "step": 668
+    },
+    {
+      "epoch": 1.4862538183837823,
+      "grad_norm": 0.006825693417340517,
+      "learning_rate": 0.0016814814814814813,
+      "loss": 0.8749,
+      "step": 669
+    },
+    {
+      "epoch": 1.488475423493474,
+      "grad_norm": 0.007616880349814892,
+      "learning_rate": 0.0016790123456790125,
+      "loss": 0.7206,
+      "step": 670
+    },
+    {
+      "epoch": 1.4906970286031658,
+      "grad_norm": 0.008109161630272865,
+      "learning_rate": 0.0016765432098765434,
+      "loss": 0.8849,
+      "step": 671
+    },
+    {
+      "epoch": 1.4929186337128575,
+      "grad_norm": 0.008072765544056892,
+      "learning_rate": 0.0016740740740740741,
+      "loss": 0.6808,
+      "step": 672
+    },
+    {
+      "epoch": 1.4951402388225494,
+      "grad_norm": 0.008430154994130135,
+      "learning_rate": 0.001671604938271605,
+      "loss": 0.8143,
+      "step": 673
+    },
+    {
+      "epoch": 1.4973618439322411,
+      "grad_norm": 0.006939015816897154,
+      "learning_rate": 0.0016691358024691358,
+      "loss": 0.7448,
+      "step": 674
+    },
+    {
+      "epoch": 1.4995834490419329,
+      "grad_norm": 0.00827766116708517,
+      "learning_rate": 0.0016666666666666668,
+      "loss": 0.8428,
+      "step": 675
+    },
+    {
+      "epoch": 1.5018050541516246,
+      "grad_norm": 0.009324070066213608,
+      "learning_rate": 0.0016641975308641975,
+      "loss": 0.8593,
+      "step": 676
+    },
+    {
+      "epoch": 1.5040266592613163,
+      "grad_norm": 0.00814896821975708,
+      "learning_rate": 0.0016617283950617285,
+      "loss": 0.7604,
+      "step": 677
+    },
+    {
+      "epoch": 1.506248264371008,
+      "grad_norm": 0.006904492620378733,
+      "learning_rate": 0.0016592592592592592,
+      "loss": 0.7365,
+      "step": 678
+    },
+    {
+      "epoch": 1.5084698694806997,
+      "grad_norm": 0.011149480938911438,
+      "learning_rate": 0.0016567901234567901,
+      "loss": 0.6607,
+      "step": 679
+    },
+    {
+      "epoch": 1.5106914745903914,
+      "grad_norm": 0.007484333124011755,
+      "learning_rate": 0.001654320987654321,
+      "loss": 0.6669,
+      "step": 680
+    },
+    {
+      "epoch": 1.5129130797000832,
+      "grad_norm": 0.015959644690155983,
+      "learning_rate": 0.0016518518518518518,
+      "loss": 0.8162,
+      "step": 681
+    },
+    {
+      "epoch": 1.515134684809775,
+      "grad_norm": 0.011469387449324131,
+      "learning_rate": 0.0016493827160493828,
+      "loss": 0.8015,
+      "step": 682
+    },
+    {
+      "epoch": 1.5173562899194668,
+      "grad_norm": 0.009255544282495975,
+      "learning_rate": 0.0016469135802469135,
+      "loss": 0.8708,
+      "step": 683
+    },
+    {
+      "epoch": 1.5195778950291585,
+      "grad_norm": 0.02047853358089924,
+      "learning_rate": 0.0016444444444444445,
+      "loss": 0.6925,
+      "step": 684
+    },
+    {
+      "epoch": 1.5217995001388505,
+      "grad_norm": 0.006029823329299688,
+      "learning_rate": 0.0016419753086419752,
+      "loss": 0.6099,
+      "step": 685
+    },
+    {
+      "epoch": 1.5240211052485422,
+      "grad_norm": 0.00622076541185379,
+      "learning_rate": 0.0016395061728395062,
+      "loss": 0.9379,
+      "step": 686
+    },
+    {
+      "epoch": 1.526242710358234,
+      "grad_norm": 0.007114652544260025,
+      "learning_rate": 0.001637037037037037,
+      "loss": 0.8084,
+      "step": 687
+    },
+    {
+      "epoch": 1.5284643154679256,
+      "grad_norm": 0.006938717793673277,
+      "learning_rate": 0.0016345679012345678,
+      "loss": 0.9378,
+      "step": 688
+    },
+    {
+      "epoch": 1.5306859205776173,
+      "grad_norm": 0.006727306172251701,
+      "learning_rate": 0.0016320987654320988,
+      "loss": 0.7257,
+      "step": 689
+    },
+    {
+      "epoch": 1.532907525687309,
+      "grad_norm": 0.006786365061998367,
+      "learning_rate": 0.0016296296296296295,
+      "loss": 0.8699,
+      "step": 690
+    },
+    {
+      "epoch": 1.5351291307970008,
+      "grad_norm": 0.007230743300169706,
+      "learning_rate": 0.0016271604938271605,
+      "loss": 0.7269,
+      "step": 691
+    },
+    {
+      "epoch": 1.5373507359066925,
+      "grad_norm": 0.0071078576147556305,
+      "learning_rate": 0.0016246913580246912,
+      "loss": 0.7598,
+      "step": 692
+    },
+    {
+      "epoch": 1.5395723410163842,
+      "grad_norm": 0.00766537431627512,
+      "learning_rate": 0.0016222222222222222,
+      "loss": 0.7221,
+      "step": 693
+    },
+    {
+      "epoch": 1.5417939461260761,
+      "grad_norm": 0.011207444593310356,
+      "learning_rate": 0.0016197530864197533,
+      "loss": 0.8521,
+      "step": 694
+    },
+    {
+      "epoch": 1.5440155512357678,
+      "grad_norm": 0.009046681225299835,
+      "learning_rate": 0.0016172839506172839,
+      "loss": 0.7945,
+      "step": 695
+    },
+    {
+      "epoch": 1.5462371563454596,
+      "grad_norm": 0.025657106190919876,
+      "learning_rate": 0.001614814814814815,
+      "loss": 0.7099,
+      "step": 696
+    },
+    {
+      "epoch": 1.5484587614551515,
+      "grad_norm": 0.00540280481800437,
+      "learning_rate": 0.0016123456790123455,
+      "loss": 0.8597,
+      "step": 697
+    },
+    {
+      "epoch": 1.5506803665648432,
+      "grad_norm": 0.0077238529920578,
+      "learning_rate": 0.0016098765432098767,
+      "loss": 0.7868,
+      "step": 698
+    },
+    {
+      "epoch": 1.552901971674535,
+      "grad_norm": 0.011015200056135654,
+      "learning_rate": 0.0016074074074074072,
+      "loss": 0.4777,
+      "step": 699
+    },
+    {
+      "epoch": 1.5551235767842266,
+      "grad_norm": 0.008766315877437592,
+      "learning_rate": 0.0016049382716049384,
+      "loss": 0.6334,
+      "step": 700
+    },
+    {
+      "epoch": 1.5573451818939184,
+      "grad_norm": 0.005418060813099146,
+      "learning_rate": 0.0016024691358024693,
+      "loss": 0.7606,
+      "step": 701
+    },
+    {
+      "epoch": 1.55956678700361,
+      "grad_norm": 0.010448934510350227,
+      "learning_rate": 0.0016,
+      "loss": 0.6833,
+      "step": 702
+    },
+    {
+      "epoch": 1.5617883921133018,
+      "grad_norm": 0.007874859496951103,
+      "learning_rate": 0.001597530864197531,
+      "loss": 0.6212,
+      "step": 703
+    },
+    {
+      "epoch": 1.5640099972229935,
+      "grad_norm": 0.01020901370793581,
+      "learning_rate": 0.0015950617283950618,
+      "loss": 0.9533,
+      "step": 704
+    },
+    {
+      "epoch": 1.5662316023326852,
+      "grad_norm": 0.006319139152765274,
+      "learning_rate": 0.0015925925925925927,
+      "loss": 0.8269,
+      "step": 705
+    },
+    {
+      "epoch": 1.5684532074423772,
+      "grad_norm": 0.017285294830799103,
+      "learning_rate": 0.0015901234567901234,
+      "loss": 1.0417,
+      "step": 706
+    },
+    {
+      "epoch": 1.5706748125520689,
+      "grad_norm": 0.009961610659956932,
+      "learning_rate": 0.0015876543209876544,
+      "loss": 0.8643,
+      "step": 707
+    },
+    {
+      "epoch": 1.5728964176617606,
+      "grad_norm": 0.008692831732332706,
+      "learning_rate": 0.0015851851851851851,
+      "loss": 0.7483,
+      "step": 708
+    },
+    {
+      "epoch": 1.5751180227714525,
+      "grad_norm": 0.010625587776303291,
+      "learning_rate": 0.001582716049382716,
+      "loss": 0.7756,
+      "step": 709
+    },
+    {
+      "epoch": 1.5773396278811442,
+      "grad_norm": 0.0054595437832176685,
+      "learning_rate": 0.001580246913580247,
+      "loss": 0.7034,
+      "step": 710
+    },
+    {
+      "epoch": 1.579561232990836,
+      "grad_norm": 0.0060311974957585335,
+      "learning_rate": 0.0015777777777777778,
+      "loss": 0.8508,
+      "step": 711
+    },
+    {
+      "epoch": 1.5817828381005277,
+      "grad_norm": 0.011815550737082958,
+      "learning_rate": 0.0015753086419753087,
+      "loss": 0.8377,
+      "step": 712
+    },
+    {
+      "epoch": 1.5840044432102194,
+      "grad_norm": 0.006426041480153799,
+      "learning_rate": 0.0015728395061728395,
+      "loss": 0.7625,
+      "step": 713
+    },
+    {
+      "epoch": 1.586226048319911,
+      "grad_norm": 0.005769714713096619,
+      "learning_rate": 0.0015703703703703704,
+      "loss": 0.8298,
+      "step": 714
+    },
+    {
+      "epoch": 1.5884476534296028,
+      "grad_norm": 0.008317188359797001,
+      "learning_rate": 0.0015679012345679011,
+      "loss": 0.7285,
+      "step": 715
+    },
+    {
+      "epoch": 1.5906692585392945,
+      "grad_norm": 0.005216104909777641,
+      "learning_rate": 0.001565432098765432,
+      "loss": 0.7872,
+      "step": 716
+    },
+    {
+      "epoch": 1.5928908636489862,
+      "grad_norm": 0.006061375141143799,
+      "learning_rate": 0.001562962962962963,
+      "loss": 0.7549,
+      "step": 717
+    },
+    {
+      "epoch": 1.5951124687586782,
+      "grad_norm": 0.008552168495953083,
+      "learning_rate": 0.0015604938271604938,
+      "loss": 0.9864,
+      "step": 718
+    },
+    {
+      "epoch": 1.59733407386837,
+      "grad_norm": 0.0080868573859334,
+      "learning_rate": 0.0015580246913580247,
+      "loss": 0.7169,
+      "step": 719
+    },
+    {
+      "epoch": 1.5995556789780616,
+      "grad_norm": 0.009083057753741741,
+      "learning_rate": 0.0015555555555555555,
+      "loss": 0.9025,
+      "step": 720
+    },
+    {
+      "epoch": 1.6017772840877536,
+      "grad_norm": 0.015129972249269485,
+      "learning_rate": 0.0015530864197530864,
+      "loss": 0.7955,
+      "step": 721
+    },
+    {
+      "epoch": 1.6039988891974453,
+      "grad_norm": 0.009628325700759888,
+      "learning_rate": 0.0015506172839506172,
+      "loss": 0.801,
+      "step": 722
+    },
+    {
+      "epoch": 1.606220494307137,
+      "grad_norm": 0.008118744008243084,
+      "learning_rate": 0.001548148148148148,
+      "loss": 0.801,
+      "step": 723
+    },
+    {
+      "epoch": 1.6084420994168287,
+      "grad_norm": 0.01229710690677166,
+      "learning_rate": 0.0015456790123456793,
+      "loss": 0.7778,
+      "step": 724
+    },
+    {
+      "epoch": 1.6106637045265204,
+      "grad_norm": 0.00783099327236414,
+      "learning_rate": 0.0015432098765432098,
+      "loss": 0.7599,
+      "step": 725
+    },
+    {
+      "epoch": 1.6128853096362121,
+      "grad_norm": 0.005414724349975586,
+      "learning_rate": 0.001540740740740741,
+      "loss": 0.6664,
+      "step": 726
+    },
+    {
+      "epoch": 1.6151069147459038,
+      "grad_norm": 0.006641815882176161,
+      "learning_rate": 0.0015382716049382715,
+      "loss": 0.769,
+      "step": 727
+    },
+    {
+      "epoch": 1.6173285198555956,
+      "grad_norm": 0.007328651379793882,
+      "learning_rate": 0.0015358024691358026,
+      "loss": 0.7119,
+      "step": 728
+    },
+    {
+      "epoch": 1.6195501249652873,
+      "grad_norm": 0.010657723993062973,
+      "learning_rate": 0.0015333333333333332,
+      "loss": 0.7682,
+      "step": 729
+    },
+    {
+      "epoch": 1.6217717300749792,
+      "grad_norm": 0.009739626199007034,
+      "learning_rate": 0.0015308641975308643,
+      "loss": 0.8293,
+      "step": 730
+    },
+    {
+      "epoch": 1.623993335184671,
+      "grad_norm": 0.0063155414536595345,
+      "learning_rate": 0.0015283950617283948,
+      "loss": 0.6139,
+      "step": 731
+    },
+    {
+      "epoch": 1.6262149402943626,
+      "grad_norm": 0.007235299795866013,
+      "learning_rate": 0.001525925925925926,
+      "loss": 0.8225,
+      "step": 732
+    },
+    {
+      "epoch": 1.6284365454040546,
+      "grad_norm": 0.008075661025941372,
+      "learning_rate": 0.001523456790123457,
+      "loss": 0.9188,
+      "step": 733
+    },
+    {
+      "epoch": 1.6306581505137463,
+      "grad_norm": 0.009136526845395565,
+      "learning_rate": 0.0015209876543209877,
+      "loss": 0.8515,
+      "step": 734
+    },
+    {
+      "epoch": 1.632879755623438,
+      "grad_norm": 0.006905075628310442,
+      "learning_rate": 0.0015185185185185187,
+      "loss": 0.6947,
+      "step": 735
+    },
+    {
+      "epoch": 1.6351013607331297,
+      "grad_norm": 0.006465461570769548,
+      "learning_rate": 0.0015160493827160494,
+      "loss": 0.7776,
+      "step": 736
+    },
+    {
+      "epoch": 1.6373229658428214,
+      "grad_norm": 0.007885467261075974,
+      "learning_rate": 0.0015135802469135803,
+      "loss": 0.8716,
+      "step": 737
+    },
+    {
+      "epoch": 1.6395445709525132,
+      "grad_norm": 0.008669420145452023,
+      "learning_rate": 0.001511111111111111,
+      "loss": 0.8524,
+      "step": 738
+    },
+    {
+      "epoch": 1.6417661760622049,
+      "grad_norm": 0.008928552269935608,
+      "learning_rate": 0.001508641975308642,
+      "loss": 0.7355,
+      "step": 739
+    },
+    {
+      "epoch": 1.6439877811718966,
+      "grad_norm": 0.007420363835990429,
+      "learning_rate": 0.001506172839506173,
+      "loss": 0.7055,
+      "step": 740
+    },
+    {
+      "epoch": 1.6462093862815883,
+      "grad_norm": 0.011267797090113163,
+      "learning_rate": 0.0015037037037037037,
+      "loss": 0.8876,
+      "step": 741
+    },
+    {
+      "epoch": 1.6484309913912802,
+      "grad_norm": 0.00850197859108448,
+      "learning_rate": 0.0015012345679012347,
+      "loss": 0.6682,
+      "step": 742
+    },
+    {
+      "epoch": 1.650652596500972,
+      "grad_norm": 0.008527017198503017,
+      "learning_rate": 0.0014987654320987656,
+      "loss": 0.6904,
+      "step": 743
+    },
+    {
+      "epoch": 1.6528742016106637,
+      "grad_norm": 0.005802887491881847,
+      "learning_rate": 0.0014962962962962963,
+      "loss": 0.862,
+      "step": 744
+    },
+    {
+      "epoch": 1.6550958067203556,
+      "grad_norm": 0.007038978394120932,
+      "learning_rate": 0.0014938271604938273,
+      "loss": 0.7665,
+      "step": 745
+    },
+    {
+      "epoch": 1.6573174118300473,
+      "grad_norm": 0.005823012441396713,
+      "learning_rate": 0.001491358024691358,
+      "loss": 0.6075,
+      "step": 746
+    },
+    {
+      "epoch": 1.659539016939739,
+      "grad_norm": 0.008316169492900372,
+      "learning_rate": 0.001488888888888889,
+      "loss": 0.6311,
+      "step": 747
+    },
+    {
+      "epoch": 1.6617606220494308,
+      "grad_norm": 0.006056457757949829,
+      "learning_rate": 0.0014864197530864197,
+      "loss": 0.7038,
+      "step": 748
+    },
+    {
+      "epoch": 1.6639822271591225,
+      "grad_norm": 0.007399104069918394,
+      "learning_rate": 0.0014839506172839507,
+      "loss": 0.6376,
+      "step": 749
+    },
+    {
+      "epoch": 1.6662038322688142,
+      "grad_norm": 0.011369763873517513,
+      "learning_rate": 0.0014814814814814814,
+      "loss": 0.699,
+      "step": 750
+    },
+    {
+      "epoch": 1.668425437378506,
+      "grad_norm": 0.007361331954598427,
+      "learning_rate": 0.0014790123456790124,
+      "loss": 0.7455,
+      "step": 751
+    },
+    {
+      "epoch": 1.6706470424881976,
+      "grad_norm": 0.007952176034450531,
+      "learning_rate": 0.0014765432098765433,
+      "loss": 0.7287,
+      "step": 752
+    },
+    {
+      "epoch": 1.6728686475978893,
+      "grad_norm": 0.029937606304883957,
+      "learning_rate": 0.001474074074074074,
+      "loss": 1.1535,
+      "step": 753
+    },
+    {
+      "epoch": 1.6750902527075813,
+      "grad_norm": 0.005133952479809523,
+      "learning_rate": 0.001471604938271605,
+      "loss": 0.9189,
+      "step": 754
+    },
+    {
+      "epoch": 1.677311857817273,
+      "grad_norm": 0.008348564617335796,
+      "learning_rate": 0.0014691358024691357,
+      "loss": 0.6906,
+      "step": 755
+    },
+    {
+      "epoch": 1.6795334629269647,
+      "grad_norm": 0.007027298677712679,
+      "learning_rate": 0.0014666666666666667,
+      "loss": 0.717,
+      "step": 756
+    },
+    {
+      "epoch": 1.6817550680366566,
+      "grad_norm": 0.008435823023319244,
+      "learning_rate": 0.0014641975308641974,
+      "loss": 0.761,
+      "step": 757
+    },
+    {
+      "epoch": 1.6839766731463484,
+      "grad_norm": 0.012493236921727657,
+      "learning_rate": 0.0014617283950617286,
+      "loss": 0.6764,
+      "step": 758
+    },
+    {
+      "epoch": 1.68619827825604,
+      "grad_norm": 0.0075070820748806,
+      "learning_rate": 0.0014592592592592593,
+      "loss": 0.7438,
+      "step": 759
+    },
+    {
+      "epoch": 1.6884198833657318,
+      "grad_norm": 0.006954481825232506,
+      "learning_rate": 0.0014567901234567903,
+      "loss": 0.8324,
+      "step": 760
+    },
+    {
+      "epoch": 1.6906414884754235,
+      "grad_norm": 0.011813154444098473,
+      "learning_rate": 0.001454320987654321,
+      "loss": 0.7317,
+      "step": 761
+    },
+    {
+      "epoch": 1.6928630935851152,
+      "grad_norm": 0.010434205643832684,
+      "learning_rate": 0.001451851851851852,
+      "loss": 0.7598,
+      "step": 762
+    },
+    {
+      "epoch": 1.695084698694807,
+      "grad_norm": 0.007777587044984102,
+      "learning_rate": 0.0014493827160493827,
+      "loss": 0.7618,
+      "step": 763
+    },
+    {
+      "epoch": 1.6973063038044987,
+      "grad_norm": 0.006949772592633963,
+      "learning_rate": 0.0014469135802469136,
+      "loss": 0.8828,
+      "step": 764
+    },
+    {
+      "epoch": 1.6995279089141904,
+      "grad_norm": 0.009230799973011017,
+      "learning_rate": 0.0014444444444444444,
+      "loss": 0.7395,
+      "step": 765
+    },
+    {
+      "epoch": 1.7017495140238823,
+      "grad_norm": 0.004457883071154356,
+      "learning_rate": 0.0014419753086419753,
+      "loss": 0.8258,
+      "step": 766
+    },
+    {
+      "epoch": 1.703971119133574,
+      "grad_norm": 0.008095591329038143,
+      "learning_rate": 0.0014395061728395063,
+      "loss": 0.6411,
+      "step": 767
+    },
+    {
+      "epoch": 1.7061927242432657,
+      "grad_norm": 0.008313817903399467,
+      "learning_rate": 0.001437037037037037,
+      "loss": 0.6612,
+      "step": 768
+    },
+    {
+      "epoch": 1.7084143293529577,
+      "grad_norm": 0.015146925114095211,
+      "learning_rate": 0.001434567901234568,
+      "loss": 0.8697,
+      "step": 769
+    },
+    {
+      "epoch": 1.7106359344626494,
+      "grad_norm": 0.00795427430421114,
+      "learning_rate": 0.0014320987654320987,
+      "loss": 0.6983,
+      "step": 770
+    },
+    {
+      "epoch": 1.712857539572341,
+      "grad_norm": 0.010255630128085613,
+      "learning_rate": 0.0014296296296296297,
+      "loss": 0.715,
+      "step": 771
+    },
+    {
+      "epoch": 1.7150791446820328,
+      "grad_norm": 0.008728211745619774,
+      "learning_rate": 0.0014271604938271604,
+      "loss": 0.8291,
+      "step": 772
+    },
+    {
+      "epoch": 1.7173007497917245,
+      "grad_norm": 0.007851282134652138,
+      "learning_rate": 0.0014246913580246916,
+      "loss": 0.6991,
+      "step": 773
+    },
+    {
+      "epoch": 1.7195223549014163,
+      "grad_norm": 0.010601981543004513,
+      "learning_rate": 0.0014222222222222223,
+      "loss": 0.8099,
+      "step": 774
+    },
+    {
+      "epoch": 1.721743960011108,
+      "grad_norm": 0.009771551936864853,
+      "learning_rate": 0.0014197530864197532,
+      "loss": 0.7372,
+      "step": 775
+    },
+    {
+      "epoch": 1.7239655651207997,
+      "grad_norm": 0.015610922127962112,
+      "learning_rate": 0.001417283950617284,
+      "loss": 0.6995,
+      "step": 776
+    },
+    {
+      "epoch": 1.7261871702304914,
+      "grad_norm": 0.00798638816922903,
+      "learning_rate": 0.001414814814814815,
+      "loss": 0.7119,
+      "step": 777
+    },
+    {
+      "epoch": 1.7284087753401833,
+      "grad_norm": 0.007400579750537872,
+      "learning_rate": 0.0014123456790123457,
+      "loss": 0.653,
+      "step": 778
+    },
+    {
+      "epoch": 1.730630380449875,
+      "grad_norm": 0.00567126739770174,
+      "learning_rate": 0.0014098765432098766,
+      "loss": 0.6784,
+      "step": 779
+    },
+    {
+      "epoch": 1.7328519855595668,
+      "grad_norm": 0.010270976461470127,
+      "learning_rate": 0.0014074074074074073,
+      "loss": 0.7221,
+      "step": 780
+    },
+    {
+      "epoch": 1.7350735906692587,
+      "grad_norm": 0.006880102213472128,
+      "learning_rate": 0.0014049382716049383,
+      "loss": 0.6944,
+      "step": 781
+    },
+    {
+      "epoch": 1.7372951957789504,
+      "grad_norm": 0.007795974612236023,
+      "learning_rate": 0.0014024691358024693,
+      "loss": 0.8301,
+      "step": 782
+    },
+    {
+      "epoch": 1.7395168008886421,
+      "grad_norm": 0.009786085225641727,
+      "learning_rate": 0.0014,
+      "loss": 0.7403,
+      "step": 783
+    },
+    {
+      "epoch": 1.7417384059983338,
+      "grad_norm": 0.009969739243388176,
+      "learning_rate": 0.001397530864197531,
+      "loss": 0.6,
+      "step": 784
+    },
+    {
+      "epoch": 1.7439600111080256,
+      "grad_norm": 0.01001895871013403,
+      "learning_rate": 0.0013950617283950617,
+      "loss": 0.7644,
+      "step": 785
+    },
+    {
+      "epoch": 1.7461816162177173,
+      "grad_norm": 0.00807132851332426,
+      "learning_rate": 0.0013925925925925926,
+      "loss": 0.7323,
+      "step": 786
+    },
+    {
+      "epoch": 1.748403221327409,
+      "grad_norm": 0.010031183250248432,
+      "learning_rate": 0.0013901234567901234,
+      "loss": 0.7332,
+      "step": 787
+    },
+    {
+      "epoch": 1.7506248264371007,
+      "grad_norm": 0.010263016447424889,
+      "learning_rate": 0.0013876543209876545,
+      "loss": 0.8214,
+      "step": 788
+    },
+    {
+      "epoch": 1.7528464315467924,
+      "grad_norm": 0.008517063222825527,
+      "learning_rate": 0.0013851851851851853,
+      "loss": 0.8323,
+      "step": 789
+    },
+    {
+      "epoch": 1.7550680366564844,
+      "grad_norm": 0.008784698322415352,
+      "learning_rate": 0.0013827160493827162,
+      "loss": 0.8068,
+      "step": 790
+    },
+    {
+      "epoch": 1.757289641766176,
+      "grad_norm": 0.010176415555179119,
+      "learning_rate": 0.001380246913580247,
+      "loss": 0.9765,
+      "step": 791
+    },
+    {
+      "epoch": 1.7595112468758678,
+      "grad_norm": 0.0070914411917328835,
+      "learning_rate": 0.001377777777777778,
+      "loss": 0.8225,
+      "step": 792
+    },
+    {
+      "epoch": 1.7617328519855595,
+      "grad_norm": 0.011972385458648205,
+      "learning_rate": 0.0013753086419753086,
+      "loss": 0.8802,
+      "step": 793
+    },
+    {
+      "epoch": 1.7639544570952514,
+      "grad_norm": 0.007360584102571011,
+      "learning_rate": 0.0013728395061728396,
+      "loss": 0.8901,
+      "step": 794
+    },
+    {
+      "epoch": 1.7661760622049432,
+      "grad_norm": 0.015350119210779667,
+      "learning_rate": 0.0013703703703703703,
+      "loss": 0.9301,
+      "step": 795
+    },
+    {
+      "epoch": 1.7683976673146349,
+      "grad_norm": 0.005848214030265808,
+      "learning_rate": 0.0013679012345679013,
+      "loss": 0.6417,
+      "step": 796
+    },
+    {
+      "epoch": 1.7706192724243266,
+      "grad_norm": 0.013009141199290752,
+      "learning_rate": 0.0013654320987654322,
+      "loss": 0.8102,
+      "step": 797
+    },
+    {
+      "epoch": 1.7728408775340183,
+      "grad_norm": 0.007184984628111124,
+      "learning_rate": 0.001362962962962963,
+      "loss": 0.8189,
+      "step": 798
+    },
+    {
+      "epoch": 1.77506248264371,
+      "grad_norm": 0.008172586560249329,
+      "learning_rate": 0.001360493827160494,
+      "loss": 0.7398,
+      "step": 799
+    },
+    {
+      "epoch": 1.7772840877534017,
+      "grad_norm": 0.0062285615131258965,
+      "learning_rate": 0.0013580246913580246,
+      "loss": 0.92,
+      "step": 800
+    },
+    {
+      "epoch": 1.7795056928630935,
+      "grad_norm": 0.0068848151713609695,
+      "learning_rate": 0.0013555555555555556,
+      "loss": 0.772,
+      "step": 801
+    },
+    {
+      "epoch": 1.7817272979727852,
+      "grad_norm": 0.009805290959775448,
+      "learning_rate": 0.0013530864197530863,
+      "loss": 0.5684,
+      "step": 802
+    },
+    {
+      "epoch": 1.783948903082477,
+      "grad_norm": 0.013168488629162312,
+      "learning_rate": 0.0013506172839506175,
+      "loss": 0.9256,
+      "step": 803
+    },
+    {
+      "epoch": 1.7861705081921688,
+      "grad_norm": 0.008684579282999039,
+      "learning_rate": 0.0013481481481481482,
+      "loss": 0.9528,
+      "step": 804
+    },
+    {
+      "epoch": 1.7883921133018605,
+      "grad_norm": 0.012764825485646725,
+      "learning_rate": 0.0013456790123456792,
+      "loss": 0.6709,
+      "step": 805
+    },
+    {
+      "epoch": 1.7906137184115525,
+      "grad_norm": 0.009115751832723618,
+      "learning_rate": 0.00134320987654321,
+      "loss": 0.9313,
+      "step": 806
+    },
+    {
+      "epoch": 1.7928353235212442,
+      "grad_norm": 0.008431021124124527,
+      "learning_rate": 0.0013407407407407409,
+      "loss": 1.0696,
+      "step": 807
+    },
+    {
+      "epoch": 1.795056928630936,
+      "grad_norm": 0.01002161018550396,
+      "learning_rate": 0.0013382716049382716,
+      "loss": 0.9138,
+      "step": 808
+    },
+    {
+      "epoch": 1.7972785337406276,
+      "grad_norm": 0.007444395218044519,
+      "learning_rate": 0.0013358024691358023,
+      "loss": 0.8264,
+      "step": 809
+    },
+    {
+      "epoch": 1.7995001388503193,
+      "grad_norm": 0.008543235249817371,
+      "learning_rate": 0.0013333333333333333,
+      "loss": 1.0303,
+      "step": 810
+    },
+    {
+      "epoch": 1.801721743960011,
+      "grad_norm": 0.00885559618473053,
+      "learning_rate": 0.0013308641975308642,
+      "loss": 0.7287,
+      "step": 811
+    },
+    {
+      "epoch": 1.8039433490697028,
+      "grad_norm": 0.01157060731202364,
+      "learning_rate": 0.0013283950617283952,
+      "loss": 0.8287,
+      "step": 812
+    },
+    {
+      "epoch": 1.8061649541793945,
+      "grad_norm": 0.011350276879966259,
+      "learning_rate": 0.001325925925925926,
+      "loss": 0.8068,
+      "step": 813
+    },
+    {
+      "epoch": 1.8083865592890862,
+      "grad_norm": 0.00892575178295374,
+      "learning_rate": 0.0013234567901234569,
+      "loss": 0.8147,
+      "step": 814
+    },
+    {
+      "epoch": 1.8106081643987781,
+      "grad_norm": 0.009176945313811302,
+      "learning_rate": 0.0013209876543209876,
+      "loss": 0.673,
+      "step": 815
+    },
+    {
+      "epoch": 1.8128297695084699,
+      "grad_norm": 0.00625951262190938,
+      "learning_rate": 0.0013185185185185186,
+      "loss": 0.7057,
+      "step": 816
+    },
+    {
+      "epoch": 1.8150513746181616,
+      "grad_norm": 0.00856202095746994,
+      "learning_rate": 0.0013160493827160493,
+      "loss": 0.8703,
+      "step": 817
+    },
+    {
+      "epoch": 1.8172729797278535,
+      "grad_norm": 0.006727920845150948,
+      "learning_rate": 0.0013135802469135802,
+      "loss": 0.6712,
+      "step": 818
+    },
+    {
+      "epoch": 1.8194945848375452,
+      "grad_norm": 0.008964741602540016,
+      "learning_rate": 0.0013111111111111112,
+      "loss": 0.7545,
+      "step": 819
+    },
+    {
+      "epoch": 1.821716189947237,
+      "grad_norm": 0.007025114260613918,
+      "learning_rate": 0.001308641975308642,
+      "loss": 0.7261,
+      "step": 820
+    },
+    {
+      "epoch": 1.8239377950569287,
+      "grad_norm": 0.006972780916839838,
+      "learning_rate": 0.0013061728395061729,
+      "loss": 0.676,
+      "step": 821
+    },
+    {
+      "epoch": 1.8261594001666204,
+      "grad_norm": 0.006812758278101683,
+      "learning_rate": 0.0013037037037037036,
+      "loss": 0.7227,
+      "step": 822
+    },
+    {
+      "epoch": 1.828381005276312,
+      "grad_norm": 0.005802335683256388,
+      "learning_rate": 0.0013012345679012346,
+      "loss": 0.7135,
+      "step": 823
+    },
+    {
+      "epoch": 1.8306026103860038,
+      "grad_norm": 0.01064333040267229,
+      "learning_rate": 0.0012987654320987653,
+      "loss": 1.083,
+      "step": 824
+    },
+    {
+      "epoch": 1.8328242154956955,
+      "grad_norm": 0.009813620708882809,
+      "learning_rate": 0.0012962962962962963,
+      "loss": 0.7178,
+      "step": 825
+    },
+    {
+      "epoch": 1.8350458206053872,
+      "grad_norm": 0.005859369412064552,
+      "learning_rate": 0.0012938271604938272,
+      "loss": 0.7861,
+      "step": 826
+    },
+    {
+      "epoch": 1.8372674257150792,
+      "grad_norm": 0.007550476584583521,
+      "learning_rate": 0.0012913580246913582,
+      "loss": 0.9533,
+      "step": 827
+    },
+    {
+      "epoch": 1.8394890308247709,
+      "grad_norm": 0.011494121514260769,
+      "learning_rate": 0.001288888888888889,
+      "loss": 0.6361,
+      "step": 828
+    },
+    {
+      "epoch": 1.8417106359344626,
+      "grad_norm": 0.01333784218877554,
+      "learning_rate": 0.0012864197530864198,
+      "loss": 0.6448,
+      "step": 829
+    },
+    {
+      "epoch": 1.8439322410441545,
+      "grad_norm": 0.0077012451365590096,
+      "learning_rate": 0.0012839506172839506,
+      "loss": 0.6669,
+      "step": 830
+    },
+    {
+      "epoch": 1.8461538461538463,
+      "grad_norm": 0.008285445161163807,
+      "learning_rate": 0.0012814814814814815,
+      "loss": 0.7666,
+      "step": 831
+    },
+    {
+      "epoch": 1.848375451263538,
+      "grad_norm": 0.010961475782096386,
+      "learning_rate": 0.0012790123456790123,
+      "loss": 0.7424,
+      "step": 832
+    },
+    {
+      "epoch": 1.8505970563732297,
+      "grad_norm": 0.009611285291612148,
+      "learning_rate": 0.0012765432098765432,
+      "loss": 0.6858,
+      "step": 833
+    },
+    {
+      "epoch": 1.8528186614829214,
+      "grad_norm": 0.0063394904136657715,
+      "learning_rate": 0.0012740740740740742,
+      "loss": 0.7984,
+      "step": 834
+    },
+    {
+      "epoch": 1.8550402665926131,
+      "grad_norm": 0.007764410227537155,
+      "learning_rate": 0.001271604938271605,
+      "loss": 0.7263,
+      "step": 835
+    },
+    {
+      "epoch": 1.8572618717023048,
+      "grad_norm": 0.008922139182686806,
+      "learning_rate": 0.0012691358024691359,
+      "loss": 0.759,
+      "step": 836
+    },
+    {
+      "epoch": 1.8594834768119965,
+      "grad_norm": 0.008557251654565334,
+      "learning_rate": 0.0012666666666666666,
+      "loss": 0.6563,
+      "step": 837
+    },
+    {
+      "epoch": 1.8617050819216883,
+      "grad_norm": 0.009035222232341766,
+      "learning_rate": 0.0012641975308641975,
+      "loss": 0.684,
+      "step": 838
+    },
+    {
+      "epoch": 1.8639266870313802,
+      "grad_norm": 0.006750497501343489,
+      "learning_rate": 0.0012617283950617283,
+      "loss": 0.7025,
+      "step": 839
+    },
+    {
+      "epoch": 1.866148292141072,
+      "grad_norm": 0.007769269403070211,
+      "learning_rate": 0.0012592592592592592,
+      "loss": 0.7128,
+      "step": 840
+    },
+    {
+      "epoch": 1.8683698972507636,
+      "grad_norm": 0.005589787382632494,
+      "learning_rate": 0.0012567901234567902,
+      "loss": 0.7813,
+      "step": 841
+    },
+    {
+      "epoch": 1.8705915023604556,
+      "grad_norm": 0.00998506136238575,
+      "learning_rate": 0.0012543209876543211,
+      "loss": 0.7923,
+      "step": 842
+    },
+    {
+      "epoch": 1.8728131074701473,
+      "grad_norm": 0.005361242685467005,
+      "learning_rate": 0.0012518518518518519,
+      "loss": 0.5826,
+      "step": 843
+    },
+    {
+      "epoch": 1.875034712579839,
+      "grad_norm": 0.010085574351251125,
+      "learning_rate": 0.0012493827160493828,
+      "loss": 0.7918,
+      "step": 844
+    },
+    {
+      "epoch": 1.8772563176895307,
+      "grad_norm": 0.006444999016821384,
+      "learning_rate": 0.0012469135802469136,
+      "loss": 0.8084,
+      "step": 845
+    },
+    {
+      "epoch": 1.8794779227992224,
+      "grad_norm": 0.008153421804308891,
+      "learning_rate": 0.0012444444444444445,
+      "loss": 0.7297,
+      "step": 846
+    },
+    {
+      "epoch": 1.8816995279089141,
+      "grad_norm": 0.008666029199957848,
+      "learning_rate": 0.0012419753086419752,
+      "loss": 0.9508,
+      "step": 847
+    },
+    {
+      "epoch": 1.8839211330186059,
+      "grad_norm": 0.006563934497535229,
+      "learning_rate": 0.0012395061728395062,
+      "loss": 0.7055,
+      "step": 848
+    },
+    {
+      "epoch": 1.8861427381282976,
+      "grad_norm": 0.008314713835716248,
+      "learning_rate": 0.0012370370370370371,
+      "loss": 0.7659,
+      "step": 849
+    },
+    {
+      "epoch": 1.8883643432379893,
+      "grad_norm": 0.0050984215922653675,
+      "learning_rate": 0.0012345679012345679,
+      "loss": 0.6926,
+      "step": 850
+    },
+    {
+      "epoch": 1.8905859483476812,
+      "grad_norm": 0.007446047849953175,
+      "learning_rate": 0.0012320987654320988,
+      "loss": 0.9914,
+      "step": 851
+    },
+    {
+      "epoch": 1.892807553457373,
+      "grad_norm": 0.007681590039283037,
+      "learning_rate": 0.0012296296296296296,
+      "loss": 0.5726,
+      "step": 852
+    },
+    {
+      "epoch": 1.8950291585670647,
+      "grad_norm": 0.010494091548025608,
+      "learning_rate": 0.0012271604938271605,
+      "loss": 0.8289,
+      "step": 853
+    },
+    {
+      "epoch": 1.8972507636767566,
+      "grad_norm": 0.01278096903115511,
+      "learning_rate": 0.0012246913580246912,
+      "loss": 0.9755,
+      "step": 854
+    },
+    {
+      "epoch": 1.8994723687864483,
+      "grad_norm": 0.007008662447333336,
+      "learning_rate": 0.0012222222222222222,
+      "loss": 0.741,
+      "step": 855
+    },
+    {
+      "epoch": 1.90169397389614,
+      "grad_norm": 0.008134209550917149,
+      "learning_rate": 0.0012197530864197532,
+      "loss": 0.7209,
+      "step": 856
+    },
+    {
+      "epoch": 1.9039155790058317,
+      "grad_norm": 0.01279931515455246,
+      "learning_rate": 0.001217283950617284,
+      "loss": 0.9444,
+      "step": 857
+    },
+    {
+      "epoch": 1.9061371841155235,
+      "grad_norm": 0.006521604955196381,
+      "learning_rate": 0.0012148148148148148,
+      "loss": 0.7992,
+      "step": 858
+    },
+    {
+      "epoch": 1.9083587892252152,
+      "grad_norm": 0.00859599094837904,
+      "learning_rate": 0.0012123456790123458,
+      "loss": 0.6704,
+      "step": 859
+    },
+    {
+      "epoch": 1.910580394334907,
+      "grad_norm": 0.013226643204689026,
+      "learning_rate": 0.0012098765432098765,
+      "loss": 0.7661,
+      "step": 860
+    },
+    {
+      "epoch": 1.9128019994445986,
+      "grad_norm": 0.00894913263618946,
+      "learning_rate": 0.0012074074074074075,
+      "loss": 0.744,
+      "step": 861
+    },
+    {
+      "epoch": 1.9150236045542903,
+      "grad_norm": 0.009207514114677906,
+      "learning_rate": 0.0012049382716049382,
+      "loss": 0.5921,
+      "step": 862
+    },
+    {
+      "epoch": 1.9172452096639823,
+      "grad_norm": 0.010871309787034988,
+      "learning_rate": 0.0012024691358024692,
+      "loss": 0.7441,
+      "step": 863
+    },
+    {
+      "epoch": 1.919466814773674,
+      "grad_norm": 0.007488877512514591,
+      "learning_rate": 0.0012000000000000001,
+      "loss": 0.7526,
+      "step": 864
+    },
+    {
+      "epoch": 1.9216884198833657,
+      "grad_norm": 0.0074579548090696335,
+      "learning_rate": 0.0011975308641975308,
+      "loss": 0.7413,
+      "step": 865
+    },
+    {
+      "epoch": 1.9239100249930576,
+      "grad_norm": 0.00982628669589758,
+      "learning_rate": 0.0011950617283950618,
+      "loss": 0.8372,
+      "step": 866
+    },
+    {
+      "epoch": 1.9261316301027493,
+      "grad_norm": 0.0074563659727573395,
+      "learning_rate": 0.0011925925925925925,
+      "loss": 1.1102,
+      "step": 867
+    },
+    {
+      "epoch": 1.928353235212441,
+      "grad_norm": 0.007274288684129715,
+      "learning_rate": 0.0011901234567901235,
+      "loss": 0.8402,
+      "step": 868
+    },
+    {
+      "epoch": 1.9305748403221328,
+      "grad_norm": 0.008403385989367962,
+      "learning_rate": 0.0011876543209876542,
+      "loss": 1.1321,
+      "step": 869
+    },
+    {
+      "epoch": 1.9327964454318245,
+      "grad_norm": 0.010185015387833118,
+      "learning_rate": 0.0011851851851851852,
+      "loss": 0.6981,
+      "step": 870
+    },
+    {
+      "epoch": 1.9350180505415162,
+      "grad_norm": 0.009552250616252422,
+      "learning_rate": 0.0011827160493827161,
+      "loss": 0.6831,
+      "step": 871
+    },
+    {
+      "epoch": 1.937239655651208,
+      "grad_norm": 0.00690442556515336,
+      "learning_rate": 0.001180246913580247,
+      "loss": 0.7935,
+      "step": 872
+    },
+    {
+      "epoch": 1.9394612607608996,
+      "grad_norm": 0.007579765748232603,
+      "learning_rate": 0.0011777777777777778,
+      "loss": 0.6622,
+      "step": 873
+    },
+    {
+      "epoch": 1.9416828658705914,
+      "grad_norm": 0.005846850108355284,
+      "learning_rate": 0.0011753086419753088,
+      "loss": 0.5541,
+      "step": 874
+    },
+    {
+      "epoch": 1.9439044709802833,
+      "grad_norm": 0.018095355480909348,
+      "learning_rate": 0.0011728395061728395,
+      "loss": 0.9458,
+      "step": 875
+    },
+    {
+      "epoch": 1.946126076089975,
+      "grad_norm": 0.00899551622569561,
+      "learning_rate": 0.0011703703703703704,
+      "loss": 1.0314,
+      "step": 876
+    },
+    {
+      "epoch": 1.9483476811996667,
+      "grad_norm": 0.010987848974764347,
+      "learning_rate": 0.0011679012345679012,
+      "loss": 0.7561,
+      "step": 877
+    },
+    {
+      "epoch": 1.9505692863093587,
+      "grad_norm": 0.01828991435468197,
+      "learning_rate": 0.0011654320987654321,
+      "loss": 0.7234,
+      "step": 878
+    },
+    {
+      "epoch": 1.9527908914190504,
+      "grad_norm": 0.0066715930588543415,
+      "learning_rate": 0.001162962962962963,
+      "loss": 0.8119,
+      "step": 879
+    },
+    {
+      "epoch": 1.955012496528742,
+      "grad_norm": 0.012585703283548355,
+      "learning_rate": 0.0011604938271604938,
+      "loss": 0.8063,
+      "step": 880
+    },
+    {
+      "epoch": 1.9572341016384338,
+      "grad_norm": 0.007742542307823896,
+      "learning_rate": 0.0011580246913580248,
+      "loss": 0.9048,
+      "step": 881
+    },
+    {
+      "epoch": 1.9594557067481255,
+      "grad_norm": 0.00921573955565691,
+      "learning_rate": 0.0011555555555555555,
+      "loss": 0.8855,
+      "step": 882
+    },
+    {
+      "epoch": 1.9616773118578172,
+      "grad_norm": 0.010648581199347973,
+      "learning_rate": 0.0011530864197530865,
+      "loss": 1.0807,
+      "step": 883
+    },
+    {
+      "epoch": 1.963898916967509,
+      "grad_norm": 0.010228232480585575,
+      "learning_rate": 0.0011506172839506172,
+      "loss": 0.8898,
+      "step": 884
+    },
+    {
+      "epoch": 1.9661205220772007,
+      "grad_norm": 0.012444967404007912,
+      "learning_rate": 0.0011481481481481481,
+      "loss": 0.7,
+      "step": 885
+    },
+    {
+      "epoch": 1.9683421271868924,
+      "grad_norm": 0.007236818317323923,
+      "learning_rate": 0.001145679012345679,
+      "loss": 0.7435,
+      "step": 886
+    },
+    {
+      "epoch": 1.9705637322965843,
+      "grad_norm": 0.01137254387140274,
+      "learning_rate": 0.00114320987654321,
+      "loss": 0.7671,
+      "step": 887
+    },
+    {
+      "epoch": 1.972785337406276,
+      "grad_norm": 0.008345142006874084,
+      "learning_rate": 0.0011407407407407408,
+      "loss": 0.7607,
+      "step": 888
+    },
+    {
+      "epoch": 1.9750069425159678,
+      "grad_norm": 0.008709809742867947,
+      "learning_rate": 0.0011382716049382717,
+      "loss": 0.7433,
+      "step": 889
+    },
+    {
+      "epoch": 1.9772285476256597,
+      "grad_norm": 0.007434303406625986,
+      "learning_rate": 0.0011358024691358025,
+      "loss": 0.7211,
+      "step": 890
+    },
+    {
+      "epoch": 1.9794501527353514,
+      "grad_norm": 0.0068624927662312984,
+      "learning_rate": 0.0011333333333333334,
+      "loss": 0.6605,
+      "step": 891
+    },
+    {
+      "epoch": 1.9816717578450431,
+      "grad_norm": 0.007105222903192043,
+      "learning_rate": 0.0011308641975308641,
+      "loss": 0.7847,
+      "step": 892
+    },
+    {
+      "epoch": 1.9838933629547348,
+      "grad_norm": 0.006928863003849983,
+      "learning_rate": 0.001128395061728395,
+      "loss": 0.8906,
+      "step": 893
+    },
+    {
+      "epoch": 1.9861149680644266,
+      "grad_norm": 0.007032196968793869,
+      "learning_rate": 0.001125925925925926,
+      "loss": 0.8525,
+      "step": 894
+    },
+    {
+      "epoch": 1.9883365731741183,
+      "grad_norm": 0.006375422701239586,
+      "learning_rate": 0.0011234567901234568,
+      "loss": 0.9003,
+      "step": 895
+    },
+    {
+      "epoch": 1.99055817828381,
+      "grad_norm": 0.005244637373834848,
+      "learning_rate": 0.0011209876543209877,
+      "loss": 0.7313,
+      "step": 896
+    },
+    {
+      "epoch": 1.9927797833935017,
+      "grad_norm": 0.007450766395777464,
+      "learning_rate": 0.0011185185185185185,
+      "loss": 0.799,
+      "step": 897
+    },
+    {
+      "epoch": 1.9950013885031934,
+      "grad_norm": 0.00844279583543539,
+      "learning_rate": 0.0011160493827160494,
+      "loss": 0.693,
+      "step": 898
+    },
+    {
+      "epoch": 1.9972229936128854,
+      "grad_norm": 0.014142382889986038,
+      "learning_rate": 0.0011135802469135802,
+      "loss": 0.8434,
+      "step": 899
+    },
+    {
+      "epoch": 1.999444598722577,
+      "grad_norm": 0.0095207579433918,
+      "learning_rate": 0.0011111111111111111,
+      "loss": 0.836,
+      "step": 900
+    },
+    {
+      "epoch": 2.001666203832269,
+      "grad_norm": 0.00861104391515255,
+      "learning_rate": 0.001108641975308642,
+      "loss": 0.9819,
+      "step": 901
+    },
+    {
+      "epoch": 2.0038878089419607,
+      "grad_norm": 0.01199290994554758,
+      "learning_rate": 0.001106172839506173,
+      "loss": 0.8446,
+      "step": 902
+    },
+    {
+      "epoch": 2.0061094140516524,
+      "grad_norm": 0.009869737550616264,
+      "learning_rate": 0.0011037037037037037,
+      "loss": 0.7414,
+      "step": 903
+    },
+    {
+      "epoch": 2.008331019161344,
+      "grad_norm": 0.006185215897858143,
+      "learning_rate": 0.0011012345679012347,
+      "loss": 0.8103,
+      "step": 904
+    },
+    {
+      "epoch": 2.010552624271036,
+      "grad_norm": 0.0077135139144957066,
+      "learning_rate": 0.0010987654320987654,
+      "loss": 0.8215,
+      "step": 905
+    },
+    {
+      "epoch": 2.0127742293807276,
+      "grad_norm": 0.007097164634615183,
+      "learning_rate": 0.0010962962962962964,
+      "loss": 0.9623,
+      "step": 906
+    },
+    {
+      "epoch": 2.0149958344904193,
+      "grad_norm": 0.010442017577588558,
+      "learning_rate": 0.0010938271604938271,
+      "loss": 0.6859,
+      "step": 907
+    },
+    {
+      "epoch": 2.017217439600111,
+      "grad_norm": 0.0054067992605268955,
+      "learning_rate": 0.001091358024691358,
+      "loss": 0.7051,
+      "step": 908
+    },
+    {
+      "epoch": 2.0194390447098027,
+      "grad_norm": 0.012054132297635078,
+      "learning_rate": 0.001088888888888889,
+      "loss": 0.7256,
+      "step": 909
+    },
+    {
+      "epoch": 2.0216606498194944,
+      "grad_norm": 0.0082894591614604,
+      "learning_rate": 0.0010864197530864198,
+      "loss": 0.7671,
+      "step": 910
+    },
+    {
+      "epoch": 2.023882254929186,
+      "grad_norm": 0.00784540269523859,
+      "learning_rate": 0.0010839506172839507,
+      "loss": 0.9397,
+      "step": 911
+    },
+    {
+      "epoch": 2.026103860038878,
+      "grad_norm": 0.007958034053444862,
+      "learning_rate": 0.0010814814814814814,
+      "loss": 0.8694,
+      "step": 912
+    },
+    {
+      "epoch": 2.02832546514857,
+      "grad_norm": 0.012851761654019356,
+      "learning_rate": 0.0010790123456790124,
+      "loss": 0.9297,
+      "step": 913
+    },
+    {
+      "epoch": 2.0305470702582618,
+      "grad_norm": 0.009936337359249592,
+      "learning_rate": 0.0010765432098765431,
+      "loss": 0.7811,
+      "step": 914
+    },
+    {
+      "epoch": 2.0327686753679535,
+      "grad_norm": 0.008913993835449219,
+      "learning_rate": 0.001074074074074074,
+      "loss": 0.6929,
+      "step": 915
+    },
+    {
+      "epoch": 2.034990280477645,
+      "grad_norm": 0.008779041469097137,
+      "learning_rate": 0.0010716049382716048,
+      "loss": 0.694,
+      "step": 916
+    },
+    {
+      "epoch": 2.037211885587337,
+      "grad_norm": 0.02905157022178173,
+      "learning_rate": 0.001069135802469136,
+      "loss": 0.9935,
+      "step": 917
+    },
+    {
+      "epoch": 2.0394334906970286,
+      "grad_norm": 0.004718302749097347,
+      "learning_rate": 0.0010666666666666667,
+      "loss": 0.6955,
+      "step": 918
+    },
+    {
+      "epoch": 2.0416550958067203,
+      "grad_norm": 0.008632673881947994,
+      "learning_rate": 0.0010641975308641977,
+      "loss": 0.6824,
+      "step": 919
+    },
+    {
+      "epoch": 2.043876700916412,
+      "grad_norm": 0.009967481717467308,
+      "learning_rate": 0.0010617283950617284,
+      "loss": 0.6765,
+      "step": 920
+    },
+    {
+      "epoch": 2.0460983060261038,
+      "grad_norm": 0.013824276626110077,
+      "learning_rate": 0.0010592592592592594,
+      "loss": 0.8293,
+      "step": 921
+    },
+    {
+      "epoch": 2.0483199111357955,
+      "grad_norm": 0.007447019685059786,
+      "learning_rate": 0.00105679012345679,
+      "loss": 0.6288,
+      "step": 922
+    },
+    {
+      "epoch": 2.050541516245487,
+      "grad_norm": 0.01773306168615818,
+      "learning_rate": 0.001054320987654321,
+      "loss": 0.8114,
+      "step": 923
+    },
+    {
+      "epoch": 2.052763121355179,
+      "grad_norm": 0.005917937494814396,
+      "learning_rate": 0.001051851851851852,
+      "loss": 0.9694,
+      "step": 924
+    },
+    {
+      "epoch": 2.054984726464871,
+      "grad_norm": 0.007322025019675493,
+      "learning_rate": 0.0010493827160493827,
+      "loss": 0.692,
+      "step": 925
+    },
+    {
+      "epoch": 2.057206331574563,
+      "grad_norm": 0.007124684751033783,
+      "learning_rate": 0.0010469135802469137,
+      "loss": 0.6237,
+      "step": 926
+    },
+    {
+      "epoch": 2.0594279366842545,
+      "grad_norm": 0.007253418676555157,
+      "learning_rate": 0.0010444444444444444,
+      "loss": 0.6917,
+      "step": 927
+    },
+    {
+      "epoch": 2.061649541793946,
+      "grad_norm": 0.011087391525506973,
+      "learning_rate": 0.0010419753086419754,
+      "loss": 0.7639,
+      "step": 928
+    },
+    {
+      "epoch": 2.063871146903638,
+      "grad_norm": 0.005655494052916765,
+      "learning_rate": 0.001039506172839506,
+      "loss": 0.564,
+      "step": 929
+    },
+    {
+      "epoch": 2.0660927520133296,
+      "grad_norm": 0.014756609685719013,
+      "learning_rate": 0.001037037037037037,
+      "loss": 0.971,
+      "step": 930
+    },
+    {
+      "epoch": 2.0683143571230214,
+      "grad_norm": 0.010251183062791824,
+      "learning_rate": 0.0010345679012345678,
+      "loss": 0.8309,
+      "step": 931
+    },
+    {
+      "epoch": 2.070535962232713,
+      "grad_norm": 0.007252897135913372,
+      "learning_rate": 0.001032098765432099,
+      "loss": 0.7928,
+      "step": 932
+    },
+    {
+      "epoch": 2.072757567342405,
+      "grad_norm": 0.007240195758640766,
+      "learning_rate": 0.0010296296296296297,
+      "loss": 0.7123,
+      "step": 933
+    },
+    {
+      "epoch": 2.0749791724520965,
+      "grad_norm": 0.0074341814033687115,
+      "learning_rate": 0.0010271604938271606,
+      "loss": 0.6665,
+      "step": 934
+    },
+    {
+      "epoch": 2.0772007775617882,
+      "grad_norm": 0.008963247761130333,
+      "learning_rate": 0.0010246913580246914,
+      "loss": 0.9262,
+      "step": 935
+    },
+    {
+      "epoch": 2.07942238267148,
+      "grad_norm": 0.009950084611773491,
+      "learning_rate": 0.0010222222222222223,
+      "loss": 0.6412,
+      "step": 936
+    },
+    {
+      "epoch": 2.081643987781172,
+      "grad_norm": 0.009436005726456642,
+      "learning_rate": 0.001019753086419753,
+      "loss": 0.748,
+      "step": 937
+    },
+    {
+      "epoch": 2.083865592890864,
+      "grad_norm": 0.008897105231881142,
+      "learning_rate": 0.001017283950617284,
+      "loss": 0.7382,
+      "step": 938
+    },
+    {
+      "epoch": 2.0860871980005555,
+      "grad_norm": 0.008523879572749138,
+      "learning_rate": 0.001014814814814815,
+      "loss": 0.7756,
+      "step": 939
+    },
+    {
+      "epoch": 2.0883088031102472,
+      "grad_norm": 0.00715121952816844,
+      "learning_rate": 0.0010123456790123457,
+      "loss": 0.7371,
+      "step": 940
+    },
+    {
+      "epoch": 2.090530408219939,
+      "grad_norm": 0.007129182107746601,
+      "learning_rate": 0.0010098765432098766,
+      "loss": 0.8859,
+      "step": 941
+    },
+    {
+      "epoch": 2.0927520133296307,
+      "grad_norm": 0.009402717463672161,
+      "learning_rate": 0.0010074074074074074,
+      "loss": 0.6308,
+      "step": 942
+    },
+    {
+      "epoch": 2.0949736184393224,
+      "grad_norm": 0.016564862802624702,
+      "learning_rate": 0.0010049382716049383,
+      "loss": 0.6037,
+      "step": 943
+    },
+    {
+      "epoch": 2.097195223549014,
+      "grad_norm": 0.009927218779921532,
+      "learning_rate": 0.001002469135802469,
+      "loss": 0.9427,
+      "step": 944
+    },
+    {
+      "epoch": 2.099416828658706,
+      "grad_norm": 0.012924822978675365,
+      "learning_rate": 0.001,
+      "loss": 0.5845,
+      "step": 945
+    },
+    {
+      "epoch": 2.1016384337683975,
+      "grad_norm": 0.0068692732602357864,
+      "learning_rate": 0.0009975308641975308,
+      "loss": 0.825,
+      "step": 946
+    },
+    {
+      "epoch": 2.1038600388780893,
+      "grad_norm": 0.006965239066630602,
+      "learning_rate": 0.000995061728395062,
+      "loss": 0.7674,
+      "step": 947
+    },
+    {
+      "epoch": 2.106081643987781,
+      "grad_norm": 0.012765421532094479,
+      "learning_rate": 0.0009925925925925927,
+      "loss": 0.7373,
+      "step": 948
+    },
+    {
+      "epoch": 2.108303249097473,
+      "grad_norm": 0.00697924941778183,
+      "learning_rate": 0.0009901234567901234,
+      "loss": 0.6757,
+      "step": 949
+    },
+    {
+      "epoch": 2.110524854207165,
+      "grad_norm": 0.008962864056229591,
+      "learning_rate": 0.0009876543209876543,
+      "loss": 0.7667,
+      "step": 950
+    },
+    {
+      "epoch": 2.1127464593168566,
+      "grad_norm": 0.007010742090642452,
+      "learning_rate": 0.000985185185185185,
+      "loss": 0.7324,
+      "step": 951
+    },
+    {
+      "epoch": 2.1149680644265483,
+      "grad_norm": 0.010677567683160305,
+      "learning_rate": 0.000982716049382716,
+      "loss": 0.9426,
+      "step": 952
+    },
+    {
+      "epoch": 2.11718966953624,
+      "grad_norm": 0.006151643116027117,
+      "learning_rate": 0.0009802469135802468,
+      "loss": 0.8181,
+      "step": 953
+    },
+    {
+      "epoch": 2.1194112746459317,
+      "grad_norm": 0.007978886365890503,
+      "learning_rate": 0.000977777777777778,
+      "loss": 0.5944,
+      "step": 954
+    },
+    {
+      "epoch": 2.1216328797556234,
+      "grad_norm": 0.005650858860462904,
+      "learning_rate": 0.0009753086419753087,
+      "loss": 0.8202,
+      "step": 955
+    },
+    {
+      "epoch": 2.123854484865315,
+      "grad_norm": 0.007249223534017801,
+      "learning_rate": 0.0009728395061728395,
+      "loss": 0.9062,
+      "step": 956
+    },
+    {
+      "epoch": 2.126076089975007,
+      "grad_norm": 0.008957468904554844,
+      "learning_rate": 0.0009703703703703704,
+      "loss": 0.7101,
+      "step": 957
+    },
+    {
+      "epoch": 2.1282976950846986,
+      "grad_norm": 0.005973477382212877,
+      "learning_rate": 0.0009679012345679012,
+      "loss": 0.8107,
+      "step": 958
+    },
+    {
+      "epoch": 2.1305193001943903,
+      "grad_norm": 0.009565635584294796,
+      "learning_rate": 0.000965432098765432,
+      "loss": 0.8101,
+      "step": 959
+    },
+    {
+      "epoch": 2.132740905304082,
+      "grad_norm": 0.010100849904119968,
+      "learning_rate": 0.0009629629629629629,
+      "loss": 0.6972,
+      "step": 960
+    },
+    {
+      "epoch": 2.134962510413774,
+      "grad_norm": 0.006807432975620031,
+      "learning_rate": 0.0009604938271604937,
+      "loss": 0.5973,
+      "step": 961
+    },
+    {
+      "epoch": 2.137184115523466,
+      "grad_norm": 0.006458532530814409,
+      "learning_rate": 0.0009580246913580248,
+      "loss": 0.7962,
+      "step": 962
+    },
+    {
+      "epoch": 2.1394057206331576,
+      "grad_norm": 0.008561724796891212,
+      "learning_rate": 0.0009555555555555556,
+      "loss": 0.7268,
+      "step": 963
+    },
+    {
+      "epoch": 2.1416273257428493,
+      "grad_norm": 0.006863057147711515,
+      "learning_rate": 0.0009530864197530865,
+      "loss": 0.7796,
+      "step": 964
+    },
+    {
+      "epoch": 2.143848930852541,
+      "grad_norm": 0.007956072688102722,
+      "learning_rate": 0.0009506172839506173,
+      "loss": 0.7257,
+      "step": 965
+    },
+    {
+      "epoch": 2.1460705359622327,
+      "grad_norm": 0.008310288190841675,
+      "learning_rate": 0.0009481481481481482,
+      "loss": 0.9291,
+      "step": 966
+    },
+    {
+      "epoch": 2.1482921410719245,
+      "grad_norm": 0.010176541283726692,
+      "learning_rate": 0.000945679012345679,
+      "loss": 0.9228,
+      "step": 967
+    },
+    {
+      "epoch": 2.150513746181616,
+      "grad_norm": 0.006355747580528259,
+      "learning_rate": 0.0009432098765432098,
+      "loss": 0.7367,
+      "step": 968
+    },
+    {
+      "epoch": 2.152735351291308,
+      "grad_norm": 0.013814382255077362,
+      "learning_rate": 0.0009407407407407408,
+      "loss": 1.0439,
+      "step": 969
+    },
+    {
+      "epoch": 2.1549569564009996,
+      "grad_norm": 0.007776509039103985,
+      "learning_rate": 0.0009382716049382716,
+      "loss": 0.7209,
+      "step": 970
+    },
+    {
+      "epoch": 2.1571785615106913,
+      "grad_norm": 0.007975906133651733,
+      "learning_rate": 0.0009358024691358025,
+      "loss": 0.7493,
+      "step": 971
+    },
+    {
+      "epoch": 2.159400166620383,
+      "grad_norm": 0.011750995181500912,
+      "learning_rate": 0.0009333333333333333,
+      "loss": 0.7508,
+      "step": 972
+    },
+    {
+      "epoch": 2.1616217717300747,
+      "grad_norm": 0.006207957398146391,
+      "learning_rate": 0.0009308641975308642,
+      "loss": 0.7657,
+      "step": 973
+    },
+    {
+      "epoch": 2.163843376839767,
+      "grad_norm": 0.008501831442117691,
+      "learning_rate": 0.000928395061728395,
+      "loss": 0.708,
+      "step": 974
+    },
+    {
+      "epoch": 2.1660649819494586,
+      "grad_norm": 0.00984508078545332,
+      "learning_rate": 0.0009259259259259259,
+      "loss": 0.7141,
+      "step": 975
+    },
+    {
+      "epoch": 2.1682865870591503,
+      "grad_norm": 0.009435421787202358,
+      "learning_rate": 0.0009234567901234567,
+      "loss": 0.7219,
+      "step": 976
+    },
+    {
+      "epoch": 2.170508192168842,
+      "grad_norm": 0.008115294389426708,
+      "learning_rate": 0.0009209876543209878,
+      "loss": 0.8451,
+      "step": 977
+    },
+    {
+      "epoch": 2.1727297972785338,
+      "grad_norm": 0.00791158340871334,
+      "learning_rate": 0.0009185185185185186,
+      "loss": 0.7585,
+      "step": 978
+    },
+    {
+      "epoch": 2.1749514023882255,
+      "grad_norm": 0.009326069615781307,
+      "learning_rate": 0.0009160493827160494,
+      "loss": 0.8518,
+      "step": 979
+    },
+    {
+      "epoch": 2.177173007497917,
+      "grad_norm": 0.009076416492462158,
+      "learning_rate": 0.0009135802469135803,
+      "loss": 0.822,
+      "step": 980
+    },
+    {
+      "epoch": 2.179394612607609,
+      "grad_norm": 0.006390303373336792,
+      "learning_rate": 0.0009111111111111111,
+      "loss": 0.756,
+      "step": 981
+    },
+    {
+      "epoch": 2.1816162177173006,
+      "grad_norm": 0.008008323609828949,
+      "learning_rate": 0.000908641975308642,
+      "loss": 0.7159,
+      "step": 982
+    },
+    {
+      "epoch": 2.1838378228269923,
+      "grad_norm": 0.011111398227512836,
+      "learning_rate": 0.0009061728395061728,
+      "loss": 0.8722,
+      "step": 983
+    },
+    {
+      "epoch": 2.186059427936684,
+      "grad_norm": 0.012215813621878624,
+      "learning_rate": 0.0009037037037037038,
+      "loss": 0.8533,
+      "step": 984
+    },
+    {
+      "epoch": 2.1882810330463762,
+      "grad_norm": 0.005620912183076143,
+      "learning_rate": 0.0009012345679012346,
+      "loss": 0.6836,
+      "step": 985
+    },
+    {
+      "epoch": 2.190502638156068,
+      "grad_norm": 0.011297494173049927,
+      "learning_rate": 0.0008987654320987655,
+      "loss": 0.7197,
+      "step": 986
+    },
+    {
+      "epoch": 2.1927242432657597,
+      "grad_norm": 0.008543256670236588,
+      "learning_rate": 0.0008962962962962963,
+      "loss": 0.8217,
+      "step": 987
+    },
+    {
+      "epoch": 2.1949458483754514,
+      "grad_norm": 0.008594110608100891,
+      "learning_rate": 0.0008938271604938271,
+      "loss": 0.7988,
+      "step": 988
+    },
+    {
+      "epoch": 2.197167453485143,
+      "grad_norm": 0.007539823185652494,
+      "learning_rate": 0.000891358024691358,
+      "loss": 0.6496,
+      "step": 989
+    },
+    {
+      "epoch": 2.199389058594835,
+      "grad_norm": 0.008808733895421028,
+      "learning_rate": 0.0008888888888888888,
+      "loss": 0.6669,
+      "step": 990
+    },
+    {
+      "epoch": 2.2016106637045265,
+      "grad_norm": 0.00771710928529501,
+      "learning_rate": 0.0008864197530864197,
+      "loss": 0.798,
+      "step": 991
+    },
+    {
+      "epoch": 2.2038322688142182,
+      "grad_norm": 0.01634322665631771,
+      "learning_rate": 0.0008839506172839507,
+      "loss": 0.7739,
+      "step": 992
+    },
+    {
+      "epoch": 2.20605387392391,
+      "grad_norm": 0.0062980130314826965,
+      "learning_rate": 0.0008814814814814816,
+      "loss": 0.6469,
+      "step": 993
+    },
+    {
+      "epoch": 2.2082754790336017,
+      "grad_norm": 0.011210683733224869,
+      "learning_rate": 0.0008790123456790124,
+      "loss": 0.736,
+      "step": 994
+    },
+    {
+      "epoch": 2.2104970841432934,
+      "grad_norm": 0.009300244972109795,
+      "learning_rate": 0.0008765432098765433,
+      "loss": 0.7872,
+      "step": 995
+    },
+    {
+      "epoch": 2.212718689252985,
+      "grad_norm": 0.004812668543308973,
+      "learning_rate": 0.0008740740740740741,
+      "loss": 0.7092,
+      "step": 996
+    },
+    {
+      "epoch": 2.214940294362677,
+      "grad_norm": 0.00885671004652977,
+      "learning_rate": 0.0008716049382716049,
+      "loss": 0.5941,
+      "step": 997
+    },
+    {
+      "epoch": 2.217161899472369,
+      "grad_norm": 0.011171343736350536,
+      "learning_rate": 0.0008691358024691358,
+      "loss": 0.7145,
+      "step": 998
+    },
+    {
+      "epoch": 2.2193835045820607,
+      "grad_norm": 0.00593544589355588,
+      "learning_rate": 0.0008666666666666666,
+      "loss": 0.7192,
+      "step": 999
+    },
+    {
+      "epoch": 2.2216051096917524,
+      "grad_norm": 0.007853595539927483,
+      "learning_rate": 0.0008641975308641976,
+      "loss": 0.8245,
+      "step": 1000
+    },
+    {
+      "epoch": 2.223826714801444,
+      "grad_norm": 0.006899894680827856,
+      "learning_rate": 0.0008617283950617284,
+      "loss": 0.6957,
+      "step": 1001
+    },
+    {
+      "epoch": 2.226048319911136,
+      "grad_norm": 0.007371754385530949,
+      "learning_rate": 0.0008592592592592593,
+      "loss": 0.8765,
+      "step": 1002
+    },
+    {
+      "epoch": 2.2282699250208275,
+      "grad_norm": 0.007087046280503273,
+      "learning_rate": 0.0008567901234567901,
+      "loss": 0.9632,
+      "step": 1003
+    },
+    {
+      "epoch": 2.2304915301305193,
+      "grad_norm": 0.008594956248998642,
+      "learning_rate": 0.000854320987654321,
+      "loss": 0.6357,
+      "step": 1004
+    },
+    {
+      "epoch": 2.232713135240211,
+      "grad_norm": 0.00901806727051735,
+      "learning_rate": 0.0008518518518518518,
+      "loss": 0.8163,
+      "step": 1005
+    },
+    {
+      "epoch": 2.2349347403499027,
+      "grad_norm": 0.010117881000041962,
+      "learning_rate": 0.0008493827160493826,
+      "loss": 0.7334,
+      "step": 1006
+    },
+    {
+      "epoch": 2.2371563454595944,
+      "grad_norm": 0.006173667497932911,
+      "learning_rate": 0.0008469135802469137,
+      "loss": 0.7894,
+      "step": 1007
+    },
+    {
+      "epoch": 2.239377950569286,
+      "grad_norm": 0.006634249817579985,
+      "learning_rate": 0.0008444444444444445,
+      "loss": 0.8352,
+      "step": 1008
+    },
+    {
+      "epoch": 2.2415995556789783,
+      "grad_norm": 0.009885517880320549,
+      "learning_rate": 0.0008419753086419754,
+      "loss": 0.7064,
+      "step": 1009
+    },
+    {
+      "epoch": 2.24382116078867,
+      "grad_norm": 0.006060949992388487,
+      "learning_rate": 0.0008395061728395062,
+      "loss": 0.6139,
+      "step": 1010
+    },
+    {
+      "epoch": 2.2460427658983617,
+      "grad_norm": 0.010281536728143692,
+      "learning_rate": 0.0008370370370370371,
+      "loss": 0.7734,
+      "step": 1011
+    },
+    {
+      "epoch": 2.2482643710080534,
+      "grad_norm": 0.009147711098194122,
+      "learning_rate": 0.0008345679012345679,
+      "loss": 0.7489,
+      "step": 1012
+    },
+    {
+      "epoch": 2.250485976117745,
+      "grad_norm": 0.00839692261070013,
+      "learning_rate": 0.0008320987654320988,
+      "loss": 0.7873,
+      "step": 1013
+    },
+    {
+      "epoch": 2.252707581227437,
+      "grad_norm": 0.011448378674685955,
+      "learning_rate": 0.0008296296296296296,
+      "loss": 0.8176,
+      "step": 1014
+    },
+    {
+      "epoch": 2.2549291863371286,
+      "grad_norm": 0.009830589406192303,
+      "learning_rate": 0.0008271604938271605,
+      "loss": 0.7302,
+      "step": 1015
+    },
+    {
+      "epoch": 2.2571507914468203,
+      "grad_norm": 0.00909352395683527,
+      "learning_rate": 0.0008246913580246914,
+      "loss": 0.6294,
+      "step": 1016
+    },
+    {
+      "epoch": 2.259372396556512,
+      "grad_norm": 0.007446724455803633,
+      "learning_rate": 0.0008222222222222222,
+      "loss": 0.7363,
+      "step": 1017
+    },
+    {
+      "epoch": 2.2615940016662037,
+      "grad_norm": 0.0077172317542135715,
+      "learning_rate": 0.0008197530864197531,
+      "loss": 0.8876,
+      "step": 1018
+    },
+    {
+      "epoch": 2.2638156067758954,
+      "grad_norm": 0.00723393214866519,
+      "learning_rate": 0.0008172839506172839,
+      "loss": 0.8141,
+      "step": 1019
+    },
+    {
+      "epoch": 2.266037211885587,
+      "grad_norm": 0.008728005923330784,
+      "learning_rate": 0.0008148148148148148,
+      "loss": 0.6813,
+      "step": 1020
+    },
+    {
+      "epoch": 2.268258816995279,
+      "grad_norm": 0.006495292764157057,
+      "learning_rate": 0.0008123456790123456,
+      "loss": 0.6796,
+      "step": 1021
+    },
+    {
+      "epoch": 2.270480422104971,
+      "grad_norm": 0.011490698903799057,
+      "learning_rate": 0.0008098765432098767,
+      "loss": 0.7595,
+      "step": 1022
+    },
+    {
+      "epoch": 2.2727020272146627,
+      "grad_norm": 0.009576416574418545,
+      "learning_rate": 0.0008074074074074075,
+      "loss": 0.8657,
+      "step": 1023
+    },
+    {
+      "epoch": 2.2749236323243545,
+      "grad_norm": 0.010087347589433193,
+      "learning_rate": 0.0008049382716049384,
+      "loss": 0.6814,
+      "step": 1024
+    },
+    {
+      "epoch": 2.277145237434046,
+      "grad_norm": 0.0059828865341842175,
+      "learning_rate": 0.0008024691358024692,
+      "loss": 0.8354,
+      "step": 1025
+    },
+    {
+      "epoch": 2.279366842543738,
+      "grad_norm": 0.007989997044205666,
+      "learning_rate": 0.0008,
+      "loss": 0.9329,
+      "step": 1026
+    },
+    {
+      "epoch": 2.2815884476534296,
+      "grad_norm": 0.007406935561448336,
+      "learning_rate": 0.0007975308641975309,
+      "loss": 0.8708,
+      "step": 1027
+    },
+    {
+      "epoch": 2.2838100527631213,
+      "grad_norm": 0.009634488262236118,
+      "learning_rate": 0.0007950617283950617,
+      "loss": 0.7884,
+      "step": 1028
+    },
+    {
+      "epoch": 2.286031657872813,
+      "grad_norm": 0.012707880698144436,
+      "learning_rate": 0.0007925925925925926,
+      "loss": 0.8503,
+      "step": 1029
+    },
+    {
+      "epoch": 2.2882532629825048,
+      "grad_norm": 0.006519515533000231,
+      "learning_rate": 0.0007901234567901235,
+      "loss": 0.6777,
+      "step": 1030
+    },
+    {
+      "epoch": 2.2904748680921965,
+      "grad_norm": 0.007584378123283386,
+      "learning_rate": 0.0007876543209876544,
+      "loss": 0.9648,
+      "step": 1031
+    },
+    {
+      "epoch": 2.292696473201888,
+      "grad_norm": 0.009531829506158829,
+      "learning_rate": 0.0007851851851851852,
+      "loss": 0.8318,
+      "step": 1032
+    },
+    {
+      "epoch": 2.2949180783115803,
+      "grad_norm": 0.008382387459278107,
+      "learning_rate": 0.000782716049382716,
+      "loss": 0.7734,
+      "step": 1033
+    },
+    {
+      "epoch": 2.297139683421272,
+      "grad_norm": 0.02264496311545372,
+      "learning_rate": 0.0007802469135802469,
+      "loss": 0.9009,
+      "step": 1034
+    },
+    {
+      "epoch": 2.2993612885309638,
+      "grad_norm": 0.011763072572648525,
+      "learning_rate": 0.0007777777777777777,
+      "loss": 0.7595,
+      "step": 1035
+    },
+    {
+      "epoch": 2.3015828936406555,
+      "grad_norm": 0.01221343595534563,
+      "learning_rate": 0.0007753086419753086,
+      "loss": 0.8407,
+      "step": 1036
+    },
+    {
+      "epoch": 2.303804498750347,
+      "grad_norm": 0.00810176320374012,
+      "learning_rate": 0.0007728395061728396,
+      "loss": 0.5631,
+      "step": 1037
+    },
+    {
+      "epoch": 2.306026103860039,
+      "grad_norm": 0.011265031062066555,
+      "learning_rate": 0.0007703703703703705,
+      "loss": 0.9348,
+      "step": 1038
+    },
+    {
+      "epoch": 2.3082477089697306,
+      "grad_norm": 0.010535339824855328,
+      "learning_rate": 0.0007679012345679013,
+      "loss": 0.6664,
+      "step": 1039
+    },
+    {
+      "epoch": 2.3104693140794224,
+      "grad_norm": 0.008275596424937248,
+      "learning_rate": 0.0007654320987654322,
+      "loss": 0.8809,
+      "step": 1040
+    },
+    {
+      "epoch": 2.312690919189114,
+      "grad_norm": 0.009080667048692703,
+      "learning_rate": 0.000762962962962963,
+      "loss": 0.7974,
+      "step": 1041
+    },
+    {
+      "epoch": 2.314912524298806,
+      "grad_norm": 0.006603560876101255,
+      "learning_rate": 0.0007604938271604939,
+      "loss": 0.6765,
+      "step": 1042
+    },
+    {
+      "epoch": 2.3171341294084975,
+      "grad_norm": 0.009494476020336151,
+      "learning_rate": 0.0007580246913580247,
+      "loss": 0.6897,
+      "step": 1043
+    },
+    {
+      "epoch": 2.319355734518189,
+      "grad_norm": 0.0070472098886966705,
+      "learning_rate": 0.0007555555555555555,
+      "loss": 0.9325,
+      "step": 1044
+    },
+    {
+      "epoch": 2.321577339627881,
+      "grad_norm": 0.007084066979587078,
+      "learning_rate": 0.0007530864197530865,
+      "loss": 0.7297,
+      "step": 1045
+    },
+    {
+      "epoch": 2.3237989447375726,
+      "grad_norm": 0.010063939727842808,
+      "learning_rate": 0.0007506172839506173,
+      "loss": 1.0324,
+      "step": 1046
+    },
+    {
+      "epoch": 2.326020549847265,
+      "grad_norm": 0.009437466971576214,
+      "learning_rate": 0.0007481481481481482,
+      "loss": 0.845,
+      "step": 1047
+    },
+    {
+      "epoch": 2.3282421549569565,
+      "grad_norm": 0.006999650038778782,
+      "learning_rate": 0.000745679012345679,
+      "loss": 0.7047,
+      "step": 1048
+    },
+    {
+      "epoch": 2.3304637600666482,
+      "grad_norm": 0.008729071356356144,
+      "learning_rate": 0.0007432098765432099,
+      "loss": 0.7548,
+      "step": 1049
+    },
+    {
+      "epoch": 2.33268536517634,
+      "grad_norm": 0.009390546008944511,
+      "learning_rate": 0.0007407407407407407,
+      "loss": 0.7937,
+      "step": 1050
+    },
+    {
+      "epoch": 2.3349069702860317,
+      "grad_norm": 0.008248205296695232,
+      "learning_rate": 0.0007382716049382717,
+      "loss": 0.6691,
+      "step": 1051
+    },
+    {
+      "epoch": 2.3371285753957234,
+      "grad_norm": 0.008162553422152996,
+      "learning_rate": 0.0007358024691358025,
+      "loss": 0.8213,
+      "step": 1052
+    },
+    {
+      "epoch": 2.339350180505415,
+      "grad_norm": 0.005831165239214897,
+      "learning_rate": 0.0007333333333333333,
+      "loss": 0.8251,
+      "step": 1053
+    },
+    {
+      "epoch": 2.341571785615107,
+      "grad_norm": 0.00917159765958786,
+      "learning_rate": 0.0007308641975308643,
+      "loss": 0.7395,
+      "step": 1054
+    },
+    {
+      "epoch": 2.3437933907247985,
+      "grad_norm": 0.007398424670100212,
+      "learning_rate": 0.0007283950617283951,
+      "loss": 0.7261,
+      "step": 1055
+    },
+    {
+      "epoch": 2.3460149958344902,
+      "grad_norm": 0.012491940520703793,
+      "learning_rate": 0.000725925925925926,
+      "loss": 0.84,
+      "step": 1056
+    },
+    {
+      "epoch": 2.3482366009441824,
+      "grad_norm": 0.0047965748235583305,
+      "learning_rate": 0.0007234567901234568,
+      "loss": 0.5854,
+      "step": 1057
+    },
+    {
+      "epoch": 2.350458206053874,
+      "grad_norm": 0.00818257499486208,
+      "learning_rate": 0.0007209876543209877,
+      "loss": 0.7076,
+      "step": 1058
+    },
+    {
+      "epoch": 2.352679811163566,
+      "grad_norm": 0.013595851138234138,
+      "learning_rate": 0.0007185185185185185,
+      "loss": 0.822,
+      "step": 1059
+    },
+    {
+      "epoch": 2.3549014162732576,
+      "grad_norm": 0.013108041137456894,
+      "learning_rate": 0.0007160493827160494,
+      "loss": 0.8705,
+      "step": 1060
+    },
+    {
+      "epoch": 2.3571230213829493,
+      "grad_norm": 0.008615655824542046,
+      "learning_rate": 0.0007135802469135802,
+      "loss": 0.5893,
+      "step": 1061
+    },
+    {
+      "epoch": 2.359344626492641,
+      "grad_norm": 0.0071796211414039135,
+      "learning_rate": 0.0007111111111111111,
+      "loss": 0.6746,
+      "step": 1062
+    },
+    {
+      "epoch": 2.3615662316023327,
+      "grad_norm": 0.009801715612411499,
+      "learning_rate": 0.000708641975308642,
+      "loss": 0.7323,
+      "step": 1063
+    },
+    {
+      "epoch": 2.3637878367120244,
+      "grad_norm": 0.006935875862836838,
+      "learning_rate": 0.0007061728395061728,
+      "loss": 0.6184,
+      "step": 1064
+    },
+    {
+      "epoch": 2.366009441821716,
+      "grad_norm": 0.009273690171539783,
+      "learning_rate": 0.0007037037037037037,
+      "loss": 0.7003,
+      "step": 1065
+    },
+    {
+      "epoch": 2.368231046931408,
+      "grad_norm": 0.008863688446581364,
+      "learning_rate": 0.0007012345679012346,
+      "loss": 0.906,
+      "step": 1066
+    },
+    {
+      "epoch": 2.3704526520410996,
+      "grad_norm": 0.00721506355330348,
+      "learning_rate": 0.0006987654320987655,
+      "loss": 0.69,
+      "step": 1067
+    },
+    {
+      "epoch": 2.3726742571507913,
+      "grad_norm": 0.010040674358606339,
+      "learning_rate": 0.0006962962962962963,
+      "loss": 0.7415,
+      "step": 1068
+    },
+    {
+      "epoch": 2.374895862260483,
+      "grad_norm": 0.04709119722247124,
+      "learning_rate": 0.0006938271604938273,
+      "loss": 0.8354,
+      "step": 1069
+    },
+    {
+      "epoch": 2.3771174673701747,
+      "grad_norm": 0.009449060074985027,
+      "learning_rate": 0.0006913580246913581,
+      "loss": 0.8285,
+      "step": 1070
+    },
+    {
+      "epoch": 2.379339072479867,
+      "grad_norm": 0.009565912187099457,
+      "learning_rate": 0.000688888888888889,
+      "loss": 0.6402,
+      "step": 1071
+    },
+    {
+      "epoch": 2.3815606775895586,
+      "grad_norm": 0.010059806518256664,
+      "learning_rate": 0.0006864197530864198,
+      "loss": 0.9116,
+      "step": 1072
+    },
+    {
+      "epoch": 2.3837822826992503,
+      "grad_norm": 0.008675451390445232,
+      "learning_rate": 0.0006839506172839506,
+      "loss": 0.7108,
+      "step": 1073
+    },
+    {
+      "epoch": 2.386003887808942,
+      "grad_norm": 0.011516880244016647,
+      "learning_rate": 0.0006814814814814815,
+      "loss": 0.7315,
+      "step": 1074
+    },
+    {
+      "epoch": 2.3882254929186337,
+      "grad_norm": 0.0067232525907456875,
+      "learning_rate": 0.0006790123456790123,
+      "loss": 0.7973,
+      "step": 1075
+    },
+    {
+      "epoch": 2.3904470980283254,
+      "grad_norm": 0.007842687889933586,
+      "learning_rate": 0.0006765432098765432,
+      "loss": 0.7277,
+      "step": 1076
+    },
+    {
+      "epoch": 2.392668703138017,
+      "grad_norm": 0.01157988142222166,
+      "learning_rate": 0.0006740740740740741,
+      "loss": 0.7497,
+      "step": 1077
+    },
+    {
+      "epoch": 2.394890308247709,
+      "grad_norm": 0.007398506626486778,
+      "learning_rate": 0.000671604938271605,
+      "loss": 0.827,
+      "step": 1078
+    },
+    {
+      "epoch": 2.3971119133574006,
+      "grad_norm": 0.005851605907082558,
+      "learning_rate": 0.0006691358024691358,
+      "loss": 0.593,
+      "step": 1079
+    },
+    {
+      "epoch": 2.3993335184670923,
+      "grad_norm": 0.006889878306537867,
+      "learning_rate": 0.0006666666666666666,
+      "loss": 0.6059,
+      "step": 1080
+    },
+    {
+      "epoch": 2.4015551235767845,
+      "grad_norm": 0.011620272882282734,
+      "learning_rate": 0.0006641975308641976,
+      "loss": 0.8598,
+      "step": 1081
+    },
+    {
+      "epoch": 2.403776728686476,
+      "grad_norm": 0.009120738133788109,
+      "learning_rate": 0.0006617283950617284,
+      "loss": 0.7034,
+      "step": 1082
+    },
+    {
+      "epoch": 2.405998333796168,
+      "grad_norm": 0.012162731029093266,
+      "learning_rate": 0.0006592592592592593,
+      "loss": 0.6221,
+      "step": 1083
+    },
+    {
+      "epoch": 2.4082199389058596,
+      "grad_norm": 0.00817908439785242,
+      "learning_rate": 0.0006567901234567901,
+      "loss": 0.8944,
+      "step": 1084
+    },
+    {
+      "epoch": 2.4104415440155513,
+      "grad_norm": 0.008096316829323769,
+      "learning_rate": 0.000654320987654321,
+      "loss": 0.7351,
+      "step": 1085
+    },
+    {
+      "epoch": 2.412663149125243,
+      "grad_norm": 0.011929173953831196,
+      "learning_rate": 0.0006518518518518518,
+      "loss": 0.6995,
+      "step": 1086
+    },
+    {
+      "epoch": 2.4148847542349348,
+      "grad_norm": 0.009247666224837303,
+      "learning_rate": 0.0006493827160493827,
+      "loss": 0.7163,
+      "step": 1087
+    },
+    {
+      "epoch": 2.4171063593446265,
+      "grad_norm": 0.00713922968134284,
+      "learning_rate": 0.0006469135802469136,
+      "loss": 0.8009,
+      "step": 1088
+    },
+    {
+      "epoch": 2.419327964454318,
+      "grad_norm": 0.008218400180339813,
+      "learning_rate": 0.0006444444444444444,
+      "loss": 0.6624,
+      "step": 1089
+    },
+    {
+      "epoch": 2.42154956956401,
+      "grad_norm": 0.014460783451795578,
+      "learning_rate": 0.0006419753086419753,
+      "loss": 0.7123,
+      "step": 1090
+    },
+    {
+      "epoch": 2.4237711746737016,
+      "grad_norm": 0.010817009955644608,
+      "learning_rate": 0.0006395061728395061,
+      "loss": 0.7479,
+      "step": 1091
+    },
+    {
+      "epoch": 2.4259927797833933,
+      "grad_norm": 0.005958953872323036,
+      "learning_rate": 0.0006370370370370371,
+      "loss": 0.7389,
+      "step": 1092
+    },
+    {
+      "epoch": 2.428214384893085,
+      "grad_norm": 0.007965938188135624,
+      "learning_rate": 0.0006345679012345679,
+      "loss": 0.7421,
+      "step": 1093
+    },
+    {
+      "epoch": 2.4304359900027768,
+      "grad_norm": 0.011952990666031837,
+      "learning_rate": 0.0006320987654320988,
+      "loss": 0.693,
+      "step": 1094
+    },
+    {
+      "epoch": 2.432657595112469,
+      "grad_norm": 0.006782431621104479,
+      "learning_rate": 0.0006296296296296296,
+      "loss": 0.7593,
+      "step": 1095
+    },
+    {
+      "epoch": 2.4348792002221606,
+      "grad_norm": 0.007503021042793989,
+      "learning_rate": 0.0006271604938271606,
+      "loss": 0.9018,
+      "step": 1096
+    },
+    {
+      "epoch": 2.4371008053318524,
+      "grad_norm": 0.009170000441372395,
+      "learning_rate": 0.0006246913580246914,
+      "loss": 0.779,
+      "step": 1097
+    },
+    {
+      "epoch": 2.439322410441544,
+      "grad_norm": 0.010611280798912048,
+      "learning_rate": 0.0006222222222222223,
+      "loss": 0.7442,
+      "step": 1098
+    },
+    {
+      "epoch": 2.441544015551236,
+      "grad_norm": 0.007238518912345171,
+      "learning_rate": 0.0006197530864197531,
+      "loss": 0.8331,
+      "step": 1099
+    },
+    {
+      "epoch": 2.4437656206609275,
+      "grad_norm": 0.007917318493127823,
+      "learning_rate": 0.0006172839506172839,
+      "loss": 0.6429,
+      "step": 1100
+    },
+    {
+      "epoch": 2.4459872257706192,
+      "grad_norm": 0.007869457826018333,
+      "learning_rate": 0.0006148148148148148,
+      "loss": 0.7167,
+      "step": 1101
+    },
+    {
+      "epoch": 2.448208830880311,
+      "grad_norm": 0.010885112918913364,
+      "learning_rate": 0.0006123456790123456,
+      "loss": 0.9405,
+      "step": 1102
+    },
+    {
+      "epoch": 2.4504304359900027,
+      "grad_norm": 0.005701167043298483,
+      "learning_rate": 0.0006098765432098766,
+      "loss": 0.6204,
+      "step": 1103
+    },
+    {
+      "epoch": 2.4526520410996944,
+      "grad_norm": 0.008726618252694607,
+      "learning_rate": 0.0006074074074074074,
+      "loss": 0.6938,
+      "step": 1104
+    },
+    {
+      "epoch": 2.4548736462093865,
+      "grad_norm": 0.009071066044270992,
+      "learning_rate": 0.0006049382716049383,
+      "loss": 0.7146,
+      "step": 1105
+    },
+    {
+      "epoch": 2.4570952513190782,
+      "grad_norm": 0.00839879922568798,
+      "learning_rate": 0.0006024691358024691,
+      "loss": 0.763,
+      "step": 1106
+    },
+    {
+      "epoch": 2.45931685642877,
+      "grad_norm": 0.010271345265209675,
+      "learning_rate": 0.0006000000000000001,
+      "loss": 0.827,
+      "step": 1107
+    },
+    {
+      "epoch": 2.4615384615384617,
+      "grad_norm": 0.005833013448864222,
+      "learning_rate": 0.0005975308641975309,
+      "loss": 0.8062,
+      "step": 1108
+    },
+    {
+      "epoch": 2.4637600666481534,
+      "grad_norm": 0.01235662680119276,
+      "learning_rate": 0.0005950617283950617,
+      "loss": 0.6257,
+      "step": 1109
+    },
+    {
+      "epoch": 2.465981671757845,
+      "grad_norm": 0.006829334422945976,
+      "learning_rate": 0.0005925925925925926,
+      "loss": 0.8064,
+      "step": 1110
+    },
+    {
+      "epoch": 2.468203276867537,
+      "grad_norm": 0.010129845701158047,
+      "learning_rate": 0.0005901234567901235,
+      "loss": 0.665,
+      "step": 1111
+    },
+    {
+      "epoch": 2.4704248819772285,
+      "grad_norm": 0.009496810846030712,
+      "learning_rate": 0.0005876543209876544,
+      "loss": 0.9311,
+      "step": 1112
+    },
+    {
+      "epoch": 2.4726464870869203,
+      "grad_norm": 0.007677487563341856,
+      "learning_rate": 0.0005851851851851852,
+      "loss": 1.0702,
+      "step": 1113
+    },
+    {
+      "epoch": 2.474868092196612,
+      "grad_norm": 0.006991180591285229,
+      "learning_rate": 0.0005827160493827161,
+      "loss": 1.0201,
+      "step": 1114
+    },
+    {
+      "epoch": 2.4770896973063037,
+      "grad_norm": 0.008119228295981884,
+      "learning_rate": 0.0005802469135802469,
+      "loss": 0.8175,
+      "step": 1115
+    },
+    {
+      "epoch": 2.4793113024159954,
+      "grad_norm": 0.006982423830777407,
+      "learning_rate": 0.0005777777777777778,
+      "loss": 0.6665,
+      "step": 1116
+    },
+    {
+      "epoch": 2.481532907525687,
+      "grad_norm": 0.006513761822134256,
+      "learning_rate": 0.0005753086419753086,
+      "loss": 0.7288,
+      "step": 1117
+    },
+    {
+      "epoch": 2.483754512635379,
+      "grad_norm": 0.009115753695368767,
+      "learning_rate": 0.0005728395061728395,
+      "loss": 0.8318,
+      "step": 1118
+    },
+    {
+      "epoch": 2.485976117745071,
+      "grad_norm": 0.007586904335767031,
+      "learning_rate": 0.0005703703703703704,
+      "loss": 0.725,
+      "step": 1119
+    },
+    {
+      "epoch": 2.4881977228547627,
+      "grad_norm": 0.006441600155085325,
+      "learning_rate": 0.0005679012345679012,
+      "loss": 0.7512,
+      "step": 1120
+    },
+    {
+      "epoch": 2.4904193279644544,
+      "grad_norm": 0.008771124295890331,
+      "learning_rate": 0.0005654320987654321,
+      "loss": 0.7566,
+      "step": 1121
+    },
+    {
+      "epoch": 2.492640933074146,
+      "grad_norm": 0.010517198592424393,
+      "learning_rate": 0.000562962962962963,
+      "loss": 0.7906,
+      "step": 1122
+    },
+    {
+      "epoch": 2.494862538183838,
+      "grad_norm": 0.009617543779313564,
+      "learning_rate": 0.0005604938271604939,
+      "loss": 1.0293,
+      "step": 1123
+    },
+    {
+      "epoch": 2.4970841432935296,
+      "grad_norm": 0.009210485965013504,
+      "learning_rate": 0.0005580246913580247,
+      "loss": 0.881,
+      "step": 1124
+    },
+    {
+      "epoch": 2.4993057484032213,
+      "grad_norm": 0.005644843447953463,
+      "learning_rate": 0.0005555555555555556,
+      "loss": 0.8017,
+      "step": 1125
+    },
+    {
+      "epoch": 2.501527353512913,
+      "grad_norm": 0.010391274467110634,
+      "learning_rate": 0.0005530864197530865,
+      "loss": 0.8173,
+      "step": 1126
+    },
+    {
+      "epoch": 2.5037489586226047,
+      "grad_norm": 0.007817761972546577,
+      "learning_rate": 0.0005506172839506173,
+      "loss": 0.6681,
+      "step": 1127
+    },
+    {
+      "epoch": 2.5059705637322964,
+      "grad_norm": 0.008231202140450478,
+      "learning_rate": 0.0005481481481481482,
+      "loss": 0.6618,
+      "step": 1128
+    },
+    {
+      "epoch": 2.5081921688419886,
+      "grad_norm": 0.008874166756868362,
+      "learning_rate": 0.000545679012345679,
+      "loss": 0.6056,
+      "step": 1129
+    },
+    {
+      "epoch": 2.5104137739516803,
+      "grad_norm": 0.01048712246119976,
+      "learning_rate": 0.0005432098765432099,
+      "loss": 0.7019,
+      "step": 1130
+    },
+    {
+      "epoch": 2.512635379061372,
+      "grad_norm": 0.008290881291031837,
+      "learning_rate": 0.0005407407407407407,
+      "loss": 0.6494,
+      "step": 1131
+    },
+    {
+      "epoch": 2.5148569841710637,
+      "grad_norm": 0.007482048589736223,
+      "learning_rate": 0.0005382716049382716,
+      "loss": 0.8287,
+      "step": 1132
+    },
+    {
+      "epoch": 2.5170785892807555,
+      "grad_norm": 0.008649013936519623,
+      "learning_rate": 0.0005358024691358024,
+      "loss": 0.8579,
+      "step": 1133
+    },
+    {
+      "epoch": 2.519300194390447,
+      "grad_norm": 0.008483688347041607,
+      "learning_rate": 0.0005333333333333334,
+      "loss": 0.6852,
+      "step": 1134
+    },
+    {
+      "epoch": 2.521521799500139,
+      "grad_norm": 0.008373193442821503,
+      "learning_rate": 0.0005308641975308642,
+      "loss": 0.7122,
+      "step": 1135
+    },
+    {
+      "epoch": 2.5237434046098306,
+      "grad_norm": 0.007378621492534876,
+      "learning_rate": 0.000528395061728395,
+      "loss": 0.8437,
+      "step": 1136
+    },
+    {
+      "epoch": 2.5259650097195223,
+      "grad_norm": 0.008011176250874996,
+      "learning_rate": 0.000525925925925926,
+      "loss": 0.5955,
+      "step": 1137
+    },
+    {
+      "epoch": 2.528186614829214,
+      "grad_norm": 0.009291508235037327,
+      "learning_rate": 0.0005234567901234568,
+      "loss": 0.8159,
+      "step": 1138
+    },
+    {
+      "epoch": 2.5304082199389057,
+      "grad_norm": 0.008907923474907875,
+      "learning_rate": 0.0005209876543209877,
+      "loss": 0.8137,
+      "step": 1139
+    },
+    {
+      "epoch": 2.5326298250485975,
+      "grad_norm": 0.008879524655640125,
+      "learning_rate": 0.0005185185185185185,
+      "loss": 0.678,
+      "step": 1140
+    },
+    {
+      "epoch": 2.534851430158289,
+      "grad_norm": 0.00674105528742075,
+      "learning_rate": 0.0005160493827160495,
+      "loss": 0.8209,
+      "step": 1141
+    },
+    {
+      "epoch": 2.537073035267981,
+      "grad_norm": 0.008492392487823963,
+      "learning_rate": 0.0005135802469135803,
+      "loss": 0.8017,
+      "step": 1142
+    },
+    {
+      "epoch": 2.5392946403776726,
+      "grad_norm": 0.012028438970446587,
+      "learning_rate": 0.0005111111111111112,
+      "loss": 0.8741,
+      "step": 1143
+    },
+    {
+      "epoch": 2.5415162454873648,
+      "grad_norm": 0.012991204857826233,
+      "learning_rate": 0.000508641975308642,
+      "loss": 0.6765,
+      "step": 1144
+    },
+    {
+      "epoch": 2.5437378505970565,
+      "grad_norm": 0.007418293505907059,
+      "learning_rate": 0.0005061728395061728,
+      "loss": 0.8896,
+      "step": 1145
+    },
+    {
+      "epoch": 2.545959455706748,
+      "grad_norm": 0.0103315319865942,
+      "learning_rate": 0.0005037037037037037,
+      "loss": 0.8652,
+      "step": 1146
+    },
+    {
+      "epoch": 2.54818106081644,
+      "grad_norm": 0.008294169791042805,
+      "learning_rate": 0.0005012345679012345,
+      "loss": 0.7071,
+      "step": 1147
+    },
+    {
+      "epoch": 2.5504026659261316,
+      "grad_norm": 0.00622114073485136,
+      "learning_rate": 0.0004987654320987654,
+      "loss": 0.8395,
+      "step": 1148
+    },
+    {
+      "epoch": 2.5526242710358233,
+      "grad_norm": 0.006999352015554905,
+      "learning_rate": 0.0004962962962962963,
+      "loss": 0.6672,
+      "step": 1149
+    },
+    {
+      "epoch": 2.554845876145515,
+      "grad_norm": 0.008257112465798855,
+      "learning_rate": 0.0004938271604938272,
+      "loss": 0.7719,
+      "step": 1150
+    },
+    {
+      "epoch": 2.5570674812552068,
+      "grad_norm": 0.0099422512575984,
+      "learning_rate": 0.000491358024691358,
+      "loss": 0.8573,
+      "step": 1151
+    },
+    {
+      "epoch": 2.5592890863648985,
+      "grad_norm": 0.013014703057706356,
+      "learning_rate": 0.000488888888888889,
+      "loss": 0.9367,
+      "step": 1152
+    },
+    {
+      "epoch": 2.5615106914745907,
+      "grad_norm": 0.006244743708521128,
+      "learning_rate": 0.00048641975308641976,
+      "loss": 0.7401,
+      "step": 1153
+    },
+    {
+      "epoch": 2.5637322965842824,
+      "grad_norm": 0.007896228693425655,
+      "learning_rate": 0.0004839506172839506,
+      "loss": 0.6035,
+      "step": 1154
+    },
+    {
+      "epoch": 2.565953901693974,
+      "grad_norm": 0.012634181417524815,
+      "learning_rate": 0.00048148148148148144,
+      "loss": 0.7003,
+      "step": 1155
+    },
+    {
+      "epoch": 2.568175506803666,
+      "grad_norm": 0.010985374450683594,
+      "learning_rate": 0.0004790123456790124,
+      "loss": 1.0121,
+      "step": 1156
+    },
+    {
+      "epoch": 2.5703971119133575,
+      "grad_norm": 0.007849064655601978,
+      "learning_rate": 0.00047654320987654324,
+      "loss": 0.8866,
+      "step": 1157
+    },
+    {
+      "epoch": 2.5726187170230492,
+      "grad_norm": 0.012782498262822628,
+      "learning_rate": 0.0004740740740740741,
+      "loss": 0.8004,
+      "step": 1158
+    },
+    {
+      "epoch": 2.574840322132741,
+      "grad_norm": 0.01364060677587986,
+      "learning_rate": 0.0004716049382716049,
+      "loss": 0.8129,
+      "step": 1159
+    },
+    {
+      "epoch": 2.5770619272424327,
+      "grad_norm": 0.0071340943686664104,
+      "learning_rate": 0.0004691358024691358,
+      "loss": 0.7682,
+      "step": 1160
+    },
+    {
+      "epoch": 2.5792835323521244,
+      "grad_norm": 0.00822313129901886,
+      "learning_rate": 0.00046666666666666666,
+      "loss": 1.0239,
+      "step": 1161
+    },
+    {
+      "epoch": 2.581505137461816,
+      "grad_norm": 0.009147636592388153,
+      "learning_rate": 0.0004641975308641975,
+      "loss": 0.7753,
+      "step": 1162
+    },
+    {
+      "epoch": 2.583726742571508,
+      "grad_norm": 0.00680557219311595,
+      "learning_rate": 0.00046172839506172835,
+      "loss": 0.8775,
+      "step": 1163
+    },
+    {
+      "epoch": 2.5859483476811995,
+      "grad_norm": 0.012209068052470684,
+      "learning_rate": 0.0004592592592592593,
+      "loss": 0.9347,
+      "step": 1164
+    },
+    {
+      "epoch": 2.5881699527908912,
+      "grad_norm": 0.008567320182919502,
+      "learning_rate": 0.00045679012345679014,
+      "loss": 0.667,
+      "step": 1165
+    },
+    {
+      "epoch": 2.590391557900583,
+      "grad_norm": 0.009676763787865639,
+      "learning_rate": 0.000454320987654321,
+      "loss": 0.8293,
+      "step": 1166
+    },
+    {
+      "epoch": 2.5926131630102747,
+      "grad_norm": 0.0067940810695290565,
+      "learning_rate": 0.0004518518518518519,
+      "loss": 0.9675,
+      "step": 1167
+    },
+    {
+      "epoch": 2.594834768119967,
+      "grad_norm": 0.006839894689619541,
+      "learning_rate": 0.0004493827160493827,
+      "loss": 0.8892,
+      "step": 1168
+    },
+    {
+      "epoch": 2.5970563732296585,
+      "grad_norm": 0.006143914069980383,
+      "learning_rate": 0.00044691358024691357,
+      "loss": 0.8653,
+      "step": 1169
+    },
+    {
+      "epoch": 2.5992779783393503,
+      "grad_norm": 0.0072947400622069836,
+      "learning_rate": 0.0004444444444444444,
+      "loss": 0.7919,
+      "step": 1170
+    },
+    {
+      "epoch": 2.601499583449042,
+      "grad_norm": 0.006502422038465738,
+      "learning_rate": 0.00044197530864197536,
+      "loss": 0.7423,
+      "step": 1171
+    },
+    {
+      "epoch": 2.6037211885587337,
+      "grad_norm": 0.006935762241482735,
+      "learning_rate": 0.0004395061728395062,
+      "loss": 0.9463,
+      "step": 1172
+    },
+    {
+      "epoch": 2.6059427936684254,
+      "grad_norm": 0.010217388160526752,
+      "learning_rate": 0.00043703703703703705,
+      "loss": 0.7173,
+      "step": 1173
+    },
+    {
+      "epoch": 2.608164398778117,
+      "grad_norm": 0.010664797388017178,
+      "learning_rate": 0.0004345679012345679,
+      "loss": 0.8849,
+      "step": 1174
+    },
+    {
+      "epoch": 2.610386003887809,
+      "grad_norm": 0.00750716682523489,
+      "learning_rate": 0.0004320987654320988,
+      "loss": 0.7131,
+      "step": 1175
+    },
+    {
+      "epoch": 2.6126076089975006,
+      "grad_norm": 0.009238378144800663,
+      "learning_rate": 0.00042962962962962963,
+      "loss": 0.6628,
+      "step": 1176
+    },
+    {
+      "epoch": 2.6148292141071927,
+      "grad_norm": 0.019038479775190353,
+      "learning_rate": 0.0004271604938271605,
+      "loss": 0.8115,
+      "step": 1177
+    },
+    {
+      "epoch": 2.6170508192168844,
+      "grad_norm": 0.008630185388028622,
+      "learning_rate": 0.0004246913580246913,
+      "loss": 0.6493,
+      "step": 1178
+    },
+    {
+      "epoch": 2.619272424326576,
+      "grad_norm": 0.0108924126252532,
+      "learning_rate": 0.00042222222222222227,
+      "loss": 0.986,
+      "step": 1179
+    },
+    {
+      "epoch": 2.621494029436268,
+      "grad_norm": 0.005341158714145422,
+      "learning_rate": 0.0004197530864197531,
+      "loss": 0.8286,
+      "step": 1180
+    },
+    {
+      "epoch": 2.6237156345459596,
+      "grad_norm": 0.008509661071002483,
+      "learning_rate": 0.00041728395061728396,
+      "loss": 0.7908,
+      "step": 1181
+    },
+    {
+      "epoch": 2.6259372396556513,
+      "grad_norm": 0.013711505569517612,
+      "learning_rate": 0.0004148148148148148,
+      "loss": 0.7929,
+      "step": 1182
+    },
+    {
+      "epoch": 2.628158844765343,
+      "grad_norm": 0.008806861005723476,
+      "learning_rate": 0.0004123456790123457,
+      "loss": 0.9112,
+      "step": 1183
+    },
+    {
+      "epoch": 2.6303804498750347,
+      "grad_norm": 0.01078883744776249,
+      "learning_rate": 0.00040987654320987654,
+      "loss": 0.5432,
+      "step": 1184
+    },
+    {
+      "epoch": 2.6326020549847264,
+      "grad_norm": 0.007406941149383783,
+      "learning_rate": 0.0004074074074074074,
+      "loss": 0.8226,
+      "step": 1185
+    },
+    {
+      "epoch": 2.634823660094418,
+      "grad_norm": 0.016257604584097862,
+      "learning_rate": 0.00040493827160493833,
+      "loss": 0.7704,
+      "step": 1186
+    },
+    {
+      "epoch": 2.63704526520411,
+      "grad_norm": 0.00873923022300005,
+      "learning_rate": 0.0004024691358024692,
+      "loss": 0.7444,
+      "step": 1187
+    },
+    {
+      "epoch": 2.6392668703138016,
+      "grad_norm": 0.010216313414275646,
+      "learning_rate": 0.0004,
+      "loss": 0.7796,
+      "step": 1188
+    },
+    {
+      "epoch": 2.6414884754234933,
+      "grad_norm": 0.015559897758066654,
+      "learning_rate": 0.00039753086419753086,
+      "loss": 0.8496,
+      "step": 1189
+    },
+    {
+      "epoch": 2.643710080533185,
+      "grad_norm": 0.007304896134883165,
+      "learning_rate": 0.00039506172839506176,
+      "loss": 0.88,
+      "step": 1190
+    },
+    {
+      "epoch": 2.6459316856428767,
+      "grad_norm": 0.007256948854774237,
+      "learning_rate": 0.0003925925925925926,
+      "loss": 0.7941,
+      "step": 1191
+    },
+    {
+      "epoch": 2.648153290752569,
+      "grad_norm": 0.0061131990514695644,
+      "learning_rate": 0.00039012345679012345,
+      "loss": 0.824,
+      "step": 1192
+    },
+    {
+      "epoch": 2.6503748958622606,
+      "grad_norm": 0.009228897280991077,
+      "learning_rate": 0.0003876543209876543,
+      "loss": 0.9218,
+      "step": 1193
+    },
+    {
+      "epoch": 2.6525965009719523,
+      "grad_norm": 0.007349466439336538,
+      "learning_rate": 0.00038518518518518524,
+      "loss": 0.674,
+      "step": 1194
+    },
+    {
+      "epoch": 2.654818106081644,
+      "grad_norm": 0.0076787457801401615,
+      "learning_rate": 0.0003827160493827161,
+      "loss": 0.7444,
+      "step": 1195
+    },
+    {
+      "epoch": 2.6570397111913358,
+      "grad_norm": 0.010441628284752369,
+      "learning_rate": 0.0003802469135802469,
+      "loss": 0.9451,
+      "step": 1196
+    },
+    {
+      "epoch": 2.6592613163010275,
+      "grad_norm": 0.01100853830575943,
+      "learning_rate": 0.00037777777777777777,
+      "loss": 0.9497,
+      "step": 1197
+    },
+    {
+      "epoch": 2.661482921410719,
+      "grad_norm": 0.008857887238264084,
+      "learning_rate": 0.00037530864197530867,
+      "loss": 0.812,
+      "step": 1198
+    },
+    {
+      "epoch": 2.663704526520411,
+      "grad_norm": 0.0065878876484930515,
+      "learning_rate": 0.0003728395061728395,
+      "loss": 0.8649,
+      "step": 1199
+    },
+    {
+      "epoch": 2.6659261316301026,
+      "grad_norm": 0.009736582636833191,
+      "learning_rate": 0.00037037037037037035,
+      "loss": 0.5912,
+      "step": 1200
+    },
+    {
+      "epoch": 2.6681477367397948,
+      "grad_norm": 0.012062900699675083,
+      "learning_rate": 0.00036790123456790125,
+      "loss": 0.9503,
+      "step": 1201
+    },
+    {
+      "epoch": 2.6703693418494865,
+      "grad_norm": 0.016475075855851173,
+      "learning_rate": 0.00036543209876543215,
+      "loss": 0.635,
+      "step": 1202
+    },
+    {
+      "epoch": 2.672590946959178,
+      "grad_norm": 0.008248421363532543,
+      "learning_rate": 0.000362962962962963,
+      "loss": 0.741,
+      "step": 1203
+    },
+    {
+      "epoch": 2.67481255206887,
+      "grad_norm": 0.006326109170913696,
+      "learning_rate": 0.00036049382716049383,
+      "loss": 0.6972,
+      "step": 1204
+    },
+    {
+      "epoch": 2.6770341571785616,
+      "grad_norm": 0.007503593806177378,
+      "learning_rate": 0.0003580246913580247,
+      "loss": 0.7367,
+      "step": 1205
+    },
+    {
+      "epoch": 2.6792557622882534,
+      "grad_norm": 0.0090335663408041,
+      "learning_rate": 0.00035555555555555557,
+      "loss": 0.8782,
+      "step": 1206
+    },
+    {
+      "epoch": 2.681477367397945,
+      "grad_norm": 0.013283466920256615,
+      "learning_rate": 0.0003530864197530864,
+      "loss": 0.8175,
+      "step": 1207
+    },
+    {
+      "epoch": 2.683698972507637,
+      "grad_norm": 0.007493285927921534,
+      "learning_rate": 0.0003506172839506173,
+      "loss": 0.7386,
+      "step": 1208
+    },
+    {
+      "epoch": 2.6859205776173285,
+      "grad_norm": 0.011668765917420387,
+      "learning_rate": 0.00034814814814814816,
+      "loss": 1.0218,
+      "step": 1209
+    },
+    {
+      "epoch": 2.68814218272702,
+      "grad_norm": 0.008925607427954674,
+      "learning_rate": 0.00034567901234567905,
+      "loss": 0.7263,
+      "step": 1210
+    },
+    {
+      "epoch": 2.690363787836712,
+      "grad_norm": 0.009130970574915409,
+      "learning_rate": 0.0003432098765432099,
+      "loss": 0.7787,
+      "step": 1211
+    },
+    {
+      "epoch": 2.6925853929464036,
+      "grad_norm": 0.006965592969208956,
+      "learning_rate": 0.00034074074074074074,
+      "loss": 0.7084,
+      "step": 1212
+    },
+    {
+      "epoch": 2.6948069980560954,
+      "grad_norm": 0.03262852504849434,
+      "learning_rate": 0.0003382716049382716,
+      "loss": 0.8719,
+      "step": 1213
+    },
+    {
+      "epoch": 2.697028603165787,
+      "grad_norm": 0.00794121716171503,
+      "learning_rate": 0.0003358024691358025,
+      "loss": 0.7461,
+      "step": 1214
+    },
+    {
+      "epoch": 2.699250208275479,
+      "grad_norm": 0.00642223609611392,
+      "learning_rate": 0.0003333333333333333,
+      "loss": 0.8822,
+      "step": 1215
+    },
+    {
+      "epoch": 2.701471813385171,
+      "grad_norm": 0.009445881471037865,
+      "learning_rate": 0.0003308641975308642,
+      "loss": 0.7372,
+      "step": 1216
+    },
+    {
+      "epoch": 2.7036934184948627,
+      "grad_norm": 0.012715525925159454,
+      "learning_rate": 0.00032839506172839506,
+      "loss": 0.6933,
+      "step": 1217
+    },
+    {
+      "epoch": 2.7059150236045544,
+      "grad_norm": 0.016722604632377625,
+      "learning_rate": 0.0003259259259259259,
+      "loss": 0.9278,
+      "step": 1218
+    },
+    {
+      "epoch": 2.708136628714246,
+      "grad_norm": 0.010087000206112862,
+      "learning_rate": 0.0003234567901234568,
+      "loss": 0.862,
+      "step": 1219
+    },
+    {
+      "epoch": 2.710358233823938,
+      "grad_norm": 0.007342969998717308,
+      "learning_rate": 0.00032098765432098765,
+      "loss": 0.8919,
+      "step": 1220
+    },
+    {
+      "epoch": 2.7125798389336295,
+      "grad_norm": 0.006463713478296995,
+      "learning_rate": 0.00031851851851851854,
+      "loss": 0.7057,
+      "step": 1221
+    },
+    {
+      "epoch": 2.7148014440433212,
+      "grad_norm": 0.008394156582653522,
+      "learning_rate": 0.0003160493827160494,
+      "loss": 0.7823,
+      "step": 1222
+    },
+    {
+      "epoch": 2.717023049153013,
+      "grad_norm": 0.007742198649793863,
+      "learning_rate": 0.0003135802469135803,
+      "loss": 0.7605,
+      "step": 1223
+    },
+    {
+      "epoch": 2.7192446542627047,
+      "grad_norm": 0.008673381991684437,
+      "learning_rate": 0.0003111111111111111,
+      "loss": 0.7759,
+      "step": 1224
+    },
+    {
+      "epoch": 2.7214662593723964,
+      "grad_norm": 0.00807158648967743,
+      "learning_rate": 0.00030864197530864197,
+      "loss": 0.8065,
+      "step": 1225
+    },
+    {
+      "epoch": 2.7236878644820885,
+      "grad_norm": 0.006811562459915876,
+      "learning_rate": 0.0003061728395061728,
+      "loss": 0.7656,
+      "step": 1226
+    },
+    {
+      "epoch": 2.7259094695917803,
+      "grad_norm": 0.015771763399243355,
+      "learning_rate": 0.0003037037037037037,
+      "loss": 0.8297,
+      "step": 1227
+    },
+    {
+      "epoch": 2.728131074701472,
+      "grad_norm": 0.011830280534923077,
+      "learning_rate": 0.00030123456790123455,
+      "loss": 0.7387,
+      "step": 1228
+    },
+    {
+      "epoch": 2.7303526798111637,
+      "grad_norm": 0.008170777931809425,
+      "learning_rate": 0.00029876543209876545,
+      "loss": 0.6999,
+      "step": 1229
+    },
+    {
+      "epoch": 2.7325742849208554,
+      "grad_norm": 0.028427088633179665,
+      "learning_rate": 0.0002962962962962963,
+      "loss": 0.7808,
+      "step": 1230
+    },
+    {
+      "epoch": 2.734795890030547,
+      "grad_norm": 0.0069966730661690235,
+      "learning_rate": 0.0002938271604938272,
+      "loss": 0.8691,
+      "step": 1231
+    },
+    {
+      "epoch": 2.737017495140239,
+      "grad_norm": 0.02192625217139721,
+      "learning_rate": 0.00029135802469135803,
+      "loss": 0.7229,
+      "step": 1232
+    },
+    {
+      "epoch": 2.7392391002499306,
+      "grad_norm": 0.0061951447278261185,
+      "learning_rate": 0.0002888888888888889,
+      "loss": 0.5309,
+      "step": 1233
+    },
+    {
+      "epoch": 2.7414607053596223,
+      "grad_norm": 0.005088411271572113,
+      "learning_rate": 0.00028641975308641977,
+      "loss": 0.8791,
+      "step": 1234
+    },
+    {
+      "epoch": 2.743682310469314,
+      "grad_norm": 0.011514096520841122,
+      "learning_rate": 0.0002839506172839506,
+      "loss": 0.6298,
+      "step": 1235
+    },
+    {
+      "epoch": 2.7459039155790057,
+      "grad_norm": 0.00723766814917326,
+      "learning_rate": 0.0002814814814814815,
+      "loss": 0.7606,
+      "step": 1236
+    },
+    {
+      "epoch": 2.7481255206886974,
+      "grad_norm": 0.009152224287390709,
+      "learning_rate": 0.00027901234567901236,
+      "loss": 0.8691,
+      "step": 1237
+    },
+    {
+      "epoch": 2.750347125798389,
+      "grad_norm": 0.010683218017220497,
+      "learning_rate": 0.00027654320987654325,
+      "loss": 0.7259,
+      "step": 1238
+    },
+    {
+      "epoch": 2.752568730908081,
+      "grad_norm": 0.014296338893473148,
+      "learning_rate": 0.0002740740740740741,
+      "loss": 0.6487,
+      "step": 1239
+    },
+    {
+      "epoch": 2.7547903360177726,
+      "grad_norm": 0.017391860485076904,
+      "learning_rate": 0.00027160493827160494,
+      "loss": 0.642,
+      "step": 1240
+    },
+    {
+      "epoch": 2.7570119411274647,
+      "grad_norm": 0.007915521040558815,
+      "learning_rate": 0.0002691358024691358,
+      "loss": 0.9457,
+      "step": 1241
+    },
+    {
+      "epoch": 2.7592335462371564,
+      "grad_norm": 0.01041811890900135,
+      "learning_rate": 0.0002666666666666667,
+      "loss": 0.7203,
+      "step": 1242
+    },
+    {
+      "epoch": 2.761455151346848,
+      "grad_norm": 0.01122908852994442,
+      "learning_rate": 0.0002641975308641975,
+      "loss": 0.911,
+      "step": 1243
+    },
+    {
+      "epoch": 2.76367675645654,
+      "grad_norm": 0.011805608868598938,
+      "learning_rate": 0.0002617283950617284,
+      "loss": 0.7091,
+      "step": 1244
+    },
+    {
+      "epoch": 2.7658983615662316,
+      "grad_norm": 0.008046625182032585,
+      "learning_rate": 0.00025925925925925926,
+      "loss": 0.6665,
+      "step": 1245
+    },
+    {
+      "epoch": 2.7681199666759233,
+      "grad_norm": 0.008950291201472282,
+      "learning_rate": 0.00025679012345679016,
+      "loss": 0.5604,
+      "step": 1246
+    },
+    {
+      "epoch": 2.770341571785615,
+      "grad_norm": 0.009476016275584698,
+      "learning_rate": 0.000254320987654321,
+      "loss": 0.6785,
+      "step": 1247
+    },
+    {
+      "epoch": 2.7725631768953067,
+      "grad_norm": 0.007629558444023132,
+      "learning_rate": 0.00025185185185185185,
+      "loss": 0.9354,
+      "step": 1248
+    },
+    {
+      "epoch": 2.7747847820049985,
+      "grad_norm": 0.011383921839296818,
+      "learning_rate": 0.0002493827160493827,
+      "loss": 0.7957,
+      "step": 1249
+    },
+    {
+      "epoch": 2.7770063871146906,
+      "grad_norm": 0.009743484668433666,
+      "learning_rate": 0.0002469135802469136,
+      "loss": 0.7978,
+      "step": 1250
+    },
+    {
+      "epoch": 2.7792279922243823,
+      "grad_norm": 0.008847765624523163,
+      "learning_rate": 0.0002444444444444445,
+      "loss": 0.9675,
+      "step": 1251
+    },
+    {
+      "epoch": 2.781449597334074,
+      "grad_norm": 0.007048389874398708,
+      "learning_rate": 0.0002419753086419753,
+      "loss": 0.8923,
+      "step": 1252
+    },
+    {
+      "epoch": 2.7836712024437658,
+      "grad_norm": 0.011864888481795788,
+      "learning_rate": 0.0002395061728395062,
+      "loss": 0.8361,
+      "step": 1253
+    },
+    {
+      "epoch": 2.7858928075534575,
+      "grad_norm": 0.007807235699146986,
+      "learning_rate": 0.00023703703703703704,
+      "loss": 0.8768,
+      "step": 1254
+    },
+    {
+      "epoch": 2.788114412663149,
+      "grad_norm": 0.006801323965191841,
+      "learning_rate": 0.0002345679012345679,
+      "loss": 0.7097,
+      "step": 1255
+    },
+    {
+      "epoch": 2.790336017772841,
+      "grad_norm": 0.005747576244175434,
+      "learning_rate": 0.00023209876543209875,
+      "loss": 0.6689,
+      "step": 1256
+    },
+    {
+      "epoch": 2.7925576228825326,
+      "grad_norm": 0.007229214999824762,
+      "learning_rate": 0.00022962962962962965,
+      "loss": 0.7671,
+      "step": 1257
+    },
+    {
+      "epoch": 2.7947792279922243,
+      "grad_norm": 0.006883475463837385,
+      "learning_rate": 0.0002271604938271605,
+      "loss": 0.7366,
+      "step": 1258
+    },
+    {
+      "epoch": 2.797000833101916,
+      "grad_norm": 0.009386280551552773,
+      "learning_rate": 0.00022469135802469136,
+      "loss": 0.7915,
+      "step": 1259
+    },
+    {
+      "epoch": 2.7992224382116078,
+      "grad_norm": 0.010489873588085175,
+      "learning_rate": 0.0002222222222222222,
+      "loss": 0.9666,
+      "step": 1260
+    },
+    {
+      "epoch": 2.8014440433212995,
+      "grad_norm": 0.009199917316436768,
+      "learning_rate": 0.0002197530864197531,
+      "loss": 0.8122,
+      "step": 1261
+    },
+    {
+      "epoch": 2.803665648430991,
+      "grad_norm": 0.00824709516018629,
+      "learning_rate": 0.00021728395061728395,
+      "loss": 0.9059,
+      "step": 1262
+    },
+    {
+      "epoch": 2.805887253540683,
+      "grad_norm": 0.008696146309375763,
+      "learning_rate": 0.00021481481481481482,
+      "loss": 0.6446,
+      "step": 1263
+    },
+    {
+      "epoch": 2.8081088586503746,
+      "grad_norm": 0.008835419081151485,
+      "learning_rate": 0.00021234567901234566,
+      "loss": 1.0159,
+      "step": 1264
+    },
+    {
+      "epoch": 2.810330463760067,
+      "grad_norm": 0.010744646191596985,
+      "learning_rate": 0.00020987654320987656,
+      "loss": 0.8361,
+      "step": 1265
+    },
+    {
+      "epoch": 2.8125520688697585,
+      "grad_norm": 0.015377935022115707,
+      "learning_rate": 0.0002074074074074074,
+      "loss": 0.8772,
+      "step": 1266
+    },
+    {
+      "epoch": 2.81477367397945,
+      "grad_norm": 0.008018149994313717,
+      "learning_rate": 0.00020493827160493827,
+      "loss": 0.8452,
+      "step": 1267
+    },
+    {
+      "epoch": 2.816995279089142,
+      "grad_norm": 0.008665296249091625,
+      "learning_rate": 0.00020246913580246917,
+      "loss": 0.7223,
+      "step": 1268
+    },
+    {
+      "epoch": 2.8192168841988337,
+      "grad_norm": 0.011783248744904995,
+      "learning_rate": 0.0002,
+      "loss": 0.9633,
+      "step": 1269
+    },
+    {
+      "epoch": 2.8214384893085254,
+      "grad_norm": 0.00910250935703516,
+      "learning_rate": 0.00019753086419753088,
+      "loss": 0.6806,
+      "step": 1270
+    },
+    {
+      "epoch": 2.823660094418217,
+      "grad_norm": 0.006656398996710777,
+      "learning_rate": 0.00019506172839506172,
+      "loss": 0.837,
+      "step": 1271
+    },
+    {
+      "epoch": 2.825881699527909,
+      "grad_norm": 0.010195860639214516,
+      "learning_rate": 0.00019259259259259262,
+      "loss": 0.6618,
+      "step": 1272
+    },
+    {
+      "epoch": 2.8281033046376005,
+      "grad_norm": 0.007249714806675911,
+      "learning_rate": 0.00019012345679012346,
+      "loss": 0.7854,
+      "step": 1273
+    },
+    {
+      "epoch": 2.8303249097472927,
+      "grad_norm": 0.010114341974258423,
+      "learning_rate": 0.00018765432098765433,
+      "loss": 0.5872,
+      "step": 1274
+    },
+    {
+      "epoch": 2.8325465148569844,
+      "grad_norm": 0.006100023165345192,
+      "learning_rate": 0.00018518518518518518,
+      "loss": 0.7327,
+      "step": 1275
+    },
+    {
+      "epoch": 2.834768119966676,
+      "grad_norm": 0.007089332211762667,
+      "learning_rate": 0.00018271604938271607,
+      "loss": 0.7358,
+      "step": 1276
+    },
+    {
+      "epoch": 2.836989725076368,
+      "grad_norm": 0.007861738093197346,
+      "learning_rate": 0.00018024691358024692,
+      "loss": 0.7768,
+      "step": 1277
+    },
+    {
+      "epoch": 2.8392113301860595,
+      "grad_norm": 0.007813625037670135,
+      "learning_rate": 0.00017777777777777779,
+      "loss": 0.8741,
+      "step": 1278
+    },
+    {
+      "epoch": 2.8414329352957512,
+      "grad_norm": 0.007935810834169388,
+      "learning_rate": 0.00017530864197530866,
+      "loss": 0.6432,
+      "step": 1279
+    },
+    {
+      "epoch": 2.843654540405443,
+      "grad_norm": 0.01766696572303772,
+      "learning_rate": 0.00017283950617283953,
+      "loss": 0.7577,
+      "step": 1280
+    },
+    {
+      "epoch": 2.8458761455151347,
+      "grad_norm": 0.0071344939060509205,
+      "learning_rate": 0.00017037037037037037,
+      "loss": 0.8434,
+      "step": 1281
+    },
+    {
+      "epoch": 2.8480977506248264,
+      "grad_norm": 0.007316224742680788,
+      "learning_rate": 0.00016790123456790124,
+      "loss": 0.6397,
+      "step": 1282
+    },
+    {
+      "epoch": 2.850319355734518,
+      "grad_norm": 0.008054982870817184,
+      "learning_rate": 0.0001654320987654321,
+      "loss": 0.6832,
+      "step": 1283
+    },
+    {
+      "epoch": 2.85254096084421,
+      "grad_norm": 0.01453061867505312,
+      "learning_rate": 0.00016296296296296295,
+      "loss": 0.773,
+      "step": 1284
+    },
+    {
+      "epoch": 2.8547625659539015,
+      "grad_norm": 0.008884952403604984,
+      "learning_rate": 0.00016049382716049382,
+      "loss": 0.8363,
+      "step": 1285
+    },
+    {
+      "epoch": 2.8569841710635933,
+      "grad_norm": 0.007511654868721962,
+      "learning_rate": 0.0001580246913580247,
+      "loss": 0.9314,
+      "step": 1286
+    },
+    {
+      "epoch": 2.859205776173285,
+      "grad_norm": 0.008190833032131195,
+      "learning_rate": 0.00015555555555555556,
+      "loss": 0.7633,
+      "step": 1287
+    },
+    {
+      "epoch": 2.8614273812829767,
+      "grad_norm": 0.006483402568846941,
+      "learning_rate": 0.0001530864197530864,
+      "loss": 0.9713,
+      "step": 1288
+    },
+    {
+      "epoch": 2.863648986392669,
+      "grad_norm": 0.007926386781036854,
+      "learning_rate": 0.00015061728395061728,
+      "loss": 0.9167,
+      "step": 1289
+    },
+    {
+      "epoch": 2.8658705915023606,
+      "grad_norm": 0.0071045164950191975,
+      "learning_rate": 0.00014814814814814815,
+      "loss": 0.7618,
+      "step": 1290
+    },
+    {
+      "epoch": 2.8680921966120523,
+      "grad_norm": 0.0076851435005664825,
+      "learning_rate": 0.00014567901234567902,
+      "loss": 0.7558,
+      "step": 1291
+    },
+    {
+      "epoch": 2.870313801721744,
+      "grad_norm": 0.006890943273901939,
+      "learning_rate": 0.00014320987654320989,
+      "loss": 0.883,
+      "step": 1292
+    },
+    {
+      "epoch": 2.8725354068314357,
+      "grad_norm": 0.007618566509336233,
+      "learning_rate": 0.00014074074074074076,
+      "loss": 0.9588,
+      "step": 1293
+    },
+    {
+      "epoch": 2.8747570119411274,
+      "grad_norm": 0.010207663290202618,
+      "learning_rate": 0.00013827160493827163,
+      "loss": 1.0954,
+      "step": 1294
+    },
+    {
+      "epoch": 2.876978617050819,
+      "grad_norm": 0.01138080283999443,
+      "learning_rate": 0.00013580246913580247,
+      "loss": 0.7984,
+      "step": 1295
+    },
+    {
+      "epoch": 2.879200222160511,
+      "grad_norm": 0.006852086633443832,
+      "learning_rate": 0.00013333333333333334,
+      "loss": 0.647,
+      "step": 1296
+    },
+    {
+      "epoch": 2.8814218272702026,
+      "grad_norm": 0.006972672417759895,
+      "learning_rate": 0.0001308641975308642,
+      "loss": 0.7138,
+      "step": 1297
+    },
+    {
+      "epoch": 2.8836434323798947,
+      "grad_norm": 0.008358401246368885,
+      "learning_rate": 0.00012839506172839508,
+      "loss": 0.6586,
+      "step": 1298
+    },
+    {
+      "epoch": 2.8858650374895864,
+      "grad_norm": 0.010818328708410263,
+      "learning_rate": 0.00012592592592592592,
+      "loss": 0.729,
+      "step": 1299
+    },
+    {
+      "epoch": 2.888086642599278,
+      "grad_norm": 0.007076184265315533,
+      "learning_rate": 0.0001234567901234568,
+      "loss": 0.6661,
+      "step": 1300
+    },
+    {
+      "epoch": 2.89030824770897,
+      "grad_norm": 0.009897168725728989,
+      "learning_rate": 0.00012098765432098765,
+      "loss": 1.1523,
+      "step": 1301
+    },
+    {
+      "epoch": 2.8925298528186616,
+      "grad_norm": 0.009953767992556095,
+      "learning_rate": 0.00011851851851851852,
+      "loss": 0.7484,
+      "step": 1302
+    },
+    {
+      "epoch": 2.8947514579283533,
+      "grad_norm": 0.010332675650715828,
+      "learning_rate": 0.00011604938271604938,
+      "loss": 0.6528,
+      "step": 1303
+    },
+    {
+      "epoch": 2.896973063038045,
+      "grad_norm": 0.006213752087205648,
+      "learning_rate": 0.00011358024691358025,
+      "loss": 0.5496,
+      "step": 1304
+    },
+    {
+      "epoch": 2.8991946681477367,
+      "grad_norm": 0.005716019310057163,
+      "learning_rate": 0.0001111111111111111,
+      "loss": 0.7742,
+      "step": 1305
+    },
+    {
+      "epoch": 2.9014162732574285,
+      "grad_norm": 0.007558442186564207,
+      "learning_rate": 0.00010864197530864197,
+      "loss": 0.7281,
+      "step": 1306
+    },
+    {
+      "epoch": 2.90363787836712,
+      "grad_norm": 0.009284201078116894,
+      "learning_rate": 0.00010617283950617283,
+      "loss": 0.7379,
+      "step": 1307
+    },
+    {
+      "epoch": 2.905859483476812,
+      "grad_norm": 0.00817166455090046,
+      "learning_rate": 0.0001037037037037037,
+      "loss": 0.8667,
+      "step": 1308
+    },
+    {
+      "epoch": 2.9080810885865036,
+      "grad_norm": 0.011280300095677376,
+      "learning_rate": 0.00010123456790123458,
+      "loss": 0.8758,
+      "step": 1309
+    },
+    {
+      "epoch": 2.9103026936961953,
+      "grad_norm": 0.006844306364655495,
+      "learning_rate": 9.876543209876544e-05,
+      "loss": 0.7103,
+      "step": 1310
+    },
+    {
+      "epoch": 2.912524298805887,
+      "grad_norm": 0.005340827163308859,
+      "learning_rate": 9.629629629629631e-05,
+      "loss": 0.6803,
+      "step": 1311
+    },
+    {
+      "epoch": 2.9147459039155788,
+      "grad_norm": 0.006157919764518738,
+      "learning_rate": 9.382716049382717e-05,
+      "loss": 0.8298,
+      "step": 1312
+    },
+    {
+      "epoch": 2.916967509025271,
+      "grad_norm": 0.019693652167916298,
+      "learning_rate": 9.135802469135804e-05,
+      "loss": 0.8528,
+      "step": 1313
+    },
+    {
+      "epoch": 2.9191891141349626,
+      "grad_norm": 0.008847893215715885,
+      "learning_rate": 8.888888888888889e-05,
+      "loss": 0.9522,
+      "step": 1314
+    },
+    {
+      "epoch": 2.9214107192446543,
+      "grad_norm": 0.03038479946553707,
+      "learning_rate": 8.641975308641976e-05,
+      "loss": 0.7119,
+      "step": 1315
+    },
+    {
+      "epoch": 2.923632324354346,
+      "grad_norm": 0.009875786490738392,
+      "learning_rate": 8.395061728395062e-05,
+      "loss": 0.5868,
+      "step": 1316
+    },
+    {
+      "epoch": 2.9258539294640378,
+      "grad_norm": 0.007000662852078676,
+      "learning_rate": 8.148148148148148e-05,
+      "loss": 0.8366,
+      "step": 1317
+    },
+    {
+      "epoch": 2.9280755345737295,
+      "grad_norm": 0.009017952717840672,
+      "learning_rate": 7.901234567901235e-05,
+      "loss": 0.856,
+      "step": 1318
+    },
+    {
+      "epoch": 2.930297139683421,
+      "grad_norm": 0.009106653742492199,
+      "learning_rate": 7.65432098765432e-05,
+      "loss": 0.844,
+      "step": 1319
+    },
+    {
+      "epoch": 2.932518744793113,
+      "grad_norm": 0.004958700388669968,
+      "learning_rate": 7.407407407407407e-05,
+      "loss": 0.7263,
+      "step": 1320
+    },
+    {
+      "epoch": 2.9347403499028046,
+      "grad_norm": 0.014845774509012699,
+      "learning_rate": 7.160493827160494e-05,
+      "loss": 0.7577,
+      "step": 1321
+    },
+    {
+      "epoch": 2.936961955012497,
+      "grad_norm": 0.008830921724438667,
+      "learning_rate": 6.913580246913581e-05,
+      "loss": 0.6669,
+      "step": 1322
+    },
+    {
+      "epoch": 2.9391835601221885,
+      "grad_norm": 0.009541644714772701,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 0.7751,
+      "step": 1323
+    },
+    {
+      "epoch": 2.9414051652318802,
+      "grad_norm": 0.008899768814444542,
+      "learning_rate": 6.419753086419754e-05,
+      "loss": 0.6798,
+      "step": 1324
+    },
+    {
+      "epoch": 2.943626770341572,
+      "grad_norm": 0.012478734366595745,
+      "learning_rate": 6.17283950617284e-05,
+      "loss": 0.8341,
+      "step": 1325
+    },
+    {
+      "epoch": 2.9458483754512637,
+      "grad_norm": 0.008129597641527653,
+      "learning_rate": 5.925925925925926e-05,
+      "loss": 0.8493,
+      "step": 1326
+    },
+    {
+      "epoch": 2.9480699805609554,
+      "grad_norm": 0.01137120183557272,
+      "learning_rate": 5.679012345679012e-05,
+      "loss": 0.7003,
+      "step": 1327
+    },
+    {
+      "epoch": 2.950291585670647,
+      "grad_norm": 0.0069569917395710945,
+      "learning_rate": 5.4320987654320986e-05,
+      "loss": 0.7322,
+      "step": 1328
+    },
+    {
+      "epoch": 2.952513190780339,
+      "grad_norm": 0.005949252285063267,
+      "learning_rate": 5.185185185185185e-05,
+      "loss": 0.8372,
+      "step": 1329
+    },
+    {
+      "epoch": 2.9547347958900305,
+      "grad_norm": 0.010746531188488007,
+      "learning_rate": 4.938271604938272e-05,
+      "loss": 0.6656,
+      "step": 1330
+    },
+    {
+      "epoch": 2.9569564009997222,
+      "grad_norm": 0.007687109522521496,
+      "learning_rate": 4.691358024691358e-05,
+      "loss": 0.8294,
+      "step": 1331
+    },
+    {
+      "epoch": 2.959178006109414,
+      "grad_norm": 0.008320269174873829,
+      "learning_rate": 4.4444444444444447e-05,
+      "loss": 0.8337,
+      "step": 1332
+    },
+    {
+      "epoch": 2.9613996112191057,
+      "grad_norm": 0.009467615745961666,
+      "learning_rate": 4.197530864197531e-05,
+      "loss": 0.8912,
+      "step": 1333
+    },
+    {
+      "epoch": 2.9636212163287974,
+      "grad_norm": 0.0069550564512610435,
+      "learning_rate": 3.950617283950617e-05,
+      "loss": 0.8289,
+      "step": 1334
+    },
+    {
+      "epoch": 2.965842821438489,
+      "grad_norm": 0.0068250177428126335,
+      "learning_rate": 3.7037037037037037e-05,
+      "loss": 0.6434,
+      "step": 1335
+    },
+    {
+      "epoch": 2.968064426548181,
+      "grad_norm": 0.005152587778866291,
+      "learning_rate": 3.456790123456791e-05,
+      "loss": 0.6858,
+      "step": 1336
+    },
+    {
+      "epoch": 2.970286031657873,
+      "grad_norm": 0.009315388277173042,
+      "learning_rate": 3.209876543209877e-05,
+      "loss": 0.6759,
+      "step": 1337
+    },
+    {
+      "epoch": 2.9725076367675647,
+      "grad_norm": 0.009328274056315422,
+      "learning_rate": 2.962962962962963e-05,
+      "loss": 0.7544,
+      "step": 1338
+    },
+    {
+      "epoch": 2.9747292418772564,
+      "grad_norm": 0.00911303423345089,
+      "learning_rate": 2.7160493827160493e-05,
+      "loss": 0.8303,
+      "step": 1339
+    },
+    {
+      "epoch": 2.976950846986948,
+      "grad_norm": 0.008570064790546894,
+      "learning_rate": 2.469135802469136e-05,
+      "loss": 0.7677,
+      "step": 1340
+    },
+    {
+      "epoch": 2.97917245209664,
+      "grad_norm": 0.007688554469496012,
+      "learning_rate": 2.2222222222222223e-05,
+      "loss": 0.6908,
+      "step": 1341
+    },
+    {
+      "epoch": 2.9813940572063315,
+      "grad_norm": 0.009369606152176857,
+      "learning_rate": 1.9753086419753087e-05,
+      "loss": 0.8807,
+      "step": 1342
+    },
+    {
+      "epoch": 2.9836156623160233,
+      "grad_norm": 0.009632132947444916,
+      "learning_rate": 1.7283950617283953e-05,
+      "loss": 0.7748,
+      "step": 1343
+    },
+    {
+      "epoch": 2.985837267425715,
+      "grad_norm": 0.008006836287677288,
+      "learning_rate": 1.4814814814814815e-05,
+      "loss": 0.8687,
+      "step": 1344
+    },
+    {
+      "epoch": 2.9880588725354067,
+      "grad_norm": 0.0064094155095517635,
+      "learning_rate": 1.234567901234568e-05,
+      "loss": 0.7026,
+      "step": 1345
+    },
+    {
+      "epoch": 2.990280477645099,
+      "grad_norm": 0.007208329625427723,
+      "learning_rate": 9.876543209876543e-06,
+      "loss": 0.7213,
+      "step": 1346
+    },
+    {
+      "epoch": 2.9925020827547906,
+      "grad_norm": 0.006725262384861708,
+      "learning_rate": 7.4074074074074075e-06,
+      "loss": 0.6405,
+      "step": 1347
+    },
+    {
+      "epoch": 2.9947236878644823,
+      "grad_norm": 0.009039103053510189,
+      "learning_rate": 4.938271604938272e-06,
+      "loss": 0.5949,
+      "step": 1348
+    },
+    {
+      "epoch": 2.996945292974174,
+      "grad_norm": 0.007107519078999758,
+      "learning_rate": 2.469135802469136e-06,
+      "loss": 0.7717,
+      "step": 1349
+    },
+    {
+      "epoch": 2.9991668980838657,
+      "grad_norm": 0.008029144257307053,
+      "learning_rate": 0.0,
+      "loss": 0.6898,
+      "step": 1350
+    },
+    {
+      "epoch": 2.9991668980838657,
+      "step": 1350,
+      "total_flos": 2.317351723938611e+16,
+      "train_loss": 0.7935600814112911,
+      "train_runtime": 1657.45,
+      "train_samples_per_second": 13.036,
+      "train_steps_per_second": 0.815
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1350,
+  "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": 2.317351723938611e+16,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}