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": 0.002830490469932556,
+      "learning_rate": 2.2222222222222223e-05,
+      "loss": 0.6744,
+      "step": 1
+    },
+    {
+      "epoch": 0.004443210219383505,
+      "grad_norm": 0.0013051991118118167,
+      "learning_rate": 4.4444444444444447e-05,
+      "loss": 0.5043,
+      "step": 2
+    },
+    {
+      "epoch": 0.006664815329075257,
+      "grad_norm": 0.0013274818193167448,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 0.7621,
+      "step": 3
+    },
+    {
+      "epoch": 0.00888642043876701,
+      "grad_norm": 0.0017257591243833303,
+      "learning_rate": 8.888888888888889e-05,
+      "loss": 0.6256,
+      "step": 4
+    },
+    {
+      "epoch": 0.011108025548458762,
+      "grad_norm": 0.001670735189691186,
+      "learning_rate": 0.0001111111111111111,
+      "loss": 0.5722,
+      "step": 5
+    },
+    {
+      "epoch": 0.013329630658150514,
+      "grad_norm": 0.0013650484615936875,
+      "learning_rate": 0.00013333333333333334,
+      "loss": 0.7089,
+      "step": 6
+    },
+    {
+      "epoch": 0.015551235767842266,
+      "grad_norm": 0.0022685532458126545,
+      "learning_rate": 0.00015555555555555556,
+      "loss": 0.6669,
+      "step": 7
+    },
+    {
+      "epoch": 0.01777284087753402,
+      "grad_norm": 0.0012487670173868537,
+      "learning_rate": 0.00017777777777777779,
+      "loss": 0.6834,
+      "step": 8
+    },
+    {
+      "epoch": 0.01999444598722577,
+      "grad_norm": 0.0018260080832988024,
+      "learning_rate": 0.0002,
+      "loss": 0.7538,
+      "step": 9
+    },
+    {
+      "epoch": 0.022216051096917523,
+      "grad_norm": 0.0013124588876962662,
+      "learning_rate": 0.0002222222222222222,
+      "loss": 0.6298,
+      "step": 10
+    },
+    {
+      "epoch": 0.024437656206609277,
+      "grad_norm": 0.0009908919455483556,
+      "learning_rate": 0.0002444444444444445,
+      "loss": 0.5524,
+      "step": 11
+    },
+    {
+      "epoch": 0.026659261316301027,
+      "grad_norm": 0.0016819137381389737,
+      "learning_rate": 0.0002666666666666667,
+      "loss": 0.6242,
+      "step": 12
+    },
+    {
+      "epoch": 0.02888086642599278,
+      "grad_norm": 0.0014569928171113133,
+      "learning_rate": 0.0002888888888888889,
+      "loss": 0.5705,
+      "step": 13
+    },
+    {
+      "epoch": 0.03110247153568453,
+      "grad_norm": 0.002944767475128174,
+      "learning_rate": 0.0003111111111111111,
+      "loss": 0.7228,
+      "step": 14
+    },
+    {
+      "epoch": 0.03332407664537628,
+      "grad_norm": 0.001788086025044322,
+      "learning_rate": 0.0003333333333333333,
+      "loss": 0.7305,
+      "step": 15
+    },
+    {
+      "epoch": 0.03554568175506804,
+      "grad_norm": 0.0013934284215793014,
+      "learning_rate": 0.00035555555555555557,
+      "loss": 0.6049,
+      "step": 16
+    },
+    {
+      "epoch": 0.03776728686475979,
+      "grad_norm": 0.0018239800119772553,
+      "learning_rate": 0.00037777777777777777,
+      "loss": 0.6171,
+      "step": 17
+    },
+    {
+      "epoch": 0.03998889197445154,
+      "grad_norm": 0.0020568512845784426,
+      "learning_rate": 0.0004,
+      "loss": 0.7091,
+      "step": 18
+    },
+    {
+      "epoch": 0.042210497084143296,
+      "grad_norm": 0.0022871694527566433,
+      "learning_rate": 0.00042222222222222227,
+      "loss": 0.8162,
+      "step": 19
+    },
+    {
+      "epoch": 0.044432102193835046,
+      "grad_norm": 0.002156606176868081,
+      "learning_rate": 0.0004444444444444444,
+      "loss": 0.7374,
+      "step": 20
+    },
+    {
+      "epoch": 0.0466537073035268,
+      "grad_norm": 0.001277489005587995,
+      "learning_rate": 0.00046666666666666666,
+      "loss": 0.5195,
+      "step": 21
+    },
+    {
+      "epoch": 0.048875312413218554,
+      "grad_norm": 0.000715097296051681,
+      "learning_rate": 0.000488888888888889,
+      "loss": 0.5719,
+      "step": 22
+    },
+    {
+      "epoch": 0.051096917522910304,
+      "grad_norm": 0.0021112209651619196,
+      "learning_rate": 0.0005111111111111112,
+      "loss": 0.6308,
+      "step": 23
+    },
+    {
+      "epoch": 0.053318522632602054,
+      "grad_norm": 0.0011713097337633371,
+      "learning_rate": 0.0005333333333333334,
+      "loss": 0.6671,
+      "step": 24
+    },
+    {
+      "epoch": 0.055540127742293804,
+      "grad_norm": 0.0019275946542620659,
+      "learning_rate": 0.0005555555555555556,
+      "loss": 0.6584,
+      "step": 25
+    },
+    {
+      "epoch": 0.05776173285198556,
+      "grad_norm": 0.0022250341717153788,
+      "learning_rate": 0.0005777777777777778,
+      "loss": 0.789,
+      "step": 26
+    },
+    {
+      "epoch": 0.05998333796167731,
+      "grad_norm": 0.0013691845815628767,
+      "learning_rate": 0.0006000000000000001,
+      "loss": 0.6083,
+      "step": 27
+    },
+    {
+      "epoch": 0.06220494307136906,
+      "grad_norm": 0.0016656300285831094,
+      "learning_rate": 0.0006222222222222223,
+      "loss": 0.7864,
+      "step": 28
+    },
+    {
+      "epoch": 0.06442654818106082,
+      "grad_norm": 0.001373667037114501,
+      "learning_rate": 0.0006444444444444444,
+      "loss": 0.7652,
+      "step": 29
+    },
+    {
+      "epoch": 0.06664815329075256,
+      "grad_norm": 0.0013376525603234768,
+      "learning_rate": 0.0006666666666666666,
+      "loss": 0.6543,
+      "step": 30
+    },
+    {
+      "epoch": 0.06886975840044432,
+      "grad_norm": 0.001769097289070487,
+      "learning_rate": 0.000688888888888889,
+      "loss": 0.6709,
+      "step": 31
+    },
+    {
+      "epoch": 0.07109136351013608,
+      "grad_norm": 0.002042358508333564,
+      "learning_rate": 0.0007111111111111111,
+      "loss": 0.7638,
+      "step": 32
+    },
+    {
+      "epoch": 0.07331296861982782,
+      "grad_norm": 0.0013311082730069757,
+      "learning_rate": 0.0007333333333333333,
+      "loss": 0.6428,
+      "step": 33
+    },
+    {
+      "epoch": 0.07553457372951958,
+      "grad_norm": 0.0014501510886475444,
+      "learning_rate": 0.0007555555555555555,
+      "loss": 0.7086,
+      "step": 34
+    },
+    {
+      "epoch": 0.07775617883921133,
+      "grad_norm": 0.0018582150805741549,
+      "learning_rate": 0.0007777777777777777,
+      "loss": 0.8002,
+      "step": 35
+    },
+    {
+      "epoch": 0.07997778394890308,
+      "grad_norm": 0.0015540813328698277,
+      "learning_rate": 0.0008,
+      "loss": 0.5421,
+      "step": 36
+    },
+    {
+      "epoch": 0.08219938905859484,
+      "grad_norm": 0.002326158108189702,
+      "learning_rate": 0.0008222222222222222,
+      "loss": 0.6808,
+      "step": 37
+    },
+    {
+      "epoch": 0.08442099416828659,
+      "grad_norm": 0.002218228066340089,
+      "learning_rate": 0.0008444444444444445,
+      "loss": 0.5874,
+      "step": 38
+    },
+    {
+      "epoch": 0.08664259927797834,
+      "grad_norm": 0.0012414696393534541,
+      "learning_rate": 0.0008666666666666666,
+      "loss": 0.7572,
+      "step": 39
+    },
+    {
+      "epoch": 0.08886420438767009,
+      "grad_norm": 0.0017535301158204675,
+      "learning_rate": 0.0008888888888888888,
+      "loss": 0.7115,
+      "step": 40
+    },
+    {
+      "epoch": 0.09108580949736185,
+      "grad_norm": 0.0016572384629398584,
+      "learning_rate": 0.0009111111111111111,
+      "loss": 0.7242,
+      "step": 41
+    },
+    {
+      "epoch": 0.0933074146070536,
+      "grad_norm": 0.0014017591020092368,
+      "learning_rate": 0.0009333333333333333,
+      "loss": 0.7907,
+      "step": 42
+    },
+    {
+      "epoch": 0.09552901971674535,
+      "grad_norm": 0.001873611006885767,
+      "learning_rate": 0.0009555555555555556,
+      "loss": 0.7169,
+      "step": 43
+    },
+    {
+      "epoch": 0.09775062482643711,
+      "grad_norm": 0.002249490935355425,
+      "learning_rate": 0.000977777777777778,
+      "loss": 0.651,
+      "step": 44
+    },
+    {
+      "epoch": 0.09997222993612885,
+      "grad_norm": 0.001455418299883604,
+      "learning_rate": 0.001,
+      "loss": 0.5291,
+      "step": 45
+    },
+    {
+      "epoch": 0.10219383504582061,
+      "grad_norm": 0.0013861401239410043,
+      "learning_rate": 0.0010222222222222223,
+      "loss": 0.7166,
+      "step": 46
+    },
+    {
+      "epoch": 0.10441544015551235,
+      "grad_norm": 0.0016446675872430205,
+      "learning_rate": 0.0010444444444444444,
+      "loss": 0.8524,
+      "step": 47
+    },
+    {
+      "epoch": 0.10663704526520411,
+      "grad_norm": 0.0018337472574785352,
+      "learning_rate": 0.0010666666666666667,
+      "loss": 0.6347,
+      "step": 48
+    },
+    {
+      "epoch": 0.10885865037489587,
+      "grad_norm": 0.0014166847104206681,
+      "learning_rate": 0.001088888888888889,
+      "loss": 0.6729,
+      "step": 49
+    },
+    {
+      "epoch": 0.11108025548458761,
+      "grad_norm": 0.002324986970052123,
+      "learning_rate": 0.0011111111111111111,
+      "loss": 0.5649,
+      "step": 50
+    },
+    {
+      "epoch": 0.11330186059427937,
+      "grad_norm": 0.0018736826023086905,
+      "learning_rate": 0.0011333333333333334,
+      "loss": 0.7072,
+      "step": 51
+    },
+    {
+      "epoch": 0.11552346570397112,
+      "grad_norm": 0.001425831695087254,
+      "learning_rate": 0.0011555555555555555,
+      "loss": 0.5661,
+      "step": 52
+    },
+    {
+      "epoch": 0.11774507081366287,
+      "grad_norm": 0.0012704561231657863,
+      "learning_rate": 0.0011777777777777778,
+      "loss": 0.65,
+      "step": 53
+    },
+    {
+      "epoch": 0.11996667592335462,
+      "grad_norm": 0.001298760762438178,
+      "learning_rate": 0.0012000000000000001,
+      "loss": 0.6336,
+      "step": 54
+    },
+    {
+      "epoch": 0.12218828103304638,
+      "grad_norm": 0.0010759654687717557,
+      "learning_rate": 0.0012222222222222222,
+      "loss": 0.5591,
+      "step": 55
+    },
+    {
+      "epoch": 0.12440988614273812,
+      "grad_norm": 0.0019087704131379724,
+      "learning_rate": 0.0012444444444444445,
+      "loss": 0.6595,
+      "step": 56
+    },
+    {
+      "epoch": 0.12663149125242987,
+      "grad_norm": 0.001623319461941719,
+      "learning_rate": 0.0012666666666666666,
+      "loss": 0.6856,
+      "step": 57
+    },
+    {
+      "epoch": 0.12885309636212164,
+      "grad_norm": 0.001485670218244195,
+      "learning_rate": 0.001288888888888889,
+      "loss": 0.5978,
+      "step": 58
+    },
+    {
+      "epoch": 0.13107470147181338,
+      "grad_norm": 0.001623534713871777,
+      "learning_rate": 0.0013111111111111112,
+      "loss": 0.5635,
+      "step": 59
+    },
+    {
+      "epoch": 0.13329630658150513,
+      "grad_norm": 0.001306689577177167,
+      "learning_rate": 0.0013333333333333333,
+      "loss": 0.7234,
+      "step": 60
+    },
+    {
+      "epoch": 0.1355179116911969,
+      "grad_norm": 0.002071897964924574,
+      "learning_rate": 0.0013555555555555556,
+      "loss": 0.7316,
+      "step": 61
+    },
+    {
+      "epoch": 0.13773951680088864,
+      "grad_norm": 0.0016734000528231263,
+      "learning_rate": 0.001377777777777778,
+      "loss": 0.6618,
+      "step": 62
+    },
+    {
+      "epoch": 0.13996112191058038,
+      "grad_norm": 0.0018130919197574258,
+      "learning_rate": 0.0014,
+      "loss": 0.6921,
+      "step": 63
+    },
+    {
+      "epoch": 0.14218272702027215,
+      "grad_norm": 0.0018388992175459862,
+      "learning_rate": 0.0014222222222222223,
+      "loss": 0.7907,
+      "step": 64
+    },
+    {
+      "epoch": 0.1444043321299639,
+      "grad_norm": 0.001330480445176363,
+      "learning_rate": 0.0014444444444444444,
+      "loss": 0.5721,
+      "step": 65
+    },
+    {
+      "epoch": 0.14662593723965564,
+      "grad_norm": 0.0015514479018747807,
+      "learning_rate": 0.0014666666666666667,
+      "loss": 0.7752,
+      "step": 66
+    },
+    {
+      "epoch": 0.1488475423493474,
+      "grad_norm": 0.0014846732374280691,
+      "learning_rate": 0.001488888888888889,
+      "loss": 0.9079,
+      "step": 67
+    },
+    {
+      "epoch": 0.15106914745903915,
+      "grad_norm": 0.0014325291849672794,
+      "learning_rate": 0.001511111111111111,
+      "loss": 0.7094,
+      "step": 68
+    },
+    {
+      "epoch": 0.1532907525687309,
+      "grad_norm": 0.0013538472121581435,
+      "learning_rate": 0.0015333333333333332,
+      "loss": 0.6732,
+      "step": 69
+    },
+    {
+      "epoch": 0.15551235767842267,
+      "grad_norm": 0.002166144549846649,
+      "learning_rate": 0.0015555555555555555,
+      "loss": 0.5317,
+      "step": 70
+    },
+    {
+      "epoch": 0.1577339627881144,
+      "grad_norm": 0.0011681554606184363,
+      "learning_rate": 0.0015777777777777778,
+      "loss": 0.5447,
+      "step": 71
+    },
+    {
+      "epoch": 0.15995556789780616,
+      "grad_norm": 0.0016300322022289038,
+      "learning_rate": 0.0016,
+      "loss": 0.6186,
+      "step": 72
+    },
+    {
+      "epoch": 0.16217717300749793,
+      "grad_norm": 0.0012133439304307103,
+      "learning_rate": 0.0016222222222222222,
+      "loss": 0.6517,
+      "step": 73
+    },
+    {
+      "epoch": 0.16439877811718967,
+      "grad_norm": 0.0017300712643191218,
+      "learning_rate": 0.0016444444444444445,
+      "loss": 0.6562,
+      "step": 74
+    },
+    {
+      "epoch": 0.1666203832268814,
+      "grad_norm": 0.0012843969743698835,
+      "learning_rate": 0.0016666666666666668,
+      "loss": 0.6749,
+      "step": 75
+    },
+    {
+      "epoch": 0.16884198833657318,
+      "grad_norm": 0.001517774653621018,
+      "learning_rate": 0.001688888888888889,
+      "loss": 0.6396,
+      "step": 76
+    },
+    {
+      "epoch": 0.17106359344626493,
+      "grad_norm": 0.0013772399397566915,
+      "learning_rate": 0.0017111111111111114,
+      "loss": 0.5603,
+      "step": 77
+    },
+    {
+      "epoch": 0.17328519855595667,
+      "grad_norm": 0.0013864693464711308,
+      "learning_rate": 0.0017333333333333333,
+      "loss": 0.5805,
+      "step": 78
+    },
+    {
+      "epoch": 0.17550680366564844,
+      "grad_norm": 0.0018722772365435958,
+      "learning_rate": 0.0017555555555555556,
+      "loss": 0.7825,
+      "step": 79
+    },
+    {
+      "epoch": 0.17772840877534019,
+      "grad_norm": 0.0014643195318058133,
+      "learning_rate": 0.0017777777777777776,
+      "loss": 0.6691,
+      "step": 80
+    },
+    {
+      "epoch": 0.17995001388503193,
+      "grad_norm": 0.001783651881851256,
+      "learning_rate": 0.0018,
+      "loss": 0.6402,
+      "step": 81
+    },
+    {
+      "epoch": 0.1821716189947237,
+      "grad_norm": 0.001158166560344398,
+      "learning_rate": 0.0018222222222222223,
+      "loss": 0.555,
+      "step": 82
+    },
+    {
+      "epoch": 0.18439322410441544,
+      "grad_norm": 0.0013353719841688871,
+      "learning_rate": 0.0018444444444444446,
+      "loss": 0.6524,
+      "step": 83
+    },
+    {
+      "epoch": 0.1866148292141072,
+      "grad_norm": 0.002519479487091303,
+      "learning_rate": 0.0018666666666666666,
+      "loss": 0.6841,
+      "step": 84
+    },
+    {
+      "epoch": 0.18883643432379896,
+      "grad_norm": 0.0019098934717476368,
+      "learning_rate": 0.001888888888888889,
+      "loss": 0.5673,
+      "step": 85
+    },
+    {
+      "epoch": 0.1910580394334907,
+      "grad_norm": 0.0013814260018989444,
+      "learning_rate": 0.0019111111111111113,
+      "loss": 0.7098,
+      "step": 86
+    },
+    {
+      "epoch": 0.19327964454318244,
+      "grad_norm": 0.0016817323630675673,
+      "learning_rate": 0.0019333333333333336,
+      "loss": 0.6082,
+      "step": 87
+    },
+    {
+      "epoch": 0.19550124965287421,
+      "grad_norm": 0.0018059983849525452,
+      "learning_rate": 0.001955555555555556,
+      "loss": 0.5896,
+      "step": 88
+    },
+    {
+      "epoch": 0.19772285476256596,
+      "grad_norm": 0.0019373574759811163,
+      "learning_rate": 0.0019777777777777775,
+      "loss": 0.7525,
+      "step": 89
+    },
+    {
+      "epoch": 0.1999444598722577,
+      "grad_norm": 0.0014255450805649161,
+      "learning_rate": 0.002,
+      "loss": 0.6427,
+      "step": 90
+    },
+    {
+      "epoch": 0.20216606498194944,
+      "grad_norm": 0.0015481573063880205,
+      "learning_rate": 0.002022222222222222,
+      "loss": 0.7504,
+      "step": 91
+    },
+    {
+      "epoch": 0.20438767009164122,
+      "grad_norm": 0.0018919891444966197,
+      "learning_rate": 0.0020444444444444447,
+      "loss": 0.679,
+      "step": 92
+    },
+    {
+      "epoch": 0.20660927520133296,
+      "grad_norm": 0.001393840298987925,
+      "learning_rate": 0.0020666666666666667,
+      "loss": 0.5829,
+      "step": 93
+    },
+    {
+      "epoch": 0.2088308803110247,
+      "grad_norm": 0.002521130722016096,
+      "learning_rate": 0.002088888888888889,
+      "loss": 0.6741,
+      "step": 94
+    },
+    {
+      "epoch": 0.21105248542071647,
+      "grad_norm": 0.001192085794173181,
+      "learning_rate": 0.0021111111111111113,
+      "loss": 0.612,
+      "step": 95
+    },
+    {
+      "epoch": 0.21327409053040822,
+      "grad_norm": 0.0020134742371737957,
+      "learning_rate": 0.0021333333333333334,
+      "loss": 0.6554,
+      "step": 96
+    },
+    {
+      "epoch": 0.21549569564009996,
+      "grad_norm": 0.001502550090663135,
+      "learning_rate": 0.0021555555555555555,
+      "loss": 0.6807,
+      "step": 97
+    },
+    {
+      "epoch": 0.21771730074979173,
+      "grad_norm": 0.0012775326613336802,
+      "learning_rate": 0.002177777777777778,
+      "loss": 0.5654,
+      "step": 98
+    },
+    {
+      "epoch": 0.21993890585948347,
+      "grad_norm": 0.0017310921102762222,
+      "learning_rate": 0.0021999999999999997,
+      "loss": 0.6411,
+      "step": 99
+    },
+    {
+      "epoch": 0.22216051096917522,
+      "grad_norm": 0.0025989811401814222,
+      "learning_rate": 0.0022222222222222222,
+      "loss": 0.7062,
+      "step": 100
+    },
+    {
+      "epoch": 0.224382116078867,
+      "grad_norm": 0.0010904415976256132,
+      "learning_rate": 0.0022444444444444443,
+      "loss": 0.6367,
+      "step": 101
+    },
+    {
+      "epoch": 0.22660372118855873,
+      "grad_norm": 0.002046885434538126,
+      "learning_rate": 0.002266666666666667,
+      "loss": 0.5193,
+      "step": 102
+    },
+    {
+      "epoch": 0.22882532629825048,
+      "grad_norm": 0.0019200810929760337,
+      "learning_rate": 0.002288888888888889,
+      "loss": 0.5122,
+      "step": 103
+    },
+    {
+      "epoch": 0.23104693140794225,
+      "grad_norm": 0.002231223974376917,
+      "learning_rate": 0.002311111111111111,
+      "loss": 0.7042,
+      "step": 104
+    },
+    {
+      "epoch": 0.233268536517634,
+      "grad_norm": 0.001883045770227909,
+      "learning_rate": 0.0023333333333333335,
+      "loss": 0.5673,
+      "step": 105
+    },
+    {
+      "epoch": 0.23549014162732573,
+      "grad_norm": 0.0014234882546588778,
+      "learning_rate": 0.0023555555555555556,
+      "loss": 0.6498,
+      "step": 106
+    },
+    {
+      "epoch": 0.2377117467370175,
+      "grad_norm": 0.0013277216348797083,
+      "learning_rate": 0.002377777777777778,
+      "loss": 0.644,
+      "step": 107
+    },
+    {
+      "epoch": 0.23993335184670925,
+      "grad_norm": 0.0019941842183470726,
+      "learning_rate": 0.0024000000000000002,
+      "loss": 0.6425,
+      "step": 108
+    },
+    {
+      "epoch": 0.242154956956401,
+      "grad_norm": 0.0017136182868853211,
+      "learning_rate": 0.0024222222222222223,
+      "loss": 0.6934,
+      "step": 109
+    },
+    {
+      "epoch": 0.24437656206609276,
+      "grad_norm": 0.0018905687611550093,
+      "learning_rate": 0.0024444444444444444,
+      "loss": 0.6896,
+      "step": 110
+    },
+    {
+      "epoch": 0.2465981671757845,
+      "grad_norm": 0.0024479280691593885,
+      "learning_rate": 0.0024666666666666665,
+      "loss": 0.6073,
+      "step": 111
+    },
+    {
+      "epoch": 0.24881977228547625,
+      "grad_norm": 0.00235267193056643,
+      "learning_rate": 0.002488888888888889,
+      "loss": 0.716,
+      "step": 112
+    },
+    {
+      "epoch": 0.251041377395168,
+      "grad_norm": 0.0016832964029163122,
+      "learning_rate": 0.002511111111111111,
+      "loss": 0.616,
+      "step": 113
+    },
+    {
+      "epoch": 0.25326298250485973,
+      "grad_norm": 0.001741166110150516,
+      "learning_rate": 0.002533333333333333,
+      "loss": 0.6916,
+      "step": 114
+    },
+    {
+      "epoch": 0.2554845876145515,
+      "grad_norm": 0.0035528475418686867,
+      "learning_rate": 0.0025555555555555557,
+      "loss": 0.5828,
+      "step": 115
+    },
+    {
+      "epoch": 0.2577061927242433,
+      "grad_norm": 0.0017036315985023975,
+      "learning_rate": 0.002577777777777778,
+      "loss": 0.7605,
+      "step": 116
+    },
+    {
+      "epoch": 0.259927797833935,
+      "grad_norm": 0.0015289601869881153,
+      "learning_rate": 0.0026000000000000003,
+      "loss": 0.6244,
+      "step": 117
+    },
+    {
+      "epoch": 0.26214940294362676,
+      "grad_norm": 0.002781358314678073,
+      "learning_rate": 0.0026222222222222224,
+      "loss": 0.6295,
+      "step": 118
+    },
+    {
+      "epoch": 0.26437100805331853,
+      "grad_norm": 0.0013995033223181963,
+      "learning_rate": 0.0026444444444444445,
+      "loss": 0.6083,
+      "step": 119
+    },
+    {
+      "epoch": 0.26659261316301025,
+      "grad_norm": 0.0064442637376487255,
+      "learning_rate": 0.0026666666666666666,
+      "loss": 0.9319,
+      "step": 120
+    },
+    {
+      "epoch": 0.268814218272702,
+      "grad_norm": 0.002406611107289791,
+      "learning_rate": 0.0026888888888888887,
+      "loss": 0.8775,
+      "step": 121
+    },
+    {
+      "epoch": 0.2710358233823938,
+      "grad_norm": 0.002220314694568515,
+      "learning_rate": 0.002711111111111111,
+      "loss": 0.6365,
+      "step": 122
+    },
+    {
+      "epoch": 0.2732574284920855,
+      "grad_norm": 0.003288067877292633,
+      "learning_rate": 0.0027333333333333333,
+      "loss": 0.703,
+      "step": 123
+    },
+    {
+      "epoch": 0.2754790336017773,
+      "grad_norm": 0.0018617436289787292,
+      "learning_rate": 0.002755555555555556,
+      "loss": 0.6479,
+      "step": 124
+    },
+    {
+      "epoch": 0.27770063871146905,
+      "grad_norm": 0.002223158022388816,
+      "learning_rate": 0.002777777777777778,
+      "loss": 0.5834,
+      "step": 125
+    },
+    {
+      "epoch": 0.27992224382116077,
+      "grad_norm": 0.002752475207671523,
+      "learning_rate": 0.0028,
+      "loss": 0.6261,
+      "step": 126
+    },
+    {
+      "epoch": 0.28214384893085254,
+      "grad_norm": 0.002583549590781331,
+      "learning_rate": 0.0028222222222222225,
+      "loss": 0.6627,
+      "step": 127
+    },
+    {
+      "epoch": 0.2843654540405443,
+      "grad_norm": 0.003046439029276371,
+      "learning_rate": 0.0028444444444444446,
+      "loss": 0.6843,
+      "step": 128
+    },
+    {
+      "epoch": 0.286587059150236,
+      "grad_norm": 0.0020824731327593327,
+      "learning_rate": 0.0028666666666666667,
+      "loss": 0.6444,
+      "step": 129
+    },
+    {
+      "epoch": 0.2888086642599278,
+      "grad_norm": 0.002808007877320051,
+      "learning_rate": 0.0028888888888888888,
+      "loss": 0.6429,
+      "step": 130
+    },
+    {
+      "epoch": 0.29103026936961957,
+      "grad_norm": 0.0021083897445350885,
+      "learning_rate": 0.002911111111111111,
+      "loss": 0.7279,
+      "step": 131
+    },
+    {
+      "epoch": 0.2932518744793113,
+      "grad_norm": 0.0023642443120479584,
+      "learning_rate": 0.0029333333333333334,
+      "loss": 0.6642,
+      "step": 132
+    },
+    {
+      "epoch": 0.29547347958900305,
+      "grad_norm": 0.002401075093075633,
+      "learning_rate": 0.0029555555555555555,
+      "loss": 0.6309,
+      "step": 133
+    },
+    {
+      "epoch": 0.2976950846986948,
+      "grad_norm": 0.0029004139360040426,
+      "learning_rate": 0.002977777777777778,
+      "loss": 0.7236,
+      "step": 134
+    },
+    {
+      "epoch": 0.29991668980838654,
+      "grad_norm": 0.003024290781468153,
+      "learning_rate": 0.003,
+      "loss": 0.7847,
+      "step": 135
+    },
+    {
+      "epoch": 0.3021382949180783,
+      "grad_norm": 0.004532734397798777,
+      "learning_rate": 0.0029975308641975312,
+      "loss": 0.6109,
+      "step": 136
+    },
+    {
+      "epoch": 0.3043599000277701,
+      "grad_norm": 0.0032856380566954613,
+      "learning_rate": 0.0029950617283950615,
+      "loss": 0.704,
+      "step": 137
+    },
+    {
+      "epoch": 0.3065815051374618,
+      "grad_norm": 0.0034370562061667442,
+      "learning_rate": 0.0029925925925925927,
+      "loss": 0.7568,
+      "step": 138
+    },
+    {
+      "epoch": 0.30880311024715357,
+      "grad_norm": 0.002310090931132436,
+      "learning_rate": 0.0029901234567901234,
+      "loss": 0.7956,
+      "step": 139
+    },
+    {
+      "epoch": 0.31102471535684534,
+      "grad_norm": 0.002518034540116787,
+      "learning_rate": 0.0029876543209876546,
+      "loss": 0.6315,
+      "step": 140
+    },
+    {
+      "epoch": 0.31324632046653705,
+      "grad_norm": 0.002741769189015031,
+      "learning_rate": 0.002985185185185185,
+      "loss": 0.5235,
+      "step": 141
+    },
+    {
+      "epoch": 0.3154679255762288,
+      "grad_norm": 0.002382303588092327,
+      "learning_rate": 0.002982716049382716,
+      "loss": 0.6484,
+      "step": 142
+    },
+    {
+      "epoch": 0.3176895306859206,
+      "grad_norm": 0.0025453900452703238,
+      "learning_rate": 0.0029802469135802472,
+      "loss": 0.5183,
+      "step": 143
+    },
+    {
+      "epoch": 0.3199111357956123,
+      "grad_norm": 0.002527153817936778,
+      "learning_rate": 0.002977777777777778,
+      "loss": 0.8045,
+      "step": 144
+    },
+    {
+      "epoch": 0.3221327409053041,
+      "grad_norm": 0.0027810451574623585,
+      "learning_rate": 0.0029753086419753087,
+      "loss": 0.6737,
+      "step": 145
+    },
+    {
+      "epoch": 0.32435434601499585,
+      "grad_norm": 0.003914841450750828,
+      "learning_rate": 0.0029728395061728394,
+      "loss": 0.6568,
+      "step": 146
+    },
+    {
+      "epoch": 0.32657595112468757,
+      "grad_norm": 0.0025024593342095613,
+      "learning_rate": 0.0029703703703703706,
+      "loss": 0.5706,
+      "step": 147
+    },
+    {
+      "epoch": 0.32879755623437934,
+      "grad_norm": 0.003010642249137163,
+      "learning_rate": 0.0029679012345679013,
+      "loss": 0.6303,
+      "step": 148
+    },
+    {
+      "epoch": 0.3310191613440711,
+      "grad_norm": 0.003823331091552973,
+      "learning_rate": 0.002965432098765432,
+      "loss": 0.5775,
+      "step": 149
+    },
+    {
+      "epoch": 0.3332407664537628,
+      "grad_norm": 0.003069646656513214,
+      "learning_rate": 0.002962962962962963,
+      "loss": 0.771,
+      "step": 150
+    },
+    {
+      "epoch": 0.3354623715634546,
+      "grad_norm": 0.0031701356638222933,
+      "learning_rate": 0.002960493827160494,
+      "loss": 0.6726,
+      "step": 151
+    },
+    {
+      "epoch": 0.33768397667314637,
+      "grad_norm": 0.0047783139161765575,
+      "learning_rate": 0.0029580246913580247,
+      "loss": 0.5175,
+      "step": 152
+    },
+    {
+      "epoch": 0.3399055817828381,
+      "grad_norm": 0.002878882922232151,
+      "learning_rate": 0.0029555555555555555,
+      "loss": 0.66,
+      "step": 153
+    },
+    {
+      "epoch": 0.34212718689252986,
+      "grad_norm": 0.002322252606973052,
+      "learning_rate": 0.0029530864197530866,
+      "loss": 0.5851,
+      "step": 154
+    },
+    {
+      "epoch": 0.3443487920022216,
+      "grad_norm": 0.004187839105725288,
+      "learning_rate": 0.0029506172839506174,
+      "loss": 0.5594,
+      "step": 155
+    },
+    {
+      "epoch": 0.34657039711191334,
+      "grad_norm": 0.0040303790010511875,
+      "learning_rate": 0.002948148148148148,
+      "loss": 0.7593,
+      "step": 156
+    },
+    {
+      "epoch": 0.3487920022216051,
+      "grad_norm": 0.004690828267484903,
+      "learning_rate": 0.002945679012345679,
+      "loss": 0.591,
+      "step": 157
+    },
+    {
+      "epoch": 0.3510136073312969,
+      "grad_norm": 0.0035938743967562914,
+      "learning_rate": 0.00294320987654321,
+      "loss": 0.7527,
+      "step": 158
+    },
+    {
+      "epoch": 0.3532352124409886,
+      "grad_norm": 0.0031942499335855246,
+      "learning_rate": 0.0029407407407407407,
+      "loss": 0.6939,
+      "step": 159
+    },
+    {
+      "epoch": 0.35545681755068037,
+      "grad_norm": 0.00442874850705266,
+      "learning_rate": 0.0029382716049382715,
+      "loss": 0.4872,
+      "step": 160
+    },
+    {
+      "epoch": 0.35767842266037214,
+      "grad_norm": 0.002961769700050354,
+      "learning_rate": 0.0029358024691358026,
+      "loss": 0.6919,
+      "step": 161
+    },
+    {
+      "epoch": 0.35990002777006386,
+      "grad_norm": 0.004248150624334812,
+      "learning_rate": 0.0029333333333333334,
+      "loss": 0.5914,
+      "step": 162
+    },
+    {
+      "epoch": 0.36212163287975563,
+      "grad_norm": 0.003655125154182315,
+      "learning_rate": 0.002930864197530864,
+      "loss": 0.6434,
+      "step": 163
+    },
+    {
+      "epoch": 0.3643432379894474,
+      "grad_norm": 0.0027984960470348597,
+      "learning_rate": 0.002928395061728395,
+      "loss": 0.6544,
+      "step": 164
+    },
+    {
+      "epoch": 0.3665648430991391,
+      "grad_norm": 0.003448409028351307,
+      "learning_rate": 0.002925925925925926,
+      "loss": 0.6135,
+      "step": 165
+    },
+    {
+      "epoch": 0.3687864482088309,
+      "grad_norm": 0.0023583476431667805,
+      "learning_rate": 0.002923456790123457,
+      "loss": 0.5549,
+      "step": 166
+    },
+    {
+      "epoch": 0.37100805331852266,
+      "grad_norm": 0.003774197306483984,
+      "learning_rate": 0.0029209876543209875,
+      "loss": 0.6155,
+      "step": 167
+    },
+    {
+      "epoch": 0.3732296584282144,
+      "grad_norm": 0.0039054304361343384,
+      "learning_rate": 0.0029185185185185186,
+      "loss": 0.4939,
+      "step": 168
+    },
+    {
+      "epoch": 0.37545126353790614,
+      "grad_norm": 0.0034814560785889626,
+      "learning_rate": 0.0029160493827160494,
+      "loss": 0.6343,
+      "step": 169
+    },
+    {
+      "epoch": 0.3776728686475979,
+      "grad_norm": 0.002560647204518318,
+      "learning_rate": 0.0029135802469135805,
+      "loss": 0.5198,
+      "step": 170
+    },
+    {
+      "epoch": 0.37989447375728963,
+      "grad_norm": 0.0023003127425909042,
+      "learning_rate": 0.002911111111111111,
+      "loss": 0.5431,
+      "step": 171
+    },
+    {
+      "epoch": 0.3821160788669814,
+      "grad_norm": 0.003486617701128125,
+      "learning_rate": 0.002908641975308642,
+      "loss": 0.4754,
+      "step": 172
+    },
+    {
+      "epoch": 0.3843376839766732,
+      "grad_norm": 0.0030546062625944614,
+      "learning_rate": 0.002906172839506173,
+      "loss": 0.6999,
+      "step": 173
+    },
+    {
+      "epoch": 0.3865592890863649,
+      "grad_norm": 0.003570267930626869,
+      "learning_rate": 0.002903703703703704,
+      "loss": 0.6752,
+      "step": 174
+    },
+    {
+      "epoch": 0.38878089419605666,
+      "grad_norm": 0.0036987727507948875,
+      "learning_rate": 0.0029012345679012346,
+      "loss": 0.6123,
+      "step": 175
+    },
+    {
+      "epoch": 0.39100249930574843,
+      "grad_norm": 0.004126350861042738,
+      "learning_rate": 0.0028987654320987654,
+      "loss": 0.5922,
+      "step": 176
+    },
+    {
+      "epoch": 0.39322410441544015,
+      "grad_norm": 0.0043579223565757275,
+      "learning_rate": 0.0028962962962962966,
+      "loss": 0.7249,
+      "step": 177
+    },
+    {
+      "epoch": 0.3954457095251319,
+      "grad_norm": 0.005049738567322493,
+      "learning_rate": 0.0028938271604938273,
+      "loss": 0.7098,
+      "step": 178
+    },
+    {
+      "epoch": 0.3976673146348237,
+      "grad_norm": 0.004803841933608055,
+      "learning_rate": 0.002891358024691358,
+      "loss": 0.6547,
+      "step": 179
+    },
+    {
+      "epoch": 0.3998889197445154,
+      "grad_norm": 0.004663265775889158,
+      "learning_rate": 0.0028888888888888888,
+      "loss": 0.5621,
+      "step": 180
+    },
+    {
+      "epoch": 0.4021105248542072,
+      "grad_norm": 0.005818499252200127,
+      "learning_rate": 0.00288641975308642,
+      "loss": 0.6772,
+      "step": 181
+    },
+    {
+      "epoch": 0.4043321299638989,
+      "grad_norm": 0.003460573498159647,
+      "learning_rate": 0.0028839506172839507,
+      "loss": 0.4637,
+      "step": 182
+    },
+    {
+      "epoch": 0.40655373507359066,
+      "grad_norm": 0.004111323039978743,
+      "learning_rate": 0.0028814814814814814,
+      "loss": 0.5993,
+      "step": 183
+    },
+    {
+      "epoch": 0.40877534018328243,
+      "grad_norm": 0.0037442215252667665,
+      "learning_rate": 0.0028790123456790126,
+      "loss": 0.6081,
+      "step": 184
+    },
+    {
+      "epoch": 0.41099694529297415,
+      "grad_norm": 0.005370614118874073,
+      "learning_rate": 0.0028765432098765433,
+      "loss": 0.628,
+      "step": 185
+    },
+    {
+      "epoch": 0.4132185504026659,
+      "grad_norm": 0.009289095178246498,
+      "learning_rate": 0.002874074074074074,
+      "loss": 0.5611,
+      "step": 186
+    },
+    {
+      "epoch": 0.4154401555123577,
+      "grad_norm": 0.013159370049834251,
+      "learning_rate": 0.0028716049382716048,
+      "loss": 0.6732,
+      "step": 187
+    },
+    {
+      "epoch": 0.4176617606220494,
+      "grad_norm": 0.008524056524038315,
+      "learning_rate": 0.002869135802469136,
+      "loss": 0.5685,
+      "step": 188
+    },
+    {
+      "epoch": 0.4198833657317412,
+      "grad_norm": 0.008815361186861992,
+      "learning_rate": 0.0028666666666666667,
+      "loss": 0.7353,
+      "step": 189
+    },
+    {
+      "epoch": 0.42210497084143295,
+      "grad_norm": 0.005370002705603838,
+      "learning_rate": 0.0028641975308641974,
+      "loss": 0.5858,
+      "step": 190
+    },
+    {
+      "epoch": 0.42432657595112466,
+      "grad_norm": 0.00529759656637907,
+      "learning_rate": 0.0028617283950617286,
+      "loss": 0.7153,
+      "step": 191
+    },
+    {
+      "epoch": 0.42654818106081643,
+      "grad_norm": 0.004369831178337336,
+      "learning_rate": 0.0028592592592592593,
+      "loss": 0.7741,
+      "step": 192
+    },
+    {
+      "epoch": 0.4287697861705082,
+      "grad_norm": 0.003865456208586693,
+      "learning_rate": 0.00285679012345679,
+      "loss": 0.6637,
+      "step": 193
+    },
+    {
+      "epoch": 0.4309913912801999,
+      "grad_norm": 0.004272242542356253,
+      "learning_rate": 0.0028543209876543208,
+      "loss": 0.4771,
+      "step": 194
+    },
+    {
+      "epoch": 0.4332129963898917,
+      "grad_norm": 0.003245055442675948,
+      "learning_rate": 0.002851851851851852,
+      "loss": 0.4521,
+      "step": 195
+    },
+    {
+      "epoch": 0.43543460149958346,
+      "grad_norm": 0.004816419444978237,
+      "learning_rate": 0.002849382716049383,
+      "loss": 0.6017,
+      "step": 196
+    },
+    {
+      "epoch": 0.4376562066092752,
+      "grad_norm": 0.004628610797226429,
+      "learning_rate": 0.0028469135802469134,
+      "loss": 0.616,
+      "step": 197
+    },
+    {
+      "epoch": 0.43987781171896695,
+      "grad_norm": 0.004131810273975134,
+      "learning_rate": 0.0028444444444444446,
+      "loss": 0.5681,
+      "step": 198
+    },
+    {
+      "epoch": 0.4420994168286587,
+      "grad_norm": 0.0029758147429674864,
+      "learning_rate": 0.0028419753086419753,
+      "loss": 0.5721,
+      "step": 199
+    },
+    {
+      "epoch": 0.44432102193835044,
+      "grad_norm": 0.0033105132170021534,
+      "learning_rate": 0.0028395061728395065,
+      "loss": 0.5717,
+      "step": 200
+    },
+    {
+      "epoch": 0.4465426270480422,
+      "grad_norm": 0.003180293133482337,
+      "learning_rate": 0.002837037037037037,
+      "loss": 0.5594,
+      "step": 201
+    },
+    {
+      "epoch": 0.448764232157734,
+      "grad_norm": 0.004509236663579941,
+      "learning_rate": 0.002834567901234568,
+      "loss": 0.5528,
+      "step": 202
+    },
+    {
+      "epoch": 0.4509858372674257,
+      "grad_norm": 0.0039048679172992706,
+      "learning_rate": 0.002832098765432099,
+      "loss": 0.6696,
+      "step": 203
+    },
+    {
+      "epoch": 0.45320744237711746,
+      "grad_norm": 0.004375668242573738,
+      "learning_rate": 0.00282962962962963,
+      "loss": 0.5643,
+      "step": 204
+    },
+    {
+      "epoch": 0.45542904748680924,
+      "grad_norm": 0.004228411708027124,
+      "learning_rate": 0.0028271604938271606,
+      "loss": 0.4618,
+      "step": 205
+    },
+    {
+      "epoch": 0.45765065259650095,
+      "grad_norm": 0.00472232885658741,
+      "learning_rate": 0.0028246913580246913,
+      "loss": 0.6037,
+      "step": 206
+    },
+    {
+      "epoch": 0.4598722577061927,
+      "grad_norm": 0.003651043400168419,
+      "learning_rate": 0.0028222222222222225,
+      "loss": 0.6719,
+      "step": 207
+    },
+    {
+      "epoch": 0.4620938628158845,
+      "grad_norm": 0.00520613556727767,
+      "learning_rate": 0.0028197530864197532,
+      "loss": 0.6436,
+      "step": 208
+    },
+    {
+      "epoch": 0.4643154679255762,
+      "grad_norm": 0.0054262257181108,
+      "learning_rate": 0.002817283950617284,
+      "loss": 0.5663,
+      "step": 209
+    },
+    {
+      "epoch": 0.466537073035268,
+      "grad_norm": 0.002916304161772132,
+      "learning_rate": 0.0028148148148148147,
+      "loss": 0.6353,
+      "step": 210
+    },
+    {
+      "epoch": 0.46875867814495975,
+      "grad_norm": 0.0056013804860413074,
+      "learning_rate": 0.002812345679012346,
+      "loss": 0.6574,
+      "step": 211
+    },
+    {
+      "epoch": 0.47098028325465147,
+      "grad_norm": 0.004248346202075481,
+      "learning_rate": 0.0028098765432098766,
+      "loss": 0.7195,
+      "step": 212
+    },
+    {
+      "epoch": 0.47320188836434324,
+      "grad_norm": 0.004659584257751703,
+      "learning_rate": 0.0028074074074074073,
+      "loss": 0.5662,
+      "step": 213
+    },
+    {
+      "epoch": 0.475423493474035,
+      "grad_norm": 0.007387971039861441,
+      "learning_rate": 0.0028049382716049385,
+      "loss": 0.5926,
+      "step": 214
+    },
+    {
+      "epoch": 0.4776450985837267,
+      "grad_norm": 0.005048746708780527,
+      "learning_rate": 0.0028024691358024692,
+      "loss": 0.6544,
+      "step": 215
+    },
+    {
+      "epoch": 0.4798667036934185,
+      "grad_norm": 0.004469868261367083,
+      "learning_rate": 0.0028,
+      "loss": 0.6088,
+      "step": 216
+    },
+    {
+      "epoch": 0.48208830880311027,
+      "grad_norm": 0.004091240931302309,
+      "learning_rate": 0.0027975308641975307,
+      "loss": 0.5936,
+      "step": 217
+    },
+    {
+      "epoch": 0.484309913912802,
+      "grad_norm": 0.004576526582241058,
+      "learning_rate": 0.002795061728395062,
+      "loss": 0.6662,
+      "step": 218
+    },
+    {
+      "epoch": 0.48653151902249375,
+      "grad_norm": 0.006232080049812794,
+      "learning_rate": 0.0027925925925925926,
+      "loss": 0.6338,
+      "step": 219
+    },
+    {
+      "epoch": 0.4887531241321855,
+      "grad_norm": 0.004759920295327902,
+      "learning_rate": 0.0027901234567901233,
+      "loss": 0.5844,
+      "step": 220
+    },
+    {
+      "epoch": 0.49097472924187724,
+      "grad_norm": 0.004427391104400158,
+      "learning_rate": 0.0027876543209876545,
+      "loss": 0.5888,
+      "step": 221
+    },
+    {
+      "epoch": 0.493196334351569,
+      "grad_norm": 0.005065519362688065,
+      "learning_rate": 0.0027851851851851852,
+      "loss": 0.6999,
+      "step": 222
+    },
+    {
+      "epoch": 0.4954179394612608,
+      "grad_norm": 0.005866500083357096,
+      "learning_rate": 0.002782716049382716,
+      "loss": 0.5718,
+      "step": 223
+    },
+    {
+      "epoch": 0.4976395445709525,
+      "grad_norm": 0.0052678692154586315,
+      "learning_rate": 0.0027802469135802467,
+      "loss": 0.5274,
+      "step": 224
+    },
+    {
+      "epoch": 0.49986114968064427,
+      "grad_norm": 0.0039049943443387747,
+      "learning_rate": 0.002777777777777778,
+      "loss": 0.7269,
+      "step": 225
+    },
+    {
+      "epoch": 0.502082754790336,
+      "grad_norm": 0.003937671426683664,
+      "learning_rate": 0.002775308641975309,
+      "loss": 0.6606,
+      "step": 226
+    },
+    {
+      "epoch": 0.5043043599000278,
+      "grad_norm": 0.002638050587847829,
+      "learning_rate": 0.0027728395061728394,
+      "loss": 0.6129,
+      "step": 227
+    },
+    {
+      "epoch": 0.5065259650097195,
+      "grad_norm": 0.0044793980196118355,
+      "learning_rate": 0.0027703703703703705,
+      "loss": 0.6376,
+      "step": 228
+    },
+    {
+      "epoch": 0.5087475701194113,
+      "grad_norm": 0.004454802256077528,
+      "learning_rate": 0.0027679012345679013,
+      "loss": 0.6064,
+      "step": 229
+    },
+    {
+      "epoch": 0.510969175229103,
+      "grad_norm": 0.003144732676446438,
+      "learning_rate": 0.0027654320987654324,
+      "loss": 0.5875,
+      "step": 230
+    },
+    {
+      "epoch": 0.5131907803387947,
+      "grad_norm": 0.004579763859510422,
+      "learning_rate": 0.0027629629629629627,
+      "loss": 0.7472,
+      "step": 231
+    },
+    {
+      "epoch": 0.5154123854484866,
+      "grad_norm": 0.007608956657350063,
+      "learning_rate": 0.002760493827160494,
+      "loss": 0.5966,
+      "step": 232
+    },
+    {
+      "epoch": 0.5176339905581783,
+      "grad_norm": 0.003314356319606304,
+      "learning_rate": 0.0027580246913580246,
+      "loss": 0.5678,
+      "step": 233
+    },
+    {
+      "epoch": 0.51985559566787,
+      "grad_norm": 0.021426383405923843,
+      "learning_rate": 0.002755555555555556,
+      "loss": 0.8359,
+      "step": 234
+    },
+    {
+      "epoch": 0.5220772007775618,
+      "grad_norm": 0.004511243663728237,
+      "learning_rate": 0.0027530864197530865,
+      "loss": 0.4909,
+      "step": 235
+    },
+    {
+      "epoch": 0.5242988058872535,
+      "grad_norm": 0.00266853766515851,
+      "learning_rate": 0.0027506172839506173,
+      "loss": 0.6042,
+      "step": 236
+    },
+    {
+      "epoch": 0.5265204109969452,
+      "grad_norm": 0.006429709494113922,
+      "learning_rate": 0.0027481481481481484,
+      "loss": 0.7027,
+      "step": 237
+    },
+    {
+      "epoch": 0.5287420161066371,
+      "grad_norm": 0.003927881829440594,
+      "learning_rate": 0.002745679012345679,
+      "loss": 0.7546,
+      "step": 238
+    },
+    {
+      "epoch": 0.5309636212163288,
+      "grad_norm": 0.0050257411785423756,
+      "learning_rate": 0.00274320987654321,
+      "loss": 0.7372,
+      "step": 239
+    },
+    {
+      "epoch": 0.5331852263260205,
+      "grad_norm": 0.003318399889394641,
+      "learning_rate": 0.0027407407407407406,
+      "loss": 0.6836,
+      "step": 240
+    },
+    {
+      "epoch": 0.5354068314357123,
+      "grad_norm": 0.005656864959746599,
+      "learning_rate": 0.002738271604938272,
+      "loss": 0.5766,
+      "step": 241
+    },
+    {
+      "epoch": 0.537628436545404,
+      "grad_norm": 0.004331695381551981,
+      "learning_rate": 0.0027358024691358025,
+      "loss": 0.5942,
+      "step": 242
+    },
+    {
+      "epoch": 0.5398500416550958,
+      "grad_norm": 0.003149270312860608,
+      "learning_rate": 0.0027333333333333333,
+      "loss": 0.5426,
+      "step": 243
+    },
+    {
+      "epoch": 0.5420716467647876,
+      "grad_norm": 0.004050451330840588,
+      "learning_rate": 0.0027308641975308644,
+      "loss": 0.5257,
+      "step": 244
+    },
+    {
+      "epoch": 0.5442932518744793,
+      "grad_norm": 0.0048174443654716015,
+      "learning_rate": 0.002728395061728395,
+      "loss": 0.569,
+      "step": 245
+    },
+    {
+      "epoch": 0.546514856984171,
+      "grad_norm": 0.004108496475964785,
+      "learning_rate": 0.002725925925925926,
+      "loss": 0.7042,
+      "step": 246
+    },
+    {
+      "epoch": 0.5487364620938628,
+      "grad_norm": 0.006572370883077383,
+      "learning_rate": 0.0027234567901234566,
+      "loss": 0.6698,
+      "step": 247
+    },
+    {
+      "epoch": 0.5509580672035546,
+      "grad_norm": 0.0034175198525190353,
+      "learning_rate": 0.002720987654320988,
+      "loss": 0.7017,
+      "step": 248
+    },
+    {
+      "epoch": 0.5531796723132463,
+      "grad_norm": 0.0031640417873859406,
+      "learning_rate": 0.0027185185185185185,
+      "loss": 0.5557,
+      "step": 249
+    },
+    {
+      "epoch": 0.5554012774229381,
+      "grad_norm": 0.003084770869463682,
+      "learning_rate": 0.0027160493827160493,
+      "loss": 0.4013,
+      "step": 250
+    },
+    {
+      "epoch": 0.5576228825326298,
+      "grad_norm": 0.004879522603005171,
+      "learning_rate": 0.0027135802469135805,
+      "loss": 0.6765,
+      "step": 251
+    },
+    {
+      "epoch": 0.5598444876423215,
+      "grad_norm": 0.0038094199262559414,
+      "learning_rate": 0.002711111111111111,
+      "loss": 0.7209,
+      "step": 252
+    },
+    {
+      "epoch": 0.5620660927520134,
+      "grad_norm": 0.004499315749853849,
+      "learning_rate": 0.002708641975308642,
+      "loss": 0.6623,
+      "step": 253
+    },
+    {
+      "epoch": 0.5642876978617051,
+      "grad_norm": 0.005681379698216915,
+      "learning_rate": 0.0027061728395061727,
+      "loss": 0.6463,
+      "step": 254
+    },
+    {
+      "epoch": 0.5665093029713968,
+      "grad_norm": 0.0031827313359826803,
+      "learning_rate": 0.002703703703703704,
+      "loss": 0.4868,
+      "step": 255
+    },
+    {
+      "epoch": 0.5687309080810886,
+      "grad_norm": 0.002911441260948777,
+      "learning_rate": 0.002701234567901235,
+      "loss": 0.6855,
+      "step": 256
+    },
+    {
+      "epoch": 0.5709525131907803,
+      "grad_norm": 0.002942469669505954,
+      "learning_rate": 0.0026987654320987653,
+      "loss": 0.6749,
+      "step": 257
+    },
+    {
+      "epoch": 0.573174118300472,
+      "grad_norm": 0.0030430902261286974,
+      "learning_rate": 0.0026962962962962965,
+      "loss": 0.5328,
+      "step": 258
+    },
+    {
+      "epoch": 0.5753957234101639,
+      "grad_norm": 0.0035275311674922705,
+      "learning_rate": 0.002693827160493827,
+      "loss": 0.8042,
+      "step": 259
+    },
+    {
+      "epoch": 0.5776173285198556,
+      "grad_norm": 0.005831616465002298,
+      "learning_rate": 0.0026913580246913584,
+      "loss": 0.5116,
+      "step": 260
+    },
+    {
+      "epoch": 0.5798389336295473,
+      "grad_norm": 0.00925903208553791,
+      "learning_rate": 0.0026888888888888887,
+      "loss": 0.6452,
+      "step": 261
+    },
+    {
+      "epoch": 0.5820605387392391,
+      "grad_norm": 0.0031285895965993404,
+      "learning_rate": 0.00268641975308642,
+      "loss": 0.5717,
+      "step": 262
+    },
+    {
+      "epoch": 0.5842821438489308,
+      "grad_norm": 0.004181864205747843,
+      "learning_rate": 0.0026839506172839506,
+      "loss": 0.608,
+      "step": 263
+    },
+    {
+      "epoch": 0.5865037489586226,
+      "grad_norm": 0.02851380594074726,
+      "learning_rate": 0.0026814814814814817,
+      "loss": 0.6639,
+      "step": 264
+    },
+    {
+      "epoch": 0.5887253540683144,
+      "grad_norm": 0.004196640104055405,
+      "learning_rate": 0.0026790123456790125,
+      "loss": 0.6573,
+      "step": 265
+    },
+    {
+      "epoch": 0.5909469591780061,
+      "grad_norm": 0.002790750004351139,
+      "learning_rate": 0.002676543209876543,
+      "loss": 0.6721,
+      "step": 266
+    },
+    {
+      "epoch": 0.5931685642876978,
+      "grad_norm": 0.004602896049618721,
+      "learning_rate": 0.0026740740740740744,
+      "loss": 0.6925,
+      "step": 267
+    },
+    {
+      "epoch": 0.5953901693973896,
+      "grad_norm": 0.008625895716249943,
+      "learning_rate": 0.0026716049382716047,
+      "loss": 0.6228,
+      "step": 268
+    },
+    {
+      "epoch": 0.5976117745070814,
+      "grad_norm": 0.00362372025847435,
+      "learning_rate": 0.002669135802469136,
+      "loss": 0.518,
+      "step": 269
+    },
+    {
+      "epoch": 0.5998333796167731,
+      "grad_norm": 0.008692633360624313,
+      "learning_rate": 0.0026666666666666666,
+      "loss": 0.6194,
+      "step": 270
+    },
+    {
+      "epoch": 0.6020549847264649,
+      "grad_norm": 0.015166512690484524,
+      "learning_rate": 0.0026641975308641977,
+      "loss": 0.7203,
+      "step": 271
+    },
+    {
+      "epoch": 0.6042765898361566,
+      "grad_norm": 0.0038714068941771984,
+      "learning_rate": 0.0026617283950617285,
+      "loss": 0.5438,
+      "step": 272
+    },
+    {
+      "epoch": 0.6064981949458483,
+      "grad_norm": 0.005590848624706268,
+      "learning_rate": 0.002659259259259259,
+      "loss": 0.5981,
+      "step": 273
+    },
+    {
+      "epoch": 0.6087198000555402,
+      "grad_norm": 0.013241161592304707,
+      "learning_rate": 0.0026567901234567904,
+      "loss": 0.7574,
+      "step": 274
+    },
+    {
+      "epoch": 0.6109414051652319,
+      "grad_norm": 0.004338677506893873,
+      "learning_rate": 0.002654320987654321,
+      "loss": 0.6665,
+      "step": 275
+    },
+    {
+      "epoch": 0.6131630102749236,
+      "grad_norm": 0.00662162946537137,
+      "learning_rate": 0.002651851851851852,
+      "loss": 0.6267,
+      "step": 276
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 0.004217423964291811,
+      "learning_rate": 0.0026493827160493826,
+      "loss": 0.6512,
+      "step": 277
+    },
+    {
+      "epoch": 0.6176062204943071,
+      "grad_norm": 0.004670712631195784,
+      "learning_rate": 0.0026469135802469138,
+      "loss": 0.6363,
+      "step": 278
+    },
+    {
+      "epoch": 0.6198278256039988,
+      "grad_norm": 0.004922980442643166,
+      "learning_rate": 0.0026444444444444445,
+      "loss": 0.6037,
+      "step": 279
+    },
+    {
+      "epoch": 0.6220494307136907,
+      "grad_norm": 0.004932676907628775,
+      "learning_rate": 0.0026419753086419752,
+      "loss": 0.576,
+      "step": 280
+    },
+    {
+      "epoch": 0.6242710358233824,
+      "grad_norm": 0.004741290118545294,
+      "learning_rate": 0.0026395061728395064,
+      "loss": 0.5563,
+      "step": 281
+    },
+    {
+      "epoch": 0.6264926409330741,
+      "grad_norm": 0.00402566883713007,
+      "learning_rate": 0.002637037037037037,
+      "loss": 0.5417,
+      "step": 282
+    },
+    {
+      "epoch": 0.6287142460427659,
+      "grad_norm": 0.003514640498906374,
+      "learning_rate": 0.002634567901234568,
+      "loss": 0.727,
+      "step": 283
+    },
+    {
+      "epoch": 0.6309358511524576,
+      "grad_norm": 0.004528773948550224,
+      "learning_rate": 0.0026320987654320986,
+      "loss": 0.5701,
+      "step": 284
+    },
+    {
+      "epoch": 0.6331574562621494,
+      "grad_norm": 0.004958092700690031,
+      "learning_rate": 0.0026296296296296298,
+      "loss": 0.6251,
+      "step": 285
+    },
+    {
+      "epoch": 0.6353790613718412,
+      "grad_norm": 0.004396567586809397,
+      "learning_rate": 0.0026271604938271605,
+      "loss": 0.6076,
+      "step": 286
+    },
+    {
+      "epoch": 0.6376006664815329,
+      "grad_norm": 0.004292050842195749,
+      "learning_rate": 0.0026246913580246912,
+      "loss": 0.5767,
+      "step": 287
+    },
+    {
+      "epoch": 0.6398222715912246,
+      "grad_norm": 0.0037857245188206434,
+      "learning_rate": 0.0026222222222222224,
+      "loss": 0.6488,
+      "step": 288
+    },
+    {
+      "epoch": 0.6420438767009164,
+      "grad_norm": 0.002987690968438983,
+      "learning_rate": 0.002619753086419753,
+      "loss": 0.5093,
+      "step": 289
+    },
+    {
+      "epoch": 0.6442654818106082,
+      "grad_norm": 0.005551345180720091,
+      "learning_rate": 0.002617283950617284,
+      "loss": 0.7926,
+      "step": 290
+    },
+    {
+      "epoch": 0.6464870869202999,
+      "grad_norm": 0.003886839607730508,
+      "learning_rate": 0.0026148148148148146,
+      "loss": 0.6786,
+      "step": 291
+    },
+    {
+      "epoch": 0.6487086920299917,
+      "grad_norm": 0.0051884097047150135,
+      "learning_rate": 0.0026123456790123458,
+      "loss": 0.6151,
+      "step": 292
+    },
+    {
+      "epoch": 0.6509302971396834,
+      "grad_norm": 0.005092722829431295,
+      "learning_rate": 0.0026098765432098765,
+      "loss": 0.7552,
+      "step": 293
+    },
+    {
+      "epoch": 0.6531519022493751,
+      "grad_norm": 0.004387577064335346,
+      "learning_rate": 0.0026074074074074072,
+      "loss": 0.6128,
+      "step": 294
+    },
+    {
+      "epoch": 0.655373507359067,
+      "grad_norm": 0.005776650737971067,
+      "learning_rate": 0.0026049382716049384,
+      "loss": 0.5585,
+      "step": 295
+    },
+    {
+      "epoch": 0.6575951124687587,
+      "grad_norm": 0.004767559934407473,
+      "learning_rate": 0.002602469135802469,
+      "loss": 0.5365,
+      "step": 296
+    },
+    {
+      "epoch": 0.6598167175784504,
+      "grad_norm": 0.002910695504397154,
+      "learning_rate": 0.0026000000000000003,
+      "loss": 0.5649,
+      "step": 297
+    },
+    {
+      "epoch": 0.6620383226881422,
+      "grad_norm": 0.003558649681508541,
+      "learning_rate": 0.0025975308641975306,
+      "loss": 0.5774,
+      "step": 298
+    },
+    {
+      "epoch": 0.6642599277978339,
+      "grad_norm": 0.003824816318228841,
+      "learning_rate": 0.002595061728395062,
+      "loss": 0.5732,
+      "step": 299
+    },
+    {
+      "epoch": 0.6664815329075257,
+      "grad_norm": 0.0028791490476578474,
+      "learning_rate": 0.0025925925925925925,
+      "loss": 0.5018,
+      "step": 300
+    },
+    {
+      "epoch": 0.6687031380172175,
+      "grad_norm": 0.0036279659252613783,
+      "learning_rate": 0.0025901234567901237,
+      "loss": 0.682,
+      "step": 301
+    },
+    {
+      "epoch": 0.6709247431269092,
+      "grad_norm": 0.004511178005486727,
+      "learning_rate": 0.0025876543209876544,
+      "loss": 0.6445,
+      "step": 302
+    },
+    {
+      "epoch": 0.6731463482366009,
+      "grad_norm": 0.00374900852330029,
+      "learning_rate": 0.002585185185185185,
+      "loss": 0.618,
+      "step": 303
+    },
+    {
+      "epoch": 0.6753679533462927,
+      "grad_norm": 0.0064397756941616535,
+      "learning_rate": 0.0025827160493827163,
+      "loss": 0.5854,
+      "step": 304
+    },
+    {
+      "epoch": 0.6775895584559845,
+      "grad_norm": 0.0034314116928726435,
+      "learning_rate": 0.002580246913580247,
+      "loss": 0.5673,
+      "step": 305
+    },
+    {
+      "epoch": 0.6798111635656762,
+      "grad_norm": 0.0028468379750847816,
+      "learning_rate": 0.002577777777777778,
+      "loss": 0.557,
+      "step": 306
+    },
+    {
+      "epoch": 0.682032768675368,
+      "grad_norm": 0.004024473484605551,
+      "learning_rate": 0.0025753086419753085,
+      "loss": 0.6834,
+      "step": 307
+    },
+    {
+      "epoch": 0.6842543737850597,
+      "grad_norm": 0.00284021208062768,
+      "learning_rate": 0.0025728395061728397,
+      "loss": 0.5258,
+      "step": 308
+    },
+    {
+      "epoch": 0.6864759788947514,
+      "grad_norm": 0.004184060264378786,
+      "learning_rate": 0.0025703703703703704,
+      "loss": 0.5752,
+      "step": 309
+    },
+    {
+      "epoch": 0.6886975840044433,
+      "grad_norm": 0.00356755661778152,
+      "learning_rate": 0.002567901234567901,
+      "loss": 0.535,
+      "step": 310
+    },
+    {
+      "epoch": 0.690919189114135,
+      "grad_norm": 0.002997822593897581,
+      "learning_rate": 0.0025654320987654323,
+      "loss": 0.5818,
+      "step": 311
+    },
+    {
+      "epoch": 0.6931407942238267,
+      "grad_norm": 0.003147626994177699,
+      "learning_rate": 0.002562962962962963,
+      "loss": 0.5176,
+      "step": 312
+    },
+    {
+      "epoch": 0.6953623993335185,
+      "grad_norm": 0.0036907889880239964,
+      "learning_rate": 0.002560493827160494,
+      "loss": 0.6186,
+      "step": 313
+    },
+    {
+      "epoch": 0.6975840044432102,
+      "grad_norm": 0.004177046474069357,
+      "learning_rate": 0.0025580246913580245,
+      "loss": 0.6218,
+      "step": 314
+    },
+    {
+      "epoch": 0.6998056095529019,
+      "grad_norm": 0.0035435082390904427,
+      "learning_rate": 0.0025555555555555557,
+      "loss": 0.7149,
+      "step": 315
+    },
+    {
+      "epoch": 0.7020272146625938,
+      "grad_norm": 0.004030556883662939,
+      "learning_rate": 0.0025530864197530864,
+      "loss": 0.6261,
+      "step": 316
+    },
+    {
+      "epoch": 0.7042488197722855,
+      "grad_norm": 0.003368956968188286,
+      "learning_rate": 0.002550617283950617,
+      "loss": 0.6349,
+      "step": 317
+    },
+    {
+      "epoch": 0.7064704248819772,
+      "grad_norm": 0.0029223670717328787,
+      "learning_rate": 0.0025481481481481483,
+      "loss": 0.5547,
+      "step": 318
+    },
+    {
+      "epoch": 0.708692029991669,
+      "grad_norm": 0.004348669201135635,
+      "learning_rate": 0.002545679012345679,
+      "loss": 0.5123,
+      "step": 319
+    },
+    {
+      "epoch": 0.7109136351013607,
+      "grad_norm": 0.012978849932551384,
+      "learning_rate": 0.00254320987654321,
+      "loss": 0.6939,
+      "step": 320
+    },
+    {
+      "epoch": 0.7131352402110525,
+      "grad_norm": 0.0044335960410535336,
+      "learning_rate": 0.0025407407407407405,
+      "loss": 0.5573,
+      "step": 321
+    },
+    {
+      "epoch": 0.7153568453207443,
+      "grad_norm": 0.0029696300625801086,
+      "learning_rate": 0.0025382716049382717,
+      "loss": 0.6562,
+      "step": 322
+    },
+    {
+      "epoch": 0.717578450430436,
+      "grad_norm": 0.003993246704339981,
+      "learning_rate": 0.0025358024691358024,
+      "loss": 0.6783,
+      "step": 323
+    },
+    {
+      "epoch": 0.7198000555401277,
+      "grad_norm": 0.0035177140962332487,
+      "learning_rate": 0.002533333333333333,
+      "loss": 0.6443,
+      "step": 324
+    },
+    {
+      "epoch": 0.7220216606498195,
+      "grad_norm": 0.006393967662006617,
+      "learning_rate": 0.0025308641975308644,
+      "loss": 0.5847,
+      "step": 325
+    },
+    {
+      "epoch": 0.7242432657595113,
+      "grad_norm": 0.003029291285201907,
+      "learning_rate": 0.002528395061728395,
+      "loss": 0.3798,
+      "step": 326
+    },
+    {
+      "epoch": 0.726464870869203,
+      "grad_norm": 0.0022584160324186087,
+      "learning_rate": 0.0025259259259259263,
+      "loss": 0.6006,
+      "step": 327
+    },
+    {
+      "epoch": 0.7286864759788948,
+      "grad_norm": 0.0029388736002147198,
+      "learning_rate": 0.0025234567901234566,
+      "loss": 0.5932,
+      "step": 328
+    },
+    {
+      "epoch": 0.7309080810885865,
+      "grad_norm": 0.0026504788547754288,
+      "learning_rate": 0.0025209876543209877,
+      "loss": 0.4713,
+      "step": 329
+    },
+    {
+      "epoch": 0.7331296861982782,
+      "grad_norm": 0.002620173152536154,
+      "learning_rate": 0.0025185185185185185,
+      "loss": 0.5698,
+      "step": 330
+    },
+    {
+      "epoch": 0.7353512913079701,
+      "grad_norm": 0.003549919929355383,
+      "learning_rate": 0.0025160493827160496,
+      "loss": 0.5552,
+      "step": 331
+    },
+    {
+      "epoch": 0.7375728964176618,
+      "grad_norm": 0.003457101294770837,
+      "learning_rate": 0.0025135802469135804,
+      "loss": 0.6285,
+      "step": 332
+    },
+    {
+      "epoch": 0.7397945015273535,
+      "grad_norm": 0.003464099019765854,
+      "learning_rate": 0.002511111111111111,
+      "loss": 0.7362,
+      "step": 333
+    },
+    {
+      "epoch": 0.7420161066370453,
+      "grad_norm": 0.0033051518257707357,
+      "learning_rate": 0.0025086419753086423,
+      "loss": 0.6319,
+      "step": 334
+    },
+    {
+      "epoch": 0.744237711746737,
+      "grad_norm": 0.003773927688598633,
+      "learning_rate": 0.002506172839506173,
+      "loss": 0.7711,
+      "step": 335
+    },
+    {
+      "epoch": 0.7464593168564287,
+      "grad_norm": 0.0026653832755982876,
+      "learning_rate": 0.0025037037037037037,
+      "loss": 0.6425,
+      "step": 336
+    },
+    {
+      "epoch": 0.7486809219661206,
+      "grad_norm": 0.002547910436987877,
+      "learning_rate": 0.0025012345679012345,
+      "loss": 0.5432,
+      "step": 337
+    },
+    {
+      "epoch": 0.7509025270758123,
+      "grad_norm": 0.00405934639275074,
+      "learning_rate": 0.0024987654320987656,
+      "loss": 0.7786,
+      "step": 338
+    },
+    {
+      "epoch": 0.753124132185504,
+      "grad_norm": 0.00281044514849782,
+      "learning_rate": 0.0024962962962962964,
+      "loss": 0.6023,
+      "step": 339
+    },
+    {
+      "epoch": 0.7553457372951958,
+      "grad_norm": 0.002545573515817523,
+      "learning_rate": 0.002493827160493827,
+      "loss": 0.407,
+      "step": 340
+    },
+    {
+      "epoch": 0.7575673424048875,
+      "grad_norm": 0.002559161279350519,
+      "learning_rate": 0.0024913580246913583,
+      "loss": 0.6799,
+      "step": 341
+    },
+    {
+      "epoch": 0.7597889475145793,
+      "grad_norm": 0.003410926554352045,
+      "learning_rate": 0.002488888888888889,
+      "loss": 0.5536,
+      "step": 342
+    },
+    {
+      "epoch": 0.7620105526242711,
+      "grad_norm": 0.002621614607051015,
+      "learning_rate": 0.0024864197530864197,
+      "loss": 0.5806,
+      "step": 343
+    },
+    {
+      "epoch": 0.7642321577339628,
+      "grad_norm": 0.005297700874507427,
+      "learning_rate": 0.0024839506172839505,
+      "loss": 0.5808,
+      "step": 344
+    },
+    {
+      "epoch": 0.7664537628436545,
+      "grad_norm": 0.003013307461515069,
+      "learning_rate": 0.0024814814814814816,
+      "loss": 0.6203,
+      "step": 345
+    },
+    {
+      "epoch": 0.7686753679533463,
+      "grad_norm": 0.00272989459335804,
+      "learning_rate": 0.0024790123456790124,
+      "loss": 0.6735,
+      "step": 346
+    },
+    {
+      "epoch": 0.7708969730630381,
+      "grad_norm": 0.004028683062642813,
+      "learning_rate": 0.002476543209876543,
+      "loss": 0.5465,
+      "step": 347
+    },
+    {
+      "epoch": 0.7731185781727298,
+      "grad_norm": 0.0042002699337899685,
+      "learning_rate": 0.0024740740740740743,
+      "loss": 0.6278,
+      "step": 348
+    },
+    {
+      "epoch": 0.7753401832824216,
+      "grad_norm": 0.0058557018637657166,
+      "learning_rate": 0.002471604938271605,
+      "loss": 0.5699,
+      "step": 349
+    },
+    {
+      "epoch": 0.7775617883921133,
+      "grad_norm": 0.0032247588969767094,
+      "learning_rate": 0.0024691358024691358,
+      "loss": 0.558,
+      "step": 350
+    },
+    {
+      "epoch": 0.779783393501805,
+      "grad_norm": 0.0032824466470628977,
+      "learning_rate": 0.0024666666666666665,
+      "loss": 0.5787,
+      "step": 351
+    },
+    {
+      "epoch": 0.7820049986114969,
+      "grad_norm": 0.0036916215904057026,
+      "learning_rate": 0.0024641975308641977,
+      "loss": 0.5193,
+      "step": 352
+    },
+    {
+      "epoch": 0.7842266037211886,
+      "grad_norm": 0.0034254384227097034,
+      "learning_rate": 0.0024617283950617284,
+      "loss": 0.6091,
+      "step": 353
+    },
+    {
+      "epoch": 0.7864482088308803,
+      "grad_norm": 0.0031547548715025187,
+      "learning_rate": 0.002459259259259259,
+      "loss": 0.5296,
+      "step": 354
+    },
+    {
+      "epoch": 0.7886698139405721,
+      "grad_norm": 0.005953786429017782,
+      "learning_rate": 0.0024567901234567903,
+      "loss": 0.5461,
+      "step": 355
+    },
+    {
+      "epoch": 0.7908914190502638,
+      "grad_norm": 0.0031025484204292297,
+      "learning_rate": 0.002454320987654321,
+      "loss": 0.6348,
+      "step": 356
+    },
+    {
+      "epoch": 0.7931130241599555,
+      "grad_norm": 0.0038411449640989304,
+      "learning_rate": 0.002451851851851852,
+      "loss": 0.5176,
+      "step": 357
+    },
+    {
+      "epoch": 0.7953346292696474,
+      "grad_norm": 0.0034447757061570883,
+      "learning_rate": 0.0024493827160493825,
+      "loss": 0.5765,
+      "step": 358
+    },
+    {
+      "epoch": 0.7975562343793391,
+      "grad_norm": 0.0037516497541218996,
+      "learning_rate": 0.0024469135802469137,
+      "loss": 0.7173,
+      "step": 359
+    },
+    {
+      "epoch": 0.7997778394890308,
+      "grad_norm": 0.003684694878757,
+      "learning_rate": 0.0024444444444444444,
+      "loss": 0.5194,
+      "step": 360
+    },
+    {
+      "epoch": 0.8019994445987225,
+      "grad_norm": 0.006603560410439968,
+      "learning_rate": 0.0024419753086419756,
+      "loss": 0.6116,
+      "step": 361
+    },
+    {
+      "epoch": 0.8042210497084143,
+      "grad_norm": 0.003936069086194038,
+      "learning_rate": 0.0024395061728395063,
+      "loss": 0.6464,
+      "step": 362
+    },
+    {
+      "epoch": 0.8064426548181061,
+      "grad_norm": 0.007097030058503151,
+      "learning_rate": 0.002437037037037037,
+      "loss": 0.6659,
+      "step": 363
+    },
+    {
+      "epoch": 0.8086642599277978,
+      "grad_norm": 0.0037213300820440054,
+      "learning_rate": 0.002434567901234568,
+      "loss": 0.5492,
+      "step": 364
+    },
+    {
+      "epoch": 0.8108858650374896,
+      "grad_norm": 0.003655084175989032,
+      "learning_rate": 0.002432098765432099,
+      "loss": 0.5408,
+      "step": 365
+    },
+    {
+      "epoch": 0.8131074701471813,
+      "grad_norm": 0.0034437603317201138,
+      "learning_rate": 0.0024296296296296297,
+      "loss": 0.7443,
+      "step": 366
+    },
+    {
+      "epoch": 0.815329075256873,
+      "grad_norm": 0.007910266518592834,
+      "learning_rate": 0.0024271604938271604,
+      "loss": 0.6611,
+      "step": 367
+    },
+    {
+      "epoch": 0.8175506803665649,
+      "grad_norm": 0.00407488364726305,
+      "learning_rate": 0.0024246913580246916,
+      "loss": 0.4984,
+      "step": 368
+    },
+    {
+      "epoch": 0.8197722854762566,
+      "grad_norm": 0.004847107920795679,
+      "learning_rate": 0.0024222222222222223,
+      "loss": 0.697,
+      "step": 369
+    },
+    {
+      "epoch": 0.8219938905859483,
+      "grad_norm": 0.006795075722038746,
+      "learning_rate": 0.002419753086419753,
+      "loss": 0.639,
+      "step": 370
+    },
+    {
+      "epoch": 0.8242154956956401,
+      "grad_norm": 0.003597404109314084,
+      "learning_rate": 0.002417283950617284,
+      "loss": 0.5233,
+      "step": 371
+    },
+    {
+      "epoch": 0.8264371008053318,
+      "grad_norm": 0.0035218833945691586,
+      "learning_rate": 0.002414814814814815,
+      "loss": 0.6507,
+      "step": 372
+    },
+    {
+      "epoch": 0.8286587059150236,
+      "grad_norm": 0.0034329844638705254,
+      "learning_rate": 0.0024123456790123457,
+      "loss": 0.5302,
+      "step": 373
+    },
+    {
+      "epoch": 0.8308803110247154,
+      "grad_norm": 0.005119191948324442,
+      "learning_rate": 0.0024098765432098764,
+      "loss": 0.8082,
+      "step": 374
+    },
+    {
+      "epoch": 0.8331019161344071,
+      "grad_norm": 0.0064886524342000484,
+      "learning_rate": 0.0024074074074074076,
+      "loss": 0.7949,
+      "step": 375
+    },
+    {
+      "epoch": 0.8353235212440988,
+      "grad_norm": 0.004545222967863083,
+      "learning_rate": 0.0024049382716049383,
+      "loss": 0.6628,
+      "step": 376
+    },
+    {
+      "epoch": 0.8375451263537906,
+      "grad_norm": 0.003230273723602295,
+      "learning_rate": 0.002402469135802469,
+      "loss": 0.546,
+      "step": 377
+    },
+    {
+      "epoch": 0.8397667314634824,
+      "grad_norm": 0.00707286037504673,
+      "learning_rate": 0.0024000000000000002,
+      "loss": 0.8847,
+      "step": 378
+    },
+    {
+      "epoch": 0.8419883365731741,
+      "grad_norm": 0.003968073055148125,
+      "learning_rate": 0.002397530864197531,
+      "loss": 0.6247,
+      "step": 379
+    },
+    {
+      "epoch": 0.8442099416828659,
+      "grad_norm": 0.0037850975058972836,
+      "learning_rate": 0.0023950617283950617,
+      "loss": 0.481,
+      "step": 380
+    },
+    {
+      "epoch": 0.8464315467925576,
+      "grad_norm": 0.0032914101611822844,
+      "learning_rate": 0.0023925925925925924,
+      "loss": 0.6331,
+      "step": 381
+    },
+    {
+      "epoch": 0.8486531519022493,
+      "grad_norm": 0.005043490324169397,
+      "learning_rate": 0.0023901234567901236,
+      "loss": 0.5586,
+      "step": 382
+    },
+    {
+      "epoch": 0.8508747570119412,
+      "grad_norm": 0.0028147133998572826,
+      "learning_rate": 0.0023876543209876543,
+      "loss": 0.6125,
+      "step": 383
+    },
+    {
+      "epoch": 0.8530963621216329,
+      "grad_norm": 0.0038198321126401424,
+      "learning_rate": 0.002385185185185185,
+      "loss": 0.5318,
+      "step": 384
+    },
+    {
+      "epoch": 0.8553179672313246,
+      "grad_norm": 0.003384563373401761,
+      "learning_rate": 0.0023827160493827162,
+      "loss": 0.7372,
+      "step": 385
+    },
+    {
+      "epoch": 0.8575395723410164,
+      "grad_norm": 0.0064719198271632195,
+      "learning_rate": 0.002380246913580247,
+      "loss": 0.5661,
+      "step": 386
+    },
+    {
+      "epoch": 0.8597611774507081,
+      "grad_norm": 0.0028373387176543474,
+      "learning_rate": 0.002377777777777778,
+      "loss": 0.577,
+      "step": 387
+    },
+    {
+      "epoch": 0.8619827825603998,
+      "grad_norm": 0.0027057346887886524,
+      "learning_rate": 0.0023753086419753084,
+      "loss": 0.6654,
+      "step": 388
+    },
+    {
+      "epoch": 0.8642043876700917,
+      "grad_norm": 0.0032722018659114838,
+      "learning_rate": 0.0023728395061728396,
+      "loss": 0.5527,
+      "step": 389
+    },
+    {
+      "epoch": 0.8664259927797834,
+      "grad_norm": 0.004043676424771547,
+      "learning_rate": 0.0023703703703703703,
+      "loss": 0.6648,
+      "step": 390
+    },
+    {
+      "epoch": 0.8686475978894751,
+      "grad_norm": 0.0032604134175926447,
+      "learning_rate": 0.0023679012345679015,
+      "loss": 0.5295,
+      "step": 391
+    },
+    {
+      "epoch": 0.8708692029991669,
+      "grad_norm": 0.0035888762213289738,
+      "learning_rate": 0.0023654320987654322,
+      "loss": 0.6486,
+      "step": 392
+    },
+    {
+      "epoch": 0.8730908081088586,
+      "grad_norm": 0.009759027510881424,
+      "learning_rate": 0.002362962962962963,
+      "loss": 0.5339,
+      "step": 393
+    },
+    {
+      "epoch": 0.8753124132185504,
+      "grad_norm": 0.0034896486904472113,
+      "learning_rate": 0.002360493827160494,
+      "loss": 0.8196,
+      "step": 394
+    },
+    {
+      "epoch": 0.8775340183282422,
+      "grad_norm": 0.0029943559784442186,
+      "learning_rate": 0.002358024691358025,
+      "loss": 0.5729,
+      "step": 395
+    },
+    {
+      "epoch": 0.8797556234379339,
+      "grad_norm": 0.0038155601359903812,
+      "learning_rate": 0.0023555555555555556,
+      "loss": 0.5233,
+      "step": 396
+    },
+    {
+      "epoch": 0.8819772285476256,
+      "grad_norm": 0.0032906325068324804,
+      "learning_rate": 0.0023530864197530863,
+      "loss": 0.5703,
+      "step": 397
+    },
+    {
+      "epoch": 0.8841988336573174,
+      "grad_norm": 0.008255512453615665,
+      "learning_rate": 0.0023506172839506175,
+      "loss": 0.6795,
+      "step": 398
+    },
+    {
+      "epoch": 0.8864204387670092,
+      "grad_norm": 0.003714282065629959,
+      "learning_rate": 0.002348148148148148,
+      "loss": 0.5704,
+      "step": 399
+    },
+    {
+      "epoch": 0.8886420438767009,
+      "grad_norm": 0.004658196121454239,
+      "learning_rate": 0.002345679012345679,
+      "loss": 0.6216,
+      "step": 400
+    },
+    {
+      "epoch": 0.8908636489863927,
+      "grad_norm": 0.003137049498036504,
+      "learning_rate": 0.00234320987654321,
+      "loss": 0.6422,
+      "step": 401
+    },
+    {
+      "epoch": 0.8930852540960844,
+      "grad_norm": 0.004380072932690382,
+      "learning_rate": 0.002340740740740741,
+      "loss": 0.6654,
+      "step": 402
+    },
+    {
+      "epoch": 0.8953068592057761,
+      "grad_norm": 0.0035470370203256607,
+      "learning_rate": 0.0023382716049382716,
+      "loss": 0.5809,
+      "step": 403
+    },
+    {
+      "epoch": 0.897528464315468,
+      "grad_norm": 0.012754486873745918,
+      "learning_rate": 0.0023358024691358024,
+      "loss": 0.5879,
+      "step": 404
+    },
+    {
+      "epoch": 0.8997500694251597,
+      "grad_norm": 0.003565397346392274,
+      "learning_rate": 0.0023333333333333335,
+      "loss": 0.5179,
+      "step": 405
+    },
+    {
+      "epoch": 0.9019716745348514,
+      "grad_norm": 0.004204631317406893,
+      "learning_rate": 0.0023308641975308643,
+      "loss": 0.7359,
+      "step": 406
+    },
+    {
+      "epoch": 0.9041932796445432,
+      "grad_norm": 0.002687248168513179,
+      "learning_rate": 0.002328395061728395,
+      "loss": 0.6678,
+      "step": 407
+    },
+    {
+      "epoch": 0.9064148847542349,
+      "grad_norm": 0.004071474075317383,
+      "learning_rate": 0.002325925925925926,
+      "loss": 0.5053,
+      "step": 408
+    },
+    {
+      "epoch": 0.9086364898639266,
+      "grad_norm": 0.004420380108058453,
+      "learning_rate": 0.002323456790123457,
+      "loss": 0.7788,
+      "step": 409
+    },
+    {
+      "epoch": 0.9108580949736185,
+      "grad_norm": 0.004801613744348288,
+      "learning_rate": 0.0023209876543209876,
+      "loss": 0.6536,
+      "step": 410
+    },
+    {
+      "epoch": 0.9130797000833102,
+      "grad_norm": 0.00353393261320889,
+      "learning_rate": 0.0023185185185185184,
+      "loss": 0.4745,
+      "step": 411
+    },
+    {
+      "epoch": 0.9153013051930019,
+      "grad_norm": 0.005573590286076069,
+      "learning_rate": 0.0023160493827160495,
+      "loss": 0.6881,
+      "step": 412
+    },
+    {
+      "epoch": 0.9175229103026937,
+      "grad_norm": 0.0035207164473831654,
+      "learning_rate": 0.0023135802469135803,
+      "loss": 0.6957,
+      "step": 413
+    },
+    {
+      "epoch": 0.9197445154123854,
+      "grad_norm": 0.003412191988900304,
+      "learning_rate": 0.002311111111111111,
+      "loss": 0.6391,
+      "step": 414
+    },
+    {
+      "epoch": 0.9219661205220772,
+      "grad_norm": 0.004509443417191505,
+      "learning_rate": 0.002308641975308642,
+      "loss": 0.6225,
+      "step": 415
+    },
+    {
+      "epoch": 0.924187725631769,
+      "grad_norm": 0.0027398800011724234,
+      "learning_rate": 0.002306172839506173,
+      "loss": 0.4486,
+      "step": 416
+    },
+    {
+      "epoch": 0.9264093307414607,
+      "grad_norm": 0.002736851805821061,
+      "learning_rate": 0.0023037037037037036,
+      "loss": 0.5405,
+      "step": 417
+    },
+    {
+      "epoch": 0.9286309358511524,
+      "grad_norm": 0.00329567096196115,
+      "learning_rate": 0.0023012345679012344,
+      "loss": 0.5796,
+      "step": 418
+    },
+    {
+      "epoch": 0.9308525409608442,
+      "grad_norm": 0.0037880768068134785,
+      "learning_rate": 0.0022987654320987655,
+      "loss": 0.5197,
+      "step": 419
+    },
+    {
+      "epoch": 0.933074146070536,
+      "grad_norm": 0.0035441648215055466,
+      "learning_rate": 0.0022962962962962963,
+      "loss": 0.6808,
+      "step": 420
+    },
+    {
+      "epoch": 0.9352957511802277,
+      "grad_norm": 0.00435903575271368,
+      "learning_rate": 0.002293827160493827,
+      "loss": 0.65,
+      "step": 421
+    },
+    {
+      "epoch": 0.9375173562899195,
+      "grad_norm": 0.003232384566217661,
+      "learning_rate": 0.002291358024691358,
+      "loss": 0.6367,
+      "step": 422
+    },
+    {
+      "epoch": 0.9397389613996112,
+      "grad_norm": 0.006525840610265732,
+      "learning_rate": 0.002288888888888889,
+      "loss": 0.634,
+      "step": 423
+    },
+    {
+      "epoch": 0.9419605665093029,
+      "grad_norm": 0.004283812828361988,
+      "learning_rate": 0.00228641975308642,
+      "loss": 0.6266,
+      "step": 424
+    },
+    {
+      "epoch": 0.9441821716189948,
+      "grad_norm": 0.00269131432287395,
+      "learning_rate": 0.0022839506172839504,
+      "loss": 0.6423,
+      "step": 425
+    },
+    {
+      "epoch": 0.9464037767286865,
+      "grad_norm": 0.005134789738804102,
+      "learning_rate": 0.0022814814814814816,
+      "loss": 0.741,
+      "step": 426
+    },
+    {
+      "epoch": 0.9486253818383782,
+      "grad_norm": 0.003960168920457363,
+      "learning_rate": 0.0022790123456790123,
+      "loss": 0.653,
+      "step": 427
+    },
+    {
+      "epoch": 0.95084698694807,
+      "grad_norm": 0.005250333808362484,
+      "learning_rate": 0.0022765432098765435,
+      "loss": 0.5763,
+      "step": 428
+    },
+    {
+      "epoch": 0.9530685920577617,
+      "grad_norm": 0.002757916459813714,
+      "learning_rate": 0.0022740740740740738,
+      "loss": 0.4882,
+      "step": 429
+    },
+    {
+      "epoch": 0.9552901971674534,
+      "grad_norm": 0.0023539310786873102,
+      "learning_rate": 0.002271604938271605,
+      "loss": 0.5957,
+      "step": 430
+    },
+    {
+      "epoch": 0.9575118022771453,
+      "grad_norm": 0.00454498128965497,
+      "learning_rate": 0.002269135802469136,
+      "loss": 0.634,
+      "step": 431
+    },
+    {
+      "epoch": 0.959733407386837,
+      "grad_norm": 0.0027770919259637594,
+      "learning_rate": 0.002266666666666667,
+      "loss": 0.6027,
+      "step": 432
+    },
+    {
+      "epoch": 0.9619550124965287,
+      "grad_norm": 0.0041671195067465305,
+      "learning_rate": 0.0022641975308641976,
+      "loss": 0.5002,
+      "step": 433
+    },
+    {
+      "epoch": 0.9641766176062205,
+      "grad_norm": 0.003652903251349926,
+      "learning_rate": 0.0022617283950617283,
+      "loss": 0.5282,
+      "step": 434
+    },
+    {
+      "epoch": 0.9663982227159122,
+      "grad_norm": 0.002368962625041604,
+      "learning_rate": 0.0022592592592592595,
+      "loss": 0.5286,
+      "step": 435
+    },
+    {
+      "epoch": 0.968619827825604,
+      "grad_norm": 0.0032470771111547947,
+      "learning_rate": 0.00225679012345679,
+      "loss": 0.5601,
+      "step": 436
+    },
+    {
+      "epoch": 0.9708414329352958,
+      "grad_norm": 0.0038956725038588047,
+      "learning_rate": 0.002254320987654321,
+      "loss": 0.669,
+      "step": 437
+    },
+    {
+      "epoch": 0.9730630380449875,
+      "grad_norm": 0.0024040553253144026,
+      "learning_rate": 0.002251851851851852,
+      "loss": 0.6164,
+      "step": 438
+    },
+    {
+      "epoch": 0.9752846431546792,
+      "grad_norm": 0.0025828545913100243,
+      "learning_rate": 0.002249382716049383,
+      "loss": 0.6219,
+      "step": 439
+    },
+    {
+      "epoch": 0.977506248264371,
+      "grad_norm": 0.005943648051470518,
+      "learning_rate": 0.0022469135802469136,
+      "loss": 0.5701,
+      "step": 440
+    },
+    {
+      "epoch": 0.9797278533740628,
+      "grad_norm": 0.0035154123324900866,
+      "learning_rate": 0.0022444444444444443,
+      "loss": 0.6773,
+      "step": 441
+    },
+    {
+      "epoch": 0.9819494584837545,
+      "grad_norm": 0.0027923276647925377,
+      "learning_rate": 0.0022419753086419755,
+      "loss": 0.6962,
+      "step": 442
+    },
+    {
+      "epoch": 0.9841710635934463,
+      "grad_norm": 0.0050351908430457115,
+      "learning_rate": 0.002239506172839506,
+      "loss": 0.5091,
+      "step": 443
+    },
+    {
+      "epoch": 0.986392668703138,
+      "grad_norm": 0.010067122988402843,
+      "learning_rate": 0.002237037037037037,
+      "loss": 0.5033,
+      "step": 444
+    },
+    {
+      "epoch": 0.9886142738128297,
+      "grad_norm": 0.003369242651388049,
+      "learning_rate": 0.002234567901234568,
+      "loss": 0.4704,
+      "step": 445
+    },
+    {
+      "epoch": 0.9908358789225216,
+      "grad_norm": 0.002988790860399604,
+      "learning_rate": 0.002232098765432099,
+      "loss": 0.7471,
+      "step": 446
+    },
+    {
+      "epoch": 0.9930574840322133,
+      "grad_norm": 0.003928947728127241,
+      "learning_rate": 0.0022296296296296296,
+      "loss": 0.7318,
+      "step": 447
+    },
+    {
+      "epoch": 0.995279089141905,
+      "grad_norm": 0.0032487758435308933,
+      "learning_rate": 0.0022271604938271603,
+      "loss": 0.6627,
+      "step": 448
+    },
+    {
+      "epoch": 0.9975006942515968,
+      "grad_norm": 0.0050149112939834595,
+      "learning_rate": 0.0022246913580246915,
+      "loss": 0.7153,
+      "step": 449
+    },
+    {
+      "epoch": 0.9997222993612885,
+      "grad_norm": 0.002519611269235611,
+      "learning_rate": 0.0022222222222222222,
+      "loss": 0.6003,
+      "step": 450
+    },
+    {
+      "epoch": 1.0019439044709804,
+      "grad_norm": 0.0036871463526040316,
+      "learning_rate": 0.002219753086419753,
+      "loss": 0.7394,
+      "step": 451
+    },
+    {
+      "epoch": 1.004165509580672,
+      "grad_norm": 0.003514339914545417,
+      "learning_rate": 0.002217283950617284,
+      "loss": 0.5648,
+      "step": 452
+    },
+    {
+      "epoch": 1.0063871146903638,
+      "grad_norm": 0.0024536640848964453,
+      "learning_rate": 0.002214814814814815,
+      "loss": 0.5651,
+      "step": 453
+    },
+    {
+      "epoch": 1.0086087198000555,
+      "grad_norm": 0.0029373124707490206,
+      "learning_rate": 0.002212345679012346,
+      "loss": 0.5751,
+      "step": 454
+    },
+    {
+      "epoch": 1.0108303249097472,
+      "grad_norm": 0.004164501093327999,
+      "learning_rate": 0.0022098765432098763,
+      "loss": 0.542,
+      "step": 455
+    },
+    {
+      "epoch": 1.013051930019439,
+      "grad_norm": 0.002176556270569563,
+      "learning_rate": 0.0022074074074074075,
+      "loss": 0.5003,
+      "step": 456
+    },
+    {
+      "epoch": 1.0152735351291309,
+      "grad_norm": 0.002453757217153907,
+      "learning_rate": 0.0022049382716049382,
+      "loss": 0.5938,
+      "step": 457
+    },
+    {
+      "epoch": 1.0174951402388226,
+      "grad_norm": 0.002585896523669362,
+      "learning_rate": 0.0022024691358024694,
+      "loss": 0.4372,
+      "step": 458
+    },
+    {
+      "epoch": 1.0197167453485143,
+      "grad_norm": 0.0036106205079704523,
+      "learning_rate": 0.0021999999999999997,
+      "loss": 0.4973,
+      "step": 459
+    },
+    {
+      "epoch": 1.021938350458206,
+      "grad_norm": 0.010448912158608437,
+      "learning_rate": 0.002197530864197531,
+      "loss": 0.6627,
+      "step": 460
+    },
+    {
+      "epoch": 1.0241599555678977,
+      "grad_norm": 0.0033705660607665777,
+      "learning_rate": 0.002195061728395062,
+      "loss": 0.6634,
+      "step": 461
+    },
+    {
+      "epoch": 1.0263815606775895,
+      "grad_norm": 0.004475906956940889,
+      "learning_rate": 0.0021925925925925928,
+      "loss": 0.6017,
+      "step": 462
+    },
+    {
+      "epoch": 1.0286031657872814,
+      "grad_norm": 0.002899282844737172,
+      "learning_rate": 0.0021901234567901235,
+      "loss": 0.6399,
+      "step": 463
+    },
+    {
+      "epoch": 1.030824770896973,
+      "grad_norm": 0.002981000579893589,
+      "learning_rate": 0.0021876543209876542,
+      "loss": 0.6914,
+      "step": 464
+    },
+    {
+      "epoch": 1.0330463760066648,
+      "grad_norm": 0.0031841108575463295,
+      "learning_rate": 0.0021851851851851854,
+      "loss": 0.6894,
+      "step": 465
+    },
+    {
+      "epoch": 1.0352679811163565,
+      "grad_norm": 0.00265736342407763,
+      "learning_rate": 0.002182716049382716,
+      "loss": 0.4852,
+      "step": 466
+    },
+    {
+      "epoch": 1.0374895862260483,
+      "grad_norm": 0.003279666183516383,
+      "learning_rate": 0.002180246913580247,
+      "loss": 0.5547,
+      "step": 467
+    },
+    {
+      "epoch": 1.03971119133574,
+      "grad_norm": 0.002579506253823638,
+      "learning_rate": 0.002177777777777778,
+      "loss": 0.5056,
+      "step": 468
+    },
+    {
+      "epoch": 1.041932796445432,
+      "grad_norm": 0.0029568190220743418,
+      "learning_rate": 0.0021753086419753088,
+      "loss": 0.6682,
+      "step": 469
+    },
+    {
+      "epoch": 1.0441544015551236,
+      "grad_norm": 0.0029284320771694183,
+      "learning_rate": 0.0021728395061728395,
+      "loss": 0.5206,
+      "step": 470
+    },
+    {
+      "epoch": 1.0463760066648153,
+      "grad_norm": 0.003473050892353058,
+      "learning_rate": 0.0021703703703703702,
+      "loss": 0.6092,
+      "step": 471
+    },
+    {
+      "epoch": 1.048597611774507,
+      "grad_norm": 0.003467805916443467,
+      "learning_rate": 0.0021679012345679014,
+      "loss": 0.7541,
+      "step": 472
+    },
+    {
+      "epoch": 1.0508192168841988,
+      "grad_norm": 0.0023657323326915503,
+      "learning_rate": 0.002165432098765432,
+      "loss": 0.588,
+      "step": 473
+    },
+    {
+      "epoch": 1.0530408219938905,
+      "grad_norm": 0.003372120438143611,
+      "learning_rate": 0.002162962962962963,
+      "loss": 0.5177,
+      "step": 474
+    },
+    {
+      "epoch": 1.0552624271035824,
+      "grad_norm": 0.002507074037566781,
+      "learning_rate": 0.002160493827160494,
+      "loss": 0.5315,
+      "step": 475
+    },
+    {
+      "epoch": 1.0574840322132741,
+      "grad_norm": 0.003847063286229968,
+      "learning_rate": 0.002158024691358025,
+      "loss": 0.6434,
+      "step": 476
+    },
+    {
+      "epoch": 1.0597056373229659,
+      "grad_norm": 0.004577492829412222,
+      "learning_rate": 0.0021555555555555555,
+      "loss": 0.6605,
+      "step": 477
+    },
+    {
+      "epoch": 1.0619272424326576,
+      "grad_norm": 0.00435985391959548,
+      "learning_rate": 0.0021530864197530863,
+      "loss": 0.4803,
+      "step": 478
+    },
+    {
+      "epoch": 1.0641488475423493,
+      "grad_norm": 0.0066704158671200275,
+      "learning_rate": 0.0021506172839506174,
+      "loss": 0.4825,
+      "step": 479
+    },
+    {
+      "epoch": 1.066370452652041,
+      "grad_norm": 0.0034896789584308863,
+      "learning_rate": 0.002148148148148148,
+      "loss": 0.4425,
+      "step": 480
+    },
+    {
+      "epoch": 1.068592057761733,
+      "grad_norm": 0.0028791052754968405,
+      "learning_rate": 0.002145679012345679,
+      "loss": 0.6792,
+      "step": 481
+    },
+    {
+      "epoch": 1.0708136628714247,
+      "grad_norm": 0.0027687002439051867,
+      "learning_rate": 0.0021432098765432096,
+      "loss": 0.6594,
+      "step": 482
+    },
+    {
+      "epoch": 1.0730352679811164,
+      "grad_norm": 0.004298859275877476,
+      "learning_rate": 0.002140740740740741,
+      "loss": 0.6979,
+      "step": 483
+    },
+    {
+      "epoch": 1.075256873090808,
+      "grad_norm": 0.0024634443689137697,
+      "learning_rate": 0.002138271604938272,
+      "loss": 0.7344,
+      "step": 484
+    },
+    {
+      "epoch": 1.0774784782004998,
+      "grad_norm": 0.0029783491045236588,
+      "learning_rate": 0.0021358024691358023,
+      "loss": 0.7162,
+      "step": 485
+    },
+    {
+      "epoch": 1.0797000833101915,
+      "grad_norm": 0.003992694895714521,
+      "learning_rate": 0.0021333333333333334,
+      "loss": 0.6333,
+      "step": 486
+    },
+    {
+      "epoch": 1.0819216884198835,
+      "grad_norm": 0.0027143885381519794,
+      "learning_rate": 0.002130864197530864,
+      "loss": 0.7107,
+      "step": 487
+    },
+    {
+      "epoch": 1.0841432935295752,
+      "grad_norm": 0.002360287122428417,
+      "learning_rate": 0.0021283950617283953,
+      "loss": 0.6408,
+      "step": 488
+    },
+    {
+      "epoch": 1.0863648986392669,
+      "grad_norm": 0.0030325008556246758,
+      "learning_rate": 0.0021259259259259256,
+      "loss": 0.6446,
+      "step": 489
+    },
+    {
+      "epoch": 1.0885865037489586,
+      "grad_norm": 0.003571762004867196,
+      "learning_rate": 0.002123456790123457,
+      "loss": 0.4972,
+      "step": 490
+    },
+    {
+      "epoch": 1.0908081088586503,
+      "grad_norm": 0.0028126256074756384,
+      "learning_rate": 0.002120987654320988,
+      "loss": 0.4611,
+      "step": 491
+    },
+    {
+      "epoch": 1.093029713968342,
+      "grad_norm": 0.0031972727738320827,
+      "learning_rate": 0.0021185185185185187,
+      "loss": 0.5785,
+      "step": 492
+    },
+    {
+      "epoch": 1.095251319078034,
+      "grad_norm": 0.0038447740953415632,
+      "learning_rate": 0.0021160493827160494,
+      "loss": 0.7667,
+      "step": 493
+    },
+    {
+      "epoch": 1.0974729241877257,
+      "grad_norm": 0.002788472454994917,
+      "learning_rate": 0.00211358024691358,
+      "loss": 0.6148,
+      "step": 494
+    },
+    {
+      "epoch": 1.0996945292974174,
+      "grad_norm": 0.003124644048511982,
+      "learning_rate": 0.0021111111111111113,
+      "loss": 0.7035,
+      "step": 495
+    },
+    {
+      "epoch": 1.1019161344071091,
+      "grad_norm": 0.003380119800567627,
+      "learning_rate": 0.002108641975308642,
+      "loss": 0.5542,
+      "step": 496
+    },
+    {
+      "epoch": 1.1041377395168008,
+      "grad_norm": 0.004016113001853228,
+      "learning_rate": 0.002106172839506173,
+      "loss": 0.5868,
+      "step": 497
+    },
+    {
+      "epoch": 1.1063593446264925,
+      "grad_norm": 0.003642839379608631,
+      "learning_rate": 0.002103703703703704,
+      "loss": 0.6001,
+      "step": 498
+    },
+    {
+      "epoch": 1.1085809497361845,
+      "grad_norm": 0.002972975606098771,
+      "learning_rate": 0.0021012345679012347,
+      "loss": 0.5375,
+      "step": 499
+    },
+    {
+      "epoch": 1.1108025548458762,
+      "grad_norm": 0.0027982559986412525,
+      "learning_rate": 0.0020987654320987655,
+      "loss": 0.6106,
+      "step": 500
+    },
+    {
+      "epoch": 1.113024159955568,
+      "grad_norm": 0.003983978182077408,
+      "learning_rate": 0.002096296296296296,
+      "loss": 0.6312,
+      "step": 501
+    },
+    {
+      "epoch": 1.1152457650652596,
+      "grad_norm": 0.0028390882071107626,
+      "learning_rate": 0.0020938271604938274,
+      "loss": 0.6555,
+      "step": 502
+    },
+    {
+      "epoch": 1.1174673701749513,
+      "grad_norm": 0.006304751615971327,
+      "learning_rate": 0.002091358024691358,
+      "loss": 0.6585,
+      "step": 503
+    },
+    {
+      "epoch": 1.119688975284643,
+      "grad_norm": 0.003519695485010743,
+      "learning_rate": 0.002088888888888889,
+      "loss": 0.5764,
+      "step": 504
+    },
+    {
+      "epoch": 1.121910580394335,
+      "grad_norm": 0.002687301719561219,
+      "learning_rate": 0.00208641975308642,
+      "loss": 0.6775,
+      "step": 505
+    },
+    {
+      "epoch": 1.1241321855040267,
+      "grad_norm": 0.004116454627364874,
+      "learning_rate": 0.0020839506172839507,
+      "loss": 0.4918,
+      "step": 506
+    },
+    {
+      "epoch": 1.1263537906137184,
+      "grad_norm": 0.0029182173311710358,
+      "learning_rate": 0.0020814814814814815,
+      "loss": 0.6922,
+      "step": 507
+    },
+    {
+      "epoch": 1.1285753957234101,
+      "grad_norm": 0.0036673294380307198,
+      "learning_rate": 0.002079012345679012,
+      "loss": 0.5987,
+      "step": 508
+    },
+    {
+      "epoch": 1.1307970008331019,
+      "grad_norm": 0.005055381450802088,
+      "learning_rate": 0.0020765432098765434,
+      "loss": 0.547,
+      "step": 509
+    },
+    {
+      "epoch": 1.1330186059427936,
+      "grad_norm": 0.003081630915403366,
+      "learning_rate": 0.002074074074074074,
+      "loss": 0.7392,
+      "step": 510
+    },
+    {
+      "epoch": 1.1352402110524855,
+      "grad_norm": 0.003554833820089698,
+      "learning_rate": 0.002071604938271605,
+      "loss": 0.666,
+      "step": 511
+    },
+    {
+      "epoch": 1.1374618161621772,
+      "grad_norm": 0.0036279240157455206,
+      "learning_rate": 0.0020691358024691356,
+      "loss": 0.5381,
+      "step": 512
+    },
+    {
+      "epoch": 1.139683421271869,
+      "grad_norm": 0.004544177558273077,
+      "learning_rate": 0.0020666666666666667,
+      "loss": 0.4533,
+      "step": 513
+    },
+    {
+      "epoch": 1.1419050263815607,
+      "grad_norm": 0.003769755130633712,
+      "learning_rate": 0.002064197530864198,
+      "loss": 0.7778,
+      "step": 514
+    },
+    {
+      "epoch": 1.1441266314912524,
+      "grad_norm": 0.0022846919018775225,
+      "learning_rate": 0.002061728395061728,
+      "loss": 0.567,
+      "step": 515
+    },
+    {
+      "epoch": 1.146348236600944,
+      "grad_norm": 0.0027041470166295767,
+      "learning_rate": 0.0020592592592592594,
+      "loss": 0.5174,
+      "step": 516
+    },
+    {
+      "epoch": 1.148569841710636,
+      "grad_norm": 0.003274301066994667,
+      "learning_rate": 0.00205679012345679,
+      "loss": 0.6534,
+      "step": 517
+    },
+    {
+      "epoch": 1.1507914468203277,
+      "grad_norm": 0.0030452990904450417,
+      "learning_rate": 0.0020543209876543213,
+      "loss": 0.5024,
+      "step": 518
+    },
+    {
+      "epoch": 1.1530130519300195,
+      "grad_norm": 0.004656861070543528,
+      "learning_rate": 0.0020518518518518516,
+      "loss": 0.5775,
+      "step": 519
+    },
+    {
+      "epoch": 1.1552346570397112,
+      "grad_norm": 0.0048223827034235,
+      "learning_rate": 0.0020493827160493827,
+      "loss": 0.5477,
+      "step": 520
+    },
+    {
+      "epoch": 1.157456262149403,
+      "grad_norm": 0.0036563219036906958,
+      "learning_rate": 0.002046913580246914,
+      "loss": 0.7265,
+      "step": 521
+    },
+    {
+      "epoch": 1.1596778672590946,
+      "grad_norm": 0.002614110242575407,
+      "learning_rate": 0.0020444444444444447,
+      "loss": 0.539,
+      "step": 522
+    },
+    {
+      "epoch": 1.1618994723687863,
+      "grad_norm": 0.006397682707756758,
+      "learning_rate": 0.0020419753086419754,
+      "loss": 0.727,
+      "step": 523
+    },
+    {
+      "epoch": 1.1641210774784783,
+      "grad_norm": 0.006625836715102196,
+      "learning_rate": 0.002039506172839506,
+      "loss": 0.4989,
+      "step": 524
+    },
+    {
+      "epoch": 1.16634268258817,
+      "grad_norm": 0.0033377420622855425,
+      "learning_rate": 0.0020370370370370373,
+      "loss": 0.5469,
+      "step": 525
+    },
+    {
+      "epoch": 1.1685642876978617,
+      "grad_norm": 0.0037035797722637653,
+      "learning_rate": 0.002034567901234568,
+      "loss": 0.6719,
+      "step": 526
+    },
+    {
+      "epoch": 1.1707858928075534,
+      "grad_norm": 0.0035533548798412085,
+      "learning_rate": 0.0020320987654320988,
+      "loss": 0.4496,
+      "step": 527
+    },
+    {
+      "epoch": 1.1730074979172451,
+      "grad_norm": 0.0026416785549372435,
+      "learning_rate": 0.00202962962962963,
+      "loss": 0.5356,
+      "step": 528
+    },
+    {
+      "epoch": 1.175229103026937,
+      "grad_norm": 0.0026185268070548773,
+      "learning_rate": 0.0020271604938271607,
+      "loss": 0.6109,
+      "step": 529
+    },
+    {
+      "epoch": 1.1774507081366288,
+      "grad_norm": 0.003254238050431013,
+      "learning_rate": 0.0020246913580246914,
+      "loss": 0.6925,
+      "step": 530
+    },
+    {
+      "epoch": 1.1796723132463205,
+      "grad_norm": 0.006051741540431976,
+      "learning_rate": 0.002022222222222222,
+      "loss": 0.5446,
+      "step": 531
+    },
+    {
+      "epoch": 1.1818939183560122,
+      "grad_norm": 0.0027108059730380774,
+      "learning_rate": 0.0020197530864197533,
+      "loss": 0.6273,
+      "step": 532
+    },
+    {
+      "epoch": 1.184115523465704,
+      "grad_norm": 0.0033522590529173613,
+      "learning_rate": 0.002017283950617284,
+      "loss": 0.6232,
+      "step": 533
+    },
+    {
+      "epoch": 1.1863371285753956,
+      "grad_norm": 0.0041407193057239056,
+      "learning_rate": 0.0020148148148148148,
+      "loss": 0.5172,
+      "step": 534
+    },
+    {
+      "epoch": 1.1885587336850874,
+      "grad_norm": 0.0039221737533807755,
+      "learning_rate": 0.002012345679012346,
+      "loss": 0.5179,
+      "step": 535
+    },
+    {
+      "epoch": 1.1907803387947793,
+      "grad_norm": 0.0050890520215034485,
+      "learning_rate": 0.0020098765432098767,
+      "loss": 0.6051,
+      "step": 536
+    },
+    {
+      "epoch": 1.193001943904471,
+      "grad_norm": 0.0036343582905828953,
+      "learning_rate": 0.0020074074074074074,
+      "loss": 0.6176,
+      "step": 537
+    },
+    {
+      "epoch": 1.1952235490141627,
+      "grad_norm": 0.0032197320833802223,
+      "learning_rate": 0.002004938271604938,
+      "loss": 0.5136,
+      "step": 538
+    },
+    {
+      "epoch": 1.1974451541238544,
+      "grad_norm": 0.0023664599284529686,
+      "learning_rate": 0.0020024691358024693,
+      "loss": 0.5228,
+      "step": 539
+    },
+    {
+      "epoch": 1.1996667592335462,
+      "grad_norm": 0.002611706266179681,
+      "learning_rate": 0.002,
+      "loss": 0.5393,
+      "step": 540
+    },
+    {
+      "epoch": 1.201888364343238,
+      "grad_norm": 0.002908397000283003,
+      "learning_rate": 0.0019975308641975308,
+      "loss": 0.5602,
+      "step": 541
+    },
+    {
+      "epoch": 1.2041099694529298,
+      "grad_norm": 0.0035483166575431824,
+      "learning_rate": 0.0019950617283950615,
+      "loss": 0.5777,
+      "step": 542
+    },
+    {
+      "epoch": 1.2063315745626215,
+      "grad_norm": 0.002436819253489375,
+      "learning_rate": 0.0019925925925925927,
+      "loss": 0.6247,
+      "step": 543
+    },
+    {
+      "epoch": 1.2085531796723132,
+      "grad_norm": 0.00271439366042614,
+      "learning_rate": 0.001990123456790124,
+      "loss": 0.634,
+      "step": 544
+    },
+    {
+      "epoch": 1.210774784782005,
+      "grad_norm": 0.0029384950175881386,
+      "learning_rate": 0.001987654320987654,
+      "loss": 0.5216,
+      "step": 545
+    },
+    {
+      "epoch": 1.2129963898916967,
+      "grad_norm": 0.0022015413269400597,
+      "learning_rate": 0.0019851851851851853,
+      "loss": 0.6649,
+      "step": 546
+    },
+    {
+      "epoch": 1.2152179950013884,
+      "grad_norm": 0.00338186277076602,
+      "learning_rate": 0.001982716049382716,
+      "loss": 0.5563,
+      "step": 547
+    },
+    {
+      "epoch": 1.2174396001110803,
+      "grad_norm": 0.002683288650587201,
+      "learning_rate": 0.001980246913580247,
+      "loss": 0.6024,
+      "step": 548
+    },
+    {
+      "epoch": 1.219661205220772,
+      "grad_norm": 0.002985814120620489,
+      "learning_rate": 0.0019777777777777775,
+      "loss": 0.4619,
+      "step": 549
+    },
+    {
+      "epoch": 1.2218828103304638,
+      "grad_norm": 0.0031337959226220846,
+      "learning_rate": 0.0019753086419753087,
+      "loss": 0.5346,
+      "step": 550
+    },
+    {
+      "epoch": 1.2241044154401555,
+      "grad_norm": 0.010030501522123814,
+      "learning_rate": 0.00197283950617284,
+      "loss": 0.6331,
+      "step": 551
+    },
+    {
+      "epoch": 1.2263260205498472,
+      "grad_norm": 0.0031471492256969213,
+      "learning_rate": 0.00197037037037037,
+      "loss": 0.5195,
+      "step": 552
+    },
+    {
+      "epoch": 1.2285476256595391,
+      "grad_norm": 0.0028481080662459135,
+      "learning_rate": 0.0019679012345679013,
+      "loss": 0.4754,
+      "step": 553
+    },
+    {
+      "epoch": 1.2307692307692308,
+      "grad_norm": 0.004545793868601322,
+      "learning_rate": 0.001965432098765432,
+      "loss": 0.6763,
+      "step": 554
+    },
+    {
+      "epoch": 1.2329908358789226,
+      "grad_norm": 0.002784636802971363,
+      "learning_rate": 0.0019629629629629632,
+      "loss": 0.5884,
+      "step": 555
+    },
+    {
+      "epoch": 1.2352124409886143,
+      "grad_norm": 0.0035187830217182636,
+      "learning_rate": 0.0019604938271604935,
+      "loss": 0.6781,
+      "step": 556
+    },
+    {
+      "epoch": 1.237434046098306,
+      "grad_norm": 0.003770441748201847,
+      "learning_rate": 0.0019580246913580247,
+      "loss": 0.6714,
+      "step": 557
+    },
+    {
+      "epoch": 1.2396556512079977,
+      "grad_norm": 0.002745373174548149,
+      "learning_rate": 0.001955555555555556,
+      "loss": 0.6145,
+      "step": 558
+    },
+    {
+      "epoch": 1.2418772563176894,
+      "grad_norm": 0.003247046610340476,
+      "learning_rate": 0.0019530864197530864,
+      "loss": 0.5698,
+      "step": 559
+    },
+    {
+      "epoch": 1.2440988614273814,
+      "grad_norm": 0.0034373749513179064,
+      "learning_rate": 0.0019506172839506173,
+      "loss": 0.7554,
+      "step": 560
+    },
+    {
+      "epoch": 1.246320466537073,
+      "grad_norm": 0.0037724527064710855,
+      "learning_rate": 0.001948148148148148,
+      "loss": 0.8468,
+      "step": 561
+    },
+    {
+      "epoch": 1.2485420716467648,
+      "grad_norm": 0.0030816916842013597,
+      "learning_rate": 0.001945679012345679,
+      "loss": 0.5393,
+      "step": 562
+    },
+    {
+      "epoch": 1.2507636767564565,
+      "grad_norm": 0.004649344366043806,
+      "learning_rate": 0.0019432098765432098,
+      "loss": 0.553,
+      "step": 563
+    },
+    {
+      "epoch": 1.2529852818661482,
+      "grad_norm": 0.002786665689200163,
+      "learning_rate": 0.0019407407407407407,
+      "loss": 0.5569,
+      "step": 564
+    },
+    {
+      "epoch": 1.2552068869758402,
+      "grad_norm": 0.0026445803232491016,
+      "learning_rate": 0.0019382716049382714,
+      "loss": 0.5139,
+      "step": 565
+    },
+    {
+      "epoch": 1.2574284920855319,
+      "grad_norm": 0.005002819932997227,
+      "learning_rate": 0.0019358024691358024,
+      "loss": 0.6082,
+      "step": 566
+    },
+    {
+      "epoch": 1.2596500971952236,
+      "grad_norm": 0.005818431731313467,
+      "learning_rate": 0.0019333333333333336,
+      "loss": 0.7579,
+      "step": 567
+    },
+    {
+      "epoch": 1.2618717023049153,
+      "grad_norm": 0.005721336230635643,
+      "learning_rate": 0.001930864197530864,
+      "loss": 0.6972,
+      "step": 568
+    },
+    {
+      "epoch": 1.264093307414607,
+      "grad_norm": 0.0032174643129110336,
+      "learning_rate": 0.0019283950617283952,
+      "loss": 0.5709,
+      "step": 569
+    },
+    {
+      "epoch": 1.2663149125242987,
+      "grad_norm": 0.00355134648270905,
+      "learning_rate": 0.0019259259259259258,
+      "loss": 0.6242,
+      "step": 570
+    },
+    {
+      "epoch": 1.2685365176339904,
+      "grad_norm": 0.0028797700069844723,
+      "learning_rate": 0.001923456790123457,
+      "loss": 0.6642,
+      "step": 571
+    },
+    {
+      "epoch": 1.2707581227436824,
+      "grad_norm": 0.0029408205300569534,
+      "learning_rate": 0.0019209876543209875,
+      "loss": 0.6446,
+      "step": 572
+    },
+    {
+      "epoch": 1.272979727853374,
+      "grad_norm": 0.00306298746727407,
+      "learning_rate": 0.0019185185185185186,
+      "loss": 0.5913,
+      "step": 573
+    },
+    {
+      "epoch": 1.2752013329630658,
+      "grad_norm": 0.004586388822644949,
+      "learning_rate": 0.0019160493827160496,
+      "loss": 0.6747,
+      "step": 574
+    },
+    {
+      "epoch": 1.2774229380727575,
+      "grad_norm": 0.0031426565255969763,
+      "learning_rate": 0.0019135802469135803,
+      "loss": 0.7594,
+      "step": 575
+    },
+    {
+      "epoch": 1.2796445431824492,
+      "grad_norm": 0.003649466671049595,
+      "learning_rate": 0.0019111111111111113,
+      "loss": 0.5871,
+      "step": 576
+    },
+    {
+      "epoch": 1.2818661482921412,
+      "grad_norm": 0.006696793716400862,
+      "learning_rate": 0.001908641975308642,
+      "loss": 0.5172,
+      "step": 577
+    },
+    {
+      "epoch": 1.284087753401833,
+      "grad_norm": 0.004110109061002731,
+      "learning_rate": 0.001906172839506173,
+      "loss": 0.6425,
+      "step": 578
+    },
+    {
+      "epoch": 1.2863093585115246,
+      "grad_norm": 0.0030410694889724255,
+      "learning_rate": 0.0019037037037037037,
+      "loss": 0.5343,
+      "step": 579
+    },
+    {
+      "epoch": 1.2885309636212163,
+      "grad_norm": 0.0044550783932209015,
+      "learning_rate": 0.0019012345679012346,
+      "loss": 0.5118,
+      "step": 580
+    },
+    {
+      "epoch": 1.290752568730908,
+      "grad_norm": 0.004111356567591429,
+      "learning_rate": 0.0018987654320987656,
+      "loss": 0.6189,
+      "step": 581
+    },
+    {
+      "epoch": 1.2929741738405998,
+      "grad_norm": 0.002536875894293189,
+      "learning_rate": 0.0018962962962962963,
+      "loss": 0.5173,
+      "step": 582
+    },
+    {
+      "epoch": 1.2951957789502915,
+      "grad_norm": 0.003421909874305129,
+      "learning_rate": 0.0018938271604938273,
+      "loss": 0.562,
+      "step": 583
+    },
+    {
+      "epoch": 1.2974173840599834,
+      "grad_norm": 0.003108728677034378,
+      "learning_rate": 0.001891358024691358,
+      "loss": 0.5725,
+      "step": 584
+    },
+    {
+      "epoch": 1.2996389891696751,
+      "grad_norm": 0.0026788951363414526,
+      "learning_rate": 0.001888888888888889,
+      "loss": 0.574,
+      "step": 585
+    },
+    {
+      "epoch": 1.3018605942793668,
+      "grad_norm": 0.0026323439087718725,
+      "learning_rate": 0.0018864197530864197,
+      "loss": 0.8206,
+      "step": 586
+    },
+    {
+      "epoch": 1.3040821993890586,
+      "grad_norm": 0.003548834938555956,
+      "learning_rate": 0.0018839506172839506,
+      "loss": 0.636,
+      "step": 587
+    },
+    {
+      "epoch": 1.3063038044987503,
+      "grad_norm": 0.004747429396957159,
+      "learning_rate": 0.0018814814814814816,
+      "loss": 0.6796,
+      "step": 588
+    },
+    {
+      "epoch": 1.3085254096084422,
+      "grad_norm": 0.0041969697922468185,
+      "learning_rate": 0.0018790123456790123,
+      "loss": 0.5519,
+      "step": 589
+    },
+    {
+      "epoch": 1.310747014718134,
+      "grad_norm": 0.003130959114059806,
+      "learning_rate": 0.0018765432098765433,
+      "loss": 0.4979,
+      "step": 590
+    },
+    {
+      "epoch": 1.3129686198278256,
+      "grad_norm": 0.0030902365688234568,
+      "learning_rate": 0.001874074074074074,
+      "loss": 0.4823,
+      "step": 591
+    },
+    {
+      "epoch": 1.3151902249375174,
+      "grad_norm": 0.0031351533252745867,
+      "learning_rate": 0.001871604938271605,
+      "loss": 0.6665,
+      "step": 592
+    },
+    {
+      "epoch": 1.317411830047209,
+      "grad_norm": 0.003788917325437069,
+      "learning_rate": 0.0018691358024691357,
+      "loss": 0.6516,
+      "step": 593
+    },
+    {
+      "epoch": 1.3196334351569008,
+      "grad_norm": 0.004900997504591942,
+      "learning_rate": 0.0018666666666666666,
+      "loss": 0.6315,
+      "step": 594
+    },
+    {
+      "epoch": 1.3218550402665925,
+      "grad_norm": 0.0032752864062786102,
+      "learning_rate": 0.0018641975308641974,
+      "loss": 0.639,
+      "step": 595
+    },
+    {
+      "epoch": 1.3240766453762844,
+      "grad_norm": 0.002660909667611122,
+      "learning_rate": 0.0018617283950617283,
+      "loss": 0.6107,
+      "step": 596
+    },
+    {
+      "epoch": 1.3262982504859762,
+      "grad_norm": 0.003523773979395628,
+      "learning_rate": 0.0018592592592592595,
+      "loss": 0.5339,
+      "step": 597
+    },
+    {
+      "epoch": 1.3285198555956679,
+      "grad_norm": 0.002650630660355091,
+      "learning_rate": 0.00185679012345679,
+      "loss": 0.5768,
+      "step": 598
+    },
+    {
+      "epoch": 1.3307414607053596,
+      "grad_norm": 0.004082327708601952,
+      "learning_rate": 0.0018543209876543212,
+      "loss": 0.462,
+      "step": 599
+    },
+    {
+      "epoch": 1.3329630658150513,
+      "grad_norm": 0.004044753964990377,
+      "learning_rate": 0.0018518518518518517,
+      "loss": 0.6988,
+      "step": 600
+    },
+    {
+      "epoch": 1.3351846709247432,
+      "grad_norm": 0.0031004957854747772,
+      "learning_rate": 0.0018493827160493829,
+      "loss": 0.6714,
+      "step": 601
+    },
+    {
+      "epoch": 1.337406276034435,
+      "grad_norm": 0.003743590787053108,
+      "learning_rate": 0.0018469135802469134,
+      "loss": 0.5599,
+      "step": 602
+    },
+    {
+      "epoch": 1.3396278811441267,
+      "grad_norm": 0.0025853265542536974,
+      "learning_rate": 0.0018444444444444446,
+      "loss": 0.4862,
+      "step": 603
+    },
+    {
+      "epoch": 1.3418494862538184,
+      "grad_norm": 0.002392064081504941,
+      "learning_rate": 0.0018419753086419755,
+      "loss": 0.6645,
+      "step": 604
+    },
+    {
+      "epoch": 1.34407109136351,
+      "grad_norm": 0.004032174590975046,
+      "learning_rate": 0.0018395061728395062,
+      "loss": 0.6411,
+      "step": 605
+    },
+    {
+      "epoch": 1.3462926964732018,
+      "grad_norm": 0.003317909548059106,
+      "learning_rate": 0.0018370370370370372,
+      "loss": 0.6159,
+      "step": 606
+    },
+    {
+      "epoch": 1.3485143015828935,
+      "grad_norm": 0.0038619996048510075,
+      "learning_rate": 0.001834567901234568,
+      "loss": 0.6031,
+      "step": 607
+    },
+    {
+      "epoch": 1.3507359066925855,
+      "grad_norm": 0.003428127383813262,
+      "learning_rate": 0.0018320987654320989,
+      "loss": 0.7644,
+      "step": 608
+    },
+    {
+      "epoch": 1.3529575118022772,
+      "grad_norm": 0.003281916258856654,
+      "learning_rate": 0.0018296296296296296,
+      "loss": 0.6009,
+      "step": 609
+    },
+    {
+      "epoch": 1.355179116911969,
+      "grad_norm": 0.0022182667162269354,
+      "learning_rate": 0.0018271604938271606,
+      "loss": 0.5008,
+      "step": 610
+    },
+    {
+      "epoch": 1.3574007220216606,
+      "grad_norm": 0.008659104816615582,
+      "learning_rate": 0.0018246913580246915,
+      "loss": 0.7363,
+      "step": 611
+    },
+    {
+      "epoch": 1.3596223271313523,
+      "grad_norm": 0.0032885768450796604,
+      "learning_rate": 0.0018222222222222223,
+      "loss": 0.6114,
+      "step": 612
+    },
+    {
+      "epoch": 1.3618439322410443,
+      "grad_norm": 0.0024236771278083324,
+      "learning_rate": 0.0018197530864197532,
+      "loss": 0.5882,
+      "step": 613
+    },
+    {
+      "epoch": 1.364065537350736,
+      "grad_norm": 0.002592614386230707,
+      "learning_rate": 0.001817283950617284,
+      "loss": 0.5504,
+      "step": 614
+    },
+    {
+      "epoch": 1.3662871424604277,
+      "grad_norm": 0.0033317788038402796,
+      "learning_rate": 0.001814814814814815,
+      "loss": 0.5076,
+      "step": 615
+    },
+    {
+      "epoch": 1.3685087475701194,
+      "grad_norm": 0.002845801878720522,
+      "learning_rate": 0.0018123456790123456,
+      "loss": 0.5616,
+      "step": 616
+    },
+    {
+      "epoch": 1.3707303526798111,
+      "grad_norm": 0.0022413271944969893,
+      "learning_rate": 0.0018098765432098766,
+      "loss": 0.4234,
+      "step": 617
+    },
+    {
+      "epoch": 1.3729519577895029,
+      "grad_norm": 0.004213643725961447,
+      "learning_rate": 0.0018074074074074075,
+      "loss": 0.6248,
+      "step": 618
+    },
+    {
+      "epoch": 1.3751735628991946,
+      "grad_norm": 0.0032481811940670013,
+      "learning_rate": 0.0018049382716049383,
+      "loss": 0.5779,
+      "step": 619
+    },
+    {
+      "epoch": 1.3773951680088863,
+      "grad_norm": 0.004150136839598417,
+      "learning_rate": 0.0018024691358024692,
+      "loss": 0.7694,
+      "step": 620
+    },
+    {
+      "epoch": 1.3796167731185782,
+      "grad_norm": 0.003350265324115753,
+      "learning_rate": 0.0018,
+      "loss": 0.7772,
+      "step": 621
+    },
+    {
+      "epoch": 1.38183837822827,
+      "grad_norm": 0.0031599882058799267,
+      "learning_rate": 0.001797530864197531,
+      "loss": 0.5887,
+      "step": 622
+    },
+    {
+      "epoch": 1.3840599833379617,
+      "grad_norm": 0.0033556423150002956,
+      "learning_rate": 0.0017950617283950616,
+      "loss": 0.6057,
+      "step": 623
+    },
+    {
+      "epoch": 1.3862815884476534,
+      "grad_norm": 0.003302312456071377,
+      "learning_rate": 0.0017925925925925926,
+      "loss": 0.5406,
+      "step": 624
+    },
+    {
+      "epoch": 1.3885031935573453,
+      "grad_norm": 0.003953338600695133,
+      "learning_rate": 0.0017901234567901233,
+      "loss": 0.5672,
+      "step": 625
+    },
+    {
+      "epoch": 1.390724798667037,
+      "grad_norm": 0.00439725024625659,
+      "learning_rate": 0.0017876543209876543,
+      "loss": 0.5253,
+      "step": 626
+    },
+    {
+      "epoch": 1.3929464037767287,
+      "grad_norm": 0.015487024560570717,
+      "learning_rate": 0.0017851851851851854,
+      "loss": 0.7185,
+      "step": 627
+    },
+    {
+      "epoch": 1.3951680088864205,
+      "grad_norm": 0.0034425046760588884,
+      "learning_rate": 0.001782716049382716,
+      "loss": 0.6243,
+      "step": 628
+    },
+    {
+      "epoch": 1.3973896139961122,
+      "grad_norm": 0.003950211219489574,
+      "learning_rate": 0.0017802469135802471,
+      "loss": 0.5918,
+      "step": 629
+    },
+    {
+      "epoch": 1.3996112191058039,
+      "grad_norm": 0.002602384425699711,
+      "learning_rate": 0.0017777777777777776,
+      "loss": 0.5422,
+      "step": 630
+    },
+    {
+      "epoch": 1.4018328242154956,
+      "grad_norm": 0.003526592394337058,
+      "learning_rate": 0.0017753086419753088,
+      "loss": 0.6647,
+      "step": 631
+    },
+    {
+      "epoch": 1.4040544293251873,
+      "grad_norm": 0.0025864916387945414,
+      "learning_rate": 0.0017728395061728393,
+      "loss": 0.6296,
+      "step": 632
+    },
+    {
+      "epoch": 1.4062760344348793,
+      "grad_norm": 0.0031570426654070616,
+      "learning_rate": 0.0017703703703703705,
+      "loss": 0.547,
+      "step": 633
+    },
+    {
+      "epoch": 1.408497639544571,
+      "grad_norm": 0.0037307501770555973,
+      "learning_rate": 0.0017679012345679015,
+      "loss": 0.6843,
+      "step": 634
+    },
+    {
+      "epoch": 1.4107192446542627,
+      "grad_norm": 0.0035500023514032364,
+      "learning_rate": 0.0017654320987654322,
+      "loss": 0.5334,
+      "step": 635
+    },
+    {
+      "epoch": 1.4129408497639544,
+      "grad_norm": 0.0032879123464226723,
+      "learning_rate": 0.0017629629629629631,
+      "loss": 0.5945,
+      "step": 636
+    },
+    {
+      "epoch": 1.4151624548736463,
+      "grad_norm": 0.0034462285693734884,
+      "learning_rate": 0.0017604938271604939,
+      "loss": 0.7461,
+      "step": 637
+    },
+    {
+      "epoch": 1.417384059983338,
+      "grad_norm": 0.0030851562041789293,
+      "learning_rate": 0.0017580246913580248,
+      "loss": 0.6538,
+      "step": 638
+    },
+    {
+      "epoch": 1.4196056650930298,
+      "grad_norm": 0.0039033882785588503,
+      "learning_rate": 0.0017555555555555556,
+      "loss": 0.4908,
+      "step": 639
+    },
+    {
+      "epoch": 1.4218272702027215,
+      "grad_norm": 0.004719994496554136,
+      "learning_rate": 0.0017530864197530865,
+      "loss": 0.6726,
+      "step": 640
+    },
+    {
+      "epoch": 1.4240488753124132,
+      "grad_norm": 0.0027313127648085356,
+      "learning_rate": 0.0017506172839506175,
+      "loss": 0.625,
+      "step": 641
+    },
+    {
+      "epoch": 1.426270480422105,
+      "grad_norm": 0.00425785081461072,
+      "learning_rate": 0.0017481481481481482,
+      "loss": 0.5663,
+      "step": 642
+    },
+    {
+      "epoch": 1.4284920855317966,
+      "grad_norm": 0.003618143033236265,
+      "learning_rate": 0.0017456790123456791,
+      "loss": 0.5657,
+      "step": 643
+    },
+    {
+      "epoch": 1.4307136906414883,
+      "grad_norm": 0.0034850751981139183,
+      "learning_rate": 0.0017432098765432099,
+      "loss": 0.4107,
+      "step": 644
+    },
+    {
+      "epoch": 1.4329352957511803,
+      "grad_norm": 0.00514020211994648,
+      "learning_rate": 0.0017407407407407408,
+      "loss": 0.4958,
+      "step": 645
+    },
+    {
+      "epoch": 1.435156900860872,
+      "grad_norm": 0.003885776735842228,
+      "learning_rate": 0.0017382716049382716,
+      "loss": 0.7162,
+      "step": 646
+    },
+    {
+      "epoch": 1.4373785059705637,
+      "grad_norm": 0.003870051121339202,
+      "learning_rate": 0.0017358024691358025,
+      "loss": 0.5913,
+      "step": 647
+    },
+    {
+      "epoch": 1.4396001110802554,
+      "grad_norm": 0.004496397916227579,
+      "learning_rate": 0.0017333333333333333,
+      "loss": 0.5303,
+      "step": 648
+    },
+    {
+      "epoch": 1.4418217161899474,
+      "grad_norm": 0.0030751321464776993,
+      "learning_rate": 0.0017308641975308642,
+      "loss": 0.5838,
+      "step": 649
+    },
+    {
+      "epoch": 1.444043321299639,
+      "grad_norm": 0.002886572852730751,
+      "learning_rate": 0.0017283950617283952,
+      "loss": 0.5713,
+      "step": 650
+    },
+    {
+      "epoch": 1.4462649264093308,
+      "grad_norm": 0.011516193859279156,
+      "learning_rate": 0.0017259259259259259,
+      "loss": 0.5521,
+      "step": 651
+    },
+    {
+      "epoch": 1.4484865315190225,
+      "grad_norm": 0.0032446719706058502,
+      "learning_rate": 0.0017234567901234568,
+      "loss": 0.6537,
+      "step": 652
+    },
+    {
+      "epoch": 1.4507081366287142,
+      "grad_norm": 0.0029585566371679306,
+      "learning_rate": 0.0017209876543209876,
+      "loss": 0.7199,
+      "step": 653
+    },
+    {
+      "epoch": 1.452929741738406,
+      "grad_norm": 0.0026364405639469624,
+      "learning_rate": 0.0017185185185185185,
+      "loss": 0.4793,
+      "step": 654
+    },
+    {
+      "epoch": 1.4551513468480977,
+      "grad_norm": 0.004324890207499266,
+      "learning_rate": 0.0017160493827160493,
+      "loss": 0.6209,
+      "step": 655
+    },
+    {
+      "epoch": 1.4573729519577894,
+      "grad_norm": 0.002921945182606578,
+      "learning_rate": 0.0017135802469135802,
+      "loss": 0.5385,
+      "step": 656
+    },
+    {
+      "epoch": 1.4595945570674813,
+      "grad_norm": 0.003601512871682644,
+      "learning_rate": 0.0017111111111111114,
+      "loss": 0.5211,
+      "step": 657
+    },
+    {
+      "epoch": 1.461816162177173,
+      "grad_norm": 0.004054198041558266,
+      "learning_rate": 0.001708641975308642,
+      "loss": 0.6397,
+      "step": 658
+    },
+    {
+      "epoch": 1.4640377672868647,
+      "grad_norm": 0.007789338007569313,
+      "learning_rate": 0.001706172839506173,
+      "loss": 0.6116,
+      "step": 659
+    },
+    {
+      "epoch": 1.4662593723965565,
+      "grad_norm": 0.002359089907258749,
+      "learning_rate": 0.0017037037037037036,
+      "loss": 0.7018,
+      "step": 660
+    },
+    {
+      "epoch": 1.4684809775062484,
+      "grad_norm": 0.0033406647853553295,
+      "learning_rate": 0.0017012345679012348,
+      "loss": 0.7088,
+      "step": 661
+    },
+    {
+      "epoch": 1.4707025826159401,
+      "grad_norm": 0.0042248377576470375,
+      "learning_rate": 0.0016987654320987653,
+      "loss": 0.5889,
+      "step": 662
+    },
+    {
+      "epoch": 1.4729241877256318,
+      "grad_norm": 0.003515084972605109,
+      "learning_rate": 0.0016962962962962964,
+      "loss": 0.526,
+      "step": 663
+    },
+    {
+      "epoch": 1.4751457928353235,
+      "grad_norm": 0.003441104432567954,
+      "learning_rate": 0.0016938271604938274,
+      "loss": 0.6096,
+      "step": 664
+    },
+    {
+      "epoch": 1.4773673979450153,
+      "grad_norm": 0.010995453223586082,
+      "learning_rate": 0.0016913580246913581,
+      "loss": 0.6834,
+      "step": 665
+    },
+    {
+      "epoch": 1.479589003054707,
+      "grad_norm": 0.002769651124253869,
+      "learning_rate": 0.001688888888888889,
+      "loss": 0.5797,
+      "step": 666
+    },
+    {
+      "epoch": 1.4818106081643987,
+      "grad_norm": 0.0071142856031656265,
+      "learning_rate": 0.0016864197530864198,
+      "loss": 0.6198,
+      "step": 667
+    },
+    {
+      "epoch": 1.4840322132740904,
+      "grad_norm": 0.0038076492492109537,
+      "learning_rate": 0.0016839506172839508,
+      "loss": 0.5671,
+      "step": 668
+    },
+    {
+      "epoch": 1.4862538183837823,
+      "grad_norm": 0.0030527866911143064,
+      "learning_rate": 0.0016814814814814813,
+      "loss": 0.709,
+      "step": 669
+    },
+    {
+      "epoch": 1.488475423493474,
+      "grad_norm": 0.003952282480895519,
+      "learning_rate": 0.0016790123456790125,
+      "loss": 0.5556,
+      "step": 670
+    },
+    {
+      "epoch": 1.4906970286031658,
+      "grad_norm": 0.002633793978020549,
+      "learning_rate": 0.0016765432098765434,
+      "loss": 0.6726,
+      "step": 671
+    },
+    {
+      "epoch": 1.4929186337128575,
+      "grad_norm": 0.003946142271161079,
+      "learning_rate": 0.0016740740740740741,
+      "loss": 0.5391,
+      "step": 672
+    },
+    {
+      "epoch": 1.4951402388225494,
+      "grad_norm": 0.002832691418007016,
+      "learning_rate": 0.001671604938271605,
+      "loss": 0.5886,
+      "step": 673
+    },
+    {
+      "epoch": 1.4973618439322411,
+      "grad_norm": 0.0036020914558321238,
+      "learning_rate": 0.0016691358024691358,
+      "loss": 0.6381,
+      "step": 674
+    },
+    {
+      "epoch": 1.4995834490419329,
+      "grad_norm": 0.003216502955183387,
+      "learning_rate": 0.0016666666666666668,
+      "loss": 0.6763,
+      "step": 675
+    },
+    {
+      "epoch": 1.5018050541516246,
+      "grad_norm": 0.004902282729744911,
+      "learning_rate": 0.0016641975308641975,
+      "loss": 0.6803,
+      "step": 676
+    },
+    {
+      "epoch": 1.5040266592613163,
+      "grad_norm": 0.0033530874643474817,
+      "learning_rate": 0.0016617283950617285,
+      "loss": 0.6021,
+      "step": 677
+    },
+    {
+      "epoch": 1.506248264371008,
+      "grad_norm": 0.0028880988247692585,
+      "learning_rate": 0.0016592592592592592,
+      "loss": 0.5753,
+      "step": 678
+    },
+    {
+      "epoch": 1.5084698694806997,
+      "grad_norm": 0.0035627970937639475,
+      "learning_rate": 0.0016567901234567901,
+      "loss": 0.4803,
+      "step": 679
+    },
+    {
+      "epoch": 1.5106914745903914,
+      "grad_norm": 0.0036246783565729856,
+      "learning_rate": 0.001654320987654321,
+      "loss": 0.5201,
+      "step": 680
+    },
+    {
+      "epoch": 1.5129130797000832,
+      "grad_norm": 0.007211472373455763,
+      "learning_rate": 0.0016518518518518518,
+      "loss": 0.5862,
+      "step": 681
+    },
+    {
+      "epoch": 1.515134684809775,
+      "grad_norm": 0.0033675231970846653,
+      "learning_rate": 0.0016493827160493828,
+      "loss": 0.5614,
+      "step": 682
+    },
+    {
+      "epoch": 1.5173562899194668,
+      "grad_norm": 0.003647651756182313,
+      "learning_rate": 0.0016469135802469135,
+      "loss": 0.6762,
+      "step": 683
+    },
+    {
+      "epoch": 1.5195778950291585,
+      "grad_norm": 0.01229399535804987,
+      "learning_rate": 0.0016444444444444445,
+      "loss": 0.5211,
+      "step": 684
+    },
+    {
+      "epoch": 1.5217995001388505,
+      "grad_norm": 0.0026124136056751013,
+      "learning_rate": 0.0016419753086419752,
+      "loss": 0.4749,
+      "step": 685
+    },
+    {
+      "epoch": 1.5240211052485422,
+      "grad_norm": 0.0031844114419072866,
+      "learning_rate": 0.0016395061728395062,
+      "loss": 0.7737,
+      "step": 686
+    },
+    {
+      "epoch": 1.526242710358234,
+      "grad_norm": 0.0031438611913472414,
+      "learning_rate": 0.001637037037037037,
+      "loss": 0.6449,
+      "step": 687
+    },
+    {
+      "epoch": 1.5284643154679256,
+      "grad_norm": 0.0029068414587527514,
+      "learning_rate": 0.0016345679012345678,
+      "loss": 0.7685,
+      "step": 688
+    },
+    {
+      "epoch": 1.5306859205776173,
+      "grad_norm": 0.0024135911371558905,
+      "learning_rate": 0.0016320987654320988,
+      "loss": 0.5223,
+      "step": 689
+    },
+    {
+      "epoch": 1.532907525687309,
+      "grad_norm": 0.003742596134543419,
+      "learning_rate": 0.0016296296296296295,
+      "loss": 0.6033,
+      "step": 690
+    },
+    {
+      "epoch": 1.5351291307970008,
+      "grad_norm": 0.0037296046502888203,
+      "learning_rate": 0.0016271604938271605,
+      "loss": 0.569,
+      "step": 691
+    },
+    {
+      "epoch": 1.5373507359066925,
+      "grad_norm": 0.0029978591483086348,
+      "learning_rate": 0.0016246913580246912,
+      "loss": 0.5678,
+      "step": 692
+    },
+    {
+      "epoch": 1.5395723410163842,
+      "grad_norm": 0.0031321558635681868,
+      "learning_rate": 0.0016222222222222222,
+      "loss": 0.5599,
+      "step": 693
+    },
+    {
+      "epoch": 1.5417939461260761,
+      "grad_norm": 0.006184721831232309,
+      "learning_rate": 0.0016197530864197533,
+      "loss": 0.6828,
+      "step": 694
+    },
+    {
+      "epoch": 1.5440155512357678,
+      "grad_norm": 0.0033378407824784517,
+      "learning_rate": 0.0016172839506172839,
+      "loss": 0.6147,
+      "step": 695
+    },
+    {
+      "epoch": 1.5462371563454596,
+      "grad_norm": 0.009592154994606972,
+      "learning_rate": 0.001614814814814815,
+      "loss": 0.4876,
+      "step": 696
+    },
+    {
+      "epoch": 1.5484587614551515,
+      "grad_norm": 0.0029813058208674192,
+      "learning_rate": 0.0016123456790123455,
+      "loss": 0.6589,
+      "step": 697
+    },
+    {
+      "epoch": 1.5506803665648432,
+      "grad_norm": 0.0029721122700721025,
+      "learning_rate": 0.0016098765432098767,
+      "loss": 0.6142,
+      "step": 698
+    },
+    {
+      "epoch": 1.552901971674535,
+      "grad_norm": 0.003871340537443757,
+      "learning_rate": 0.0016074074074074072,
+      "loss": 0.3377,
+      "step": 699
+    },
+    {
+      "epoch": 1.5551235767842266,
+      "grad_norm": 0.0040636686608195305,
+      "learning_rate": 0.0016049382716049384,
+      "loss": 0.4972,
+      "step": 700
+    },
+    {
+      "epoch": 1.5573451818939184,
+      "grad_norm": 0.002279713749885559,
+      "learning_rate": 0.0016024691358024693,
+      "loss": 0.5894,
+      "step": 701
+    },
+    {
+      "epoch": 1.55956678700361,
+      "grad_norm": 0.00906870886683464,
+      "learning_rate": 0.0016,
+      "loss": 0.5124,
+      "step": 702
+    },
+    {
+      "epoch": 1.5617883921133018,
+      "grad_norm": 0.0029125467408448458,
+      "learning_rate": 0.001597530864197531,
+      "loss": 0.4712,
+      "step": 703
+    },
+    {
+      "epoch": 1.5640099972229935,
+      "grad_norm": 0.004257185850292444,
+      "learning_rate": 0.0015950617283950618,
+      "loss": 0.6892,
+      "step": 704
+    },
+    {
+      "epoch": 1.5662316023326852,
+      "grad_norm": 0.003058369504287839,
+      "learning_rate": 0.0015925925925925927,
+      "loss": 0.6638,
+      "step": 705
+    },
+    {
+      "epoch": 1.5684532074423772,
+      "grad_norm": 0.00876684207469225,
+      "learning_rate": 0.0015901234567901234,
+      "loss": 0.8143,
+      "step": 706
+    },
+    {
+      "epoch": 1.5706748125520689,
+      "grad_norm": 0.004514805041253567,
+      "learning_rate": 0.0015876543209876544,
+      "loss": 0.6552,
+      "step": 707
+    },
+    {
+      "epoch": 1.5728964176617606,
+      "grad_norm": 0.0033371017780154943,
+      "learning_rate": 0.0015851851851851851,
+      "loss": 0.5918,
+      "step": 708
+    },
+    {
+      "epoch": 1.5751180227714525,
+      "grad_norm": 0.004203807096928358,
+      "learning_rate": 0.001582716049382716,
+      "loss": 0.5464,
+      "step": 709
+    },
+    {
+      "epoch": 1.5773396278811442,
+      "grad_norm": 0.00238715554587543,
+      "learning_rate": 0.001580246913580247,
+      "loss": 0.5678,
+      "step": 710
+    },
+    {
+      "epoch": 1.579561232990836,
+      "grad_norm": 0.003361933631822467,
+      "learning_rate": 0.0015777777777777778,
+      "loss": 0.6751,
+      "step": 711
+    },
+    {
+      "epoch": 1.5817828381005277,
+      "grad_norm": 0.005657128058373928,
+      "learning_rate": 0.0015753086419753087,
+      "loss": 0.6271,
+      "step": 712
+    },
+    {
+      "epoch": 1.5840044432102194,
+      "grad_norm": 0.003028120379894972,
+      "learning_rate": 0.0015728395061728395,
+      "loss": 0.5869,
+      "step": 713
+    },
+    {
+      "epoch": 1.586226048319911,
+      "grad_norm": 0.0022292290814220905,
+      "learning_rate": 0.0015703703703703704,
+      "loss": 0.6638,
+      "step": 714
+    },
+    {
+      "epoch": 1.5884476534296028,
+      "grad_norm": 0.005532798357307911,
+      "learning_rate": 0.0015679012345679011,
+      "loss": 0.62,
+      "step": 715
+    },
+    {
+      "epoch": 1.5906692585392945,
+      "grad_norm": 0.0030338081996887922,
+      "learning_rate": 0.001565432098765432,
+      "loss": 0.5918,
+      "step": 716
+    },
+    {
+      "epoch": 1.5928908636489862,
+      "grad_norm": 0.002758693415671587,
+      "learning_rate": 0.001562962962962963,
+      "loss": 0.5983,
+      "step": 717
+    },
+    {
+      "epoch": 1.5951124687586782,
+      "grad_norm": 0.003892328590154648,
+      "learning_rate": 0.0015604938271604938,
+      "loss": 0.7643,
+      "step": 718
+    },
+    {
+      "epoch": 1.59733407386837,
+      "grad_norm": 0.003598182462155819,
+      "learning_rate": 0.0015580246913580247,
+      "loss": 0.5539,
+      "step": 719
+    },
+    {
+      "epoch": 1.5995556789780616,
+      "grad_norm": 0.0035891798324882984,
+      "learning_rate": 0.0015555555555555555,
+      "loss": 0.6757,
+      "step": 720
+    },
+    {
+      "epoch": 1.6017772840877536,
+      "grad_norm": 0.0055450694635510445,
+      "learning_rate": 0.0015530864197530864,
+      "loss": 0.5755,
+      "step": 721
+    },
+    {
+      "epoch": 1.6039988891974453,
+      "grad_norm": 0.0034944158978760242,
+      "learning_rate": 0.0015506172839506172,
+      "loss": 0.5402,
+      "step": 722
+    },
+    {
+      "epoch": 1.606220494307137,
+      "grad_norm": 0.0032796862069517374,
+      "learning_rate": 0.001548148148148148,
+      "loss": 0.6529,
+      "step": 723
+    },
+    {
+      "epoch": 1.6084420994168287,
+      "grad_norm": 0.004137154668569565,
+      "learning_rate": 0.0015456790123456793,
+      "loss": 0.5997,
+      "step": 724
+    },
+    {
+      "epoch": 1.6106637045265204,
+      "grad_norm": 0.003077402478083968,
+      "learning_rate": 0.0015432098765432098,
+      "loss": 0.5607,
+      "step": 725
+    },
+    {
+      "epoch": 1.6128853096362121,
+      "grad_norm": 0.0024837322998791933,
+      "learning_rate": 0.001540740740740741,
+      "loss": 0.4938,
+      "step": 726
+    },
+    {
+      "epoch": 1.6151069147459038,
+      "grad_norm": 0.0028923966456204653,
+      "learning_rate": 0.0015382716049382715,
+      "loss": 0.5682,
+      "step": 727
+    },
+    {
+      "epoch": 1.6173285198555956,
+      "grad_norm": 0.0036173651460558176,
+      "learning_rate": 0.0015358024691358026,
+      "loss": 0.5425,
+      "step": 728
+    },
+    {
+      "epoch": 1.6195501249652873,
+      "grad_norm": 0.00272387289442122,
+      "learning_rate": 0.0015333333333333332,
+      "loss": 0.5502,
+      "step": 729
+    },
+    {
+      "epoch": 1.6217717300749792,
+      "grad_norm": 0.005459922831505537,
+      "learning_rate": 0.0015308641975308643,
+      "loss": 0.6427,
+      "step": 730
+    },
+    {
+      "epoch": 1.623993335184671,
+      "grad_norm": 0.0024996125139296055,
+      "learning_rate": 0.0015283950617283948,
+      "loss": 0.4652,
+      "step": 731
+    },
+    {
+      "epoch": 1.6262149402943626,
+      "grad_norm": 0.002894905162975192,
+      "learning_rate": 0.001525925925925926,
+      "loss": 0.5918,
+      "step": 732
+    },
+    {
+      "epoch": 1.6284365454040546,
+      "grad_norm": 0.002637655008584261,
+      "learning_rate": 0.001523456790123457,
+      "loss": 0.7916,
+      "step": 733
+    },
+    {
+      "epoch": 1.6306581505137463,
+      "grad_norm": 0.003807837376371026,
+      "learning_rate": 0.0015209876543209877,
+      "loss": 0.6596,
+      "step": 734
+    },
+    {
+      "epoch": 1.632879755623438,
+      "grad_norm": 0.0031546615064144135,
+      "learning_rate": 0.0015185185185185187,
+      "loss": 0.5268,
+      "step": 735
+    },
+    {
+      "epoch": 1.6351013607331297,
+      "grad_norm": 0.002546357223764062,
+      "learning_rate": 0.0015160493827160494,
+      "loss": 0.5968,
+      "step": 736
+    },
+    {
+      "epoch": 1.6373229658428214,
+      "grad_norm": 0.003256875555962324,
+      "learning_rate": 0.0015135802469135803,
+      "loss": 0.6944,
+      "step": 737
+    },
+    {
+      "epoch": 1.6395445709525132,
+      "grad_norm": 0.0026704976335167885,
+      "learning_rate": 0.001511111111111111,
+      "loss": 0.6394,
+      "step": 738
+    },
+    {
+      "epoch": 1.6417661760622049,
+      "grad_norm": 0.0033118342980742455,
+      "learning_rate": 0.001508641975308642,
+      "loss": 0.5765,
+      "step": 739
+    },
+    {
+      "epoch": 1.6439877811718966,
+      "grad_norm": 0.0022287603933364153,
+      "learning_rate": 0.001506172839506173,
+      "loss": 0.5156,
+      "step": 740
+    },
+    {
+      "epoch": 1.6462093862815883,
+      "grad_norm": 0.0037926218938082457,
+      "learning_rate": 0.0015037037037037037,
+      "loss": 0.6705,
+      "step": 741
+    },
+    {
+      "epoch": 1.6484309913912802,
+      "grad_norm": 0.0030129130464047194,
+      "learning_rate": 0.0015012345679012347,
+      "loss": 0.5043,
+      "step": 742
+    },
+    {
+      "epoch": 1.650652596500972,
+      "grad_norm": 0.0029962975531816483,
+      "learning_rate": 0.0014987654320987656,
+      "loss": 0.486,
+      "step": 743
+    },
+    {
+      "epoch": 1.6528742016106637,
+      "grad_norm": 0.0029747833032160997,
+      "learning_rate": 0.0014962962962962963,
+      "loss": 0.7188,
+      "step": 744
+    },
+    {
+      "epoch": 1.6550958067203556,
+      "grad_norm": 0.0021846694871783257,
+      "learning_rate": 0.0014938271604938273,
+      "loss": 0.6036,
+      "step": 745
+    },
+    {
+      "epoch": 1.6573174118300473,
+      "grad_norm": 0.0029039906803518534,
+      "learning_rate": 0.001491358024691358,
+      "loss": 0.4769,
+      "step": 746
+    },
+    {
+      "epoch": 1.659539016939739,
+      "grad_norm": 0.0031887327786535025,
+      "learning_rate": 0.001488888888888889,
+      "loss": 0.4571,
+      "step": 747
+    },
+    {
+      "epoch": 1.6617606220494308,
+      "grad_norm": 0.0024829183239489794,
+      "learning_rate": 0.0014864197530864197,
+      "loss": 0.5437,
+      "step": 748
+    },
+    {
+      "epoch": 1.6639822271591225,
+      "grad_norm": 0.0026573697105050087,
+      "learning_rate": 0.0014839506172839507,
+      "loss": 0.4695,
+      "step": 749
+    },
+    {
+      "epoch": 1.6662038322688142,
+      "grad_norm": 0.004018016159534454,
+      "learning_rate": 0.0014814814814814814,
+      "loss": 0.5496,
+      "step": 750
+    },
+    {
+      "epoch": 1.668425437378506,
+      "grad_norm": 0.0031915605068206787,
+      "learning_rate": 0.0014790123456790124,
+      "loss": 0.5897,
+      "step": 751
+    },
+    {
+      "epoch": 1.6706470424881976,
+      "grad_norm": 0.0019371211528778076,
+      "learning_rate": 0.0014765432098765433,
+      "loss": 0.5734,
+      "step": 752
+    },
+    {
+      "epoch": 1.6728686475978893,
+      "grad_norm": 0.023410236462950706,
+      "learning_rate": 0.001474074074074074,
+      "loss": 0.829,
+      "step": 753
+    },
+    {
+      "epoch": 1.6750902527075813,
+      "grad_norm": 0.001892826403491199,
+      "learning_rate": 0.001471604938271605,
+      "loss": 0.7567,
+      "step": 754
+    },
+    {
+      "epoch": 1.677311857817273,
+      "grad_norm": 0.004156641196459532,
+      "learning_rate": 0.0014691358024691357,
+      "loss": 0.5387,
+      "step": 755
+    },
+    {
+      "epoch": 1.6795334629269647,
+      "grad_norm": 0.0030142583418637514,
+      "learning_rate": 0.0014666666666666667,
+      "loss": 0.5668,
+      "step": 756
+    },
+    {
+      "epoch": 1.6817550680366566,
+      "grad_norm": 0.004339707084000111,
+      "learning_rate": 0.0014641975308641974,
+      "loss": 0.5851,
+      "step": 757
+    },
+    {
+      "epoch": 1.6839766731463484,
+      "grad_norm": 0.0036378875374794006,
+      "learning_rate": 0.0014617283950617286,
+      "loss": 0.4967,
+      "step": 758
+    },
+    {
+      "epoch": 1.68619827825604,
+      "grad_norm": 0.0033033231738954782,
+      "learning_rate": 0.0014592592592592593,
+      "loss": 0.5743,
+      "step": 759
+    },
+    {
+      "epoch": 1.6884198833657318,
+      "grad_norm": 0.0028085929807275534,
+      "learning_rate": 0.0014567901234567903,
+      "loss": 0.6463,
+      "step": 760
+    },
+    {
+      "epoch": 1.6906414884754235,
+      "grad_norm": 0.005265321582555771,
+      "learning_rate": 0.001454320987654321,
+      "loss": 0.5585,
+      "step": 761
+    },
+    {
+      "epoch": 1.6928630935851152,
+      "grad_norm": 0.005091564729809761,
+      "learning_rate": 0.001451851851851852,
+      "loss": 0.5628,
+      "step": 762
+    },
+    {
+      "epoch": 1.695084698694807,
+      "grad_norm": 0.003041359828785062,
+      "learning_rate": 0.0014493827160493827,
+      "loss": 0.5766,
+      "step": 763
+    },
+    {
+      "epoch": 1.6973063038044987,
+      "grad_norm": 0.0026597119867801666,
+      "learning_rate": 0.0014469135802469136,
+      "loss": 0.7254,
+      "step": 764
+    },
+    {
+      "epoch": 1.6995279089141904,
+      "grad_norm": 0.003554214723408222,
+      "learning_rate": 0.0014444444444444444,
+      "loss": 0.5711,
+      "step": 765
+    },
+    {
+      "epoch": 1.7017495140238823,
+      "grad_norm": 0.0022794667165726423,
+      "learning_rate": 0.0014419753086419753,
+      "loss": 0.6111,
+      "step": 766
+    },
+    {
+      "epoch": 1.703971119133574,
+      "grad_norm": 0.00306728295981884,
+      "learning_rate": 0.0014395061728395063,
+      "loss": 0.4681,
+      "step": 767
+    },
+    {
+      "epoch": 1.7061927242432657,
+      "grad_norm": 0.0037019511219114065,
+      "learning_rate": 0.001437037037037037,
+      "loss": 0.5045,
+      "step": 768
+    },
+    {
+      "epoch": 1.7084143293529577,
+      "grad_norm": 0.010336251929402351,
+      "learning_rate": 0.001434567901234568,
+      "loss": 0.6809,
+      "step": 769
+    },
+    {
+      "epoch": 1.7106359344626494,
+      "grad_norm": 0.0039057794492691755,
+      "learning_rate": 0.0014320987654320987,
+      "loss": 0.537,
+      "step": 770
+    },
+    {
+      "epoch": 1.712857539572341,
+      "grad_norm": 0.005506658926606178,
+      "learning_rate": 0.0014296296296296297,
+      "loss": 0.5169,
+      "step": 771
+    },
+    {
+      "epoch": 1.7150791446820328,
+      "grad_norm": 0.0042714932933449745,
+      "learning_rate": 0.0014271604938271604,
+      "loss": 0.6398,
+      "step": 772
+    },
+    {
+      "epoch": 1.7173007497917245,
+      "grad_norm": 0.0034559788182377815,
+      "learning_rate": 0.0014246913580246916,
+      "loss": 0.5793,
+      "step": 773
+    },
+    {
+      "epoch": 1.7195223549014163,
+      "grad_norm": 0.0028325633611530066,
+      "learning_rate": 0.0014222222222222223,
+      "loss": 0.6201,
+      "step": 774
+    },
+    {
+      "epoch": 1.721743960011108,
+      "grad_norm": 0.0031491892877966166,
+      "learning_rate": 0.0014197530864197532,
+      "loss": 0.5324,
+      "step": 775
+    },
+    {
+      "epoch": 1.7239655651207997,
+      "grad_norm": 0.003899619448930025,
+      "learning_rate": 0.001417283950617284,
+      "loss": 0.4846,
+      "step": 776
+    },
+    {
+      "epoch": 1.7261871702304914,
+      "grad_norm": 0.005161529406905174,
+      "learning_rate": 0.001414814814814815,
+      "loss": 0.5526,
+      "step": 777
+    },
+    {
+      "epoch": 1.7284087753401833,
+      "grad_norm": 0.0032485160045325756,
+      "learning_rate": 0.0014123456790123457,
+      "loss": 0.5336,
+      "step": 778
+    },
+    {
+      "epoch": 1.730630380449875,
+      "grad_norm": 0.0028458782471716404,
+      "learning_rate": 0.0014098765432098766,
+      "loss": 0.4992,
+      "step": 779
+    },
+    {
+      "epoch": 1.7328519855595668,
+      "grad_norm": 0.004183284472674131,
+      "learning_rate": 0.0014074074074074073,
+      "loss": 0.5624,
+      "step": 780
+    },
+    {
+      "epoch": 1.7350735906692587,
+      "grad_norm": 0.0023061821702867746,
+      "learning_rate": 0.0014049382716049383,
+      "loss": 0.5448,
+      "step": 781
+    },
+    {
+      "epoch": 1.7372951957789504,
+      "grad_norm": 0.003513550153002143,
+      "learning_rate": 0.0014024691358024693,
+      "loss": 0.681,
+      "step": 782
+    },
+    {
+      "epoch": 1.7395168008886421,
+      "grad_norm": 0.0030224856454879045,
+      "learning_rate": 0.0014,
+      "loss": 0.6037,
+      "step": 783
+    },
+    {
+      "epoch": 1.7417384059983338,
+      "grad_norm": 0.0029998235404491425,
+      "learning_rate": 0.001397530864197531,
+      "loss": 0.4593,
+      "step": 784
+    },
+    {
+      "epoch": 1.7439600111080256,
+      "grad_norm": 0.003832211485132575,
+      "learning_rate": 0.0013950617283950617,
+      "loss": 0.5945,
+      "step": 785
+    },
+    {
+      "epoch": 1.7461816162177173,
+      "grad_norm": 0.003351503750309348,
+      "learning_rate": 0.0013925925925925926,
+      "loss": 0.5424,
+      "step": 786
+    },
+    {
+      "epoch": 1.748403221327409,
+      "grad_norm": 0.004922722466289997,
+      "learning_rate": 0.0013901234567901234,
+      "loss": 0.5702,
+      "step": 787
+    },
+    {
+      "epoch": 1.7506248264371007,
+      "grad_norm": 0.0033356482163071632,
+      "learning_rate": 0.0013876543209876545,
+      "loss": 0.6295,
+      "step": 788
+    },
+    {
+      "epoch": 1.7528464315467924,
+      "grad_norm": 0.0028782282024621964,
+      "learning_rate": 0.0013851851851851853,
+      "loss": 0.6662,
+      "step": 789
+    },
+    {
+      "epoch": 1.7550680366564844,
+      "grad_norm": 0.0029406151734292507,
+      "learning_rate": 0.0013827160493827162,
+      "loss": 0.6298,
+      "step": 790
+    },
+    {
+      "epoch": 1.757289641766176,
+      "grad_norm": 0.005104298237711191,
+      "learning_rate": 0.001380246913580247,
+      "loss": 0.7489,
+      "step": 791
+    },
+    {
+      "epoch": 1.7595112468758678,
+      "grad_norm": 0.0033603517804294825,
+      "learning_rate": 0.001377777777777778,
+      "loss": 0.6464,
+      "step": 792
+    },
+    {
+      "epoch": 1.7617328519855595,
+      "grad_norm": 0.003993791062384844,
+      "learning_rate": 0.0013753086419753086,
+      "loss": 0.7019,
+      "step": 793
+    },
+    {
+      "epoch": 1.7639544570952514,
+      "grad_norm": 0.0031684539280831814,
+      "learning_rate": 0.0013728395061728396,
+      "loss": 0.6741,
+      "step": 794
+    },
+    {
+      "epoch": 1.7661760622049432,
+      "grad_norm": 0.005214819684624672,
+      "learning_rate": 0.0013703703703703703,
+      "loss": 0.7585,
+      "step": 795
+    },
+    {
+      "epoch": 1.7683976673146349,
+      "grad_norm": 0.0023873383179306984,
+      "learning_rate": 0.0013679012345679013,
+      "loss": 0.477,
+      "step": 796
+    },
+    {
+      "epoch": 1.7706192724243266,
+      "grad_norm": 0.008193212561309338,
+      "learning_rate": 0.0013654320987654322,
+      "loss": 0.5233,
+      "step": 797
+    },
+    {
+      "epoch": 1.7728408775340183,
+      "grad_norm": 0.0024355570785701275,
+      "learning_rate": 0.001362962962962963,
+      "loss": 0.6476,
+      "step": 798
+    },
+    {
+      "epoch": 1.77506248264371,
+      "grad_norm": 0.003934087231755257,
+      "learning_rate": 0.001360493827160494,
+      "loss": 0.5401,
+      "step": 799
+    },
+    {
+      "epoch": 1.7772840877534017,
+      "grad_norm": 0.0030379316303879023,
+      "learning_rate": 0.0013580246913580246,
+      "loss": 0.7182,
+      "step": 800
+    },
+    {
+      "epoch": 1.7795056928630935,
+      "grad_norm": 0.0031336776446551085,
+      "learning_rate": 0.0013555555555555556,
+      "loss": 0.576,
+      "step": 801
+    },
+    {
+      "epoch": 1.7817272979727852,
+      "grad_norm": 0.0028493087738752365,
+      "learning_rate": 0.0013530864197530863,
+      "loss": 0.4405,
+      "step": 802
+    },
+    {
+      "epoch": 1.783948903082477,
+      "grad_norm": 0.0059678577817976475,
+      "learning_rate": 0.0013506172839506175,
+      "loss": 0.7326,
+      "step": 803
+    },
+    {
+      "epoch": 1.7861705081921688,
+      "grad_norm": 0.0037832173984497786,
+      "learning_rate": 0.0013481481481481482,
+      "loss": 0.7904,
+      "step": 804
+    },
+    {
+      "epoch": 1.7883921133018605,
+      "grad_norm": 0.005568805616348982,
+      "learning_rate": 0.0013456790123456792,
+      "loss": 0.4736,
+      "step": 805
+    },
+    {
+      "epoch": 1.7906137184115525,
+      "grad_norm": 0.003628462553024292,
+      "learning_rate": 0.00134320987654321,
+      "loss": 0.7379,
+      "step": 806
+    },
+    {
+      "epoch": 1.7928353235212442,
+      "grad_norm": 0.0037421484012156725,
+      "learning_rate": 0.0013407407407407409,
+      "loss": 0.8635,
+      "step": 807
+    },
+    {
+      "epoch": 1.795056928630936,
+      "grad_norm": 0.0035124209243804216,
+      "learning_rate": 0.0013382716049382716,
+      "loss": 0.7215,
+      "step": 808
+    },
+    {
+      "epoch": 1.7972785337406276,
+      "grad_norm": 0.0027946464251726866,
+      "learning_rate": 0.0013358024691358023,
+      "loss": 0.6327,
+      "step": 809
+    },
+    {
+      "epoch": 1.7995001388503193,
+      "grad_norm": 0.0035201977007091045,
+      "learning_rate": 0.0013333333333333333,
+      "loss": 0.8533,
+      "step": 810
+    },
+    {
+      "epoch": 1.801721743960011,
+      "grad_norm": 0.002365663181990385,
+      "learning_rate": 0.0013308641975308642,
+      "loss": 0.56,
+      "step": 811
+    },
+    {
+      "epoch": 1.8039433490697028,
+      "grad_norm": 0.004486533813178539,
+      "learning_rate": 0.0013283950617283952,
+      "loss": 0.6485,
+      "step": 812
+    },
+    {
+      "epoch": 1.8061649541793945,
+      "grad_norm": 0.0047790007665753365,
+      "learning_rate": 0.001325925925925926,
+      "loss": 0.6221,
+      "step": 813
+    },
+    {
+      "epoch": 1.8083865592890862,
+      "grad_norm": 0.003134514670819044,
+      "learning_rate": 0.0013234567901234569,
+      "loss": 0.627,
+      "step": 814
+    },
+    {
+      "epoch": 1.8106081643987781,
+      "grad_norm": 0.00474384892731905,
+      "learning_rate": 0.0013209876543209876,
+      "loss": 0.4979,
+      "step": 815
+    },
+    {
+      "epoch": 1.8128297695084699,
+      "grad_norm": 0.002987217390909791,
+      "learning_rate": 0.0013185185185185186,
+      "loss": 0.5478,
+      "step": 816
+    },
+    {
+      "epoch": 1.8150513746181616,
+      "grad_norm": 0.004350063391029835,
+      "learning_rate": 0.0013160493827160493,
+      "loss": 0.6971,
+      "step": 817
+    },
+    {
+      "epoch": 1.8172729797278535,
+      "grad_norm": 0.0028352115768939257,
+      "learning_rate": 0.0013135802469135802,
+      "loss": 0.5121,
+      "step": 818
+    },
+    {
+      "epoch": 1.8194945848375452,
+      "grad_norm": 0.003356591099873185,
+      "learning_rate": 0.0013111111111111112,
+      "loss": 0.5808,
+      "step": 819
+    },
+    {
+      "epoch": 1.821716189947237,
+      "grad_norm": 0.002301991917192936,
+      "learning_rate": 0.001308641975308642,
+      "loss": 0.5578,
+      "step": 820
+    },
+    {
+      "epoch": 1.8239377950569287,
+      "grad_norm": 0.002217639936134219,
+      "learning_rate": 0.0013061728395061729,
+      "loss": 0.5005,
+      "step": 821
+    },
+    {
+      "epoch": 1.8261594001666204,
+      "grad_norm": 0.003184704342857003,
+      "learning_rate": 0.0013037037037037036,
+      "loss": 0.5521,
+      "step": 822
+    },
+    {
+      "epoch": 1.828381005276312,
+      "grad_norm": 0.0025311317294836044,
+      "learning_rate": 0.0013012345679012346,
+      "loss": 0.5431,
+      "step": 823
+    },
+    {
+      "epoch": 1.8306026103860038,
+      "grad_norm": 0.0038392578717321157,
+      "learning_rate": 0.0012987654320987653,
+      "loss": 0.7709,
+      "step": 824
+    },
+    {
+      "epoch": 1.8328242154956955,
+      "grad_norm": 0.00471165357157588,
+      "learning_rate": 0.0012962962962962963,
+      "loss": 0.515,
+      "step": 825
+    },
+    {
+      "epoch": 1.8350458206053872,
+      "grad_norm": 0.00271948822773993,
+      "learning_rate": 0.0012938271604938272,
+      "loss": 0.631,
+      "step": 826
+    },
+    {
+      "epoch": 1.8372674257150792,
+      "grad_norm": 0.003444887697696686,
+      "learning_rate": 0.0012913580246913582,
+      "loss": 0.7599,
+      "step": 827
+    },
+    {
+      "epoch": 1.8394890308247709,
+      "grad_norm": 0.004435122944414616,
+      "learning_rate": 0.001288888888888889,
+      "loss": 0.4582,
+      "step": 828
+    },
+    {
+      "epoch": 1.8417106359344626,
+      "grad_norm": 0.004231697879731655,
+      "learning_rate": 0.0012864197530864198,
+      "loss": 0.4969,
+      "step": 829
+    },
+    {
+      "epoch": 1.8439322410441545,
+      "grad_norm": 0.004003169015049934,
+      "learning_rate": 0.0012839506172839506,
+      "loss": 0.4949,
+      "step": 830
+    },
+    {
+      "epoch": 1.8461538461538463,
+      "grad_norm": 0.002871347591280937,
+      "learning_rate": 0.0012814814814814815,
+      "loss": 0.5691,
+      "step": 831
+    },
+    {
+      "epoch": 1.848375451263538,
+      "grad_norm": 0.0037768930196762085,
+      "learning_rate": 0.0012790123456790123,
+      "loss": 0.5709,
+      "step": 832
+    },
+    {
+      "epoch": 1.8505970563732297,
+      "grad_norm": 0.003291292116045952,
+      "learning_rate": 0.0012765432098765432,
+      "loss": 0.5088,
+      "step": 833
+    },
+    {
+      "epoch": 1.8528186614829214,
+      "grad_norm": 0.0028356255497783422,
+      "learning_rate": 0.0012740740740740742,
+      "loss": 0.6365,
+      "step": 834
+    },
+    {
+      "epoch": 1.8550402665926131,
+      "grad_norm": 0.002764333738014102,
+      "learning_rate": 0.001271604938271605,
+      "loss": 0.5681,
+      "step": 835
+    },
+    {
+      "epoch": 1.8572618717023048,
+      "grad_norm": 0.0034658394288271666,
+      "learning_rate": 0.0012691358024691359,
+      "loss": 0.6079,
+      "step": 836
+    },
+    {
+      "epoch": 1.8594834768119965,
+      "grad_norm": 0.0034142339136451483,
+      "learning_rate": 0.0012666666666666666,
+      "loss": 0.5121,
+      "step": 837
+    },
+    {
+      "epoch": 1.8617050819216883,
+      "grad_norm": 0.00351393548771739,
+      "learning_rate": 0.0012641975308641975,
+      "loss": 0.5633,
+      "step": 838
+    },
+    {
+      "epoch": 1.8639266870313802,
+      "grad_norm": 0.002946634776890278,
+      "learning_rate": 0.0012617283950617283,
+      "loss": 0.5348,
+      "step": 839
+    },
+    {
+      "epoch": 1.866148292141072,
+      "grad_norm": 0.0031421848107129335,
+      "learning_rate": 0.0012592592592592592,
+      "loss": 0.512,
+      "step": 840
+    },
+    {
+      "epoch": 1.8683698972507636,
+      "grad_norm": 0.0020834184251725674,
+      "learning_rate": 0.0012567901234567902,
+      "loss": 0.6114,
+      "step": 841
+    },
+    {
+      "epoch": 1.8705915023604556,
+      "grad_norm": 0.0052506448701024055,
+      "learning_rate": 0.0012543209876543211,
+      "loss": 0.5895,
+      "step": 842
+    },
+    {
+      "epoch": 1.8728131074701473,
+      "grad_norm": 0.002586257178336382,
+      "learning_rate": 0.0012518518518518519,
+      "loss": 0.4924,
+      "step": 843
+    },
+    {
+      "epoch": 1.875034712579839,
+      "grad_norm": 0.005527224391698837,
+      "learning_rate": 0.0012493827160493828,
+      "loss": 0.5824,
+      "step": 844
+    },
+    {
+      "epoch": 1.8772563176895307,
+      "grad_norm": 0.002853000769391656,
+      "learning_rate": 0.0012469135802469136,
+      "loss": 0.6263,
+      "step": 845
+    },
+    {
+      "epoch": 1.8794779227992224,
+      "grad_norm": 0.0035244813188910484,
+      "learning_rate": 0.0012444444444444445,
+      "loss": 0.5518,
+      "step": 846
+    },
+    {
+      "epoch": 1.8816995279089141,
+      "grad_norm": 0.00501825287938118,
+      "learning_rate": 0.0012419753086419752,
+      "loss": 0.7406,
+      "step": 847
+    },
+    {
+      "epoch": 1.8839211330186059,
+      "grad_norm": 0.0029411388095468283,
+      "learning_rate": 0.0012395061728395062,
+      "loss": 0.5762,
+      "step": 848
+    },
+    {
+      "epoch": 1.8861427381282976,
+      "grad_norm": 0.004388920497149229,
+      "learning_rate": 0.0012370370370370371,
+      "loss": 0.6418,
+      "step": 849
+    },
+    {
+      "epoch": 1.8883643432379893,
+      "grad_norm": 0.0025772335939109325,
+      "learning_rate": 0.0012345679012345679,
+      "loss": 0.4931,
+      "step": 850
+    },
+    {
+      "epoch": 1.8905859483476812,
+      "grad_norm": 0.0032501844689249992,
+      "learning_rate": 0.0012320987654320988,
+      "loss": 0.7966,
+      "step": 851
+    },
+    {
+      "epoch": 1.892807553457373,
+      "grad_norm": 0.002551119541749358,
+      "learning_rate": 0.0012296296296296296,
+      "loss": 0.4498,
+      "step": 852
+    },
+    {
+      "epoch": 1.8950291585670647,
+      "grad_norm": 0.0032290250528603792,
+      "learning_rate": 0.0012271604938271605,
+      "loss": 0.5856,
+      "step": 853
+    },
+    {
+      "epoch": 1.8972507636767566,
+      "grad_norm": 0.004684037994593382,
+      "learning_rate": 0.0012246913580246912,
+      "loss": 0.7181,
+      "step": 854
+    },
+    {
+      "epoch": 1.8994723687864483,
+      "grad_norm": 0.0029012972954660654,
+      "learning_rate": 0.0012222222222222222,
+      "loss": 0.5601,
+      "step": 855
+    },
+    {
+      "epoch": 1.90169397389614,
+      "grad_norm": 0.002747067715972662,
+      "learning_rate": 0.0012197530864197532,
+      "loss": 0.5593,
+      "step": 856
+    },
+    {
+      "epoch": 1.9039155790058317,
+      "grad_norm": 0.003568375715985894,
+      "learning_rate": 0.001217283950617284,
+      "loss": 0.7568,
+      "step": 857
+    },
+    {
+      "epoch": 1.9061371841155235,
+      "grad_norm": 0.0028879425954073668,
+      "learning_rate": 0.0012148148148148148,
+      "loss": 0.5961,
+      "step": 858
+    },
+    {
+      "epoch": 1.9083587892252152,
+      "grad_norm": 0.0026781039778143167,
+      "learning_rate": 0.0012123456790123458,
+      "loss": 0.5069,
+      "step": 859
+    },
+    {
+      "epoch": 1.910580394334907,
+      "grad_norm": 0.005707379896193743,
+      "learning_rate": 0.0012098765432098765,
+      "loss": 0.5974,
+      "step": 860
+    },
+    {
+      "epoch": 1.9128019994445986,
+      "grad_norm": 0.0032613372895866632,
+      "learning_rate": 0.0012074074074074075,
+      "loss": 0.5567,
+      "step": 861
+    },
+    {
+      "epoch": 1.9150236045542903,
+      "grad_norm": 0.0032240943983197212,
+      "learning_rate": 0.0012049382716049382,
+      "loss": 0.4579,
+      "step": 862
+    },
+    {
+      "epoch": 1.9172452096639823,
+      "grad_norm": 0.0034050929825752974,
+      "learning_rate": 0.0012024691358024692,
+      "loss": 0.5795,
+      "step": 863
+    },
+    {
+      "epoch": 1.919466814773674,
+      "grad_norm": 0.002941009821370244,
+      "learning_rate": 0.0012000000000000001,
+      "loss": 0.5617,
+      "step": 864
+    },
+    {
+      "epoch": 1.9216884198833657,
+      "grad_norm": 0.003575817449018359,
+      "learning_rate": 0.0011975308641975308,
+      "loss": 0.6024,
+      "step": 865
+    },
+    {
+      "epoch": 1.9239100249930576,
+      "grad_norm": 0.003908672835677862,
+      "learning_rate": 0.0011950617283950618,
+      "loss": 0.6763,
+      "step": 866
+    },
+    {
+      "epoch": 1.9261316301027493,
+      "grad_norm": 0.0037765023298561573,
+      "learning_rate": 0.0011925925925925925,
+      "loss": 0.8389,
+      "step": 867
+    },
+    {
+      "epoch": 1.928353235212441,
+      "grad_norm": 0.0037054913118481636,
+      "learning_rate": 0.0011901234567901235,
+      "loss": 0.6546,
+      "step": 868
+    },
+    {
+      "epoch": 1.9305748403221328,
+      "grad_norm": 0.003900451585650444,
+      "learning_rate": 0.0011876543209876542,
+      "loss": 0.9324,
+      "step": 869
+    },
+    {
+      "epoch": 1.9327964454318245,
+      "grad_norm": 0.004091076552867889,
+      "learning_rate": 0.0011851851851851852,
+      "loss": 0.5661,
+      "step": 870
+    },
+    {
+      "epoch": 1.9350180505415162,
+      "grad_norm": 0.0036010409239679575,
+      "learning_rate": 0.0011827160493827161,
+      "loss": 0.5512,
+      "step": 871
+    },
+    {
+      "epoch": 1.937239655651208,
+      "grad_norm": 0.002555584069341421,
+      "learning_rate": 0.001180246913580247,
+      "loss": 0.5976,
+      "step": 872
+    },
+    {
+      "epoch": 1.9394612607608996,
+      "grad_norm": 0.002963186940178275,
+      "learning_rate": 0.0011777777777777778,
+      "loss": 0.5241,
+      "step": 873
+    },
+    {
+      "epoch": 1.9416828658705914,
+      "grad_norm": 0.002275889739394188,
+      "learning_rate": 0.0011753086419753088,
+      "loss": 0.4056,
+      "step": 874
+    },
+    {
+      "epoch": 1.9439044709802833,
+      "grad_norm": 0.004920678213238716,
+      "learning_rate": 0.0011728395061728395,
+      "loss": 0.768,
+      "step": 875
+    },
+    {
+      "epoch": 1.946126076089975,
+      "grad_norm": 0.003980088047683239,
+      "learning_rate": 0.0011703703703703704,
+      "loss": 0.823,
+      "step": 876
+    },
+    {
+      "epoch": 1.9483476811996667,
+      "grad_norm": 0.006616650614887476,
+      "learning_rate": 0.0011679012345679012,
+      "loss": 0.5191,
+      "step": 877
+    },
+    {
+      "epoch": 1.9505692863093587,
+      "grad_norm": 0.009624580852687359,
+      "learning_rate": 0.0011654320987654321,
+      "loss": 0.5275,
+      "step": 878
+    },
+    {
+      "epoch": 1.9527908914190504,
+      "grad_norm": 0.0033147030044347048,
+      "learning_rate": 0.001162962962962963,
+      "loss": 0.6125,
+      "step": 879
+    },
+    {
+      "epoch": 1.955012496528742,
+      "grad_norm": 0.004352504387497902,
+      "learning_rate": 0.0011604938271604938,
+      "loss": 0.6313,
+      "step": 880
+    },
+    {
+      "epoch": 1.9572341016384338,
+      "grad_norm": 0.0033672174904495478,
+      "learning_rate": 0.0011580246913580248,
+      "loss": 0.749,
+      "step": 881
+    },
+    {
+      "epoch": 1.9594557067481255,
+      "grad_norm": 0.003146470058709383,
+      "learning_rate": 0.0011555555555555555,
+      "loss": 0.6821,
+      "step": 882
+    },
+    {
+      "epoch": 1.9616773118578172,
+      "grad_norm": 0.004087381064891815,
+      "learning_rate": 0.0011530864197530865,
+      "loss": 0.7134,
+      "step": 883
+    },
+    {
+      "epoch": 1.963898916967509,
+      "grad_norm": 0.005879619624465704,
+      "learning_rate": 0.0011506172839506172,
+      "loss": 0.6927,
+      "step": 884
+    },
+    {
+      "epoch": 1.9661205220772007,
+      "grad_norm": 0.003711952827870846,
+      "learning_rate": 0.0011481481481481481,
+      "loss": 0.5362,
+      "step": 885
+    },
+    {
+      "epoch": 1.9683421271868924,
+      "grad_norm": 0.002924004103988409,
+      "learning_rate": 0.001145679012345679,
+      "loss": 0.6104,
+      "step": 886
+    },
+    {
+      "epoch": 1.9705637322965843,
+      "grad_norm": 0.0042654345743358135,
+      "learning_rate": 0.00114320987654321,
+      "loss": 0.6043,
+      "step": 887
+    },
+    {
+      "epoch": 1.972785337406276,
+      "grad_norm": 0.0030271972063928843,
+      "learning_rate": 0.0011407407407407408,
+      "loss": 0.6018,
+      "step": 888
+    },
+    {
+      "epoch": 1.9750069425159678,
+      "grad_norm": 0.0038058331701904535,
+      "learning_rate": 0.0011382716049382717,
+      "loss": 0.5575,
+      "step": 889
+    },
+    {
+      "epoch": 1.9772285476256597,
+      "grad_norm": 0.0035539015661925077,
+      "learning_rate": 0.0011358024691358025,
+      "loss": 0.6049,
+      "step": 890
+    },
+    {
+      "epoch": 1.9794501527353514,
+      "grad_norm": 0.0024260911159217358,
+      "learning_rate": 0.0011333333333333334,
+      "loss": 0.544,
+      "step": 891
+    },
+    {
+      "epoch": 1.9816717578450431,
+      "grad_norm": 0.0033254839945584536,
+      "learning_rate": 0.0011308641975308641,
+      "loss": 0.6019,
+      "step": 892
+    },
+    {
+      "epoch": 1.9838933629547348,
+      "grad_norm": 0.003086360404267907,
+      "learning_rate": 0.001128395061728395,
+      "loss": 0.7242,
+      "step": 893
+    },
+    {
+      "epoch": 1.9861149680644266,
+      "grad_norm": 0.004596822429448366,
+      "learning_rate": 0.001125925925925926,
+      "loss": 0.6787,
+      "step": 894
+    },
+    {
+      "epoch": 1.9883365731741183,
+      "grad_norm": 0.0026326498482376337,
+      "learning_rate": 0.0011234567901234568,
+      "loss": 0.7331,
+      "step": 895
+    },
+    {
+      "epoch": 1.99055817828381,
+      "grad_norm": 0.002320569707080722,
+      "learning_rate": 0.0011209876543209877,
+      "loss": 0.5561,
+      "step": 896
+    },
+    {
+      "epoch": 1.9927797833935017,
+      "grad_norm": 0.0028732616920024157,
+      "learning_rate": 0.0011185185185185185,
+      "loss": 0.6496,
+      "step": 897
+    },
+    {
+      "epoch": 1.9950013885031934,
+      "grad_norm": 0.0031755072996020317,
+      "learning_rate": 0.0011160493827160494,
+      "loss": 0.5572,
+      "step": 898
+    },
+    {
+      "epoch": 1.9972229936128854,
+      "grad_norm": 0.008104439824819565,
+      "learning_rate": 0.0011135802469135802,
+      "loss": 0.6715,
+      "step": 899
+    },
+    {
+      "epoch": 1.999444598722577,
+      "grad_norm": 0.0031844815239310265,
+      "learning_rate": 0.0011111111111111111,
+      "loss": 0.6162,
+      "step": 900
+    },
+    {
+      "epoch": 2.001666203832269,
+      "grad_norm": 0.0032060628291219473,
+      "learning_rate": 0.001108641975308642,
+      "loss": 0.8407,
+      "step": 901
+    },
+    {
+      "epoch": 2.0038878089419607,
+      "grad_norm": 0.004372204653918743,
+      "learning_rate": 0.001106172839506173,
+      "loss": 0.6518,
+      "step": 902
+    },
+    {
+      "epoch": 2.0061094140516524,
+      "grad_norm": 0.004050056450068951,
+      "learning_rate": 0.0011037037037037037,
+      "loss": 0.5921,
+      "step": 903
+    },
+    {
+      "epoch": 2.008331019161344,
+      "grad_norm": 0.0027456870302557945,
+      "learning_rate": 0.0011012345679012347,
+      "loss": 0.62,
+      "step": 904
+    },
+    {
+      "epoch": 2.010552624271036,
+      "grad_norm": 0.003426674986258149,
+      "learning_rate": 0.0010987654320987654,
+      "loss": 0.5978,
+      "step": 905
+    },
+    {
+      "epoch": 2.0127742293807276,
+      "grad_norm": 0.0026935499627143145,
+      "learning_rate": 0.0010962962962962964,
+      "loss": 0.7481,
+      "step": 906
+    },
+    {
+      "epoch": 2.0149958344904193,
+      "grad_norm": 0.003700273809954524,
+      "learning_rate": 0.0010938271604938271,
+      "loss": 0.5227,
+      "step": 907
+    },
+    {
+      "epoch": 2.017217439600111,
+      "grad_norm": 0.002455906942486763,
+      "learning_rate": 0.001091358024691358,
+      "loss": 0.5486,
+      "step": 908
+    },
+    {
+      "epoch": 2.0194390447098027,
+      "grad_norm": 0.005244240630418062,
+      "learning_rate": 0.001088888888888889,
+      "loss": 0.5229,
+      "step": 909
+    },
+    {
+      "epoch": 2.0216606498194944,
+      "grad_norm": 0.0026684950571507215,
+      "learning_rate": 0.0010864197530864198,
+      "loss": 0.6032,
+      "step": 910
+    },
+    {
+      "epoch": 2.023882254929186,
+      "grad_norm": 0.003200845792889595,
+      "learning_rate": 0.0010839506172839507,
+      "loss": 0.7075,
+      "step": 911
+    },
+    {
+      "epoch": 2.026103860038878,
+      "grad_norm": 0.0034468970261514187,
+      "learning_rate": 0.0010814814814814814,
+      "loss": 0.6763,
+      "step": 912
+    },
+    {
+      "epoch": 2.02832546514857,
+      "grad_norm": 0.0038897364865988493,
+      "learning_rate": 0.0010790123456790124,
+      "loss": 0.6753,
+      "step": 913
+    },
+    {
+      "epoch": 2.0305470702582618,
+      "grad_norm": 0.004033284727483988,
+      "learning_rate": 0.0010765432098765431,
+      "loss": 0.585,
+      "step": 914
+    },
+    {
+      "epoch": 2.0327686753679535,
+      "grad_norm": 0.0034496188163757324,
+      "learning_rate": 0.001074074074074074,
+      "loss": 0.5053,
+      "step": 915
+    },
+    {
+      "epoch": 2.034990280477645,
+      "grad_norm": 0.0031487636733800173,
+      "learning_rate": 0.0010716049382716048,
+      "loss": 0.4998,
+      "step": 916
+    },
+    {
+      "epoch": 2.037211885587337,
+      "grad_norm": 0.010006478987634182,
+      "learning_rate": 0.001069135802469136,
+      "loss": 0.7352,
+      "step": 917
+    },
+    {
+      "epoch": 2.0394334906970286,
+      "grad_norm": 0.0017054947093129158,
+      "learning_rate": 0.0010666666666666667,
+      "loss": 0.5026,
+      "step": 918
+    },
+    {
+      "epoch": 2.0416550958067203,
+      "grad_norm": 0.002679955679923296,
+      "learning_rate": 0.0010641975308641977,
+      "loss": 0.5098,
+      "step": 919
+    },
+    {
+      "epoch": 2.043876700916412,
+      "grad_norm": 0.0028986730612814426,
+      "learning_rate": 0.0010617283950617284,
+      "loss": 0.5148,
+      "step": 920
+    },
+    {
+      "epoch": 2.0460983060261038,
+      "grad_norm": 0.0033778445795178413,
+      "learning_rate": 0.0010592592592592594,
+      "loss": 0.6228,
+      "step": 921
+    },
+    {
+      "epoch": 2.0483199111357955,
+      "grad_norm": 0.0032284266781061888,
+      "learning_rate": 0.00105679012345679,
+      "loss": 0.486,
+      "step": 922
+    },
+    {
+      "epoch": 2.050541516245487,
+      "grad_norm": 0.005231673363596201,
+      "learning_rate": 0.001054320987654321,
+      "loss": 0.6242,
+      "step": 923
+    },
+    {
+      "epoch": 2.052763121355179,
+      "grad_norm": 0.0021800033282488585,
+      "learning_rate": 0.001051851851851852,
+      "loss": 0.791,
+      "step": 924
+    },
+    {
+      "epoch": 2.054984726464871,
+      "grad_norm": 0.002410687506198883,
+      "learning_rate": 0.0010493827160493827,
+      "loss": 0.4821,
+      "step": 925
+    },
+    {
+      "epoch": 2.057206331574563,
+      "grad_norm": 0.002446606522426009,
+      "learning_rate": 0.0010469135802469137,
+      "loss": 0.4773,
+      "step": 926
+    },
+    {
+      "epoch": 2.0594279366842545,
+      "grad_norm": 0.0032503134571015835,
+      "learning_rate": 0.0010444444444444444,
+      "loss": 0.56,
+      "step": 927
+    },
+    {
+      "epoch": 2.061649541793946,
+      "grad_norm": 0.003430373501032591,
+      "learning_rate": 0.0010419753086419754,
+      "loss": 0.5977,
+      "step": 928
+    },
+    {
+      "epoch": 2.063871146903638,
+      "grad_norm": 0.002888357499614358,
+      "learning_rate": 0.001039506172839506,
+      "loss": 0.4282,
+      "step": 929
+    },
+    {
+      "epoch": 2.0660927520133296,
+      "grad_norm": 0.0051438468508422375,
+      "learning_rate": 0.001037037037037037,
+      "loss": 0.74,
+      "step": 930
+    },
+    {
+      "epoch": 2.0683143571230214,
+      "grad_norm": 0.002881046384572983,
+      "learning_rate": 0.0010345679012345678,
+      "loss": 0.6234,
+      "step": 931
+    },
+    {
+      "epoch": 2.070535962232713,
+      "grad_norm": 0.003086875891312957,
+      "learning_rate": 0.001032098765432099,
+      "loss": 0.5821,
+      "step": 932
+    },
+    {
+      "epoch": 2.072757567342405,
+      "grad_norm": 0.002403114689514041,
+      "learning_rate": 0.0010296296296296297,
+      "loss": 0.5231,
+      "step": 933
+    },
+    {
+      "epoch": 2.0749791724520965,
+      "grad_norm": 0.003417134517803788,
+      "learning_rate": 0.0010271604938271606,
+      "loss": 0.5367,
+      "step": 934
+    },
+    {
+      "epoch": 2.0772007775617882,
+      "grad_norm": 0.0030301024671643972,
+      "learning_rate": 0.0010246913580246914,
+      "loss": 0.717,
+      "step": 935
+    },
+    {
+      "epoch": 2.07942238267148,
+      "grad_norm": 0.0032155404333025217,
+      "learning_rate": 0.0010222222222222223,
+      "loss": 0.4748,
+      "step": 936
+    },
+    {
+      "epoch": 2.081643987781172,
+      "grad_norm": 0.003064756980165839,
+      "learning_rate": 0.001019753086419753,
+      "loss": 0.6103,
+      "step": 937
+    },
+    {
+      "epoch": 2.083865592890864,
+      "grad_norm": 0.0030713852029293776,
+      "learning_rate": 0.001017283950617284,
+      "loss": 0.5911,
+      "step": 938
+    },
+    {
+      "epoch": 2.0860871980005555,
+      "grad_norm": 0.0025016344152390957,
+      "learning_rate": 0.001014814814814815,
+      "loss": 0.6379,
+      "step": 939
+    },
+    {
+      "epoch": 2.0883088031102472,
+      "grad_norm": 0.002508206060156226,
+      "learning_rate": 0.0010123456790123457,
+      "loss": 0.5557,
+      "step": 940
+    },
+    {
+      "epoch": 2.090530408219939,
+      "grad_norm": 0.0027723240200430155,
+      "learning_rate": 0.0010098765432098766,
+      "loss": 0.6908,
+      "step": 941
+    },
+    {
+      "epoch": 2.0927520133296307,
+      "grad_norm": 0.0024629731196910143,
+      "learning_rate": 0.0010074074074074074,
+      "loss": 0.4695,
+      "step": 942
+    },
+    {
+      "epoch": 2.0949736184393224,
+      "grad_norm": 0.004411566071212292,
+      "learning_rate": 0.0010049382716049383,
+      "loss": 0.4918,
+      "step": 943
+    },
+    {
+      "epoch": 2.097195223549014,
+      "grad_norm": 0.0032425241079181433,
+      "learning_rate": 0.001002469135802469,
+      "loss": 0.7062,
+      "step": 944
+    },
+    {
+      "epoch": 2.099416828658706,
+      "grad_norm": 0.0037023716140538454,
+      "learning_rate": 0.001,
+      "loss": 0.4463,
+      "step": 945
+    },
+    {
+      "epoch": 2.1016384337683975,
+      "grad_norm": 0.0033389173913747072,
+      "learning_rate": 0.0009975308641975308,
+      "loss": 0.6592,
+      "step": 946
+    },
+    {
+      "epoch": 2.1038600388780893,
+      "grad_norm": 0.002863410161808133,
+      "learning_rate": 0.000995061728395062,
+      "loss": 0.5698,
+      "step": 947
+    },
+    {
+      "epoch": 2.106081643987781,
+      "grad_norm": 0.002832047175616026,
+      "learning_rate": 0.0009925925925925927,
+      "loss": 0.5297,
+      "step": 948
+    },
+    {
+      "epoch": 2.108303249097473,
+      "grad_norm": 0.002443015808239579,
+      "learning_rate": 0.0009901234567901234,
+      "loss": 0.5083,
+      "step": 949
+    },
+    {
+      "epoch": 2.110524854207165,
+      "grad_norm": 0.003680097870528698,
+      "learning_rate": 0.0009876543209876543,
+      "loss": 0.6119,
+      "step": 950
+    },
+    {
+      "epoch": 2.1127464593168566,
+      "grad_norm": 0.0026307208463549614,
+      "learning_rate": 0.000985185185185185,
+      "loss": 0.5917,
+      "step": 951
+    },
+    {
+      "epoch": 2.1149680644265483,
+      "grad_norm": 0.006063534878194332,
+      "learning_rate": 0.000982716049382716,
+      "loss": 0.7092,
+      "step": 952
+    },
+    {
+      "epoch": 2.11718966953624,
+      "grad_norm": 0.0022731171920895576,
+      "learning_rate": 0.0009802469135802468,
+      "loss": 0.6964,
+      "step": 953
+    },
+    {
+      "epoch": 2.1194112746459317,
+      "grad_norm": 0.0027985614724457264,
+      "learning_rate": 0.000977777777777778,
+      "loss": 0.4312,
+      "step": 954
+    },
+    {
+      "epoch": 2.1216328797556234,
+      "grad_norm": 0.0022834171541035175,
+      "learning_rate": 0.0009753086419753087,
+      "loss": 0.6673,
+      "step": 955
+    },
+    {
+      "epoch": 2.123854484865315,
+      "grad_norm": 0.003239828860387206,
+      "learning_rate": 0.0009728395061728395,
+      "loss": 0.696,
+      "step": 956
+    },
+    {
+      "epoch": 2.126076089975007,
+      "grad_norm": 0.003194565186277032,
+      "learning_rate": 0.0009703703703703704,
+      "loss": 0.5976,
+      "step": 957
+    },
+    {
+      "epoch": 2.1282976950846986,
+      "grad_norm": 0.002975456416606903,
+      "learning_rate": 0.0009679012345679012,
+      "loss": 0.6193,
+      "step": 958
+    },
+    {
+      "epoch": 2.1305193001943903,
+      "grad_norm": 0.0038265048060566187,
+      "learning_rate": 0.000965432098765432,
+      "loss": 0.6149,
+      "step": 959
+    },
+    {
+      "epoch": 2.132740905304082,
+      "grad_norm": 0.004864771384745836,
+      "learning_rate": 0.0009629629629629629,
+      "loss": 0.5857,
+      "step": 960
+    },
+    {
+      "epoch": 2.134962510413774,
+      "grad_norm": 0.0027290659490972757,
+      "learning_rate": 0.0009604938271604937,
+      "loss": 0.4769,
+      "step": 961
+    },
+    {
+      "epoch": 2.137184115523466,
+      "grad_norm": 0.0031398101709783077,
+      "learning_rate": 0.0009580246913580248,
+      "loss": 0.609,
+      "step": 962
+    },
+    {
+      "epoch": 2.1394057206331576,
+      "grad_norm": 0.0030726627446711063,
+      "learning_rate": 0.0009555555555555556,
+      "loss": 0.5457,
+      "step": 963
+    },
+    {
+      "epoch": 2.1416273257428493,
+      "grad_norm": 0.003792215371504426,
+      "learning_rate": 0.0009530864197530865,
+      "loss": 0.6251,
+      "step": 964
+    },
+    {
+      "epoch": 2.143848930852541,
+      "grad_norm": 0.003504569409415126,
+      "learning_rate": 0.0009506172839506173,
+      "loss": 0.5433,
+      "step": 965
+    },
+    {
+      "epoch": 2.1460705359622327,
+      "grad_norm": 0.002695186994969845,
+      "learning_rate": 0.0009481481481481482,
+      "loss": 0.7355,
+      "step": 966
+    },
+    {
+      "epoch": 2.1482921410719245,
+      "grad_norm": 0.00423799641430378,
+      "learning_rate": 0.000945679012345679,
+      "loss": 0.707,
+      "step": 967
+    },
+    {
+      "epoch": 2.150513746181616,
+      "grad_norm": 0.0020570242777466774,
+      "learning_rate": 0.0009432098765432098,
+      "loss": 0.554,
+      "step": 968
+    },
+    {
+      "epoch": 2.152735351291308,
+      "grad_norm": 0.008327562361955643,
+      "learning_rate": 0.0009407407407407408,
+      "loss": 0.7368,
+      "step": 969
+    },
+    {
+      "epoch": 2.1549569564009996,
+      "grad_norm": 0.003192309755831957,
+      "learning_rate": 0.0009382716049382716,
+      "loss": 0.5697,
+      "step": 970
+    },
+    {
+      "epoch": 2.1571785615106913,
+      "grad_norm": 0.003064699936658144,
+      "learning_rate": 0.0009358024691358025,
+      "loss": 0.6009,
+      "step": 971
+    },
+    {
+      "epoch": 2.159400166620383,
+      "grad_norm": 0.004011985845863819,
+      "learning_rate": 0.0009333333333333333,
+      "loss": 0.591,
+      "step": 972
+    },
+    {
+      "epoch": 2.1616217717300747,
+      "grad_norm": 0.0022674642968922853,
+      "learning_rate": 0.0009308641975308642,
+      "loss": 0.5875,
+      "step": 973
+    },
+    {
+      "epoch": 2.163843376839767,
+      "grad_norm": 0.0029301177710294724,
+      "learning_rate": 0.000928395061728395,
+      "loss": 0.5329,
+      "step": 974
+    },
+    {
+      "epoch": 2.1660649819494586,
+      "grad_norm": 0.003778566373512149,
+      "learning_rate": 0.0009259259259259259,
+      "loss": 0.5196,
+      "step": 975
+    },
+    {
+      "epoch": 2.1682865870591503,
+      "grad_norm": 0.006877492181956768,
+      "learning_rate": 0.0009234567901234567,
+      "loss": 0.531,
+      "step": 976
+    },
+    {
+      "epoch": 2.170508192168842,
+      "grad_norm": 0.0031633051112294197,
+      "learning_rate": 0.0009209876543209878,
+      "loss": 0.6596,
+      "step": 977
+    },
+    {
+      "epoch": 2.1727297972785338,
+      "grad_norm": 0.003557397285476327,
+      "learning_rate": 0.0009185185185185186,
+      "loss": 0.5644,
+      "step": 978
+    },
+    {
+      "epoch": 2.1749514023882255,
+      "grad_norm": 0.003621811978518963,
+      "learning_rate": 0.0009160493827160494,
+      "loss": 0.6401,
+      "step": 979
+    },
+    {
+      "epoch": 2.177173007497917,
+      "grad_norm": 0.0038098511286079884,
+      "learning_rate": 0.0009135802469135803,
+      "loss": 0.6542,
+      "step": 980
+    },
+    {
+      "epoch": 2.179394612607609,
+      "grad_norm": 0.002912967000156641,
+      "learning_rate": 0.0009111111111111111,
+      "loss": 0.5677,
+      "step": 981
+    },
+    {
+      "epoch": 2.1816162177173006,
+      "grad_norm": 0.0035600329283624887,
+      "learning_rate": 0.000908641975308642,
+      "loss": 0.538,
+      "step": 982
+    },
+    {
+      "epoch": 2.1838378228269923,
+      "grad_norm": 0.0066943420097231865,
+      "learning_rate": 0.0009061728395061728,
+      "loss": 0.5981,
+      "step": 983
+    },
+    {
+      "epoch": 2.186059427936684,
+      "grad_norm": 0.0033611019607633352,
+      "learning_rate": 0.0009037037037037038,
+      "loss": 0.6569,
+      "step": 984
+    },
+    {
+      "epoch": 2.1882810330463762,
+      "grad_norm": 0.002343050902709365,
+      "learning_rate": 0.0009012345679012346,
+      "loss": 0.5339,
+      "step": 985
+    },
+    {
+      "epoch": 2.190502638156068,
+      "grad_norm": 0.0032377007883042097,
+      "learning_rate": 0.0008987654320987655,
+      "loss": 0.5378,
+      "step": 986
+    },
+    {
+      "epoch": 2.1927242432657597,
+      "grad_norm": 0.0030594319105148315,
+      "learning_rate": 0.0008962962962962963,
+      "loss": 0.5795,
+      "step": 987
+    },
+    {
+      "epoch": 2.1949458483754514,
+      "grad_norm": 0.0029608525801450014,
+      "learning_rate": 0.0008938271604938271,
+      "loss": 0.5973,
+      "step": 988
+    },
+    {
+      "epoch": 2.197167453485143,
+      "grad_norm": 0.0032938173972070217,
+      "learning_rate": 0.000891358024691358,
+      "loss": 0.4822,
+      "step": 989
+    },
+    {
+      "epoch": 2.199389058594835,
+      "grad_norm": 0.0032497644424438477,
+      "learning_rate": 0.0008888888888888888,
+      "loss": 0.5267,
+      "step": 990
+    },
+    {
+      "epoch": 2.2016106637045265,
+      "grad_norm": 0.0032970972824841738,
+      "learning_rate": 0.0008864197530864197,
+      "loss": 0.6145,
+      "step": 991
+    },
+    {
+      "epoch": 2.2038322688142182,
+      "grad_norm": 0.005727086216211319,
+      "learning_rate": 0.0008839506172839507,
+      "loss": 0.5812,
+      "step": 992
+    },
+    {
+      "epoch": 2.20605387392391,
+      "grad_norm": 0.002563419518992305,
+      "learning_rate": 0.0008814814814814816,
+      "loss": 0.4931,
+      "step": 993
+    },
+    {
+      "epoch": 2.2082754790336017,
+      "grad_norm": 0.004069584421813488,
+      "learning_rate": 0.0008790123456790124,
+      "loss": 0.5933,
+      "step": 994
+    },
+    {
+      "epoch": 2.2104970841432934,
+      "grad_norm": 0.0038564614951610565,
+      "learning_rate": 0.0008765432098765433,
+      "loss": 0.559,
+      "step": 995
+    },
+    {
+      "epoch": 2.212718689252985,
+      "grad_norm": 0.00243211816996336,
+      "learning_rate": 0.0008740740740740741,
+      "loss": 0.5612,
+      "step": 996
+    },
+    {
+      "epoch": 2.214940294362677,
+      "grad_norm": 0.002944133710116148,
+      "learning_rate": 0.0008716049382716049,
+      "loss": 0.4462,
+      "step": 997
+    },
+    {
+      "epoch": 2.217161899472369,
+      "grad_norm": 0.004629380535334349,
+      "learning_rate": 0.0008691358024691358,
+      "loss": 0.5722,
+      "step": 998
+    },
+    {
+      "epoch": 2.2193835045820607,
+      "grad_norm": 0.0027174127753823996,
+      "learning_rate": 0.0008666666666666666,
+      "loss": 0.5904,
+      "step": 999
+    },
+    {
+      "epoch": 2.2216051096917524,
+      "grad_norm": 0.0032827944960445166,
+      "learning_rate": 0.0008641975308641976,
+      "loss": 0.6582,
+      "step": 1000
+    },
+    {
+      "epoch": 2.223826714801444,
+      "grad_norm": 0.0030610065441578627,
+      "learning_rate": 0.0008617283950617284,
+      "loss": 0.542,
+      "step": 1001
+    },
+    {
+      "epoch": 2.226048319911136,
+      "grad_norm": 0.003364963224157691,
+      "learning_rate": 0.0008592592592592593,
+      "loss": 0.6866,
+      "step": 1002
+    },
+    {
+      "epoch": 2.2282699250208275,
+      "grad_norm": 0.0030600426252931356,
+      "learning_rate": 0.0008567901234567901,
+      "loss": 0.7132,
+      "step": 1003
+    },
+    {
+      "epoch": 2.2304915301305193,
+      "grad_norm": 0.003552238689735532,
+      "learning_rate": 0.000854320987654321,
+      "loss": 0.5064,
+      "step": 1004
+    },
+    {
+      "epoch": 2.232713135240211,
+      "grad_norm": 0.0041660708375275135,
+      "learning_rate": 0.0008518518518518518,
+      "loss": 0.6445,
+      "step": 1005
+    },
+    {
+      "epoch": 2.2349347403499027,
+      "grad_norm": 0.0053717633709311485,
+      "learning_rate": 0.0008493827160493826,
+      "loss": 0.5408,
+      "step": 1006
+    },
+    {
+      "epoch": 2.2371563454595944,
+      "grad_norm": 0.00260202307254076,
+      "learning_rate": 0.0008469135802469137,
+      "loss": 0.6055,
+      "step": 1007
+    },
+    {
+      "epoch": 2.239377950569286,
+      "grad_norm": 0.002337533049285412,
+      "learning_rate": 0.0008444444444444445,
+      "loss": 0.6636,
+      "step": 1008
+    },
+    {
+      "epoch": 2.2415995556789783,
+      "grad_norm": 0.003964942414313555,
+      "learning_rate": 0.0008419753086419754,
+      "loss": 0.577,
+      "step": 1009
+    },
+    {
+      "epoch": 2.24382116078867,
+      "grad_norm": 0.0029365664813667536,
+      "learning_rate": 0.0008395061728395062,
+      "loss": 0.4955,
+      "step": 1010
+    },
+    {
+      "epoch": 2.2460427658983617,
+      "grad_norm": 0.004111704416573048,
+      "learning_rate": 0.0008370370370370371,
+      "loss": 0.5974,
+      "step": 1011
+    },
+    {
+      "epoch": 2.2482643710080534,
+      "grad_norm": 0.00641743466258049,
+      "learning_rate": 0.0008345679012345679,
+      "loss": 0.5814,
+      "step": 1012
+    },
+    {
+      "epoch": 2.250485976117745,
+      "grad_norm": 0.0032692174427211285,
+      "learning_rate": 0.0008320987654320988,
+      "loss": 0.5916,
+      "step": 1013
+    },
+    {
+      "epoch": 2.252707581227437,
+      "grad_norm": 0.007102708797901869,
+      "learning_rate": 0.0008296296296296296,
+      "loss": 0.6731,
+      "step": 1014
+    },
+    {
+      "epoch": 2.2549291863371286,
+      "grad_norm": 0.004556303843855858,
+      "learning_rate": 0.0008271604938271605,
+      "loss": 0.6026,
+      "step": 1015
+    },
+    {
+      "epoch": 2.2571507914468203,
+      "grad_norm": 0.004175705835223198,
+      "learning_rate": 0.0008246913580246914,
+      "loss": 0.5041,
+      "step": 1016
+    },
+    {
+      "epoch": 2.259372396556512,
+      "grad_norm": 0.0031485306099057198,
+      "learning_rate": 0.0008222222222222222,
+      "loss": 0.5685,
+      "step": 1017
+    },
+    {
+      "epoch": 2.2615940016662037,
+      "grad_norm": 0.0030771722085773945,
+      "learning_rate": 0.0008197530864197531,
+      "loss": 0.651,
+      "step": 1018
+    },
+    {
+      "epoch": 2.2638156067758954,
+      "grad_norm": 0.0028245358262211084,
+      "learning_rate": 0.0008172839506172839,
+      "loss": 0.6397,
+      "step": 1019
+    },
+    {
+      "epoch": 2.266037211885587,
+      "grad_norm": 0.0038441738579422235,
+      "learning_rate": 0.0008148148148148148,
+      "loss": 0.5486,
+      "step": 1020
+    },
+    {
+      "epoch": 2.268258816995279,
+      "grad_norm": 0.0029330954421311617,
+      "learning_rate": 0.0008123456790123456,
+      "loss": 0.5605,
+      "step": 1021
+    },
+    {
+      "epoch": 2.270480422104971,
+      "grad_norm": 0.0047192759811878204,
+      "learning_rate": 0.0008098765432098767,
+      "loss": 0.5762,
+      "step": 1022
+    },
+    {
+      "epoch": 2.2727020272146627,
+      "grad_norm": 0.004585467744618654,
+      "learning_rate": 0.0008074074074074075,
+      "loss": 0.6548,
+      "step": 1023
+    },
+    {
+      "epoch": 2.2749236323243545,
+      "grad_norm": 0.004374127369374037,
+      "learning_rate": 0.0008049382716049384,
+      "loss": 0.5263,
+      "step": 1024
+    },
+    {
+      "epoch": 2.277145237434046,
+      "grad_norm": 0.0028513516299426556,
+      "learning_rate": 0.0008024691358024692,
+      "loss": 0.6538,
+      "step": 1025
+    },
+    {
+      "epoch": 2.279366842543738,
+      "grad_norm": 0.0029575314838439226,
+      "learning_rate": 0.0008,
+      "loss": 0.7316,
+      "step": 1026
+    },
+    {
+      "epoch": 2.2815884476534296,
+      "grad_norm": 0.0030439943075180054,
+      "learning_rate": 0.0007975308641975309,
+      "loss": 0.6945,
+      "step": 1027
+    },
+    {
+      "epoch": 2.2838100527631213,
+      "grad_norm": 0.003041263669729233,
+      "learning_rate": 0.0007950617283950617,
+      "loss": 0.5912,
+      "step": 1028
+    },
+    {
+      "epoch": 2.286031657872813,
+      "grad_norm": 0.004571863915771246,
+      "learning_rate": 0.0007925925925925926,
+      "loss": 0.6742,
+      "step": 1029
+    },
+    {
+      "epoch": 2.2882532629825048,
+      "grad_norm": 0.002520073438063264,
+      "learning_rate": 0.0007901234567901235,
+      "loss": 0.5184,
+      "step": 1030
+    },
+    {
+      "epoch": 2.2904748680921965,
+      "grad_norm": 0.003077463945373893,
+      "learning_rate": 0.0007876543209876544,
+      "loss": 0.7712,
+      "step": 1031
+    },
+    {
+      "epoch": 2.292696473201888,
+      "grad_norm": 0.0033735015895217657,
+      "learning_rate": 0.0007851851851851852,
+      "loss": 0.6573,
+      "step": 1032
+    },
+    {
+      "epoch": 2.2949180783115803,
+      "grad_norm": 0.0035345982760190964,
+      "learning_rate": 0.000782716049382716,
+      "loss": 0.5874,
+      "step": 1033
+    },
+    {
+      "epoch": 2.297139683421272,
+      "grad_norm": 0.009167306125164032,
+      "learning_rate": 0.0007802469135802469,
+      "loss": 0.6943,
+      "step": 1034
+    },
+    {
+      "epoch": 2.2993612885309638,
+      "grad_norm": 0.0029318418819457293,
+      "learning_rate": 0.0007777777777777777,
+      "loss": 0.5472,
+      "step": 1035
+    },
+    {
+      "epoch": 2.3015828936406555,
+      "grad_norm": 0.006966088432818651,
+      "learning_rate": 0.0007753086419753086,
+      "loss": 0.6548,
+      "step": 1036
+    },
+    {
+      "epoch": 2.303804498750347,
+      "grad_norm": 0.0032308287918567657,
+      "learning_rate": 0.0007728395061728396,
+      "loss": 0.4253,
+      "step": 1037
+    },
+    {
+      "epoch": 2.306026103860039,
+      "grad_norm": 0.0036757299676537514,
+      "learning_rate": 0.0007703703703703705,
+      "loss": 0.7802,
+      "step": 1038
+    },
+    {
+      "epoch": 2.3082477089697306,
+      "grad_norm": 0.003490788396447897,
+      "learning_rate": 0.0007679012345679013,
+      "loss": 0.5233,
+      "step": 1039
+    },
+    {
+      "epoch": 2.3104693140794224,
+      "grad_norm": 0.0038194290827959776,
+      "learning_rate": 0.0007654320987654322,
+      "loss": 0.6783,
+      "step": 1040
+    },
+    {
+      "epoch": 2.312690919189114,
+      "grad_norm": 0.0034465447533875704,
+      "learning_rate": 0.000762962962962963,
+      "loss": 0.5965,
+      "step": 1041
+    },
+    {
+      "epoch": 2.314912524298806,
+      "grad_norm": 0.0025339655112475157,
+      "learning_rate": 0.0007604938271604939,
+      "loss": 0.5269,
+      "step": 1042
+    },
+    {
+      "epoch": 2.3171341294084975,
+      "grad_norm": 0.003927841316908598,
+      "learning_rate": 0.0007580246913580247,
+      "loss": 0.5006,
+      "step": 1043
+    },
+    {
+      "epoch": 2.319355734518189,
+      "grad_norm": 0.0037475882563740015,
+      "learning_rate": 0.0007555555555555555,
+      "loss": 0.736,
+      "step": 1044
+    },
+    {
+      "epoch": 2.321577339627881,
+      "grad_norm": 0.0032975671347230673,
+      "learning_rate": 0.0007530864197530865,
+      "loss": 0.553,
+      "step": 1045
+    },
+    {
+      "epoch": 2.3237989447375726,
+      "grad_norm": 0.0042790621519088745,
+      "learning_rate": 0.0007506172839506173,
+      "loss": 0.8489,
+      "step": 1046
+    },
+    {
+      "epoch": 2.326020549847265,
+      "grad_norm": 0.005435307510197163,
+      "learning_rate": 0.0007481481481481482,
+      "loss": 0.6543,
+      "step": 1047
+    },
+    {
+      "epoch": 2.3282421549569565,
+      "grad_norm": 0.003635212080553174,
+      "learning_rate": 0.000745679012345679,
+      "loss": 0.5125,
+      "step": 1048
+    },
+    {
+      "epoch": 2.3304637600666482,
+      "grad_norm": 0.003130897181108594,
+      "learning_rate": 0.0007432098765432099,
+      "loss": 0.5689,
+      "step": 1049
+    },
+    {
+      "epoch": 2.33268536517634,
+      "grad_norm": 0.0036147863138467073,
+      "learning_rate": 0.0007407407407407407,
+      "loss": 0.6393,
+      "step": 1050
+    },
+    {
+      "epoch": 2.3349069702860317,
+      "grad_norm": 0.0032868238631635904,
+      "learning_rate": 0.0007382716049382717,
+      "loss": 0.5392,
+      "step": 1051
+    },
+    {
+      "epoch": 2.3371285753957234,
+      "grad_norm": 0.0031462872866541147,
+      "learning_rate": 0.0007358024691358025,
+      "loss": 0.6029,
+      "step": 1052
+    },
+    {
+      "epoch": 2.339350180505415,
+      "grad_norm": 0.002507926896214485,
+      "learning_rate": 0.0007333333333333333,
+      "loss": 0.6052,
+      "step": 1053
+    },
+    {
+      "epoch": 2.341571785615107,
+      "grad_norm": 0.0035951449535787106,
+      "learning_rate": 0.0007308641975308643,
+      "loss": 0.5695,
+      "step": 1054
+    },
+    {
+      "epoch": 2.3437933907247985,
+      "grad_norm": 0.003026745980605483,
+      "learning_rate": 0.0007283950617283951,
+      "loss": 0.5688,
+      "step": 1055
+    },
+    {
+      "epoch": 2.3460149958344902,
+      "grad_norm": 0.003848330583423376,
+      "learning_rate": 0.000725925925925926,
+      "loss": 0.5459,
+      "step": 1056
+    },
+    {
+      "epoch": 2.3482366009441824,
+      "grad_norm": 0.0019765642937272787,
+      "learning_rate": 0.0007234567901234568,
+      "loss": 0.4718,
+      "step": 1057
+    },
+    {
+      "epoch": 2.350458206053874,
+      "grad_norm": 0.002635281765833497,
+      "learning_rate": 0.0007209876543209877,
+      "loss": 0.567,
+      "step": 1058
+    },
+    {
+      "epoch": 2.352679811163566,
+      "grad_norm": 0.005967604462057352,
+      "learning_rate": 0.0007185185185185185,
+      "loss": 0.5747,
+      "step": 1059
+    },
+    {
+      "epoch": 2.3549014162732576,
+      "grad_norm": 0.00438288738951087,
+      "learning_rate": 0.0007160493827160494,
+      "loss": 0.6649,
+      "step": 1060
+    },
+    {
+      "epoch": 2.3571230213829493,
+      "grad_norm": 0.0029575556982308626,
+      "learning_rate": 0.0007135802469135802,
+      "loss": 0.4728,
+      "step": 1061
+    },
+    {
+      "epoch": 2.359344626492641,
+      "grad_norm": 0.002983496058732271,
+      "learning_rate": 0.0007111111111111111,
+      "loss": 0.5236,
+      "step": 1062
+    },
+    {
+      "epoch": 2.3615662316023327,
+      "grad_norm": 0.004713087808340788,
+      "learning_rate": 0.000708641975308642,
+      "loss": 0.5725,
+      "step": 1063
+    },
+    {
+      "epoch": 2.3637878367120244,
+      "grad_norm": 0.0026739819440990686,
+      "learning_rate": 0.0007061728395061728,
+      "loss": 0.4965,
+      "step": 1064
+    },
+    {
+      "epoch": 2.366009441821716,
+      "grad_norm": 0.002458624541759491,
+      "learning_rate": 0.0007037037037037037,
+      "loss": 0.5445,
+      "step": 1065
+    },
+    {
+      "epoch": 2.368231046931408,
+      "grad_norm": 0.003967548720538616,
+      "learning_rate": 0.0007012345679012346,
+      "loss": 0.6989,
+      "step": 1066
+    },
+    {
+      "epoch": 2.3704526520410996,
+      "grad_norm": 0.003330156672745943,
+      "learning_rate": 0.0006987654320987655,
+      "loss": 0.5463,
+      "step": 1067
+    },
+    {
+      "epoch": 2.3726742571507913,
+      "grad_norm": 0.003007511142641306,
+      "learning_rate": 0.0006962962962962963,
+      "loss": 0.5579,
+      "step": 1068
+    },
+    {
+      "epoch": 2.374895862260483,
+      "grad_norm": 0.016308341175317764,
+      "learning_rate": 0.0006938271604938273,
+      "loss": 0.5285,
+      "step": 1069
+    },
+    {
+      "epoch": 2.3771174673701747,
+      "grad_norm": 0.0036616057623177767,
+      "learning_rate": 0.0006913580246913581,
+      "loss": 0.6642,
+      "step": 1070
+    },
+    {
+      "epoch": 2.379339072479867,
+      "grad_norm": 0.0035557104274630547,
+      "learning_rate": 0.000688888888888889,
+      "loss": 0.5241,
+      "step": 1071
+    },
+    {
+      "epoch": 2.3815606775895586,
+      "grad_norm": 0.0035361377522349358,
+      "learning_rate": 0.0006864197530864198,
+      "loss": 0.648,
+      "step": 1072
+    },
+    {
+      "epoch": 2.3837822826992503,
+      "grad_norm": 0.003730503376573324,
+      "learning_rate": 0.0006839506172839506,
+      "loss": 0.514,
+      "step": 1073
+    },
+    {
+      "epoch": 2.386003887808942,
+      "grad_norm": 0.004067933186888695,
+      "learning_rate": 0.0006814814814814815,
+      "loss": 0.5257,
+      "step": 1074
+    },
+    {
+      "epoch": 2.3882254929186337,
+      "grad_norm": 0.0020785662345588207,
+      "learning_rate": 0.0006790123456790123,
+      "loss": 0.6464,
+      "step": 1075
+    },
+    {
+      "epoch": 2.3904470980283254,
+      "grad_norm": 0.002413610927760601,
+      "learning_rate": 0.0006765432098765432,
+      "loss": 0.5519,
+      "step": 1076
+    },
+    {
+      "epoch": 2.392668703138017,
+      "grad_norm": 0.007650860119611025,
+      "learning_rate": 0.0006740740740740741,
+      "loss": 0.5797,
+      "step": 1077
+    },
+    {
+      "epoch": 2.394890308247709,
+      "grad_norm": 0.002815013052895665,
+      "learning_rate": 0.000671604938271605,
+      "loss": 0.6294,
+      "step": 1078
+    },
+    {
+      "epoch": 2.3971119133574006,
+      "grad_norm": 0.0022983052767813206,
+      "learning_rate": 0.0006691358024691358,
+      "loss": 0.4428,
+      "step": 1079
+    },
+    {
+      "epoch": 2.3993335184670923,
+      "grad_norm": 0.003167651128023863,
+      "learning_rate": 0.0006666666666666666,
+      "loss": 0.467,
+      "step": 1080
+    },
+    {
+      "epoch": 2.4015551235767845,
+      "grad_norm": 0.00482570007443428,
+      "learning_rate": 0.0006641975308641976,
+      "loss": 0.6744,
+      "step": 1081
+    },
+    {
+      "epoch": 2.403776728686476,
+      "grad_norm": 0.005482307635247707,
+      "learning_rate": 0.0006617283950617284,
+      "loss": 0.5776,
+      "step": 1082
+    },
+    {
+      "epoch": 2.405998333796168,
+      "grad_norm": 0.0035053386818617582,
+      "learning_rate": 0.0006592592592592593,
+      "loss": 0.4751,
+      "step": 1083
+    },
+    {
+      "epoch": 2.4082199389058596,
+      "grad_norm": 0.0036530292127281427,
+      "learning_rate": 0.0006567901234567901,
+      "loss": 0.7041,
+      "step": 1084
+    },
+    {
+      "epoch": 2.4104415440155513,
+      "grad_norm": 0.004132540430873632,
+      "learning_rate": 0.000654320987654321,
+      "loss": 0.5617,
+      "step": 1085
+    },
+    {
+      "epoch": 2.412663149125243,
+      "grad_norm": 0.004573510028421879,
+      "learning_rate": 0.0006518518518518518,
+      "loss": 0.4957,
+      "step": 1086
+    },
+    {
+      "epoch": 2.4148847542349348,
+      "grad_norm": 0.004237889777868986,
+      "learning_rate": 0.0006493827160493827,
+      "loss": 0.5725,
+      "step": 1087
+    },
+    {
+      "epoch": 2.4171063593446265,
+      "grad_norm": 0.00312399142421782,
+      "learning_rate": 0.0006469135802469136,
+      "loss": 0.5986,
+      "step": 1088
+    },
+    {
+      "epoch": 2.419327964454318,
+      "grad_norm": 0.003359878435730934,
+      "learning_rate": 0.0006444444444444444,
+      "loss": 0.5232,
+      "step": 1089
+    },
+    {
+      "epoch": 2.42154956956401,
+      "grad_norm": 0.005386237986385822,
+      "learning_rate": 0.0006419753086419753,
+      "loss": 0.4914,
+      "step": 1090
+    },
+    {
+      "epoch": 2.4237711746737016,
+      "grad_norm": 0.009064272977411747,
+      "learning_rate": 0.0006395061728395061,
+      "loss": 0.5483,
+      "step": 1091
+    },
+    {
+      "epoch": 2.4259927797833933,
+      "grad_norm": 0.0025790943764150143,
+      "learning_rate": 0.0006370370370370371,
+      "loss": 0.5408,
+      "step": 1092
+    },
+    {
+      "epoch": 2.428214384893085,
+      "grad_norm": 0.004174118861556053,
+      "learning_rate": 0.0006345679012345679,
+      "loss": 0.5581,
+      "step": 1093
+    },
+    {
+      "epoch": 2.4304359900027768,
+      "grad_norm": 0.004470566753298044,
+      "learning_rate": 0.0006320987654320988,
+      "loss": 0.5416,
+      "step": 1094
+    },
+    {
+      "epoch": 2.432657595112469,
+      "grad_norm": 0.003054042812436819,
+      "learning_rate": 0.0006296296296296296,
+      "loss": 0.5733,
+      "step": 1095
+    },
+    {
+      "epoch": 2.4348792002221606,
+      "grad_norm": 0.005424006376415491,
+      "learning_rate": 0.0006271604938271606,
+      "loss": 0.7099,
+      "step": 1096
+    },
+    {
+      "epoch": 2.4371008053318524,
+      "grad_norm": 0.004061280284076929,
+      "learning_rate": 0.0006246913580246914,
+      "loss": 0.6355,
+      "step": 1097
+    },
+    {
+      "epoch": 2.439322410441544,
+      "grad_norm": 0.004855146631598473,
+      "learning_rate": 0.0006222222222222223,
+      "loss": 0.5481,
+      "step": 1098
+    },
+    {
+      "epoch": 2.441544015551236,
+      "grad_norm": 0.002916454104706645,
+      "learning_rate": 0.0006197530864197531,
+      "loss": 0.6916,
+      "step": 1099
+    },
+    {
+      "epoch": 2.4437656206609275,
+      "grad_norm": 0.003550430526956916,
+      "learning_rate": 0.0006172839506172839,
+      "loss": 0.5293,
+      "step": 1100
+    },
+    {
+      "epoch": 2.4459872257706192,
+      "grad_norm": 0.00397482980042696,
+      "learning_rate": 0.0006148148148148148,
+      "loss": 0.5562,
+      "step": 1101
+    },
+    {
+      "epoch": 2.448208830880311,
+      "grad_norm": 0.006127642467617989,
+      "learning_rate": 0.0006123456790123456,
+      "loss": 0.7633,
+      "step": 1102
+    },
+    {
+      "epoch": 2.4504304359900027,
+      "grad_norm": 0.002872834214940667,
+      "learning_rate": 0.0006098765432098766,
+      "loss": 0.4499,
+      "step": 1103
+    },
+    {
+      "epoch": 2.4526520410996944,
+      "grad_norm": 0.0024833932984620333,
+      "learning_rate": 0.0006074074074074074,
+      "loss": 0.5494,
+      "step": 1104
+    },
+    {
+      "epoch": 2.4548736462093865,
+      "grad_norm": 0.0026254167314618826,
+      "learning_rate": 0.0006049382716049383,
+      "loss": 0.5676,
+      "step": 1105
+    },
+    {
+      "epoch": 2.4570952513190782,
+      "grad_norm": 0.0036874699871987104,
+      "learning_rate": 0.0006024691358024691,
+      "loss": 0.5925,
+      "step": 1106
+    },
+    {
+      "epoch": 2.45931685642877,
+      "grad_norm": 0.005146813113242388,
+      "learning_rate": 0.0006000000000000001,
+      "loss": 0.6584,
+      "step": 1107
+    },
+    {
+      "epoch": 2.4615384615384617,
+      "grad_norm": 0.0026747530791908503,
+      "learning_rate": 0.0005975308641975309,
+      "loss": 0.6409,
+      "step": 1108
+    },
+    {
+      "epoch": 2.4637600666481534,
+      "grad_norm": 0.004561138339340687,
+      "learning_rate": 0.0005950617283950617,
+      "loss": 0.4832,
+      "step": 1109
+    },
+    {
+      "epoch": 2.465981671757845,
+      "grad_norm": 0.0025099292397499084,
+      "learning_rate": 0.0005925925925925926,
+      "loss": 0.6076,
+      "step": 1110
+    },
+    {
+      "epoch": 2.468203276867537,
+      "grad_norm": 0.005029266234487295,
+      "learning_rate": 0.0005901234567901235,
+      "loss": 0.5094,
+      "step": 1111
+    },
+    {
+      "epoch": 2.4704248819772285,
+      "grad_norm": 0.0034568537957966328,
+      "learning_rate": 0.0005876543209876544,
+      "loss": 0.6715,
+      "step": 1112
+    },
+    {
+      "epoch": 2.4726464870869203,
+      "grad_norm": 0.0031753075309097767,
+      "learning_rate": 0.0005851851851851852,
+      "loss": 0.8845,
+      "step": 1113
+    },
+    {
+      "epoch": 2.474868092196612,
+      "grad_norm": 0.003458821913227439,
+      "learning_rate": 0.0005827160493827161,
+      "loss": 0.8207,
+      "step": 1114
+    },
+    {
+      "epoch": 2.4770896973063037,
+      "grad_norm": 0.0038800067268311977,
+      "learning_rate": 0.0005802469135802469,
+      "loss": 0.6511,
+      "step": 1115
+    },
+    {
+      "epoch": 2.4793113024159954,
+      "grad_norm": 0.002908184193074703,
+      "learning_rate": 0.0005777777777777778,
+      "loss": 0.5199,
+      "step": 1116
+    },
+    {
+      "epoch": 2.481532907525687,
+      "grad_norm": 0.0035255809780210257,
+      "learning_rate": 0.0005753086419753086,
+      "loss": 0.5549,
+      "step": 1117
+    },
+    {
+      "epoch": 2.483754512635379,
+      "grad_norm": 0.003498326987028122,
+      "learning_rate": 0.0005728395061728395,
+      "loss": 0.6323,
+      "step": 1118
+    },
+    {
+      "epoch": 2.485976117745071,
+      "grad_norm": 0.003741385880857706,
+      "learning_rate": 0.0005703703703703704,
+      "loss": 0.5352,
+      "step": 1119
+    },
+    {
+      "epoch": 2.4881977228547627,
+      "grad_norm": 0.003014420159161091,
+      "learning_rate": 0.0005679012345679012,
+      "loss": 0.5901,
+      "step": 1120
+    },
+    {
+      "epoch": 2.4904193279644544,
+      "grad_norm": 0.0028464593924582005,
+      "learning_rate": 0.0005654320987654321,
+      "loss": 0.5738,
+      "step": 1121
+    },
+    {
+      "epoch": 2.492640933074146,
+      "grad_norm": 0.003817921504378319,
+      "learning_rate": 0.000562962962962963,
+      "loss": 0.611,
+      "step": 1122
+    },
+    {
+      "epoch": 2.494862538183838,
+      "grad_norm": 0.004369066096842289,
+      "learning_rate": 0.0005604938271604939,
+      "loss": 0.8018,
+      "step": 1123
+    },
+    {
+      "epoch": 2.4970841432935296,
+      "grad_norm": 0.003815658623352647,
+      "learning_rate": 0.0005580246913580247,
+      "loss": 0.6316,
+      "step": 1124
+    },
+    {
+      "epoch": 2.4993057484032213,
+      "grad_norm": 0.0025187828578054905,
+      "learning_rate": 0.0005555555555555556,
+      "loss": 0.6251,
+      "step": 1125
+    },
+    {
+      "epoch": 2.501527353512913,
+      "grad_norm": 0.0034318508114665747,
+      "learning_rate": 0.0005530864197530865,
+      "loss": 0.6364,
+      "step": 1126
+    },
+    {
+      "epoch": 2.5037489586226047,
+      "grad_norm": 0.003666200442239642,
+      "learning_rate": 0.0005506172839506173,
+      "loss": 0.529,
+      "step": 1127
+    },
+    {
+      "epoch": 2.5059705637322964,
+      "grad_norm": 0.0033737020567059517,
+      "learning_rate": 0.0005481481481481482,
+      "loss": 0.499,
+      "step": 1128
+    },
+    {
+      "epoch": 2.5081921688419886,
+      "grad_norm": 0.004506078083068132,
+      "learning_rate": 0.000545679012345679,
+      "loss": 0.4437,
+      "step": 1129
+    },
+    {
+      "epoch": 2.5104137739516803,
+      "grad_norm": 0.004632357507944107,
+      "learning_rate": 0.0005432098765432099,
+      "loss": 0.5765,
+      "step": 1130
+    },
+    {
+      "epoch": 2.512635379061372,
+      "grad_norm": 0.0025815838016569614,
+      "learning_rate": 0.0005407407407407407,
+      "loss": 0.5222,
+      "step": 1131
+    },
+    {
+      "epoch": 2.5148569841710637,
+      "grad_norm": 0.002767681609839201,
+      "learning_rate": 0.0005382716049382716,
+      "loss": 0.6225,
+      "step": 1132
+    },
+    {
+      "epoch": 2.5170785892807555,
+      "grad_norm": 0.0037222790997475386,
+      "learning_rate": 0.0005358024691358024,
+      "loss": 0.685,
+      "step": 1133
+    },
+    {
+      "epoch": 2.519300194390447,
+      "grad_norm": 0.0034634550102055073,
+      "learning_rate": 0.0005333333333333334,
+      "loss": 0.5363,
+      "step": 1134
+    },
+    {
+      "epoch": 2.521521799500139,
+      "grad_norm": 0.0035786996595561504,
+      "learning_rate": 0.0005308641975308642,
+      "loss": 0.5579,
+      "step": 1135
+    },
+    {
+      "epoch": 2.5237434046098306,
+      "grad_norm": 0.0031474498100578785,
+      "learning_rate": 0.000528395061728395,
+      "loss": 0.6542,
+      "step": 1136
+    },
+    {
+      "epoch": 2.5259650097195223,
+      "grad_norm": 0.0023730038665235043,
+      "learning_rate": 0.000525925925925926,
+      "loss": 0.4511,
+      "step": 1137
+    },
+    {
+      "epoch": 2.528186614829214,
+      "grad_norm": 0.003312505781650543,
+      "learning_rate": 0.0005234567901234568,
+      "loss": 0.6185,
+      "step": 1138
+    },
+    {
+      "epoch": 2.5304082199389057,
+      "grad_norm": 0.004715273156762123,
+      "learning_rate": 0.0005209876543209877,
+      "loss": 0.6263,
+      "step": 1139
+    },
+    {
+      "epoch": 2.5326298250485975,
+      "grad_norm": 0.0033239570911973715,
+      "learning_rate": 0.0005185185185185185,
+      "loss": 0.5316,
+      "step": 1140
+    },
+    {
+      "epoch": 2.534851430158289,
+      "grad_norm": 0.003266816260293126,
+      "learning_rate": 0.0005160493827160495,
+      "loss": 0.6333,
+      "step": 1141
+    },
+    {
+      "epoch": 2.537073035267981,
+      "grad_norm": 0.0032662188168615103,
+      "learning_rate": 0.0005135802469135803,
+      "loss": 0.6425,
+      "step": 1142
+    },
+    {
+      "epoch": 2.5392946403776726,
+      "grad_norm": 0.0029916493222117424,
+      "learning_rate": 0.0005111111111111112,
+      "loss": 0.6858,
+      "step": 1143
+    },
+    {
+      "epoch": 2.5415162454873648,
+      "grad_norm": 0.003962009213864803,
+      "learning_rate": 0.000508641975308642,
+      "loss": 0.506,
+      "step": 1144
+    },
+    {
+      "epoch": 2.5437378505970565,
+      "grad_norm": 0.003180817700922489,
+      "learning_rate": 0.0005061728395061728,
+      "loss": 0.6995,
+      "step": 1145
+    },
+    {
+      "epoch": 2.545959455706748,
+      "grad_norm": 0.004911080002784729,
+      "learning_rate": 0.0005037037037037037,
+      "loss": 0.6845,
+      "step": 1146
+    },
+    {
+      "epoch": 2.54818106081644,
+      "grad_norm": 0.0032035394106060266,
+      "learning_rate": 0.0005012345679012345,
+      "loss": 0.5308,
+      "step": 1147
+    },
+    {
+      "epoch": 2.5504026659261316,
+      "grad_norm": 0.00372600881382823,
+      "learning_rate": 0.0004987654320987654,
+      "loss": 0.6681,
+      "step": 1148
+    },
+    {
+      "epoch": 2.5526242710358233,
+      "grad_norm": 0.002806935226544738,
+      "learning_rate": 0.0004962962962962963,
+      "loss": 0.5279,
+      "step": 1149
+    },
+    {
+      "epoch": 2.554845876145515,
+      "grad_norm": 0.002871694741770625,
+      "learning_rate": 0.0004938271604938272,
+      "loss": 0.6015,
+      "step": 1150
+    },
+    {
+      "epoch": 2.5570674812552068,
+      "grad_norm": 0.004236211068928242,
+      "learning_rate": 0.000491358024691358,
+      "loss": 0.7049,
+      "step": 1151
+    },
+    {
+      "epoch": 2.5592890863648985,
+      "grad_norm": 0.003672770457342267,
+      "learning_rate": 0.000488888888888889,
+      "loss": 0.7154,
+      "step": 1152
+    },
+    {
+      "epoch": 2.5615106914745907,
+      "grad_norm": 0.0026950067840516567,
+      "learning_rate": 0.00048641975308641976,
+      "loss": 0.5824,
+      "step": 1153
+    },
+    {
+      "epoch": 2.5637322965842824,
+      "grad_norm": 0.003053544322028756,
+      "learning_rate": 0.0004839506172839506,
+      "loss": 0.4743,
+      "step": 1154
+    },
+    {
+      "epoch": 2.565953901693974,
+      "grad_norm": 0.006343662738800049,
+      "learning_rate": 0.00048148148148148144,
+      "loss": 0.5234,
+      "step": 1155
+    },
+    {
+      "epoch": 2.568175506803666,
+      "grad_norm": 0.002904749009758234,
+      "learning_rate": 0.0004790123456790124,
+      "loss": 0.8594,
+      "step": 1156
+    },
+    {
+      "epoch": 2.5703971119133575,
+      "grad_norm": 0.003864643629640341,
+      "learning_rate": 0.00047654320987654324,
+      "loss": 0.6856,
+      "step": 1157
+    },
+    {
+      "epoch": 2.5726187170230492,
+      "grad_norm": 0.005410465411841869,
+      "learning_rate": 0.0004740740740740741,
+      "loss": 0.5771,
+      "step": 1158
+    },
+    {
+      "epoch": 2.574840322132741,
+      "grad_norm": 0.0040793525986373425,
+      "learning_rate": 0.0004716049382716049,
+      "loss": 0.5899,
+      "step": 1159
+    },
+    {
+      "epoch": 2.5770619272424327,
+      "grad_norm": 0.003131742123514414,
+      "learning_rate": 0.0004691358024691358,
+      "loss": 0.5691,
+      "step": 1160
+    },
+    {
+      "epoch": 2.5792835323521244,
+      "grad_norm": 0.0028007898945361376,
+      "learning_rate": 0.00046666666666666666,
+      "loss": 0.8081,
+      "step": 1161
+    },
+    {
+      "epoch": 2.581505137461816,
+      "grad_norm": 0.0028667037840932608,
+      "learning_rate": 0.0004641975308641975,
+      "loss": 0.6398,
+      "step": 1162
+    },
+    {
+      "epoch": 2.583726742571508,
+      "grad_norm": 0.002876229351386428,
+      "learning_rate": 0.00046172839506172835,
+      "loss": 0.6649,
+      "step": 1163
+    },
+    {
+      "epoch": 2.5859483476811995,
+      "grad_norm": 0.0037671527825295925,
+      "learning_rate": 0.0004592592592592593,
+      "loss": 0.7111,
+      "step": 1164
+    },
+    {
+      "epoch": 2.5881699527908912,
+      "grad_norm": 0.004723631776869297,
+      "learning_rate": 0.00045679012345679014,
+      "loss": 0.4845,
+      "step": 1165
+    },
+    {
+      "epoch": 2.590391557900583,
+      "grad_norm": 0.004219359718263149,
+      "learning_rate": 0.000454320987654321,
+      "loss": 0.6386,
+      "step": 1166
+    },
+    {
+      "epoch": 2.5926131630102747,
+      "grad_norm": 0.002575064543634653,
+      "learning_rate": 0.0004518518518518519,
+      "loss": 0.79,
+      "step": 1167
+    },
+    {
+      "epoch": 2.594834768119967,
+      "grad_norm": 0.0030822402331978083,
+      "learning_rate": 0.0004493827160493827,
+      "loss": 0.6529,
+      "step": 1168
+    },
+    {
+      "epoch": 2.5970563732296585,
+      "grad_norm": 0.0026442541275173426,
+      "learning_rate": 0.00044691358024691357,
+      "loss": 0.6858,
+      "step": 1169
+    },
+    {
+      "epoch": 2.5992779783393503,
+      "grad_norm": 0.004252356477081776,
+      "learning_rate": 0.0004444444444444444,
+      "loss": 0.6116,
+      "step": 1170
+    },
+    {
+      "epoch": 2.601499583449042,
+      "grad_norm": 0.002785630989819765,
+      "learning_rate": 0.00044197530864197536,
+      "loss": 0.5769,
+      "step": 1171
+    },
+    {
+      "epoch": 2.6037211885587337,
+      "grad_norm": 0.0027077009435743093,
+      "learning_rate": 0.0004395061728395062,
+      "loss": 0.7772,
+      "step": 1172
+    },
+    {
+      "epoch": 2.6059427936684254,
+      "grad_norm": 0.0035961696412414312,
+      "learning_rate": 0.00043703703703703705,
+      "loss": 0.5475,
+      "step": 1173
+    },
+    {
+      "epoch": 2.608164398778117,
+      "grad_norm": 0.0038466020487248898,
+      "learning_rate": 0.0004345679012345679,
+      "loss": 0.661,
+      "step": 1174
+    },
+    {
+      "epoch": 2.610386003887809,
+      "grad_norm": 0.0033505731262266636,
+      "learning_rate": 0.0004320987654320988,
+      "loss": 0.5725,
+      "step": 1175
+    },
+    {
+      "epoch": 2.6126076089975006,
+      "grad_norm": 0.004187949933111668,
+      "learning_rate": 0.00042962962962962963,
+      "loss": 0.5103,
+      "step": 1176
+    },
+    {
+      "epoch": 2.6148292141071927,
+      "grad_norm": 0.009911083616316319,
+      "learning_rate": 0.0004271604938271605,
+      "loss": 0.6615,
+      "step": 1177
+    },
+    {
+      "epoch": 2.6170508192168844,
+      "grad_norm": 0.003106280229985714,
+      "learning_rate": 0.0004246913580246913,
+      "loss": 0.4906,
+      "step": 1178
+    },
+    {
+      "epoch": 2.619272424326576,
+      "grad_norm": 0.0043709734454751015,
+      "learning_rate": 0.00042222222222222227,
+      "loss": 0.7297,
+      "step": 1179
+    },
+    {
+      "epoch": 2.621494029436268,
+      "grad_norm": 0.0024783313274383545,
+      "learning_rate": 0.0004197530864197531,
+      "loss": 0.654,
+      "step": 1180
+    },
+    {
+      "epoch": 2.6237156345459596,
+      "grad_norm": 0.002610607771202922,
+      "learning_rate": 0.00041728395061728396,
+      "loss": 0.5909,
+      "step": 1181
+    },
+    {
+      "epoch": 2.6259372396556513,
+      "grad_norm": 0.006081472150981426,
+      "learning_rate": 0.0004148148148148148,
+      "loss": 0.6009,
+      "step": 1182
+    },
+    {
+      "epoch": 2.628158844765343,
+      "grad_norm": 0.0038016068283468485,
+      "learning_rate": 0.0004123456790123457,
+      "loss": 0.7115,
+      "step": 1183
+    },
+    {
+      "epoch": 2.6303804498750347,
+      "grad_norm": 0.0031236426439136267,
+      "learning_rate": 0.00040987654320987654,
+      "loss": 0.41,
+      "step": 1184
+    },
+    {
+      "epoch": 2.6326020549847264,
+      "grad_norm": 0.003347929334267974,
+      "learning_rate": 0.0004074074074074074,
+      "loss": 0.6007,
+      "step": 1185
+    },
+    {
+      "epoch": 2.634823660094418,
+      "grad_norm": 0.00643836660310626,
+      "learning_rate": 0.00040493827160493833,
+      "loss": 0.5467,
+      "step": 1186
+    },
+    {
+      "epoch": 2.63704526520411,
+      "grad_norm": 0.0026751633267849684,
+      "learning_rate": 0.0004024691358024692,
+      "loss": 0.5962,
+      "step": 1187
+    },
+    {
+      "epoch": 2.6392668703138016,
+      "grad_norm": 0.005678553134202957,
+      "learning_rate": 0.0004,
+      "loss": 0.6097,
+      "step": 1188
+    },
+    {
+      "epoch": 2.6414884754234933,
+      "grad_norm": 0.007360636722296476,
+      "learning_rate": 0.00039753086419753086,
+      "loss": 0.6594,
+      "step": 1189
+    },
+    {
+      "epoch": 2.643710080533185,
+      "grad_norm": 0.0031560687348246574,
+      "learning_rate": 0.00039506172839506176,
+      "loss": 0.6907,
+      "step": 1190
+    },
+    {
+      "epoch": 2.6459316856428767,
+      "grad_norm": 0.002814547624439001,
+      "learning_rate": 0.0003925925925925926,
+      "loss": 0.6022,
+      "step": 1191
+    },
+    {
+      "epoch": 2.648153290752569,
+      "grad_norm": 0.0033059653360396624,
+      "learning_rate": 0.00039012345679012345,
+      "loss": 0.6396,
+      "step": 1192
+    },
+    {
+      "epoch": 2.6503748958622606,
+      "grad_norm": 0.004498746711760759,
+      "learning_rate": 0.0003876543209876543,
+      "loss": 0.6909,
+      "step": 1193
+    },
+    {
+      "epoch": 2.6525965009719523,
+      "grad_norm": 0.0029875014442950487,
+      "learning_rate": 0.00038518518518518524,
+      "loss": 0.485,
+      "step": 1194
+    },
+    {
+      "epoch": 2.654818106081644,
+      "grad_norm": 0.0031488672830164433,
+      "learning_rate": 0.0003827160493827161,
+      "loss": 0.5441,
+      "step": 1195
+    },
+    {
+      "epoch": 2.6570397111913358,
+      "grad_norm": 0.003440035041421652,
+      "learning_rate": 0.0003802469135802469,
+      "loss": 0.7539,
+      "step": 1196
+    },
+    {
+      "epoch": 2.6592613163010275,
+      "grad_norm": 0.0036845544818788767,
+      "learning_rate": 0.00037777777777777777,
+      "loss": 0.7387,
+      "step": 1197
+    },
+    {
+      "epoch": 2.661482921410719,
+      "grad_norm": 0.0035468495916575193,
+      "learning_rate": 0.00037530864197530867,
+      "loss": 0.6392,
+      "step": 1198
+    },
+    {
+      "epoch": 2.663704526520411,
+      "grad_norm": 0.0034712713677436113,
+      "learning_rate": 0.0003728395061728395,
+      "loss": 0.6942,
+      "step": 1199
+    },
+    {
+      "epoch": 2.6659261316301026,
+      "grad_norm": 0.0027431834023445845,
+      "learning_rate": 0.00037037037037037035,
+      "loss": 0.433,
+      "step": 1200
+    },
+    {
+      "epoch": 2.6681477367397948,
+      "grad_norm": 0.011069080792367458,
+      "learning_rate": 0.00036790123456790125,
+      "loss": 0.6633,
+      "step": 1201
+    },
+    {
+      "epoch": 2.6703693418494865,
+      "grad_norm": 0.008258162997663021,
+      "learning_rate": 0.00036543209876543215,
+      "loss": 0.4325,
+      "step": 1202
+    },
+    {
+      "epoch": 2.672590946959178,
+      "grad_norm": 0.003486806061118841,
+      "learning_rate": 0.000362962962962963,
+      "loss": 0.5709,
+      "step": 1203
+    },
+    {
+      "epoch": 2.67481255206887,
+      "grad_norm": 0.003325529396533966,
+      "learning_rate": 0.00036049382716049383,
+      "loss": 0.5366,
+      "step": 1204
+    },
+    {
+      "epoch": 2.6770341571785616,
+      "grad_norm": 0.0030097831040620804,
+      "learning_rate": 0.0003580246913580247,
+      "loss": 0.577,
+      "step": 1205
+    },
+    {
+      "epoch": 2.6792557622882534,
+      "grad_norm": 0.0038523420225828886,
+      "learning_rate": 0.00035555555555555557,
+      "loss": 0.7107,
+      "step": 1206
+    },
+    {
+      "epoch": 2.681477367397945,
+      "grad_norm": 0.0037631280720233917,
+      "learning_rate": 0.0003530864197530864,
+      "loss": 0.6634,
+      "step": 1207
+    },
+    {
+      "epoch": 2.683698972507637,
+      "grad_norm": 0.0037562635261565447,
+      "learning_rate": 0.0003506172839506173,
+      "loss": 0.5952,
+      "step": 1208
+    },
+    {
+      "epoch": 2.6859205776173285,
+      "grad_norm": 0.004943422507494688,
+      "learning_rate": 0.00034814814814814816,
+      "loss": 0.7673,
+      "step": 1209
+    },
+    {
+      "epoch": 2.68814218272702,
+      "grad_norm": 0.0029032754246145487,
+      "learning_rate": 0.00034567901234567905,
+      "loss": 0.5417,
+      "step": 1210
+    },
+    {
+      "epoch": 2.690363787836712,
+      "grad_norm": 0.0029870313592255116,
+      "learning_rate": 0.0003432098765432099,
+      "loss": 0.5779,
+      "step": 1211
+    },
+    {
+      "epoch": 2.6925853929464036,
+      "grad_norm": 0.003654767759144306,
+      "learning_rate": 0.00034074074074074074,
+      "loss": 0.5605,
+      "step": 1212
+    },
+    {
+      "epoch": 2.6948069980560954,
+      "grad_norm": 0.00890091247856617,
+      "learning_rate": 0.0003382716049382716,
+      "loss": 0.637,
+      "step": 1213
+    },
+    {
+      "epoch": 2.697028603165787,
+      "grad_norm": 0.003046605037525296,
+      "learning_rate": 0.0003358024691358025,
+      "loss": 0.5595,
+      "step": 1214
+    },
+    {
+      "epoch": 2.699250208275479,
+      "grad_norm": 0.0023681134916841984,
+      "learning_rate": 0.0003333333333333333,
+      "loss": 0.6964,
+      "step": 1215
+    },
+    {
+      "epoch": 2.701471813385171,
+      "grad_norm": 0.0029847752302885056,
+      "learning_rate": 0.0003308641975308642,
+      "loss": 0.6097,
+      "step": 1216
+    },
+    {
+      "epoch": 2.7036934184948627,
+      "grad_norm": 0.005827764980494976,
+      "learning_rate": 0.00032839506172839506,
+      "loss": 0.5802,
+      "step": 1217
+    },
+    {
+      "epoch": 2.7059150236045544,
+      "grad_norm": 0.008738023228943348,
+      "learning_rate": 0.0003259259259259259,
+      "loss": 0.7378,
+      "step": 1218
+    },
+    {
+      "epoch": 2.708136628714246,
+      "grad_norm": 0.004061623010784388,
+      "learning_rate": 0.0003234567901234568,
+      "loss": 0.6561,
+      "step": 1219
+    },
+    {
+      "epoch": 2.710358233823938,
+      "grad_norm": 0.003264094004407525,
+      "learning_rate": 0.00032098765432098765,
+      "loss": 0.6734,
+      "step": 1220
+    },
+    {
+      "epoch": 2.7125798389336295,
+      "grad_norm": 0.002878582803532481,
+      "learning_rate": 0.00031851851851851854,
+      "loss": 0.5257,
+      "step": 1221
+    },
+    {
+      "epoch": 2.7148014440433212,
+      "grad_norm": 0.0038063591346144676,
+      "learning_rate": 0.0003160493827160494,
+      "loss": 0.6192,
+      "step": 1222
+    },
+    {
+      "epoch": 2.717023049153013,
+      "grad_norm": 0.0025181041564792395,
+      "learning_rate": 0.0003135802469135803,
+      "loss": 0.616,
+      "step": 1223
+    },
+    {
+      "epoch": 2.7192446542627047,
+      "grad_norm": 0.003332714084535837,
+      "learning_rate": 0.0003111111111111111,
+      "loss": 0.5767,
+      "step": 1224
+    },
+    {
+      "epoch": 2.7214662593723964,
+      "grad_norm": 0.004335831850767136,
+      "learning_rate": 0.00030864197530864197,
+      "loss": 0.6023,
+      "step": 1225
+    },
+    {
+      "epoch": 2.7236878644820885,
+      "grad_norm": 0.0024720283690840006,
+      "learning_rate": 0.0003061728395061728,
+      "loss": 0.5528,
+      "step": 1226
+    },
+    {
+      "epoch": 2.7259094695917803,
+      "grad_norm": 0.005799434147775173,
+      "learning_rate": 0.0003037037037037037,
+      "loss": 0.6276,
+      "step": 1227
+    },
+    {
+      "epoch": 2.728131074701472,
+      "grad_norm": 0.006058346480131149,
+      "learning_rate": 0.00030123456790123455,
+      "loss": 0.5782,
+      "step": 1228
+    },
+    {
+      "epoch": 2.7303526798111637,
+      "grad_norm": 0.0030225799418985844,
+      "learning_rate": 0.00029876543209876545,
+      "loss": 0.5824,
+      "step": 1229
+    },
+    {
+      "epoch": 2.7325742849208554,
+      "grad_norm": 0.01594926044344902,
+      "learning_rate": 0.0002962962962962963,
+      "loss": 0.6547,
+      "step": 1230
+    },
+    {
+      "epoch": 2.734795890030547,
+      "grad_norm": 0.0027183282654732466,
+      "learning_rate": 0.0002938271604938272,
+      "loss": 0.6479,
+      "step": 1231
+    },
+    {
+      "epoch": 2.737017495140239,
+      "grad_norm": 0.010722714476287365,
+      "learning_rate": 0.00029135802469135803,
+      "loss": 0.5311,
+      "step": 1232
+    },
+    {
+      "epoch": 2.7392391002499306,
+      "grad_norm": 0.0027126965578645468,
+      "learning_rate": 0.0002888888888888889,
+      "loss": 0.4191,
+      "step": 1233
+    },
+    {
+      "epoch": 2.7414607053596223,
+      "grad_norm": 0.0027542004827409983,
+      "learning_rate": 0.00028641975308641977,
+      "loss": 0.6691,
+      "step": 1234
+    },
+    {
+      "epoch": 2.743682310469314,
+      "grad_norm": 0.003709719516336918,
+      "learning_rate": 0.0002839506172839506,
+      "loss": 0.4843,
+      "step": 1235
+    },
+    {
+      "epoch": 2.7459039155790057,
+      "grad_norm": 0.0035991393961012363,
+      "learning_rate": 0.0002814814814814815,
+      "loss": 0.6054,
+      "step": 1236
+    },
+    {
+      "epoch": 2.7481255206886974,
+      "grad_norm": 0.002744116587564349,
+      "learning_rate": 0.00027901234567901236,
+      "loss": 0.6879,
+      "step": 1237
+    },
+    {
+      "epoch": 2.750347125798389,
+      "grad_norm": 0.004210659768432379,
+      "learning_rate": 0.00027654320987654325,
+      "loss": 0.5684,
+      "step": 1238
+    },
+    {
+      "epoch": 2.752568730908081,
+      "grad_norm": 0.0034985432866960764,
+      "learning_rate": 0.0002740740740740741,
+      "loss": 0.4931,
+      "step": 1239
+    },
+    {
+      "epoch": 2.7547903360177726,
+      "grad_norm": 0.012942667119204998,
+      "learning_rate": 0.00027160493827160494,
+      "loss": 0.5202,
+      "step": 1240
+    },
+    {
+      "epoch": 2.7570119411274647,
+      "grad_norm": 0.0034006896894425154,
+      "learning_rate": 0.0002691358024691358,
+      "loss": 0.7363,
+      "step": 1241
+    },
+    {
+      "epoch": 2.7592335462371564,
+      "grad_norm": 0.002795079257339239,
+      "learning_rate": 0.0002666666666666667,
+      "loss": 0.5581,
+      "step": 1242
+    },
+    {
+      "epoch": 2.761455151346848,
+      "grad_norm": 0.0036135506816208363,
+      "learning_rate": 0.0002641975308641975,
+      "loss": 0.6858,
+      "step": 1243
+    },
+    {
+      "epoch": 2.76367675645654,
+      "grad_norm": 0.0034595513716340065,
+      "learning_rate": 0.0002617283950617284,
+      "loss": 0.5736,
+      "step": 1244
+    },
+    {
+      "epoch": 2.7658983615662316,
+      "grad_norm": 0.003965721465647221,
+      "learning_rate": 0.00025925925925925926,
+      "loss": 0.4689,
+      "step": 1245
+    },
+    {
+      "epoch": 2.7681199666759233,
+      "grad_norm": 0.004251439590007067,
+      "learning_rate": 0.00025679012345679016,
+      "loss": 0.4203,
+      "step": 1246
+    },
+    {
+      "epoch": 2.770341571785615,
+      "grad_norm": 0.004119568970054388,
+      "learning_rate": 0.000254320987654321,
+      "loss": 0.5342,
+      "step": 1247
+    },
+    {
+      "epoch": 2.7725631768953067,
+      "grad_norm": 0.0036970556247979403,
+      "learning_rate": 0.00025185185185185185,
+      "loss": 0.7664,
+      "step": 1248
+    },
+    {
+      "epoch": 2.7747847820049985,
+      "grad_norm": 0.005256022792309523,
+      "learning_rate": 0.0002493827160493827,
+      "loss": 0.6459,
+      "step": 1249
+    },
+    {
+      "epoch": 2.7770063871146906,
+      "grad_norm": 0.004164376296103001,
+      "learning_rate": 0.0002469135802469136,
+      "loss": 0.626,
+      "step": 1250
+    },
+    {
+      "epoch": 2.7792279922243823,
+      "grad_norm": 0.003417801344767213,
+      "learning_rate": 0.0002444444444444445,
+      "loss": 0.7121,
+      "step": 1251
+    },
+    {
+      "epoch": 2.781449597334074,
+      "grad_norm": 0.0032853151205927134,
+      "learning_rate": 0.0002419753086419753,
+      "loss": 0.6877,
+      "step": 1252
+    },
+    {
+      "epoch": 2.7836712024437658,
+      "grad_norm": 0.0036194906570017338,
+      "learning_rate": 0.0002395061728395062,
+      "loss": 0.6578,
+      "step": 1253
+    },
+    {
+      "epoch": 2.7858928075534575,
+      "grad_norm": 0.0030540688894689083,
+      "learning_rate": 0.00023703703703703704,
+      "loss": 0.7105,
+      "step": 1254
+    },
+    {
+      "epoch": 2.788114412663149,
+      "grad_norm": 0.002393856178969145,
+      "learning_rate": 0.0002345679012345679,
+      "loss": 0.57,
+      "step": 1255
+    },
+    {
+      "epoch": 2.790336017772841,
+      "grad_norm": 0.0026177624240517616,
+      "learning_rate": 0.00023209876543209875,
+      "loss": 0.548,
+      "step": 1256
+    },
+    {
+      "epoch": 2.7925576228825326,
+      "grad_norm": 0.0030573352705687284,
+      "learning_rate": 0.00022962962962962965,
+      "loss": 0.5298,
+      "step": 1257
+    },
+    {
+      "epoch": 2.7947792279922243,
+      "grad_norm": 0.0029401860665529966,
+      "learning_rate": 0.0002271604938271605,
+      "loss": 0.5248,
+      "step": 1258
+    },
+    {
+      "epoch": 2.797000833101916,
+      "grad_norm": 0.0033551547676324844,
+      "learning_rate": 0.00022469135802469136,
+      "loss": 0.6293,
+      "step": 1259
+    },
+    {
+      "epoch": 2.7992224382116078,
+      "grad_norm": 0.003985394258052111,
+      "learning_rate": 0.0002222222222222222,
+      "loss": 0.7248,
+      "step": 1260
+    },
+    {
+      "epoch": 2.8014440433212995,
+      "grad_norm": 0.0034075267612934113,
+      "learning_rate": 0.0002197530864197531,
+      "loss": 0.6553,
+      "step": 1261
+    },
+    {
+      "epoch": 2.803665648430991,
+      "grad_norm": 0.0031692166812717915,
+      "learning_rate": 0.00021728395061728395,
+      "loss": 0.7047,
+      "step": 1262
+    },
+    {
+      "epoch": 2.805887253540683,
+      "grad_norm": 0.003578015835955739,
+      "learning_rate": 0.00021481481481481482,
+      "loss": 0.4964,
+      "step": 1263
+    },
+    {
+      "epoch": 2.8081088586503746,
+      "grad_norm": 0.0034794467501342297,
+      "learning_rate": 0.00021234567901234566,
+      "loss": 0.8116,
+      "step": 1264
+    },
+    {
+      "epoch": 2.810330463760067,
+      "grad_norm": 0.003644336946308613,
+      "learning_rate": 0.00020987654320987656,
+      "loss": 0.6715,
+      "step": 1265
+    },
+    {
+      "epoch": 2.8125520688697585,
+      "grad_norm": 0.004753046203404665,
+      "learning_rate": 0.0002074074074074074,
+      "loss": 0.6864,
+      "step": 1266
+    },
+    {
+      "epoch": 2.81477367397945,
+      "grad_norm": 0.0034688313025981188,
+      "learning_rate": 0.00020493827160493827,
+      "loss": 0.6755,
+      "step": 1267
+    },
+    {
+      "epoch": 2.816995279089142,
+      "grad_norm": 0.003245983738452196,
+      "learning_rate": 0.00020246913580246917,
+      "loss": 0.5707,
+      "step": 1268
+    },
+    {
+      "epoch": 2.8192168841988337,
+      "grad_norm": 0.003639776026830077,
+      "learning_rate": 0.0002,
+      "loss": 0.7689,
+      "step": 1269
+    },
+    {
+      "epoch": 2.8214384893085254,
+      "grad_norm": 0.00351712410338223,
+      "learning_rate": 0.00019753086419753088,
+      "loss": 0.5325,
+      "step": 1270
+    },
+    {
+      "epoch": 2.823660094418217,
+      "grad_norm": 0.002562894020229578,
+      "learning_rate": 0.00019506172839506172,
+      "loss": 0.626,
+      "step": 1271
+    },
+    {
+      "epoch": 2.825881699527909,
+      "grad_norm": 0.003236958524212241,
+      "learning_rate": 0.00019259259259259262,
+      "loss": 0.4666,
+      "step": 1272
+    },
+    {
+      "epoch": 2.8281033046376005,
+      "grad_norm": 0.0036686372477561235,
+      "learning_rate": 0.00019012345679012346,
+      "loss": 0.6194,
+      "step": 1273
+    },
+    {
+      "epoch": 2.8303249097472927,
+      "grad_norm": 0.0032101618126034737,
+      "learning_rate": 0.00018765432098765433,
+      "loss": 0.4723,
+      "step": 1274
+    },
+    {
+      "epoch": 2.8325465148569844,
+      "grad_norm": 0.0028950897976756096,
+      "learning_rate": 0.00018518518518518518,
+      "loss": 0.5364,
+      "step": 1275
+    },
+    {
+      "epoch": 2.834768119966676,
+      "grad_norm": 0.004062660038471222,
+      "learning_rate": 0.00018271604938271607,
+      "loss": 0.5738,
+      "step": 1276
+    },
+    {
+      "epoch": 2.836989725076368,
+      "grad_norm": 0.003023022785782814,
+      "learning_rate": 0.00018024691358024692,
+      "loss": 0.6205,
+      "step": 1277
+    },
+    {
+      "epoch": 2.8392113301860595,
+      "grad_norm": 0.003414308186620474,
+      "learning_rate": 0.00017777777777777779,
+      "loss": 0.7143,
+      "step": 1278
+    },
+    {
+      "epoch": 2.8414329352957512,
+      "grad_norm": 0.002360349288210273,
+      "learning_rate": 0.00017530864197530866,
+      "loss": 0.485,
+      "step": 1279
+    },
+    {
+      "epoch": 2.843654540405443,
+      "grad_norm": 0.0031187781132757664,
+      "learning_rate": 0.00017283950617283953,
+      "loss": 0.5694,
+      "step": 1280
+    },
+    {
+      "epoch": 2.8458761455151347,
+      "grad_norm": 0.0037832462694495916,
+      "learning_rate": 0.00017037037037037037,
+      "loss": 0.6778,
+      "step": 1281
+    },
+    {
+      "epoch": 2.8480977506248264,
+      "grad_norm": 0.003156352788209915,
+      "learning_rate": 0.00016790123456790124,
+      "loss": 0.4658,
+      "step": 1282
+    },
+    {
+      "epoch": 2.850319355734518,
+      "grad_norm": 0.003529968671500683,
+      "learning_rate": 0.0001654320987654321,
+      "loss": 0.5322,
+      "step": 1283
+    },
+    {
+      "epoch": 2.85254096084421,
+      "grad_norm": 0.005195183213800192,
+      "learning_rate": 0.00016296296296296295,
+      "loss": 0.6196,
+      "step": 1284
+    },
+    {
+      "epoch": 2.8547625659539015,
+      "grad_norm": 0.0050943163223564625,
+      "learning_rate": 0.00016049382716049382,
+      "loss": 0.5934,
+      "step": 1285
+    },
+    {
+      "epoch": 2.8569841710635933,
+      "grad_norm": 0.003997886553406715,
+      "learning_rate": 0.0001580246913580247,
+      "loss": 0.7084,
+      "step": 1286
+    },
+    {
+      "epoch": 2.859205776173285,
+      "grad_norm": 0.0034028280060738325,
+      "learning_rate": 0.00015555555555555556,
+      "loss": 0.6223,
+      "step": 1287
+    },
+    {
+      "epoch": 2.8614273812829767,
+      "grad_norm": 0.0027945416513830423,
+      "learning_rate": 0.0001530864197530864,
+      "loss": 0.6469,
+      "step": 1288
+    },
+    {
+      "epoch": 2.863648986392669,
+      "grad_norm": 0.002943682251498103,
+      "learning_rate": 0.00015061728395061728,
+      "loss": 0.6867,
+      "step": 1289
+    },
+    {
+      "epoch": 2.8658705915023606,
+      "grad_norm": 0.0031444213818758726,
+      "learning_rate": 0.00014814814814814815,
+      "loss": 0.519,
+      "step": 1290
+    },
+    {
+      "epoch": 2.8680921966120523,
+      "grad_norm": 0.0026177004911005497,
+      "learning_rate": 0.00014567901234567902,
+      "loss": 0.6096,
+      "step": 1291
+    },
+    {
+      "epoch": 2.870313801721744,
+      "grad_norm": 0.002935094991698861,
+      "learning_rate": 0.00014320987654320989,
+      "loss": 0.6899,
+      "step": 1292
+    },
+    {
+      "epoch": 2.8725354068314357,
+      "grad_norm": 0.0028896513395011425,
+      "learning_rate": 0.00014074074074074076,
+      "loss": 0.7482,
+      "step": 1293
+    },
+    {
+      "epoch": 2.8747570119411274,
+      "grad_norm": 0.00413087010383606,
+      "learning_rate": 0.00013827160493827163,
+      "loss": 0.8511,
+      "step": 1294
+    },
+    {
+      "epoch": 2.876978617050819,
+      "grad_norm": 0.003740635933354497,
+      "learning_rate": 0.00013580246913580247,
+      "loss": 0.6157,
+      "step": 1295
+    },
+    {
+      "epoch": 2.879200222160511,
+      "grad_norm": 0.002238227752968669,
+      "learning_rate": 0.00013333333333333334,
+      "loss": 0.5067,
+      "step": 1296
+    },
+    {
+      "epoch": 2.8814218272702026,
+      "grad_norm": 0.0026507952716201544,
+      "learning_rate": 0.0001308641975308642,
+      "loss": 0.5867,
+      "step": 1297
+    },
+    {
+      "epoch": 2.8836434323798947,
+      "grad_norm": 0.0034062659833580256,
+      "learning_rate": 0.00012839506172839508,
+      "loss": 0.5274,
+      "step": 1298
+    },
+    {
+      "epoch": 2.8858650374895864,
+      "grad_norm": 0.0057266768999397755,
+      "learning_rate": 0.00012592592592592592,
+      "loss": 0.5766,
+      "step": 1299
+    },
+    {
+      "epoch": 2.888086642599278,
+      "grad_norm": 0.0024858866818249226,
+      "learning_rate": 0.0001234567901234568,
+      "loss": 0.5223,
+      "step": 1300
+    },
+    {
+      "epoch": 2.89030824770897,
+      "grad_norm": 0.0042802924290299416,
+      "learning_rate": 0.00012098765432098765,
+      "loss": 0.8749,
+      "step": 1301
+    },
+    {
+      "epoch": 2.8925298528186616,
+      "grad_norm": 0.003919759299606085,
+      "learning_rate": 0.00011851851851851852,
+      "loss": 0.5534,
+      "step": 1302
+    },
+    {
+      "epoch": 2.8947514579283533,
+      "grad_norm": 0.0030064480379223824,
+      "learning_rate": 0.00011604938271604938,
+      "loss": 0.4913,
+      "step": 1303
+    },
+    {
+      "epoch": 2.896973063038045,
+      "grad_norm": 0.0021056190598756075,
+      "learning_rate": 0.00011358024691358025,
+      "loss": 0.407,
+      "step": 1304
+    },
+    {
+      "epoch": 2.8991946681477367,
+      "grad_norm": 0.002747428370639682,
+      "learning_rate": 0.0001111111111111111,
+      "loss": 0.573,
+      "step": 1305
+    },
+    {
+      "epoch": 2.9014162732574285,
+      "grad_norm": 0.0034187480341643095,
+      "learning_rate": 0.00010864197530864197,
+      "loss": 0.555,
+      "step": 1306
+    },
+    {
+      "epoch": 2.90363787836712,
+      "grad_norm": 0.0028815080877393484,
+      "learning_rate": 0.00010617283950617283,
+      "loss": 0.5424,
+      "step": 1307
+    },
+    {
+      "epoch": 2.905859483476812,
+      "grad_norm": 0.004015353973954916,
+      "learning_rate": 0.0001037037037037037,
+      "loss": 0.646,
+      "step": 1308
+    },
+    {
+      "epoch": 2.9080810885865036,
+      "grad_norm": 0.0040834201499819756,
+      "learning_rate": 0.00010123456790123458,
+      "loss": 0.6989,
+      "step": 1309
+    },
+    {
+      "epoch": 2.9103026936961953,
+      "grad_norm": 0.0023935800418257713,
+      "learning_rate": 9.876543209876544e-05,
+      "loss": 0.5851,
+      "step": 1310
+    },
+    {
+      "epoch": 2.912524298805887,
+      "grad_norm": 0.0023539229296147823,
+      "learning_rate": 9.629629629629631e-05,
+      "loss": 0.4969,
+      "step": 1311
+    },
+    {
+      "epoch": 2.9147459039155788,
+      "grad_norm": 0.002933288924396038,
+      "learning_rate": 9.382716049382717e-05,
+      "loss": 0.6214,
+      "step": 1312
+    },
+    {
+      "epoch": 2.916967509025271,
+      "grad_norm": 0.004053778015077114,
+      "learning_rate": 9.135802469135804e-05,
+      "loss": 0.6662,
+      "step": 1313
+    },
+    {
+      "epoch": 2.9191891141349626,
+      "grad_norm": 0.0033682617358863354,
+      "learning_rate": 8.888888888888889e-05,
+      "loss": 0.7816,
+      "step": 1314
+    },
+    {
+      "epoch": 2.9214107192446543,
+      "grad_norm": 0.010227394290268421,
+      "learning_rate": 8.641975308641976e-05,
+      "loss": 0.5627,
+      "step": 1315
+    },
+    {
+      "epoch": 2.923632324354346,
+      "grad_norm": 0.003825727617368102,
+      "learning_rate": 8.395061728395062e-05,
+      "loss": 0.4702,
+      "step": 1316
+    },
+    {
+      "epoch": 2.9258539294640378,
+      "grad_norm": 0.003311006585136056,
+      "learning_rate": 8.148148148148148e-05,
+      "loss": 0.5711,
+      "step": 1317
+    },
+    {
+      "epoch": 2.9280755345737295,
+      "grad_norm": 0.004043097607791424,
+      "learning_rate": 7.901234567901235e-05,
+      "loss": 0.6537,
+      "step": 1318
+    },
+    {
+      "epoch": 2.930297139683421,
+      "grad_norm": 0.002848139265552163,
+      "learning_rate": 7.65432098765432e-05,
+      "loss": 0.6193,
+      "step": 1319
+    },
+    {
+      "epoch": 2.932518744793113,
+      "grad_norm": 0.002361015183851123,
+      "learning_rate": 7.407407407407407e-05,
+      "loss": 0.5331,
+      "step": 1320
+    },
+    {
+      "epoch": 2.9347403499028046,
+      "grad_norm": 0.004088804591447115,
+      "learning_rate": 7.160493827160494e-05,
+      "loss": 0.6029,
+      "step": 1321
+    },
+    {
+      "epoch": 2.936961955012497,
+      "grad_norm": 0.002269380958750844,
+      "learning_rate": 6.913580246913581e-05,
+      "loss": 0.4867,
+      "step": 1322
+    },
+    {
+      "epoch": 2.9391835601221885,
+      "grad_norm": 0.0039354427717626095,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 0.5974,
+      "step": 1323
+    },
+    {
+      "epoch": 2.9414051652318802,
+      "grad_norm": 0.004085162654519081,
+      "learning_rate": 6.419753086419754e-05,
+      "loss": 0.5142,
+      "step": 1324
+    },
+    {
+      "epoch": 2.943626770341572,
+      "grad_norm": 0.005218549631536007,
+      "learning_rate": 6.17283950617284e-05,
+      "loss": 0.6215,
+      "step": 1325
+    },
+    {
+      "epoch": 2.9458483754512637,
+      "grad_norm": 0.002707709791138768,
+      "learning_rate": 5.925925925925926e-05,
+      "loss": 0.6414,
+      "step": 1326
+    },
+    {
+      "epoch": 2.9480699805609554,
+      "grad_norm": 0.005078699439764023,
+      "learning_rate": 5.679012345679012e-05,
+      "loss": 0.5196,
+      "step": 1327
+    },
+    {
+      "epoch": 2.950291585670647,
+      "grad_norm": 0.0025226271245628595,
+      "learning_rate": 5.4320987654320986e-05,
+      "loss": 0.5715,
+      "step": 1328
+    },
+    {
+      "epoch": 2.952513190780339,
+      "grad_norm": 0.002955828094854951,
+      "learning_rate": 5.185185185185185e-05,
+      "loss": 0.6267,
+      "step": 1329
+    },
+    {
+      "epoch": 2.9547347958900305,
+      "grad_norm": 0.004617676604539156,
+      "learning_rate": 4.938271604938272e-05,
+      "loss": 0.5121,
+      "step": 1330
+    },
+    {
+      "epoch": 2.9569564009997222,
+      "grad_norm": 0.003442096756771207,
+      "learning_rate": 4.691358024691358e-05,
+      "loss": 0.6568,
+      "step": 1331
+    },
+    {
+      "epoch": 2.959178006109414,
+      "grad_norm": 0.003014144254848361,
+      "learning_rate": 4.4444444444444447e-05,
+      "loss": 0.6131,
+      "step": 1332
+    },
+    {
+      "epoch": 2.9613996112191057,
+      "grad_norm": 0.0031045388896018267,
+      "learning_rate": 4.197530864197531e-05,
+      "loss": 0.67,
+      "step": 1333
+    },
+    {
+      "epoch": 2.9636212163287974,
+      "grad_norm": 0.0025501151103526354,
+      "learning_rate": 3.950617283950617e-05,
+      "loss": 0.6671,
+      "step": 1334
+    },
+    {
+      "epoch": 2.965842821438489,
+      "grad_norm": 0.0036570553202182055,
+      "learning_rate": 3.7037037037037037e-05,
+      "loss": 0.5173,
+      "step": 1335
+    },
+    {
+      "epoch": 2.968064426548181,
+      "grad_norm": 0.0016833347035571933,
+      "learning_rate": 3.456790123456791e-05,
+      "loss": 0.4934,
+      "step": 1336
+    },
+    {
+      "epoch": 2.970286031657873,
+      "grad_norm": 0.003212490351870656,
+      "learning_rate": 3.209876543209877e-05,
+      "loss": 0.4975,
+      "step": 1337
+    },
+    {
+      "epoch": 2.9725076367675647,
+      "grad_norm": 0.0029618414118885994,
+      "learning_rate": 2.962962962962963e-05,
+      "loss": 0.5626,
+      "step": 1338
+    },
+    {
+      "epoch": 2.9747292418772564,
+      "grad_norm": 0.004005649592727423,
+      "learning_rate": 2.7160493827160493e-05,
+      "loss": 0.6448,
+      "step": 1339
+    },
+    {
+      "epoch": 2.976950846986948,
+      "grad_norm": 0.0033238399773836136,
+      "learning_rate": 2.469135802469136e-05,
+      "loss": 0.6304,
+      "step": 1340
+    },
+    {
+      "epoch": 2.97917245209664,
+      "grad_norm": 0.0032543411944061518,
+      "learning_rate": 2.2222222222222223e-05,
+      "loss": 0.5257,
+      "step": 1341
+    },
+    {
+      "epoch": 2.9813940572063315,
+      "grad_norm": 0.003472758922725916,
+      "learning_rate": 1.9753086419753087e-05,
+      "loss": 0.6997,
+      "step": 1342
+    },
+    {
+      "epoch": 2.9836156623160233,
+      "grad_norm": 0.003682138863950968,
+      "learning_rate": 1.7283950617283953e-05,
+      "loss": 0.6162,
+      "step": 1343
+    },
+    {
+      "epoch": 2.985837267425715,
+      "grad_norm": 0.0032471835147589445,
+      "learning_rate": 1.4814814814814815e-05,
+      "loss": 0.6362,
+      "step": 1344
+    },
+    {
+      "epoch": 2.9880588725354067,
+      "grad_norm": 0.0025808580685406923,
+      "learning_rate": 1.234567901234568e-05,
+      "loss": 0.5575,
+      "step": 1345
+    },
+    {
+      "epoch": 2.990280477645099,
+      "grad_norm": 0.00250285305082798,
+      "learning_rate": 9.876543209876543e-06,
+      "loss": 0.5708,
+      "step": 1346
+    },
+    {
+      "epoch": 2.9925020827547906,
+      "grad_norm": 0.00289158639498055,
+      "learning_rate": 7.4074074074074075e-06,
+      "loss": 0.5226,
+      "step": 1347
+    },
+    {
+      "epoch": 2.9947236878644823,
+      "grad_norm": 0.0033831135369837284,
+      "learning_rate": 4.938271604938272e-06,
+      "loss": 0.4506,
+      "step": 1348
+    },
+    {
+      "epoch": 2.996945292974174,
+      "grad_norm": 0.004179411567747593,
+      "learning_rate": 2.469135802469136e-06,
+      "loss": 0.601,
+      "step": 1349
+    },
+    {
+      "epoch": 2.9991668980838657,
+      "grad_norm": 0.002582500921562314,
+      "learning_rate": 0.0,
+      "loss": 0.5285,
+      "step": 1350
+    },
+    {
+      "epoch": 2.9991668980838657,
+      "step": 1350,
+      "total_flos": 5.0668751263105024e+17,
+      "train_loss": 0.6102046124140421,
+      "train_runtime": 4348.7477,
+      "train_samples_per_second": 4.968,
+      "train_steps_per_second": 0.31
+    }
+  ],
+  "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": 5.0668751263105024e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}