diff --git "a/checkpoint-1668/trainer_state.json" "b/checkpoint-1668/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1668/trainer_state.json"
@@ -0,0 +1,11709 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.7510130571814498,
+  "eval_steps": 500,
+  "global_step": 1668,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00045024763619990995,
+      "grad_norm": 0.0765572264790535,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.1244,
+      "step": 1
+    },
+    {
+      "epoch": 0.0009004952723998199,
+      "grad_norm": 0.06779415905475616,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.029,
+      "step": 2
+    },
+    {
+      "epoch": 0.0013507429085997298,
+      "grad_norm": 0.10786198079586029,
+      "learning_rate": 6e-06,
+      "loss": 1.1956,
+      "step": 3
+    },
+    {
+      "epoch": 0.0018009905447996398,
+      "grad_norm": 0.09927327185869217,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 1.0711,
+      "step": 4
+    },
+    {
+      "epoch": 0.0022512381809995496,
+      "grad_norm": 0.08395445346832275,
+      "learning_rate": 1e-05,
+      "loss": 1.003,
+      "step": 5
+    },
+    {
+      "epoch": 0.0027014858171994596,
+      "grad_norm": 0.08248680830001831,
+      "learning_rate": 1.2e-05,
+      "loss": 1.1811,
+      "step": 6
+    },
+    {
+      "epoch": 0.0031517334533993696,
+      "grad_norm": 0.08581385016441345,
+      "learning_rate": 1.4000000000000001e-05,
+      "loss": 1.1521,
+      "step": 7
+    },
+    {
+      "epoch": 0.0036019810895992796,
+      "grad_norm": 0.08581811934709549,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 1.0545,
+      "step": 8
+    },
+    {
+      "epoch": 0.00405222872579919,
+      "grad_norm": 0.06538940221071243,
+      "learning_rate": 1.8e-05,
+      "loss": 0.909,
+      "step": 9
+    },
+    {
+      "epoch": 0.004502476361999099,
+      "grad_norm": 0.12085781246423721,
+      "learning_rate": 2e-05,
+      "loss": 1.3947,
+      "step": 10
+    },
+    {
+      "epoch": 0.00495272399819901,
+      "grad_norm": 0.09600106626749039,
+      "learning_rate": 2.2000000000000003e-05,
+      "loss": 1.0388,
+      "step": 11
+    },
+    {
+      "epoch": 0.005402971634398919,
+      "grad_norm": 0.09513469785451889,
+      "learning_rate": 2.4e-05,
+      "loss": 0.9652,
+      "step": 12
+    },
+    {
+      "epoch": 0.00585321927059883,
+      "grad_norm": 0.11549637466669083,
+      "learning_rate": 2.6000000000000002e-05,
+      "loss": 0.957,
+      "step": 13
+    },
+    {
+      "epoch": 0.006303466906798739,
+      "grad_norm": 0.1442708671092987,
+      "learning_rate": 2.8000000000000003e-05,
+      "loss": 1.6483,
+      "step": 14
+    },
+    {
+      "epoch": 0.00675371454299865,
+      "grad_norm": 0.11006711423397064,
+      "learning_rate": 3e-05,
+      "loss": 0.96,
+      "step": 15
+    },
+    {
+      "epoch": 0.007203962179198559,
+      "grad_norm": 0.18399226665496826,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 0.9475,
+      "step": 16
+    },
+    {
+      "epoch": 0.00765420981539847,
+      "grad_norm": 0.1473955661058426,
+      "learning_rate": 3.4000000000000007e-05,
+      "loss": 1.2357,
+      "step": 17
+    },
+    {
+      "epoch": 0.00810445745159838,
+      "grad_norm": 0.16420996189117432,
+      "learning_rate": 3.6e-05,
+      "loss": 1.2338,
+      "step": 18
+    },
+    {
+      "epoch": 0.008554705087798289,
+      "grad_norm": 0.1456652283668518,
+      "learning_rate": 3.8e-05,
+      "loss": 1.2819,
+      "step": 19
+    },
+    {
+      "epoch": 0.009004952723998198,
+      "grad_norm": 0.12699273228645325,
+      "learning_rate": 4e-05,
+      "loss": 1.0816,
+      "step": 20
+    },
+    {
+      "epoch": 0.00945520036019811,
+      "grad_norm": 0.09541643410921097,
+      "learning_rate": 4.2e-05,
+      "loss": 0.6852,
+      "step": 21
+    },
+    {
+      "epoch": 0.00990544799639802,
+      "grad_norm": 0.10875753313302994,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 0.8732,
+      "step": 22
+    },
+    {
+      "epoch": 0.010355695632597929,
+      "grad_norm": 0.14051833748817444,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 1.2255,
+      "step": 23
+    },
+    {
+      "epoch": 0.010805943268797838,
+      "grad_norm": 0.10268246382474899,
+      "learning_rate": 4.8e-05,
+      "loss": 1.0155,
+      "step": 24
+    },
+    {
+      "epoch": 0.01125619090499775,
+      "grad_norm": 0.11052165925502777,
+      "learning_rate": 5e-05,
+      "loss": 1.0785,
+      "step": 25
+    },
+    {
+      "epoch": 0.01170643854119766,
+      "grad_norm": 0.0866745263338089,
+      "learning_rate": 5.2000000000000004e-05,
+      "loss": 0.9994,
+      "step": 26
+    },
+    {
+      "epoch": 0.012156686177397569,
+      "grad_norm": 0.1572750061750412,
+      "learning_rate": 5.4000000000000005e-05,
+      "loss": 1.2012,
+      "step": 27
+    },
+    {
+      "epoch": 0.012606933813597478,
+      "grad_norm": 0.12944263219833374,
+      "learning_rate": 5.6000000000000006e-05,
+      "loss": 1.1422,
+      "step": 28
+    },
+    {
+      "epoch": 0.013057181449797388,
+      "grad_norm": 0.14454616606235504,
+      "learning_rate": 5.8e-05,
+      "loss": 1.0463,
+      "step": 29
+    },
+    {
+      "epoch": 0.0135074290859973,
+      "grad_norm": 0.0934099555015564,
+      "learning_rate": 6e-05,
+      "loss": 1.0211,
+      "step": 30
+    },
+    {
+      "epoch": 0.013957676722197209,
+      "grad_norm": 0.18798087537288666,
+      "learning_rate": 6.2e-05,
+      "loss": 1.2729,
+      "step": 31
+    },
+    {
+      "epoch": 0.014407924358397118,
+      "grad_norm": 0.14255733788013458,
+      "learning_rate": 6.400000000000001e-05,
+      "loss": 0.722,
+      "step": 32
+    },
+    {
+      "epoch": 0.014858171994597028,
+      "grad_norm": 0.11062135547399521,
+      "learning_rate": 6.6e-05,
+      "loss": 0.8126,
+      "step": 33
+    },
+    {
+      "epoch": 0.01530841963079694,
+      "grad_norm": 0.21483464539051056,
+      "learning_rate": 6.800000000000001e-05,
+      "loss": 1.2268,
+      "step": 34
+    },
+    {
+      "epoch": 0.01575866726699685,
+      "grad_norm": 0.11179517209529877,
+      "learning_rate": 7e-05,
+      "loss": 0.9284,
+      "step": 35
+    },
+    {
+      "epoch": 0.01620891490319676,
+      "grad_norm": 0.15556995570659637,
+      "learning_rate": 7.2e-05,
+      "loss": 0.9832,
+      "step": 36
+    },
+    {
+      "epoch": 0.016659162539396668,
+      "grad_norm": 0.0957222655415535,
+      "learning_rate": 7.4e-05,
+      "loss": 0.6844,
+      "step": 37
+    },
+    {
+      "epoch": 0.017109410175596577,
+      "grad_norm": 0.08807898312807083,
+      "learning_rate": 7.6e-05,
+      "loss": 0.6882,
+      "step": 38
+    },
+    {
+      "epoch": 0.017559657811796487,
+      "grad_norm": 0.21804407238960266,
+      "learning_rate": 7.800000000000001e-05,
+      "loss": 1.2437,
+      "step": 39
+    },
+    {
+      "epoch": 0.018009905447996397,
+      "grad_norm": 0.06721039116382599,
+      "learning_rate": 8e-05,
+      "loss": 0.5842,
+      "step": 40
+    },
+    {
+      "epoch": 0.01846015308419631,
+      "grad_norm": 0.0984923243522644,
+      "learning_rate": 8.2e-05,
+      "loss": 0.9689,
+      "step": 41
+    },
+    {
+      "epoch": 0.01891040072039622,
+      "grad_norm": 0.13112875819206238,
+      "learning_rate": 8.4e-05,
+      "loss": 0.914,
+      "step": 42
+    },
+    {
+      "epoch": 0.01936064835659613,
+      "grad_norm": 0.11981655657291412,
+      "learning_rate": 8.6e-05,
+      "loss": 0.8496,
+      "step": 43
+    },
+    {
+      "epoch": 0.01981089599279604,
+      "grad_norm": 0.19388309121131897,
+      "learning_rate": 8.800000000000001e-05,
+      "loss": 1.1577,
+      "step": 44
+    },
+    {
+      "epoch": 0.020261143628995948,
+      "grad_norm": 0.17431718111038208,
+      "learning_rate": 9e-05,
+      "loss": 0.7394,
+      "step": 45
+    },
+    {
+      "epoch": 0.020711391265195857,
+      "grad_norm": 0.11217185854911804,
+      "learning_rate": 9.200000000000001e-05,
+      "loss": 0.8329,
+      "step": 46
+    },
+    {
+      "epoch": 0.021161638901395767,
+      "grad_norm": 0.10824249684810638,
+      "learning_rate": 9.4e-05,
+      "loss": 0.7787,
+      "step": 47
+    },
+    {
+      "epoch": 0.021611886537595677,
+      "grad_norm": 0.1337151825428009,
+      "learning_rate": 9.6e-05,
+      "loss": 0.9841,
+      "step": 48
+    },
+    {
+      "epoch": 0.022062134173795586,
+      "grad_norm": 0.1323201060295105,
+      "learning_rate": 9.8e-05,
+      "loss": 0.6849,
+      "step": 49
+    },
+    {
+      "epoch": 0.0225123818099955,
+      "grad_norm": 0.10257723182439804,
+      "learning_rate": 0.0001,
+      "loss": 0.61,
+      "step": 50
+    },
+    {
+      "epoch": 0.02296262944619541,
+      "grad_norm": 0.13418203592300415,
+      "learning_rate": 0.00010200000000000001,
+      "loss": 0.4748,
+      "step": 51
+    },
+    {
+      "epoch": 0.02341287708239532,
+      "grad_norm": 0.19270585477352142,
+      "learning_rate": 0.00010400000000000001,
+      "loss": 0.8814,
+      "step": 52
+    },
+    {
+      "epoch": 0.023863124718595228,
+      "grad_norm": 0.1021466851234436,
+      "learning_rate": 0.00010600000000000002,
+      "loss": 0.7988,
+      "step": 53
+    },
+    {
+      "epoch": 0.024313372354795137,
+      "grad_norm": 0.11183752864599228,
+      "learning_rate": 0.00010800000000000001,
+      "loss": 1.0283,
+      "step": 54
+    },
+    {
+      "epoch": 0.024763619990995047,
+      "grad_norm": 0.09874755144119263,
+      "learning_rate": 0.00011000000000000002,
+      "loss": 0.7967,
+      "step": 55
+    },
+    {
+      "epoch": 0.025213867627194957,
+      "grad_norm": 0.11574210971593857,
+      "learning_rate": 0.00011200000000000001,
+      "loss": 0.5071,
+      "step": 56
+    },
+    {
+      "epoch": 0.025664115263394866,
+      "grad_norm": 0.15173684060573578,
+      "learning_rate": 0.00011399999999999999,
+      "loss": 0.7338,
+      "step": 57
+    },
+    {
+      "epoch": 0.026114362899594776,
+      "grad_norm": 0.23482930660247803,
+      "learning_rate": 0.000116,
+      "loss": 1.1597,
+      "step": 58
+    },
+    {
+      "epoch": 0.02656461053579469,
+      "grad_norm": 0.10613696277141571,
+      "learning_rate": 0.000118,
+      "loss": 0.5098,
+      "step": 59
+    },
+    {
+      "epoch": 0.0270148581719946,
+      "grad_norm": 0.10165957361459732,
+      "learning_rate": 0.00012,
+      "loss": 0.672,
+      "step": 60
+    },
+    {
+      "epoch": 0.027465105808194508,
+      "grad_norm": 0.15795522928237915,
+      "learning_rate": 0.000122,
+      "loss": 0.7772,
+      "step": 61
+    },
+    {
+      "epoch": 0.027915353444394417,
+      "grad_norm": 0.13971729576587677,
+      "learning_rate": 0.000124,
+      "loss": 0.6143,
+      "step": 62
+    },
+    {
+      "epoch": 0.028365601080594327,
+      "grad_norm": 0.12256252765655518,
+      "learning_rate": 0.000126,
+      "loss": 0.657,
+      "step": 63
+    },
+    {
+      "epoch": 0.028815848716794237,
+      "grad_norm": 0.13188689947128296,
+      "learning_rate": 0.00012800000000000002,
+      "loss": 0.7105,
+      "step": 64
+    },
+    {
+      "epoch": 0.029266096352994146,
+      "grad_norm": 0.12793396413326263,
+      "learning_rate": 0.00013000000000000002,
+      "loss": 0.6478,
+      "step": 65
+    },
+    {
+      "epoch": 0.029716343989194056,
+      "grad_norm": 0.13094471395015717,
+      "learning_rate": 0.000132,
+      "loss": 0.8335,
+      "step": 66
+    },
+    {
+      "epoch": 0.030166591625393965,
+      "grad_norm": 0.31635135412216187,
+      "learning_rate": 0.000134,
+      "loss": 1.1427,
+      "step": 67
+    },
+    {
+      "epoch": 0.03061683926159388,
+      "grad_norm": 0.13020838797092438,
+      "learning_rate": 0.00013600000000000003,
+      "loss": 0.7015,
+      "step": 68
+    },
+    {
+      "epoch": 0.031067086897793788,
+      "grad_norm": 0.17031438648700714,
+      "learning_rate": 0.000138,
+      "loss": 1.1009,
+      "step": 69
+    },
+    {
+      "epoch": 0.0315173345339937,
+      "grad_norm": 0.2181975245475769,
+      "learning_rate": 0.00014,
+      "loss": 0.5874,
+      "step": 70
+    },
+    {
+      "epoch": 0.031967582170193604,
+      "grad_norm": 0.24470588564872742,
+      "learning_rate": 0.000142,
+      "loss": 0.8178,
+      "step": 71
+    },
+    {
+      "epoch": 0.03241782980639352,
+      "grad_norm": 0.13803082704544067,
+      "learning_rate": 0.000144,
+      "loss": 0.717,
+      "step": 72
+    },
+    {
+      "epoch": 0.03286807744259343,
+      "grad_norm": 0.12839782238006592,
+      "learning_rate": 0.000146,
+      "loss": 0.6458,
+      "step": 73
+    },
+    {
+      "epoch": 0.033318325078793336,
+      "grad_norm": 0.1400393396615982,
+      "learning_rate": 0.000148,
+      "loss": 0.5043,
+      "step": 74
+    },
+    {
+      "epoch": 0.03376857271499325,
+      "grad_norm": 0.11069265007972717,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 0.6656,
+      "step": 75
+    },
+    {
+      "epoch": 0.034218820351193155,
+      "grad_norm": 0.1982526183128357,
+      "learning_rate": 0.000152,
+      "loss": 0.5909,
+      "step": 76
+    },
+    {
+      "epoch": 0.03466906798739307,
+      "grad_norm": 0.14941537380218506,
+      "learning_rate": 0.000154,
+      "loss": 0.5221,
+      "step": 77
+    },
+    {
+      "epoch": 0.035119315623592974,
+      "grad_norm": 0.1304963082075119,
+      "learning_rate": 0.00015600000000000002,
+      "loss": 0.5228,
+      "step": 78
+    },
+    {
+      "epoch": 0.03556956325979289,
+      "grad_norm": 0.10047885775566101,
+      "learning_rate": 0.00015800000000000002,
+      "loss": 0.4865,
+      "step": 79
+    },
+    {
+      "epoch": 0.03601981089599279,
+      "grad_norm": 0.15201841294765472,
+      "learning_rate": 0.00016,
+      "loss": 0.6745,
+      "step": 80
+    },
+    {
+      "epoch": 0.036470058532192706,
+      "grad_norm": 0.1672833114862442,
+      "learning_rate": 0.000162,
+      "loss": 0.6863,
+      "step": 81
+    },
+    {
+      "epoch": 0.03692030616839262,
+      "grad_norm": 0.21495819091796875,
+      "learning_rate": 0.000164,
+      "loss": 0.9159,
+      "step": 82
+    },
+    {
+      "epoch": 0.037370553804592525,
+      "grad_norm": 0.1072831079363823,
+      "learning_rate": 0.000166,
+      "loss": 0.5166,
+      "step": 83
+    },
+    {
+      "epoch": 0.03782080144079244,
+      "grad_norm": 0.1589314341545105,
+      "learning_rate": 0.000168,
+      "loss": 1.0129,
+      "step": 84
+    },
+    {
+      "epoch": 0.038271049076992344,
+      "grad_norm": 0.09406781196594238,
+      "learning_rate": 0.00017,
+      "loss": 0.5907,
+      "step": 85
+    },
+    {
+      "epoch": 0.03872129671319226,
+      "grad_norm": 0.23734940588474274,
+      "learning_rate": 0.000172,
+      "loss": 0.9875,
+      "step": 86
+    },
+    {
+      "epoch": 0.039171544349392164,
+      "grad_norm": 0.114694744348526,
+      "learning_rate": 0.000174,
+      "loss": 0.4937,
+      "step": 87
+    },
+    {
+      "epoch": 0.03962179198559208,
+      "grad_norm": 0.2216087430715561,
+      "learning_rate": 0.00017600000000000002,
+      "loss": 0.5199,
+      "step": 88
+    },
+    {
+      "epoch": 0.04007203962179198,
+      "grad_norm": 0.14134696125984192,
+      "learning_rate": 0.00017800000000000002,
+      "loss": 0.5823,
+      "step": 89
+    },
+    {
+      "epoch": 0.040522287257991896,
+      "grad_norm": 0.1532876044511795,
+      "learning_rate": 0.00018,
+      "loss": 0.4445,
+      "step": 90
+    },
+    {
+      "epoch": 0.04097253489419181,
+      "grad_norm": 0.13293257355690002,
+      "learning_rate": 0.000182,
+      "loss": 0.7258,
+      "step": 91
+    },
+    {
+      "epoch": 0.041422782530391715,
+      "grad_norm": 0.127973273396492,
+      "learning_rate": 0.00018400000000000003,
+      "loss": 0.5618,
+      "step": 92
+    },
+    {
+      "epoch": 0.04187303016659163,
+      "grad_norm": 0.10355206578969955,
+      "learning_rate": 0.00018600000000000002,
+      "loss": 0.5749,
+      "step": 93
+    },
+    {
+      "epoch": 0.042323277802791534,
+      "grad_norm": 0.16975559294223785,
+      "learning_rate": 0.000188,
+      "loss": 0.59,
+      "step": 94
+    },
+    {
+      "epoch": 0.04277352543899145,
+      "grad_norm": 0.156317338347435,
+      "learning_rate": 0.00019,
+      "loss": 0.756,
+      "step": 95
+    },
+    {
+      "epoch": 0.04322377307519135,
+      "grad_norm": 0.16262571513652802,
+      "learning_rate": 0.000192,
+      "loss": 0.7395,
+      "step": 96
+    },
+    {
+      "epoch": 0.043674020711391266,
+      "grad_norm": 0.18147265911102295,
+      "learning_rate": 0.000194,
+      "loss": 0.5907,
+      "step": 97
+    },
+    {
+      "epoch": 0.04412426834759117,
+      "grad_norm": 0.13861382007598877,
+      "learning_rate": 0.000196,
+      "loss": 0.5812,
+      "step": 98
+    },
+    {
+      "epoch": 0.044574515983791085,
+      "grad_norm": 0.1534164696931839,
+      "learning_rate": 0.00019800000000000002,
+      "loss": 0.6238,
+      "step": 99
+    },
+    {
+      "epoch": 0.045024763619991,
+      "grad_norm": 0.19617322087287903,
+      "learning_rate": 0.0002,
+      "loss": 0.6167,
+      "step": 100
+    },
+    {
+      "epoch": 0.045475011256190905,
+      "grad_norm": 0.14344817399978638,
+      "learning_rate": 0.00019999989030462093,
+      "loss": 0.6478,
+      "step": 101
+    },
+    {
+      "epoch": 0.04592525889239082,
+      "grad_norm": 0.13441966474056244,
+      "learning_rate": 0.00019999956121872435,
+      "loss": 0.8529,
+      "step": 102
+    },
+    {
+      "epoch": 0.046375506528590724,
+      "grad_norm": 0.15605680644512177,
+      "learning_rate": 0.00019999901274303226,
+      "loss": 0.5875,
+      "step": 103
+    },
+    {
+      "epoch": 0.04682575416479064,
+      "grad_norm": 0.15065008401870728,
+      "learning_rate": 0.00019999824487874798,
+      "loss": 0.7964,
+      "step": 104
+    },
+    {
+      "epoch": 0.04727600180099054,
+      "grad_norm": 0.1270512342453003,
+      "learning_rate": 0.0001999972576275561,
+      "loss": 0.7603,
+      "step": 105
+    },
+    {
+      "epoch": 0.047726249437190456,
+      "grad_norm": 0.11631572991609573,
+      "learning_rate": 0.0001999960509916226,
+      "loss": 0.6237,
+      "step": 106
+    },
+    {
+      "epoch": 0.04817649707339036,
+      "grad_norm": 0.12819796800613403,
+      "learning_rate": 0.00019999462497359466,
+      "loss": 0.8404,
+      "step": 107
+    },
+    {
+      "epoch": 0.048626744709590275,
+      "grad_norm": 0.3845088481903076,
+      "learning_rate": 0.0001999929795766009,
+      "loss": 0.8845,
+      "step": 108
+    },
+    {
+      "epoch": 0.04907699234579019,
+      "grad_norm": 0.19142518937587738,
+      "learning_rate": 0.00019999111480425112,
+      "loss": 0.8079,
+      "step": 109
+    },
+    {
+      "epoch": 0.049527239981990094,
+      "grad_norm": 0.14023594558238983,
+      "learning_rate": 0.0001999890306606365,
+      "loss": 0.4691,
+      "step": 110
+    },
+    {
+      "epoch": 0.04997748761819001,
+      "grad_norm": 0.21261820197105408,
+      "learning_rate": 0.00019998672715032946,
+      "loss": 1.233,
+      "step": 111
+    },
+    {
+      "epoch": 0.05042773525438991,
+      "grad_norm": 0.16879242658615112,
+      "learning_rate": 0.00019998420427838363,
+      "loss": 0.7716,
+      "step": 112
+    },
+    {
+      "epoch": 0.050877982890589826,
+      "grad_norm": 0.2122250497341156,
+      "learning_rate": 0.000199981462050334,
+      "loss": 0.8041,
+      "step": 113
+    },
+    {
+      "epoch": 0.05132823052678973,
+      "grad_norm": 0.20105648040771484,
+      "learning_rate": 0.0001999785004721968,
+      "loss": 0.7503,
+      "step": 114
+    },
+    {
+      "epoch": 0.051778478162989645,
+      "grad_norm": 0.13634870946407318,
+      "learning_rate": 0.0001999753195504694,
+      "loss": 0.6349,
+      "step": 115
+    },
+    {
+      "epoch": 0.05222872579918955,
+      "grad_norm": 0.18042653799057007,
+      "learning_rate": 0.00019997191929213047,
+      "loss": 0.5866,
+      "step": 116
+    },
+    {
+      "epoch": 0.052678973435389465,
+      "grad_norm": 0.17580613493919373,
+      "learning_rate": 0.00019996829970463984,
+      "loss": 0.6823,
+      "step": 117
+    },
+    {
+      "epoch": 0.05312922107158938,
+      "grad_norm": 0.2443765252828598,
+      "learning_rate": 0.00019996446079593855,
+      "loss": 1.1471,
+      "step": 118
+    },
+    {
+      "epoch": 0.053579468707789284,
+      "grad_norm": 0.1729765236377716,
+      "learning_rate": 0.0001999604025744489,
+      "loss": 0.5781,
+      "step": 119
+    },
+    {
+      "epoch": 0.0540297163439892,
+      "grad_norm": 0.10787572711706161,
+      "learning_rate": 0.00019995612504907414,
+      "loss": 0.4322,
+      "step": 120
+    },
+    {
+      "epoch": 0.0544799639801891,
+      "grad_norm": 0.1601431667804718,
+      "learning_rate": 0.00019995162822919883,
+      "loss": 0.7133,
+      "step": 121
+    },
+    {
+      "epoch": 0.054930211616389016,
+      "grad_norm": 0.12010475248098373,
+      "learning_rate": 0.00019994691212468855,
+      "loss": 0.5991,
+      "step": 122
+    },
+    {
+      "epoch": 0.05538045925258892,
+      "grad_norm": 0.19738584756851196,
+      "learning_rate": 0.00019994197674588998,
+      "loss": 0.6837,
+      "step": 123
+    },
+    {
+      "epoch": 0.055830706888788835,
+      "grad_norm": 0.19277459383010864,
+      "learning_rate": 0.00019993682210363094,
+      "loss": 0.8575,
+      "step": 124
+    },
+    {
+      "epoch": 0.05628095452498874,
+      "grad_norm": 0.14095377922058105,
+      "learning_rate": 0.00019993144820922018,
+      "loss": 0.6536,
+      "step": 125
+    },
+    {
+      "epoch": 0.056731202161188654,
+      "grad_norm": 0.20605207979679108,
+      "learning_rate": 0.0001999258550744476,
+      "loss": 0.7939,
+      "step": 126
+    },
+    {
+      "epoch": 0.05718144979738857,
+      "grad_norm": 0.15481193363666534,
+      "learning_rate": 0.00019992004271158397,
+      "loss": 0.6178,
+      "step": 127
+    },
+    {
+      "epoch": 0.05763169743358847,
+      "grad_norm": 0.17113305628299713,
+      "learning_rate": 0.00019991401113338104,
+      "loss": 0.8284,
+      "step": 128
+    },
+    {
+      "epoch": 0.058081945069788386,
+      "grad_norm": 0.21664205193519592,
+      "learning_rate": 0.0001999077603530716,
+      "loss": 0.7236,
+      "step": 129
+    },
+    {
+      "epoch": 0.05853219270598829,
+      "grad_norm": 0.12617772817611694,
+      "learning_rate": 0.0001999012903843693,
+      "loss": 0.4423,
+      "step": 130
+    },
+    {
+      "epoch": 0.058982440342188205,
+      "grad_norm": 0.11639740318059921,
+      "learning_rate": 0.00019989460124146858,
+      "loss": 0.5195,
+      "step": 131
+    },
+    {
+      "epoch": 0.05943268797838811,
+      "grad_norm": 0.19009053707122803,
+      "learning_rate": 0.00019988769293904484,
+      "loss": 0.7141,
+      "step": 132
+    },
+    {
+      "epoch": 0.059882935614588025,
+      "grad_norm": 0.22553405165672302,
+      "learning_rate": 0.00019988056549225428,
+      "loss": 0.584,
+      "step": 133
+    },
+    {
+      "epoch": 0.06033318325078793,
+      "grad_norm": 0.1268945187330246,
+      "learning_rate": 0.0001998732189167338,
+      "loss": 0.7948,
+      "step": 134
+    },
+    {
+      "epoch": 0.060783430886987844,
+      "grad_norm": 0.22973321378231049,
+      "learning_rate": 0.00019986565322860115,
+      "loss": 0.4834,
+      "step": 135
+    },
+    {
+      "epoch": 0.06123367852318776,
+      "grad_norm": 0.11541842669248581,
+      "learning_rate": 0.00019985786844445476,
+      "loss": 0.6104,
+      "step": 136
+    },
+    {
+      "epoch": 0.06168392615938766,
+      "grad_norm": 0.19445981085300446,
+      "learning_rate": 0.0001998498645813737,
+      "loss": 0.8488,
+      "step": 137
+    },
+    {
+      "epoch": 0.062134173795587576,
+      "grad_norm": 0.14637619256973267,
+      "learning_rate": 0.00019984164165691773,
+      "loss": 0.4229,
+      "step": 138
+    },
+    {
+      "epoch": 0.06258442143178748,
+      "grad_norm": 0.13113552331924438,
+      "learning_rate": 0.00019983319968912714,
+      "loss": 0.4772,
+      "step": 139
+    },
+    {
+      "epoch": 0.0630346690679874,
+      "grad_norm": 0.15641582012176514,
+      "learning_rate": 0.00019982453869652289,
+      "loss": 0.8693,
+      "step": 140
+    },
+    {
+      "epoch": 0.06348491670418731,
+      "grad_norm": 0.14171209931373596,
+      "learning_rate": 0.00019981565869810636,
+      "loss": 0.5126,
+      "step": 141
+    },
+    {
+      "epoch": 0.06393516434038721,
+      "grad_norm": 0.243012934923172,
+      "learning_rate": 0.00019980655971335945,
+      "loss": 0.754,
+      "step": 142
+    },
+    {
+      "epoch": 0.06438541197658712,
+      "grad_norm": 0.12468169629573822,
+      "learning_rate": 0.0001997972417622445,
+      "loss": 0.6749,
+      "step": 143
+    },
+    {
+      "epoch": 0.06483565961278703,
+      "grad_norm": 0.1856045126914978,
+      "learning_rate": 0.0001997877048652042,
+      "loss": 0.7246,
+      "step": 144
+    },
+    {
+      "epoch": 0.06528590724898695,
+      "grad_norm": 0.11403091996908188,
+      "learning_rate": 0.00019977794904316168,
+      "loss": 0.4308,
+      "step": 145
+    },
+    {
+      "epoch": 0.06573615488518686,
+      "grad_norm": 0.11661963164806366,
+      "learning_rate": 0.00019976797431752025,
+      "loss": 0.8359,
+      "step": 146
+    },
+    {
+      "epoch": 0.06618640252138676,
+      "grad_norm": 0.1257256120443344,
+      "learning_rate": 0.0001997577807101636,
+      "loss": 0.4442,
+      "step": 147
+    },
+    {
+      "epoch": 0.06663665015758667,
+      "grad_norm": 0.1263304203748703,
+      "learning_rate": 0.00019974736824345551,
+      "loss": 0.8348,
+      "step": 148
+    },
+    {
+      "epoch": 0.06708689779378658,
+      "grad_norm": 0.13273175060749054,
+      "learning_rate": 0.00019973673694024,
+      "loss": 0.6628,
+      "step": 149
+    },
+    {
+      "epoch": 0.0675371454299865,
+      "grad_norm": 0.15919096767902374,
+      "learning_rate": 0.00019972588682384116,
+      "loss": 0.5862,
+      "step": 150
+    },
+    {
+      "epoch": 0.0679873930661864,
+      "grad_norm": 0.16459324955940247,
+      "learning_rate": 0.00019971481791806312,
+      "loss": 0.3376,
+      "step": 151
+    },
+    {
+      "epoch": 0.06843764070238631,
+      "grad_norm": 0.1261291652917862,
+      "learning_rate": 0.00019970353024719006,
+      "loss": 0.5961,
+      "step": 152
+    },
+    {
+      "epoch": 0.06888788833858622,
+      "grad_norm": 0.1355976164340973,
+      "learning_rate": 0.0001996920238359861,
+      "loss": 0.6367,
+      "step": 153
+    },
+    {
+      "epoch": 0.06933813597478614,
+      "grad_norm": 0.16610701382160187,
+      "learning_rate": 0.0001996802987096952,
+      "loss": 0.7141,
+      "step": 154
+    },
+    {
+      "epoch": 0.06978838361098605,
+      "grad_norm": 0.16071228682994843,
+      "learning_rate": 0.00019966835489404122,
+      "loss": 0.5332,
+      "step": 155
+    },
+    {
+      "epoch": 0.07023863124718595,
+      "grad_norm": 0.09267164766788483,
+      "learning_rate": 0.0001996561924152278,
+      "loss": 0.434,
+      "step": 156
+    },
+    {
+      "epoch": 0.07068887888338586,
+      "grad_norm": 0.14471803605556488,
+      "learning_rate": 0.0001996438112999383,
+      "loss": 0.8514,
+      "step": 157
+    },
+    {
+      "epoch": 0.07113912651958577,
+      "grad_norm": 0.10718607902526855,
+      "learning_rate": 0.00019963121157533576,
+      "loss": 0.707,
+      "step": 158
+    },
+    {
+      "epoch": 0.07158937415578569,
+      "grad_norm": 0.1329723447561264,
+      "learning_rate": 0.00019961839326906273,
+      "loss": 0.5713,
+      "step": 159
+    },
+    {
+      "epoch": 0.07203962179198559,
+      "grad_norm": 0.14699022471904755,
+      "learning_rate": 0.0001996053564092415,
+      "loss": 0.9221,
+      "step": 160
+    },
+    {
+      "epoch": 0.0724898694281855,
+      "grad_norm": 0.14206327497959137,
+      "learning_rate": 0.00019959210102447363,
+      "loss": 0.6067,
+      "step": 161
+    },
+    {
+      "epoch": 0.07294011706438541,
+      "grad_norm": 0.14892014861106873,
+      "learning_rate": 0.00019957862714384028,
+      "loss": 0.5244,
+      "step": 162
+    },
+    {
+      "epoch": 0.07339036470058533,
+      "grad_norm": 0.13372868299484253,
+      "learning_rate": 0.0001995649347969019,
+      "loss": 0.6268,
+      "step": 163
+    },
+    {
+      "epoch": 0.07384061233678524,
+      "grad_norm": 0.11254440993070602,
+      "learning_rate": 0.00019955102401369814,
+      "loss": 0.3449,
+      "step": 164
+    },
+    {
+      "epoch": 0.07429085997298514,
+      "grad_norm": 0.1538407951593399,
+      "learning_rate": 0.0001995368948247481,
+      "loss": 0.6215,
+      "step": 165
+    },
+    {
+      "epoch": 0.07474110760918505,
+      "grad_norm": 0.12231205403804779,
+      "learning_rate": 0.00019952254726104984,
+      "loss": 0.6257,
+      "step": 166
+    },
+    {
+      "epoch": 0.07519135524538496,
+      "grad_norm": 0.2967234253883362,
+      "learning_rate": 0.0001995079813540806,
+      "loss": 0.801,
+      "step": 167
+    },
+    {
+      "epoch": 0.07564160288158488,
+      "grad_norm": 0.16710905730724335,
+      "learning_rate": 0.00019949319713579664,
+      "loss": 0.5782,
+      "step": 168
+    },
+    {
+      "epoch": 0.07609185051778478,
+      "grad_norm": 0.10462953895330429,
+      "learning_rate": 0.0001994781946386332,
+      "loss": 0.5159,
+      "step": 169
+    },
+    {
+      "epoch": 0.07654209815398469,
+      "grad_norm": 0.11842755228281021,
+      "learning_rate": 0.00019946297389550433,
+      "loss": 0.5246,
+      "step": 170
+    },
+    {
+      "epoch": 0.0769923457901846,
+      "grad_norm": 0.1503244787454605,
+      "learning_rate": 0.00019944753493980293,
+      "loss": 0.5239,
+      "step": 171
+    },
+    {
+      "epoch": 0.07744259342638452,
+      "grad_norm": 0.11417451500892639,
+      "learning_rate": 0.00019943187780540066,
+      "loss": 0.3971,
+      "step": 172
+    },
+    {
+      "epoch": 0.07789284106258443,
+      "grad_norm": 0.11363469809293747,
+      "learning_rate": 0.00019941600252664784,
+      "loss": 0.6202,
+      "step": 173
+    },
+    {
+      "epoch": 0.07834308869878433,
+      "grad_norm": 0.19468806684017181,
+      "learning_rate": 0.0001993999091383733,
+      "loss": 1.0165,
+      "step": 174
+    },
+    {
+      "epoch": 0.07879333633498424,
+      "grad_norm": 0.1858697235584259,
+      "learning_rate": 0.00019938359767588453,
+      "loss": 0.5116,
+      "step": 175
+    },
+    {
+      "epoch": 0.07924358397118415,
+      "grad_norm": 0.17502325773239136,
+      "learning_rate": 0.00019936706817496734,
+      "loss": 0.864,
+      "step": 176
+    },
+    {
+      "epoch": 0.07969383160738407,
+      "grad_norm": 0.11531347781419754,
+      "learning_rate": 0.0001993503206718859,
+      "loss": 0.498,
+      "step": 177
+    },
+    {
+      "epoch": 0.08014407924358397,
+      "grad_norm": 0.16447703540325165,
+      "learning_rate": 0.0001993333552033827,
+      "loss": 0.701,
+      "step": 178
+    },
+    {
+      "epoch": 0.08059432687978388,
+      "grad_norm": 0.14537443220615387,
+      "learning_rate": 0.00019931617180667845,
+      "loss": 0.7476,
+      "step": 179
+    },
+    {
+      "epoch": 0.08104457451598379,
+      "grad_norm": 0.11203992366790771,
+      "learning_rate": 0.0001992987705194719,
+      "loss": 0.693,
+      "step": 180
+    },
+    {
+      "epoch": 0.0814948221521837,
+      "grad_norm": 0.1476212739944458,
+      "learning_rate": 0.00019928115137993982,
+      "loss": 0.5739,
+      "step": 181
+    },
+    {
+      "epoch": 0.08194506978838362,
+      "grad_norm": 0.13168933987617493,
+      "learning_rate": 0.00019926331442673707,
+      "loss": 0.8419,
+      "step": 182
+    },
+    {
+      "epoch": 0.08239531742458352,
+      "grad_norm": 0.18853585422039032,
+      "learning_rate": 0.00019924525969899622,
+      "loss": 0.65,
+      "step": 183
+    },
+    {
+      "epoch": 0.08284556506078343,
+      "grad_norm": 0.48106467723846436,
+      "learning_rate": 0.00019922698723632767,
+      "loss": 0.9028,
+      "step": 184
+    },
+    {
+      "epoch": 0.08329581269698334,
+      "grad_norm": 0.14343899488449097,
+      "learning_rate": 0.0001992084970788195,
+      "loss": 0.6108,
+      "step": 185
+    },
+    {
+      "epoch": 0.08374606033318326,
+      "grad_norm": 0.09865623712539673,
+      "learning_rate": 0.00019918978926703748,
+      "loss": 0.4703,
+      "step": 186
+    },
+    {
+      "epoch": 0.08419630796938316,
+      "grad_norm": 0.13469426333904266,
+      "learning_rate": 0.00019917086384202475,
+      "loss": 0.7372,
+      "step": 187
+    },
+    {
+      "epoch": 0.08464655560558307,
+      "grad_norm": 0.10762239992618561,
+      "learning_rate": 0.00019915172084530197,
+      "loss": 0.5506,
+      "step": 188
+    },
+    {
+      "epoch": 0.08509680324178298,
+      "grad_norm": 0.17493847012519836,
+      "learning_rate": 0.0001991323603188671,
+      "loss": 0.5832,
+      "step": 189
+    },
+    {
+      "epoch": 0.0855470508779829,
+      "grad_norm": 0.1554277092218399,
+      "learning_rate": 0.00019911278230519536,
+      "loss": 0.5535,
+      "step": 190
+    },
+    {
+      "epoch": 0.08599729851418281,
+      "grad_norm": 0.13917969167232513,
+      "learning_rate": 0.00019909298684723904,
+      "loss": 0.6591,
+      "step": 191
+    },
+    {
+      "epoch": 0.0864475461503827,
+      "grad_norm": 0.16856342554092407,
+      "learning_rate": 0.00019907297398842765,
+      "loss": 1.1022,
+      "step": 192
+    },
+    {
+      "epoch": 0.08689779378658262,
+      "grad_norm": 0.13101813197135925,
+      "learning_rate": 0.00019905274377266747,
+      "loss": 0.7233,
+      "step": 193
+    },
+    {
+      "epoch": 0.08734804142278253,
+      "grad_norm": 0.16108326613903046,
+      "learning_rate": 0.00019903229624434178,
+      "loss": 0.5351,
+      "step": 194
+    },
+    {
+      "epoch": 0.08779828905898245,
+      "grad_norm": 0.18426254391670227,
+      "learning_rate": 0.0001990116314483105,
+      "loss": 0.8561,
+      "step": 195
+    },
+    {
+      "epoch": 0.08824853669518234,
+      "grad_norm": 0.13305288553237915,
+      "learning_rate": 0.00019899074942991035,
+      "loss": 0.7561,
+      "step": 196
+    },
+    {
+      "epoch": 0.08869878433138226,
+      "grad_norm": 0.1580049991607666,
+      "learning_rate": 0.00019896965023495452,
+      "loss": 0.8103,
+      "step": 197
+    },
+    {
+      "epoch": 0.08914903196758217,
+      "grad_norm": 0.16388972103595734,
+      "learning_rate": 0.00019894833390973266,
+      "loss": 0.7726,
+      "step": 198
+    },
+    {
+      "epoch": 0.08959927960378208,
+      "grad_norm": 0.14917367696762085,
+      "learning_rate": 0.0001989268005010109,
+      "loss": 0.7033,
+      "step": 199
+    },
+    {
+      "epoch": 0.090049527239982,
+      "grad_norm": 0.1342577487230301,
+      "learning_rate": 0.00019890505005603146,
+      "loss": 0.8173,
+      "step": 200
+    },
+    {
+      "epoch": 0.0904997748761819,
+      "grad_norm": 0.182293102145195,
+      "learning_rate": 0.00019888308262251285,
+      "loss": 0.4909,
+      "step": 201
+    },
+    {
+      "epoch": 0.09095002251238181,
+      "grad_norm": 0.18417170643806458,
+      "learning_rate": 0.0001988608982486496,
+      "loss": 0.7985,
+      "step": 202
+    },
+    {
+      "epoch": 0.09140027014858172,
+      "grad_norm": 0.20273232460021973,
+      "learning_rate": 0.00019883849698311217,
+      "loss": 0.8457,
+      "step": 203
+    },
+    {
+      "epoch": 0.09185051778478164,
+      "grad_norm": 0.14214643836021423,
+      "learning_rate": 0.00019881587887504685,
+      "loss": 0.4491,
+      "step": 204
+    },
+    {
+      "epoch": 0.09230076542098153,
+      "grad_norm": 0.16013310849666595,
+      "learning_rate": 0.0001987930439740757,
+      "loss": 0.5407,
+      "step": 205
+    },
+    {
+      "epoch": 0.09275101305718145,
+      "grad_norm": 0.09869419783353806,
+      "learning_rate": 0.0001987699923302963,
+      "loss": 0.446,
+      "step": 206
+    },
+    {
+      "epoch": 0.09320126069338136,
+      "grad_norm": 0.19170308113098145,
+      "learning_rate": 0.00019874672399428198,
+      "loss": 0.6148,
+      "step": 207
+    },
+    {
+      "epoch": 0.09365150832958127,
+      "grad_norm": 0.13894422352313995,
+      "learning_rate": 0.00019872323901708118,
+      "loss": 0.522,
+      "step": 208
+    },
+    {
+      "epoch": 0.09410175596578119,
+      "grad_norm": 0.11728322505950928,
+      "learning_rate": 0.00019869953745021786,
+      "loss": 0.6586,
+      "step": 209
+    },
+    {
+      "epoch": 0.09455200360198109,
+      "grad_norm": 0.136150062084198,
+      "learning_rate": 0.00019867561934569104,
+      "loss": 0.5717,
+      "step": 210
+    },
+    {
+      "epoch": 0.095002251238181,
+      "grad_norm": 0.12023138999938965,
+      "learning_rate": 0.00019865148475597477,
+      "loss": 0.5623,
+      "step": 211
+    },
+    {
+      "epoch": 0.09545249887438091,
+      "grad_norm": 0.1110725998878479,
+      "learning_rate": 0.0001986271337340182,
+      "loss": 0.7082,
+      "step": 212
+    },
+    {
+      "epoch": 0.09590274651058082,
+      "grad_norm": 0.1555807739496231,
+      "learning_rate": 0.00019860256633324515,
+      "loss": 0.5501,
+      "step": 213
+    },
+    {
+      "epoch": 0.09635299414678072,
+      "grad_norm": 0.12648093700408936,
+      "learning_rate": 0.00019857778260755428,
+      "loss": 0.5439,
+      "step": 214
+    },
+    {
+      "epoch": 0.09680324178298064,
+      "grad_norm": 0.2394242286682129,
+      "learning_rate": 0.00019855278261131875,
+      "loss": 0.6875,
+      "step": 215
+    },
+    {
+      "epoch": 0.09725348941918055,
+      "grad_norm": 0.10676310956478119,
+      "learning_rate": 0.0001985275663993863,
+      "loss": 0.612,
+      "step": 216
+    },
+    {
+      "epoch": 0.09770373705538046,
+      "grad_norm": 0.1496703326702118,
+      "learning_rate": 0.0001985021340270789,
+      "loss": 0.4285,
+      "step": 217
+    },
+    {
+      "epoch": 0.09815398469158038,
+      "grad_norm": 0.2819899618625641,
+      "learning_rate": 0.00019847648555019288,
+      "loss": 0.6139,
+      "step": 218
+    },
+    {
+      "epoch": 0.09860423232778028,
+      "grad_norm": 0.09466250985860825,
+      "learning_rate": 0.0001984506210249986,
+      "loss": 0.4133,
+      "step": 219
+    },
+    {
+      "epoch": 0.09905447996398019,
+      "grad_norm": 0.12264137715101242,
+      "learning_rate": 0.00019842454050824043,
+      "loss": 0.4683,
+      "step": 220
+    },
+    {
+      "epoch": 0.0995047276001801,
+      "grad_norm": 0.13499128818511963,
+      "learning_rate": 0.00019839824405713666,
+      "loss": 0.6537,
+      "step": 221
+    },
+    {
+      "epoch": 0.09995497523638001,
+      "grad_norm": 0.13991668820381165,
+      "learning_rate": 0.0001983717317293792,
+      "loss": 0.4452,
+      "step": 222
+    },
+    {
+      "epoch": 0.10040522287257991,
+      "grad_norm": 0.2020612359046936,
+      "learning_rate": 0.0001983450035831337,
+      "loss": 0.9447,
+      "step": 223
+    },
+    {
+      "epoch": 0.10085547050877983,
+      "grad_norm": 0.1975604146718979,
+      "learning_rate": 0.00019831805967703922,
+      "loss": 0.6073,
+      "step": 224
+    },
+    {
+      "epoch": 0.10130571814497974,
+      "grad_norm": 0.16907307505607605,
+      "learning_rate": 0.00019829090007020823,
+      "loss": 0.6806,
+      "step": 225
+    },
+    {
+      "epoch": 0.10175596578117965,
+      "grad_norm": 0.18565411865711212,
+      "learning_rate": 0.00019826352482222638,
+      "loss": 0.934,
+      "step": 226
+    },
+    {
+      "epoch": 0.10220621341737957,
+      "grad_norm": 0.14842809736728668,
+      "learning_rate": 0.0001982359339931524,
+      "loss": 0.806,
+      "step": 227
+    },
+    {
+      "epoch": 0.10265646105357946,
+      "grad_norm": 0.17171354591846466,
+      "learning_rate": 0.00019820812764351808,
+      "loss": 0.8474,
+      "step": 228
+    },
+    {
+      "epoch": 0.10310670868977938,
+      "grad_norm": 0.12110444903373718,
+      "learning_rate": 0.00019818010583432796,
+      "loss": 0.7888,
+      "step": 229
+    },
+    {
+      "epoch": 0.10355695632597929,
+      "grad_norm": 0.1511252224445343,
+      "learning_rate": 0.00019815186862705923,
+      "loss": 0.5006,
+      "step": 230
+    },
+    {
+      "epoch": 0.1040072039621792,
+      "grad_norm": 0.19050273299217224,
+      "learning_rate": 0.00019812341608366183,
+      "loss": 0.6127,
+      "step": 231
+    },
+    {
+      "epoch": 0.1044574515983791,
+      "grad_norm": 0.2948392927646637,
+      "learning_rate": 0.00019809474826655793,
+      "loss": 0.8423,
+      "step": 232
+    },
+    {
+      "epoch": 0.10490769923457902,
+      "grad_norm": 0.1426222175359726,
+      "learning_rate": 0.0001980658652386421,
+      "loss": 0.4675,
+      "step": 233
+    },
+    {
+      "epoch": 0.10535794687077893,
+      "grad_norm": 0.14354975521564484,
+      "learning_rate": 0.00019803676706328104,
+      "loss": 0.4095,
+      "step": 234
+    },
+    {
+      "epoch": 0.10580819450697884,
+      "grad_norm": 0.13434763252735138,
+      "learning_rate": 0.00019800745380431343,
+      "loss": 0.4253,
+      "step": 235
+    },
+    {
+      "epoch": 0.10625844214317876,
+      "grad_norm": 0.16992104053497314,
+      "learning_rate": 0.00019797792552604988,
+      "loss": 0.6168,
+      "step": 236
+    },
+    {
+      "epoch": 0.10670868977937865,
+      "grad_norm": 0.17915000021457672,
+      "learning_rate": 0.0001979481822932727,
+      "loss": 0.5936,
+      "step": 237
+    },
+    {
+      "epoch": 0.10715893741557857,
+      "grad_norm": 0.15283551812171936,
+      "learning_rate": 0.00019791822417123574,
+      "loss": 0.5042,
+      "step": 238
+    },
+    {
+      "epoch": 0.10760918505177848,
+      "grad_norm": 0.20631375908851624,
+      "learning_rate": 0.00019788805122566446,
+      "loss": 0.8967,
+      "step": 239
+    },
+    {
+      "epoch": 0.1080594326879784,
+      "grad_norm": 0.16916035115718842,
+      "learning_rate": 0.00019785766352275542,
+      "loss": 0.5099,
+      "step": 240
+    },
+    {
+      "epoch": 0.10850968032417829,
+      "grad_norm": 0.16004551947116852,
+      "learning_rate": 0.00019782706112917646,
+      "loss": 0.7917,
+      "step": 241
+    },
+    {
+      "epoch": 0.1089599279603782,
+      "grad_norm": 0.1461247056722641,
+      "learning_rate": 0.00019779624411206642,
+      "loss": 0.5996,
+      "step": 242
+    },
+    {
+      "epoch": 0.10941017559657812,
+      "grad_norm": 0.11229803413152695,
+      "learning_rate": 0.00019776521253903496,
+      "loss": 0.4637,
+      "step": 243
+    },
+    {
+      "epoch": 0.10986042323277803,
+      "grad_norm": 0.12437206506729126,
+      "learning_rate": 0.0001977339664781625,
+      "loss": 0.7709,
+      "step": 244
+    },
+    {
+      "epoch": 0.11031067086897794,
+      "grad_norm": 0.16254368424415588,
+      "learning_rate": 0.000197702505998,
+      "loss": 0.5494,
+      "step": 245
+    },
+    {
+      "epoch": 0.11076091850517784,
+      "grad_norm": 0.1651155799627304,
+      "learning_rate": 0.00019767083116756885,
+      "loss": 0.6674,
+      "step": 246
+    },
+    {
+      "epoch": 0.11121116614137776,
+      "grad_norm": 0.13120125234127045,
+      "learning_rate": 0.00019763894205636072,
+      "loss": 0.7428,
+      "step": 247
+    },
+    {
+      "epoch": 0.11166141377757767,
+      "grad_norm": 0.12071948498487473,
+      "learning_rate": 0.00019760683873433735,
+      "loss": 0.4826,
+      "step": 248
+    },
+    {
+      "epoch": 0.11211166141377758,
+      "grad_norm": 0.12815825641155243,
+      "learning_rate": 0.00019757452127193046,
+      "loss": 0.4422,
+      "step": 249
+    },
+    {
+      "epoch": 0.11256190904997748,
+      "grad_norm": 0.09186654537916183,
+      "learning_rate": 0.00019754198974004158,
+      "loss": 0.5586,
+      "step": 250
+    },
+    {
+      "epoch": 0.1130121566861774,
+      "grad_norm": 0.11398129910230637,
+      "learning_rate": 0.0001975092442100419,
+      "loss": 0.3335,
+      "step": 251
+    },
+    {
+      "epoch": 0.11346240432237731,
+      "grad_norm": 0.174946129322052,
+      "learning_rate": 0.00019747628475377208,
+      "loss": 0.5802,
+      "step": 252
+    },
+    {
+      "epoch": 0.11391265195857722,
+      "grad_norm": 0.16683144867420197,
+      "learning_rate": 0.00019744311144354208,
+      "loss": 0.817,
+      "step": 253
+    },
+    {
+      "epoch": 0.11436289959477713,
+      "grad_norm": 0.13282062113285065,
+      "learning_rate": 0.00019740972435213115,
+      "loss": 0.6512,
+      "step": 254
+    },
+    {
+      "epoch": 0.11481314723097703,
+      "grad_norm": 0.15675008296966553,
+      "learning_rate": 0.00019737612355278742,
+      "loss": 0.6616,
+      "step": 255
+    },
+    {
+      "epoch": 0.11526339486717695,
+      "grad_norm": 0.1303519308567047,
+      "learning_rate": 0.00019734230911922797,
+      "loss": 0.5815,
+      "step": 256
+    },
+    {
+      "epoch": 0.11571364250337686,
+      "grad_norm": 0.15984652936458588,
+      "learning_rate": 0.00019730828112563854,
+      "loss": 0.8434,
+      "step": 257
+    },
+    {
+      "epoch": 0.11616389013957677,
+      "grad_norm": 0.15283840894699097,
+      "learning_rate": 0.00019727403964667342,
+      "loss": 0.6213,
+      "step": 258
+    },
+    {
+      "epoch": 0.11661413777577667,
+      "grad_norm": 0.20689162611961365,
+      "learning_rate": 0.0001972395847574552,
+      "loss": 0.7758,
+      "step": 259
+    },
+    {
+      "epoch": 0.11706438541197658,
+      "grad_norm": 0.2301361858844757,
+      "learning_rate": 0.00019720491653357476,
+      "loss": 0.6383,
+      "step": 260
+    },
+    {
+      "epoch": 0.1175146330481765,
+      "grad_norm": 0.10308758914470673,
+      "learning_rate": 0.00019717003505109095,
+      "loss": 0.5113,
+      "step": 261
+    },
+    {
+      "epoch": 0.11796488068437641,
+      "grad_norm": 0.13094520568847656,
+      "learning_rate": 0.00019713494038653057,
+      "loss": 0.739,
+      "step": 262
+    },
+    {
+      "epoch": 0.11841512832057632,
+      "grad_norm": 0.19437825679779053,
+      "learning_rate": 0.00019709963261688805,
+      "loss": 0.7061,
+      "step": 263
+    },
+    {
+      "epoch": 0.11886537595677622,
+      "grad_norm": 0.16305358707904816,
+      "learning_rate": 0.00019706411181962533,
+      "loss": 0.4889,
+      "step": 264
+    },
+    {
+      "epoch": 0.11931562359297614,
+      "grad_norm": 0.1489294022321701,
+      "learning_rate": 0.0001970283780726718,
+      "loss": 0.6531,
+      "step": 265
+    },
+    {
+      "epoch": 0.11976587122917605,
+      "grad_norm": 0.1914891004562378,
+      "learning_rate": 0.00019699243145442397,
+      "loss": 0.7552,
+      "step": 266
+    },
+    {
+      "epoch": 0.12021611886537596,
+      "grad_norm": 0.1331598460674286,
+      "learning_rate": 0.00019695627204374547,
+      "loss": 0.506,
+      "step": 267
+    },
+    {
+      "epoch": 0.12066636650157586,
+      "grad_norm": 0.1297321915626526,
+      "learning_rate": 0.00019691989991996663,
+      "loss": 0.8836,
+      "step": 268
+    },
+    {
+      "epoch": 0.12111661413777577,
+      "grad_norm": 0.1162014752626419,
+      "learning_rate": 0.00019688331516288452,
+      "loss": 0.5927,
+      "step": 269
+    },
+    {
+      "epoch": 0.12156686177397569,
+      "grad_norm": 0.12953799962997437,
+      "learning_rate": 0.0001968465178527628,
+      "loss": 0.5941,
+      "step": 270
+    },
+    {
+      "epoch": 0.1220171094101756,
+      "grad_norm": 0.10634926706552505,
+      "learning_rate": 0.00019680950807033127,
+      "loss": 0.4479,
+      "step": 271
+    },
+    {
+      "epoch": 0.12246735704637551,
+      "grad_norm": 0.1020803153514862,
+      "learning_rate": 0.00019677228589678601,
+      "loss": 0.3629,
+      "step": 272
+    },
+    {
+      "epoch": 0.12291760468257541,
+      "grad_norm": 0.1202622652053833,
+      "learning_rate": 0.00019673485141378906,
+      "loss": 0.64,
+      "step": 273
+    },
+    {
+      "epoch": 0.12336785231877533,
+      "grad_norm": 0.10577255487442017,
+      "learning_rate": 0.0001966972047034682,
+      "loss": 0.4982,
+      "step": 274
+    },
+    {
+      "epoch": 0.12381809995497524,
+      "grad_norm": 0.24492061138153076,
+      "learning_rate": 0.00019665934584841682,
+      "loss": 0.5144,
+      "step": 275
+    },
+    {
+      "epoch": 0.12426834759117515,
+      "grad_norm": 0.1439315229654312,
+      "learning_rate": 0.00019662127493169372,
+      "loss": 0.5807,
+      "step": 276
+    },
+    {
+      "epoch": 0.12471859522737505,
+      "grad_norm": 0.17341749370098114,
+      "learning_rate": 0.00019658299203682305,
+      "loss": 0.7106,
+      "step": 277
+    },
+    {
+      "epoch": 0.12516884286357496,
+      "grad_norm": 0.1397465020418167,
+      "learning_rate": 0.00019654449724779388,
+      "loss": 0.5442,
+      "step": 278
+    },
+    {
+      "epoch": 0.12561909049977488,
+      "grad_norm": 0.1980588436126709,
+      "learning_rate": 0.0001965057906490602,
+      "loss": 0.4957,
+      "step": 279
+    },
+    {
+      "epoch": 0.1260693381359748,
+      "grad_norm": 0.17171113193035126,
+      "learning_rate": 0.0001964668723255408,
+      "loss": 0.7713,
+      "step": 280
+    },
+    {
+      "epoch": 0.1265195857721747,
+      "grad_norm": 0.1723875254392624,
+      "learning_rate": 0.00019642774236261883,
+      "loss": 0.6049,
+      "step": 281
+    },
+    {
+      "epoch": 0.12696983340837462,
+      "grad_norm": 0.15818177163600922,
+      "learning_rate": 0.00019638840084614182,
+      "loss": 0.7384,
+      "step": 282
+    },
+    {
+      "epoch": 0.12742008104457453,
+      "grad_norm": 0.10646625608205795,
+      "learning_rate": 0.0001963488478624214,
+      "loss": 0.5811,
+      "step": 283
+    },
+    {
+      "epoch": 0.12787032868077441,
+      "grad_norm": 0.1276092380285263,
+      "learning_rate": 0.00019630908349823318,
+      "loss": 0.6216,
+      "step": 284
+    },
+    {
+      "epoch": 0.12832057631697433,
+      "grad_norm": 0.16972172260284424,
+      "learning_rate": 0.0001962691078408165,
+      "loss": 0.6525,
+      "step": 285
+    },
+    {
+      "epoch": 0.12877082395317424,
+      "grad_norm": 0.12826530635356903,
+      "learning_rate": 0.00019622892097787426,
+      "loss": 0.5282,
+      "step": 286
+    },
+    {
+      "epoch": 0.12922107158937415,
+      "grad_norm": 0.1766362488269806,
+      "learning_rate": 0.0001961885229975727,
+      "loss": 0.5094,
+      "step": 287
+    },
+    {
+      "epoch": 0.12967131922557407,
+      "grad_norm": 0.1271311342716217,
+      "learning_rate": 0.00019614791398854132,
+      "loss": 0.64,
+      "step": 288
+    },
+    {
+      "epoch": 0.13012156686177398,
+      "grad_norm": 0.10138287395238876,
+      "learning_rate": 0.00019610709403987246,
+      "loss": 0.5693,
+      "step": 289
+    },
+    {
+      "epoch": 0.1305718144979739,
+      "grad_norm": 0.12682873010635376,
+      "learning_rate": 0.00019606606324112135,
+      "loss": 0.4281,
+      "step": 290
+    },
+    {
+      "epoch": 0.1310220621341738,
+      "grad_norm": 0.1417691856622696,
+      "learning_rate": 0.00019602482168230577,
+      "loss": 0.7213,
+      "step": 291
+    },
+    {
+      "epoch": 0.13147230977037372,
+      "grad_norm": 0.09300810843706131,
+      "learning_rate": 0.00019598336945390586,
+      "loss": 0.4848,
+      "step": 292
+    },
+    {
+      "epoch": 0.1319225574065736,
+      "grad_norm": 0.14398658275604248,
+      "learning_rate": 0.00019594170664686403,
+      "loss": 0.97,
+      "step": 293
+    },
+    {
+      "epoch": 0.13237280504277352,
+      "grad_norm": 0.1314631700515747,
+      "learning_rate": 0.00019589983335258456,
+      "loss": 0.8713,
+      "step": 294
+    },
+    {
+      "epoch": 0.13282305267897343,
+      "grad_norm": 0.1376807987689972,
+      "learning_rate": 0.00019585774966293364,
+      "loss": 0.7095,
+      "step": 295
+    },
+    {
+      "epoch": 0.13327330031517334,
+      "grad_norm": 0.1281300187110901,
+      "learning_rate": 0.000195815455670239,
+      "loss": 0.4981,
+      "step": 296
+    },
+    {
+      "epoch": 0.13372354795137326,
+      "grad_norm": 0.15598444640636444,
+      "learning_rate": 0.00019577295146728973,
+      "loss": 0.9382,
+      "step": 297
+    },
+    {
+      "epoch": 0.13417379558757317,
+      "grad_norm": 0.1730741262435913,
+      "learning_rate": 0.0001957302371473361,
+      "loss": 0.7257,
+      "step": 298
+    },
+    {
+      "epoch": 0.13462404322377308,
+      "grad_norm": 0.13347448408603668,
+      "learning_rate": 0.00019568731280408946,
+      "loss": 0.5904,
+      "step": 299
+    },
+    {
+      "epoch": 0.135074290859973,
+      "grad_norm": 0.14394089579582214,
+      "learning_rate": 0.00019564417853172174,
+      "loss": 0.821,
+      "step": 300
+    },
+    {
+      "epoch": 0.1355245384961729,
+      "grad_norm": 0.10720130056142807,
+      "learning_rate": 0.00019560083442486564,
+      "loss": 0.5467,
+      "step": 301
+    },
+    {
+      "epoch": 0.1359747861323728,
+      "grad_norm": 0.1632530391216278,
+      "learning_rate": 0.0001955572805786141,
+      "loss": 0.8096,
+      "step": 302
+    },
+    {
+      "epoch": 0.1364250337685727,
+      "grad_norm": 0.11133529245853424,
+      "learning_rate": 0.0001955135170885202,
+      "loss": 0.4978,
+      "step": 303
+    },
+    {
+      "epoch": 0.13687528140477262,
+      "grad_norm": 0.14525838196277618,
+      "learning_rate": 0.000195469544050597,
+      "loss": 0.7541,
+      "step": 304
+    },
+    {
+      "epoch": 0.13732552904097253,
+      "grad_norm": 0.11284986883401871,
+      "learning_rate": 0.00019542536156131734,
+      "loss": 0.5897,
+      "step": 305
+    },
+    {
+      "epoch": 0.13777577667717245,
+      "grad_norm": 0.10966871678829193,
+      "learning_rate": 0.00019538096971761344,
+      "loss": 0.3643,
+      "step": 306
+    },
+    {
+      "epoch": 0.13822602431337236,
+      "grad_norm": 0.12545877695083618,
+      "learning_rate": 0.00019533636861687696,
+      "loss": 0.6678,
+      "step": 307
+    },
+    {
+      "epoch": 0.13867627194957227,
+      "grad_norm": 0.18901687860488892,
+      "learning_rate": 0.00019529155835695857,
+      "loss": 0.6655,
+      "step": 308
+    },
+    {
+      "epoch": 0.13912651958577218,
+      "grad_norm": 0.12730500102043152,
+      "learning_rate": 0.00019524653903616783,
+      "loss": 0.5449,
+      "step": 309
+    },
+    {
+      "epoch": 0.1395767672219721,
+      "grad_norm": 0.1989346295595169,
+      "learning_rate": 0.00019520131075327298,
+      "loss": 0.8237,
+      "step": 310
+    },
+    {
+      "epoch": 0.14002701485817198,
+      "grad_norm": 0.11587393283843994,
+      "learning_rate": 0.0001951558736075007,
+      "loss": 0.5054,
+      "step": 311
+    },
+    {
+      "epoch": 0.1404772624943719,
+      "grad_norm": 0.12517288327217102,
+      "learning_rate": 0.00019511022769853588,
+      "loss": 0.6994,
+      "step": 312
+    },
+    {
+      "epoch": 0.1409275101305718,
+      "grad_norm": 0.140543594956398,
+      "learning_rate": 0.00019506437312652146,
+      "loss": 0.4638,
+      "step": 313
+    },
+    {
+      "epoch": 0.14137775776677172,
+      "grad_norm": 0.11124718934297562,
+      "learning_rate": 0.00019501830999205805,
+      "loss": 0.5011,
+      "step": 314
+    },
+    {
+      "epoch": 0.14182800540297164,
+      "grad_norm": 0.13298822939395905,
+      "learning_rate": 0.00019497203839620398,
+      "loss": 0.5607,
+      "step": 315
+    },
+    {
+      "epoch": 0.14227825303917155,
+      "grad_norm": 0.12221076339483261,
+      "learning_rate": 0.00019492555844047486,
+      "loss": 0.7314,
+      "step": 316
+    },
+    {
+      "epoch": 0.14272850067537146,
+      "grad_norm": 0.15603968501091003,
+      "learning_rate": 0.00019487887022684336,
+      "loss": 0.6031,
+      "step": 317
+    },
+    {
+      "epoch": 0.14317874831157137,
+      "grad_norm": 0.11718322336673737,
+      "learning_rate": 0.00019483197385773915,
+      "loss": 0.4385,
+      "step": 318
+    },
+    {
+      "epoch": 0.1436289959477713,
+      "grad_norm": 0.19184814393520355,
+      "learning_rate": 0.00019478486943604853,
+      "loss": 0.9228,
+      "step": 319
+    },
+    {
+      "epoch": 0.14407924358397117,
+      "grad_norm": 0.1626308113336563,
+      "learning_rate": 0.0001947375570651142,
+      "loss": 0.5238,
+      "step": 320
+    },
+    {
+      "epoch": 0.14452949122017109,
+      "grad_norm": 0.13151338696479797,
+      "learning_rate": 0.00019469003684873516,
+      "loss": 0.4863,
+      "step": 321
+    },
+    {
+      "epoch": 0.144979738856371,
+      "grad_norm": 0.17465566098690033,
+      "learning_rate": 0.0001946423088911664,
+      "loss": 0.8271,
+      "step": 322
+    },
+    {
+      "epoch": 0.1454299864925709,
+      "grad_norm": 0.11772558093070984,
+      "learning_rate": 0.00019459437329711867,
+      "loss": 0.4249,
+      "step": 323
+    },
+    {
+      "epoch": 0.14588023412877082,
+      "grad_norm": 0.12523289024829865,
+      "learning_rate": 0.00019454623017175812,
+      "loss": 0.6053,
+      "step": 324
+    },
+    {
+      "epoch": 0.14633048176497074,
+      "grad_norm": 0.1224207729101181,
+      "learning_rate": 0.00019449787962070642,
+      "loss": 0.3191,
+      "step": 325
+    },
+    {
+      "epoch": 0.14678072940117065,
+      "grad_norm": 0.13695955276489258,
+      "learning_rate": 0.00019444932175004018,
+      "loss": 0.7099,
+      "step": 326
+    },
+    {
+      "epoch": 0.14723097703737056,
+      "grad_norm": 0.11768505722284317,
+      "learning_rate": 0.0001944005566662909,
+      "loss": 0.5006,
+      "step": 327
+    },
+    {
+      "epoch": 0.14768122467357048,
+      "grad_norm": 0.13041026890277863,
+      "learning_rate": 0.0001943515844764446,
+      "loss": 0.7371,
+      "step": 328
+    },
+    {
+      "epoch": 0.14813147230977036,
+      "grad_norm": 0.14333194494247437,
+      "learning_rate": 0.00019430240528794178,
+      "loss": 0.564,
+      "step": 329
+    },
+    {
+      "epoch": 0.14858171994597028,
+      "grad_norm": 0.22557055950164795,
+      "learning_rate": 0.00019425301920867707,
+      "loss": 0.8325,
+      "step": 330
+    },
+    {
+      "epoch": 0.1490319675821702,
+      "grad_norm": 0.09881255030632019,
+      "learning_rate": 0.0001942034263469989,
+      "loss": 0.3952,
+      "step": 331
+    },
+    {
+      "epoch": 0.1494822152183701,
+      "grad_norm": 0.17186418175697327,
+      "learning_rate": 0.0001941536268117095,
+      "loss": 0.6872,
+      "step": 332
+    },
+    {
+      "epoch": 0.14993246285457001,
+      "grad_norm": 0.10947912931442261,
+      "learning_rate": 0.00019410362071206437,
+      "loss": 0.7905,
+      "step": 333
+    },
+    {
+      "epoch": 0.15038271049076993,
+      "grad_norm": 0.1941973865032196,
+      "learning_rate": 0.00019405340815777233,
+      "loss": 1.194,
+      "step": 334
+    },
+    {
+      "epoch": 0.15083295812696984,
+      "grad_norm": 0.14739827811717987,
+      "learning_rate": 0.00019400298925899504,
+      "loss": 0.9983,
+      "step": 335
+    },
+    {
+      "epoch": 0.15128320576316975,
+      "grad_norm": 0.16295821964740753,
+      "learning_rate": 0.00019395236412634694,
+      "loss": 0.8202,
+      "step": 336
+    },
+    {
+      "epoch": 0.15173345339936967,
+      "grad_norm": 0.11084377765655518,
+      "learning_rate": 0.00019390153287089488,
+      "loss": 0.4248,
+      "step": 337
+    },
+    {
+      "epoch": 0.15218370103556955,
+      "grad_norm": 0.14127840101718903,
+      "learning_rate": 0.00019385049560415794,
+      "loss": 0.7224,
+      "step": 338
+    },
+    {
+      "epoch": 0.15263394867176946,
+      "grad_norm": 0.2182367444038391,
+      "learning_rate": 0.00019379925243810713,
+      "loss": 0.5326,
+      "step": 339
+    },
+    {
+      "epoch": 0.15308419630796938,
+      "grad_norm": 0.13091087341308594,
+      "learning_rate": 0.00019374780348516526,
+      "loss": 0.4927,
+      "step": 340
+    },
+    {
+      "epoch": 0.1535344439441693,
+      "grad_norm": 0.16064320504665375,
+      "learning_rate": 0.00019369614885820656,
+      "loss": 0.7003,
+      "step": 341
+    },
+    {
+      "epoch": 0.1539846915803692,
+      "grad_norm": 0.09630954265594482,
+      "learning_rate": 0.00019364428867055654,
+      "loss": 0.3861,
+      "step": 342
+    },
+    {
+      "epoch": 0.15443493921656912,
+      "grad_norm": 0.12148114293813705,
+      "learning_rate": 0.00019359222303599164,
+      "loss": 0.537,
+      "step": 343
+    },
+    {
+      "epoch": 0.15488518685276903,
+      "grad_norm": 0.14204011857509613,
+      "learning_rate": 0.000193539952068739,
+      "loss": 0.8411,
+      "step": 344
+    },
+    {
+      "epoch": 0.15533543448896894,
+      "grad_norm": 0.14506953954696655,
+      "learning_rate": 0.00019348747588347637,
+      "loss": 0.5471,
+      "step": 345
+    },
+    {
+      "epoch": 0.15578568212516886,
+      "grad_norm": 0.08959248661994934,
+      "learning_rate": 0.0001934347945953316,
+      "loss": 0.5253,
+      "step": 346
+    },
+    {
+      "epoch": 0.15623592976136874,
+      "grad_norm": 0.09563452005386353,
+      "learning_rate": 0.00019338190831988262,
+      "loss": 0.5858,
+      "step": 347
+    },
+    {
+      "epoch": 0.15668617739756865,
+      "grad_norm": 0.12575723230838776,
+      "learning_rate": 0.00019332881717315696,
+      "loss": 0.681,
+      "step": 348
+    },
+    {
+      "epoch": 0.15713642503376857,
+      "grad_norm": 0.10907140374183655,
+      "learning_rate": 0.00019327552127163173,
+      "loss": 0.3553,
+      "step": 349
+    },
+    {
+      "epoch": 0.15758667266996848,
+      "grad_norm": 0.11749302595853806,
+      "learning_rate": 0.00019322202073223323,
+      "loss": 0.3902,
+      "step": 350
+    },
+    {
+      "epoch": 0.1580369203061684,
+      "grad_norm": 0.13981446623802185,
+      "learning_rate": 0.00019316831567233665,
+      "loss": 0.8966,
+      "step": 351
+    },
+    {
+      "epoch": 0.1584871679423683,
+      "grad_norm": 0.10998506844043732,
+      "learning_rate": 0.00019311440620976597,
+      "loss": 0.6452,
+      "step": 352
+    },
+    {
+      "epoch": 0.15893741557856822,
+      "grad_norm": 0.1666320413351059,
+      "learning_rate": 0.00019306029246279354,
+      "loss": 0.6171,
+      "step": 353
+    },
+    {
+      "epoch": 0.15938766321476813,
+      "grad_norm": 0.2394436001777649,
+      "learning_rate": 0.00019300597455013994,
+      "loss": 0.9888,
+      "step": 354
+    },
+    {
+      "epoch": 0.15983791085096805,
+      "grad_norm": 0.12341636419296265,
+      "learning_rate": 0.00019295145259097364,
+      "loss": 0.7458,
+      "step": 355
+    },
+    {
+      "epoch": 0.16028815848716793,
+      "grad_norm": 0.10844842344522476,
+      "learning_rate": 0.00019289672670491078,
+      "loss": 0.4957,
+      "step": 356
+    },
+    {
+      "epoch": 0.16073840612336784,
+      "grad_norm": 0.12860339879989624,
+      "learning_rate": 0.0001928417970120149,
+      "loss": 0.5372,
+      "step": 357
+    },
+    {
+      "epoch": 0.16118865375956776,
+      "grad_norm": 0.1366884410381317,
+      "learning_rate": 0.00019278666363279666,
+      "loss": 0.6109,
+      "step": 358
+    },
+    {
+      "epoch": 0.16163890139576767,
+      "grad_norm": 0.12508247792720795,
+      "learning_rate": 0.00019273132668821364,
+      "loss": 0.7143,
+      "step": 359
+    },
+    {
+      "epoch": 0.16208914903196758,
+      "grad_norm": 0.14982745051383972,
+      "learning_rate": 0.00019267578629966994,
+      "loss": 0.8145,
+      "step": 360
+    },
+    {
+      "epoch": 0.1625393966681675,
+      "grad_norm": 0.18936455249786377,
+      "learning_rate": 0.00019262004258901604,
+      "loss": 0.6182,
+      "step": 361
+    },
+    {
+      "epoch": 0.1629896443043674,
+      "grad_norm": 0.12429028004407883,
+      "learning_rate": 0.00019256409567854847,
+      "loss": 0.4624,
+      "step": 362
+    },
+    {
+      "epoch": 0.16343989194056732,
+      "grad_norm": 0.1498599797487259,
+      "learning_rate": 0.00019250794569100962,
+      "loss": 0.7231,
+      "step": 363
+    },
+    {
+      "epoch": 0.16389013957676724,
+      "grad_norm": 0.10616910457611084,
+      "learning_rate": 0.0001924515927495874,
+      "loss": 0.5302,
+      "step": 364
+    },
+    {
+      "epoch": 0.16434038721296712,
+      "grad_norm": 0.1957257241010666,
+      "learning_rate": 0.0001923950369779149,
+      "loss": 0.483,
+      "step": 365
+    },
+    {
+      "epoch": 0.16479063484916703,
+      "grad_norm": 0.1274428367614746,
+      "learning_rate": 0.00019233827850007027,
+      "loss": 0.6651,
+      "step": 366
+    },
+    {
+      "epoch": 0.16524088248536695,
+      "grad_norm": 0.13893334567546844,
+      "learning_rate": 0.00019228131744057636,
+      "loss": 0.6261,
+      "step": 367
+    },
+    {
+      "epoch": 0.16569113012156686,
+      "grad_norm": 0.12449665367603302,
+      "learning_rate": 0.0001922241539244005,
+      "loss": 0.4765,
+      "step": 368
+    },
+    {
+      "epoch": 0.16614137775776677,
+      "grad_norm": 0.16900379955768585,
+      "learning_rate": 0.0001921667880769541,
+      "loss": 0.6946,
+      "step": 369
+    },
+    {
+      "epoch": 0.16659162539396669,
+      "grad_norm": 0.1074359118938446,
+      "learning_rate": 0.00019210922002409263,
+      "loss": 0.4501,
+      "step": 370
+    },
+    {
+      "epoch": 0.1670418730301666,
+      "grad_norm": 0.13378627598285675,
+      "learning_rate": 0.00019205144989211496,
+      "loss": 0.5609,
+      "step": 371
+    },
+    {
+      "epoch": 0.1674921206663665,
+      "grad_norm": 0.09647689014673233,
+      "learning_rate": 0.00019199347780776355,
+      "loss": 0.3497,
+      "step": 372
+    },
+    {
+      "epoch": 0.16794236830256642,
+      "grad_norm": 0.1407075673341751,
+      "learning_rate": 0.00019193530389822363,
+      "loss": 0.5982,
+      "step": 373
+    },
+    {
+      "epoch": 0.1683926159387663,
+      "grad_norm": 0.09926603734493256,
+      "learning_rate": 0.00019187692829112353,
+      "loss": 0.5566,
+      "step": 374
+    },
+    {
+      "epoch": 0.16884286357496622,
+      "grad_norm": 0.1123383492231369,
+      "learning_rate": 0.00019181835111453386,
+      "loss": 0.5333,
+      "step": 375
+    },
+    {
+      "epoch": 0.16929311121116614,
+      "grad_norm": 0.12356512993574142,
+      "learning_rate": 0.0001917595724969676,
+      "loss": 0.6167,
+      "step": 376
+    },
+    {
+      "epoch": 0.16974335884736605,
+      "grad_norm": 0.13913817703723907,
+      "learning_rate": 0.0001917005925673795,
+      "loss": 0.7434,
+      "step": 377
+    },
+    {
+      "epoch": 0.17019360648356596,
+      "grad_norm": 0.10188717395067215,
+      "learning_rate": 0.00019164141145516615,
+      "loss": 0.5166,
+      "step": 378
+    },
+    {
+      "epoch": 0.17064385411976588,
+      "grad_norm": 0.14303578436374664,
+      "learning_rate": 0.0001915820292901654,
+      "loss": 0.4715,
+      "step": 379
+    },
+    {
+      "epoch": 0.1710941017559658,
+      "grad_norm": 0.19386319816112518,
+      "learning_rate": 0.0001915224462026563,
+      "loss": 0.7322,
+      "step": 380
+    },
+    {
+      "epoch": 0.1715443493921657,
+      "grad_norm": 0.21589113771915436,
+      "learning_rate": 0.00019146266232335856,
+      "loss": 0.6367,
+      "step": 381
+    },
+    {
+      "epoch": 0.17199459702836561,
+      "grad_norm": 0.12046749144792557,
+      "learning_rate": 0.00019140267778343254,
+      "loss": 0.5864,
+      "step": 382
+    },
+    {
+      "epoch": 0.1724448446645655,
+      "grad_norm": 0.09039587527513504,
+      "learning_rate": 0.00019134249271447873,
+      "loss": 0.3398,
+      "step": 383
+    },
+    {
+      "epoch": 0.1728950923007654,
+      "grad_norm": 0.24143409729003906,
+      "learning_rate": 0.00019128210724853766,
+      "loss": 0.685,
+      "step": 384
+    },
+    {
+      "epoch": 0.17334533993696533,
+      "grad_norm": 0.14764373004436493,
+      "learning_rate": 0.00019122152151808942,
+      "loss": 0.8001,
+      "step": 385
+    },
+    {
+      "epoch": 0.17379558757316524,
+      "grad_norm": 0.1054905503988266,
+      "learning_rate": 0.0001911607356560535,
+      "loss": 0.5227,
+      "step": 386
+    },
+    {
+      "epoch": 0.17424583520936515,
+      "grad_norm": 0.13460716605186462,
+      "learning_rate": 0.0001910997497957885,
+      "loss": 0.7139,
+      "step": 387
+    },
+    {
+      "epoch": 0.17469608284556506,
+      "grad_norm": 0.15011832118034363,
+      "learning_rate": 0.00019103856407109177,
+      "loss": 0.8376,
+      "step": 388
+    },
+    {
+      "epoch": 0.17514633048176498,
+      "grad_norm": 0.11285755038261414,
+      "learning_rate": 0.00019097717861619909,
+      "loss": 0.5861,
+      "step": 389
+    },
+    {
+      "epoch": 0.1755965781179649,
+      "grad_norm": 0.1388164460659027,
+      "learning_rate": 0.00019091559356578447,
+      "loss": 0.6639,
+      "step": 390
+    },
+    {
+      "epoch": 0.1760468257541648,
+      "grad_norm": 0.1127384752035141,
+      "learning_rate": 0.00019085380905495985,
+      "loss": 0.7341,
+      "step": 391
+    },
+    {
+      "epoch": 0.1764970733903647,
+      "grad_norm": 0.12747950851917267,
+      "learning_rate": 0.00019079182521927477,
+      "loss": 0.8167,
+      "step": 392
+    },
+    {
+      "epoch": 0.1769473210265646,
+      "grad_norm": 0.11133919656276703,
+      "learning_rate": 0.00019072964219471597,
+      "loss": 0.4676,
+      "step": 393
+    },
+    {
+      "epoch": 0.17739756866276452,
+      "grad_norm": 0.12316610664129257,
+      "learning_rate": 0.00019066726011770726,
+      "loss": 0.6086,
+      "step": 394
+    },
+    {
+      "epoch": 0.17784781629896443,
+      "grad_norm": 0.1366506665945053,
+      "learning_rate": 0.0001906046791251092,
+      "loss": 0.5879,
+      "step": 395
+    },
+    {
+      "epoch": 0.17829806393516434,
+      "grad_norm": 0.21371358633041382,
+      "learning_rate": 0.0001905418993542187,
+      "loss": 0.6283,
+      "step": 396
+    },
+    {
+      "epoch": 0.17874831157136425,
+      "grad_norm": 0.10454566776752472,
+      "learning_rate": 0.00019047892094276874,
+      "loss": 0.4797,
+      "step": 397
+    },
+    {
+      "epoch": 0.17919855920756417,
+      "grad_norm": 0.12460599839687347,
+      "learning_rate": 0.00019041574402892813,
+      "loss": 0.7523,
+      "step": 398
+    },
+    {
+      "epoch": 0.17964880684376408,
+      "grad_norm": 0.1775173544883728,
+      "learning_rate": 0.00019035236875130122,
+      "loss": 0.6573,
+      "step": 399
+    },
+    {
+      "epoch": 0.180099054479964,
+      "grad_norm": 0.10196325927972794,
+      "learning_rate": 0.00019028879524892752,
+      "loss": 0.58,
+      "step": 400
+    },
+    {
+      "epoch": 0.18054930211616388,
+      "grad_norm": 0.12372785061597824,
+      "learning_rate": 0.00019022502366128135,
+      "loss": 0.4318,
+      "step": 401
+    },
+    {
+      "epoch": 0.1809995497523638,
+      "grad_norm": 0.11982943117618561,
+      "learning_rate": 0.00019016105412827174,
+      "loss": 0.4654,
+      "step": 402
+    },
+    {
+      "epoch": 0.1814497973885637,
+      "grad_norm": 0.17599032819271088,
+      "learning_rate": 0.0001900968867902419,
+      "loss": 0.7423,
+      "step": 403
+    },
+    {
+      "epoch": 0.18190004502476362,
+      "grad_norm": 0.1327388435602188,
+      "learning_rate": 0.0001900325217879691,
+      "loss": 0.7153,
+      "step": 404
+    },
+    {
+      "epoch": 0.18235029266096353,
+      "grad_norm": 0.15787407755851746,
+      "learning_rate": 0.00018996795926266413,
+      "loss": 0.7349,
+      "step": 405
+    },
+    {
+      "epoch": 0.18280054029716344,
+      "grad_norm": 0.14755015075206757,
+      "learning_rate": 0.00018990319935597124,
+      "loss": 0.5413,
+      "step": 406
+    },
+    {
+      "epoch": 0.18325078793336336,
+      "grad_norm": 0.1527538001537323,
+      "learning_rate": 0.00018983824220996767,
+      "loss": 0.709,
+      "step": 407
+    },
+    {
+      "epoch": 0.18370103556956327,
+      "grad_norm": 0.14745236933231354,
+      "learning_rate": 0.0001897730879671634,
+      "loss": 0.5934,
+      "step": 408
+    },
+    {
+      "epoch": 0.18415128320576318,
+      "grad_norm": 0.14777810871601105,
+      "learning_rate": 0.00018970773677050081,
+      "loss": 0.7119,
+      "step": 409
+    },
+    {
+      "epoch": 0.18460153084196307,
+      "grad_norm": 0.1468253880739212,
+      "learning_rate": 0.00018964218876335443,
+      "loss": 0.6637,
+      "step": 410
+    },
+    {
+      "epoch": 0.18505177847816298,
+      "grad_norm": 0.17522287368774414,
+      "learning_rate": 0.00018957644408953045,
+      "loss": 0.5787,
+      "step": 411
+    },
+    {
+      "epoch": 0.1855020261143629,
+      "grad_norm": 0.17414158582687378,
+      "learning_rate": 0.00018951050289326665,
+      "loss": 0.6306,
+      "step": 412
+    },
+    {
+      "epoch": 0.1859522737505628,
+      "grad_norm": 0.11318681389093399,
+      "learning_rate": 0.00018944436531923196,
+      "loss": 0.537,
+      "step": 413
+    },
+    {
+      "epoch": 0.18640252138676272,
+      "grad_norm": 0.10259708762168884,
+      "learning_rate": 0.00018937803151252604,
+      "loss": 0.5882,
+      "step": 414
+    },
+    {
+      "epoch": 0.18685276902296263,
+      "grad_norm": 0.167019322514534,
+      "learning_rate": 0.00018931150161867916,
+      "loss": 0.7422,
+      "step": 415
+    },
+    {
+      "epoch": 0.18730301665916255,
+      "grad_norm": 0.13366097211837769,
+      "learning_rate": 0.0001892447757836518,
+      "loss": 0.6495,
+      "step": 416
+    },
+    {
+      "epoch": 0.18775326429536246,
+      "grad_norm": 0.13971048593521118,
+      "learning_rate": 0.00018917785415383418,
+      "loss": 1.0094,
+      "step": 417
+    },
+    {
+      "epoch": 0.18820351193156237,
+      "grad_norm": 0.21706287562847137,
+      "learning_rate": 0.00018911073687604625,
+      "loss": 0.7533,
+      "step": 418
+    },
+    {
+      "epoch": 0.18865375956776226,
+      "grad_norm": 0.10582134127616882,
+      "learning_rate": 0.00018904342409753705,
+      "loss": 0.5725,
+      "step": 419
+    },
+    {
+      "epoch": 0.18910400720396217,
+      "grad_norm": 0.12243637442588806,
+      "learning_rate": 0.00018897591596598465,
+      "loss": 0.7237,
+      "step": 420
+    },
+    {
+      "epoch": 0.18955425484016208,
+      "grad_norm": 0.13889282941818237,
+      "learning_rate": 0.00018890821262949566,
+      "loss": 0.6723,
+      "step": 421
+    },
+    {
+      "epoch": 0.190004502476362,
+      "grad_norm": 0.15518023073673248,
+      "learning_rate": 0.0001888403142366049,
+      "loss": 0.4716,
+      "step": 422
+    },
+    {
+      "epoch": 0.1904547501125619,
+      "grad_norm": 0.2321695238351822,
+      "learning_rate": 0.0001887722209362752,
+      "loss": 0.5559,
+      "step": 423
+    },
+    {
+      "epoch": 0.19090499774876182,
+      "grad_norm": 0.1782892644405365,
+      "learning_rate": 0.00018870393287789697,
+      "loss": 0.6209,
+      "step": 424
+    },
+    {
+      "epoch": 0.19135524538496174,
+      "grad_norm": 0.1700783222913742,
+      "learning_rate": 0.00018863545021128783,
+      "loss": 0.4632,
+      "step": 425
+    },
+    {
+      "epoch": 0.19180549302116165,
+      "grad_norm": 0.08484381437301636,
+      "learning_rate": 0.0001885667730866925,
+      "loss": 0.3636,
+      "step": 426
+    },
+    {
+      "epoch": 0.19225574065736156,
+      "grad_norm": 0.07963482290506363,
+      "learning_rate": 0.00018849790165478223,
+      "loss": 0.371,
+      "step": 427
+    },
+    {
+      "epoch": 0.19270598829356145,
+      "grad_norm": 0.13142360746860504,
+      "learning_rate": 0.00018842883606665458,
+      "loss": 0.4427,
+      "step": 428
+    },
+    {
+      "epoch": 0.19315623592976136,
+      "grad_norm": 0.13863639533519745,
+      "learning_rate": 0.00018835957647383303,
+      "loss": 0.5489,
+      "step": 429
+    },
+    {
+      "epoch": 0.19360648356596127,
+      "grad_norm": 0.14126621186733246,
+      "learning_rate": 0.00018829012302826677,
+      "loss": 0.5678,
+      "step": 430
+    },
+    {
+      "epoch": 0.1940567312021612,
+      "grad_norm": 0.15903249382972717,
+      "learning_rate": 0.0001882204758823302,
+      "loss": 0.9308,
+      "step": 431
+    },
+    {
+      "epoch": 0.1945069788383611,
+      "grad_norm": 0.14325307309627533,
+      "learning_rate": 0.00018815063518882273,
+      "loss": 0.7108,
+      "step": 432
+    },
+    {
+      "epoch": 0.194957226474561,
+      "grad_norm": 0.13106994330883026,
+      "learning_rate": 0.0001880806011009684,
+      "loss": 0.6671,
+      "step": 433
+    },
+    {
+      "epoch": 0.19540747411076093,
+      "grad_norm": 0.1287631243467331,
+      "learning_rate": 0.00018801037377241554,
+      "loss": 0.5737,
+      "step": 434
+    },
+    {
+      "epoch": 0.19585772174696084,
+      "grad_norm": 0.188482865691185,
+      "learning_rate": 0.00018793995335723642,
+      "loss": 0.6767,
+      "step": 435
+    },
+    {
+      "epoch": 0.19630796938316075,
+      "grad_norm": 0.1125263124704361,
+      "learning_rate": 0.00018786934000992688,
+      "loss": 0.5825,
+      "step": 436
+    },
+    {
+      "epoch": 0.19675821701936064,
+      "grad_norm": 0.11809971183538437,
+      "learning_rate": 0.0001877985338854061,
+      "loss": 0.5825,
+      "step": 437
+    },
+    {
+      "epoch": 0.19720846465556055,
+      "grad_norm": 0.1736549735069275,
+      "learning_rate": 0.00018772753513901622,
+      "loss": 0.5259,
+      "step": 438
+    },
+    {
+      "epoch": 0.19765871229176046,
+      "grad_norm": 0.1427125781774521,
+      "learning_rate": 0.00018765634392652186,
+      "loss": 0.6848,
+      "step": 439
+    },
+    {
+      "epoch": 0.19810895992796038,
+      "grad_norm": 0.15245462954044342,
+      "learning_rate": 0.00018758496040411,
+      "loss": 0.5486,
+      "step": 440
+    },
+    {
+      "epoch": 0.1985592075641603,
+      "grad_norm": 0.10975594073534012,
+      "learning_rate": 0.00018751338472838944,
+      "loss": 0.526,
+      "step": 441
+    },
+    {
+      "epoch": 0.1990094552003602,
+      "grad_norm": 0.1165640652179718,
+      "learning_rate": 0.00018744161705639068,
+      "loss": 0.5775,
+      "step": 442
+    },
+    {
+      "epoch": 0.19945970283656012,
+      "grad_norm": 0.1155453622341156,
+      "learning_rate": 0.00018736965754556528,
+      "loss": 0.55,
+      "step": 443
+    },
+    {
+      "epoch": 0.19990995047276003,
+      "grad_norm": 0.13406625390052795,
+      "learning_rate": 0.0001872975063537858,
+      "loss": 0.6853,
+      "step": 444
+    },
+    {
+      "epoch": 0.20036019810895994,
+      "grad_norm": 0.14500440657138824,
+      "learning_rate": 0.0001872251636393453,
+      "loss": 0.6365,
+      "step": 445
+    },
+    {
+      "epoch": 0.20081044574515983,
+      "grad_norm": 0.10520116239786148,
+      "learning_rate": 0.00018715262956095697,
+      "loss": 0.4706,
+      "step": 446
+    },
+    {
+      "epoch": 0.20126069338135974,
+      "grad_norm": 0.12547025084495544,
+      "learning_rate": 0.00018707990427775388,
+      "loss": 0.42,
+      "step": 447
+    },
+    {
+      "epoch": 0.20171094101755965,
+      "grad_norm": 0.12145490944385529,
+      "learning_rate": 0.00018700698794928863,
+      "loss": 0.5166,
+      "step": 448
+    },
+    {
+      "epoch": 0.20216118865375957,
+      "grad_norm": 0.09916845709085464,
+      "learning_rate": 0.0001869338807355328,
+      "loss": 0.466,
+      "step": 449
+    },
+    {
+      "epoch": 0.20261143628995948,
+      "grad_norm": 0.1008843407034874,
+      "learning_rate": 0.00018686058279687698,
+      "loss": 0.6246,
+      "step": 450
+    },
+    {
+      "epoch": 0.2030616839261594,
+      "grad_norm": 0.11514821648597717,
+      "learning_rate": 0.00018678709429413,
+      "loss": 0.35,
+      "step": 451
+    },
+    {
+      "epoch": 0.2035119315623593,
+      "grad_norm": 0.11495738476514816,
+      "learning_rate": 0.0001867134153885189,
+      "loss": 0.432,
+      "step": 452
+    },
+    {
+      "epoch": 0.20396217919855922,
+      "grad_norm": 0.14869830012321472,
+      "learning_rate": 0.00018663954624168833,
+      "loss": 0.7917,
+      "step": 453
+    },
+    {
+      "epoch": 0.20441242683475913,
+      "grad_norm": 0.15117718279361725,
+      "learning_rate": 0.0001865654870157004,
+      "loss": 0.6199,
+      "step": 454
+    },
+    {
+      "epoch": 0.20486267447095902,
+      "grad_norm": 0.1845121830701828,
+      "learning_rate": 0.00018649123787303423,
+      "loss": 0.7555,
+      "step": 455
+    },
+    {
+      "epoch": 0.20531292210715893,
+      "grad_norm": 0.22003404796123505,
+      "learning_rate": 0.00018641679897658553,
+      "loss": 0.77,
+      "step": 456
+    },
+    {
+      "epoch": 0.20576316974335884,
+      "grad_norm": 0.13569730520248413,
+      "learning_rate": 0.00018634217048966637,
+      "loss": 0.5725,
+      "step": 457
+    },
+    {
+      "epoch": 0.20621341737955876,
+      "grad_norm": 0.15057680010795593,
+      "learning_rate": 0.00018626735257600477,
+      "loss": 0.6863,
+      "step": 458
+    },
+    {
+      "epoch": 0.20666366501575867,
+      "grad_norm": 0.14343906939029694,
+      "learning_rate": 0.0001861923453997443,
+      "loss": 0.4548,
+      "step": 459
+    },
+    {
+      "epoch": 0.20711391265195858,
+      "grad_norm": 0.11216168850660324,
+      "learning_rate": 0.00018611714912544379,
+      "loss": 0.7167,
+      "step": 460
+    },
+    {
+      "epoch": 0.2075641602881585,
+      "grad_norm": 0.11779499053955078,
+      "learning_rate": 0.00018604176391807692,
+      "loss": 0.6868,
+      "step": 461
+    },
+    {
+      "epoch": 0.2080144079243584,
+      "grad_norm": 0.14333891868591309,
+      "learning_rate": 0.00018596618994303184,
+      "loss": 0.7725,
+      "step": 462
+    },
+    {
+      "epoch": 0.20846465556055832,
+      "grad_norm": 0.09890694171190262,
+      "learning_rate": 0.00018589042736611092,
+      "loss": 0.4177,
+      "step": 463
+    },
+    {
+      "epoch": 0.2089149031967582,
+      "grad_norm": 0.13302220404148102,
+      "learning_rate": 0.0001858144763535302,
+      "loss": 0.5341,
+      "step": 464
+    },
+    {
+      "epoch": 0.20936515083295812,
+      "grad_norm": 0.08379435539245605,
+      "learning_rate": 0.0001857383370719192,
+      "loss": 0.3646,
+      "step": 465
+    },
+    {
+      "epoch": 0.20981539846915803,
+      "grad_norm": 0.18817852437496185,
+      "learning_rate": 0.00018566200968832047,
+      "loss": 0.6227,
+      "step": 466
+    },
+    {
+      "epoch": 0.21026564610535795,
+      "grad_norm": 0.149583101272583,
+      "learning_rate": 0.00018558549437018923,
+      "loss": 0.6685,
+      "step": 467
+    },
+    {
+      "epoch": 0.21071589374155786,
+      "grad_norm": 0.11600211262702942,
+      "learning_rate": 0.000185508791285393,
+      "loss": 0.554,
+      "step": 468
+    },
+    {
+      "epoch": 0.21116614137775777,
+      "grad_norm": 0.12885649502277374,
+      "learning_rate": 0.00018543190060221128,
+      "loss": 0.606,
+      "step": 469
+    },
+    {
+      "epoch": 0.21161638901395768,
+      "grad_norm": 0.21655143797397614,
+      "learning_rate": 0.00018535482248933512,
+      "loss": 0.5278,
+      "step": 470
+    },
+    {
+      "epoch": 0.2120666366501576,
+      "grad_norm": 0.1451300084590912,
+      "learning_rate": 0.00018527755711586678,
+      "loss": 0.5828,
+      "step": 471
+    },
+    {
+      "epoch": 0.2125168842863575,
+      "grad_norm": 0.18852467834949493,
+      "learning_rate": 0.00018520010465131935,
+      "loss": 0.8346,
+      "step": 472
+    },
+    {
+      "epoch": 0.2129671319225574,
+      "grad_norm": 0.09463053941726685,
+      "learning_rate": 0.0001851224652656164,
+      "loss": 0.3929,
+      "step": 473
+    },
+    {
+      "epoch": 0.2134173795587573,
+      "grad_norm": 0.10494886338710785,
+      "learning_rate": 0.00018504463912909151,
+      "loss": 0.5152,
+      "step": 474
+    },
+    {
+      "epoch": 0.21386762719495722,
+      "grad_norm": 0.18627098202705383,
+      "learning_rate": 0.00018496662641248807,
+      "loss": 0.9133,
+      "step": 475
+    },
+    {
+      "epoch": 0.21431787483115713,
+      "grad_norm": 0.11310123652219772,
+      "learning_rate": 0.00018488842728695877,
+      "loss": 0.6909,
+      "step": 476
+    },
+    {
+      "epoch": 0.21476812246735705,
+      "grad_norm": 0.20986901223659515,
+      "learning_rate": 0.00018481004192406525,
+      "loss": 0.4951,
+      "step": 477
+    },
+    {
+      "epoch": 0.21521837010355696,
+      "grad_norm": 0.14960235357284546,
+      "learning_rate": 0.00018473147049577774,
+      "loss": 0.4111,
+      "step": 478
+    },
+    {
+      "epoch": 0.21566861773975687,
+      "grad_norm": 0.09463857859373093,
+      "learning_rate": 0.00018465271317447476,
+      "loss": 0.5775,
+      "step": 479
+    },
+    {
+      "epoch": 0.2161188653759568,
+      "grad_norm": 0.11150664836168289,
+      "learning_rate": 0.0001845737701329425,
+      "loss": 0.5166,
+      "step": 480
+    },
+    {
+      "epoch": 0.2165691130121567,
+      "grad_norm": 0.11134067177772522,
+      "learning_rate": 0.00018449464154437478,
+      "loss": 0.746,
+      "step": 481
+    },
+    {
+      "epoch": 0.21701936064835659,
+      "grad_norm": 0.1153070479631424,
+      "learning_rate": 0.00018441532758237234,
+      "loss": 0.4524,
+      "step": 482
+    },
+    {
+      "epoch": 0.2174696082845565,
+      "grad_norm": 0.15530085563659668,
+      "learning_rate": 0.00018433582842094276,
+      "loss": 0.486,
+      "step": 483
+    },
+    {
+      "epoch": 0.2179198559207564,
+      "grad_norm": 0.10047878324985504,
+      "learning_rate": 0.00018425614423449974,
+      "loss": 0.7484,
+      "step": 484
+    },
+    {
+      "epoch": 0.21837010355695632,
+      "grad_norm": 0.11846239119768143,
+      "learning_rate": 0.00018417627519786315,
+      "loss": 0.7191,
+      "step": 485
+    },
+    {
+      "epoch": 0.21882035119315624,
+      "grad_norm": 0.1119120642542839,
+      "learning_rate": 0.0001840962214862582,
+      "loss": 0.5552,
+      "step": 486
+    },
+    {
+      "epoch": 0.21927059882935615,
+      "grad_norm": 0.1144852340221405,
+      "learning_rate": 0.00018401598327531536,
+      "loss": 0.6059,
+      "step": 487
+    },
+    {
+      "epoch": 0.21972084646555606,
+      "grad_norm": 0.12210698425769806,
+      "learning_rate": 0.00018393556074106982,
+      "loss": 0.5376,
+      "step": 488
+    },
+    {
+      "epoch": 0.22017109410175598,
+      "grad_norm": 0.11316890269517899,
+      "learning_rate": 0.0001838549540599612,
+      "loss": 0.3963,
+      "step": 489
+    },
+    {
+      "epoch": 0.2206213417379559,
+      "grad_norm": 0.10670144110918045,
+      "learning_rate": 0.00018377416340883314,
+      "loss": 0.3346,
+      "step": 490
+    },
+    {
+      "epoch": 0.22107158937415577,
+      "grad_norm": 0.13043299317359924,
+      "learning_rate": 0.0001836931889649328,
+      "loss": 0.6036,
+      "step": 491
+    },
+    {
+      "epoch": 0.2215218370103557,
+      "grad_norm": 0.10817930847406387,
+      "learning_rate": 0.00018361203090591071,
+      "loss": 0.4809,
+      "step": 492
+    },
+    {
+      "epoch": 0.2219720846465556,
+      "grad_norm": 0.12122377008199692,
+      "learning_rate": 0.00018353068940982008,
+      "loss": 0.4964,
+      "step": 493
+    },
+    {
+      "epoch": 0.2224223322827555,
+      "grad_norm": 0.16302037239074707,
+      "learning_rate": 0.00018344916465511666,
+      "loss": 0.7115,
+      "step": 494
+    },
+    {
+      "epoch": 0.22287257991895543,
+      "grad_norm": 0.12112080305814743,
+      "learning_rate": 0.0001833674568206582,
+      "loss": 0.3796,
+      "step": 495
+    },
+    {
+      "epoch": 0.22332282755515534,
+      "grad_norm": 0.14118660986423492,
+      "learning_rate": 0.0001832855660857042,
+      "loss": 0.5618,
+      "step": 496
+    },
+    {
+      "epoch": 0.22377307519135525,
+      "grad_norm": 0.2936384975910187,
+      "learning_rate": 0.00018320349262991535,
+      "loss": 0.9151,
+      "step": 497
+    },
+    {
+      "epoch": 0.22422332282755517,
+      "grad_norm": 0.21970796585083008,
+      "learning_rate": 0.0001831212366333532,
+      "loss": 0.6244,
+      "step": 498
+    },
+    {
+      "epoch": 0.22467357046375508,
+      "grad_norm": 0.12831395864486694,
+      "learning_rate": 0.00018303879827647975,
+      "loss": 0.5017,
+      "step": 499
+    },
+    {
+      "epoch": 0.22512381809995496,
+      "grad_norm": 0.11645712703466415,
+      "learning_rate": 0.00018295617774015724,
+      "loss": 0.4761,
+      "step": 500
+    },
+    {
+      "epoch": 0.22557406573615488,
+      "grad_norm": 0.10314258933067322,
+      "learning_rate": 0.00018287337520564746,
+      "loss": 0.7429,
+      "step": 501
+    },
+    {
+      "epoch": 0.2260243133723548,
+      "grad_norm": 0.1125650405883789,
+      "learning_rate": 0.00018279039085461148,
+      "loss": 0.716,
+      "step": 502
+    },
+    {
+      "epoch": 0.2264745610085547,
+      "grad_norm": 0.10940995812416077,
+      "learning_rate": 0.00018270722486910933,
+      "loss": 0.4912,
+      "step": 503
+    },
+    {
+      "epoch": 0.22692480864475462,
+      "grad_norm": 0.12597322463989258,
+      "learning_rate": 0.0001826238774315995,
+      "loss": 0.6679,
+      "step": 504
+    },
+    {
+      "epoch": 0.22737505628095453,
+      "grad_norm": 0.10549933463335037,
+      "learning_rate": 0.00018254034872493854,
+      "loss": 0.3847,
+      "step": 505
+    },
+    {
+      "epoch": 0.22782530391715444,
+      "grad_norm": 0.16294513642787933,
+      "learning_rate": 0.00018245663893238075,
+      "loss": 0.8137,
+      "step": 506
+    },
+    {
+      "epoch": 0.22827555155335436,
+      "grad_norm": 0.11669060587882996,
+      "learning_rate": 0.00018237274823757762,
+      "loss": 0.4301,
+      "step": 507
+    },
+    {
+      "epoch": 0.22872579918955427,
+      "grad_norm": 0.09987634420394897,
+      "learning_rate": 0.00018228867682457765,
+      "loss": 0.5634,
+      "step": 508
+    },
+    {
+      "epoch": 0.22917604682575415,
+      "grad_norm": 0.16850747168064117,
+      "learning_rate": 0.00018220442487782568,
+      "loss": 0.9629,
+      "step": 509
+    },
+    {
+      "epoch": 0.22962629446195407,
+      "grad_norm": 0.20761728286743164,
+      "learning_rate": 0.00018211999258216274,
+      "loss": 0.7455,
+      "step": 510
+    },
+    {
+      "epoch": 0.23007654209815398,
+      "grad_norm": 0.0917520597577095,
+      "learning_rate": 0.0001820353801228255,
+      "loss": 0.314,
+      "step": 511
+    },
+    {
+      "epoch": 0.2305267897343539,
+      "grad_norm": 0.11821037530899048,
+      "learning_rate": 0.00018195058768544583,
+      "loss": 0.4385,
+      "step": 512
+    },
+    {
+      "epoch": 0.2309770373705538,
+      "grad_norm": 0.24366021156311035,
+      "learning_rate": 0.00018186561545605054,
+      "loss": 0.8605,
+      "step": 513
+    },
+    {
+      "epoch": 0.23142728500675372,
+      "grad_norm": 0.10586755722761154,
+      "learning_rate": 0.00018178046362106085,
+      "loss": 0.5973,
+      "step": 514
+    },
+    {
+      "epoch": 0.23187753264295363,
+      "grad_norm": 0.13302288949489594,
+      "learning_rate": 0.00018169513236729196,
+      "loss": 0.6625,
+      "step": 515
+    },
+    {
+      "epoch": 0.23232778027915355,
+      "grad_norm": 0.13257688283920288,
+      "learning_rate": 0.00018160962188195278,
+      "loss": 0.5042,
+      "step": 516
+    },
+    {
+      "epoch": 0.23277802791535346,
+      "grad_norm": 0.24843749403953552,
+      "learning_rate": 0.00018152393235264548,
+      "loss": 0.5555,
+      "step": 517
+    },
+    {
+      "epoch": 0.23322827555155334,
+      "grad_norm": 0.1322741061449051,
+      "learning_rate": 0.00018143806396736487,
+      "loss": 0.6156,
+      "step": 518
+    },
+    {
+      "epoch": 0.23367852318775326,
+      "grad_norm": 0.12498054653406143,
+      "learning_rate": 0.0001813520169144983,
+      "loss": 0.7439,
+      "step": 519
+    },
+    {
+      "epoch": 0.23412877082395317,
+      "grad_norm": 0.11136923730373383,
+      "learning_rate": 0.00018126579138282503,
+      "loss": 0.3396,
+      "step": 520
+    },
+    {
+      "epoch": 0.23457901846015308,
+      "grad_norm": 0.10236714035272598,
+      "learning_rate": 0.00018117938756151593,
+      "loss": 0.4114,
+      "step": 521
+    },
+    {
+      "epoch": 0.235029266096353,
+      "grad_norm": 0.14691545069217682,
+      "learning_rate": 0.000181092805640133,
+      "loss": 0.7668,
+      "step": 522
+    },
+    {
+      "epoch": 0.2354795137325529,
+      "grad_norm": 0.22314922511577606,
+      "learning_rate": 0.000181006045808629,
+      "loss": 0.797,
+      "step": 523
+    },
+    {
+      "epoch": 0.23592976136875282,
+      "grad_norm": 0.14861617982387543,
+      "learning_rate": 0.0001809191082573469,
+      "loss": 0.7364,
+      "step": 524
+    },
+    {
+      "epoch": 0.23638000900495273,
+      "grad_norm": 0.13408222794532776,
+      "learning_rate": 0.00018083199317701971,
+      "loss": 0.457,
+      "step": 525
+    },
+    {
+      "epoch": 0.23683025664115265,
+      "grad_norm": 0.14997214078903198,
+      "learning_rate": 0.00018074470075876985,
+      "loss": 0.6158,
+      "step": 526
+    },
+    {
+      "epoch": 0.23728050427735253,
+      "grad_norm": 0.19004739820957184,
+      "learning_rate": 0.00018065723119410884,
+      "loss": 0.5877,
+      "step": 527
+    },
+    {
+      "epoch": 0.23773075191355245,
+      "grad_norm": 0.12186384201049805,
+      "learning_rate": 0.0001805695846749368,
+      "loss": 0.6983,
+      "step": 528
+    },
+    {
+      "epoch": 0.23818099954975236,
+      "grad_norm": 0.12739919126033783,
+      "learning_rate": 0.0001804817613935421,
+      "loss": 0.8073,
+      "step": 529
+    },
+    {
+      "epoch": 0.23863124718595227,
+      "grad_norm": 0.17672795057296753,
+      "learning_rate": 0.0001803937615426009,
+      "loss": 0.8065,
+      "step": 530
+    },
+    {
+      "epoch": 0.23908149482215219,
+      "grad_norm": 0.11641567945480347,
+      "learning_rate": 0.00018030558531517671,
+      "loss": 0.4231,
+      "step": 531
+    },
+    {
+      "epoch": 0.2395317424583521,
+      "grad_norm": 0.13499368727207184,
+      "learning_rate": 0.00018021723290472007,
+      "loss": 0.6501,
+      "step": 532
+    },
+    {
+      "epoch": 0.239981990094552,
+      "grad_norm": 0.131595179438591,
+      "learning_rate": 0.000180128704505068,
+      "loss": 0.8086,
+      "step": 533
+    },
+    {
+      "epoch": 0.24043223773075192,
+      "grad_norm": 0.12846219539642334,
+      "learning_rate": 0.0001800400003104436,
+      "loss": 0.6585,
+      "step": 534
+    },
+    {
+      "epoch": 0.24088248536695184,
+      "grad_norm": 0.10282287746667862,
+      "learning_rate": 0.00017995112051545574,
+      "loss": 0.6292,
+      "step": 535
+    },
+    {
+      "epoch": 0.24133273300315172,
+      "grad_norm": 0.12396449595689774,
+      "learning_rate": 0.00017986206531509836,
+      "loss": 0.6804,
+      "step": 536
+    },
+    {
+      "epoch": 0.24178298063935164,
+      "grad_norm": 0.09174476563930511,
+      "learning_rate": 0.00017977283490475045,
+      "loss": 0.5154,
+      "step": 537
+    },
+    {
+      "epoch": 0.24223322827555155,
+      "grad_norm": 0.1682201623916626,
+      "learning_rate": 0.00017968342948017523,
+      "loss": 0.602,
+      "step": 538
+    },
+    {
+      "epoch": 0.24268347591175146,
+      "grad_norm": 0.12803952395915985,
+      "learning_rate": 0.00017959384923751993,
+      "loss": 0.6052,
+      "step": 539
+    },
+    {
+      "epoch": 0.24313372354795137,
+      "grad_norm": 0.11690130829811096,
+      "learning_rate": 0.00017950409437331538,
+      "loss": 0.5479,
+      "step": 540
+    },
+    {
+      "epoch": 0.2435839711841513,
+      "grad_norm": 0.08725091814994812,
+      "learning_rate": 0.00017941416508447536,
+      "loss": 0.5069,
+      "step": 541
+    },
+    {
+      "epoch": 0.2440342188203512,
+      "grad_norm": 0.13879774510860443,
+      "learning_rate": 0.00017932406156829651,
+      "loss": 0.758,
+      "step": 542
+    },
+    {
+      "epoch": 0.2444844664565511,
+      "grad_norm": 0.11410533636808395,
+      "learning_rate": 0.00017923378402245758,
+      "loss": 0.5867,
+      "step": 543
+    },
+    {
+      "epoch": 0.24493471409275103,
+      "grad_norm": 0.14098933339118958,
+      "learning_rate": 0.00017914333264501916,
+      "loss": 0.6009,
+      "step": 544
+    },
+    {
+      "epoch": 0.2453849617289509,
+      "grad_norm": 0.10953424870967865,
+      "learning_rate": 0.00017905270763442325,
+      "loss": 0.6121,
+      "step": 545
+    },
+    {
+      "epoch": 0.24583520936515083,
+      "grad_norm": 0.1401321440935135,
+      "learning_rate": 0.00017896190918949268,
+      "loss": 0.7023,
+      "step": 546
+    },
+    {
+      "epoch": 0.24628545700135074,
+      "grad_norm": 0.11982724070549011,
+      "learning_rate": 0.0001788709375094309,
+      "loss": 0.6408,
+      "step": 547
+    },
+    {
+      "epoch": 0.24673570463755065,
+      "grad_norm": 0.08310122787952423,
+      "learning_rate": 0.00017877979279382135,
+      "loss": 0.348,
+      "step": 548
+    },
+    {
+      "epoch": 0.24718595227375056,
+      "grad_norm": 0.07799794524908066,
+      "learning_rate": 0.00017868847524262708,
+      "loss": 0.3912,
+      "step": 549
+    },
+    {
+      "epoch": 0.24763619990995048,
+      "grad_norm": 0.15976564586162567,
+      "learning_rate": 0.00017859698505619045,
+      "loss": 0.761,
+      "step": 550
+    },
+    {
+      "epoch": 0.2480864475461504,
+      "grad_norm": 0.09265519678592682,
+      "learning_rate": 0.0001785053224352324,
+      "loss": 0.2937,
+      "step": 551
+    },
+    {
+      "epoch": 0.2485366951823503,
+      "grad_norm": 0.1177322193980217,
+      "learning_rate": 0.00017841348758085225,
+      "loss": 0.5512,
+      "step": 552
+    },
+    {
+      "epoch": 0.24898694281855022,
+      "grad_norm": 0.15256206691265106,
+      "learning_rate": 0.0001783214806945272,
+      "loss": 0.5349,
+      "step": 553
+    },
+    {
+      "epoch": 0.2494371904547501,
+      "grad_norm": 0.09825648367404938,
+      "learning_rate": 0.00017822930197811189,
+      "loss": 0.6699,
+      "step": 554
+    },
+    {
+      "epoch": 0.24988743809095001,
+      "grad_norm": 0.10745353251695633,
+      "learning_rate": 0.0001781369516338378,
+      "loss": 0.4782,
+      "step": 555
+    },
+    {
+      "epoch": 0.25033768572714993,
+      "grad_norm": 0.08675112575292587,
+      "learning_rate": 0.00017804442986431318,
+      "loss": 0.455,
+      "step": 556
+    },
+    {
+      "epoch": 0.25078793336334987,
+      "grad_norm": 0.09504085034132004,
+      "learning_rate": 0.00017795173687252215,
+      "loss": 0.4956,
+      "step": 557
+    },
+    {
+      "epoch": 0.25123818099954975,
+      "grad_norm": 0.10465075075626373,
+      "learning_rate": 0.0001778588728618246,
+      "loss": 0.4874,
+      "step": 558
+    },
+    {
+      "epoch": 0.25168842863574964,
+      "grad_norm": 0.15467704832553864,
+      "learning_rate": 0.0001777658380359556,
+      "loss": 0.5053,
+      "step": 559
+    },
+    {
+      "epoch": 0.2521386762719496,
+      "grad_norm": 0.12720686197280884,
+      "learning_rate": 0.00017767263259902495,
+      "loss": 0.6615,
+      "step": 560
+    },
+    {
+      "epoch": 0.25258892390814947,
+      "grad_norm": 0.13382335007190704,
+      "learning_rate": 0.00017757925675551675,
+      "loss": 0.95,
+      "step": 561
+    },
+    {
+      "epoch": 0.2530391715443494,
+      "grad_norm": 0.13226279616355896,
+      "learning_rate": 0.000177485710710289,
+      "loss": 0.6256,
+      "step": 562
+    },
+    {
+      "epoch": 0.2534894191805493,
+      "grad_norm": 0.15509527921676636,
+      "learning_rate": 0.00017739199466857303,
+      "loss": 0.6895,
+      "step": 563
+    },
+    {
+      "epoch": 0.25393966681674923,
+      "grad_norm": 0.13538683950901031,
+      "learning_rate": 0.00017729810883597322,
+      "loss": 0.3785,
+      "step": 564
+    },
+    {
+      "epoch": 0.2543899144529491,
+      "grad_norm": 0.1653541624546051,
+      "learning_rate": 0.00017720405341846638,
+      "loss": 0.5272,
+      "step": 565
+    },
+    {
+      "epoch": 0.25484016208914906,
+      "grad_norm": 0.11467984318733215,
+      "learning_rate": 0.0001771098286224014,
+      "loss": 0.5432,
+      "step": 566
+    },
+    {
+      "epoch": 0.25529040972534894,
+      "grad_norm": 0.10456334799528122,
+      "learning_rate": 0.00017701543465449885,
+      "loss": 0.6011,
+      "step": 567
+    },
+    {
+      "epoch": 0.25574065736154883,
+      "grad_norm": 0.12927591800689697,
+      "learning_rate": 0.00017692087172185028,
+      "loss": 0.6474,
+      "step": 568
+    },
+    {
+      "epoch": 0.25619090499774877,
+      "grad_norm": 0.13111372292041779,
+      "learning_rate": 0.00017682614003191807,
+      "loss": 0.7342,
+      "step": 569
+    },
+    {
+      "epoch": 0.25664115263394865,
+      "grad_norm": 0.14993315935134888,
+      "learning_rate": 0.00017673123979253477,
+      "loss": 0.7094,
+      "step": 570
+    },
+    {
+      "epoch": 0.2570914002701486,
+      "grad_norm": 0.12955696880817413,
+      "learning_rate": 0.00017663617121190274,
+      "loss": 0.7034,
+      "step": 571
+    },
+    {
+      "epoch": 0.2575416479063485,
+      "grad_norm": 0.2662518322467804,
+      "learning_rate": 0.00017654093449859367,
+      "loss": 0.5198,
+      "step": 572
+    },
+    {
+      "epoch": 0.2579918955425484,
+      "grad_norm": 0.14272138476371765,
+      "learning_rate": 0.00017644552986154812,
+      "loss": 0.3671,
+      "step": 573
+    },
+    {
+      "epoch": 0.2584421431787483,
+      "grad_norm": 0.13986584544181824,
+      "learning_rate": 0.00017634995751007501,
+      "loss": 0.411,
+      "step": 574
+    },
+    {
+      "epoch": 0.25889239081494825,
+      "grad_norm": 0.1407669186592102,
+      "learning_rate": 0.00017625421765385125,
+      "loss": 0.7302,
+      "step": 575
+    },
+    {
+      "epoch": 0.25934263845114813,
+      "grad_norm": 0.1373807042837143,
+      "learning_rate": 0.0001761583105029213,
+      "loss": 0.6851,
+      "step": 576
+    },
+    {
+      "epoch": 0.259792886087348,
+      "grad_norm": 0.15483181178569794,
+      "learning_rate": 0.0001760622362676965,
+      "loss": 0.8242,
+      "step": 577
+    },
+    {
+      "epoch": 0.26024313372354796,
+      "grad_norm": 0.1679830104112625,
+      "learning_rate": 0.00017596599515895488,
+      "loss": 0.8496,
+      "step": 578
+    },
+    {
+      "epoch": 0.26069338135974784,
+      "grad_norm": 0.16874641180038452,
+      "learning_rate": 0.00017586958738784056,
+      "loss": 0.7768,
+      "step": 579
+    },
+    {
+      "epoch": 0.2611436289959478,
+      "grad_norm": 0.11139454692602158,
+      "learning_rate": 0.00017577301316586326,
+      "loss": 0.6863,
+      "step": 580
+    },
+    {
+      "epoch": 0.26159387663214767,
+      "grad_norm": 0.10413756966590881,
+      "learning_rate": 0.00017567627270489788,
+      "loss": 0.6664,
+      "step": 581
+    },
+    {
+      "epoch": 0.2620441242683476,
+      "grad_norm": 0.17609253525733948,
+      "learning_rate": 0.00017557936621718408,
+      "loss": 0.6043,
+      "step": 582
+    },
+    {
+      "epoch": 0.2624943719045475,
+      "grad_norm": 0.1377457231283188,
+      "learning_rate": 0.00017548229391532572,
+      "loss": 0.6071,
+      "step": 583
+    },
+    {
+      "epoch": 0.26294461954074744,
+      "grad_norm": 0.30970466136932373,
+      "learning_rate": 0.00017538505601229044,
+      "loss": 0.8397,
+      "step": 584
+    },
+    {
+      "epoch": 0.2633948671769473,
+      "grad_norm": 0.10679148137569427,
+      "learning_rate": 0.00017528765272140927,
+      "loss": 0.6117,
+      "step": 585
+    },
+    {
+      "epoch": 0.2638451148131472,
+      "grad_norm": 0.13156072795391083,
+      "learning_rate": 0.000175190084256376,
+      "loss": 0.3631,
+      "step": 586
+    },
+    {
+      "epoch": 0.26429536244934715,
+      "grad_norm": 0.08383766561746597,
+      "learning_rate": 0.0001750923508312468,
+      "loss": 0.385,
+      "step": 587
+    },
+    {
+      "epoch": 0.26474561008554703,
+      "grad_norm": 0.11846126616001129,
+      "learning_rate": 0.00017499445266043982,
+      "loss": 0.72,
+      "step": 588
+    },
+    {
+      "epoch": 0.265195857721747,
+      "grad_norm": 0.10805469006299973,
+      "learning_rate": 0.00017489638995873453,
+      "loss": 0.5234,
+      "step": 589
+    },
+    {
+      "epoch": 0.26564610535794686,
+      "grad_norm": 0.11644116789102554,
+      "learning_rate": 0.00017479816294127152,
+      "loss": 0.6428,
+      "step": 590
+    },
+    {
+      "epoch": 0.2660963529941468,
+      "grad_norm": 0.14633668959140778,
+      "learning_rate": 0.00017469977182355171,
+      "loss": 0.6604,
+      "step": 591
+    },
+    {
+      "epoch": 0.2665466006303467,
+      "grad_norm": 0.13522794842720032,
+      "learning_rate": 0.00017460121682143616,
+      "loss": 0.5411,
+      "step": 592
+    },
+    {
+      "epoch": 0.2669968482665466,
+      "grad_norm": 0.2391389161348343,
+      "learning_rate": 0.00017450249815114546,
+      "loss": 1.0195,
+      "step": 593
+    },
+    {
+      "epoch": 0.2674470959027465,
+      "grad_norm": 0.2023504376411438,
+      "learning_rate": 0.0001744036160292592,
+      "loss": 0.8917,
+      "step": 594
+    },
+    {
+      "epoch": 0.2678973435389464,
+      "grad_norm": 0.09261943399906158,
+      "learning_rate": 0.00017430457067271564,
+      "loss": 0.6363,
+      "step": 595
+    },
+    {
+      "epoch": 0.26834759117514634,
+      "grad_norm": 0.11652254313230515,
+      "learning_rate": 0.00017420536229881113,
+      "loss": 0.8312,
+      "step": 596
+    },
+    {
+      "epoch": 0.2687978388113462,
+      "grad_norm": 0.11802475899457932,
+      "learning_rate": 0.0001741059911251997,
+      "loss": 0.4705,
+      "step": 597
+    },
+    {
+      "epoch": 0.26924808644754616,
+      "grad_norm": 0.16600105166435242,
+      "learning_rate": 0.00017400645736989246,
+      "loss": 0.732,
+      "step": 598
+    },
+    {
+      "epoch": 0.26969833408374605,
+      "grad_norm": 0.1356017142534256,
+      "learning_rate": 0.00017390676125125735,
+      "loss": 0.7958,
+      "step": 599
+    },
+    {
+      "epoch": 0.270148581719946,
+      "grad_norm": 0.18415898084640503,
+      "learning_rate": 0.00017380690298801838,
+      "loss": 0.6729,
+      "step": 600
+    },
+    {
+      "epoch": 0.2705988293561459,
+      "grad_norm": 0.14587146043777466,
+      "learning_rate": 0.0001737068827992554,
+      "loss": 0.6265,
+      "step": 601
+    },
+    {
+      "epoch": 0.2710490769923458,
+      "grad_norm": 0.11254272609949112,
+      "learning_rate": 0.0001736067009044034,
+      "loss": 0.5362,
+      "step": 602
+    },
+    {
+      "epoch": 0.2714993246285457,
+      "grad_norm": 0.148624449968338,
+      "learning_rate": 0.00017350635752325224,
+      "loss": 0.4632,
+      "step": 603
+    },
+    {
+      "epoch": 0.2719495722647456,
+      "grad_norm": 0.1539754867553711,
+      "learning_rate": 0.00017340585287594604,
+      "loss": 0.7544,
+      "step": 604
+    },
+    {
+      "epoch": 0.27239981990094553,
+      "grad_norm": 0.1998700052499771,
+      "learning_rate": 0.00017330518718298264,
+      "loss": 0.8636,
+      "step": 605
+    },
+    {
+      "epoch": 0.2728500675371454,
+      "grad_norm": 0.08187783509492874,
+      "learning_rate": 0.00017320436066521334,
+      "loss": 0.5339,
+      "step": 606
+    },
+    {
+      "epoch": 0.27330031517334535,
+      "grad_norm": 0.1311655342578888,
+      "learning_rate": 0.00017310337354384215,
+      "loss": 0.4696,
+      "step": 607
+    },
+    {
+      "epoch": 0.27375056280954524,
+      "grad_norm": 0.12701299786567688,
+      "learning_rate": 0.00017300222604042553,
+      "loss": 0.5371,
+      "step": 608
+    },
+    {
+      "epoch": 0.2742008104457452,
+      "grad_norm": 0.12272265553474426,
+      "learning_rate": 0.00017290091837687173,
+      "loss": 0.4253,
+      "step": 609
+    },
+    {
+      "epoch": 0.27465105808194507,
+      "grad_norm": 0.12956619262695312,
+      "learning_rate": 0.00017279945077544037,
+      "loss": 0.7137,
+      "step": 610
+    },
+    {
+      "epoch": 0.275101305718145,
+      "grad_norm": 0.18266145884990692,
+      "learning_rate": 0.00017269782345874203,
+      "loss": 0.7314,
+      "step": 611
+    },
+    {
+      "epoch": 0.2755515533543449,
+      "grad_norm": 0.11342166364192963,
+      "learning_rate": 0.00017259603664973767,
+      "loss": 0.4994,
+      "step": 612
+    },
+    {
+      "epoch": 0.2760018009905448,
+      "grad_norm": 0.14500822126865387,
+      "learning_rate": 0.00017249409057173808,
+      "loss": 0.556,
+      "step": 613
+    },
+    {
+      "epoch": 0.2764520486267447,
+      "grad_norm": 0.12207160145044327,
+      "learning_rate": 0.00017239198544840355,
+      "loss": 0.6999,
+      "step": 614
+    },
+    {
+      "epoch": 0.2769022962629446,
+      "grad_norm": 0.13425995409488678,
+      "learning_rate": 0.00017228972150374334,
+      "loss": 0.6007,
+      "step": 615
+    },
+    {
+      "epoch": 0.27735254389914454,
+      "grad_norm": 0.10573825240135193,
+      "learning_rate": 0.00017218729896211504,
+      "loss": 0.499,
+      "step": 616
+    },
+    {
+      "epoch": 0.27780279153534443,
+      "grad_norm": 0.2626184821128845,
+      "learning_rate": 0.00017208471804822426,
+      "loss": 0.6642,
+      "step": 617
+    },
+    {
+      "epoch": 0.27825303917154437,
+      "grad_norm": 0.13032759726047516,
+      "learning_rate": 0.00017198197898712404,
+      "loss": 0.5887,
+      "step": 618
+    },
+    {
+      "epoch": 0.27870328680774425,
+      "grad_norm": 0.15816746652126312,
+      "learning_rate": 0.00017187908200421434,
+      "loss": 0.7826,
+      "step": 619
+    },
+    {
+      "epoch": 0.2791535344439442,
+      "grad_norm": 0.1325405091047287,
+      "learning_rate": 0.0001717760273252417,
+      "loss": 0.4861,
+      "step": 620
+    },
+    {
+      "epoch": 0.2796037820801441,
+      "grad_norm": 0.13173474371433258,
+      "learning_rate": 0.00017167281517629855,
+      "loss": 0.5396,
+      "step": 621
+    },
+    {
+      "epoch": 0.28005402971634397,
+      "grad_norm": 0.21644869446754456,
+      "learning_rate": 0.0001715694457838228,
+      "loss": 1.0289,
+      "step": 622
+    },
+    {
+      "epoch": 0.2805042773525439,
+      "grad_norm": 0.11475660651922226,
+      "learning_rate": 0.00017146591937459733,
+      "loss": 0.6039,
+      "step": 623
+    },
+    {
+      "epoch": 0.2809545249887438,
+      "grad_norm": 0.11032216250896454,
+      "learning_rate": 0.00017136223617574952,
+      "loss": 0.5801,
+      "step": 624
+    },
+    {
+      "epoch": 0.28140477262494373,
+      "grad_norm": 0.12464266270399094,
+      "learning_rate": 0.00017125839641475072,
+      "loss": 0.5861,
+      "step": 625
+    },
+    {
+      "epoch": 0.2818550202611436,
+      "grad_norm": 0.13577550649642944,
+      "learning_rate": 0.00017115440031941574,
+      "loss": 0.5484,
+      "step": 626
+    },
+    {
+      "epoch": 0.28230526789734356,
+      "grad_norm": 0.1740119904279709,
+      "learning_rate": 0.00017105024811790248,
+      "loss": 0.7445,
+      "step": 627
+    },
+    {
+      "epoch": 0.28275551553354344,
+      "grad_norm": 0.13359907269477844,
+      "learning_rate": 0.0001709459400387112,
+      "loss": 0.6523,
+      "step": 628
+    },
+    {
+      "epoch": 0.2832057631697434,
+      "grad_norm": 0.11157403886318207,
+      "learning_rate": 0.00017084147631068417,
+      "loss": 0.4318,
+      "step": 629
+    },
+    {
+      "epoch": 0.28365601080594327,
+      "grad_norm": 0.1431920975446701,
+      "learning_rate": 0.00017073685716300518,
+      "loss": 0.619,
+      "step": 630
+    },
+    {
+      "epoch": 0.28410625844214316,
+      "grad_norm": 0.16049373149871826,
+      "learning_rate": 0.00017063208282519894,
+      "loss": 0.7827,
+      "step": 631
+    },
+    {
+      "epoch": 0.2845565060783431,
+      "grad_norm": 0.16427528858184814,
+      "learning_rate": 0.00017052715352713075,
+      "loss": 0.5588,
+      "step": 632
+    },
+    {
+      "epoch": 0.285006753714543,
+      "grad_norm": 0.11942577362060547,
+      "learning_rate": 0.0001704220694990057,
+      "loss": 0.4881,
+      "step": 633
+    },
+    {
+      "epoch": 0.2854570013507429,
+      "grad_norm": 0.1363128125667572,
+      "learning_rate": 0.00017031683097136843,
+      "loss": 0.4988,
+      "step": 634
+    },
+    {
+      "epoch": 0.2859072489869428,
+      "grad_norm": 0.09995927661657333,
+      "learning_rate": 0.00017021143817510262,
+      "loss": 0.4271,
+      "step": 635
+    },
+    {
+      "epoch": 0.28635749662314275,
+      "grad_norm": 0.1725994050502777,
+      "learning_rate": 0.00017010589134143026,
+      "loss": 0.665,
+      "step": 636
+    },
+    {
+      "epoch": 0.28680774425934263,
+      "grad_norm": 0.11268609017133713,
+      "learning_rate": 0.0001700001907019114,
+      "loss": 0.4814,
+      "step": 637
+    },
+    {
+      "epoch": 0.2872579918955426,
+      "grad_norm": 0.1391492486000061,
+      "learning_rate": 0.00016989433648844342,
+      "loss": 0.4747,
+      "step": 638
+    },
+    {
+      "epoch": 0.28770823953174246,
+      "grad_norm": 0.11034171283245087,
+      "learning_rate": 0.00016978832893326074,
+      "loss": 0.6336,
+      "step": 639
+    },
+    {
+      "epoch": 0.28815848716794235,
+      "grad_norm": 0.18257318437099457,
+      "learning_rate": 0.00016968216826893407,
+      "loss": 0.5116,
+      "step": 640
+    },
+    {
+      "epoch": 0.2886087348041423,
+      "grad_norm": 0.1501189023256302,
+      "learning_rate": 0.00016957585472837015,
+      "loss": 0.5854,
+      "step": 641
+    },
+    {
+      "epoch": 0.28905898244034217,
+      "grad_norm": 0.15408477187156677,
+      "learning_rate": 0.00016946938854481104,
+      "loss": 0.4866,
+      "step": 642
+    },
+    {
+      "epoch": 0.2895092300765421,
+      "grad_norm": 0.13506793975830078,
+      "learning_rate": 0.00016936276995183373,
+      "loss": 0.5125,
+      "step": 643
+    },
+    {
+      "epoch": 0.289959477712742,
+      "grad_norm": 0.1239505186676979,
+      "learning_rate": 0.00016925599918334955,
+      "loss": 0.5272,
+      "step": 644
+    },
+    {
+      "epoch": 0.29040972534894194,
+      "grad_norm": 0.11671876162290573,
+      "learning_rate": 0.00016914907647360368,
+      "loss": 0.4362,
+      "step": 645
+    },
+    {
+      "epoch": 0.2908599729851418,
+      "grad_norm": 0.11398175358772278,
+      "learning_rate": 0.0001690420020571747,
+      "loss": 0.4647,
+      "step": 646
+    },
+    {
+      "epoch": 0.29131022062134176,
+      "grad_norm": 0.12950275838375092,
+      "learning_rate": 0.0001689347761689739,
+      "loss": 0.793,
+      "step": 647
+    },
+    {
+      "epoch": 0.29176046825754165,
+      "grad_norm": 0.23010940849781036,
+      "learning_rate": 0.0001688273990442451,
+      "loss": 1.0406,
+      "step": 648
+    },
+    {
+      "epoch": 0.29221071589374153,
+      "grad_norm": 0.11109314858913422,
+      "learning_rate": 0.00016871987091856368,
+      "loss": 0.5695,
+      "step": 649
+    },
+    {
+      "epoch": 0.2926609635299415,
+      "grad_norm": 0.13457128405570984,
+      "learning_rate": 0.00016861219202783645,
+      "loss": 0.6236,
+      "step": 650
+    },
+    {
+      "epoch": 0.29311121116614136,
+      "grad_norm": 0.11238634586334229,
+      "learning_rate": 0.00016850436260830093,
+      "loss": 0.4599,
+      "step": 651
+    },
+    {
+      "epoch": 0.2935614588023413,
+      "grad_norm": 0.19172637164592743,
+      "learning_rate": 0.00016839638289652492,
+      "loss": 0.7568,
+      "step": 652
+    },
+    {
+      "epoch": 0.2940117064385412,
+      "grad_norm": 0.0784100815653801,
+      "learning_rate": 0.00016828825312940592,
+      "loss": 0.3631,
+      "step": 653
+    },
+    {
+      "epoch": 0.29446195407474113,
+      "grad_norm": 0.15234927833080292,
+      "learning_rate": 0.00016817997354417067,
+      "loss": 0.8102,
+      "step": 654
+    },
+    {
+      "epoch": 0.294912201710941,
+      "grad_norm": 0.15923357009887695,
+      "learning_rate": 0.00016807154437837454,
+      "loss": 0.4755,
+      "step": 655
+    },
+    {
+      "epoch": 0.29536244934714095,
+      "grad_norm": 0.10125742852687836,
+      "learning_rate": 0.0001679629658699011,
+      "loss": 0.4429,
+      "step": 656
+    },
+    {
+      "epoch": 0.29581269698334084,
+      "grad_norm": 0.09144002944231033,
+      "learning_rate": 0.00016785423825696157,
+      "loss": 0.412,
+      "step": 657
+    },
+    {
+      "epoch": 0.2962629446195407,
+      "grad_norm": 0.10798758268356323,
+      "learning_rate": 0.00016774536177809429,
+      "loss": 0.6716,
+      "step": 658
+    },
+    {
+      "epoch": 0.29671319225574067,
+      "grad_norm": 0.11133555322885513,
+      "learning_rate": 0.0001676363366721642,
+      "loss": 0.6726,
+      "step": 659
+    },
+    {
+      "epoch": 0.29716343989194055,
+      "grad_norm": 0.10137926042079926,
+      "learning_rate": 0.00016752716317836229,
+      "loss": 0.7359,
+      "step": 660
+    },
+    {
+      "epoch": 0.2976136875281405,
+      "grad_norm": 0.14925476908683777,
+      "learning_rate": 0.0001674178415362051,
+      "loss": 0.5926,
+      "step": 661
+    },
+    {
+      "epoch": 0.2980639351643404,
+      "grad_norm": 0.12005611509084702,
+      "learning_rate": 0.00016730837198553424,
+      "loss": 0.6263,
+      "step": 662
+    },
+    {
+      "epoch": 0.2985141828005403,
+      "grad_norm": 0.14329755306243896,
+      "learning_rate": 0.00016719875476651577,
+      "loss": 0.7062,
+      "step": 663
+    },
+    {
+      "epoch": 0.2989644304367402,
+      "grad_norm": 0.10556526482105255,
+      "learning_rate": 0.00016708899011963978,
+      "loss": 0.5423,
+      "step": 664
+    },
+    {
+      "epoch": 0.29941467807294014,
+      "grad_norm": 0.144564688205719,
+      "learning_rate": 0.0001669790782857197,
+      "loss": 0.7399,
+      "step": 665
+    },
+    {
+      "epoch": 0.29986492570914003,
+      "grad_norm": 0.10654689371585846,
+      "learning_rate": 0.00016686901950589196,
+      "loss": 0.5196,
+      "step": 666
+    },
+    {
+      "epoch": 0.3003151733453399,
+      "grad_norm": 0.20161013305187225,
+      "learning_rate": 0.00016675881402161536,
+      "loss": 0.5729,
+      "step": 667
+    },
+    {
+      "epoch": 0.30076542098153985,
+      "grad_norm": 0.08455012738704681,
+      "learning_rate": 0.00016664846207467055,
+      "loss": 0.4474,
+      "step": 668
+    },
+    {
+      "epoch": 0.30121566861773974,
+      "grad_norm": 0.11929601430892944,
+      "learning_rate": 0.0001665379639071595,
+      "loss": 0.4378,
+      "step": 669
+    },
+    {
+      "epoch": 0.3016659162539397,
+      "grad_norm": 0.10755518823862076,
+      "learning_rate": 0.00016642731976150495,
+      "loss": 0.3609,
+      "step": 670
+    },
+    {
+      "epoch": 0.30211616389013957,
+      "grad_norm": 0.15590333938598633,
+      "learning_rate": 0.00016631652988044997,
+      "loss": 0.7459,
+      "step": 671
+    },
+    {
+      "epoch": 0.3025664115263395,
+      "grad_norm": 0.13067883253097534,
+      "learning_rate": 0.00016620559450705727,
+      "loss": 0.5964,
+      "step": 672
+    },
+    {
+      "epoch": 0.3030166591625394,
+      "grad_norm": 0.1570952981710434,
+      "learning_rate": 0.00016609451388470886,
+      "loss": 0.7302,
+      "step": 673
+    },
+    {
+      "epoch": 0.30346690679873933,
+      "grad_norm": 0.13876138627529144,
+      "learning_rate": 0.00016598328825710533,
+      "loss": 0.7263,
+      "step": 674
+    },
+    {
+      "epoch": 0.3039171544349392,
+      "grad_norm": 0.11717363446950912,
+      "learning_rate": 0.00016587191786826547,
+      "loss": 0.5958,
+      "step": 675
+    },
+    {
+      "epoch": 0.3043674020711391,
+      "grad_norm": 0.17613142728805542,
+      "learning_rate": 0.00016576040296252556,
+      "loss": 0.4596,
+      "step": 676
+    },
+    {
+      "epoch": 0.30481764970733904,
+      "grad_norm": 0.11680793017148972,
+      "learning_rate": 0.00016564874378453902,
+      "loss": 0.7218,
+      "step": 677
+    },
+    {
+      "epoch": 0.30526789734353893,
+      "grad_norm": 0.1843346506357193,
+      "learning_rate": 0.00016553694057927577,
+      "loss": 0.5575,
+      "step": 678
+    },
+    {
+      "epoch": 0.30571814497973887,
+      "grad_norm": 0.10092949867248535,
+      "learning_rate": 0.0001654249935920217,
+      "loss": 0.3303,
+      "step": 679
+    },
+    {
+      "epoch": 0.30616839261593876,
+      "grad_norm": 0.10840975493192673,
+      "learning_rate": 0.0001653129030683782,
+      "loss": 0.5746,
+      "step": 680
+    },
+    {
+      "epoch": 0.3066186402521387,
+      "grad_norm": 0.1275509148836136,
+      "learning_rate": 0.00016520066925426144,
+      "loss": 0.5451,
+      "step": 681
+    },
+    {
+      "epoch": 0.3070688878883386,
+      "grad_norm": 0.10627155005931854,
+      "learning_rate": 0.00016508829239590212,
+      "loss": 0.8482,
+      "step": 682
+    },
+    {
+      "epoch": 0.3075191355245385,
+      "grad_norm": 0.12347329407930374,
+      "learning_rate": 0.00016497577273984462,
+      "loss": 0.7731,
+      "step": 683
+    },
+    {
+      "epoch": 0.3079693831607384,
+      "grad_norm": 0.09412003308534622,
+      "learning_rate": 0.00016486311053294668,
+      "loss": 0.4788,
+      "step": 684
+    },
+    {
+      "epoch": 0.3084196307969383,
+      "grad_norm": 0.1440611183643341,
+      "learning_rate": 0.00016475030602237876,
+      "loss": 0.4363,
+      "step": 685
+    },
+    {
+      "epoch": 0.30886987843313823,
+      "grad_norm": 0.19373439252376556,
+      "learning_rate": 0.0001646373594556236,
+      "loss": 0.6485,
+      "step": 686
+    },
+    {
+      "epoch": 0.3093201260693381,
+      "grad_norm": 0.12926699221134186,
+      "learning_rate": 0.00016452427108047544,
+      "loss": 0.6408,
+      "step": 687
+    },
+    {
+      "epoch": 0.30977037370553806,
+      "grad_norm": 0.13196153938770294,
+      "learning_rate": 0.0001644110411450398,
+      "loss": 0.5908,
+      "step": 688
+    },
+    {
+      "epoch": 0.31022062134173795,
+      "grad_norm": 0.0886620581150055,
+      "learning_rate": 0.0001642976698977326,
+      "loss": 0.4056,
+      "step": 689
+    },
+    {
+      "epoch": 0.3106708689779379,
+      "grad_norm": 0.13059855997562408,
+      "learning_rate": 0.00016418415758727998,
+      "loss": 0.6418,
+      "step": 690
+    },
+    {
+      "epoch": 0.31112111661413777,
+      "grad_norm": 0.10823266953229904,
+      "learning_rate": 0.0001640705044627174,
+      "loss": 0.3523,
+      "step": 691
+    },
+    {
+      "epoch": 0.3115713642503377,
+      "grad_norm": 0.15798276662826538,
+      "learning_rate": 0.0001639567107733893,
+      "loss": 0.7451,
+      "step": 692
+    },
+    {
+      "epoch": 0.3120216118865376,
+      "grad_norm": 0.2084863781929016,
+      "learning_rate": 0.00016384277676894857,
+      "loss": 0.4863,
+      "step": 693
+    },
+    {
+      "epoch": 0.3124718595227375,
+      "grad_norm": 0.11761056631803513,
+      "learning_rate": 0.00016372870269935584,
+      "loss": 0.5848,
+      "step": 694
+    },
+    {
+      "epoch": 0.3129221071589374,
+      "grad_norm": 0.11858902126550674,
+      "learning_rate": 0.00016361448881487914,
+      "loss": 0.502,
+      "step": 695
+    },
+    {
+      "epoch": 0.3133723547951373,
+      "grad_norm": 0.10026060789823532,
+      "learning_rate": 0.00016350013536609308,
+      "loss": 0.5716,
+      "step": 696
+    },
+    {
+      "epoch": 0.31382260243133725,
+      "grad_norm": 0.12254612147808075,
+      "learning_rate": 0.00016338564260387863,
+      "loss": 0.7918,
+      "step": 697
+    },
+    {
+      "epoch": 0.31427285006753714,
+      "grad_norm": 0.10671459138393402,
+      "learning_rate": 0.0001632710107794223,
+      "loss": 0.2939,
+      "step": 698
+    },
+    {
+      "epoch": 0.3147230977037371,
+      "grad_norm": 0.11760569363832474,
+      "learning_rate": 0.0001631562401442157,
+      "loss": 0.5094,
+      "step": 699
+    },
+    {
+      "epoch": 0.31517334533993696,
+      "grad_norm": 0.1579584926366806,
+      "learning_rate": 0.00016304133095005507,
+      "loss": 0.5976,
+      "step": 700
+    },
+    {
+      "epoch": 0.3156235929761369,
+      "grad_norm": 0.09953977167606354,
+      "learning_rate": 0.0001629262834490405,
+      "loss": 0.3675,
+      "step": 701
+    },
+    {
+      "epoch": 0.3160738406123368,
+      "grad_norm": 0.13274246454238892,
+      "learning_rate": 0.0001628110978935756,
+      "loss": 0.3452,
+      "step": 702
+    },
+    {
+      "epoch": 0.3165240882485367,
+      "grad_norm": 0.0844864770770073,
+      "learning_rate": 0.00016269577453636682,
+      "loss": 0.4606,
+      "step": 703
+    },
+    {
+      "epoch": 0.3169743358847366,
+      "grad_norm": 0.12627467513084412,
+      "learning_rate": 0.0001625803136304229,
+      "loss": 0.4089,
+      "step": 704
+    },
+    {
+      "epoch": 0.3174245835209365,
+      "grad_norm": 0.2038150429725647,
+      "learning_rate": 0.00016246471542905454,
+      "loss": 0.8118,
+      "step": 705
+    },
+    {
+      "epoch": 0.31787483115713644,
+      "grad_norm": 0.09169654548168182,
+      "learning_rate": 0.00016234898018587337,
+      "loss": 0.3385,
+      "step": 706
+    },
+    {
+      "epoch": 0.3183250787933363,
+      "grad_norm": 0.11134479939937592,
+      "learning_rate": 0.00016223310815479186,
+      "loss": 0.5299,
+      "step": 707
+    },
+    {
+      "epoch": 0.31877532642953627,
+      "grad_norm": 0.10228997468948364,
+      "learning_rate": 0.00016211709959002257,
+      "loss": 0.5512,
+      "step": 708
+    },
+    {
+      "epoch": 0.31922557406573615,
+      "grad_norm": 0.10541591048240662,
+      "learning_rate": 0.00016200095474607753,
+      "loss": 0.4205,
+      "step": 709
+    },
+    {
+      "epoch": 0.3196758217019361,
+      "grad_norm": 0.1062127873301506,
+      "learning_rate": 0.0001618846738777678,
+      "loss": 0.4366,
+      "step": 710
+    },
+    {
+      "epoch": 0.320126069338136,
+      "grad_norm": 0.1272103488445282,
+      "learning_rate": 0.0001617682572402029,
+      "loss": 0.6047,
+      "step": 711
+    },
+    {
+      "epoch": 0.32057631697433586,
+      "grad_norm": 0.13013339042663574,
+      "learning_rate": 0.00016165170508879008,
+      "loss": 0.607,
+      "step": 712
+    },
+    {
+      "epoch": 0.3210265646105358,
+      "grad_norm": 0.1604539155960083,
+      "learning_rate": 0.00016153501767923408,
+      "loss": 0.49,
+      "step": 713
+    },
+    {
+      "epoch": 0.3214768122467357,
+      "grad_norm": 0.14530783891677856,
+      "learning_rate": 0.00016141819526753628,
+      "loss": 0.6825,
+      "step": 714
+    },
+    {
+      "epoch": 0.32192705988293563,
+      "grad_norm": 0.12872616946697235,
+      "learning_rate": 0.0001613012381099942,
+      "loss": 0.7003,
+      "step": 715
+    },
+    {
+      "epoch": 0.3223773075191355,
+      "grad_norm": 0.15016387403011322,
+      "learning_rate": 0.0001611841464632011,
+      "loss": 0.7731,
+      "step": 716
+    },
+    {
+      "epoch": 0.32282755515533545,
+      "grad_norm": 0.15742337703704834,
+      "learning_rate": 0.0001610669205840452,
+      "loss": 0.7139,
+      "step": 717
+    },
+    {
+      "epoch": 0.32327780279153534,
+      "grad_norm": 0.20721325278282166,
+      "learning_rate": 0.00016094956072970927,
+      "loss": 0.7004,
+      "step": 718
+    },
+    {
+      "epoch": 0.3237280504277353,
+      "grad_norm": 0.17191503942012787,
+      "learning_rate": 0.0001608320671576699,
+      "loss": 0.3094,
+      "step": 719
+    },
+    {
+      "epoch": 0.32417829806393517,
+      "grad_norm": 0.1103680282831192,
+      "learning_rate": 0.00016071444012569726,
+      "loss": 0.7272,
+      "step": 720
+    },
+    {
+      "epoch": 0.32462854570013505,
+      "grad_norm": 0.1476084440946579,
+      "learning_rate": 0.00016059667989185407,
+      "loss": 0.6619,
+      "step": 721
+    },
+    {
+      "epoch": 0.325078793336335,
+      "grad_norm": 0.13148553669452667,
+      "learning_rate": 0.00016047878671449545,
+      "loss": 0.4449,
+      "step": 722
+    },
+    {
+      "epoch": 0.3255290409725349,
+      "grad_norm": 0.16558244824409485,
+      "learning_rate": 0.00016036076085226814,
+      "loss": 0.7282,
+      "step": 723
+    },
+    {
+      "epoch": 0.3259792886087348,
+      "grad_norm": 0.13639886677265167,
+      "learning_rate": 0.00016024260256410997,
+      "loss": 0.6096,
+      "step": 724
+    },
+    {
+      "epoch": 0.3264295362449347,
+      "grad_norm": 0.13836698234081268,
+      "learning_rate": 0.00016012431210924932,
+      "loss": 0.7583,
+      "step": 725
+    },
+    {
+      "epoch": 0.32687978388113464,
+      "grad_norm": 0.0774713084101677,
+      "learning_rate": 0.00016000588974720444,
+      "loss": 0.4317,
+      "step": 726
+    },
+    {
+      "epoch": 0.32733003151733453,
+      "grad_norm": 0.1188255026936531,
+      "learning_rate": 0.00015988733573778317,
+      "loss": 0.5266,
+      "step": 727
+    },
+    {
+      "epoch": 0.32778027915353447,
+      "grad_norm": 0.12945453822612762,
+      "learning_rate": 0.00015976865034108192,
+      "loss": 0.6844,
+      "step": 728
+    },
+    {
+      "epoch": 0.32823052678973436,
+      "grad_norm": 0.13470758497714996,
+      "learning_rate": 0.0001596498338174856,
+      "loss": 0.3834,
+      "step": 729
+    },
+    {
+      "epoch": 0.32868077442593424,
+      "grad_norm": 0.11171048879623413,
+      "learning_rate": 0.0001595308864276666,
+      "loss": 0.4864,
+      "step": 730
+    },
+    {
+      "epoch": 0.3291310220621342,
+      "grad_norm": 0.09181814640760422,
+      "learning_rate": 0.00015941180843258453,
+      "loss": 0.371,
+      "step": 731
+    },
+    {
+      "epoch": 0.32958126969833407,
+      "grad_norm": 0.14527185261249542,
+      "learning_rate": 0.00015929260009348552,
+      "loss": 0.4827,
+      "step": 732
+    },
+    {
+      "epoch": 0.330031517334534,
+      "grad_norm": 0.1399325281381607,
+      "learning_rate": 0.00015917326167190165,
+      "loss": 0.5008,
+      "step": 733
+    },
+    {
+      "epoch": 0.3304817649707339,
+      "grad_norm": 0.11961071938276291,
+      "learning_rate": 0.00015905379342965033,
+      "loss": 0.4737,
+      "step": 734
+    },
+    {
+      "epoch": 0.33093201260693383,
+      "grad_norm": 0.14174236357212067,
+      "learning_rate": 0.0001589341956288339,
+      "loss": 0.4565,
+      "step": 735
+    },
+    {
+      "epoch": 0.3313822602431337,
+      "grad_norm": 0.10174470394849777,
+      "learning_rate": 0.0001588144685318389,
+      "loss": 0.5517,
+      "step": 736
+    },
+    {
+      "epoch": 0.33183250787933366,
+      "grad_norm": 0.201952263712883,
+      "learning_rate": 0.0001586946124013354,
+      "loss": 0.6922,
+      "step": 737
+    },
+    {
+      "epoch": 0.33228275551553355,
+      "grad_norm": 0.09435506910085678,
+      "learning_rate": 0.00015857462750027684,
+      "loss": 0.4666,
+      "step": 738
+    },
+    {
+      "epoch": 0.33273300315173343,
+      "grad_norm": 0.1485670655965805,
+      "learning_rate": 0.0001584545140918989,
+      "loss": 0.7306,
+      "step": 739
+    },
+    {
+      "epoch": 0.33318325078793337,
+      "grad_norm": 0.11810648441314697,
+      "learning_rate": 0.00015833427243971927,
+      "loss": 0.5796,
+      "step": 740
+    },
+    {
+      "epoch": 0.33363349842413326,
+      "grad_norm": 0.13161268830299377,
+      "learning_rate": 0.00015821390280753713,
+      "loss": 0.6046,
+      "step": 741
+    },
+    {
+      "epoch": 0.3340837460603332,
+      "grad_norm": 0.14327849447727203,
+      "learning_rate": 0.00015809340545943222,
+      "loss": 0.8739,
+      "step": 742
+    },
+    {
+      "epoch": 0.3345339936965331,
+      "grad_norm": 0.11349102109670639,
+      "learning_rate": 0.00015797278065976465,
+      "loss": 0.5193,
+      "step": 743
+    },
+    {
+      "epoch": 0.334984241332733,
+      "grad_norm": 0.1669684499502182,
+      "learning_rate": 0.00015785202867317407,
+      "loss": 0.579,
+      "step": 744
+    },
+    {
+      "epoch": 0.3354344889689329,
+      "grad_norm": 0.12111262232065201,
+      "learning_rate": 0.00015773114976457917,
+      "loss": 0.5242,
+      "step": 745
+    },
+    {
+      "epoch": 0.33588473660513285,
+      "grad_norm": 0.199104443192482,
+      "learning_rate": 0.00015761014419917713,
+      "loss": 0.7214,
+      "step": 746
+    },
+    {
+      "epoch": 0.33633498424133274,
+      "grad_norm": 0.13180191814899445,
+      "learning_rate": 0.00015748901224244292,
+      "loss": 0.8565,
+      "step": 747
+    },
+    {
+      "epoch": 0.3367852318775326,
+      "grad_norm": 0.15166360139846802,
+      "learning_rate": 0.00015736775416012893,
+      "loss": 0.6953,
+      "step": 748
+    },
+    {
+      "epoch": 0.33723547951373256,
+      "grad_norm": 0.10592274367809296,
+      "learning_rate": 0.0001572463702182641,
+      "loss": 0.4042,
+      "step": 749
+    },
+    {
+      "epoch": 0.33768572714993245,
+      "grad_norm": 0.10570541769266129,
+      "learning_rate": 0.00015712486068315368,
+      "loss": 0.3681,
+      "step": 750
+    },
+    {
+      "epoch": 0.3381359747861324,
+      "grad_norm": 0.12998966872692108,
+      "learning_rate": 0.00015700322582137827,
+      "loss": 0.6283,
+      "step": 751
+    },
+    {
+      "epoch": 0.3385862224223323,
+      "grad_norm": 0.1482112556695938,
+      "learning_rate": 0.0001568814658997936,
+      "loss": 0.9408,
+      "step": 752
+    },
+    {
+      "epoch": 0.3390364700585322,
+      "grad_norm": 0.11294540017843246,
+      "learning_rate": 0.00015675958118552963,
+      "loss": 0.3744,
+      "step": 753
+    },
+    {
+      "epoch": 0.3394867176947321,
+      "grad_norm": 0.1013224646449089,
+      "learning_rate": 0.00015663757194599013,
+      "loss": 0.5521,
+      "step": 754
+    },
+    {
+      "epoch": 0.33993696533093204,
+      "grad_norm": 0.13891354203224182,
+      "learning_rate": 0.00015651543844885218,
+      "loss": 0.8211,
+      "step": 755
+    },
+    {
+      "epoch": 0.3403872129671319,
+      "grad_norm": 0.16383050382137299,
+      "learning_rate": 0.00015639318096206535,
+      "loss": 0.6412,
+      "step": 756
+    },
+    {
+      "epoch": 0.3408374606033318,
+      "grad_norm": 0.11336307227611542,
+      "learning_rate": 0.00015627079975385122,
+      "loss": 0.4068,
+      "step": 757
+    },
+    {
+      "epoch": 0.34128770823953175,
+      "grad_norm": 0.11039210110902786,
+      "learning_rate": 0.0001561482950927029,
+      "loss": 0.5694,
+      "step": 758
+    },
+    {
+      "epoch": 0.34173795587573164,
+      "grad_norm": 0.117017962038517,
+      "learning_rate": 0.00015602566724738428,
+      "loss": 0.9038,
+      "step": 759
+    },
+    {
+      "epoch": 0.3421882035119316,
+      "grad_norm": 0.14746886491775513,
+      "learning_rate": 0.00015590291648692953,
+      "loss": 0.5036,
+      "step": 760
+    },
+    {
+      "epoch": 0.34263845114813146,
+      "grad_norm": 0.1578681915998459,
+      "learning_rate": 0.00015578004308064246,
+      "loss": 0.5716,
+      "step": 761
+    },
+    {
+      "epoch": 0.3430886987843314,
+      "grad_norm": 0.14602695405483246,
+      "learning_rate": 0.000155657047298096,
+      "loss": 0.6038,
+      "step": 762
+    },
+    {
+      "epoch": 0.3435389464205313,
+      "grad_norm": 0.11272371560335159,
+      "learning_rate": 0.0001555339294091315,
+      "loss": 0.5282,
+      "step": 763
+    },
+    {
+      "epoch": 0.34398919405673123,
+      "grad_norm": 0.11612574756145477,
+      "learning_rate": 0.0001554106896838582,
+      "loss": 0.571,
+      "step": 764
+    },
+    {
+      "epoch": 0.3444394416929311,
+      "grad_norm": 0.101449154317379,
+      "learning_rate": 0.00015528732839265272,
+      "loss": 0.4935,
+      "step": 765
+    },
+    {
+      "epoch": 0.344889689329131,
+      "grad_norm": 0.12047034502029419,
+      "learning_rate": 0.00015516384580615834,
+      "loss": 0.6796,
+      "step": 766
+    },
+    {
+      "epoch": 0.34533993696533094,
+      "grad_norm": 0.1183471530675888,
+      "learning_rate": 0.0001550402421952844,
+      "loss": 0.4991,
+      "step": 767
+    },
+    {
+      "epoch": 0.3457901846015308,
+      "grad_norm": 0.13030461966991425,
+      "learning_rate": 0.0001549165178312058,
+      "loss": 0.5976,
+      "step": 768
+    },
+    {
+      "epoch": 0.34624043223773077,
+      "grad_norm": 0.11863750964403152,
+      "learning_rate": 0.0001547926729853624,
+      "loss": 0.7309,
+      "step": 769
+    },
+    {
+      "epoch": 0.34669067987393065,
+      "grad_norm": 0.11039796471595764,
+      "learning_rate": 0.00015466870792945828,
+      "loss": 0.4566,
+      "step": 770
+    },
+    {
+      "epoch": 0.3471409275101306,
+      "grad_norm": 0.15371926128864288,
+      "learning_rate": 0.0001545446229354614,
+      "loss": 0.7352,
+      "step": 771
+    },
+    {
+      "epoch": 0.3475911751463305,
+      "grad_norm": 0.11060057580471039,
+      "learning_rate": 0.00015442041827560274,
+      "loss": 0.6647,
+      "step": 772
+    },
+    {
+      "epoch": 0.3480414227825304,
+      "grad_norm": 0.09247463941574097,
+      "learning_rate": 0.0001542960942223758,
+      "loss": 0.4154,
+      "step": 773
+    },
+    {
+      "epoch": 0.3484916704187303,
+      "grad_norm": 0.2172059416770935,
+      "learning_rate": 0.00015417165104853614,
+      "loss": 0.5676,
+      "step": 774
+    },
+    {
+      "epoch": 0.3489419180549302,
+      "grad_norm": 0.10758895426988602,
+      "learning_rate": 0.0001540470890271005,
+      "loss": 0.3691,
+      "step": 775
+    },
+    {
+      "epoch": 0.34939216569113013,
+      "grad_norm": 0.09961098432540894,
+      "learning_rate": 0.0001539224084313465,
+      "loss": 0.3453,
+      "step": 776
+    },
+    {
+      "epoch": 0.34984241332733,
+      "grad_norm": 0.15455502271652222,
+      "learning_rate": 0.0001537976095348118,
+      "loss": 0.7118,
+      "step": 777
+    },
+    {
+      "epoch": 0.35029266096352996,
+      "grad_norm": 0.1055678129196167,
+      "learning_rate": 0.0001536726926112937,
+      "loss": 0.3567,
+      "step": 778
+    },
+    {
+      "epoch": 0.35074290859972984,
+      "grad_norm": 0.16685661673545837,
+      "learning_rate": 0.00015354765793484834,
+      "loss": 0.4576,
+      "step": 779
+    },
+    {
+      "epoch": 0.3511931562359298,
+      "grad_norm": 0.10810405015945435,
+      "learning_rate": 0.00015342250577979024,
+      "loss": 0.5218,
+      "step": 780
+    },
+    {
+      "epoch": 0.35164340387212967,
+      "grad_norm": 0.14653025567531586,
+      "learning_rate": 0.00015329723642069172,
+      "loss": 0.4442,
+      "step": 781
+    },
+    {
+      "epoch": 0.3520936515083296,
+      "grad_norm": 0.10543451458215714,
+      "learning_rate": 0.00015317185013238212,
+      "loss": 0.6262,
+      "step": 782
+    },
+    {
+      "epoch": 0.3525438991445295,
+      "grad_norm": 0.12601809203624725,
+      "learning_rate": 0.0001530463471899474,
+      "loss": 0.4633,
+      "step": 783
+    },
+    {
+      "epoch": 0.3529941467807294,
+      "grad_norm": 0.08294503390789032,
+      "learning_rate": 0.0001529207278687294,
+      "loss": 0.4173,
+      "step": 784
+    },
+    {
+      "epoch": 0.3534443944169293,
+      "grad_norm": 0.14413581788539886,
+      "learning_rate": 0.0001527949924443253,
+      "loss": 0.4392,
+      "step": 785
+    },
+    {
+      "epoch": 0.3538946420531292,
+      "grad_norm": 0.13317470252513885,
+      "learning_rate": 0.000152669141192587,
+      "loss": 0.3581,
+      "step": 786
+    },
+    {
+      "epoch": 0.35434488968932915,
+      "grad_norm": 0.085340715944767,
+      "learning_rate": 0.00015254317438962055,
+      "loss": 0.3489,
+      "step": 787
+    },
+    {
+      "epoch": 0.35479513732552903,
+      "grad_norm": 0.10696078091859818,
+      "learning_rate": 0.0001524170923117854,
+      "loss": 0.5384,
+      "step": 788
+    },
+    {
+      "epoch": 0.35524538496172897,
+      "grad_norm": 0.22438734769821167,
+      "learning_rate": 0.00015229089523569406,
+      "loss": 1.3758,
+      "step": 789
+    },
+    {
+      "epoch": 0.35569563259792886,
+      "grad_norm": 0.10124437510967255,
+      "learning_rate": 0.00015216458343821122,
+      "loss": 0.6308,
+      "step": 790
+    },
+    {
+      "epoch": 0.3561458802341288,
+      "grad_norm": 0.12516243755817413,
+      "learning_rate": 0.0001520381571964533,
+      "loss": 0.5511,
+      "step": 791
+    },
+    {
+      "epoch": 0.3565961278703287,
+      "grad_norm": 0.10918477922677994,
+      "learning_rate": 0.00015191161678778774,
+      "loss": 0.7287,
+      "step": 792
+    },
+    {
+      "epoch": 0.35704637550652857,
+      "grad_norm": 0.12352471053600311,
+      "learning_rate": 0.00015178496248983254,
+      "loss": 0.7522,
+      "step": 793
+    },
+    {
+      "epoch": 0.3574966231427285,
+      "grad_norm": 0.10196998715400696,
+      "learning_rate": 0.00015165819458045555,
+      "loss": 0.6178,
+      "step": 794
+    },
+    {
+      "epoch": 0.3579468707789284,
+      "grad_norm": 0.13494299352169037,
+      "learning_rate": 0.0001515313133377738,
+      "loss": 0.6626,
+      "step": 795
+    },
+    {
+      "epoch": 0.35839711841512834,
+      "grad_norm": 0.115663543343544,
+      "learning_rate": 0.00015140431904015302,
+      "loss": 0.371,
+      "step": 796
+    },
+    {
+      "epoch": 0.3588473660513282,
+      "grad_norm": 0.11560159921646118,
+      "learning_rate": 0.000151277211966207,
+      "loss": 0.5869,
+      "step": 797
+    },
+    {
+      "epoch": 0.35929761368752816,
+      "grad_norm": 0.11919040232896805,
+      "learning_rate": 0.00015114999239479685,
+      "loss": 0.6465,
+      "step": 798
+    },
+    {
+      "epoch": 0.35974786132372805,
+      "grad_norm": 0.1117527186870575,
+      "learning_rate": 0.00015102266060503062,
+      "loss": 0.6769,
+      "step": 799
+    },
+    {
+      "epoch": 0.360198108959928,
+      "grad_norm": 0.09560609608888626,
+      "learning_rate": 0.00015089521687626243,
+      "loss": 0.439,
+      "step": 800
+    },
+    {
+      "epoch": 0.3606483565961279,
+      "grad_norm": 0.09249652177095413,
+      "learning_rate": 0.0001507676614880921,
+      "loss": 0.6611,
+      "step": 801
+    },
+    {
+      "epoch": 0.36109860423232776,
+      "grad_norm": 0.10189029574394226,
+      "learning_rate": 0.0001506399947203643,
+      "loss": 0.4699,
+      "step": 802
+    },
+    {
+      "epoch": 0.3615488518685277,
+      "grad_norm": 0.15552063286304474,
+      "learning_rate": 0.00015051221685316815,
+      "loss": 0.34,
+      "step": 803
+    },
+    {
+      "epoch": 0.3619990995047276,
+      "grad_norm": 0.11871030181646347,
+      "learning_rate": 0.00015038432816683655,
+      "loss": 0.3739,
+      "step": 804
+    },
+    {
+      "epoch": 0.3624493471409275,
+      "grad_norm": 0.1611144244670868,
+      "learning_rate": 0.00015025632894194533,
+      "loss": 0.4427,
+      "step": 805
+    },
+    {
+      "epoch": 0.3628995947771274,
+      "grad_norm": 0.13491782546043396,
+      "learning_rate": 0.00015012821945931302,
+      "loss": 0.5327,
+      "step": 806
+    },
+    {
+      "epoch": 0.36334984241332735,
+      "grad_norm": 0.11987339705228806,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 0.465,
+      "step": 807
+    },
+    {
+      "epoch": 0.36380009004952724,
+      "grad_norm": 0.10991059243679047,
+      "learning_rate": 0.00014987167084530792,
+      "loss": 0.3928,
+      "step": 808
+    },
+    {
+      "epoch": 0.3642503376857272,
+      "grad_norm": 0.11346585303544998,
+      "learning_rate": 0.00014974323227677903,
+      "loss": 0.4172,
+      "step": 809
+    },
+    {
+      "epoch": 0.36470058532192706,
+      "grad_norm": 0.10804543644189835,
+      "learning_rate": 0.00014961468457619576,
+      "loss": 0.486,
+      "step": 810
+    },
+    {
+      "epoch": 0.36515083295812695,
+      "grad_norm": 0.1287837028503418,
+      "learning_rate": 0.00014948602802557982,
+      "loss": 0.7842,
+      "step": 811
+    },
+    {
+      "epoch": 0.3656010805943269,
+      "grad_norm": 0.189126119017601,
+      "learning_rate": 0.00014935726290719179,
+      "loss": 0.5055,
+      "step": 812
+    },
+    {
+      "epoch": 0.3660513282305268,
+      "grad_norm": 0.14750684797763824,
+      "learning_rate": 0.0001492283895035305,
+      "loss": 0.7233,
+      "step": 813
+    },
+    {
+      "epoch": 0.3665015758667267,
+      "grad_norm": 0.15690705180168152,
+      "learning_rate": 0.00014909940809733222,
+      "loss": 0.6194,
+      "step": 814
+    },
+    {
+      "epoch": 0.3669518235029266,
+      "grad_norm": 0.19372795522212982,
+      "learning_rate": 0.00014897031897157027,
+      "loss": 0.6624,
+      "step": 815
+    },
+    {
+      "epoch": 0.36740207113912654,
+      "grad_norm": 0.10648975521326065,
+      "learning_rate": 0.00014884112240945427,
+      "loss": 0.4062,
+      "step": 816
+    },
+    {
+      "epoch": 0.3678523187753264,
+      "grad_norm": 0.09759756177663803,
+      "learning_rate": 0.00014871181869442954,
+      "loss": 0.3868,
+      "step": 817
+    },
+    {
+      "epoch": 0.36830256641152637,
+      "grad_norm": 0.10551328957080841,
+      "learning_rate": 0.00014858240811017643,
+      "loss": 0.515,
+      "step": 818
+    },
+    {
+      "epoch": 0.36875281404772625,
+      "grad_norm": 0.1455947756767273,
+      "learning_rate": 0.00014845289094060987,
+      "loss": 0.5502,
+      "step": 819
+    },
+    {
+      "epoch": 0.36920306168392614,
+      "grad_norm": 0.10850723832845688,
+      "learning_rate": 0.00014832326746987848,
+      "loss": 0.5984,
+      "step": 820
+    },
+    {
+      "epoch": 0.3696533093201261,
+      "grad_norm": 0.10516699403524399,
+      "learning_rate": 0.00014819353798236427,
+      "loss": 0.4705,
+      "step": 821
+    },
+    {
+      "epoch": 0.37010355695632596,
+      "grad_norm": 0.12209156155586243,
+      "learning_rate": 0.00014806370276268163,
+      "loss": 0.6246,
+      "step": 822
+    },
+    {
+      "epoch": 0.3705538045925259,
+      "grad_norm": 0.12474829703569412,
+      "learning_rate": 0.00014793376209567714,
+      "loss": 0.6095,
+      "step": 823
+    },
+    {
+      "epoch": 0.3710040522287258,
+      "grad_norm": 0.131979301571846,
+      "learning_rate": 0.00014780371626642858,
+      "loss": 0.7422,
+      "step": 824
+    },
+    {
+      "epoch": 0.37145429986492573,
+      "grad_norm": 0.2378464937210083,
+      "learning_rate": 0.0001476735655602445,
+      "loss": 0.6664,
+      "step": 825
+    },
+    {
+      "epoch": 0.3719045475011256,
+      "grad_norm": 0.10609517246484756,
+      "learning_rate": 0.00014754331026266347,
+      "loss": 0.5718,
+      "step": 826
+    },
+    {
+      "epoch": 0.37235479513732556,
+      "grad_norm": 0.14600014686584473,
+      "learning_rate": 0.0001474129506594536,
+      "loss": 0.7753,
+      "step": 827
+    },
+    {
+      "epoch": 0.37280504277352544,
+      "grad_norm": 0.1877153515815735,
+      "learning_rate": 0.00014728248703661182,
+      "loss": 0.7968,
+      "step": 828
+    },
+    {
+      "epoch": 0.3732552904097253,
+      "grad_norm": 0.10621737688779831,
+      "learning_rate": 0.00014715191968036327,
+      "loss": 0.7492,
+      "step": 829
+    },
+    {
+      "epoch": 0.37370553804592527,
+      "grad_norm": 0.11296433210372925,
+      "learning_rate": 0.0001470212488771606,
+      "loss": 0.3859,
+      "step": 830
+    },
+    {
+      "epoch": 0.37415578568212515,
+      "grad_norm": 0.13514156639575958,
+      "learning_rate": 0.00014689047491368355,
+      "loss": 0.8557,
+      "step": 831
+    },
+    {
+      "epoch": 0.3746060333183251,
+      "grad_norm": 0.13606959581375122,
+      "learning_rate": 0.0001467595980768381,
+      "loss": 0.4246,
+      "step": 832
+    },
+    {
+      "epoch": 0.375056280954525,
+      "grad_norm": 0.11795323342084885,
+      "learning_rate": 0.0001466286186537559,
+      "loss": 0.5286,
+      "step": 833
+    },
+    {
+      "epoch": 0.3755065285907249,
+      "grad_norm": 0.10288207232952118,
+      "learning_rate": 0.00014649753693179372,
+      "loss": 0.4672,
+      "step": 834
+    },
+    {
+      "epoch": 0.3759567762269248,
+      "grad_norm": 0.15558631718158722,
+      "learning_rate": 0.00014636635319853275,
+      "loss": 0.5923,
+      "step": 835
+    },
+    {
+      "epoch": 0.37640702386312475,
+      "grad_norm": 0.1984454244375229,
+      "learning_rate": 0.00014623506774177797,
+      "loss": 0.8204,
+      "step": 836
+    },
+    {
+      "epoch": 0.37685727149932463,
+      "grad_norm": 0.1156761646270752,
+      "learning_rate": 0.0001461036808495575,
+      "loss": 0.5447,
+      "step": 837
+    },
+    {
+      "epoch": 0.3773075191355245,
+      "grad_norm": 0.09257693588733673,
+      "learning_rate": 0.00014597219281012208,
+      "loss": 0.4283,
+      "step": 838
+    },
+    {
+      "epoch": 0.37775776677172446,
+      "grad_norm": 0.1182890310883522,
+      "learning_rate": 0.00014584060391194435,
+      "loss": 0.6197,
+      "step": 839
+    },
+    {
+      "epoch": 0.37820801440792434,
+      "grad_norm": 0.11148649454116821,
+      "learning_rate": 0.00014570891444371816,
+      "loss": 0.3818,
+      "step": 840
+    },
+    {
+      "epoch": 0.3786582620441243,
+      "grad_norm": 0.12036056816577911,
+      "learning_rate": 0.00014557712469435798,
+      "loss": 0.9072,
+      "step": 841
+    },
+    {
+      "epoch": 0.37910850968032417,
+      "grad_norm": 0.11290227621793747,
+      "learning_rate": 0.00014544523495299842,
+      "loss": 0.623,
+      "step": 842
+    },
+    {
+      "epoch": 0.3795587573165241,
+      "grad_norm": 0.1249767541885376,
+      "learning_rate": 0.00014531324550899334,
+      "loss": 0.6404,
+      "step": 843
+    },
+    {
+      "epoch": 0.380009004952724,
+      "grad_norm": 0.15120069682598114,
+      "learning_rate": 0.0001451811566519154,
+      "loss": 0.4482,
+      "step": 844
+    },
+    {
+      "epoch": 0.38045925258892394,
+      "grad_norm": 0.10740632563829422,
+      "learning_rate": 0.00014504896867155532,
+      "loss": 0.3576,
+      "step": 845
+    },
+    {
+      "epoch": 0.3809095002251238,
+      "grad_norm": 0.11806316673755646,
+      "learning_rate": 0.00014491668185792134,
+      "loss": 0.4433,
+      "step": 846
+    },
+    {
+      "epoch": 0.3813597478613237,
+      "grad_norm": 0.14729344844818115,
+      "learning_rate": 0.00014478429650123853,
+      "loss": 0.4373,
+      "step": 847
+    },
+    {
+      "epoch": 0.38180999549752365,
+      "grad_norm": 0.08761313557624817,
+      "learning_rate": 0.00014465181289194802,
+      "loss": 0.5744,
+      "step": 848
+    },
+    {
+      "epoch": 0.38226024313372353,
+      "grad_norm": 0.08612076193094254,
+      "learning_rate": 0.0001445192313207067,
+      "loss": 0.4019,
+      "step": 849
+    },
+    {
+      "epoch": 0.3827104907699235,
+      "grad_norm": 0.11369466781616211,
+      "learning_rate": 0.0001443865520783863,
+      "loss": 0.3979,
+      "step": 850
+    },
+    {
+      "epoch": 0.38316073840612336,
+      "grad_norm": 0.13293461501598358,
+      "learning_rate": 0.00014425377545607276,
+      "loss": 0.6473,
+      "step": 851
+    },
+    {
+      "epoch": 0.3836109860423233,
+      "grad_norm": 0.12485184520483017,
+      "learning_rate": 0.00014412090174506568,
+      "loss": 0.7031,
+      "step": 852
+    },
+    {
+      "epoch": 0.3840612336785232,
+      "grad_norm": 0.16592583060264587,
+      "learning_rate": 0.0001439879312368778,
+      "loss": 0.9979,
+      "step": 853
+    },
+    {
+      "epoch": 0.3845114813147231,
+      "grad_norm": 0.1345318853855133,
+      "learning_rate": 0.00014385486422323405,
+      "loss": 0.4682,
+      "step": 854
+    },
+    {
+      "epoch": 0.384961728950923,
+      "grad_norm": 0.10853058099746704,
+      "learning_rate": 0.00014372170099607125,
+      "loss": 0.4729,
+      "step": 855
+    },
+    {
+      "epoch": 0.3854119765871229,
+      "grad_norm": 0.12104907631874084,
+      "learning_rate": 0.00014358844184753712,
+      "loss": 0.575,
+      "step": 856
+    },
+    {
+      "epoch": 0.38586222422332284,
+      "grad_norm": 0.11846081167459488,
+      "learning_rate": 0.00014345508706998996,
+      "loss": 0.676,
+      "step": 857
+    },
+    {
+      "epoch": 0.3863124718595227,
+      "grad_norm": 0.12110213935375214,
+      "learning_rate": 0.0001433216369559978,
+      "loss": 0.6396,
+      "step": 858
+    },
+    {
+      "epoch": 0.38676271949572266,
+      "grad_norm": 0.12762798368930817,
+      "learning_rate": 0.00014318809179833793,
+      "loss": 0.3068,
+      "step": 859
+    },
+    {
+      "epoch": 0.38721296713192255,
+      "grad_norm": 0.11746729165315628,
+      "learning_rate": 0.00014305445188999598,
+      "loss": 0.665,
+      "step": 860
+    },
+    {
+      "epoch": 0.3876632147681225,
+      "grad_norm": 0.17210255563259125,
+      "learning_rate": 0.0001429207175241656,
+      "loss": 0.8916,
+      "step": 861
+    },
+    {
+      "epoch": 0.3881134624043224,
+      "grad_norm": 0.12675803899765015,
+      "learning_rate": 0.00014278688899424765,
+      "loss": 0.5564,
+      "step": 862
+    },
+    {
+      "epoch": 0.3885637100405223,
+      "grad_norm": 0.14210274815559387,
+      "learning_rate": 0.00014265296659384956,
+      "loss": 0.6256,
+      "step": 863
+    },
+    {
+      "epoch": 0.3890139576767222,
+      "grad_norm": 0.11456255614757538,
+      "learning_rate": 0.00014251895061678465,
+      "loss": 0.6183,
+      "step": 864
+    },
+    {
+      "epoch": 0.3894642053129221,
+      "grad_norm": 0.1801631599664688,
+      "learning_rate": 0.00014238484135707164,
+      "loss": 1.0253,
+      "step": 865
+    },
+    {
+      "epoch": 0.389914452949122,
+      "grad_norm": 0.09097810834646225,
+      "learning_rate": 0.00014225063910893387,
+      "loss": 0.4098,
+      "step": 866
+    },
+    {
+      "epoch": 0.3903647005853219,
+      "grad_norm": 0.1634313315153122,
+      "learning_rate": 0.00014211634416679857,
+      "loss": 0.7349,
+      "step": 867
+    },
+    {
+      "epoch": 0.39081494822152185,
+      "grad_norm": 0.0949796736240387,
+      "learning_rate": 0.00014198195682529652,
+      "loss": 0.6894,
+      "step": 868
+    },
+    {
+      "epoch": 0.39126519585772174,
+      "grad_norm": 0.18297404050827026,
+      "learning_rate": 0.0001418474773792611,
+      "loss": 0.7267,
+      "step": 869
+    },
+    {
+      "epoch": 0.3917154434939217,
+      "grad_norm": 0.10333032160997391,
+      "learning_rate": 0.0001417129061237278,
+      "loss": 0.6559,
+      "step": 870
+    },
+    {
+      "epoch": 0.39216569113012156,
+      "grad_norm": 0.18238197267055511,
+      "learning_rate": 0.0001415782433539335,
+      "loss": 0.5942,
+      "step": 871
+    },
+    {
+      "epoch": 0.3926159387663215,
+      "grad_norm": 0.0907868891954422,
+      "learning_rate": 0.0001414434893653159,
+      "loss": 0.5297,
+      "step": 872
+    },
+    {
+      "epoch": 0.3930661864025214,
+      "grad_norm": 0.18751130998134613,
+      "learning_rate": 0.00014130864445351273,
+      "loss": 0.7636,
+      "step": 873
+    },
+    {
+      "epoch": 0.3935164340387213,
+      "grad_norm": 0.10867595672607422,
+      "learning_rate": 0.00014117370891436135,
+      "loss": 0.4767,
+      "step": 874
+    },
+    {
+      "epoch": 0.3939666816749212,
+      "grad_norm": 0.15845608711242676,
+      "learning_rate": 0.0001410386830438978,
+      "loss": 0.7434,
+      "step": 875
+    },
+    {
+      "epoch": 0.3944169293111211,
+      "grad_norm": 0.14045387506484985,
+      "learning_rate": 0.00014090356713835638,
+      "loss": 0.4809,
+      "step": 876
+    },
+    {
+      "epoch": 0.39486717694732104,
+      "grad_norm": 0.11211243271827698,
+      "learning_rate": 0.00014076836149416887,
+      "loss": 0.6769,
+      "step": 877
+    },
+    {
+      "epoch": 0.3953174245835209,
+      "grad_norm": 0.15294021368026733,
+      "learning_rate": 0.00014063306640796404,
+      "loss": 0.6243,
+      "step": 878
+    },
+    {
+      "epoch": 0.39576767221972087,
+      "grad_norm": 0.10523733496665955,
+      "learning_rate": 0.00014049768217656676,
+      "loss": 0.6238,
+      "step": 879
+    },
+    {
+      "epoch": 0.39621791985592075,
+      "grad_norm": 0.1284034103155136,
+      "learning_rate": 0.0001403622090969975,
+      "loss": 0.4213,
+      "step": 880
+    },
+    {
+      "epoch": 0.3966681674921207,
+      "grad_norm": 0.11438385397195816,
+      "learning_rate": 0.00014022664746647168,
+      "loss": 0.5444,
+      "step": 881
+    },
+    {
+      "epoch": 0.3971184151283206,
+      "grad_norm": 0.09823127835988998,
+      "learning_rate": 0.000140090997582399,
+      "loss": 0.5754,
+      "step": 882
+    },
+    {
+      "epoch": 0.39756866276452046,
+      "grad_norm": 0.12103039771318436,
+      "learning_rate": 0.0001399552597423828,
+      "loss": 0.3748,
+      "step": 883
+    },
+    {
+      "epoch": 0.3980189104007204,
+      "grad_norm": 0.13115237653255463,
+      "learning_rate": 0.00013981943424421932,
+      "loss": 0.589,
+      "step": 884
+    },
+    {
+      "epoch": 0.3984691580369203,
+      "grad_norm": 0.1328432261943817,
+      "learning_rate": 0.00013968352138589714,
+      "loss": 0.6935,
+      "step": 885
+    },
+    {
+      "epoch": 0.39891940567312023,
+      "grad_norm": 0.12178020924329758,
+      "learning_rate": 0.00013954752146559654,
+      "loss": 0.4687,
+      "step": 886
+    },
+    {
+      "epoch": 0.3993696533093201,
+      "grad_norm": 0.08869553357362747,
+      "learning_rate": 0.00013941143478168872,
+      "loss": 0.4652,
+      "step": 887
+    },
+    {
+      "epoch": 0.39981990094552006,
+      "grad_norm": 0.11473149061203003,
+      "learning_rate": 0.0001392752616327354,
+      "loss": 0.4415,
+      "step": 888
+    },
+    {
+      "epoch": 0.40027014858171994,
+      "grad_norm": 0.1047920286655426,
+      "learning_rate": 0.00013913900231748777,
+      "loss": 0.5512,
+      "step": 889
+    },
+    {
+      "epoch": 0.4007203962179199,
+      "grad_norm": 0.09504768997430801,
+      "learning_rate": 0.00013900265713488624,
+      "loss": 0.4918,
+      "step": 890
+    },
+    {
+      "epoch": 0.40117064385411977,
+      "grad_norm": 0.09980174899101257,
+      "learning_rate": 0.00013886622638405952,
+      "loss": 0.496,
+      "step": 891
+    },
+    {
+      "epoch": 0.40162089149031965,
+      "grad_norm": 0.10838398337364197,
+      "learning_rate": 0.00013872971036432406,
+      "loss": 0.5742,
+      "step": 892
+    },
+    {
+      "epoch": 0.4020711391265196,
+      "grad_norm": 0.15895016491413116,
+      "learning_rate": 0.00013859310937518342,
+      "loss": 0.4869,
+      "step": 893
+    },
+    {
+      "epoch": 0.4025213867627195,
+      "grad_norm": 0.10939373075962067,
+      "learning_rate": 0.0001384564237163275,
+      "loss": 0.5913,
+      "step": 894
+    },
+    {
+      "epoch": 0.4029716343989194,
+      "grad_norm": 0.11063853651285172,
+      "learning_rate": 0.00013831965368763205,
+      "loss": 0.5093,
+      "step": 895
+    },
+    {
+      "epoch": 0.4034218820351193,
+      "grad_norm": 0.15515856444835663,
+      "learning_rate": 0.00013818279958915786,
+      "loss": 0.7627,
+      "step": 896
+    },
+    {
+      "epoch": 0.40387212967131925,
+      "grad_norm": 0.1372167319059372,
+      "learning_rate": 0.00013804586172115016,
+      "loss": 0.4953,
+      "step": 897
+    },
+    {
+      "epoch": 0.40432237730751913,
+      "grad_norm": 0.18656258285045624,
+      "learning_rate": 0.00013790884038403795,
+      "loss": 0.9423,
+      "step": 898
+    },
+    {
+      "epoch": 0.4047726249437191,
+      "grad_norm": 0.17795826494693756,
+      "learning_rate": 0.0001377717358784334,
+      "loss": 0.7234,
+      "step": 899
+    },
+    {
+      "epoch": 0.40522287257991896,
+      "grad_norm": 0.11300812661647797,
+      "learning_rate": 0.0001376345485051312,
+      "loss": 0.6728,
+      "step": 900
+    },
+    {
+      "epoch": 0.40567312021611884,
+      "grad_norm": 0.1559130698442459,
+      "learning_rate": 0.00013749727856510768,
+      "loss": 0.5094,
+      "step": 901
+    },
+    {
+      "epoch": 0.4061233678523188,
+      "grad_norm": 0.19531500339508057,
+      "learning_rate": 0.0001373599263595204,
+      "loss": 0.9277,
+      "step": 902
+    },
+    {
+      "epoch": 0.40657361548851867,
+      "grad_norm": 0.09455003589391708,
+      "learning_rate": 0.00013722249218970747,
+      "loss": 0.3287,
+      "step": 903
+    },
+    {
+      "epoch": 0.4070238631247186,
+      "grad_norm": 0.1947425901889801,
+      "learning_rate": 0.00013708497635718674,
+      "loss": 0.4564,
+      "step": 904
+    },
+    {
+      "epoch": 0.4074741107609185,
+      "grad_norm": 0.13219277560710907,
+      "learning_rate": 0.00013694737916365517,
+      "loss": 0.4765,
+      "step": 905
+    },
+    {
+      "epoch": 0.40792435839711844,
+      "grad_norm": 0.10248047113418579,
+      "learning_rate": 0.00013680970091098832,
+      "loss": 0.6187,
+      "step": 906
+    },
+    {
+      "epoch": 0.4083746060333183,
+      "grad_norm": 0.08886689692735672,
+      "learning_rate": 0.0001366719419012396,
+      "loss": 0.4869,
+      "step": 907
+    },
+    {
+      "epoch": 0.40882485366951826,
+      "grad_norm": 0.14525815844535828,
+      "learning_rate": 0.00013653410243663952,
+      "loss": 0.5753,
+      "step": 908
+    },
+    {
+      "epoch": 0.40927510130571815,
+      "grad_norm": 0.10542677342891693,
+      "learning_rate": 0.0001363961828195951,
+      "loss": 0.4275,
+      "step": 909
+    },
+    {
+      "epoch": 0.40972534894191803,
+      "grad_norm": 0.1681213229894638,
+      "learning_rate": 0.00013625818335268926,
+      "loss": 0.4951,
+      "step": 910
+    },
+    {
+      "epoch": 0.410175596578118,
+      "grad_norm": 0.09946253895759583,
+      "learning_rate": 0.00013612010433868005,
+      "loss": 0.4537,
+      "step": 911
+    },
+    {
+      "epoch": 0.41062584421431786,
+      "grad_norm": 0.11235876381397247,
+      "learning_rate": 0.0001359819460805001,
+      "loss": 0.6303,
+      "step": 912
+    },
+    {
+      "epoch": 0.4110760918505178,
+      "grad_norm": 0.10638522356748581,
+      "learning_rate": 0.00013584370888125583,
+      "loss": 0.511,
+      "step": 913
+    },
+    {
+      "epoch": 0.4115263394867177,
+      "grad_norm": 0.14852966368198395,
+      "learning_rate": 0.0001357053930442269,
+      "loss": 0.5443,
+      "step": 914
+    },
+    {
+      "epoch": 0.4119765871229176,
+      "grad_norm": 0.10808053612709045,
+      "learning_rate": 0.0001355669988728655,
+      "loss": 0.3655,
+      "step": 915
+    },
+    {
+      "epoch": 0.4124268347591175,
+      "grad_norm": 0.15118542313575745,
+      "learning_rate": 0.00013542852667079556,
+      "loss": 0.5236,
+      "step": 916
+    },
+    {
+      "epoch": 0.41287708239531745,
+      "grad_norm": 0.1098078191280365,
+      "learning_rate": 0.0001352899767418124,
+      "loss": 0.5078,
+      "step": 917
+    },
+    {
+      "epoch": 0.41332733003151734,
+      "grad_norm": 0.10399964451789856,
+      "learning_rate": 0.00013515134938988169,
+      "loss": 0.631,
+      "step": 918
+    },
+    {
+      "epoch": 0.4137775776677172,
+      "grad_norm": 0.09587126225233078,
+      "learning_rate": 0.00013501264491913906,
+      "loss": 0.4485,
+      "step": 919
+    },
+    {
+      "epoch": 0.41422782530391716,
+      "grad_norm": 0.1310843825340271,
+      "learning_rate": 0.00013487386363388932,
+      "loss": 0.8821,
+      "step": 920
+    },
+    {
+      "epoch": 0.41467807294011705,
+      "grad_norm": 0.13849076628684998,
+      "learning_rate": 0.0001347350058386057,
+      "loss": 0.7235,
+      "step": 921
+    },
+    {
+      "epoch": 0.415128320576317,
+      "grad_norm": 0.11911311745643616,
+      "learning_rate": 0.00013459607183792947,
+      "loss": 0.6408,
+      "step": 922
+    },
+    {
+      "epoch": 0.4155785682125169,
+      "grad_norm": 0.11103199422359467,
+      "learning_rate": 0.00013445706193666892,
+      "loss": 0.5971,
+      "step": 923
+    },
+    {
+      "epoch": 0.4160288158487168,
+      "grad_norm": 0.11604959517717361,
+      "learning_rate": 0.00013431797643979895,
+      "loss": 0.7261,
+      "step": 924
+    },
+    {
+      "epoch": 0.4164790634849167,
+      "grad_norm": 0.1506742686033249,
+      "learning_rate": 0.00013417881565246027,
+      "loss": 0.4864,
+      "step": 925
+    },
+    {
+      "epoch": 0.41692931112111664,
+      "grad_norm": 0.12140141427516937,
+      "learning_rate": 0.00013403957987995882,
+      "loss": 0.6282,
+      "step": 926
+    },
+    {
+      "epoch": 0.4173795587573165,
+      "grad_norm": 0.0939534604549408,
+      "learning_rate": 0.000133900269427765,
+      "loss": 0.4966,
+      "step": 927
+    },
+    {
+      "epoch": 0.4178298063935164,
+      "grad_norm": 0.14521941542625427,
+      "learning_rate": 0.00013376088460151305,
+      "loss": 0.5198,
+      "step": 928
+    },
+    {
+      "epoch": 0.41828005402971635,
+      "grad_norm": 0.1109466478228569,
+      "learning_rate": 0.0001336214257070004,
+      "loss": 0.4807,
+      "step": 929
+    },
+    {
+      "epoch": 0.41873030166591624,
+      "grad_norm": 0.18470987677574158,
+      "learning_rate": 0.00013348189305018704,
+      "loss": 1.0278,
+      "step": 930
+    },
+    {
+      "epoch": 0.4191805493021162,
+      "grad_norm": 0.11823201179504395,
+      "learning_rate": 0.00013334228693719465,
+      "loss": 0.7853,
+      "step": 931
+    },
+    {
+      "epoch": 0.41963079693831606,
+      "grad_norm": 0.10978804528713226,
+      "learning_rate": 0.00013320260767430615,
+      "loss": 0.8322,
+      "step": 932
+    },
+    {
+      "epoch": 0.420081044574516,
+      "grad_norm": 0.12781614065170288,
+      "learning_rate": 0.00013306285556796495,
+      "loss": 0.814,
+      "step": 933
+    },
+    {
+      "epoch": 0.4205312922107159,
+      "grad_norm": 0.17657306790351868,
+      "learning_rate": 0.00013292303092477426,
+      "loss": 0.6681,
+      "step": 934
+    },
+    {
+      "epoch": 0.42098153984691583,
+      "grad_norm": 0.15384222567081451,
+      "learning_rate": 0.00013278313405149642,
+      "loss": 0.4704,
+      "step": 935
+    },
+    {
+      "epoch": 0.4214317874831157,
+      "grad_norm": 0.1087196096777916,
+      "learning_rate": 0.00013264316525505218,
+      "loss": 0.575,
+      "step": 936
+    },
+    {
+      "epoch": 0.4218820351193156,
+      "grad_norm": 0.13700257241725922,
+      "learning_rate": 0.00013250312484252023,
+      "loss": 0.6139,
+      "step": 937
+    },
+    {
+      "epoch": 0.42233228275551554,
+      "grad_norm": 0.18106839060783386,
+      "learning_rate": 0.0001323630131211363,
+      "loss": 0.7156,
+      "step": 938
+    },
+    {
+      "epoch": 0.4227825303917154,
+      "grad_norm": 0.12380406260490417,
+      "learning_rate": 0.00013222283039829248,
+      "loss": 0.9108,
+      "step": 939
+    },
+    {
+      "epoch": 0.42323277802791537,
+      "grad_norm": 0.18230657279491425,
+      "learning_rate": 0.00013208257698153677,
+      "loss": 0.442,
+      "step": 940
+    },
+    {
+      "epoch": 0.42368302566411525,
+      "grad_norm": 0.15106737613677979,
+      "learning_rate": 0.00013194225317857217,
+      "loss": 0.6779,
+      "step": 941
+    },
+    {
+      "epoch": 0.4241332733003152,
+      "grad_norm": 0.1334100365638733,
+      "learning_rate": 0.00013180185929725617,
+      "loss": 0.8568,
+      "step": 942
+    },
+    {
+      "epoch": 0.4245835209365151,
+      "grad_norm": 0.08325067907571793,
+      "learning_rate": 0.00013166139564559993,
+      "loss": 0.4378,
+      "step": 943
+    },
+    {
+      "epoch": 0.425033768572715,
+      "grad_norm": 0.10389431565999985,
+      "learning_rate": 0.00013152086253176773,
+      "loss": 0.5099,
+      "step": 944
+    },
+    {
+      "epoch": 0.4254840162089149,
+      "grad_norm": 0.11528710275888443,
+      "learning_rate": 0.0001313802602640763,
+      "loss": 0.5181,
+      "step": 945
+    },
+    {
+      "epoch": 0.4259342638451148,
+      "grad_norm": 0.10791566222906113,
+      "learning_rate": 0.00013123958915099392,
+      "loss": 0.4125,
+      "step": 946
+    },
+    {
+      "epoch": 0.42638451148131473,
+      "grad_norm": 0.0962347611784935,
+      "learning_rate": 0.00013109884950114007,
+      "loss": 0.4247,
+      "step": 947
+    },
+    {
+      "epoch": 0.4268347591175146,
+      "grad_norm": 0.1013663187623024,
+      "learning_rate": 0.00013095804162328452,
+      "loss": 0.5027,
+      "step": 948
+    },
+    {
+      "epoch": 0.42728500675371456,
+      "grad_norm": 0.08481322228908539,
+      "learning_rate": 0.00013081716582634675,
+      "loss": 0.4425,
+      "step": 949
+    },
+    {
+      "epoch": 0.42773525438991444,
+      "grad_norm": 0.09992804378271103,
+      "learning_rate": 0.0001306762224193952,
+      "loss": 0.4265,
+      "step": 950
+    },
+    {
+      "epoch": 0.4281855020261144,
+      "grad_norm": 0.1000409871339798,
+      "learning_rate": 0.0001305352117116467,
+      "loss": 0.4453,
+      "step": 951
+    },
+    {
+      "epoch": 0.42863574966231427,
+      "grad_norm": 0.13956154882907867,
+      "learning_rate": 0.00013039413401246577,
+      "loss": 0.3944,
+      "step": 952
+    },
+    {
+      "epoch": 0.4290859972985142,
+      "grad_norm": 0.10688282549381256,
+      "learning_rate": 0.00013025298963136378,
+      "loss": 0.4619,
+      "step": 953
+    },
+    {
+      "epoch": 0.4295362449347141,
+      "grad_norm": 0.14733999967575073,
+      "learning_rate": 0.00013011177887799845,
+      "loss": 0.5179,
+      "step": 954
+    },
+    {
+      "epoch": 0.429986492570914,
+      "grad_norm": 0.11598526686429977,
+      "learning_rate": 0.00012997050206217317,
+      "loss": 0.3756,
+      "step": 955
+    },
+    {
+      "epoch": 0.4304367402071139,
+      "grad_norm": 0.0924430638551712,
+      "learning_rate": 0.00012982915949383617,
+      "loss": 0.4439,
+      "step": 956
+    },
+    {
+      "epoch": 0.4308869878433138,
+      "grad_norm": 0.14701209962368011,
+      "learning_rate": 0.00012968775148308002,
+      "loss": 0.6424,
+      "step": 957
+    },
+    {
+      "epoch": 0.43133723547951375,
+      "grad_norm": 0.1830032765865326,
+      "learning_rate": 0.0001295462783401408,
+      "loss": 0.7465,
+      "step": 958
+    },
+    {
+      "epoch": 0.43178748311571363,
+      "grad_norm": 0.157825767993927,
+      "learning_rate": 0.00012940474037539756,
+      "loss": 0.6951,
+      "step": 959
+    },
+    {
+      "epoch": 0.4322377307519136,
+      "grad_norm": 0.14804725348949432,
+      "learning_rate": 0.00012926313789937143,
+      "loss": 0.6299,
+      "step": 960
+    },
+    {
+      "epoch": 0.43268797838811346,
+      "grad_norm": 0.11605916172266006,
+      "learning_rate": 0.00012912147122272523,
+      "loss": 0.4111,
+      "step": 961
+    },
+    {
+      "epoch": 0.4331382260243134,
+      "grad_norm": 0.08589927107095718,
+      "learning_rate": 0.00012897974065626252,
+      "loss": 0.5103,
+      "step": 962
+    },
+    {
+      "epoch": 0.4335884736605133,
+      "grad_norm": 0.18623948097229004,
+      "learning_rate": 0.00012883794651092704,
+      "loss": 0.3807,
+      "step": 963
+    },
+    {
+      "epoch": 0.43403872129671317,
+      "grad_norm": 0.12046798318624496,
+      "learning_rate": 0.00012869608909780214,
+      "loss": 0.4913,
+      "step": 964
+    },
+    {
+      "epoch": 0.4344889689329131,
+      "grad_norm": 0.13895408809185028,
+      "learning_rate": 0.00012855416872810973,
+      "loss": 0.3356,
+      "step": 965
+    },
+    {
+      "epoch": 0.434939216569113,
+      "grad_norm": 0.08843182027339935,
+      "learning_rate": 0.0001284121857132101,
+      "loss": 0.4028,
+      "step": 966
+    },
+    {
+      "epoch": 0.43538946420531294,
+      "grad_norm": 0.14155325293540955,
+      "learning_rate": 0.00012827014036460085,
+      "loss": 0.468,
+      "step": 967
+    },
+    {
+      "epoch": 0.4358397118415128,
+      "grad_norm": 0.14368608593940735,
+      "learning_rate": 0.00012812803299391628,
+      "loss": 0.7905,
+      "step": 968
+    },
+    {
+      "epoch": 0.43628995947771276,
+      "grad_norm": 0.15499171614646912,
+      "learning_rate": 0.0001279858639129269,
+      "loss": 0.6043,
+      "step": 969
+    },
+    {
+      "epoch": 0.43674020711391265,
+      "grad_norm": 0.18597197532653809,
+      "learning_rate": 0.0001278436334335385,
+      "loss": 0.8312,
+      "step": 970
+    },
+    {
+      "epoch": 0.4371904547501126,
+      "grad_norm": 0.10050120949745178,
+      "learning_rate": 0.0001277013418677916,
+      "loss": 0.3709,
+      "step": 971
+    },
+    {
+      "epoch": 0.4376407023863125,
+      "grad_norm": 0.16297680139541626,
+      "learning_rate": 0.00012755898952786077,
+      "loss": 0.6841,
+      "step": 972
+    },
+    {
+      "epoch": 0.43809095002251236,
+      "grad_norm": 0.12217584997415543,
+      "learning_rate": 0.00012741657672605384,
+      "loss": 0.7174,
+      "step": 973
+    },
+    {
+      "epoch": 0.4385411976587123,
+      "grad_norm": 0.0945337563753128,
+      "learning_rate": 0.0001272741037748114,
+      "loss": 0.402,
+      "step": 974
+    },
+    {
+      "epoch": 0.4389914452949122,
+      "grad_norm": 0.09897486120462418,
+      "learning_rate": 0.0001271315709867059,
+      "loss": 0.542,
+      "step": 975
+    },
+    {
+      "epoch": 0.4394416929311121,
+      "grad_norm": 0.13974529504776,
+      "learning_rate": 0.00012698897867444113,
+      "loss": 0.9472,
+      "step": 976
+    },
+    {
+      "epoch": 0.439891940567312,
+      "grad_norm": 0.13686338067054749,
+      "learning_rate": 0.0001268463271508514,
+      "loss": 0.5166,
+      "step": 977
+    },
+    {
+      "epoch": 0.44034218820351195,
+      "grad_norm": 0.14328137040138245,
+      "learning_rate": 0.000126703616728901,
+      "loss": 0.7918,
+      "step": 978
+    },
+    {
+      "epoch": 0.44079243583971184,
+      "grad_norm": 0.11810629814863205,
+      "learning_rate": 0.0001265608477216834,
+      "loss": 0.4506,
+      "step": 979
+    },
+    {
+      "epoch": 0.4412426834759118,
+      "grad_norm": 0.13406547904014587,
+      "learning_rate": 0.00012641802044242065,
+      "loss": 0.5489,
+      "step": 980
+    },
+    {
+      "epoch": 0.44169293111211166,
+      "grad_norm": 0.12685255706310272,
+      "learning_rate": 0.00012627513520446253,
+      "loss": 0.4508,
+      "step": 981
+    },
+    {
+      "epoch": 0.44214317874831155,
+      "grad_norm": 0.10661809146404266,
+      "learning_rate": 0.00012613219232128608,
+      "loss": 0.4359,
+      "step": 982
+    },
+    {
+      "epoch": 0.4425934263845115,
+      "grad_norm": 0.15434306859970093,
+      "learning_rate": 0.00012598919210649476,
+      "loss": 0.6928,
+      "step": 983
+    },
+    {
+      "epoch": 0.4430436740207114,
+      "grad_norm": 0.15799319744110107,
+      "learning_rate": 0.00012584613487381788,
+      "loss": 0.84,
+      "step": 984
+    },
+    {
+      "epoch": 0.4434939216569113,
+      "grad_norm": 0.15313151478767395,
+      "learning_rate": 0.0001257030209371097,
+      "loss": 0.7133,
+      "step": 985
+    },
+    {
+      "epoch": 0.4439441692931112,
+      "grad_norm": 0.13239026069641113,
+      "learning_rate": 0.00012555985061034902,
+      "loss": 0.3829,
+      "step": 986
+    },
+    {
+      "epoch": 0.44439441692931114,
+      "grad_norm": 0.13071371614933014,
+      "learning_rate": 0.00012541662420763834,
+      "loss": 0.7179,
+      "step": 987
+    },
+    {
+      "epoch": 0.444844664565511,
+      "grad_norm": 0.11222220957279205,
+      "learning_rate": 0.00012527334204320308,
+      "loss": 0.4961,
+      "step": 988
+    },
+    {
+      "epoch": 0.44529491220171097,
+      "grad_norm": 0.10775894671678543,
+      "learning_rate": 0.00012513000443139112,
+      "loss": 0.6711,
+      "step": 989
+    },
+    {
+      "epoch": 0.44574515983791085,
+      "grad_norm": 0.19599351286888123,
+      "learning_rate": 0.0001249866116866719,
+      "loss": 0.5903,
+      "step": 990
+    },
+    {
+      "epoch": 0.44619540747411074,
+      "grad_norm": 0.16542845964431763,
+      "learning_rate": 0.00012484316412363585,
+      "loss": 0.7201,
+      "step": 991
+    },
+    {
+      "epoch": 0.4466456551103107,
+      "grad_norm": 0.10988330096006393,
+      "learning_rate": 0.0001246996620569937,
+      "loss": 0.5231,
+      "step": 992
+    },
+    {
+      "epoch": 0.44709590274651057,
+      "grad_norm": 0.17398272454738617,
+      "learning_rate": 0.0001245561058015757,
+      "loss": 0.8279,
+      "step": 993
+    },
+    {
+      "epoch": 0.4475461503827105,
+      "grad_norm": 0.11568157374858856,
+      "learning_rate": 0.000124412495672331,
+      "loss": 0.3994,
+      "step": 994
+    },
+    {
+      "epoch": 0.4479963980189104,
+      "grad_norm": 0.2278115302324295,
+      "learning_rate": 0.00012426883198432697,
+      "loss": 0.7765,
+      "step": 995
+    },
+    {
+      "epoch": 0.44844664565511033,
+      "grad_norm": 0.12829448282718658,
+      "learning_rate": 0.00012412511505274844,
+      "loss": 0.7713,
+      "step": 996
+    },
+    {
+      "epoch": 0.4488968932913102,
+      "grad_norm": 0.13341756165027618,
+      "learning_rate": 0.0001239813451928971,
+      "loss": 0.5739,
+      "step": 997
+    },
+    {
+      "epoch": 0.44934714092751016,
+      "grad_norm": 0.11134277284145355,
+      "learning_rate": 0.00012383752272019074,
+      "loss": 0.4434,
+      "step": 998
+    },
+    {
+      "epoch": 0.44979738856371004,
+      "grad_norm": 0.12369354814291,
+      "learning_rate": 0.00012369364795016254,
+      "loss": 0.4361,
+      "step": 999
+    },
+    {
+      "epoch": 0.45024763619990993,
+      "grad_norm": 0.15551245212554932,
+      "learning_rate": 0.00012354972119846047,
+      "loss": 0.7093,
+      "step": 1000
+    },
+    {
+      "epoch": 0.45069788383610987,
+      "grad_norm": 0.14374928176403046,
+      "learning_rate": 0.0001234057427808465,
+      "loss": 0.6275,
+      "step": 1001
+    },
+    {
+      "epoch": 0.45114813147230975,
+      "grad_norm": 0.22181017696857452,
+      "learning_rate": 0.00012326171301319604,
+      "loss": 0.6078,
+      "step": 1002
+    },
+    {
+      "epoch": 0.4515983791085097,
+      "grad_norm": 0.09431958198547363,
+      "learning_rate": 0.000123117632211497,
+      "loss": 0.5025,
+      "step": 1003
+    },
+    {
+      "epoch": 0.4520486267447096,
+      "grad_norm": 0.26845240592956543,
+      "learning_rate": 0.00012297350069184935,
+      "loss": 0.6794,
+      "step": 1004
+    },
+    {
+      "epoch": 0.4524988743809095,
+      "grad_norm": 0.11008667945861816,
+      "learning_rate": 0.00012282931877046442,
+      "loss": 0.4678,
+      "step": 1005
+    },
+    {
+      "epoch": 0.4529491220171094,
+      "grad_norm": 0.10794171690940857,
+      "learning_rate": 0.00012268508676366394,
+      "loss": 0.3083,
+      "step": 1006
+    },
+    {
+      "epoch": 0.45339936965330935,
+      "grad_norm": 0.12231296300888062,
+      "learning_rate": 0.0001225408049878796,
+      "loss": 0.6962,
+      "step": 1007
+    },
+    {
+      "epoch": 0.45384961728950923,
+      "grad_norm": 0.11282064020633698,
+      "learning_rate": 0.00012239647375965234,
+      "loss": 0.6915,
+      "step": 1008
+    },
+    {
+      "epoch": 0.4542998649257091,
+      "grad_norm": 0.18748655915260315,
+      "learning_rate": 0.00012225209339563145,
+      "loss": 0.7161,
+      "step": 1009
+    },
+    {
+      "epoch": 0.45475011256190906,
+      "grad_norm": 0.1754457652568817,
+      "learning_rate": 0.0001221076642125742,
+      "loss": 0.6861,
+      "step": 1010
+    },
+    {
+      "epoch": 0.45520036019810894,
+      "grad_norm": 0.11903553456068039,
+      "learning_rate": 0.00012196318652734477,
+      "loss": 0.5809,
+      "step": 1011
+    },
+    {
+      "epoch": 0.4556506078343089,
+      "grad_norm": 0.1223655641078949,
+      "learning_rate": 0.00012181866065691392,
+      "loss": 0.5618,
+      "step": 1012
+    },
+    {
+      "epoch": 0.45610085547050877,
+      "grad_norm": 0.11474163085222244,
+      "learning_rate": 0.00012167408691835809,
+      "loss": 0.4485,
+      "step": 1013
+    },
+    {
+      "epoch": 0.4565511031067087,
+      "grad_norm": 0.1002950370311737,
+      "learning_rate": 0.00012152946562885859,
+      "loss": 0.4542,
+      "step": 1014
+    },
+    {
+      "epoch": 0.4570013507429086,
+      "grad_norm": 0.1637374758720398,
+      "learning_rate": 0.00012138479710570123,
+      "loss": 0.6778,
+      "step": 1015
+    },
+    {
+      "epoch": 0.45745159837910854,
+      "grad_norm": 0.21878598630428314,
+      "learning_rate": 0.00012124008166627536,
+      "loss": 0.7022,
+      "step": 1016
+    },
+    {
+      "epoch": 0.4579018460153084,
+      "grad_norm": 0.11327499896287918,
+      "learning_rate": 0.00012109531962807332,
+      "loss": 0.3138,
+      "step": 1017
+    },
+    {
+      "epoch": 0.4583520936515083,
+      "grad_norm": 0.1296899914741516,
+      "learning_rate": 0.0001209505113086896,
+      "loss": 0.8202,
+      "step": 1018
+    },
+    {
+      "epoch": 0.45880234128770825,
+      "grad_norm": 0.17711815237998962,
+      "learning_rate": 0.00012080565702582027,
+      "loss": 0.8224,
+      "step": 1019
+    },
+    {
+      "epoch": 0.45925258892390813,
+      "grad_norm": 0.15500164031982422,
+      "learning_rate": 0.00012066075709726226,
+      "loss": 0.5952,
+      "step": 1020
+    },
+    {
+      "epoch": 0.4597028365601081,
+      "grad_norm": 0.1448982208967209,
+      "learning_rate": 0.00012051581184091263,
+      "loss": 0.8513,
+      "step": 1021
+    },
+    {
+      "epoch": 0.46015308419630796,
+      "grad_norm": 0.13475260138511658,
+      "learning_rate": 0.00012037082157476783,
+      "loss": 0.5366,
+      "step": 1022
+    },
+    {
+      "epoch": 0.4606033318325079,
+      "grad_norm": 0.12645705044269562,
+      "learning_rate": 0.00012022578661692314,
+      "loss": 0.3769,
+      "step": 1023
+    },
+    {
+      "epoch": 0.4610535794687078,
+      "grad_norm": 0.1389506459236145,
+      "learning_rate": 0.00012008070728557186,
+      "loss": 0.5807,
+      "step": 1024
+    },
+    {
+      "epoch": 0.4615038271049077,
+      "grad_norm": 0.10616742819547653,
+      "learning_rate": 0.00011993558389900464,
+      "loss": 0.5971,
+      "step": 1025
+    },
+    {
+      "epoch": 0.4619540747411076,
+      "grad_norm": 0.11959128081798553,
+      "learning_rate": 0.00011979041677560871,
+      "loss": 0.4767,
+      "step": 1026
+    },
+    {
+      "epoch": 0.4624043223773075,
+      "grad_norm": 0.17118290066719055,
+      "learning_rate": 0.00011964520623386741,
+      "loss": 0.6314,
+      "step": 1027
+    },
+    {
+      "epoch": 0.46285457001350744,
+      "grad_norm": 0.17040640115737915,
+      "learning_rate": 0.0001194999525923592,
+      "loss": 0.6397,
+      "step": 1028
+    },
+    {
+      "epoch": 0.4633048176497073,
+      "grad_norm": 0.12378225475549698,
+      "learning_rate": 0.00011935465616975716,
+      "loss": 0.5451,
+      "step": 1029
+    },
+    {
+      "epoch": 0.46375506528590726,
+      "grad_norm": 0.10181150585412979,
+      "learning_rate": 0.00011920931728482823,
+      "loss": 0.4511,
+      "step": 1030
+    },
+    {
+      "epoch": 0.46420531292210715,
+      "grad_norm": 0.13490979373455048,
+      "learning_rate": 0.00011906393625643244,
+      "loss": 0.5588,
+      "step": 1031
+    },
+    {
+      "epoch": 0.4646555605583071,
+      "grad_norm": 0.11228366941213608,
+      "learning_rate": 0.00011891851340352236,
+      "loss": 0.5573,
+      "step": 1032
+    },
+    {
+      "epoch": 0.465105808194507,
+      "grad_norm": 0.12945735454559326,
+      "learning_rate": 0.00011877304904514232,
+      "loss": 0.5581,
+      "step": 1033
+    },
+    {
+      "epoch": 0.4655560558307069,
+      "grad_norm": 0.11461125314235687,
+      "learning_rate": 0.00011862754350042763,
+      "loss": 0.4796,
+      "step": 1034
+    },
+    {
+      "epoch": 0.4660063034669068,
+      "grad_norm": 0.12311673909425735,
+      "learning_rate": 0.00011848199708860406,
+      "loss": 0.6905,
+      "step": 1035
+    },
+    {
+      "epoch": 0.4664565511031067,
+      "grad_norm": 0.19080358743667603,
+      "learning_rate": 0.00011833641012898691,
+      "loss": 0.5518,
+      "step": 1036
+    },
+    {
+      "epoch": 0.4669067987393066,
+      "grad_norm": 0.1149805560708046,
+      "learning_rate": 0.00011819078294098058,
+      "loss": 0.3845,
+      "step": 1037
+    },
+    {
+      "epoch": 0.4673570463755065,
+      "grad_norm": 0.10079693794250488,
+      "learning_rate": 0.00011804511584407763,
+      "loss": 0.4201,
+      "step": 1038
+    },
+    {
+      "epoch": 0.46780729401170645,
+      "grad_norm": 0.18696637451648712,
+      "learning_rate": 0.00011789940915785825,
+      "loss": 0.8889,
+      "step": 1039
+    },
+    {
+      "epoch": 0.46825754164790634,
+      "grad_norm": 0.22084087133407593,
+      "learning_rate": 0.00011775366320198939,
+      "loss": 1.0403,
+      "step": 1040
+    },
+    {
+      "epoch": 0.4687077892841063,
+      "grad_norm": 0.11905587464570999,
+      "learning_rate": 0.00011760787829622424,
+      "loss": 0.6145,
+      "step": 1041
+    },
+    {
+      "epoch": 0.46915803692030617,
+      "grad_norm": 0.09251664578914642,
+      "learning_rate": 0.00011746205476040138,
+      "loss": 0.498,
+      "step": 1042
+    },
+    {
+      "epoch": 0.4696082845565061,
+      "grad_norm": 0.17941583693027496,
+      "learning_rate": 0.0001173161929144442,
+      "loss": 0.4337,
+      "step": 1043
+    },
+    {
+      "epoch": 0.470058532192706,
+      "grad_norm": 0.11491622030735016,
+      "learning_rate": 0.00011717029307836012,
+      "loss": 0.608,
+      "step": 1044
+    },
+    {
+      "epoch": 0.4705087798289059,
+      "grad_norm": 0.1131272241473198,
+      "learning_rate": 0.00011702435557223987,
+      "loss": 0.43,
+      "step": 1045
+    },
+    {
+      "epoch": 0.4709590274651058,
+      "grad_norm": 0.1037481352686882,
+      "learning_rate": 0.00011687838071625685,
+      "loss": 0.3655,
+      "step": 1046
+    },
+    {
+      "epoch": 0.4714092751013057,
+      "grad_norm": 0.11314115673303604,
+      "learning_rate": 0.0001167323688306664,
+      "loss": 0.4968,
+      "step": 1047
+    },
+    {
+      "epoch": 0.47185952273750564,
+      "grad_norm": 0.12661971151828766,
+      "learning_rate": 0.00011658632023580516,
+      "loss": 0.5765,
+      "step": 1048
+    },
+    {
+      "epoch": 0.47230977037370553,
+      "grad_norm": 0.13820421695709229,
+      "learning_rate": 0.00011644023525209016,
+      "loss": 0.4683,
+      "step": 1049
+    },
+    {
+      "epoch": 0.47276001800990547,
+      "grad_norm": 0.10357666015625,
+      "learning_rate": 0.00011629411420001841,
+      "loss": 0.5269,
+      "step": 1050
+    },
+    {
+      "epoch": 0.47321026564610535,
+      "grad_norm": 0.14196226000785828,
+      "learning_rate": 0.000116147957400166,
+      "loss": 0.5179,
+      "step": 1051
+    },
+    {
+      "epoch": 0.4736605132823053,
+      "grad_norm": 0.11081206053495407,
+      "learning_rate": 0.00011600176517318741,
+      "loss": 0.5293,
+      "step": 1052
+    },
+    {
+      "epoch": 0.4741107609185052,
+      "grad_norm": 0.24888749420642853,
+      "learning_rate": 0.00011585553783981486,
+      "loss": 0.6826,
+      "step": 1053
+    },
+    {
+      "epoch": 0.47456100855470507,
+      "grad_norm": 0.19636018574237823,
+      "learning_rate": 0.00011570927572085766,
+      "loss": 0.6607,
+      "step": 1054
+    },
+    {
+      "epoch": 0.475011256190905,
+      "grad_norm": 0.1011781096458435,
+      "learning_rate": 0.00011556297913720137,
+      "loss": 0.3881,
+      "step": 1055
+    },
+    {
+      "epoch": 0.4754615038271049,
+      "grad_norm": 0.19313034415245056,
+      "learning_rate": 0.00011541664840980715,
+      "loss": 0.5694,
+      "step": 1056
+    },
+    {
+      "epoch": 0.47591175146330483,
+      "grad_norm": 0.0900929644703865,
+      "learning_rate": 0.00011527028385971109,
+      "loss": 0.5063,
+      "step": 1057
+    },
+    {
+      "epoch": 0.4763619990995047,
+      "grad_norm": 0.2087116539478302,
+      "learning_rate": 0.00011512388580802349,
+      "loss": 0.9138,
+      "step": 1058
+    },
+    {
+      "epoch": 0.47681224673570466,
+      "grad_norm": 0.12531067430973053,
+      "learning_rate": 0.00011497745457592816,
+      "loss": 0.7122,
+      "step": 1059
+    },
+    {
+      "epoch": 0.47726249437190454,
+      "grad_norm": 0.14638039469718933,
+      "learning_rate": 0.00011483099048468168,
+      "loss": 0.773,
+      "step": 1060
+    },
+    {
+      "epoch": 0.4777127420081045,
+      "grad_norm": 0.12527123093605042,
+      "learning_rate": 0.00011468449385561272,
+      "loss": 0.5656,
+      "step": 1061
+    },
+    {
+      "epoch": 0.47816298964430437,
+      "grad_norm": 0.12665505707263947,
+      "learning_rate": 0.00011453796501012142,
+      "loss": 0.8902,
+      "step": 1062
+    },
+    {
+      "epoch": 0.47861323728050426,
+      "grad_norm": 0.11346755921840668,
+      "learning_rate": 0.0001143914042696784,
+      "loss": 0.6606,
+      "step": 1063
+    },
+    {
+      "epoch": 0.4790634849167042,
+      "grad_norm": 0.14631487429141998,
+      "learning_rate": 0.00011424481195582446,
+      "loss": 0.9405,
+      "step": 1064
+    },
+    {
+      "epoch": 0.4795137325529041,
+      "grad_norm": 0.09528237581253052,
+      "learning_rate": 0.0001140981883901696,
+      "loss": 0.6703,
+      "step": 1065
+    },
+    {
+      "epoch": 0.479963980189104,
+      "grad_norm": 0.17060686647891998,
+      "learning_rate": 0.00011395153389439233,
+      "loss": 0.5835,
+      "step": 1066
+    },
+    {
+      "epoch": 0.4804142278253039,
+      "grad_norm": 0.08870841562747955,
+      "learning_rate": 0.00011380484879023905,
+      "loss": 0.6098,
+      "step": 1067
+    },
+    {
+      "epoch": 0.48086447546150385,
+      "grad_norm": 0.1041567251086235,
+      "learning_rate": 0.00011365813339952336,
+      "loss": 0.5288,
+      "step": 1068
+    },
+    {
+      "epoch": 0.48131472309770373,
+      "grad_norm": 0.13088594377040863,
+      "learning_rate": 0.00011351138804412524,
+      "loss": 0.6405,
+      "step": 1069
+    },
+    {
+      "epoch": 0.4817649707339037,
+      "grad_norm": 0.17708148062229156,
+      "learning_rate": 0.00011336461304599049,
+      "loss": 0.6575,
+      "step": 1070
+    },
+    {
+      "epoch": 0.48221521837010356,
+      "grad_norm": 0.09144304692745209,
+      "learning_rate": 0.00011321780872712984,
+      "loss": 0.348,
+      "step": 1071
+    },
+    {
+      "epoch": 0.48266546600630345,
+      "grad_norm": 0.13802246749401093,
+      "learning_rate": 0.00011307097540961839,
+      "loss": 0.4644,
+      "step": 1072
+    },
+    {
+      "epoch": 0.4831157136425034,
+      "grad_norm": 0.1594877541065216,
+      "learning_rate": 0.0001129241134155949,
+      "loss": 0.7823,
+      "step": 1073
+    },
+    {
+      "epoch": 0.48356596127870327,
+      "grad_norm": 0.11637808382511139,
+      "learning_rate": 0.00011277722306726104,
+      "loss": 0.3233,
+      "step": 1074
+    },
+    {
+      "epoch": 0.4840162089149032,
+      "grad_norm": 0.13459697365760803,
+      "learning_rate": 0.00011263030468688058,
+      "loss": 0.8098,
+      "step": 1075
+    },
+    {
+      "epoch": 0.4844664565511031,
+      "grad_norm": 0.18186055123806,
+      "learning_rate": 0.00011248335859677893,
+      "loss": 0.9328,
+      "step": 1076
+    },
+    {
+      "epoch": 0.48491670418730304,
+      "grad_norm": 0.1338060200214386,
+      "learning_rate": 0.0001123363851193422,
+      "loss": 0.9175,
+      "step": 1077
+    },
+    {
+      "epoch": 0.4853669518235029,
+      "grad_norm": 0.13376158475875854,
+      "learning_rate": 0.00011218938457701661,
+      "loss": 0.4527,
+      "step": 1078
+    },
+    {
+      "epoch": 0.48581719945970286,
+      "grad_norm": 0.11007722467184067,
+      "learning_rate": 0.00011204235729230781,
+      "loss": 0.5122,
+      "step": 1079
+    },
+    {
+      "epoch": 0.48626744709590275,
+      "grad_norm": 0.14273898303508759,
+      "learning_rate": 0.00011189530358778005,
+      "loss": 0.867,
+      "step": 1080
+    },
+    {
+      "epoch": 0.48671769473210263,
+      "grad_norm": 0.10047457367181778,
+      "learning_rate": 0.00011174822378605551,
+      "loss": 0.4891,
+      "step": 1081
+    },
+    {
+      "epoch": 0.4871679423683026,
+      "grad_norm": 0.1287500560283661,
+      "learning_rate": 0.00011160111820981381,
+      "loss": 0.4082,
+      "step": 1082
+    },
+    {
+      "epoch": 0.48761819000450246,
+      "grad_norm": 0.22643907368183136,
+      "learning_rate": 0.00011145398718179088,
+      "loss": 0.5939,
+      "step": 1083
+    },
+    {
+      "epoch": 0.4880684376407024,
+      "grad_norm": 0.138490229845047,
+      "learning_rate": 0.00011130683102477864,
+      "loss": 0.4414,
+      "step": 1084
+    },
+    {
+      "epoch": 0.4885186852769023,
+      "grad_norm": 0.1002749353647232,
+      "learning_rate": 0.00011115965006162405,
+      "loss": 0.4362,
+      "step": 1085
+    },
+    {
+      "epoch": 0.4889689329131022,
+      "grad_norm": 0.1642259657382965,
+      "learning_rate": 0.00011101244461522861,
+      "loss": 0.7746,
+      "step": 1086
+    },
+    {
+      "epoch": 0.4894191805493021,
+      "grad_norm": 0.15255208313465118,
+      "learning_rate": 0.00011086521500854745,
+      "loss": 0.4927,
+      "step": 1087
+    },
+    {
+      "epoch": 0.48986942818550205,
+      "grad_norm": 0.10778426378965378,
+      "learning_rate": 0.00011071796156458869,
+      "loss": 0.2738,
+      "step": 1088
+    },
+    {
+      "epoch": 0.49031967582170194,
+      "grad_norm": 0.09920228272676468,
+      "learning_rate": 0.00011057068460641283,
+      "loss": 0.6391,
+      "step": 1089
+    },
+    {
+      "epoch": 0.4907699234579018,
+      "grad_norm": 0.16423043608665466,
+      "learning_rate": 0.00011042338445713186,
+      "loss": 0.6938,
+      "step": 1090
+    },
+    {
+      "epoch": 0.49122017109410177,
+      "grad_norm": 0.19821050763130188,
+      "learning_rate": 0.00011027606143990867,
+      "loss": 0.7274,
+      "step": 1091
+    },
+    {
+      "epoch": 0.49167041873030165,
+      "grad_norm": 0.16431446373462677,
+      "learning_rate": 0.00011012871587795639,
+      "loss": 0.5661,
+      "step": 1092
+    },
+    {
+      "epoch": 0.4921206663665016,
+      "grad_norm": 0.13417373597621918,
+      "learning_rate": 0.00010998134809453757,
+      "loss": 0.9536,
+      "step": 1093
+    },
+    {
+      "epoch": 0.4925709140027015,
+      "grad_norm": 0.14543132483959198,
+      "learning_rate": 0.00010983395841296348,
+      "loss": 0.7977,
+      "step": 1094
+    },
+    {
+      "epoch": 0.4930211616389014,
+      "grad_norm": 0.10510304570198059,
+      "learning_rate": 0.00010968654715659348,
+      "loss": 0.5033,
+      "step": 1095
+    },
+    {
+      "epoch": 0.4934714092751013,
+      "grad_norm": 0.19212384521961212,
+      "learning_rate": 0.00010953911464883422,
+      "loss": 0.6667,
+      "step": 1096
+    },
+    {
+      "epoch": 0.49392165691130124,
+      "grad_norm": 0.1124984547495842,
+      "learning_rate": 0.00010939166121313902,
+      "loss": 0.7691,
+      "step": 1097
+    },
+    {
+      "epoch": 0.49437190454750113,
+      "grad_norm": 0.16706517338752747,
+      "learning_rate": 0.00010924418717300709,
+      "loss": 0.7701,
+      "step": 1098
+    },
+    {
+      "epoch": 0.494822152183701,
+      "grad_norm": 0.12648510932922363,
+      "learning_rate": 0.0001090966928519828,
+      "loss": 0.6337,
+      "step": 1099
+    },
+    {
+      "epoch": 0.49527239981990095,
+      "grad_norm": 0.11845281720161438,
+      "learning_rate": 0.00010894917857365512,
+      "loss": 0.7376,
+      "step": 1100
+    },
+    {
+      "epoch": 0.49572264745610084,
+      "grad_norm": 0.10094430297613144,
+      "learning_rate": 0.00010880164466165674,
+      "loss": 0.6233,
+      "step": 1101
+    },
+    {
+      "epoch": 0.4961728950923008,
+      "grad_norm": 0.16018596291542053,
+      "learning_rate": 0.00010865409143966338,
+      "loss": 0.8062,
+      "step": 1102
+    },
+    {
+      "epoch": 0.49662314272850067,
+      "grad_norm": 0.17076793313026428,
+      "learning_rate": 0.00010850651923139319,
+      "loss": 0.7614,
+      "step": 1103
+    },
+    {
+      "epoch": 0.4970733903647006,
+      "grad_norm": 0.1536262333393097,
+      "learning_rate": 0.000108358928360606,
+      "loss": 0.4389,
+      "step": 1104
+    },
+    {
+      "epoch": 0.4975236380009005,
+      "grad_norm": 0.14144901931285858,
+      "learning_rate": 0.00010821131915110248,
+      "loss": 0.5877,
+      "step": 1105
+    },
+    {
+      "epoch": 0.49797388563710043,
+      "grad_norm": 0.13459810614585876,
+      "learning_rate": 0.00010806369192672361,
+      "loss": 1.0005,
+      "step": 1106
+    },
+    {
+      "epoch": 0.4984241332733003,
+      "grad_norm": 0.13351495563983917,
+      "learning_rate": 0.00010791604701134989,
+      "loss": 0.515,
+      "step": 1107
+    },
+    {
+      "epoch": 0.4988743809095002,
+      "grad_norm": 0.14650779962539673,
+      "learning_rate": 0.00010776838472890065,
+      "loss": 0.7792,
+      "step": 1108
+    },
+    {
+      "epoch": 0.49932462854570014,
+      "grad_norm": 0.10190403461456299,
+      "learning_rate": 0.00010762070540333323,
+      "loss": 0.5357,
+      "step": 1109
+    },
+    {
+      "epoch": 0.49977487618190003,
+      "grad_norm": 0.17262737452983856,
+      "learning_rate": 0.00010747300935864243,
+      "loss": 0.8398,
+      "step": 1110
+    },
+    {
+      "epoch": 0.5002251238180999,
+      "grad_norm": 0.1292974203824997,
+      "learning_rate": 0.00010732529691885977,
+      "loss": 0.7456,
+      "step": 1111
+    },
+    {
+      "epoch": 0.5006753714542999,
+      "grad_norm": 0.12028542160987854,
+      "learning_rate": 0.00010717756840805265,
+      "loss": 0.686,
+      "step": 1112
+    },
+    {
+      "epoch": 0.5011256190904998,
+      "grad_norm": 0.13336226344108582,
+      "learning_rate": 0.00010702982415032379,
+      "loss": 0.5588,
+      "step": 1113
+    },
+    {
+      "epoch": 0.5015758667266997,
+      "grad_norm": 0.14129920303821564,
+      "learning_rate": 0.00010688206446981041,
+      "loss": 0.7421,
+      "step": 1114
+    },
+    {
+      "epoch": 0.5020261143628996,
+      "grad_norm": 0.1906348317861557,
+      "learning_rate": 0.00010673428969068364,
+      "loss": 0.6188,
+      "step": 1115
+    },
+    {
+      "epoch": 0.5024763619990995,
+      "grad_norm": 0.12442957609891891,
+      "learning_rate": 0.00010658650013714767,
+      "loss": 0.5726,
+      "step": 1116
+    },
+    {
+      "epoch": 0.5029266096352994,
+      "grad_norm": 0.15517951548099518,
+      "learning_rate": 0.00010643869613343906,
+      "loss": 0.5136,
+      "step": 1117
+    },
+    {
+      "epoch": 0.5033768572714993,
+      "grad_norm": 0.10180553793907166,
+      "learning_rate": 0.00010629087800382622,
+      "loss": 0.4248,
+      "step": 1118
+    },
+    {
+      "epoch": 0.5038271049076992,
+      "grad_norm": 0.20831447839736938,
+      "learning_rate": 0.00010614304607260846,
+      "loss": 0.5419,
+      "step": 1119
+    },
+    {
+      "epoch": 0.5042773525438992,
+      "grad_norm": 0.12829376757144928,
+      "learning_rate": 0.0001059952006641153,
+      "loss": 0.5596,
+      "step": 1120
+    },
+    {
+      "epoch": 0.5047276001800991,
+      "grad_norm": 0.16832290589809418,
+      "learning_rate": 0.00010584734210270598,
+      "loss": 0.7504,
+      "step": 1121
+    },
+    {
+      "epoch": 0.5051778478162989,
+      "grad_norm": 0.17684654891490936,
+      "learning_rate": 0.00010569947071276847,
+      "loss": 0.7047,
+      "step": 1122
+    },
+    {
+      "epoch": 0.5056280954524989,
+      "grad_norm": 0.10829100757837296,
+      "learning_rate": 0.00010555158681871898,
+      "loss": 0.4419,
+      "step": 1123
+    },
+    {
+      "epoch": 0.5060783430886988,
+      "grad_norm": 0.13380323350429535,
+      "learning_rate": 0.00010540369074500105,
+      "loss": 0.6631,
+      "step": 1124
+    },
+    {
+      "epoch": 0.5065285907248986,
+      "grad_norm": 0.09989731013774872,
+      "learning_rate": 0.00010525578281608503,
+      "loss": 0.6089,
+      "step": 1125
+    },
+    {
+      "epoch": 0.5069788383610986,
+      "grad_norm": 0.11359839141368866,
+      "learning_rate": 0.00010510786335646725,
+      "loss": 0.8323,
+      "step": 1126
+    },
+    {
+      "epoch": 0.5074290859972985,
+      "grad_norm": 0.11239267885684967,
+      "learning_rate": 0.00010495993269066935,
+      "loss": 0.4295,
+      "step": 1127
+    },
+    {
+      "epoch": 0.5078793336334985,
+      "grad_norm": 0.1148342713713646,
+      "learning_rate": 0.00010481199114323748,
+      "loss": 0.4813,
+      "step": 1128
+    },
+    {
+      "epoch": 0.5083295812696983,
+      "grad_norm": 0.18241752684116364,
+      "learning_rate": 0.00010466403903874176,
+      "loss": 0.5096,
+      "step": 1129
+    },
+    {
+      "epoch": 0.5087798289058982,
+      "grad_norm": 0.133066788315773,
+      "learning_rate": 0.00010451607670177544,
+      "loss": 0.4113,
+      "step": 1130
+    },
+    {
+      "epoch": 0.5092300765420982,
+      "grad_norm": 0.11908308416604996,
+      "learning_rate": 0.00010436810445695422,
+      "loss": 0.6327,
+      "step": 1131
+    },
+    {
+      "epoch": 0.5096803241782981,
+      "grad_norm": 0.11956717818975449,
+      "learning_rate": 0.0001042201226289155,
+      "loss": 0.6731,
+      "step": 1132
+    },
+    {
+      "epoch": 0.510130571814498,
+      "grad_norm": 0.0867386981844902,
+      "learning_rate": 0.00010407213154231774,
+      "loss": 0.4159,
+      "step": 1133
+    },
+    {
+      "epoch": 0.5105808194506979,
+      "grad_norm": 0.13541851937770844,
+      "learning_rate": 0.00010392413152183973,
+      "loss": 0.4486,
+      "step": 1134
+    },
+    {
+      "epoch": 0.5110310670868978,
+      "grad_norm": 0.13011570274829865,
+      "learning_rate": 0.00010377612289217983,
+      "loss": 0.6851,
+      "step": 1135
+    },
+    {
+      "epoch": 0.5114813147230977,
+      "grad_norm": 0.11855112761259079,
+      "learning_rate": 0.00010362810597805526,
+      "loss": 0.6893,
+      "step": 1136
+    },
+    {
+      "epoch": 0.5119315623592976,
+      "grad_norm": 0.15503288805484772,
+      "learning_rate": 0.00010348008110420149,
+      "loss": 0.6545,
+      "step": 1137
+    },
+    {
+      "epoch": 0.5123818099954975,
+      "grad_norm": 0.10951442271471024,
+      "learning_rate": 0.00010333204859537143,
+      "loss": 0.7358,
+      "step": 1138
+    },
+    {
+      "epoch": 0.5128320576316975,
+      "grad_norm": 0.1102096363902092,
+      "learning_rate": 0.00010318400877633467,
+      "loss": 0.5409,
+      "step": 1139
+    },
+    {
+      "epoch": 0.5132823052678973,
+      "grad_norm": 0.11721492558717728,
+      "learning_rate": 0.0001030359619718769,
+      "loss": 0.5645,
+      "step": 1140
+    },
+    {
+      "epoch": 0.5137325529040973,
+      "grad_norm": 0.1417810171842575,
+      "learning_rate": 0.00010288790850679918,
+      "loss": 0.5552,
+      "step": 1141
+    },
+    {
+      "epoch": 0.5141828005402972,
+      "grad_norm": 0.13170863687992096,
+      "learning_rate": 0.00010273984870591706,
+      "loss": 0.5917,
+      "step": 1142
+    },
+    {
+      "epoch": 0.514633048176497,
+      "grad_norm": 0.12531635165214539,
+      "learning_rate": 0.00010259178289406011,
+      "loss": 0.542,
+      "step": 1143
+    },
+    {
+      "epoch": 0.515083295812697,
+      "grad_norm": 0.1103934794664383,
+      "learning_rate": 0.000102443711396071,
+      "loss": 0.5718,
+      "step": 1144
+    },
+    {
+      "epoch": 0.5155335434488969,
+      "grad_norm": 0.11113052070140839,
+      "learning_rate": 0.00010229563453680496,
+      "loss": 0.4596,
+      "step": 1145
+    },
+    {
+      "epoch": 0.5159837910850968,
+      "grad_norm": 0.11733509600162506,
+      "learning_rate": 0.0001021475526411289,
+      "loss": 0.571,
+      "step": 1146
+    },
+    {
+      "epoch": 0.5164340387212967,
+      "grad_norm": 0.12669098377227783,
+      "learning_rate": 0.00010199946603392079,
+      "loss": 0.5691,
+      "step": 1147
+    },
+    {
+      "epoch": 0.5168842863574966,
+      "grad_norm": 0.11662288755178452,
+      "learning_rate": 0.00010185137504006901,
+      "loss": 0.5428,
+      "step": 1148
+    },
+    {
+      "epoch": 0.5173345339936966,
+      "grad_norm": 0.14257319271564484,
+      "learning_rate": 0.00010170327998447149,
+      "loss": 0.5023,
+      "step": 1149
+    },
+    {
+      "epoch": 0.5177847816298965,
+      "grad_norm": 0.10827930271625519,
+      "learning_rate": 0.0001015551811920351,
+      "loss": 0.5743,
+      "step": 1150
+    },
+    {
+      "epoch": 0.5182350292660963,
+      "grad_norm": 0.13265033066272736,
+      "learning_rate": 0.0001014070789876749,
+      "loss": 0.7526,
+      "step": 1151
+    },
+    {
+      "epoch": 0.5186852769022963,
+      "grad_norm": 0.15612582862377167,
+      "learning_rate": 0.00010125897369631344,
+      "loss": 0.5247,
+      "step": 1152
+    },
+    {
+      "epoch": 0.5191355245384962,
+      "grad_norm": 0.12089035660028458,
+      "learning_rate": 0.00010111086564288004,
+      "loss": 0.472,
+      "step": 1153
+    },
+    {
+      "epoch": 0.519585772174696,
+      "grad_norm": 0.11215074360370636,
+      "learning_rate": 0.00010096275515231011,
+      "loss": 0.6251,
+      "step": 1154
+    },
+    {
+      "epoch": 0.520036019810896,
+      "grad_norm": 0.1849210560321808,
+      "learning_rate": 0.00010081464254954431,
+      "loss": 0.4897,
+      "step": 1155
+    },
+    {
+      "epoch": 0.5204862674470959,
+      "grad_norm": 0.13270075619220734,
+      "learning_rate": 0.00010066652815952806,
+      "loss": 0.5874,
+      "step": 1156
+    },
+    {
+      "epoch": 0.5209365150832959,
+      "grad_norm": 0.10537293553352356,
+      "learning_rate": 0.00010051841230721065,
+      "loss": 0.6798,
+      "step": 1157
+    },
+    {
+      "epoch": 0.5213867627194957,
+      "grad_norm": 0.11898064613342285,
+      "learning_rate": 0.00010037029531754454,
+      "loss": 0.4658,
+      "step": 1158
+    },
+    {
+      "epoch": 0.5218370103556956,
+      "grad_norm": 0.11760295927524567,
+      "learning_rate": 0.00010022217751548471,
+      "loss": 0.4596,
+      "step": 1159
+    },
+    {
+      "epoch": 0.5222872579918956,
+      "grad_norm": 0.14772509038448334,
+      "learning_rate": 0.00010007405922598794,
+      "loss": 0.7379,
+      "step": 1160
+    },
+    {
+      "epoch": 0.5227375056280954,
+      "grad_norm": 0.13493098318576813,
+      "learning_rate": 9.992594077401208e-05,
+      "loss": 0.8497,
+      "step": 1161
+    },
+    {
+      "epoch": 0.5231877532642953,
+      "grad_norm": 0.13291716575622559,
+      "learning_rate": 9.977782248451534e-05,
+      "loss": 0.7088,
+      "step": 1162
+    },
+    {
+      "epoch": 0.5236380009004953,
+      "grad_norm": 0.14112362265586853,
+      "learning_rate": 9.962970468245549e-05,
+      "loss": 0.5106,
+      "step": 1163
+    },
+    {
+      "epoch": 0.5240882485366952,
+      "grad_norm": 0.1459013968706131,
+      "learning_rate": 9.948158769278939e-05,
+      "loss": 0.6222,
+      "step": 1164
+    },
+    {
+      "epoch": 0.524538496172895,
+      "grad_norm": 0.14151379466056824,
+      "learning_rate": 9.933347184047195e-05,
+      "loss": 0.5922,
+      "step": 1165
+    },
+    {
+      "epoch": 0.524988743809095,
+      "grad_norm": 0.172861710190773,
+      "learning_rate": 9.918535745045573e-05,
+      "loss": 0.4753,
+      "step": 1166
+    },
+    {
+      "epoch": 0.5254389914452949,
+      "grad_norm": 0.09632491320371628,
+      "learning_rate": 9.903724484768991e-05,
+      "loss": 0.5194,
+      "step": 1167
+    },
+    {
+      "epoch": 0.5258892390814949,
+      "grad_norm": 0.12447653710842133,
+      "learning_rate": 9.888913435711999e-05,
+      "loss": 0.5278,
+      "step": 1168
+    },
+    {
+      "epoch": 0.5263394867176947,
+      "grad_norm": 0.19318704307079315,
+      "learning_rate": 9.874102630368657e-05,
+      "loss": 0.6995,
+      "step": 1169
+    },
+    {
+      "epoch": 0.5267897343538946,
+      "grad_norm": 0.08978021144866943,
+      "learning_rate": 9.859292101232514e-05,
+      "loss": 0.4182,
+      "step": 1170
+    },
+    {
+      "epoch": 0.5272399819900946,
+      "grad_norm": 0.14166943728923798,
+      "learning_rate": 9.844481880796491e-05,
+      "loss": 0.7091,
+      "step": 1171
+    },
+    {
+      "epoch": 0.5276902296262944,
+      "grad_norm": 0.1576286405324936,
+      "learning_rate": 9.829672001552854e-05,
+      "loss": 0.7152,
+      "step": 1172
+    },
+    {
+      "epoch": 0.5281404772624944,
+      "grad_norm": 0.1699717938899994,
+      "learning_rate": 9.814862495993101e-05,
+      "loss": 0.6032,
+      "step": 1173
+    },
+    {
+      "epoch": 0.5285907248986943,
+      "grad_norm": 0.14461272954940796,
+      "learning_rate": 9.800053396607921e-05,
+      "loss": 0.7823,
+      "step": 1174
+    },
+    {
+      "epoch": 0.5290409725348942,
+      "grad_norm": 0.11832782626152039,
+      "learning_rate": 9.785244735887114e-05,
+      "loss": 0.7521,
+      "step": 1175
+    },
+    {
+      "epoch": 0.5294912201710941,
+      "grad_norm": 0.10224983096122742,
+      "learning_rate": 9.770436546319505e-05,
+      "loss": 0.5624,
+      "step": 1176
+    },
+    {
+      "epoch": 0.529941467807294,
+      "grad_norm": 0.15683287382125854,
+      "learning_rate": 9.755628860392902e-05,
+      "loss": 0.4633,
+      "step": 1177
+    },
+    {
+      "epoch": 0.530391715443494,
+      "grad_norm": 0.0979384332895279,
+      "learning_rate": 9.740821710593989e-05,
+      "loss": 0.3497,
+      "step": 1178
+    },
+    {
+      "epoch": 0.5308419630796938,
+      "grad_norm": 0.20489826798439026,
+      "learning_rate": 9.726015129408296e-05,
+      "loss": 0.9516,
+      "step": 1179
+    },
+    {
+      "epoch": 0.5312922107158937,
+      "grad_norm": 0.1254095435142517,
+      "learning_rate": 9.711209149320083e-05,
+      "loss": 0.708,
+      "step": 1180
+    },
+    {
+      "epoch": 0.5317424583520937,
+      "grad_norm": 0.14179953932762146,
+      "learning_rate": 9.696403802812312e-05,
+      "loss": 0.7913,
+      "step": 1181
+    },
+    {
+      "epoch": 0.5321927059882936,
+      "grad_norm": 0.1712818741798401,
+      "learning_rate": 9.681599122366534e-05,
+      "loss": 0.6506,
+      "step": 1182
+    },
+    {
+      "epoch": 0.5326429536244934,
+      "grad_norm": 0.1529311239719391,
+      "learning_rate": 9.66679514046286e-05,
+      "loss": 0.4993,
+      "step": 1183
+    },
+    {
+      "epoch": 0.5330932012606934,
+      "grad_norm": 0.1869378387928009,
+      "learning_rate": 9.651991889579851e-05,
+      "loss": 0.6196,
+      "step": 1184
+    },
+    {
+      "epoch": 0.5335434488968933,
+      "grad_norm": 0.11333781480789185,
+      "learning_rate": 9.637189402194476e-05,
+      "loss": 0.6486,
+      "step": 1185
+    },
+    {
+      "epoch": 0.5339936965330933,
+      "grad_norm": 0.11273683607578278,
+      "learning_rate": 9.622387710782019e-05,
+      "loss": 0.5976,
+      "step": 1186
+    },
+    {
+      "epoch": 0.5344439441692931,
+      "grad_norm": 0.1207125335931778,
+      "learning_rate": 9.60758684781603e-05,
+      "loss": 0.6036,
+      "step": 1187
+    },
+    {
+      "epoch": 0.534894191805493,
+      "grad_norm": 0.10002782195806503,
+      "learning_rate": 9.592786845768227e-05,
+      "loss": 0.5571,
+      "step": 1188
+    },
+    {
+      "epoch": 0.535344439441693,
+      "grad_norm": 0.19848157465457916,
+      "learning_rate": 9.577987737108454e-05,
+      "loss": 0.6273,
+      "step": 1189
+    },
+    {
+      "epoch": 0.5357946870778928,
+      "grad_norm": 0.1344083994626999,
+      "learning_rate": 9.563189554304579e-05,
+      "loss": 0.5679,
+      "step": 1190
+    },
+    {
+      "epoch": 0.5362449347140927,
+      "grad_norm": 0.09527532756328583,
+      "learning_rate": 9.548392329822457e-05,
+      "loss": 0.4674,
+      "step": 1191
+    },
+    {
+      "epoch": 0.5366951823502927,
+      "grad_norm": 0.22306379675865173,
+      "learning_rate": 9.533596096125825e-05,
+      "loss": 0.5244,
+      "step": 1192
+    },
+    {
+      "epoch": 0.5371454299864926,
+      "grad_norm": 0.12183703482151031,
+      "learning_rate": 9.518800885676257e-05,
+      "loss": 0.5404,
+      "step": 1193
+    },
+    {
+      "epoch": 0.5375956776226924,
+      "grad_norm": 0.12995265424251556,
+      "learning_rate": 9.504006730933069e-05,
+      "loss": 0.4851,
+      "step": 1194
+    },
+    {
+      "epoch": 0.5380459252588924,
+      "grad_norm": 0.2176143378019333,
+      "learning_rate": 9.489213664353276e-05,
+      "loss": 0.5169,
+      "step": 1195
+    },
+    {
+      "epoch": 0.5384961728950923,
+      "grad_norm": 0.14356349408626556,
+      "learning_rate": 9.474421718391499e-05,
+      "loss": 0.5394,
+      "step": 1196
+    },
+    {
+      "epoch": 0.5389464205312922,
+      "grad_norm": 0.09458054602146149,
+      "learning_rate": 9.459630925499898e-05,
+      "loss": 0.4527,
+      "step": 1197
+    },
+    {
+      "epoch": 0.5393966681674921,
+      "grad_norm": 0.10279335826635361,
+      "learning_rate": 9.444841318128103e-05,
+      "loss": 0.4283,
+      "step": 1198
+    },
+    {
+      "epoch": 0.539846915803692,
+      "grad_norm": 0.13163302838802338,
+      "learning_rate": 9.430052928723153e-05,
+      "loss": 0.6693,
+      "step": 1199
+    },
+    {
+      "epoch": 0.540297163439892,
+      "grad_norm": 0.10580742359161377,
+      "learning_rate": 9.415265789729403e-05,
+      "loss": 0.4613,
+      "step": 1200
+    },
+    {
+      "epoch": 0.5407474110760918,
+      "grad_norm": 0.16433700919151306,
+      "learning_rate": 9.400479933588469e-05,
+      "loss": 0.5168,
+      "step": 1201
+    },
+    {
+      "epoch": 0.5411976587122918,
+      "grad_norm": 0.12860211730003357,
+      "learning_rate": 9.385695392739156e-05,
+      "loss": 0.6029,
+      "step": 1202
+    },
+    {
+      "epoch": 0.5416479063484917,
+      "grad_norm": 0.14047440886497498,
+      "learning_rate": 9.370912199617378e-05,
+      "loss": 0.4468,
+      "step": 1203
+    },
+    {
+      "epoch": 0.5420981539846916,
+      "grad_norm": 0.10742084681987762,
+      "learning_rate": 9.356130386656095e-05,
+      "loss": 0.5495,
+      "step": 1204
+    },
+    {
+      "epoch": 0.5425484016208915,
+      "grad_norm": 0.10444051772356033,
+      "learning_rate": 9.341349986285234e-05,
+      "loss": 0.6552,
+      "step": 1205
+    },
+    {
+      "epoch": 0.5429986492570914,
+      "grad_norm": 0.15860235691070557,
+      "learning_rate": 9.326571030931637e-05,
+      "loss": 0.6142,
+      "step": 1206
+    },
+    {
+      "epoch": 0.5434488968932913,
+      "grad_norm": 0.10210921615362167,
+      "learning_rate": 9.31179355301896e-05,
+      "loss": 0.4398,
+      "step": 1207
+    },
+    {
+      "epoch": 0.5438991445294912,
+      "grad_norm": 0.11654400080442429,
+      "learning_rate": 9.297017584967625e-05,
+      "loss": 0.5731,
+      "step": 1208
+    },
+    {
+      "epoch": 0.5443493921656911,
+      "grad_norm": 0.12270856648683548,
+      "learning_rate": 9.282243159194735e-05,
+      "loss": 0.3934,
+      "step": 1209
+    },
+    {
+      "epoch": 0.5447996398018911,
+      "grad_norm": 0.20615768432617188,
+      "learning_rate": 9.267470308114025e-05,
+      "loss": 0.7443,
+      "step": 1210
+    },
+    {
+      "epoch": 0.545249887438091,
+      "grad_norm": 0.12823031842708588,
+      "learning_rate": 9.252699064135758e-05,
+      "loss": 0.5324,
+      "step": 1211
+    },
+    {
+      "epoch": 0.5457001350742908,
+      "grad_norm": 0.10071907192468643,
+      "learning_rate": 9.237929459666681e-05,
+      "loss": 0.3863,
+      "step": 1212
+    },
+    {
+      "epoch": 0.5461503827104908,
+      "grad_norm": 0.13511130213737488,
+      "learning_rate": 9.223161527109937e-05,
+      "loss": 0.5122,
+      "step": 1213
+    },
+    {
+      "epoch": 0.5466006303466907,
+      "grad_norm": 0.17224298417568207,
+      "learning_rate": 9.208395298865014e-05,
+      "loss": 0.7788,
+      "step": 1214
+    },
+    {
+      "epoch": 0.5470508779828905,
+      "grad_norm": 0.11435385793447495,
+      "learning_rate": 9.193630807327641e-05,
+      "loss": 0.368,
+      "step": 1215
+    },
+    {
+      "epoch": 0.5475011256190905,
+      "grad_norm": 0.1763196587562561,
+      "learning_rate": 9.178868084889757e-05,
+      "loss": 0.8301,
+      "step": 1216
+    },
+    {
+      "epoch": 0.5479513732552904,
+      "grad_norm": 0.17762482166290283,
+      "learning_rate": 9.164107163939403e-05,
+      "loss": 0.5631,
+      "step": 1217
+    },
+    {
+      "epoch": 0.5484016208914904,
+      "grad_norm": 0.10984458029270172,
+      "learning_rate": 9.149348076860685e-05,
+      "loss": 0.5252,
+      "step": 1218
+    },
+    {
+      "epoch": 0.5488518685276902,
+      "grad_norm": 0.1165989339351654,
+      "learning_rate": 9.134590856033664e-05,
+      "loss": 0.5397,
+      "step": 1219
+    },
+    {
+      "epoch": 0.5493021161638901,
+      "grad_norm": 0.12493187189102173,
+      "learning_rate": 9.119835533834331e-05,
+      "loss": 0.5137,
+      "step": 1220
+    },
+    {
+      "epoch": 0.5497523638000901,
+      "grad_norm": 0.15735436975955963,
+      "learning_rate": 9.10508214263449e-05,
+      "loss": 0.4559,
+      "step": 1221
+    },
+    {
+      "epoch": 0.55020261143629,
+      "grad_norm": 0.22769418358802795,
+      "learning_rate": 9.090330714801723e-05,
+      "loss": 0.724,
+      "step": 1222
+    },
+    {
+      "epoch": 0.5506528590724898,
+      "grad_norm": 0.14711758494377136,
+      "learning_rate": 9.075581282699295e-05,
+      "loss": 0.4097,
+      "step": 1223
+    },
+    {
+      "epoch": 0.5511031067086898,
+      "grad_norm": 0.1470324993133545,
+      "learning_rate": 9.060833878686099e-05,
+      "loss": 0.6273,
+      "step": 1224
+    },
+    {
+      "epoch": 0.5515533543448897,
+      "grad_norm": 0.11119558662176132,
+      "learning_rate": 9.046088535116582e-05,
+      "loss": 0.3838,
+      "step": 1225
+    },
+    {
+      "epoch": 0.5520036019810896,
+      "grad_norm": 0.15826478600502014,
+      "learning_rate": 9.031345284340653e-05,
+      "loss": 0.5832,
+      "step": 1226
+    },
+    {
+      "epoch": 0.5524538496172895,
+      "grad_norm": 0.15013498067855835,
+      "learning_rate": 9.016604158703654e-05,
+      "loss": 0.6492,
+      "step": 1227
+    },
+    {
+      "epoch": 0.5529040972534894,
+      "grad_norm": 0.11475347727537155,
+      "learning_rate": 9.001865190546245e-05,
+      "loss": 0.5188,
+      "step": 1228
+    },
+    {
+      "epoch": 0.5533543448896894,
+      "grad_norm": 0.14742490649223328,
+      "learning_rate": 8.987128412204364e-05,
+      "loss": 0.9857,
+      "step": 1229
+    },
+    {
+      "epoch": 0.5538045925258892,
+      "grad_norm": 0.1260511726140976,
+      "learning_rate": 8.972393856009133e-05,
+      "loss": 0.6227,
+      "step": 1230
+    },
+    {
+      "epoch": 0.5542548401620891,
+      "grad_norm": 0.11928174644708633,
+      "learning_rate": 8.957661554286818e-05,
+      "loss": 0.3414,
+      "step": 1231
+    },
+    {
+      "epoch": 0.5547050877982891,
+      "grad_norm": 0.14346943795681,
+      "learning_rate": 8.942931539358718e-05,
+      "loss": 0.5204,
+      "step": 1232
+    },
+    {
+      "epoch": 0.5551553354344889,
+      "grad_norm": 0.12197433412075043,
+      "learning_rate": 8.928203843541132e-05,
+      "loss": 0.6769,
+      "step": 1233
+    },
+    {
+      "epoch": 0.5556055830706889,
+      "grad_norm": 0.09538273513317108,
+      "learning_rate": 8.913478499145254e-05,
+      "loss": 0.4859,
+      "step": 1234
+    },
+    {
+      "epoch": 0.5560558307068888,
+      "grad_norm": 0.15390203893184662,
+      "learning_rate": 8.89875553847714e-05,
+      "loss": 0.5945,
+      "step": 1235
+    },
+    {
+      "epoch": 0.5565060783430887,
+      "grad_norm": 0.22316904366016388,
+      "learning_rate": 8.884034993837596e-05,
+      "loss": 0.614,
+      "step": 1236
+    },
+    {
+      "epoch": 0.5569563259792886,
+      "grad_norm": 0.11698149889707565,
+      "learning_rate": 8.869316897522141e-05,
+      "loss": 0.562,
+      "step": 1237
+    },
+    {
+      "epoch": 0.5574065736154885,
+      "grad_norm": 0.13198155164718628,
+      "learning_rate": 8.854601281820914e-05,
+      "loss": 0.4711,
+      "step": 1238
+    },
+    {
+      "epoch": 0.5578568212516885,
+      "grad_norm": 0.15348884463310242,
+      "learning_rate": 8.839888179018623e-05,
+      "loss": 0.8197,
+      "step": 1239
+    },
+    {
+      "epoch": 0.5583070688878884,
+      "grad_norm": 0.13334012031555176,
+      "learning_rate": 8.82517762139445e-05,
+      "loss": 0.396,
+      "step": 1240
+    },
+    {
+      "epoch": 0.5587573165240882,
+      "grad_norm": 0.10079912841320038,
+      "learning_rate": 8.810469641222001e-05,
+      "loss": 0.404,
+      "step": 1241
+    },
+    {
+      "epoch": 0.5592075641602882,
+      "grad_norm": 0.12311876565217972,
+      "learning_rate": 8.795764270769221e-05,
+      "loss": 0.4955,
+      "step": 1242
+    },
+    {
+      "epoch": 0.5596578117964881,
+      "grad_norm": 0.11680926382541656,
+      "learning_rate": 8.781061542298342e-05,
+      "loss": 0.5608,
+      "step": 1243
+    },
+    {
+      "epoch": 0.5601080594326879,
+      "grad_norm": 0.14139245450496674,
+      "learning_rate": 8.766361488065784e-05,
+      "loss": 0.5722,
+      "step": 1244
+    },
+    {
+      "epoch": 0.5605583070688879,
+      "grad_norm": 0.181406632065773,
+      "learning_rate": 8.751664140322114e-05,
+      "loss": 0.6087,
+      "step": 1245
+    },
+    {
+      "epoch": 0.5610085547050878,
+      "grad_norm": 0.1275119036436081,
+      "learning_rate": 8.736969531311944e-05,
+      "loss": 0.38,
+      "step": 1246
+    },
+    {
+      "epoch": 0.5614588023412878,
+      "grad_norm": 0.13071183860301971,
+      "learning_rate": 8.722277693273901e-05,
+      "loss": 0.7024,
+      "step": 1247
+    },
+    {
+      "epoch": 0.5619090499774876,
+      "grad_norm": 0.08929245173931122,
+      "learning_rate": 8.707588658440511e-05,
+      "loss": 0.304,
+      "step": 1248
+    },
+    {
+      "epoch": 0.5623592976136875,
+      "grad_norm": 0.10706759244203568,
+      "learning_rate": 8.69290245903816e-05,
+      "loss": 0.7391,
+      "step": 1249
+    },
+    {
+      "epoch": 0.5628095452498875,
+      "grad_norm": 0.11461903899908066,
+      "learning_rate": 8.678219127287018e-05,
+      "loss": 0.3783,
+      "step": 1250
+    },
+    {
+      "epoch": 0.5632597928860873,
+      "grad_norm": 0.15995772182941437,
+      "learning_rate": 8.663538695400952e-05,
+      "loss": 0.67,
+      "step": 1251
+    },
+    {
+      "epoch": 0.5637100405222872,
+      "grad_norm": 0.10434066504240036,
+      "learning_rate": 8.648861195587477e-05,
+      "loss": 0.5749,
+      "step": 1252
+    },
+    {
+      "epoch": 0.5641602881584872,
+      "grad_norm": 0.13994228839874268,
+      "learning_rate": 8.634186660047663e-05,
+      "loss": 0.5196,
+      "step": 1253
+    },
+    {
+      "epoch": 0.5646105357946871,
+      "grad_norm": 0.13562047481536865,
+      "learning_rate": 8.619515120976098e-05,
+      "loss": 0.6121,
+      "step": 1254
+    },
+    {
+      "epoch": 0.565060783430887,
+      "grad_norm": 0.11193189024925232,
+      "learning_rate": 8.604846610560771e-05,
+      "loss": 0.315,
+      "step": 1255
+    },
+    {
+      "epoch": 0.5655110310670869,
+      "grad_norm": 0.10108640044927597,
+      "learning_rate": 8.590181160983044e-05,
+      "loss": 0.5027,
+      "step": 1256
+    },
+    {
+      "epoch": 0.5659612787032868,
+      "grad_norm": 0.10204221308231354,
+      "learning_rate": 8.575518804417552e-05,
+      "loss": 0.49,
+      "step": 1257
+    },
+    {
+      "epoch": 0.5664115263394868,
+      "grad_norm": 0.13836467266082764,
+      "learning_rate": 8.560859573032161e-05,
+      "loss": 0.7208,
+      "step": 1258
+    },
+    {
+      "epoch": 0.5668617739756866,
+      "grad_norm": 0.10946767777204514,
+      "learning_rate": 8.546203498987862e-05,
+      "loss": 0.5405,
+      "step": 1259
+    },
+    {
+      "epoch": 0.5673120216118865,
+      "grad_norm": 0.15283167362213135,
+      "learning_rate": 8.531550614438729e-05,
+      "loss": 0.9043,
+      "step": 1260
+    },
+    {
+      "epoch": 0.5677622692480865,
+      "grad_norm": 0.162896066904068,
+      "learning_rate": 8.516900951531833e-05,
+      "loss": 0.7976,
+      "step": 1261
+    },
+    {
+      "epoch": 0.5682125168842863,
+      "grad_norm": 0.16917985677719116,
+      "learning_rate": 8.502254542407186e-05,
+      "loss": 0.7434,
+      "step": 1262
+    },
+    {
+      "epoch": 0.5686627645204863,
+      "grad_norm": 0.17442438006401062,
+      "learning_rate": 8.487611419197655e-05,
+      "loss": 0.6368,
+      "step": 1263
+    },
+    {
+      "epoch": 0.5691130121566862,
+      "grad_norm": 0.19976301491260529,
+      "learning_rate": 8.472971614028896e-05,
+      "loss": 0.5351,
+      "step": 1264
+    },
+    {
+      "epoch": 0.5695632597928861,
+      "grad_norm": 0.14719092845916748,
+      "learning_rate": 8.458335159019289e-05,
+      "loss": 0.5733,
+      "step": 1265
+    },
+    {
+      "epoch": 0.570013507429086,
+      "grad_norm": 0.0837155431509018,
+      "learning_rate": 8.443702086279866e-05,
+      "loss": 0.3826,
+      "step": 1266
+    },
+    {
+      "epoch": 0.5704637550652859,
+      "grad_norm": 0.12264306843280792,
+      "learning_rate": 8.429072427914235e-05,
+      "loss": 0.3846,
+      "step": 1267
+    },
+    {
+      "epoch": 0.5709140027014858,
+      "grad_norm": 0.12002010643482208,
+      "learning_rate": 8.414446216018516e-05,
+      "loss": 0.4964,
+      "step": 1268
+    },
+    {
+      "epoch": 0.5713642503376857,
+      "grad_norm": 0.1603572964668274,
+      "learning_rate": 8.399823482681262e-05,
+      "loss": 0.7489,
+      "step": 1269
+    },
+    {
+      "epoch": 0.5718144979738856,
+      "grad_norm": 0.09958768635988235,
+      "learning_rate": 8.385204259983403e-05,
+      "loss": 0.4807,
+      "step": 1270
+    },
+    {
+      "epoch": 0.5722647456100856,
+      "grad_norm": 0.12137196213006973,
+      "learning_rate": 8.370588579998161e-05,
+      "loss": 0.6686,
+      "step": 1271
+    },
+    {
+      "epoch": 0.5727149932462855,
+      "grad_norm": 0.1373930275440216,
+      "learning_rate": 8.355976474790987e-05,
+      "loss": 0.6284,
+      "step": 1272
+    },
+    {
+      "epoch": 0.5731652408824853,
+      "grad_norm": 0.1516687422990799,
+      "learning_rate": 8.341367976419487e-05,
+      "loss": 0.6249,
+      "step": 1273
+    },
+    {
+      "epoch": 0.5736154885186853,
+      "grad_norm": 0.16523873805999756,
+      "learning_rate": 8.32676311693336e-05,
+      "loss": 0.7283,
+      "step": 1274
+    },
+    {
+      "epoch": 0.5740657361548852,
+      "grad_norm": 0.13635611534118652,
+      "learning_rate": 8.312161928374318e-05,
+      "loss": 0.5864,
+      "step": 1275
+    },
+    {
+      "epoch": 0.5745159837910851,
+      "grad_norm": 0.13822101056575775,
+      "learning_rate": 8.297564442776014e-05,
+      "loss": 0.5161,
+      "step": 1276
+    },
+    {
+      "epoch": 0.574966231427285,
+      "grad_norm": 0.17994721233844757,
+      "learning_rate": 8.28297069216399e-05,
+      "loss": 0.9364,
+      "step": 1277
+    },
+    {
+      "epoch": 0.5754164790634849,
+      "grad_norm": 0.11868420243263245,
+      "learning_rate": 8.26838070855558e-05,
+      "loss": 0.3709,
+      "step": 1278
+    },
+    {
+      "epoch": 0.5758667266996849,
+      "grad_norm": 0.12870121002197266,
+      "learning_rate": 8.253794523959863e-05,
+      "loss": 0.6866,
+      "step": 1279
+    },
+    {
+      "epoch": 0.5763169743358847,
+      "grad_norm": 0.27837392687797546,
+      "learning_rate": 8.239212170377577e-05,
+      "loss": 0.8873,
+      "step": 1280
+    },
+    {
+      "epoch": 0.5767672219720846,
+      "grad_norm": 0.129364013671875,
+      "learning_rate": 8.224633679801062e-05,
+      "loss": 0.6094,
+      "step": 1281
+    },
+    {
+      "epoch": 0.5772174696082846,
+      "grad_norm": 0.12454076111316681,
+      "learning_rate": 8.210059084214176e-05,
+      "loss": 0.8812,
+      "step": 1282
+    },
+    {
+      "epoch": 0.5776677172444845,
+      "grad_norm": 0.12894383072853088,
+      "learning_rate": 8.195488415592238e-05,
+      "loss": 0.4985,
+      "step": 1283
+    },
+    {
+      "epoch": 0.5781179648806843,
+      "grad_norm": 0.11755508184432983,
+      "learning_rate": 8.180921705901942e-05,
+      "loss": 0.5535,
+      "step": 1284
+    },
+    {
+      "epoch": 0.5785682125168843,
+      "grad_norm": 0.15028131008148193,
+      "learning_rate": 8.16635898710131e-05,
+      "loss": 0.9306,
+      "step": 1285
+    },
+    {
+      "epoch": 0.5790184601530842,
+      "grad_norm": 0.14813221991062164,
+      "learning_rate": 8.151800291139598e-05,
+      "loss": 0.8058,
+      "step": 1286
+    },
+    {
+      "epoch": 0.579468707789284,
+      "grad_norm": 0.11500914394855499,
+      "learning_rate": 8.137245649957239e-05,
+      "loss": 0.4821,
+      "step": 1287
+    },
+    {
+      "epoch": 0.579918955425484,
+      "grad_norm": 0.1422368288040161,
+      "learning_rate": 8.122695095485769e-05,
+      "loss": 0.7432,
+      "step": 1288
+    },
+    {
+      "epoch": 0.5803692030616839,
+      "grad_norm": 0.13462093472480774,
+      "learning_rate": 8.108148659647765e-05,
+      "loss": 0.6238,
+      "step": 1289
+    },
+    {
+      "epoch": 0.5808194506978839,
+      "grad_norm": 0.10843046754598618,
+      "learning_rate": 8.093606374356759e-05,
+      "loss": 0.3637,
+      "step": 1290
+    },
+    {
+      "epoch": 0.5812696983340837,
+      "grad_norm": 0.10636607557535172,
+      "learning_rate": 8.079068271517183e-05,
+      "loss": 0.2598,
+      "step": 1291
+    },
+    {
+      "epoch": 0.5817199459702836,
+      "grad_norm": 0.14134688675403595,
+      "learning_rate": 8.064534383024285e-05,
+      "loss": 0.5579,
+      "step": 1292
+    },
+    {
+      "epoch": 0.5821701936064836,
+      "grad_norm": 0.15534378588199615,
+      "learning_rate": 8.050004740764083e-05,
+      "loss": 0.7021,
+      "step": 1293
+    },
+    {
+      "epoch": 0.5826204412426835,
+      "grad_norm": 0.15912222862243652,
+      "learning_rate": 8.035479376613261e-05,
+      "loss": 0.5505,
+      "step": 1294
+    },
+    {
+      "epoch": 0.5830706888788834,
+      "grad_norm": 0.12676149606704712,
+      "learning_rate": 8.020958322439133e-05,
+      "loss": 0.3903,
+      "step": 1295
+    },
+    {
+      "epoch": 0.5835209365150833,
+      "grad_norm": 0.11101707071065903,
+      "learning_rate": 8.00644161009954e-05,
+      "loss": 0.588,
+      "step": 1296
+    },
+    {
+      "epoch": 0.5839711841512832,
+      "grad_norm": 0.1225343719124794,
+      "learning_rate": 7.991929271442817e-05,
+      "loss": 0.352,
+      "step": 1297
+    },
+    {
+      "epoch": 0.5844214317874831,
+      "grad_norm": 0.09751452505588531,
+      "learning_rate": 7.977421338307687e-05,
+      "loss": 0.5621,
+      "step": 1298
+    },
+    {
+      "epoch": 0.584871679423683,
+      "grad_norm": 0.1758609563112259,
+      "learning_rate": 7.962917842523217e-05,
+      "loss": 0.4345,
+      "step": 1299
+    },
+    {
+      "epoch": 0.585321927059883,
+      "grad_norm": 0.14067481458187103,
+      "learning_rate": 7.94841881590874e-05,
+      "loss": 0.6434,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5857721746960829,
+      "grad_norm": 0.15941548347473145,
+      "learning_rate": 7.933924290273775e-05,
+      "loss": 0.9133,
+      "step": 1301
+    },
+    {
+      "epoch": 0.5862224223322827,
+      "grad_norm": 0.14021165668964386,
+      "learning_rate": 7.919434297417975e-05,
+      "loss": 0.5827,
+      "step": 1302
+    },
+    {
+      "epoch": 0.5866726699684827,
+      "grad_norm": 0.1197437196969986,
+      "learning_rate": 7.904948869131041e-05,
+      "loss": 0.3414,
+      "step": 1303
+    },
+    {
+      "epoch": 0.5871229176046826,
+      "grad_norm": 0.1488257348537445,
+      "learning_rate": 7.89046803719267e-05,
+      "loss": 0.7457,
+      "step": 1304
+    },
+    {
+      "epoch": 0.5875731652408824,
+      "grad_norm": 0.14734642207622528,
+      "learning_rate": 7.875991833372464e-05,
+      "loss": 0.4718,
+      "step": 1305
+    },
+    {
+      "epoch": 0.5880234128770824,
+      "grad_norm": 0.2054227888584137,
+      "learning_rate": 7.86152028942988e-05,
+      "loss": 0.4495,
+      "step": 1306
+    },
+    {
+      "epoch": 0.5884736605132823,
+      "grad_norm": 0.13476184010505676,
+      "learning_rate": 7.847053437114141e-05,
+      "loss": 0.4934,
+      "step": 1307
+    },
+    {
+      "epoch": 0.5889239081494823,
+      "grad_norm": 0.10180338472127914,
+      "learning_rate": 7.832591308164195e-05,
+      "loss": 0.2717,
+      "step": 1308
+    },
+    {
+      "epoch": 0.5893741557856821,
+      "grad_norm": 0.1598125845193863,
+      "learning_rate": 7.818133934308606e-05,
+      "loss": 0.8283,
+      "step": 1309
+    },
+    {
+      "epoch": 0.589824403421882,
+      "grad_norm": 0.14444632828235626,
+      "learning_rate": 7.803681347265524e-05,
+      "loss": 0.5142,
+      "step": 1310
+    },
+    {
+      "epoch": 0.590274651058082,
+      "grad_norm": 0.13305717706680298,
+      "learning_rate": 7.789233578742582e-05,
+      "loss": 0.4168,
+      "step": 1311
+    },
+    {
+      "epoch": 0.5907248986942819,
+      "grad_norm": 0.18450669944286346,
+      "learning_rate": 7.774790660436858e-05,
+      "loss": 0.682,
+      "step": 1312
+    },
+    {
+      "epoch": 0.5911751463304817,
+      "grad_norm": 0.14253094792366028,
+      "learning_rate": 7.760352624034771e-05,
+      "loss": 0.7342,
+      "step": 1313
+    },
+    {
+      "epoch": 0.5916253939666817,
+      "grad_norm": 0.10962200909852982,
+      "learning_rate": 7.745919501212043e-05,
+      "loss": 0.6097,
+      "step": 1314
+    },
+    {
+      "epoch": 0.5920756416028816,
+      "grad_norm": 0.1199650838971138,
+      "learning_rate": 7.731491323633608e-05,
+      "loss": 0.4628,
+      "step": 1315
+    },
+    {
+      "epoch": 0.5925258892390814,
+      "grad_norm": 0.14307717978954315,
+      "learning_rate": 7.717068122953561e-05,
+      "loss": 0.6345,
+      "step": 1316
+    },
+    {
+      "epoch": 0.5929761368752814,
+      "grad_norm": 0.14080429077148438,
+      "learning_rate": 7.702649930815066e-05,
+      "loss": 0.6036,
+      "step": 1317
+    },
+    {
+      "epoch": 0.5934263845114813,
+      "grad_norm": 0.13833670318126678,
+      "learning_rate": 7.688236778850306e-05,
+      "loss": 0.4574,
+      "step": 1318
+    },
+    {
+      "epoch": 0.5938766321476813,
+      "grad_norm": 0.18195441365242004,
+      "learning_rate": 7.6738286986804e-05,
+      "loss": 0.8573,
+      "step": 1319
+    },
+    {
+      "epoch": 0.5943268797838811,
+      "grad_norm": 0.1248680129647255,
+      "learning_rate": 7.659425721915352e-05,
+      "loss": 0.4484,
+      "step": 1320
+    },
+    {
+      "epoch": 0.594777127420081,
+      "grad_norm": 0.1658993661403656,
+      "learning_rate": 7.645027880153957e-05,
+      "loss": 0.4752,
+      "step": 1321
+    },
+    {
+      "epoch": 0.595227375056281,
+      "grad_norm": 0.1413148194551468,
+      "learning_rate": 7.630635204983751e-05,
+      "loss": 0.4605,
+      "step": 1322
+    },
+    {
+      "epoch": 0.5956776226924808,
+      "grad_norm": 0.18926557898521423,
+      "learning_rate": 7.616247727980927e-05,
+      "loss": 0.6113,
+      "step": 1323
+    },
+    {
+      "epoch": 0.5961278703286808,
+      "grad_norm": 0.11979208141565323,
+      "learning_rate": 7.60186548071029e-05,
+      "loss": 0.4808,
+      "step": 1324
+    },
+    {
+      "epoch": 0.5965781179648807,
+      "grad_norm": 0.1612721085548401,
+      "learning_rate": 7.587488494725157e-05,
+      "loss": 0.7139,
+      "step": 1325
+    },
+    {
+      "epoch": 0.5970283656010806,
+      "grad_norm": 0.12615184485912323,
+      "learning_rate": 7.573116801567301e-05,
+      "loss": 0.6044,
+      "step": 1326
+    },
+    {
+      "epoch": 0.5974786132372805,
+      "grad_norm": 0.13410833477973938,
+      "learning_rate": 7.558750432766903e-05,
+      "loss": 0.6535,
+      "step": 1327
+    },
+    {
+      "epoch": 0.5979288608734804,
+      "grad_norm": 0.14799728989601135,
+      "learning_rate": 7.54438941984243e-05,
+      "loss": 0.6315,
+      "step": 1328
+    },
+    {
+      "epoch": 0.5983791085096803,
+      "grad_norm": 0.10088810324668884,
+      "learning_rate": 7.530033794300632e-05,
+      "loss": 0.3169,
+      "step": 1329
+    },
+    {
+      "epoch": 0.5988293561458803,
+      "grad_norm": 0.16833072900772095,
+      "learning_rate": 7.515683587636413e-05,
+      "loss": 0.967,
+      "step": 1330
+    },
+    {
+      "epoch": 0.5992796037820801,
+      "grad_norm": 0.120406374335289,
+      "learning_rate": 7.501338831332814e-05,
+      "loss": 0.6433,
+      "step": 1331
+    },
+    {
+      "epoch": 0.5997298514182801,
+      "grad_norm": 0.1405513435602188,
+      "learning_rate": 7.48699955686089e-05,
+      "loss": 0.3741,
+      "step": 1332
+    },
+    {
+      "epoch": 0.60018009905448,
+      "grad_norm": 0.15536221861839294,
+      "learning_rate": 7.472665795679694e-05,
+      "loss": 0.8232,
+      "step": 1333
+    },
+    {
+      "epoch": 0.6006303466906798,
+      "grad_norm": 0.09815354645252228,
+      "learning_rate": 7.45833757923617e-05,
+      "loss": 0.6233,
+      "step": 1334
+    },
+    {
+      "epoch": 0.6010805943268798,
+      "grad_norm": 0.14648522436618805,
+      "learning_rate": 7.4440149389651e-05,
+      "loss": 0.5803,
+      "step": 1335
+    },
+    {
+      "epoch": 0.6015308419630797,
+      "grad_norm": 0.1205916479229927,
+      "learning_rate": 7.42969790628903e-05,
+      "loss": 0.5749,
+      "step": 1336
+    },
+    {
+      "epoch": 0.6019810895992797,
+      "grad_norm": 0.1672033965587616,
+      "learning_rate": 7.415386512618217e-05,
+      "loss": 0.5891,
+      "step": 1337
+    },
+    {
+      "epoch": 0.6024313372354795,
+      "grad_norm": 0.1229754388332367,
+      "learning_rate": 7.401080789350525e-05,
+      "loss": 0.4212,
+      "step": 1338
+    },
+    {
+      "epoch": 0.6028815848716794,
+      "grad_norm": 0.17313526570796967,
+      "learning_rate": 7.386780767871397e-05,
+      "loss": 0.4635,
+      "step": 1339
+    },
+    {
+      "epoch": 0.6033318325078794,
+      "grad_norm": 0.1634226143360138,
+      "learning_rate": 7.372486479553748e-05,
+      "loss": 0.5606,
+      "step": 1340
+    },
+    {
+      "epoch": 0.6037820801440792,
+      "grad_norm": 0.19118663668632507,
+      "learning_rate": 7.358197955757939e-05,
+      "loss": 0.5548,
+      "step": 1341
+    },
+    {
+      "epoch": 0.6042323277802791,
+      "grad_norm": 0.11755997687578201,
+      "learning_rate": 7.343915227831661e-05,
+      "loss": 0.3795,
+      "step": 1342
+    },
+    {
+      "epoch": 0.6046825754164791,
+      "grad_norm": 0.12181942909955978,
+      "learning_rate": 7.329638327109903e-05,
+      "loss": 0.3837,
+      "step": 1343
+    },
+    {
+      "epoch": 0.605132823052679,
+      "grad_norm": 0.1529816836118698,
+      "learning_rate": 7.315367284914863e-05,
+      "loss": 0.3302,
+      "step": 1344
+    },
+    {
+      "epoch": 0.6055830706888788,
+      "grad_norm": 0.10931789129972458,
+      "learning_rate": 7.301102132555891e-05,
+      "loss": 0.3735,
+      "step": 1345
+    },
+    {
+      "epoch": 0.6060333183250788,
+      "grad_norm": 0.13579382002353668,
+      "learning_rate": 7.286842901329412e-05,
+      "loss": 0.5139,
+      "step": 1346
+    },
+    {
+      "epoch": 0.6064835659612787,
+      "grad_norm": 0.13477551937103271,
+      "learning_rate": 7.272589622518863e-05,
+      "loss": 0.6656,
+      "step": 1347
+    },
+    {
+      "epoch": 0.6069338135974787,
+      "grad_norm": 0.197172150015831,
+      "learning_rate": 7.258342327394617e-05,
+      "loss": 0.6956,
+      "step": 1348
+    },
+    {
+      "epoch": 0.6073840612336785,
+      "grad_norm": 0.10106527805328369,
+      "learning_rate": 7.244101047213928e-05,
+      "loss": 0.4282,
+      "step": 1349
+    },
+    {
+      "epoch": 0.6078343088698784,
+      "grad_norm": 0.1479652374982834,
+      "learning_rate": 7.229865813220844e-05,
+      "loss": 0.7254,
+      "step": 1350
+    },
+    {
+      "epoch": 0.6082845565060784,
+      "grad_norm": 0.09884002059698105,
+      "learning_rate": 7.215636656646152e-05,
+      "loss": 0.4459,
+      "step": 1351
+    },
+    {
+      "epoch": 0.6087348041422782,
+      "grad_norm": 0.10854051262140274,
+      "learning_rate": 7.201413608707312e-05,
+      "loss": 0.5774,
+      "step": 1352
+    },
+    {
+      "epoch": 0.6091850517784781,
+      "grad_norm": 0.11682359129190445,
+      "learning_rate": 7.187196700608373e-05,
+      "loss": 0.4736,
+      "step": 1353
+    },
+    {
+      "epoch": 0.6096352994146781,
+      "grad_norm": 0.10375698655843735,
+      "learning_rate": 7.17298596353992e-05,
+      "loss": 0.3074,
+      "step": 1354
+    },
+    {
+      "epoch": 0.610085547050878,
+      "grad_norm": 0.08852627128362656,
+      "learning_rate": 7.158781428678988e-05,
+      "loss": 0.3952,
+      "step": 1355
+    },
+    {
+      "epoch": 0.6105357946870779,
+      "grad_norm": 0.13759958744049072,
+      "learning_rate": 7.144583127189029e-05,
+      "loss": 0.4269,
+      "step": 1356
+    },
+    {
+      "epoch": 0.6109860423232778,
+      "grad_norm": 0.1308608502149582,
+      "learning_rate": 7.13039109021979e-05,
+      "loss": 0.5267,
+      "step": 1357
+    },
+    {
+      "epoch": 0.6114362899594777,
+      "grad_norm": 0.19154761731624603,
+      "learning_rate": 7.116205348907299e-05,
+      "loss": 0.6196,
+      "step": 1358
+    },
+    {
+      "epoch": 0.6118865375956776,
+      "grad_norm": 0.12577466666698456,
+      "learning_rate": 7.10202593437375e-05,
+      "loss": 0.7268,
+      "step": 1359
+    },
+    {
+      "epoch": 0.6123367852318775,
+      "grad_norm": 0.2084265649318695,
+      "learning_rate": 7.087852877727481e-05,
+      "loss": 1.0728,
+      "step": 1360
+    },
+    {
+      "epoch": 0.6127870328680775,
+      "grad_norm": 0.12076105922460556,
+      "learning_rate": 7.07368621006286e-05,
+      "loss": 0.4582,
+      "step": 1361
+    },
+    {
+      "epoch": 0.6132372805042774,
+      "grad_norm": 0.12156103551387787,
+      "learning_rate": 7.05952596246025e-05,
+      "loss": 0.6327,
+      "step": 1362
+    },
+    {
+      "epoch": 0.6136875281404772,
+      "grad_norm": 0.19326069951057434,
+      "learning_rate": 7.04537216598592e-05,
+      "loss": 0.598,
+      "step": 1363
+    },
+    {
+      "epoch": 0.6141377757766772,
+      "grad_norm": 0.14157316088676453,
+      "learning_rate": 7.031224851692e-05,
+      "loss": 0.3945,
+      "step": 1364
+    },
+    {
+      "epoch": 0.6145880234128771,
+      "grad_norm": 0.1685905158519745,
+      "learning_rate": 7.017084050616386e-05,
+      "loss": 0.7197,
+      "step": 1365
+    },
+    {
+      "epoch": 0.615038271049077,
+      "grad_norm": 0.15271985530853271,
+      "learning_rate": 7.002949793782687e-05,
+      "loss": 0.5754,
+      "step": 1366
+    },
+    {
+      "epoch": 0.6154885186852769,
+      "grad_norm": 0.13049724698066711,
+      "learning_rate": 6.988822112200156e-05,
+      "loss": 0.5222,
+      "step": 1367
+    },
+    {
+      "epoch": 0.6159387663214768,
+      "grad_norm": 0.16381117701530457,
+      "learning_rate": 6.974701036863627e-05,
+      "loss": 0.7024,
+      "step": 1368
+    },
+    {
+      "epoch": 0.6163890139576768,
+      "grad_norm": 0.16484807431697845,
+      "learning_rate": 6.960586598753427e-05,
+      "loss": 0.5949,
+      "step": 1369
+    },
+    {
+      "epoch": 0.6168392615938766,
+      "grad_norm": 0.11741825938224792,
+      "learning_rate": 6.946478828835332e-05,
+      "loss": 0.4621,
+      "step": 1370
+    },
+    {
+      "epoch": 0.6172895092300765,
+      "grad_norm": 0.12466549873352051,
+      "learning_rate": 6.932377758060482e-05,
+      "loss": 0.5636,
+      "step": 1371
+    },
+    {
+      "epoch": 0.6177397568662765,
+      "grad_norm": 0.15033484995365143,
+      "learning_rate": 6.91828341736533e-05,
+      "loss": 0.4897,
+      "step": 1372
+    },
+    {
+      "epoch": 0.6181900045024764,
+      "grad_norm": 0.20052161812782288,
+      "learning_rate": 6.904195837671552e-05,
+      "loss": 0.8442,
+      "step": 1373
+    },
+    {
+      "epoch": 0.6186402521386762,
+      "grad_norm": 0.11016356945037842,
+      "learning_rate": 6.890115049885994e-05,
+      "loss": 0.496,
+      "step": 1374
+    },
+    {
+      "epoch": 0.6190904997748762,
+      "grad_norm": 0.09439466148614883,
+      "learning_rate": 6.87604108490061e-05,
+      "loss": 0.6055,
+      "step": 1375
+    },
+    {
+      "epoch": 0.6195407474110761,
+      "grad_norm": 0.10878894478082657,
+      "learning_rate": 6.861973973592372e-05,
+      "loss": 0.3725,
+      "step": 1376
+    },
+    {
+      "epoch": 0.619990995047276,
+      "grad_norm": 0.11336513608694077,
+      "learning_rate": 6.847913746823227e-05,
+      "loss": 0.438,
+      "step": 1377
+    },
+    {
+      "epoch": 0.6204412426834759,
+      "grad_norm": 0.11527767032384872,
+      "learning_rate": 6.833860435440007e-05,
+      "loss": 0.7187,
+      "step": 1378
+    },
+    {
+      "epoch": 0.6208914903196758,
+      "grad_norm": 0.1567215472459793,
+      "learning_rate": 6.819814070274384e-05,
+      "loss": 0.9158,
+      "step": 1379
+    },
+    {
+      "epoch": 0.6213417379558758,
+      "grad_norm": 0.0919230580329895,
+      "learning_rate": 6.805774682142782e-05,
+      "loss": 0.3885,
+      "step": 1380
+    },
+    {
+      "epoch": 0.6217919855920756,
+      "grad_norm": 0.10463282465934753,
+      "learning_rate": 6.791742301846326e-05,
+      "loss": 0.5682,
+      "step": 1381
+    },
+    {
+      "epoch": 0.6222422332282755,
+      "grad_norm": 0.1968832015991211,
+      "learning_rate": 6.777716960170753e-05,
+      "loss": 0.5105,
+      "step": 1382
+    },
+    {
+      "epoch": 0.6226924808644755,
+      "grad_norm": 0.16856680810451508,
+      "learning_rate": 6.763698687886373e-05,
+      "loss": 0.4798,
+      "step": 1383
+    },
+    {
+      "epoch": 0.6231427285006754,
+      "grad_norm": 0.13112439215183258,
+      "learning_rate": 6.749687515747976e-05,
+      "loss": 0.497,
+      "step": 1384
+    },
+    {
+      "epoch": 0.6235929761368753,
+      "grad_norm": 0.14991061389446259,
+      "learning_rate": 6.735683474494784e-05,
+      "loss": 0.6191,
+      "step": 1385
+    },
+    {
+      "epoch": 0.6240432237730752,
+      "grad_norm": 0.16807954013347626,
+      "learning_rate": 6.721686594850363e-05,
+      "loss": 0.515,
+      "step": 1386
+    },
+    {
+      "epoch": 0.6244934714092751,
+      "grad_norm": 0.11303338408470154,
+      "learning_rate": 6.707696907522578e-05,
+      "loss": 0.6496,
+      "step": 1387
+    },
+    {
+      "epoch": 0.624943719045475,
+      "grad_norm": 0.12270212173461914,
+      "learning_rate": 6.693714443203507e-05,
+      "loss": 0.9359,
+      "step": 1388
+    },
+    {
+      "epoch": 0.6253939666816749,
+      "grad_norm": 0.11557002365589142,
+      "learning_rate": 6.679739232569388e-05,
+      "loss": 0.5223,
+      "step": 1389
+    },
+    {
+      "epoch": 0.6258442143178748,
+      "grad_norm": 0.14341603219509125,
+      "learning_rate": 6.665771306280537e-05,
+      "loss": 0.3838,
+      "step": 1390
+    },
+    {
+      "epoch": 0.6262944619540748,
+      "grad_norm": 0.12248849868774414,
+      "learning_rate": 6.651810694981298e-05,
+      "loss": 0.5884,
+      "step": 1391
+    },
+    {
+      "epoch": 0.6267447095902746,
+      "grad_norm": 0.14705607295036316,
+      "learning_rate": 6.637857429299958e-05,
+      "loss": 0.6783,
+      "step": 1392
+    },
+    {
+      "epoch": 0.6271949572264746,
+      "grad_norm": 0.13700012862682343,
+      "learning_rate": 6.623911539848698e-05,
+      "loss": 0.4563,
+      "step": 1393
+    },
+    {
+      "epoch": 0.6276452048626745,
+      "grad_norm": 0.10728766769170761,
+      "learning_rate": 6.609973057223501e-05,
+      "loss": 0.4547,
+      "step": 1394
+    },
+    {
+      "epoch": 0.6280954524988743,
+      "grad_norm": 0.12912385165691376,
+      "learning_rate": 6.59604201200412e-05,
+      "loss": 0.5126,
+      "step": 1395
+    },
+    {
+      "epoch": 0.6285457001350743,
+      "grad_norm": 0.10579320043325424,
+      "learning_rate": 6.582118434753975e-05,
+      "loss": 0.3019,
+      "step": 1396
+    },
+    {
+      "epoch": 0.6289959477712742,
+      "grad_norm": 0.08309445530176163,
+      "learning_rate": 6.568202356020108e-05,
+      "loss": 0.34,
+      "step": 1397
+    },
+    {
+      "epoch": 0.6294461954074742,
+      "grad_norm": 0.10809546709060669,
+      "learning_rate": 6.554293806333109e-05,
+      "loss": 0.4162,
+      "step": 1398
+    },
+    {
+      "epoch": 0.629896443043674,
+      "grad_norm": 0.16474172472953796,
+      "learning_rate": 6.540392816207055e-05,
+      "loss": 0.6822,
+      "step": 1399
+    },
+    {
+      "epoch": 0.6303466906798739,
+      "grad_norm": 0.10012883692979813,
+      "learning_rate": 6.526499416139431e-05,
+      "loss": 0.3945,
+      "step": 1400
+    },
+    {
+      "epoch": 0.6307969383160739,
+      "grad_norm": 0.12698769569396973,
+      "learning_rate": 6.512613636611069e-05,
+      "loss": 0.5951,
+      "step": 1401
+    },
+    {
+      "epoch": 0.6312471859522738,
+      "grad_norm": 0.10972292721271515,
+      "learning_rate": 6.498735508086093e-05,
+      "loss": 0.5275,
+      "step": 1402
+    },
+    {
+      "epoch": 0.6316974335884736,
+      "grad_norm": 0.1074480339884758,
+      "learning_rate": 6.48486506101183e-05,
+      "loss": 0.6018,
+      "step": 1403
+    },
+    {
+      "epoch": 0.6321476812246736,
+      "grad_norm": 0.2023674100637436,
+      "learning_rate": 6.471002325818761e-05,
+      "loss": 0.4781,
+      "step": 1404
+    },
+    {
+      "epoch": 0.6325979288608735,
+      "grad_norm": 0.2261338233947754,
+      "learning_rate": 6.45714733292044e-05,
+      "loss": 0.5779,
+      "step": 1405
+    },
+    {
+      "epoch": 0.6330481764970733,
+      "grad_norm": 0.09085024893283844,
+      "learning_rate": 6.443300112713453e-05,
+      "loss": 0.3756,
+      "step": 1406
+    },
+    {
+      "epoch": 0.6334984241332733,
+      "grad_norm": 0.11965364962816238,
+      "learning_rate": 6.42946069557731e-05,
+      "loss": 0.4776,
+      "step": 1407
+    },
+    {
+      "epoch": 0.6339486717694732,
+      "grad_norm": 0.1240624338388443,
+      "learning_rate": 6.415629111874418e-05,
+      "loss": 0.6869,
+      "step": 1408
+    },
+    {
+      "epoch": 0.6343989194056732,
+      "grad_norm": 0.11597741395235062,
+      "learning_rate": 6.40180539194999e-05,
+      "loss": 0.549,
+      "step": 1409
+    },
+    {
+      "epoch": 0.634849167041873,
+      "grad_norm": 0.22984369099140167,
+      "learning_rate": 6.387989566131996e-05,
+      "loss": 0.7818,
+      "step": 1410
+    },
+    {
+      "epoch": 0.6352994146780729,
+      "grad_norm": 0.10499876737594604,
+      "learning_rate": 6.374181664731075e-05,
+      "loss": 0.6748,
+      "step": 1411
+    },
+    {
+      "epoch": 0.6357496623142729,
+      "grad_norm": 0.13295234739780426,
+      "learning_rate": 6.360381718040494e-05,
+      "loss": 0.4241,
+      "step": 1412
+    },
+    {
+      "epoch": 0.6361999099504727,
+      "grad_norm": 0.10862339287996292,
+      "learning_rate": 6.34658975633605e-05,
+      "loss": 0.6123,
+      "step": 1413
+    },
+    {
+      "epoch": 0.6366501575866726,
+      "grad_norm": 0.1305660903453827,
+      "learning_rate": 6.332805809876041e-05,
+      "loss": 0.6284,
+      "step": 1414
+    },
+    {
+      "epoch": 0.6371004052228726,
+      "grad_norm": 0.15044735372066498,
+      "learning_rate": 6.319029908901168e-05,
+      "loss": 0.5772,
+      "step": 1415
+    },
+    {
+      "epoch": 0.6375506528590725,
+      "grad_norm": 0.13762564957141876,
+      "learning_rate": 6.305262083634488e-05,
+      "loss": 0.6292,
+      "step": 1416
+    },
+    {
+      "epoch": 0.6380009004952724,
+      "grad_norm": 0.15244434773921967,
+      "learning_rate": 6.29150236428133e-05,
+      "loss": 0.543,
+      "step": 1417
+    },
+    {
+      "epoch": 0.6384511481314723,
+      "grad_norm": 0.11114734411239624,
+      "learning_rate": 6.277750781029254e-05,
+      "loss": 0.4352,
+      "step": 1418
+    },
+    {
+      "epoch": 0.6389013957676722,
+      "grad_norm": 0.204766184091568,
+      "learning_rate": 6.26400736404796e-05,
+      "loss": 0.6156,
+      "step": 1419
+    },
+    {
+      "epoch": 0.6393516434038722,
+      "grad_norm": 0.21171844005584717,
+      "learning_rate": 6.250272143489237e-05,
+      "loss": 0.7111,
+      "step": 1420
+    },
+    {
+      "epoch": 0.639801891040072,
+      "grad_norm": 0.10301060229539871,
+      "learning_rate": 6.236545149486881e-05,
+      "loss": 0.4855,
+      "step": 1421
+    },
+    {
+      "epoch": 0.640252138676272,
+      "grad_norm": 0.12298280745744705,
+      "learning_rate": 6.222826412156659e-05,
+      "loss": 0.5579,
+      "step": 1422
+    },
+    {
+      "epoch": 0.6407023863124719,
+      "grad_norm": 0.14201900362968445,
+      "learning_rate": 6.209115961596208e-05,
+      "loss": 0.813,
+      "step": 1423
+    },
+    {
+      "epoch": 0.6411526339486717,
+      "grad_norm": 0.12120926380157471,
+      "learning_rate": 6.195413827884986e-05,
+      "loss": 0.4257,
+      "step": 1424
+    },
+    {
+      "epoch": 0.6416028815848717,
+      "grad_norm": 0.13606584072113037,
+      "learning_rate": 6.181720041084217e-05,
+      "loss": 0.7456,
+      "step": 1425
+    },
+    {
+      "epoch": 0.6420531292210716,
+      "grad_norm": 0.10782089829444885,
+      "learning_rate": 6.168034631236795e-05,
+      "loss": 0.6499,
+      "step": 1426
+    },
+    {
+      "epoch": 0.6425033768572715,
+      "grad_norm": 0.09734063595533371,
+      "learning_rate": 6.15435762836725e-05,
+      "loss": 0.6112,
+      "step": 1427
+    },
+    {
+      "epoch": 0.6429536244934714,
+      "grad_norm": 0.0944141149520874,
+      "learning_rate": 6.140689062481657e-05,
+      "loss": 0.367,
+      "step": 1428
+    },
+    {
+      "epoch": 0.6434038721296713,
+      "grad_norm": 0.10263581573963165,
+      "learning_rate": 6.127028963567593e-05,
+      "loss": 0.6239,
+      "step": 1429
+    },
+    {
+      "epoch": 0.6438541197658713,
+      "grad_norm": 0.14555425941944122,
+      "learning_rate": 6.113377361594049e-05,
+      "loss": 0.72,
+      "step": 1430
+    },
+    {
+      "epoch": 0.6443043674020711,
+      "grad_norm": 0.1000761091709137,
+      "learning_rate": 6.099734286511378e-05,
+      "loss": 0.4955,
+      "step": 1431
+    },
+    {
+      "epoch": 0.644754615038271,
+      "grad_norm": 0.10147224366664886,
+      "learning_rate": 6.0860997682512224e-05,
+      "loss": 0.3785,
+      "step": 1432
+    },
+    {
+      "epoch": 0.645204862674471,
+      "grad_norm": 0.11730487644672394,
+      "learning_rate": 6.0724738367264625e-05,
+      "loss": 0.491,
+      "step": 1433
+    },
+    {
+      "epoch": 0.6456551103106709,
+      "grad_norm": 0.1680234670639038,
+      "learning_rate": 6.0588565218311265e-05,
+      "loss": 0.7816,
+      "step": 1434
+    },
+    {
+      "epoch": 0.6461053579468707,
+      "grad_norm": 0.12494342029094696,
+      "learning_rate": 6.04524785344035e-05,
+      "loss": 0.6096,
+      "step": 1435
+    },
+    {
+      "epoch": 0.6465556055830707,
+      "grad_norm": 0.15204067528247833,
+      "learning_rate": 6.0316478614102875e-05,
+      "loss": 0.4838,
+      "step": 1436
+    },
+    {
+      "epoch": 0.6470058532192706,
+      "grad_norm": 0.11809913069009781,
+      "learning_rate": 6.018056575578075e-05,
+      "loss": 0.7296,
+      "step": 1437
+    },
+    {
+      "epoch": 0.6474561008554706,
+      "grad_norm": 0.10827966779470444,
+      "learning_rate": 6.0044740257617225e-05,
+      "loss": 0.4804,
+      "step": 1438
+    },
+    {
+      "epoch": 0.6479063484916704,
+      "grad_norm": 0.113580621778965,
+      "learning_rate": 5.990900241760102e-05,
+      "loss": 0.4071,
+      "step": 1439
+    },
+    {
+      "epoch": 0.6483565961278703,
+      "grad_norm": 0.15822423994541168,
+      "learning_rate": 5.9773352533528335e-05,
+      "loss": 0.5573,
+      "step": 1440
+    },
+    {
+      "epoch": 0.6488068437640703,
+      "grad_norm": 0.10006440430879593,
+      "learning_rate": 5.963779090300255e-05,
+      "loss": 0.6999,
+      "step": 1441
+    },
+    {
+      "epoch": 0.6492570914002701,
+      "grad_norm": 0.11854738742113113,
+      "learning_rate": 5.9502317823433275e-05,
+      "loss": 0.6299,
+      "step": 1442
+    },
+    {
+      "epoch": 0.64970733903647,
+      "grad_norm": 0.16326195001602173,
+      "learning_rate": 5.9366933592035975e-05,
+      "loss": 0.5282,
+      "step": 1443
+    },
+    {
+      "epoch": 0.65015758667267,
+      "grad_norm": 0.13654035329818726,
+      "learning_rate": 5.923163850583113e-05,
+      "loss": 0.6435,
+      "step": 1444
+    },
+    {
+      "epoch": 0.6506078343088699,
+      "grad_norm": 0.12476153671741486,
+      "learning_rate": 5.909643286164368e-05,
+      "loss": 0.6193,
+      "step": 1445
+    },
+    {
+      "epoch": 0.6510580819450698,
+      "grad_norm": 0.11656292527914047,
+      "learning_rate": 5.8961316956102234e-05,
+      "loss": 0.4084,
+      "step": 1446
+    },
+    {
+      "epoch": 0.6515083295812697,
+      "grad_norm": 0.1265832483768463,
+      "learning_rate": 5.882629108563871e-05,
+      "loss": 0.6598,
+      "step": 1447
+    },
+    {
+      "epoch": 0.6519585772174696,
+      "grad_norm": 0.16534487903118134,
+      "learning_rate": 5.869135554648729e-05,
+      "loss": 0.4481,
+      "step": 1448
+    },
+    {
+      "epoch": 0.6524088248536695,
+      "grad_norm": 0.11791814863681793,
+      "learning_rate": 5.855651063468413e-05,
+      "loss": 0.6835,
+      "step": 1449
+    },
+    {
+      "epoch": 0.6528590724898694,
+      "grad_norm": 0.18122895061969757,
+      "learning_rate": 5.84217566460665e-05,
+      "loss": 0.7914,
+      "step": 1450
+    },
+    {
+      "epoch": 0.6533093201260693,
+      "grad_norm": 0.11034449934959412,
+      "learning_rate": 5.828709387627218e-05,
+      "loss": 0.5055,
+      "step": 1451
+    },
+    {
+      "epoch": 0.6537595677622693,
+      "grad_norm": 0.1272686868906021,
+      "learning_rate": 5.8152522620738916e-05,
+      "loss": 0.4003,
+      "step": 1452
+    },
+    {
+      "epoch": 0.6542098153984691,
+      "grad_norm": 0.13408847153186798,
+      "learning_rate": 5.8018043174703497e-05,
+      "loss": 0.4764,
+      "step": 1453
+    },
+    {
+      "epoch": 0.6546600630346691,
+      "grad_norm": 0.10295604169368744,
+      "learning_rate": 5.788365583320145e-05,
+      "loss": 0.6708,
+      "step": 1454
+    },
+    {
+      "epoch": 0.655110310670869,
+      "grad_norm": 0.13045580685138702,
+      "learning_rate": 5.7749360891066176e-05,
+      "loss": 0.5123,
+      "step": 1455
+    },
+    {
+      "epoch": 0.6555605583070689,
+      "grad_norm": 0.20646020770072937,
+      "learning_rate": 5.761515864292836e-05,
+      "loss": 0.4016,
+      "step": 1456
+    },
+    {
+      "epoch": 0.6560108059432688,
+      "grad_norm": 0.11828052252531052,
+      "learning_rate": 5.7481049383215345e-05,
+      "loss": 0.6746,
+      "step": 1457
+    },
+    {
+      "epoch": 0.6564610535794687,
+      "grad_norm": 0.1225493773818016,
+      "learning_rate": 5.73470334061505e-05,
+      "loss": 0.3736,
+      "step": 1458
+    },
+    {
+      "epoch": 0.6569113012156687,
+      "grad_norm": 0.11636580526828766,
+      "learning_rate": 5.721311100575235e-05,
+      "loss": 0.5181,
+      "step": 1459
+    },
+    {
+      "epoch": 0.6573615488518685,
+      "grad_norm": 0.19074717164039612,
+      "learning_rate": 5.707928247583444e-05,
+      "loss": 0.6376,
+      "step": 1460
+    },
+    {
+      "epoch": 0.6578117964880684,
+      "grad_norm": 0.17381444573402405,
+      "learning_rate": 5.694554811000408e-05,
+      "loss": 0.7946,
+      "step": 1461
+    },
+    {
+      "epoch": 0.6582620441242684,
+      "grad_norm": 0.11473318934440613,
+      "learning_rate": 5.681190820166213e-05,
+      "loss": 0.6838,
+      "step": 1462
+    },
+    {
+      "epoch": 0.6587122917604683,
+      "grad_norm": 0.19790330529212952,
+      "learning_rate": 5.667836304400221e-05,
+      "loss": 0.6935,
+      "step": 1463
+    },
+    {
+      "epoch": 0.6591625393966681,
+      "grad_norm": 0.14861299097537994,
+      "learning_rate": 5.6544912930010064e-05,
+      "loss": 0.4259,
+      "step": 1464
+    },
+    {
+      "epoch": 0.6596127870328681,
+      "grad_norm": 0.1718035191297531,
+      "learning_rate": 5.6411558152462894e-05,
+      "loss": 0.7112,
+      "step": 1465
+    },
+    {
+      "epoch": 0.660063034669068,
+      "grad_norm": 0.14902283251285553,
+      "learning_rate": 5.62782990039288e-05,
+      "loss": 0.3711,
+      "step": 1466
+    },
+    {
+      "epoch": 0.6605132823052678,
+      "grad_norm": 0.20114299654960632,
+      "learning_rate": 5.614513577676592e-05,
+      "loss": 0.6803,
+      "step": 1467
+    },
+    {
+      "epoch": 0.6609635299414678,
+      "grad_norm": 0.18261073529720306,
+      "learning_rate": 5.6012068763122236e-05,
+      "loss": 0.6786,
+      "step": 1468
+    },
+    {
+      "epoch": 0.6614137775776677,
+      "grad_norm": 0.1697646528482437,
+      "learning_rate": 5.587909825493434e-05,
+      "loss": 0.6679,
+      "step": 1469
+    },
+    {
+      "epoch": 0.6618640252138677,
+      "grad_norm": 0.1117134764790535,
+      "learning_rate": 5.574622454392731e-05,
+      "loss": 0.6785,
+      "step": 1470
+    },
+    {
+      "epoch": 0.6623142728500675,
+      "grad_norm": 0.10873869806528091,
+      "learning_rate": 5.561344792161374e-05,
+      "loss": 0.5648,
+      "step": 1471
+    },
+    {
+      "epoch": 0.6627645204862674,
+      "grad_norm": 0.1491149365901947,
+      "learning_rate": 5.54807686792933e-05,
+      "loss": 0.7219,
+      "step": 1472
+    },
+    {
+      "epoch": 0.6632147681224674,
+      "grad_norm": 0.1503085494041443,
+      "learning_rate": 5.5348187108051986e-05,
+      "loss": 0.5678,
+      "step": 1473
+    },
+    {
+      "epoch": 0.6636650157586673,
+      "grad_norm": 0.12257836759090424,
+      "learning_rate": 5.521570349876151e-05,
+      "loss": 0.7089,
+      "step": 1474
+    },
+    {
+      "epoch": 0.6641152633948672,
+      "grad_norm": 0.1146334707736969,
+      "learning_rate": 5.5083318142078655e-05,
+      "loss": 0.4076,
+      "step": 1475
+    },
+    {
+      "epoch": 0.6645655110310671,
+      "grad_norm": 0.10977943241596222,
+      "learning_rate": 5.495103132844467e-05,
+      "loss": 0.5901,
+      "step": 1476
+    },
+    {
+      "epoch": 0.665015758667267,
+      "grad_norm": 0.14321941137313843,
+      "learning_rate": 5.4818843348084634e-05,
+      "loss": 0.5048,
+      "step": 1477
+    },
+    {
+      "epoch": 0.6654660063034669,
+      "grad_norm": 0.12193970382213593,
+      "learning_rate": 5.468675449100664e-05,
+      "loss": 0.6742,
+      "step": 1478
+    },
+    {
+      "epoch": 0.6659162539396668,
+      "grad_norm": 0.10547365248203278,
+      "learning_rate": 5.4554765047001613e-05,
+      "loss": 0.525,
+      "step": 1479
+    },
+    {
+      "epoch": 0.6663665015758667,
+      "grad_norm": 0.14464342594146729,
+      "learning_rate": 5.442287530564203e-05,
+      "loss": 0.994,
+      "step": 1480
+    },
+    {
+      "epoch": 0.6668167492120667,
+      "grad_norm": 0.10600848495960236,
+      "learning_rate": 5.429108555628186e-05,
+      "loss": 0.4751,
+      "step": 1481
+    },
+    {
+      "epoch": 0.6672669968482665,
+      "grad_norm": 0.10199996083974838,
+      "learning_rate": 5.4159396088055645e-05,
+      "loss": 0.5001,
+      "step": 1482
+    },
+    {
+      "epoch": 0.6677172444844665,
+      "grad_norm": 0.087745800614357,
+      "learning_rate": 5.40278071898779e-05,
+      "loss": 0.5097,
+      "step": 1483
+    },
+    {
+      "epoch": 0.6681674921206664,
+      "grad_norm": 0.12530991435050964,
+      "learning_rate": 5.38963191504425e-05,
+      "loss": 0.5492,
+      "step": 1484
+    },
+    {
+      "epoch": 0.6686177397568662,
+      "grad_norm": 0.09561583399772644,
+      "learning_rate": 5.376493225822209e-05,
+      "loss": 0.4015,
+      "step": 1485
+    },
+    {
+      "epoch": 0.6690679873930662,
+      "grad_norm": 0.10222489386796951,
+      "learning_rate": 5.363364680146725e-05,
+      "loss": 0.4284,
+      "step": 1486
+    },
+    {
+      "epoch": 0.6695182350292661,
+      "grad_norm": 0.20286443829536438,
+      "learning_rate": 5.350246306820632e-05,
+      "loss": 1.016,
+      "step": 1487
+    },
+    {
+      "epoch": 0.669968482665466,
+      "grad_norm": 0.10691449046134949,
+      "learning_rate": 5.3371381346244135e-05,
+      "loss": 0.4236,
+      "step": 1488
+    },
+    {
+      "epoch": 0.6704187303016659,
+      "grad_norm": 0.19654229283332825,
+      "learning_rate": 5.324040192316193e-05,
+      "loss": 0.9537,
+      "step": 1489
+    },
+    {
+      "epoch": 0.6708689779378658,
+      "grad_norm": 0.11526825278997421,
+      "learning_rate": 5.310952508631646e-05,
+      "loss": 0.59,
+      "step": 1490
+    },
+    {
+      "epoch": 0.6713192255740658,
+      "grad_norm": 0.11821702122688293,
+      "learning_rate": 5.297875112283941e-05,
+      "loss": 0.4774,
+      "step": 1491
+    },
+    {
+      "epoch": 0.6717694732102657,
+      "grad_norm": 0.18998759984970093,
+      "learning_rate": 5.2848080319636765e-05,
+      "loss": 0.6456,
+      "step": 1492
+    },
+    {
+      "epoch": 0.6722197208464655,
+      "grad_norm": 0.17253439128398895,
+      "learning_rate": 5.271751296338823e-05,
+      "loss": 0.8013,
+      "step": 1493
+    },
+    {
+      "epoch": 0.6726699684826655,
+      "grad_norm": 0.1001097708940506,
+      "learning_rate": 5.25870493405464e-05,
+      "loss": 0.346,
+      "step": 1494
+    },
+    {
+      "epoch": 0.6731202161188654,
+      "grad_norm": 0.12708154320716858,
+      "learning_rate": 5.2456689737336575e-05,
+      "loss": 0.4669,
+      "step": 1495
+    },
+    {
+      "epoch": 0.6735704637550652,
+      "grad_norm": 0.09766164422035217,
+      "learning_rate": 5.232643443975553e-05,
+      "loss": 0.7082,
+      "step": 1496
+    },
+    {
+      "epoch": 0.6740207113912652,
+      "grad_norm": 0.11283717304468155,
+      "learning_rate": 5.219628373357143e-05,
+      "loss": 0.6487,
+      "step": 1497
+    },
+    {
+      "epoch": 0.6744709590274651,
+      "grad_norm": 0.11566527932882309,
+      "learning_rate": 5.2066237904322855e-05,
+      "loss": 0.4366,
+      "step": 1498
+    },
+    {
+      "epoch": 0.6749212066636651,
+      "grad_norm": 0.09786967933177948,
+      "learning_rate": 5.193629723731837e-05,
+      "loss": 0.458,
+      "step": 1499
+    },
+    {
+      "epoch": 0.6753714542998649,
+      "grad_norm": 0.12495690584182739,
+      "learning_rate": 5.180646201763577e-05,
+      "loss": 0.4467,
+      "step": 1500
+    },
+    {
+      "epoch": 0.6758217019360648,
+      "grad_norm": 0.12726067006587982,
+      "learning_rate": 5.167673253012153e-05,
+      "loss": 0.556,
+      "step": 1501
+    },
+    {
+      "epoch": 0.6762719495722648,
+      "grad_norm": 0.12084569782018661,
+      "learning_rate": 5.154710905939015e-05,
+      "loss": 0.4797,
+      "step": 1502
+    },
+    {
+      "epoch": 0.6767221972084646,
+      "grad_norm": 0.17686428129673004,
+      "learning_rate": 5.141759188982357e-05,
+      "loss": 0.752,
+      "step": 1503
+    },
+    {
+      "epoch": 0.6771724448446645,
+      "grad_norm": 0.1723306030035019,
+      "learning_rate": 5.1288181305570506e-05,
+      "loss": 0.4121,
+      "step": 1504
+    },
+    {
+      "epoch": 0.6776226924808645,
+      "grad_norm": 0.10812796652317047,
+      "learning_rate": 5.115887759054572e-05,
+      "loss": 0.4184,
+      "step": 1505
+    },
+    {
+      "epoch": 0.6780729401170644,
+      "grad_norm": 0.1417206972837448,
+      "learning_rate": 5.102968102842973e-05,
+      "loss": 0.7234,
+      "step": 1506
+    },
+    {
+      "epoch": 0.6785231877532643,
+      "grad_norm": 0.1694130301475525,
+      "learning_rate": 5.090059190266779e-05,
+      "loss": 0.3912,
+      "step": 1507
+    },
+    {
+      "epoch": 0.6789734353894642,
+      "grad_norm": 0.10867101699113846,
+      "learning_rate": 5.0771610496469525e-05,
+      "loss": 0.5186,
+      "step": 1508
+    },
+    {
+      "epoch": 0.6794236830256641,
+      "grad_norm": 0.1154758408665657,
+      "learning_rate": 5.064273709280821e-05,
+      "loss": 0.309,
+      "step": 1509
+    },
+    {
+      "epoch": 0.6798739306618641,
+      "grad_norm": 0.13570335507392883,
+      "learning_rate": 5.05139719744202e-05,
+      "loss": 0.5334,
+      "step": 1510
+    },
+    {
+      "epoch": 0.6803241782980639,
+      "grad_norm": 0.1160716563463211,
+      "learning_rate": 5.038531542380425e-05,
+      "loss": 0.6093,
+      "step": 1511
+    },
+    {
+      "epoch": 0.6807744259342638,
+      "grad_norm": 0.11643337458372116,
+      "learning_rate": 5.025676772322099e-05,
+      "loss": 0.5983,
+      "step": 1512
+    },
+    {
+      "epoch": 0.6812246735704638,
+      "grad_norm": 0.1893714815378189,
+      "learning_rate": 5.012832915469208e-05,
+      "loss": 0.6929,
+      "step": 1513
+    },
+    {
+      "epoch": 0.6816749212066636,
+      "grad_norm": 0.17537082731723785,
+      "learning_rate": 5.000000000000002e-05,
+      "loss": 0.6325,
+      "step": 1514
+    },
+    {
+      "epoch": 0.6821251688428636,
+      "grad_norm": 0.1313885599374771,
+      "learning_rate": 4.9871780540687e-05,
+      "loss": 0.4508,
+      "step": 1515
+    },
+    {
+      "epoch": 0.6825754164790635,
+      "grad_norm": 0.12365297973155975,
+      "learning_rate": 4.9743671058054707e-05,
+      "loss": 0.4929,
+      "step": 1516
+    },
+    {
+      "epoch": 0.6830256641152634,
+      "grad_norm": 0.1398656815290451,
+      "learning_rate": 4.961567183316349e-05,
+      "loss": 0.5697,
+      "step": 1517
+    },
+    {
+      "epoch": 0.6834759117514633,
+      "grad_norm": 0.104002445936203,
+      "learning_rate": 4.948778314683184e-05,
+      "loss": 0.4957,
+      "step": 1518
+    },
+    {
+      "epoch": 0.6839261593876632,
+      "grad_norm": 0.11461395025253296,
+      "learning_rate": 4.93600052796357e-05,
+      "loss": 0.5311,
+      "step": 1519
+    },
+    {
+      "epoch": 0.6843764070238632,
+      "grad_norm": 0.11953675746917725,
+      "learning_rate": 4.923233851190795e-05,
+      "loss": 0.5434,
+      "step": 1520
+    },
+    {
+      "epoch": 0.684826654660063,
+      "grad_norm": 0.1791170984506607,
+      "learning_rate": 4.9104783123737566e-05,
+      "loss": 0.577,
+      "step": 1521
+    },
+    {
+      "epoch": 0.6852769022962629,
+      "grad_norm": 0.11953389644622803,
+      "learning_rate": 4.8977339394969427e-05,
+      "loss": 0.5692,
+      "step": 1522
+    },
+    {
+      "epoch": 0.6857271499324629,
+      "grad_norm": 0.11368032544851303,
+      "learning_rate": 4.885000760520317e-05,
+      "loss": 0.4507,
+      "step": 1523
+    },
+    {
+      "epoch": 0.6861773975686628,
+      "grad_norm": 0.11295335739850998,
+      "learning_rate": 4.8722788033793e-05,
+      "loss": 0.3427,
+      "step": 1524
+    },
+    {
+      "epoch": 0.6866276452048626,
+      "grad_norm": 0.11858644336462021,
+      "learning_rate": 4.8595680959847e-05,
+      "loss": 0.34,
+      "step": 1525
+    },
+    {
+      "epoch": 0.6870778928410626,
+      "grad_norm": 0.1761258840560913,
+      "learning_rate": 4.846868666222623e-05,
+      "loss": 0.7268,
+      "step": 1526
+    },
+    {
+      "epoch": 0.6875281404772625,
+      "grad_norm": 0.12842504680156708,
+      "learning_rate": 4.834180541954447e-05,
+      "loss": 0.5065,
+      "step": 1527
+    },
+    {
+      "epoch": 0.6879783881134625,
+      "grad_norm": 0.2049914449453354,
+      "learning_rate": 4.821503751016746e-05,
+      "loss": 0.6762,
+      "step": 1528
+    },
+    {
+      "epoch": 0.6884286357496623,
+      "grad_norm": 0.09843791276216507,
+      "learning_rate": 4.8088383212212266e-05,
+      "loss": 0.401,
+      "step": 1529
+    },
+    {
+      "epoch": 0.6888788833858622,
+      "grad_norm": 0.10201644897460938,
+      "learning_rate": 4.796184280354671e-05,
+      "loss": 0.6004,
+      "step": 1530
+    },
+    {
+      "epoch": 0.6893291310220622,
+      "grad_norm": 0.11396285891532898,
+      "learning_rate": 4.7835416561788805e-05,
+      "loss": 0.4575,
+      "step": 1531
+    },
+    {
+      "epoch": 0.689779378658262,
+      "grad_norm": 0.14770056307315826,
+      "learning_rate": 4.7709104764305914e-05,
+      "loss": 0.7727,
+      "step": 1532
+    },
+    {
+      "epoch": 0.6902296262944619,
+      "grad_norm": 0.22169174253940582,
+      "learning_rate": 4.7582907688214595e-05,
+      "loss": 0.7092,
+      "step": 1533
+    },
+    {
+      "epoch": 0.6906798739306619,
+      "grad_norm": 0.15397916734218597,
+      "learning_rate": 4.7456825610379475e-05,
+      "loss": 0.4696,
+      "step": 1534
+    },
+    {
+      "epoch": 0.6911301215668618,
+      "grad_norm": 0.09904127568006516,
+      "learning_rate": 4.733085880741301e-05,
+      "loss": 0.4691,
+      "step": 1535
+    },
+    {
+      "epoch": 0.6915803692030617,
+      "grad_norm": 0.16268642246723175,
+      "learning_rate": 4.720500755567472e-05,
+      "loss": 0.4491,
+      "step": 1536
+    },
+    {
+      "epoch": 0.6920306168392616,
+      "grad_norm": 0.18245473504066467,
+      "learning_rate": 4.7079272131270623e-05,
+      "loss": 0.702,
+      "step": 1537
+    },
+    {
+      "epoch": 0.6924808644754615,
+      "grad_norm": 0.1282515972852707,
+      "learning_rate": 4.695365281005262e-05,
+      "loss": 0.6678,
+      "step": 1538
+    },
+    {
+      "epoch": 0.6929311121116614,
+      "grad_norm": 0.12507499754428864,
+      "learning_rate": 4.682814986761792e-05,
+      "loss": 0.2855,
+      "step": 1539
+    },
+    {
+      "epoch": 0.6933813597478613,
+      "grad_norm": 0.09926587343215942,
+      "learning_rate": 4.6702763579308284e-05,
+      "loss": 0.7932,
+      "step": 1540
+    },
+    {
+      "epoch": 0.6938316073840612,
+      "grad_norm": 0.1379023641347885,
+      "learning_rate": 4.657749422020979e-05,
+      "loss": 0.4925,
+      "step": 1541
+    },
+    {
+      "epoch": 0.6942818550202612,
+      "grad_norm": 0.09524741023778915,
+      "learning_rate": 4.645234206515171e-05,
+      "loss": 0.4577,
+      "step": 1542
+    },
+    {
+      "epoch": 0.694732102656461,
+      "grad_norm": 0.11450893431901932,
+      "learning_rate": 4.6327307388706345e-05,
+      "loss": 0.3425,
+      "step": 1543
+    },
+    {
+      "epoch": 0.695182350292661,
+      "grad_norm": 0.12477904558181763,
+      "learning_rate": 4.620239046518823e-05,
+      "loss": 0.6392,
+      "step": 1544
+    },
+    {
+      "epoch": 0.6956325979288609,
+      "grad_norm": 0.12293538451194763,
+      "learning_rate": 4.607759156865358e-05,
+      "loss": 0.5896,
+      "step": 1545
+    },
+    {
+      "epoch": 0.6960828455650608,
+      "grad_norm": 0.09063368290662766,
+      "learning_rate": 4.595291097289952e-05,
+      "loss": 0.3269,
+      "step": 1546
+    },
+    {
+      "epoch": 0.6965330932012607,
+      "grad_norm": 0.10746719688177109,
+      "learning_rate": 4.582834895146392e-05,
+      "loss": 0.4959,
+      "step": 1547
+    },
+    {
+      "epoch": 0.6969833408374606,
+      "grad_norm": 0.12925972044467926,
+      "learning_rate": 4.570390577762419e-05,
+      "loss": 0.6926,
+      "step": 1548
+    },
+    {
+      "epoch": 0.6974335884736605,
+      "grad_norm": 0.13169018924236298,
+      "learning_rate": 4.5579581724397255e-05,
+      "loss": 0.4284,
+      "step": 1549
+    },
+    {
+      "epoch": 0.6978838361098604,
+      "grad_norm": 0.1480969935655594,
+      "learning_rate": 4.545537706453861e-05,
+      "loss": 0.3167,
+      "step": 1550
+    },
+    {
+      "epoch": 0.6983340837460603,
+      "grad_norm": 0.1721503585577011,
+      "learning_rate": 4.533129207054168e-05,
+      "loss": 0.7798,
+      "step": 1551
+    },
+    {
+      "epoch": 0.6987843313822603,
+      "grad_norm": 0.13256415724754333,
+      "learning_rate": 4.520732701463762e-05,
+      "loss": 0.8859,
+      "step": 1552
+    },
+    {
+      "epoch": 0.6992345790184602,
+      "grad_norm": 0.10677511245012283,
+      "learning_rate": 4.508348216879421e-05,
+      "loss": 0.6176,
+      "step": 1553
+    },
+    {
+      "epoch": 0.69968482665466,
+      "grad_norm": 0.10240598022937775,
+      "learning_rate": 4.495975780471562e-05,
+      "loss": 0.4319,
+      "step": 1554
+    },
+    {
+      "epoch": 0.70013507429086,
+      "grad_norm": 0.13021352887153625,
+      "learning_rate": 4.483615419384167e-05,
+      "loss": 0.4961,
+      "step": 1555
+    },
+    {
+      "epoch": 0.7005853219270599,
+      "grad_norm": 0.1073610931634903,
+      "learning_rate": 4.471267160734731e-05,
+      "loss": 0.4581,
+      "step": 1556
+    },
+    {
+      "epoch": 0.7010355695632597,
+      "grad_norm": 0.1581767052412033,
+      "learning_rate": 4.45893103161418e-05,
+      "loss": 0.4322,
+      "step": 1557
+    },
+    {
+      "epoch": 0.7014858171994597,
+      "grad_norm": 0.1240011602640152,
+      "learning_rate": 4.446607059086855e-05,
+      "loss": 0.4042,
+      "step": 1558
+    },
+    {
+      "epoch": 0.7019360648356596,
+      "grad_norm": 0.13136258721351624,
+      "learning_rate": 4.434295270190403e-05,
+      "loss": 0.6886,
+      "step": 1559
+    },
+    {
+      "epoch": 0.7023863124718596,
+      "grad_norm": 0.15671755373477936,
+      "learning_rate": 4.421995691935755e-05,
+      "loss": 0.5517,
+      "step": 1560
+    },
+    {
+      "epoch": 0.7028365601080594,
+      "grad_norm": 0.13269691169261932,
+      "learning_rate": 4.409708351307049e-05,
+      "loss": 0.4656,
+      "step": 1561
+    },
+    {
+      "epoch": 0.7032868077442593,
+      "grad_norm": 0.09884883463382721,
+      "learning_rate": 4.3974332752615734e-05,
+      "loss": 0.415,
+      "step": 1562
+    },
+    {
+      "epoch": 0.7037370553804593,
+      "grad_norm": 0.10842882096767426,
+      "learning_rate": 4.385170490729712e-05,
+      "loss": 0.33,
+      "step": 1563
+    },
+    {
+      "epoch": 0.7041873030166592,
+      "grad_norm": 0.12758079171180725,
+      "learning_rate": 4.372920024614884e-05,
+      "loss": 0.5314,
+      "step": 1564
+    },
+    {
+      "epoch": 0.704637550652859,
+      "grad_norm": 0.10838529467582703,
+      "learning_rate": 4.360681903793468e-05,
+      "loss": 0.4407,
+      "step": 1565
+    },
+    {
+      "epoch": 0.705087798289059,
+      "grad_norm": 0.16309107840061188,
+      "learning_rate": 4.348456155114786e-05,
+      "loss": 0.6031,
+      "step": 1566
+    },
+    {
+      "epoch": 0.7055380459252589,
+      "grad_norm": 0.1477869600057602,
+      "learning_rate": 4.33624280540099e-05,
+      "loss": 0.7672,
+      "step": 1567
+    },
+    {
+      "epoch": 0.7059882935614588,
+      "grad_norm": 0.12081977725028992,
+      "learning_rate": 4.3240418814470415e-05,
+      "loss": 0.4702,
+      "step": 1568
+    },
+    {
+      "epoch": 0.7064385411976587,
+      "grad_norm": 0.11842213571071625,
+      "learning_rate": 4.3118534100206434e-05,
+      "loss": 0.5524,
+      "step": 1569
+    },
+    {
+      "epoch": 0.7068887888338586,
+      "grad_norm": 0.13214458525180817,
+      "learning_rate": 4.2996774178621736e-05,
+      "loss": 0.3502,
+      "step": 1570
+    },
+    {
+      "epoch": 0.7073390364700586,
+      "grad_norm": 0.1304139643907547,
+      "learning_rate": 4.287513931684635e-05,
+      "loss": 0.4634,
+      "step": 1571
+    },
+    {
+      "epoch": 0.7077892841062584,
+      "grad_norm": 0.1663137972354889,
+      "learning_rate": 4.275362978173594e-05,
+      "loss": 0.6364,
+      "step": 1572
+    },
+    {
+      "epoch": 0.7082395317424584,
+      "grad_norm": 0.13673043251037598,
+      "learning_rate": 4.26322458398711e-05,
+      "loss": 0.5587,
+      "step": 1573
+    },
+    {
+      "epoch": 0.7086897793786583,
+      "grad_norm": 0.13097482919692993,
+      "learning_rate": 4.251098775755709e-05,
+      "loss": 0.6171,
+      "step": 1574
+    },
+    {
+      "epoch": 0.7091400270148581,
+      "grad_norm": 0.16620616614818573,
+      "learning_rate": 4.238985580082293e-05,
+      "loss": 0.6491,
+      "step": 1575
+    },
+    {
+      "epoch": 0.7095902746510581,
+      "grad_norm": 0.22634264826774597,
+      "learning_rate": 4.226885023542083e-05,
+      "loss": 0.7484,
+      "step": 1576
+    },
+    {
+      "epoch": 0.710040522287258,
+      "grad_norm": 0.11268556118011475,
+      "learning_rate": 4.2147971326825966e-05,
+      "loss": 0.5402,
+      "step": 1577
+    },
+    {
+      "epoch": 0.7104907699234579,
+      "grad_norm": 0.16073636710643768,
+      "learning_rate": 4.202721934023537e-05,
+      "loss": 0.8349,
+      "step": 1578
+    },
+    {
+      "epoch": 0.7109410175596578,
+      "grad_norm": 0.12581096589565277,
+      "learning_rate": 4.19065945405678e-05,
+      "loss": 0.6952,
+      "step": 1579
+    },
+    {
+      "epoch": 0.7113912651958577,
+      "grad_norm": 0.10183834284543991,
+      "learning_rate": 4.17860971924629e-05,
+      "loss": 0.5344,
+      "step": 1580
+    },
+    {
+      "epoch": 0.7118415128320577,
+      "grad_norm": 0.16986067593097687,
+      "learning_rate": 4.1665727560280735e-05,
+      "loss": 0.852,
+      "step": 1581
+    },
+    {
+      "epoch": 0.7122917604682576,
+      "grad_norm": 0.19713398814201355,
+      "learning_rate": 4.154548590810113e-05,
+      "loss": 0.5469,
+      "step": 1582
+    },
+    {
+      "epoch": 0.7127420081044574,
+      "grad_norm": 0.11787287145853043,
+      "learning_rate": 4.14253724997232e-05,
+      "loss": 0.5503,
+      "step": 1583
+    },
+    {
+      "epoch": 0.7131922557406574,
+      "grad_norm": 0.1299823522567749,
+      "learning_rate": 4.130538759866457e-05,
+      "loss": 0.4898,
+      "step": 1584
+    },
+    {
+      "epoch": 0.7136425033768573,
+      "grad_norm": 0.1284799426794052,
+      "learning_rate": 4.118553146816115e-05,
+      "loss": 0.5782,
+      "step": 1585
+    },
+    {
+      "epoch": 0.7140927510130571,
+      "grad_norm": 0.13369742035865784,
+      "learning_rate": 4.106580437116612e-05,
+      "loss": 0.5072,
+      "step": 1586
+    },
+    {
+      "epoch": 0.7145429986492571,
+      "grad_norm": 0.17579106986522675,
+      "learning_rate": 4.094620657034969e-05,
+      "loss": 0.8465,
+      "step": 1587
+    },
+    {
+      "epoch": 0.714993246285457,
+      "grad_norm": 0.22650310397148132,
+      "learning_rate": 4.082673832809838e-05,
+      "loss": 0.6008,
+      "step": 1588
+    },
+    {
+      "epoch": 0.715443493921657,
+      "grad_norm": 0.234883651137352,
+      "learning_rate": 4.070739990651449e-05,
+      "loss": 0.5887,
+      "step": 1589
+    },
+    {
+      "epoch": 0.7158937415578568,
+      "grad_norm": 0.13419248163700104,
+      "learning_rate": 4.058819156741546e-05,
+      "loss": 0.6063,
+      "step": 1590
+    },
+    {
+      "epoch": 0.7163439891940567,
+      "grad_norm": 0.09254826605319977,
+      "learning_rate": 4.046911357233343e-05,
+      "loss": 0.3823,
+      "step": 1591
+    },
+    {
+      "epoch": 0.7167942368302567,
+      "grad_norm": 0.1679757833480835,
+      "learning_rate": 4.03501661825144e-05,
+      "loss": 0.5351,
+      "step": 1592
+    },
+    {
+      "epoch": 0.7172444844664565,
+      "grad_norm": 0.1017613559961319,
+      "learning_rate": 4.023134965891809e-05,
+      "loss": 0.5625,
+      "step": 1593
+    },
+    {
+      "epoch": 0.7176947321026564,
+      "grad_norm": 0.19212448596954346,
+      "learning_rate": 4.011266426221687e-05,
+      "loss": 0.783,
+      "step": 1594
+    },
+    {
+      "epoch": 0.7181449797388564,
+      "grad_norm": 0.1037936583161354,
+      "learning_rate": 3.999411025279557e-05,
+      "loss": 0.5249,
+      "step": 1595
+    },
+    {
+      "epoch": 0.7185952273750563,
+      "grad_norm": 0.21334043145179749,
+      "learning_rate": 3.987568789075072e-05,
+      "loss": 0.7089,
+      "step": 1596
+    },
+    {
+      "epoch": 0.7190454750112562,
+      "grad_norm": 0.09836383163928986,
+      "learning_rate": 3.9757397435890044e-05,
+      "loss": 0.457,
+      "step": 1597
+    },
+    {
+      "epoch": 0.7194957226474561,
+      "grad_norm": 0.10441449284553528,
+      "learning_rate": 3.963923914773187e-05,
+      "loss": 0.4625,
+      "step": 1598
+    },
+    {
+      "epoch": 0.719945970283656,
+      "grad_norm": 0.1090429425239563,
+      "learning_rate": 3.9521213285504556e-05,
+      "loss": 0.4876,
+      "step": 1599
+    },
+    {
+      "epoch": 0.720396217919856,
+      "grad_norm": 0.11702052503824234,
+      "learning_rate": 3.9403320108145945e-05,
+      "loss": 0.5034,
+      "step": 1600
+    },
+    {
+      "epoch": 0.7208464655560558,
+      "grad_norm": 0.12491095811128616,
+      "learning_rate": 3.9285559874302756e-05,
+      "loss": 0.6502,
+      "step": 1601
+    },
+    {
+      "epoch": 0.7212967131922557,
+      "grad_norm": 0.15113623440265656,
+      "learning_rate": 3.9167932842330114e-05,
+      "loss": 0.4628,
+      "step": 1602
+    },
+    {
+      "epoch": 0.7217469608284557,
+      "grad_norm": 0.1599442958831787,
+      "learning_rate": 3.905043927029074e-05,
+      "loss": 0.579,
+      "step": 1603
+    },
+    {
+      "epoch": 0.7221972084646555,
+      "grad_norm": 0.10640469938516617,
+      "learning_rate": 3.893307941595481e-05,
+      "loss": 0.4568,
+      "step": 1604
+    },
+    {
+      "epoch": 0.7226474561008555,
+      "grad_norm": 0.14028172194957733,
+      "learning_rate": 3.8815853536798904e-05,
+      "loss": 0.6441,
+      "step": 1605
+    },
+    {
+      "epoch": 0.7230977037370554,
+      "grad_norm": 0.11768586188554764,
+      "learning_rate": 3.86987618900058e-05,
+      "loss": 0.5078,
+      "step": 1606
+    },
+    {
+      "epoch": 0.7235479513732553,
+      "grad_norm": 0.17457154393196106,
+      "learning_rate": 3.858180473246373e-05,
+      "loss": 0.8523,
+      "step": 1607
+    },
+    {
+      "epoch": 0.7239981990094552,
+      "grad_norm": 0.1039859876036644,
+      "learning_rate": 3.8464982320765905e-05,
+      "loss": 0.3675,
+      "step": 1608
+    },
+    {
+      "epoch": 0.7244484466456551,
+      "grad_norm": 0.13688620924949646,
+      "learning_rate": 3.834829491120991e-05,
+      "loss": 0.6705,
+      "step": 1609
+    },
+    {
+      "epoch": 0.724898694281855,
+      "grad_norm": 0.1294519603252411,
+      "learning_rate": 3.823174275979716e-05,
+      "loss": 0.4342,
+      "step": 1610
+    },
+    {
+      "epoch": 0.7253489419180549,
+      "grad_norm": 0.14002221822738647,
+      "learning_rate": 3.8115326122232196e-05,
+      "loss": 0.782,
+      "step": 1611
+    },
+    {
+      "epoch": 0.7257991895542548,
+      "grad_norm": 0.14345231652259827,
+      "learning_rate": 3.79990452539225e-05,
+      "loss": 0.8031,
+      "step": 1612
+    },
+    {
+      "epoch": 0.7262494371904548,
+      "grad_norm": 0.12528835237026215,
+      "learning_rate": 3.788290040997746e-05,
+      "loss": 0.5075,
+      "step": 1613
+    },
+    {
+      "epoch": 0.7266996848266547,
+      "grad_norm": 0.183663472533226,
+      "learning_rate": 3.7766891845208155e-05,
+      "loss": 0.8724,
+      "step": 1614
+    },
+    {
+      "epoch": 0.7271499324628545,
+      "grad_norm": 0.09628823399543762,
+      "learning_rate": 3.7651019814126654e-05,
+      "loss": 0.3916,
+      "step": 1615
+    },
+    {
+      "epoch": 0.7276001800990545,
+      "grad_norm": 0.15968765318393707,
+      "learning_rate": 3.753528457094548e-05,
+      "loss": 0.5802,
+      "step": 1616
+    },
+    {
+      "epoch": 0.7280504277352544,
+      "grad_norm": 0.10458499938249588,
+      "learning_rate": 3.7419686369577075e-05,
+      "loss": 0.585,
+      "step": 1617
+    },
+    {
+      "epoch": 0.7285006753714544,
+      "grad_norm": 0.1110527366399765,
+      "learning_rate": 3.730422546363324e-05,
+      "loss": 0.6312,
+      "step": 1618
+    },
+    {
+      "epoch": 0.7289509230076542,
+      "grad_norm": 0.10445191711187363,
+      "learning_rate": 3.7188902106424416e-05,
+      "loss": 0.3715,
+      "step": 1619
+    },
+    {
+      "epoch": 0.7294011706438541,
+      "grad_norm": 0.19541087746620178,
+      "learning_rate": 3.7073716550959534e-05,
+      "loss": 0.5043,
+      "step": 1620
+    },
+    {
+      "epoch": 0.7298514182800541,
+      "grad_norm": 0.15454646944999695,
+      "learning_rate": 3.695866904994496e-05,
+      "loss": 0.4605,
+      "step": 1621
+    },
+    {
+      "epoch": 0.7303016659162539,
+      "grad_norm": 0.12476633489131927,
+      "learning_rate": 3.684375985578431e-05,
+      "loss": 0.7631,
+      "step": 1622
+    },
+    {
+      "epoch": 0.7307519135524538,
+      "grad_norm": 0.1335902065038681,
+      "learning_rate": 3.672898922057774e-05,
+      "loss": 0.6115,
+      "step": 1623
+    },
+    {
+      "epoch": 0.7312021611886538,
+      "grad_norm": 0.21652035415172577,
+      "learning_rate": 3.6614357396121404e-05,
+      "loss": 0.6122,
+      "step": 1624
+    },
+    {
+      "epoch": 0.7316524088248537,
+      "grad_norm": 0.22802138328552246,
+      "learning_rate": 3.649986463390694e-05,
+      "loss": 0.4561,
+      "step": 1625
+    },
+    {
+      "epoch": 0.7321026564610535,
+      "grad_norm": 0.12539541721343994,
+      "learning_rate": 3.638551118512089e-05,
+      "loss": 0.3666,
+      "step": 1626
+    },
+    {
+      "epoch": 0.7325529040972535,
+      "grad_norm": 0.14440281689167023,
+      "learning_rate": 3.627129730064416e-05,
+      "loss": 0.7506,
+      "step": 1627
+    },
+    {
+      "epoch": 0.7330031517334534,
+      "grad_norm": 0.15482786297798157,
+      "learning_rate": 3.615722323105143e-05,
+      "loss": 0.6112,
+      "step": 1628
+    },
+    {
+      "epoch": 0.7334533993696533,
+      "grad_norm": 0.09844894707202911,
+      "learning_rate": 3.604328922661072e-05,
+      "loss": 0.4506,
+      "step": 1629
+    },
+    {
+      "epoch": 0.7339036470058532,
+      "grad_norm": 0.09876194596290588,
+      "learning_rate": 3.5929495537282595e-05,
+      "loss": 0.3801,
+      "step": 1630
+    },
+    {
+      "epoch": 0.7343538946420531,
+      "grad_norm": 0.11300911009311676,
+      "learning_rate": 3.581584241272005e-05,
+      "loss": 0.2742,
+      "step": 1631
+    },
+    {
+      "epoch": 0.7348041422782531,
+      "grad_norm": 0.12158750742673874,
+      "learning_rate": 3.57023301022674e-05,
+      "loss": 0.485,
+      "step": 1632
+    },
+    {
+      "epoch": 0.7352543899144529,
+      "grad_norm": 0.10776393860578537,
+      "learning_rate": 3.558895885496023e-05,
+      "loss": 0.6717,
+      "step": 1633
+    },
+    {
+      "epoch": 0.7357046375506529,
+      "grad_norm": 0.14828453958034515,
+      "learning_rate": 3.547572891952457e-05,
+      "loss": 0.7198,
+      "step": 1634
+    },
+    {
+      "epoch": 0.7361548851868528,
+      "grad_norm": 0.11934065073728561,
+      "learning_rate": 3.5362640544376415e-05,
+      "loss": 0.5431,
+      "step": 1635
+    },
+    {
+      "epoch": 0.7366051328230527,
+      "grad_norm": 0.12670141458511353,
+      "learning_rate": 3.5249693977621226e-05,
+      "loss": 0.7049,
+      "step": 1636
+    },
+    {
+      "epoch": 0.7370553804592526,
+      "grad_norm": 0.13355158269405365,
+      "learning_rate": 3.513688946705336e-05,
+      "loss": 0.7022,
+      "step": 1637
+    },
+    {
+      "epoch": 0.7375056280954525,
+      "grad_norm": 0.14985230565071106,
+      "learning_rate": 3.5024227260155385e-05,
+      "loss": 0.8045,
+      "step": 1638
+    },
+    {
+      "epoch": 0.7379558757316524,
+      "grad_norm": 0.12919121980667114,
+      "learning_rate": 3.491170760409792e-05,
+      "loss": 0.5669,
+      "step": 1639
+    },
+    {
+      "epoch": 0.7384061233678523,
+      "grad_norm": 0.17263253033161163,
+      "learning_rate": 3.479933074573858e-05,
+      "loss": 0.6408,
+      "step": 1640
+    },
+    {
+      "epoch": 0.7388563710040522,
+      "grad_norm": 0.09497567266225815,
+      "learning_rate": 3.468709693162183e-05,
+      "loss": 0.3811,
+      "step": 1641
+    },
+    {
+      "epoch": 0.7393066186402522,
+      "grad_norm": 0.08906082063913345,
+      "learning_rate": 3.457500640797831e-05,
+      "loss": 0.3825,
+      "step": 1642
+    },
+    {
+      "epoch": 0.7397568662764521,
+      "grad_norm": 0.16931599378585815,
+      "learning_rate": 3.446305942072425e-05,
+      "loss": 1.0083,
+      "step": 1643
+    },
+    {
+      "epoch": 0.7402071139126519,
+      "grad_norm": 0.17544637620449066,
+      "learning_rate": 3.4351256215461e-05,
+      "loss": 0.5478,
+      "step": 1644
+    },
+    {
+      "epoch": 0.7406573615488519,
+      "grad_norm": 0.20658282935619354,
+      "learning_rate": 3.4239597037474494e-05,
+      "loss": 0.4695,
+      "step": 1645
+    },
+    {
+      "epoch": 0.7411076091850518,
+      "grad_norm": 0.14529536664485931,
+      "learning_rate": 3.412808213173454e-05,
+      "loss": 0.8659,
+      "step": 1646
+    },
+    {
+      "epoch": 0.7415578568212516,
+      "grad_norm": 0.12230098992586136,
+      "learning_rate": 3.401671174289469e-05,
+      "loss": 0.5459,
+      "step": 1647
+    },
+    {
+      "epoch": 0.7420081044574516,
+      "grad_norm": 0.16703882813453674,
+      "learning_rate": 3.390548611529116e-05,
+      "loss": 0.6082,
+      "step": 1648
+    },
+    {
+      "epoch": 0.7424583520936515,
+      "grad_norm": 0.10900110006332397,
+      "learning_rate": 3.379440549294271e-05,
+      "loss": 0.6259,
+      "step": 1649
+    },
+    {
+      "epoch": 0.7429085997298515,
+      "grad_norm": 0.1300138235092163,
+      "learning_rate": 3.368347011955006e-05,
+      "loss": 0.4008,
+      "step": 1650
+    },
+    {
+      "epoch": 0.7433588473660513,
+      "grad_norm": 0.22459514439105988,
+      "learning_rate": 3.357268023849507e-05,
+      "loss": 0.7222,
+      "step": 1651
+    },
+    {
+      "epoch": 0.7438090950022512,
+      "grad_norm": 0.12489008903503418,
+      "learning_rate": 3.3462036092840535e-05,
+      "loss": 0.699,
+      "step": 1652
+    },
+    {
+      "epoch": 0.7442593426384512,
+      "grad_norm": 0.12445378303527832,
+      "learning_rate": 3.3351537925329456e-05,
+      "loss": 0.5377,
+      "step": 1653
+    },
+    {
+      "epoch": 0.7447095902746511,
+      "grad_norm": 0.10186782479286194,
+      "learning_rate": 3.324118597838464e-05,
+      "loss": 0.3425,
+      "step": 1654
+    },
+    {
+      "epoch": 0.7451598379108509,
+      "grad_norm": 0.11316341161727905,
+      "learning_rate": 3.3130980494108034e-05,
+      "loss": 0.4823,
+      "step": 1655
+    },
+    {
+      "epoch": 0.7456100855470509,
+      "grad_norm": 0.10849397629499435,
+      "learning_rate": 3.302092171428033e-05,
+      "loss": 0.4251,
+      "step": 1656
+    },
+    {
+      "epoch": 0.7460603331832508,
+      "grad_norm": 0.13365057110786438,
+      "learning_rate": 3.291100988036022e-05,
+      "loss": 0.6921,
+      "step": 1657
+    },
+    {
+      "epoch": 0.7465105808194507,
+      "grad_norm": 0.14498238265514374,
+      "learning_rate": 3.2801245233484236e-05,
+      "loss": 0.5422,
+      "step": 1658
+    },
+    {
+      "epoch": 0.7469608284556506,
+      "grad_norm": 0.123689666390419,
+      "learning_rate": 3.269162801446578e-05,
+      "loss": 0.759,
+      "step": 1659
+    },
+    {
+      "epoch": 0.7474110760918505,
+      "grad_norm": 0.16270355880260468,
+      "learning_rate": 3.258215846379492e-05,
+      "loss": 0.5498,
+      "step": 1660
+    },
+    {
+      "epoch": 0.7478613237280505,
+      "grad_norm": 0.1305166780948639,
+      "learning_rate": 3.2472836821637744e-05,
+      "loss": 0.5375,
+      "step": 1661
+    },
+    {
+      "epoch": 0.7483115713642503,
+      "grad_norm": 0.2051253616809845,
+      "learning_rate": 3.236366332783586e-05,
+      "loss": 0.6834,
+      "step": 1662
+    },
+    {
+      "epoch": 0.7487618190004502,
+      "grad_norm": 0.13869157433509827,
+      "learning_rate": 3.225463822190572e-05,
+      "loss": 0.5859,
+      "step": 1663
+    },
+    {
+      "epoch": 0.7492120666366502,
+      "grad_norm": 0.10008493810892105,
+      "learning_rate": 3.214576174303846e-05,
+      "loss": 0.4643,
+      "step": 1664
+    },
+    {
+      "epoch": 0.74966231427285,
+      "grad_norm": 0.13287660479545593,
+      "learning_rate": 3.2037034130098905e-05,
+      "loss": 0.6476,
+      "step": 1665
+    },
+    {
+      "epoch": 0.75011256190905,
+      "grad_norm": 0.1328587532043457,
+      "learning_rate": 3.1928455621625494e-05,
+      "loss": 0.6488,
+      "step": 1666
+    },
+    {
+      "epoch": 0.7505628095452499,
+      "grad_norm": 0.11050630360841751,
+      "learning_rate": 3.182002645582935e-05,
+      "loss": 0.6299,
+      "step": 1667
+    },
+    {
+      "epoch": 0.7510130571814498,
+      "grad_norm": 0.1361268013715744,
+      "learning_rate": 3.1711746870594086e-05,
+      "loss": 0.5877,
+      "step": 1668
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2221,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 556,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 625913609846784.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}